]> git.ipfire.org Git - thirdparty/gcc.git/blame - gcc/ChangeLog
gimple-pretty-print.c (dump_ssaname_info): Always print "# " before the info, not...
[thirdparty/gcc.git] / gcc / ChangeLog
CommitLineData
d09c0e9b
JJ
12013-10-23 Jakub Jelinek <jakub@redhat.com>
2
3 * gimple-pretty-print.c (dump_ssaname_info): Always print "# " before
4 the info, not after it.
5 (gump_gimple_phi): Add COMMENT argument, if true, print "# " after
6 dump_ssaname_info call.
7 (pp_gimple_stmt_1): Adjust caller.
8 (dump_phi_nodes): Likewise. Don't print "# " here.
9
561400f0
JH
102013-10-22 Jan Hubicka <jh@suse.cz>
11
12 * i386.h (TARGET_MISALIGNED_MOVE_STRING_PROLOGUES_EPILOGUES): New tuning flag.
13 * x86-tune.def (TARGET_MISALIGNED_MOVE_STRING_PROLOGUES): Define it.
14 * i386.c (expand_small_movmem_or_setmem): New function.
15 (expand_set_or_movmem_prologue_epilogue_by_misaligned_moves): New function
16 (alg_usable_p): Add support for value ranges; cleanup.
17 (ix86_expand_set_or_movmem): Add support for misaligned moves.
18
7a1dd0fa
SA
192013-10-22 Sterling Augustine <saugustine@google.com>
20
21 * doc/invoke.texi: Document -ggnu-pubnames.
22 * common.opt: Add new option -ggnu-pubnames and modify -gpubnames
23 logic.
24 * dwarf2out.c: Include gdb/gdb-index.h.
25 (DEBUG_PUBNAMES_SECTION, DEBUG_PUBTYPES_SECTION): Handle
26 debug_generate_pub_sections.
27 (is_java, output_pubtables, output_pubname): New functions.
28 (include_pubname_in_output): Handle debug_generate_pub_sections at
29 level 2.
30 (size_of_pubnames): Use new local space_for_flags based on
31 debug_generate_pub_sections.
32 (output_pubnames): Unify pubnames and pubtypes output logic.
33 Genericize comments. Call output_pubname.
34 (dwarf2out_finish): Move logic to output_pubnames and call it.
35
9ae93484
UB
362013-10-22 Uros Bizjak <ubizjak@gmail.com>
37
38 PR target/58779
39 * config/i386/i386.c (put_condition_code) <case GTU, case LEU>:
40 Remove CCCmode handling.
41 <case LTU>: Return 'c' suffix for CCCmode.
42 <case GEU>: Return 'nc' suffix for CCCmode.
43 (ix86_cc_mode) <case GTU, case LEU>: Do not generate overflow checks.
44 * config/i386/i386.md (*sub<mode>3_cconly_overflow): Remove.
45 (*sub<mode>3_cc_overflow): Ditto.
46 (*subsi3_zext_cc_overflow): Ditto.
47
ade97eb1
SE
482013-10-22 Steve Ellcey <sellcey@mips.com>
49
50 * config/mips/mips.c (mips_rtx_costs): Fix cost estimate for nor
51 (AND (NOT OP1) (NOT OP2)).
52
843d68c6
BS
532013-10-22 Bill Schmidt <wschmidt@vnet.ibm.com>
54
55 * config/rs6000/rs6000.c (altivec_expand_vec_perm_const): Reverse
56 meaning of merge-high and merge-low masks for little endian; avoid
57 use of vector-pack masks for little endian for mismatched modes.
58
574733d3
JBG
592013-10-22 Jan-Benedict Glaw <jbglaw@lug-owl.de>
60
61 * config/tilepro/tilepro.c: Include "tree.h".
62
1adb2fb9
AS
632013-10-22 Andreas Schwab <schwab@suse.de>
64
65 * config/m68k/m68k.c (notice_update_cc): Handle register conflict
66 with PRE_DEC.
67
f2f39141
PC
682013-10-22 Paolo Carlini <paolo.carlini@oracle.com>
69
70 * doc/contrib.texi ([Fran@,{c}ois Dumont], [Tim Shen],
71 [Ed Smith-Rowland]): New entries.
72 ([Stephen M. Webb]): Update.
73
8223bb78
AM
742013-10-22 Andrew MacLeod <amacleod@redhat.com>
75
76 * tree-ssa-ter.h: Remove duplicate copy of file contents.
77
1e08df0e
MP
782013-10-21 Marek Polacek <polacek@redhat.com>
79
80 PR middle-end/58809
81 * fold-const.c (fold_range_test): Return 0 if the type is not
82 an integral type.
83
67b5215c
RS
842013-10-21 Richard Sandiford <rdsandiford@googlemail.com>
85
86 * system.h: Move hwint.h include further down.
87 * hwint.h (sext_hwi, zext_hwi): Define unconditionally. Add
88 gcc_checking_asserts.
89 * hwint.c (sext_hwi, zext_hwi): Delete ENABLE_CHECKING versions.
90
0cad6830
BE
912013-10-21 Bernd Edlinger <bernd.edlinger@hotmail.de>
92
93 Fix volatile issues in optimize_bit_field_compare.
94 * fold-const.c (optimize_bit_field_compare): Bail out if
95 lvolatilep or rvolatilep.
96
5dfd32cd
BE
972013-10-21 Bernd Edlinger <bernd.edlinger@hotmail.de>
98
99 Fix DECL_BIT_FIELD depencency on flag_strict_volatile_bitfields
100 and get_inner_reference returning different pmode for non-volatile
101 bit-field members dependent on flag_strict_volatile_bitfields.
102 * stor-layout.c (layout_decl): Remove special handling of
103 flag_strict_volatile_bitfields.
104 * expr.c (get_inner_reference): Don't use DECL_BIT_FIELD
105 if flag_strict_volatile_bitfields > 0 and TREE_THIS_VOLATILE.
106
0bff3610
PM
1072013-10-21 Paulo Matos <pmatos@broadcom.com>
108
109 * ipa-inline.c (edge_badness): Cap edge->count at max_count for badness
110 calculations.
111
b5c4ff78
JL
1122013-10-21 Jeff Law <law@redhat.com>
113
f2f39141
PC
114 * tree-ssa-threadedge.c (thread_through_normal_block): New
115 argument VISITED. Remove VISISTED as a local variable. When we
116 have a threadable jump, verify the destination of the jump has not
117 been visised.
118 (thread_across_edge): Allocate VISITED bitmap once at function
119 scope and use it throughout. Make sure to set appropriate bits in
120 VISITED for E (start of jump thread path).
121 * tree-ssa-threadupdate.c (mark_threaded_blocks): Reject threading
122 through a joiner if any edge on the path has a recorded jump thread.
b5c4ff78 123
ce0770fc
ILT
1242013-10-21 Ian Lance Taylor <iant@google.com>
125
126 * doc/invoke.texi (Optimize Options): For -fno-toplevel-reorder,
127 don't imply that attributes can solve all problems.
128 (Directory Options): Fix typo.
129
0e7504c8
KT
1302013-10-21 Kyrylo Tkachov <kyrylo.tkachov@arm.com>
131
132 * config/arm/arm.c (cortexa9_extra_costs): Update mult costs for
133 extend and extend_add.
134
2445aeac
RB
1352013-10-21 Richard Biener <rguenther@suse.de>
136
137 PR tree-optimization/58794
138 * fold-const.c (operand_equal_p): Compare FIELD_DECL operand
139 of COMPONENT_REFs with OEP_CONSTANT_ADDRESS_OF left in place.
140
d480e6a5
RB
1412013-10-21 Richard Biener <rguenther@suse.de>
142
143 PR middle-end/58742
144 * fold-const.c (fold_binary_loc): Fold ((T) (X /[ex] C)) * C
145 to (T) X for sign-changing conversions (or no conversion).
146
54f9aba0
UB
1472013-10-20 Uros Bizjak <ubizjak@gmail.com>
148
149 * config/i386/i386.md (kxnor<mode>): Add FLAGS_REG clobber.
150
d6c6ba3c
JH
1512013-10-20 Jan Hubicka <jh@suse.cz>
152
153 * config/i386/i386-tune.def: Add comment; organize into categories
154
d606b917
MZ
1552013-10-21 Michael Zolotukhin <michael.v.zolotukhin@gmail.com>
156
157 * config/i386/i386.c (expand_set_or_movmem_via_loop): Add issetmem
158 argument. Update function comment.
159 (expand_set_or_movmem_via_rep): New function combining
160 expand_movmem_via_rep_mov and expand_setmem_via_rep_stos.
161 (expand_movmem_via_rep_mov): Remove.
162 expand_setmem_via_rep_stos): Remove.
163 (expand_movmem_epilogue): Update calls correspondingly.
164 (expand_setmem_epilogue_via_loop): Likewise.
165 (emit_memset): New.
166 (expand_setmem_epilogue): Add VEC_VALUE argument, refactor.
167 (expand_set_or_movmem_prologue): New function combining
168 expand_movmem_prologue and expand_setmem_prologue.
169 (expand_movmem_prologue): Remove.
170 (expand_setmem_prologue): Remove.
171 (expand_set_or_movmem_constant_prologue): New function combining
172 expand_constant_movmem_prologue and expand_constant_setmem_prologue.
173 (expand_constant_movmem_prologue): Remove.
174 (expand_constant_setmem_prologue): Remove.
175 (promote_duplicated_reg): Allow vector-const0 value.
176 (ix86_expand_set_or_movmem): New function combining ix86_expand_movmem
177 and ix86_expand_setmem.
178 (ix86_expand_movmem): Call ix86_expand_set_or_movmem.
179 (ix86_expand_setmem): Call ix86_expand_set_or_movmem.
180
4d648807
DN
1812013-10-21 Diego Novillo <dnovillo@google.com>
182
183 * asan.c: Include tree.h
184 * bb-reorder.c: Likewise.
185 * cfgcleanup.c: Likewise.
186 * cfgloopmanip.c: Likewise.
187 * data-streamer-in.c: Likewise.
188 * data-streamer-out.c: Likewise.
189 * data-streamer.c: Likewise.
190 * dwarf2cfi.c: Likewise.
191 * graphite-blocking.c: Likewise.
192 * graphite-clast-to-gimple.c: Likewise.
193 * graphite-dependences.c: Likewise.
194 * graphite-interchange.c: Likewise.
195 * graphite-optimize-isl.c: Likewise.
196 * graphite-poly.c: Likewise.
197 * graphite-scop-detection.c: Likewise.
198 * graphite-sese-to-poly.c: Likewise.
199 * graphite.c: Likewise.
200 * ipa-devirt.c: Likewise.
201 * ipa-profile.c: Likewise.
202 * ipa.c: Likewise.
203 * ira.c: Likewise.
204 * loop-init.c: Likewise.
205 * loop-unroll.c: Likewise.
206 * lower-subreg.c: Likewise.
207 * lto/lto-object.c: Likewise.
208 * recog.c: Likewise.
209 * reginfo.c: Likewise.
210 * tree-loop-distribution.c: Likewise.
211 * tree-parloops.c: Likewise.
212 * tree-ssa-strlen.c: Likewise.
213 * tree-streamer.c: Likewise.
214 * value-prof.c: Likewise.
215 * target-globals.c: Likewise.
216 * expr.h: Include tree-core.h instead of tree.h.
217 * gimple.h: Likewise.
218 * ipa-prop.h: Likewise.
219 * ipa-utils.h: Likewise.
220 * lto-streamer.h: Likewise.
221 * streamer-hooks.h: Likewise.
222 * ipa-reference.h: Include cgraph.h instead of tree.h.
223 * cgraph.h: Include basic-block.h instead of tree.h.
224 * tree-streamer.h: Do not include tree.h.
225 * genattrtab.c (write_header): Generate inclusion of
226 tree.h.
227 * genautomata.c (main): Likewise.
228 * genemit.c: Likewise.
229 * genopinit.c: Likewise.
230 * genoutput.c (output_prologue): Likewise.
231 * genpeep.c: Likewise.
232
dd346f1f
BS
2332013-10-20 Bill Schmidt <wschmidt@linux.vnet.ibm.com>
234
235 * config/rs6000/altivec.md (vec_unpacku_hi_v16qi): Adjust for
236 little endian.
237 (vec_unpacku_hi_v8hi): Likewise.
238 (vec_unpacku_lo_v16qi): Likewise.
239 (vec_unpacku_lo_v8hi): Likewise.
240
915e8e6e
JH
2412013-10-20 Jan Hubicka <jh@suse.cz>
242
243 * config/i386/x86-tune.def (X86_TUNE_SLOW_IMUL_IMM32_MEM,
244 X86_TUNE_SLOW_IMUL_IMM8): Keep enabled only for K8 and AMDFAM10.
245 (X86_TUNE_USE_VECTOR_FP_CONVERTS): Disable for generic.
246
0f37323c
RS
2472013-10-20 Richard Sandiford <rdsandiford@googlemail.com>
248
249 * config/mips/mips.h (ISA_HAS_WSBH): Define.
250 * config/mips/mips.md (UNSPEC_WSBH, UNSPEC_DSBH, UNSPEC_DSHD): New
251 constants.
252 (bswaphi2, bswapsi2, bswapdi2, wsbh, dsbh, dshd): New patterns.
253
84935d25
JDA
2542013-10-19 John David Anglin <danglin@gcc.gnu.org>
255
256 PR target/58603
257 * system.h: Undef m_slot.
258
d7d1451a
BS
2592013-10-19 Bill Schmidt <wschmidt@linux.vnet.ibm.com>
260
261 * config/rs6000/rs6000.c (vspltis_constant): Make sure we check
262 all elements for both endian flavors.
263
6eb793dd
UB
2642013-10-19 Uros Bizjak <ubizjak@gmail.com>
265
266 PR target/58792
267 * config/i386/i386.c (ix86_function_value_regno): Add DX_REG,
268 ST1_REG and XMM1_REG for 32bit and 64bit targets. Also add DI_REG
269 and SI_REG for 64bit SYSV ABI targets.
270
c07757e5
UB
2712013-10-19 Uros Bizjak <ubizjak@gmail.com>
272
273 * mode-switching.c (create_pre_exit): Rename maybe_builtin_apply
274 to multi_reg_return. Clarify that we are skipping USEs of multiple
275 return registers. Use bool type where appropriate.
276
41ee845b
JH
2772013-10-18 Jan Hubicka <jh@suse.cz>
278
279 * config/i386/i386.h (ACCUMULATE_OUTGOING_ARGS): Disable accumulation
280 for cold functions.
281 * x86-tune.def (X86_TUNE_USE_LEAVE): Update comment.
282 (X86_TUNE_PUSH_MEMORY): Likewise.
283 (X86_TUNE_AVX256_UNALIGNED_LOAD_OPTIMAL,
284 X86_TUNE_AVX256_UNALIGNED_STORE_OPTIMAL): New.
c07757e5
UB
285 (X86_TUNE_ACCUMULATE_OUTGOING_ARGS, X86_TUNE_ALWAYS_FANCY_MATH_387):
286 New.
41ee845b
JH
287 * i386.c (x86_accumulate_outgoing_args, x86_arch_always_fancy_math_387,
288 x86_avx256_split_unaligned_load, x86_avx256_split_unaligned_store):
289 Remove.
290 (ix86_option_override_internal): Update to use tune features instead
291 of variables.
292
6f978a2a
CH
2932013-10-18 Cong Hou <congh@google.com>
294
295 PR tree-optimization/58508
296 * tree-vect-loop-manip.c (vect_loop_versioning): Hoist loop invariant
297 statement that contains data refs with zero-step.
298
4484a35a
AM
2992013-10-18 Andrew MacLeod <amacleod@redhat.com>
300
c07757e5
UB
301 * tree-ssa.h: Don't include gimple-low.h, tree-ssa-address.h,
302 sbitmap.h, tree-ssa-threadedge.h, tree-ssa-dom.h and tree-cfgcleanup.h.
4484a35a
AM
303 * gimple-low.c (gimple_check_call_arg,
304 gimple_check_call_matching_types): Move to cgraph.c.
305 * gimple-low.h: Remove prototype.
306 * cgraph.c: (gimple_check_call_arg, gimple_check_call_matching_types):
307 Relocate from gimple-low.c.
308 * cgraph.h: Add prototype. Don't include basic-block.h.
309 * gimplify.c: Add gimple-low to include list.
310 * omp-low.c: Add gimple-low and tree-cfgcleanup.h to include list.
311 * tree-eh.c: Add gimple-low to include list.
312 * tree-nested.c: Likewise.
313 * cfgexpand.c: Add tree-ssa-address.h to include list.
314 * expr.c: Likewise.
315 * gimple-fold.c: Likewise.
316 * gimple-ssa-strength-reduction.c: Likewise.
317 * trans-mem.c: Likewise.
318 * tree-mudflap.c: Likewise.
319 * tree-ssa-loop-ivopts.c: Likewise.
320 * tree-ssa-dom.c: Include tree-ssa-threadedge.h and tree-ssa-dom.h.
321 (degenerate_phi_result): Move to tree-phinodes.c.
322 * tree-ssa-loop-ch.c: Include tree-ssa-threadedge.h.
323 * tree-ssa-threadedge.c: Likewise.
324 * tree-vrp.c: Likewise.
325 * tree-phinodes.c (degenerate_phi_result): Relocate here.
326 * tree-ssa-dom.h (degenerate_phi_result): Remove Prototype.
327 * tree-phinodes.h (degenerate_phi_result): Add prototype.
328 * tree-ssa-copy.c: Include tree-ssa-dom.h.
329 * tree-ssa-forwprop.c: Likewise.
330 * tree-cfgcleanup.c (execute_cleanup_cfg_post_optimizing,
331 pass_data_cleanup_cfg_post_optimizing,
332 make_pass_cleanup_cfg_post_optimizing): Relocate from tree-optimize.c.
333 * tree-optimize.c: Delete File.
334 * graphite.c: Include tree-cfgcleanup.h.
335 * passes.c: Likewise.
336 * tree-cfg.c: Likewise.
337 * tree-profile.c: Likewise.
338 * tree-ssa-dse.c: Likewise.
339 * tree-ssa-loop-ivcanon.c: Likewise.
340 * tree-switch-conversion.c: Don't include tree-ssa-operands.h.
341 * tree-outof-ssa.c: Include sbitmap.h.
342 * tree-ssa-live.c: Likewise.
343 * tree-ssa-propagate.c: Likewise.
344 * tree-ssa-structalias.c: Likewise.
345 * tree-stdarg.c: Likewise.
346 * Makefile.in (OBJS): Delete tree-optimize.o.
347 * basic-block.h (gcov_type, gcov_type_unsigned): Move to coretypes.h.
348 * coretypes.h (gcov_type, gcov_type_unsigned): Relocate here.
349 * varasm.c: Include basic-block.h.
350 * cfgloop.h: Include function.h instead of basic-block.h
351 (bb_loop_depth): Move to cfgloop.c.
352 * cfgloop.c (bb_loop_depth): Relocate from cfgloop.h.
353
bfaa17c1
TJ
3542013-10-18 Teresa Johnson <tejohnson@google.com>
355
356 * predict.c (probably_never_executed): Compare frequency-based
357 count to number of training runs.
358 * params.def (UNLIKELY_BB_COUNT_FRACTION): New parameter.
359
f58883bd
KT
3602013-10-18 Kyrylo Tkachov <kyrylo.tkachov@arm.com>
361
362 * config/arm/arm.c (cortexa9_extra_costs): New table.
363 (arm_cortex_a9_tune): Use cortexa9_extra_costs.
364
01e127b1
JL
3652013-10-18 Jeff Law <law@redhat.com>
366
367 * tree-ssa-threadupdate.c: Do not include "tm.h" or "tm_p.h".
368
369 * tree-ssa-threadupdate.c: Include "dbgcnt.h".
c07757e5
UB
370 (register_jump_thread): Add "registered_jump_thread" debug
371 counter support.
01e127b1
JL
372 * dbgcnt.def (registered_jump_thread): New debug counter.
373
f41b8848
AM
3742013-10-18 Andrew MacLeod <amacleod@redhat.com>
375
376 * config/rs6000/rs6000.c: Include cgraph.h.
377
adc7a812
TJ
3782013-10-18 Teresa Johnson <tejohnson@google.com>
379
380 * tree-ssa-tail-merge.c (replace_block_by): Update edge
381 weights during merging.
382
377d9792
AM
3832013-10-18 Andrew MacLeod <amacleod@redhat.com>
384
385 * tree-cfg.h: Rename from tree-flow.h. Remove #includes.
386 * tree-ssa.h: Relocate required #includes from tree-cfg.h.
387 * tree-ssa-operands.h: Remove prototype.
388 * tree-ssa-operands.c (virtual_operand_p): Move to gimple.c.
389 * gimple.c (virtual_operand_p): Relocate from gimple.c.
390 * gimple.h: Add prototype.
391 * gimple-ssa.h: Include tree-ssa-operands.h.
392 * tree-dump.c: Add tree-cfg.h to include list.
393 * tree-ssa-alias.c: Add ipa-reference.h to include list.
394 * config/alpha/alpha.c: Include gimple-ssa.h instead of tree-flow.h.
395 * config/i386/i386.c: Don't include tree-flow.h.
396 * config/rs6000/rs6000.c: Likewise.
397
8ad8afaf
JBG
3982013-10-18 Jan-Benedict Glaw <jbglaw@lug-owl.de>
399
400 * config/frv/frv.c (frv_init_cumulative_args): Fix wrong cast.
401
c0e081a9
RB
4022013-10-18 Richard Biener <rguenther@suse.de>
403
404 * stor-layout.c (layout_type): Do not change TYPE_PRECISION
405 or TYPE_UNSIGNED of integral types.
406 (set_min_and_max_values_for_integral_type): Leave TYPE_MIN/MAX_VALUE
407 NULL_TREE for zero-precision integral types.
408
054fceb5
JG
4092013-10-18 James Greenhalgh <james.greenhalgh@arm.com>
410
411 * config/aarch64/arm_neon.h
412 (vcvt<ds>_n_<fsu><32,64>_<fsu><32,64>): Correct argument types.
413
70e8b89b
ST
4142013-10-17 Sriraman Tallam <tmsriram@google.com>
415
416 * opth-gen.awk: Define target_flags_explicit.
417
3170766c 4182013-10-17 Michael Meissner <meissner@linux.vnet.ibm.com>
5845f602
MM
419
420 * config/rs6000/rs6000.c (enum rs6000_reload_reg_type): Add new
421 fields to the reg_addr array that describes the valid addressing
422 mode for any register, general purpose registers, floating point
423 registers, and Altivec registers.
424 (FIRST_RELOAD_REG_CLASS): Likewise.
425 (LAST_RELOAD_REG_CLASS): Likewise.
426 (struct reload_reg_map_type): Likewise.
427 (reload_reg_map_type): Likewise.
428 (RELOAD_REG_VALID): Likewise.
429 (RELOAD_REG_MULTIPLE): Likewise.
430 (RELOAD_REG_INDEXED): Likewise.
431 (RELOAD_REG_OFFSET): Likewise.
432 (RELOAD_REG_PRE_INCDEC): Likewise.
433 (RELOAD_REG_PRE_MODIFY): Likewise.
434 (reg_addr): Likewise.
435 (mode_supports_pre_incdec_p): New helper functions to say whether
436 a given mode supports PRE_INC, PRE_DEC, and PRE_MODIFY.
437 (mode_supports_pre_modify_p): Likewise.
438 (rs6000_debug_vector_unit): Rearrange the -mdebug=reg output to
439 print the valid address mode bits for each mode.
440 (rs6000_debug_print_mode): Likewise.
441 (rs6000_debug_reg_global): Likewise.
442 (rs6000_setup_reg_addr_masks): New function to set up the address
443 mask bits for each type.
444 (rs6000_init_hard_regno_mode_ok): Use memset to clear arrays.
445 Call rs6000_setup_reg_addr_masks to set up the address mask bits.
446 (rs6000_legitimate_address_p): Use mode_supports_pre_incdec_p and
447 mode_supports_pre_modify_p to determine if PRE_INC, PRE_DEC, and
448 PRE_MODIFY are supported.
3170766c
MM
449 (rs6000_output_move_128bit): Change to use {src,dest}_vmx_p for altivec
450 registers, instead of {src,dest}_av_p.
5845f602
MM
451 (rs6000_print_options_internal): Tweak the debug output slightly.
452
82e33890
UB
4532013-10-17 Uros Bizjak <ubizjak@gmail.com>
454
455 * config/i386/sse.md (*vec_widen_smult_even_v8si): Remove
456 isa attribute.
457
0645c1a2
AM
4582013-10-17 Andrew MacLeod <amacleod@redhat.com>
459
460 * tree-flow.h (struct omp_region): Move to omp-low.c.
461 Remove omp_ prototypes and variables.
462 * gimple.h (omp_reduction_init): Move prototype to omp-low.h.
463 (copy_var_decl): Relocate prototype from tree-flow.h.
464 * gimple.c (copy_var_decl): Relocate from omp-low.c.
465 * tree.h: Move prototype to omp-low.h.
466 * omp-low.h: New File. Relocate prototypes here.
467 * omp-low.c (struct omp_region): Make local here.
468 (root_omp_region): Make static.
469 (copy_var_decl) Move to gimple.c.
470 (new_omp_region): Make static.
471 (make_gimple_omp_edges): New. Refactored from tree-cfg.c make_edges.
472 * tree-cfg.c: Include omp-low.h.
473 (make_edges): Factor out OMP specific bits to make_gimple_omp_edges.
474 * gimplify.c: Include omp-low.h.
475 * tree-parloops.c: Likewise.
476
6f1abb55
UB
4772013-10-17 Uros Bizjak <ubizjak@gmail.com>
478
479 * config/i386/i386.c (ix86_fixup_binary_operands): When both source
480 operands are in memory, prefer to force non-matched operand 1 to
481 the register.
482
1ddbbbc0
MM
4832013-10-17 Michael Meissner <meissner@linux.vnet.ibm.com>
484
485 PR target/58673
486 * config/rs6000/rs6000.c (rs6000_legitimate_address_p): Only
487 restrict TImode addresses to single indirect registers if both
488 -mquad-memory and -mvsx-timode are used.
489 (rs6000_output_move_128bit): Use quad_load_store_p to determine if
490 we should emit load/store quad. Remove using %y for quad memory
491 addresses.
492
493 * config/rs6000/rs6000.md (mov<mode>_ppc64, TI/PTImode): Add
494 constraints to allow load/store quad on machines where TImode is
495 not allowed in VSX registers. Use 'n' instead of 'F' constraint
496 for TImode to load integer constants.
497
f541a481
KT
4982013-10-17 Kyrylo Tkachov <kyrylo.tkachov@arm.com>
499
500 * config/aarch64/aarch64.c (aarch64_print_operand): Handle 'c'.
501
51bb310d
MS
5022013-10-17 Marcus Shawcroft <marcus.shawcroft@arm.com>
503
504 * config/aarch64/aarch64.c (aarch64_preferred_reload_class): Adjust
505 handling of STACK_REG.
506
f4bed77b
RB
5072013-10-17 Richard Biener <rguenther@suse.de>
508
509 PR tree-optimization/58143
510 * tree-ssa-loop-im.c (arith_code_with_undefined_signed_overflow):
511 New function.
512 (rewrite_to_defined_overflow): Likewise.
513 (move_computations_dom_walker::before_dom): Rewrite stmts
514 with undefined signed overflow that are not always executed
515 into unsigned arithmetic.
516
d358fb9d
MM
5172013-10-16 Michael Meissner <meissner@linux.vnet.ibm.com>
518
519 PR target/57756
520 * config/rs6000/rs6000.opt (rs6000_isa_flags_explicit): Move the
521 explicit isa flag to be an options variable, instead of using
522 global_options_set. Remove define from rs6000.h.
523 * config/rs6000/rs6000.h (rs6000_isa_flags_explicit): Likewise.
524
525 * config/rs6000/rs6000.c (rs6000_option_override_internal):
526 Initialize rs6000_isa_flags_explicit.
527 (rs6000_function_specific_save): Add gcc_options* parameter, so
528 that the powerpc builds after the 2013-10-15 changes.
529 (rs6000_function_specific_restore): Likewise.
530
c71aed47
DD
5312013-10-16 DJ Delorie <dj@redhat.com>
532
533 * config/rl78/rl78.c (rl78_alloc_address_registers_macax): Verify
534 op is a REG before checking REGNO.
535 (rl78_alloc_physical_registers): Verify pattern is a SET before
536 checking SET_SRC.
537
90d3bd51
BS
5382013-10-16 Bill Schmidt <wschmidt@linux.vnet.ibm.com>
539
540 * gcc/config/rs6000/vector.md (vec_unpacks_hi_v4sf): Correct for
541 endianness.
542 (vec_unpacks_lo_v4sf): Likewise.
543 (vec_unpacks_float_hi_v4si): Likewise.
544 (vec_unpacks_float_lo_v4si): Likewise.
545 (vec_unpacku_float_hi_v4si): Likewise.
546 (vec_unpacku_float_lo_v4si): Likewise.
547
de75c876
BS
5482013-10-16 Bill Schmidt <wschmidt@linux.vnet.ibm.com>
549
550 * config/rs6000/vsx.md (vsx_concat_<mode>): Adjust output for LE.
551 (vsx_concat_v2sf): Likewise.
552
21216998
JG
5532013-10-16 James Greenhalgh <james.greenhalgh@arm.com>
554
555 * config/aarch64/aarch64.md
556 (*mov<mode>_aarch64): Fix output template for DUP (element) Scalar.
557
1ef88893 5582013-10-16 Andrew MacLeod <amacleod@redhat.com>
7ab4168e 559
1ef88893 560 PR tree-optimization/58697
7ab4168e 561 * cfgloop.c (get_estimated_loop_iterations_int): Rename from
1ef88893
AM
562 estimated_loop_iterations_int.
563 (max_stmt_executions_int): Call get_max_loop_iterations_int.
7ab4168e 564 (get_max_loop_iterations_int): New. HWINT version of
1ef88893
AM
565 get_max_loop_iterations.
566 * cfgloop.h: Add prototypes.
567 * loop-iv.c (find_simple_exit): call get_estimated_loop_iterations_int.
7ab4168e 568 * loop-unroll.c (decide_peel_once_rolling): Call
1ef88893
AM
569 get_estimated_loop_iterations_int.
570 * tree-ssa-loop-niter.c (estimated_loop_iterations_int): Add back.
571 * tree-ssa-loop-niter.h: Tweak prototypes.
572
c22df64f
DM
5732013-10-16 David Malcolm <dmalcolm@redhat.com>
574
575 * gengtype-parse.c (struct_field_seq): Ignore access-control
576 keywords ("public:" etc).
577
f24bb080
MS
5782013-10-16 Marcus Shawcroft <marcus.shawcroft@arm.com>
579
580 * config/aarch64/aarch64.c (aarch64_regno_regclass): Classify
581 FRAME_POINTER_REGNUM and ARG_POINTER_REGNUM as POINTER_REGS.
582
c281bf26
YR
5832013-10-16 Yvan Roux <yvan.roux@linaro.org>
584
585 * config/arm/arm.opt (mlra): New option.
586 * config/arm/arm.c (arm_lra_p): New function.
587 (TARGET_LRA_P): Define.
588
5806f481
PM
5892013-10-16 Paulo Matos <pmatos@broadcom.com>
590
591 * tree-core.h (tree_code_name): Remove.
592 * tree.h (get_tree_code_name): New prototype.
593 * tree.c (tree_code_name): Make static.
594 (get_tree_code_name): New function.
595 (dump_tree_statistics, tree_check_failed, tree_not_check_failed,
596 tree_class_check_failed, tree_range_check_failed,
597 tree_not_class_check_failed, omp_clause_check_failed,
598 tree_contains_struct_check_failed, tree_operand_check_failed): Use new
599 wrapper get_tree_code_name instead of calling tree_code_name directly.
600 * tree-vrp.c (dump_asserts_for): Likewise.
601 * tree-dump.c (dequeue_and_dump): Likewise.
602 * tree-pretty-print.c (do_niy, dump_generic_node): Likewise.
603 * tree-pretty-print.h (pp_unsupported_tree): Likewise.
604 * lto-streamer-out.c (lto_write_tree, DFS_write_tree): Likewise.
605 * tree-ssa-dom.c (print_expr_hash_elt): Likewise.
606 * gimple-pretty-print.c (dump_unary_rhs, dump_binary_rhs,
607 dump_ternary_rhs, dump_gimple_assign, dump_gimple_cond,
608 dump_gimple_omp_for): Likewise.
609 * tree-vect-data-refs.c (vect_create_data_ref_ptr): Likewise.
610 * tree-ssa-pre.c (print_pre_expr): Likewise.
611 * ipa-prop.c (ipa_print_node_jump_functions_for_edge): Likewise.
612 * print-tree.c (print_node_brief, print_node): Likewise.
613 * gimple.c (gimple_check_failed): Likewise.
614 * lto-streamer.c (lto_tag_name, print_lto_report): Likewise.
615 * config/frv/frv.c (frv_init_cumulative_args): Likewise.
616 * config/mep/mep.c (mep_validate_vliw): Likewise.
617 * config/iq2000/iq2000.c (init_cumulative_args): Likewise.
618 * config/rs6000/rs6000.c (init_cumulative_args): Likewise.
619
82c9df56
GG
6202013-10-16 Ganesh Gopalasubramanian <Ganesh.Gopalasubramanian@amd.com>
621
622 * config/i386/i386.c (ix86_option_override_internal): Enable FMA4
623 for AMD bdver3.
624
b82d0df9
HPN
6252013-10-16 Hans-Peter Nilsson <hp@axis.com>
626
627 * config/cris/t-elfmulti (MULTILIB_OPTIONS, MULTILIB_DIRNAMES)
628 (MULTILIB_MATCHES): Add multilib for -march=v8.
629
bf7b5747
ST
6302013-10-15 Sriraman Tallam <tmsriram@google.com>
631
632 PR target/57756
633 * optc-save-gen.awk: Add extra parameter to the save and restore
634 target calls.
635 * opth-gen.awk: Generate new TARGET_* macros to accept a parameter.
636 * tree.c (build_optimization_node): New parameter. Add extra parameter
637 to call to cl_optimization_save.
638 (build_target_option_node): New parameter. Add extra parameter
639 to call to cl_target_option_save.
640 * tree.h (build_optimization_node): New parameter.
641 (build_target_option_node): New parameter.
642 * c-family/c-common.c (handle_optimize_attribute): Fix calls to
643 build_optimization_node and build_target_option_node.
644 * c-family/c-pragma.c (handle_pragma_optimize): Ditto.
645 (handle_pragma_push_options): Ditto.
646 * toplev.c (process_options): Ditto.
647 * opts.c (init_options_struct): Check for opts_set non-null.
648 * target.def (target_option.save): New parameter.
649 (target_option.restore): New parameter.
650 * tm.texi: Generate.
651 * config/i386/i386-c.c (ix86_target_macros_internal): Ditto.
652 (ix86_pragma_target_parse): Ditto.
653 * config/i386/i386-protos.h (ix86_valid_target_attribute_tree): New
654 parameters.
655 * config/rs6000/rs6000.c (rs6000_option_override_internal): Fix calls
656 to build_optimization_node and build_target_option_node.
657 (rs6000_valid_attribute_p): Ditto.
658 (rs6000_pragma_target_parse): Ditto.
659 * config/i386/i386.opt (x_ix86_target_flags_explicit): New TargetSave
660 data.
661 * config/i386/i386.h:
662 TARGET_64BIT_P: New Macro
663 TARGET_MMX_P: New Macro.
664 TARGET_3DNOW_P: New Macro.
665 TARGET_3DNOW_A_P: New Macro.
666 TARGET_SSE_P: New Macro.
667 TARGET_SSE2_P: New Macro.
668 TARGET_SSE3_P: New Macro.
669 TARGET_SSSE3_P: New Macro.
670 TARGET_SSE4_1_P: New Macro.
671 TARGET_SSE4_2_P: New Macro.
672 TARGET_AVX_P: New Macro.
673 TARGET_AVX2_P: New Macro.
674 TARGET_AVX512F_P: New Macro.
675 TARGET_AVX512PF_P: New Macro.
676 TARGET_AVX512ER_P: New Macro.
677 TARGET_AVX512CD_P: New Macro.
678 TARGET_FMA_P: New Macro.
679 TARGET_SSE4A_P: New Macro.
680 TARGET_FMA4_P: New Macro.
681 TARGET_XOP_P: New Macro.
682 TARGET_LWP_P: New Macro.
683 TARGET_ABM_P: New Macro.
684 TARGET_BMI_P: New Macro.
685 TARGET_BMI2_P: New Macro.
686 TARGET_LZCNT_P: New Macro.
687 TARGET_TBM_P: New Macro.
688 TARGET_POPCNT_P: New Macro.
689 TARGET_SAHF_P: New Macro.
690 TARGET_MOVBE_P: New Macro.
691 TARGET_CRC32_P: New Macro.
692 TARGET_AES_P: New Macro.
693 TARGET_PCLMUL_P: New Macro.
694 TARGET_CMPXCHG16B_P: New Macro.
695 TARGET_FSGSBASE_P: New Macro.
696 TARGET_RDRND_P: New Macro.
697 TARGET_F16C_P: New Macro.
698 TARGET_RTM_P: New Macro.
699 TARGET_HLE_P: New Macro.
700 TARGET_RDSEED_P: New Macro.
701 TARGET_PRFCHW_P: New Macro.
702 TARGET_ADX_P: New Macro.
703 TARGET_FXSR_P: New Macro.
704 TARGET_XSAVE_P: New Macro.
705 TARGET_XSAVEOPT_P: New Macro.
706 TARGET_LP64_P: New Macro.
707 TARGET_X32_P: New Macro.
708 TARGET_FPMATH_DEFAULT_P: New Macro.
709 TARGET_FLOAT_RETURNS_IN_80387_P: New Macro.
710 * config/i386/i386.c (ix86_option_override_internal): New parameters.
711 opts and opts_set.
712 Change ix86_tune_string to access opts->x_ix86_tune_string.
713 Change ix86_isa_flags to access opts->x_ix86_isa_flags.
714 Change ix86_arch_string to access opts->x_ix86_arch_string.
715 Change ix86_stringop_alg to access opts->x_ix86_stringop_alg.
716 Change ix86_pmode to access opts->x_ix86_pmode.
717 Change ix86_abi to access opts->x_ix86_abi.
718 Change ix86_cmodel to access opts->x_ix86_cmodel.
719 Change ix86_asm_dialect to access opts->x_ix86_asm_dialect.
720 Change ix86_isa_flags_explicit to access
721 opts->x_ix86_isa_flags_explicit.
722 Change ix86_dump_tunes to access opts->x_ix86_dump_tunes.
723 Change ix86_regparm to access opts->x_ix86_regparm.
724 Change ix86_branch_cost to access opts->x_ix86_branch_cost.
725 Change ix86_preferred_stack_boundary_arg to access
726 opts->x_ix86_preferred_stack_boundary_arg.
727 Change ix86_force_align_arg_pointer to access
728 opts->x_ix86_force_align_arg_pointer.
729 Change ix86_incoming_stack_boundar_arg to access
730 opts->x_ix86_incoming_stack_boundar_arg.
731 Change ix86_fpmath to access opts->x_ix86_fpmath.
732 Change ix86_veclibabi_type to access opts->x_ix86_veclibabi_type.
733 Change ix86_recip_name to access opts->x_ix86_recip_name.
734 Change ix86_stack_protector_guard to access
735 opts->x_ix86_stack_protector_guard.
736 Change ix86_tune_memcpy_strategy to access
737 opts->x_ix86_tune_memcpy_strategy.
738 Change ix86_tune_memset_strategy to access
739 opts->x_ix86_tune_memset_strategy.
740 Change global_options to access opts.
741 Change global_options_set to access opts_set.
6f1abb55
UB
742 Change TARGET_64BIT to TARGET_64BIT_P (opts->...).
743 Change TARGET_MMX to TARGET_MMX_P (opts->...).
744 Change TARGET_3DNOW to TARGET_3DNOW_P (opts->...).
745 Change TARGET_3DNOW_A to TARGET_3DNOW_A_P (opts->...).
746 Change TARGET_SSE to TARGET_SSE_P (opts->...).
747 Change TARGET_SSE2 to TARGET_SSE2_P (opts->...).
748 Change TARGET_SSE3 to TARGET_SSE3_P (opts->...).
749 Change TARGET_SSSE3 to TARGET_SSSE3_P (opts->...).
750 Change TARGET_SSE4_1 to TARGET_SSE4_1_P (opts->...).
751 Change TARGET_SSE4_2 to TARGET_SSE4_2_P (opts->...).
752 Change TARGET_AVX to TARGET_AVX_P (opts->...).
753 Change TARGET_AVX2 to TARGET_AVX2_P (opts->...).
754 Change TARGET_AVX512F to TARGET_AVX512F_P (opts->...).
755 Change TARGET_AVX512PF to TARGET_AVX512PF_P (opts->...).
756 Change TARGET_AVX512ER to TARGET_AVX512ER_P (opts->...).
757 Change TARGET_AVX512CD to TARGET_AVX512CD_P (opts->...).
758 Change TARGET_FMA to TARGET_FMA_P (opts->...).
759 Change TARGET_SSE4A to TARGET_SSE4A_P (opts->...).
760 Change TARGET_FMA4 to TARGET_FMA4_P (opts->...).
761 Change TARGET_XOP to TARGET_XOP_P (opts->...).
762 Change TARGET_LWP to TARGET_LWP_P (opts->...).
763 Change TARGET_ABM to TARGET_ABM_P (opts->...).
764 Change TARGET_BMI to TARGET_BMI_P (opts->...).
765 Change TARGET_BMI2 to TARGET_BMI2_P (opts->...).
766 Change TARGET_LZCNT to TARGET_LZCNT_P (opts->...).
767 Change TARGET_TBM to TARGET_TBM_P (opts->...).
768 Change TARGET_POPCNT to TARGET_POPCNT_P (opts->...).
769 Change TARGET_SAHF to TARGET_SAHF_P (opts->...).
770 Change TARGET_MOVBE to TARGET_MOVBE_P (opts->...).
771 Change TARGET_CRC32 to TARGET_CRC32_P (opts->...).
772 Change TARGET_AES to TARGET_AES_P (opts->...).
773 Change TARGET_PCLMUL to TARGET_PCLMUL_P (opts->...).
774 Change TARGET_CMPXCHG16B to TARGET_CMPXCHG16B_P (opts->...).
775 Change TARGET_FSGSBASE to TARGET_FSGSBASE_P (opts->...).
776 Change TARGET_RDRND to TARGET_RDRND_P (opts->...).
777 Change TARGET_F16C to TARGET_F16C_P (opts->...).
778 Change TARGET_RTM to TARGET_RTM_P (opts->...).
779 Change TARGET_HLE to TARGET_HLE_P (opts->...).
780 Change TARGET_RDSEED to TARGET_RDSEED_P (opts->...).
781 Change TARGET_PRFCHW to TARGET_PRFCHW_P (opts->...).
782 Change TARGET_ADX to TARGET_ADX_P (opts->...).
783 Change TARGET_FXSR to TARGET_FXSR_P (opts->...).
784 Change TARGET_XSAVE to TARGET_XSAVE_P (opts->...).
785 Change TARGET_XSAVEOPT to TARGET_XSAVEOPT_P (opts->...).
786 Change TARGET_LP64 to TARGET_LP64_P (opts->...).
787 Change TARGET_X32 to TARGET_X32_P (opts->...).
788 Change TARGET_FPMATH_DEFAULT to TARGET_FPMATH_DEFAULT_P (opts->...).
789 Change TARGET_FLOAT_RETURNS_IN_80387 to
790 TARGET_FLOAT_RETURNS_IN_80387_P (opts->...).
bf7b5747
ST
791 (ix86_function_specific_save): New parameter. Use opts-> fields
792 to replace global fields.
793 (ix86_function_specific_restore): Ditto.
794 (ix86_valid_target_attribute_inner_p): New parameters.
795 Fix recursive call.
796 Fix call to ix86_handle_option and set_option.
797 (ix86_valid_target_attribute_tree): New parameters.
798 Change global_options to access opts.
799 Change global_options_set to access opts_set.
800 Fix call to ix86_valid_target_attribute_inner_p.
801 Change ix86_tune_string to access opts->x_ix86_tune_string.
802 Change ix86_arch_string to access opts->x_ix86_arch_string.
803 Change ix86_fpmath to access opts->x_ix86_fpmath
804 Fix call to ix86_option_override_internal.
805 Fix call to ix86_add_new_builtins.
806 Fix calls to build_optimization_node and build_target_option_node.
807 (ix86_valid_target_attribute_p): Remove access to global_options.
808 Use new gcc_options structure func_options.
809 Fix call to ix86_valid_target_attribute_tree.
810 Fix call to build_optimization_node.
811 (get_builtin_code_for_version): Fix call to
812 ix86_valid_target_attribute_tree.
813
459260ec
DM
8142013-10-15 David Malcolm <dmalcolm@redhat.com>
815
816 * Makefile.in (PICFLAG): New.
817 (enable_host_shared): New.
818 (INTERNAL_CFLAGS): Use PICFLAG.
819 (LIBIBERTY): Use pic build of libiberty.a if configured with
820 --enable-host-shared.
821 * configure.ac: Add --enable-host-shared, setting up new
822 PICFLAG variable.
823 * configure: Regenerate.
6f1abb55 824 * doc/install.texi (--enable-shared): Add note contrasting it with ...
459260ec
DM
825 (--enable-host-shared): New option.
826
16022041
RS
8272013-10-15 Richard Biener <rguenther@suse.de>
828
829 * tree-tailcall.c (find_tail_calls): Don't use tail-call recursion
830 for built-in functions.
831
b114bfb4
ZC
8322013-10-15 Zhenqiang Chen <zhenqiang.chen@arm.com>
833
834 * tree-ssa-reassoc.c: Include rtl.h and tm_p.h.
835 (optimize_range_tests_1): New function,
836 extracted from optimize_range_tests.
837 (optimize_range_tests_xor): Similarly.
838 (optimize_range_tests_diff): New function.
839 (optimize_range_tests): Use optimize_range_tests_1.
840
69b8f2f9
CH
8412013-10-15 Cong Hou <congh@google.com>
842
7ab4168e
MS
843 * tree-vect-loop.c (vect_is_simple_reduction_1): Relax the
844 requirement of the reduction pattern so that one operand of the
69b8f2f9
CH
845 reduction operation can come from outside of the loop.
846
8ff5a4e4
JG
8472013-10-15 James Greenhalgh <james.greenhalgh@arm.com>
848
849 * config/arm/neon-schedgen.ml: Remove.
850 * config/arm/cortex-a9-neon.md: Remove comment regarding
851 neon-schedgen.ml.
852
52b69ad2
JG
8532013-10-15 James Greenhalgh <james.greenhalgh@arm.com>
854
855 * config/arm/types: Remove old neon types.
856
e7c6a556
JG
8572013-10-15 James Greenhalgh <james.greenhalgh@arm.com>
858
859 * config/arm/cortex-a7.md
860 (cortex_a7_neon_type): New.
861 (cortex_a7_neon_mul): Update for new types.
862 (cortex_a7_neon_mla): Likewise.
863 (cortex_a7_neon): Likewise.
864
795dd8d5
JG
8652013-10-15 James Greenhalgh <james.greenhalgh@arm.com>
866
867 * config/arm/cortex-a15-neon.md
868 (cortex_a15_neon_type): New,
869
870 (cortex_a15_neon_int_1): Remove.
871 (cortex_a15_neon_int_2): Likewise.
872 (cortex_a15_neon_int_3): Likewise.
873 (cortex_a15_neon_int_4): Likewise.
874 (cortex_a15_neon_int_5): Likewise.
875 (cortex_a15_neon_vqneg_vqabs): Likewise.
876 (cortex_a15_neon_vmov): Likewise.
877 (cortex_a15_neon_vaba): Likewise.
878 (cortex_a15_neon_vaba_qqq): Likewise.
879 (cortex_a15_neon_mul_ddd_8_16_qdd_16_8_long_32_16_long): Likewise.
880 (cortex_a15_neon_mul_qqq_8_16_32_ddd_32): Likewise.
881 (cortex_a15_neon_mul_qdd_64_32_long_qqd_16_ddd_32_scalar_64_32_long_scalar):
882 Likewise.
883 (cortex_a15_neon_mla_ddd_8_16_qdd_16_8_long_32_16_long): Likewise.
884 (cortex_a15_neon_mla_qqq_8_16): Likewise.
885 (cortex_a15_neon_mla_ddd_32_qqd_16_ddd_32_scalar): Likewise.
886 (cortex_a15_neon_mla_qqq_32_qqd_32_scalar): Likewise.
887 (cortex_a15_neon_mul_ddd_16_scalar_32_16_long_scalar): Likewise.
888 (cortex_a15_neon_mul_qqd_32_scalar): Likewise.
889 (cortex_a15_neon_mla_ddd_16_scalar_qdd_32_16_long_scalar): Likewise.
890 (cortex_a15_neon_shift_1): Likewise.
891 (cortex_a15_neon_shift_2): Likewise.
892 (cortex_a15_neon_shift_3): Likewise.
893 (cortex_a15_neon_vshl_ddd): Likewise.
894 (cortex_a15_neon_vqshl_vrshl_vqrshl_qqq): Likewise.
895 (cortex_a15_neon_vsra_vrsra): Likewise.
896 (cortex_a15_neon_fp_vmla_ddd_scalar): Likewise.
897 (cortex_a15_neon_fp_vmla_qqq_scalar): Likewise.
898 (cortex_a15_neon_bp_3cycle): Likewise.
899 (cortex_a15_neon_ldm_2): Likewise.
900 (cortex_a15_neon_stm_2): Likewise.
901 (cortex_a15_neon_mcr): Likewise.
902 (cortex_a15_neon_mrc): Likewise.
903 (cortex_a15_neon_fp_vadd_ddd_vabs_dd): Likewise.
904 (cortex_a15_neon_fp_vadd_qqq_vabs_qq): Likewise.
905 (cortex_a15_neon_fp_vmul_ddd): Likewise.
906 (cortex_a15_neon_fp_vmul_qqd): Likewise.
907 (cortex_a15_neon_fp_vmla_ddd): Likewise.
908 (cortex_a15_neon_fp_vmla_qqq): Likewise.
909 (cortex_a15_neon_fp_vmla_ddd_scalar): Likewise.
910 (cortex_a15_neon_fp_vmla_qqq_scalar): Likewise.
911 (cortex_a15_neon_fp_vrecps_vrsqrts_ddd): Likewise.
912 (cortex_a15_neon_fp_vrecps_vrsqrts_qqq): Likewise.
913 (cortex_a15_neon_bp_simple): Likewise.
914 (cortex_a15_neon_bp_2cycle): Likewise.
915 (cortex_a15_neon_bp_3cycle): Likewise.
916 (cortex_a15_neon_vld1_1_2_regs): Likewise.
917 (cortex_a15_neon_vld1_3_4_regs): Likewise.
918 (cortex_a15_neon_vld2_2_regs_vld1_vld2_all_lanes): Likewise.
919 (cortex_a15_neon_vld2_4_regs): Likewise.
920 (cortex_a15_neon_vld3_vld4): Likewise.
921 (cortex_a15_neon_vst1_1_2_regs_vst2_2_regs): Likewise.
922 (cortex_a15_neon_vst1_3_4_regs): Likewise.
923 (cortex_a15_neon_vst2_4_regs_vst3_vst4): Rename to...
924 (cortex_a15_neon_vst2_4_regs_vst3): ...This, update for new attributes.
925 (cortex_a15_neon_vst3_vst4): Rename to...
926 (cortex_a15_neon_vst4): This, update for new attributes.
927 (cortex_a15_neon_vld1_vld2_lane): Update for new attributes.
928 (cortex_a15_neon_vld3_vld4_lane): Likewise.
929 (cortex_a15_neon_vst1_vst2_lane): Likewise.
930 (cortex_a15_neon_vst3_vst4_lane): Likewise.
931 (cortex_a15_neon_vld3_vld4_all_lanes): Likewise.
932 (cortex_a15_neon_ldm_2): Likewise.
933 (cortex_a15_neon_stm_2): Likewise.
934 (cortex_a15_neon_mcr): Likewise.
935 (cortex_a15_neon_mcr_2_mcrr): Likewise.
936 (cortex_a15_neon_mrc): Likewise.
937 (cortex_a15_neon_mrrc): Likewise.
938
939 (cortex_a15_neon_abd): New.
940 (cortex_a15_neon_abd_q): Likewise.
941 (cortex_a15_neon_aba): Likewise.
942 (cortex_a15_neon_aba_q): Likewise.
943 (cortex_a15_neon_acc): Likewise.
944 (cortex_a15_neon_acc_q): Likewise.
945 (cortex_a15_neon_arith_basic): Likewise.
946 (cortex_a15_neon_arith_complex): Likewise.
947 (cortex_a15_neon_multiply): Likewise.
948 (cortex_a15_neon_multiply_q): Likewise.
949 (cortex_a15_neon_mla): Likewise.
950 (cortex_a15_neon_mla_q): Likewise.
951 (cortex_a15_neon_sat_mla_long): Likewise.
952 (cortex_a15_neon_shift_acc): Likewise.
953 (cortex_a15_neon_shift_imm_basic): Likewise.
954 (cortex_a15_neon_shift_imm_complex): Likewise.
955 (cortex_a15_neon_shift_reg_basic): Likewise.
956 (cortex_a15_neon_shift_reg_basic_q): Likewise.
957 (cortex_a15_neon_shift_reg_complex): Likewise.
958 (cortex_a15_neon_shift_reg_complex_q): Likewise.
959 (cortex_a15_neon_fp_negabs): Likewise
960 (cortex_a15_neon_fp_arith): Likewise
961 (cortex_a15_neon_fp_arith_q): Likewise
962 (cortex_a15_neon_fp_cvt_int): Likewise
963 (cortex_a15_neon_fp_cvt_int_q): Likewise
964 (cortex_a15_neon_fp_cvt_16): Likewise
965 (cortex_a15_neon_fp_mul): Likewise
966 (cortex_a15_neon_fp_mul_q): Likewise
967 (cortex_a15_neon_fp_mla): Likewise
968 (cortex_a15_neon_fp_mla_q): Likewise
969 (cortex_a15_neon_fp_recps_rsqrte): Likewise.
970 (cortex_a15_neon_fp_recps_rsqrte_q): Likewise.
971 (cortex_a15_neon_bitops): Likewise.
972 (cortex_a15_neon_bitops_q): Likewise.
973 (cortex_a15_neon_from_gp): Likewise.
974 (cortex_a15_neon_from_gp_q): Likewise.
975 (cortex_a15_neon_tbl3_tbl4): Likewise.
976 (cortex_a15_neon_zip_q): Likewise.
977 (cortex_a15_neon_to_gp): Likewise.
978 (cortex_a15_neon_load_a): Likewise.
979 (cortex_a15_neon_load_b): Likewise.
980 (cortex_a15_neon_load_c): Likewise.
981 (cortex_a15_neon_load_d): Likewise.
982 (cortex_a15_neon_load_e): Likewise.
983 (cortex_a15_neon_load_f): Likewise.
984 (cortex_a15_neon_store_a): Likewise.
985 (cortex_a15_neon_store_b): Likewise.
986 (cortex_a15_neon_store_c): Likewise.
987 (cortex_a15_neon_store_d): Likewise.
988 (cortex_a15_neon_store_e): Likewise.
989 (cortex_a15_neon_store_f): Likewise.
990 (cortex_a15_neon_store_g): Likewise.
991 (cortex_a15_neon_store_h): Likewise.
992 (cortex_a15_vfp_to_from_gp): Likewise.
993
bc5e57e3
JG
9942013-10-15 James Greenhalgh <james.greenhalgh@arm.com>
995
996 * config/arm/cortex-a9-neon.md (cortex_a9_neon_type): New.
997
998 (cortex_a9_neon_vshl_ddd): Remove.
999 (cortex_a9_neon_vst3_vst4): Likewise.
1000 (cortex_a9_neon_vld3_vld4_all_lanes): Likewise.
1001
1002 (cortex_a9_neon_bit_ops_q): New.
1003
1004 (cortex_a9_neon_int_1): Use cortex_a8_neon_type.
1005 (cortex_a9_neon_int_2): Likewise.
1006 (cortex_a9_neon_int_3): Likewise.
1007 (cortex_a9_neon_int_4): Likewise.
1008 (cortex_a9_neon_int_5): Likewise.
1009 (cortex_a9_neon_vqneg_vqabs): Likewise.
1010 (cortex_a9_neon_vmov): Likewise.
1011 (cortex_a9_neon_vaba): Likewise.
1012 (cortex_a9_neon_vaba_qqq): Likewise.
1013 (cortex_a9_neon_shift_1): Likewise.
1014 (cortex_a9_neon_shift_2): Likewise.
1015 (cortex_a9_neon_shift_3): Likewise.
1016 (cortex_a9_neon_vqshl_vrshl_vqrshl_qqq): Likewise.
1017 (cortex_a9_neon_vsra_vrsra): Likewise.
1018 (cortex_a9_neon_mul_ddd_8_16_qdd_16_8_long_32_16_long): Likewise.
1019 (cortex_a9_neon_mul_qqq_8_16_32_ddd_32): Likewise.
1020 (cortex_a9_neon_mul_qdd_64_32_long_qqd_16_ddd_32_scalar_64_32_long_scalar):
1021 Likewise.
1022 (cortex_a9_neon_mla_ddd_8_16_qdd_16_8_long_32_16_long): Likewise.
1023 (cortex_a9_neon_mla_qqq_8_16): Likewise.
1024 (cortex_a9_neon_mla_ddd_32_qqd_16_ddd_32_scalar_qdd_64_32_long_scalar_qdd_64_32_long):
1025 Likewise.
1026 (cortex_a9_neon_mla_qqq_32_qqd_32_scalar): Likewise.
1027 (cortex_a9_neon_mul_ddd_16_scalar_32_16_long_scalar): Likewise.
1028 (cortex_a9_neon_mul_qqd_32_scalar): Likewise.
1029 (cortex_a9_neon_mla_ddd_16_scalar_qdd_32_16_long_scalar): Likewise.
1030 (cortex_a9_neon_fp_vadd_ddd_vabs_dd): Likewise.
1031 (cortex_a9_neon_fp_vadd_qqq_vabs_qq): Likewise.
1032 (cortex_a9_neon_fp_vsum): Likewise.
1033 (cortex_a9_neon_fp_vmul_ddd): Likewise.
1034 (cortex_a9_neon_fp_vmul_qqd): Likewise.
1035 (cortex_a9_neon_fp_vmla_ddd): Likewise.
1036 (cortex_a9_neon_fp_vmla_qqq): Likewise.
1037 (cortex_a9_neon_fp_vmla_ddd_scalar): Likewise.
1038 (cortex_a9_neon_fp_vmla_qqq_scalar): Likewise.
1039 (cortex_a9_neon_fp_vrecps_vrsqrts_ddd): Likewise.
1040 (cortex_a9_neon_fp_vrecps_vrsqrts_qqq): Likewise.
1041 (cortex_a9_neon_bp_simple): Likewise.
1042 (cortex_a9_neon_bp_2cycle): Likewise.
1043 (cortex_a9_neon_bp_3cycle): Likewise.
1044 (cortex_a9_neon_ldr): Likewise.
1045 (cortex_a9_neon_str): Likewise.
1046 (cortex_a9_neon_vld1_1_2_regs): Likewise.
1047 (cortex_a9_neon_vld1_3_4_regs): Likewise.
1048 (cortex_a9_neon_vld2_2_regs_vld1_vld2_all_lanes): Likewise.
1049 (cortex_a9_neon_vld2_4_regs): Likewise.
1050 (cortex_a9_neon_vld3_vld4): Likewise.
1051 (cortex_a9_neon_vld1_vld2_lane): Likewise.
1052 (cortex_a9_neon_vld3_vld4_lane): Likewise.
1053 (cortex_a9_neon_vld3_vld4_all_lanes): Likewise.
1054 (cortex_a9_neon_vst1_1_2_regs_vst2_2_regs): Likewise.
1055 (cortex_a9_neon_vst1_3_4_regs): Likewise.
1056 (cortex_a9_neon_vst2_4_regs_vst3_vst4): Likewise.
1057 (cortex_a9_neon_vst1_vst2_lane): Likewise.
1058 (cortex_a9_neon_vst3_vst4_lane): Likewise.
1059 (cortex_a9_neon_mcr): Likewise.
1060 (cortex_a9_neon_mcr_2_mcrr): Likewise.
1061 (cortex_a9_neon_mrc): Likewise.
1062 (cortex_a9_neon_mrrc): Likewise.
1063
9deda8cb
JG
10642013-10-15 James Greenhalgh <james.greenhalgh@arm.com>
1065
1066 * config/arm/cortex-a8-neon.md (cortex_a8_neon_type): New.
1067
1068 (cortex_a8_neon_vshl_ddd): Remove.
1069 (cortex_a8_neon_vst3_vst4): Likewise.
1070 (cortex_a8_neon_vld3_vld4_all_lanes): Likewise.
1071
1072 (cortex_a8_neon_bit_ops_q): New.
1073
1074 (cortex_a8_neon_int_1): Use cortex_a8_neon_type.
1075 (cortex_a8_neon_int_2): Likewise..
1076 (cortex_a8_neon_int_3): Likewise.
1077 (cortex_a8_neon_int_5): Likewise.
1078 (cortex_a8_neon_vqneg_vqabs): Likewise.
1079 (cortex_a8_neon_int_4): Likewise.
1080 (cortex_a8_neon_vaba): Likewise.
1081 (cortex_a8_neon_vaba_qqq): Likewise.
1082 (cortex_a8_neon_shift_1): Likewise.
1083 (cortex_a8_neon_shift_2): Likewise.
1084 (cortex_a8_neon_shift_3): Likewise.
1085 (cortex_a8_neon_vqshl_vrshl_vqrshl_qqq): Likewise.
1086 (cortex_a8_neon_vsra_vrsra): Likewise.
1087 (cortex_a8_neon_mul_ddd_8_16_qdd_16_8_long_32_16_long): Likewise.
1088 (cortex_a8_neon_mul_qqq_8_16_32_ddd_32): Likewise.
1089 (cortex_a8_neon_mul_qdd_64_32_long_qqd_16_ddd_32_scalar_64_32_long_scalar):
1090 Likewise.
1091 (cortex_a8_neon_mla_ddd_8_16_qdd_16_8_long_32_16_long): Likewise.
1092 (cortex_a8_neon_mla_qqq_8_16): Likewise.
1093 (cortex_a8_neon_mla_ddd_32_qqd_16_ddd_32_scalar_qdd_64_32_long_scalar_qdd_64_32_long):
1094 Likewise.
1095 (cortex_a8_neon_mla_qqq_32_qqd_32_scalar): Likewise.
1096 (cortex_a8_neon_mul_ddd_16_scalar_32_16_long_scalar): Likewise.
1097 (cortex_a8_neon_mul_qqd_32_scalar): Likewise.
1098 (cortex_a8_neon_mla_ddd_16_scalar_qdd_32_16_long_scalar): Likewise.
1099 (cortex_a8_neon_fp_vadd_ddd_vabs_dd): Likewise.
1100 (cortex_a8_neon_fp_vadd_qqq_vabs_qq): Likewise.
1101 (cortex_a8_neon_fp_vsum): Likewise.
1102 (cortex_a8_neon_fp_vmul_ddd): Likewise.
1103 (cortex_a8_neon_fp_vmul_qqd): Likewise.
1104 (cortex_a8_neon_fp_vmla_ddd): Likewise.
1105 (cortex_a8_neon_fp_vmla_qqq): Likewise.
1106 (cortex_a8_neon_fp_vmla_ddd_scalar): Likewise.
1107 (cortex_a8_neon_fp_vmla_qqq_scalar): Likewise.
1108 (cortex_a8_neon_fp_vrecps_vrsqrts_ddd): Likewise.
1109 (cortex_a8_neon_fp_vrecps_vrsqrts_qqq): Likewise.
1110 (cortex_a8_neon_bp_simple): Likewise.
1111 (cortex_a8_neon_bp_2cycle): Likewise.
1112 (cortex_a8_neon_bp_3cycle): Likewise.
1113 (cortex_a8_neon_ldr): Likewise.
1114 (cortex_a8_neon_str): Likewise.
1115 (cortex_a8_neon_vld1_1_2_regs): Likewise.
1116 (cortex_a8_neon_vld1_3_4_regs): Likewise.
1117 (cortex_a8_neon_vld2_2_regs_vld1_vld2_all_lanes): Likewise.
1118 (cortex_a8_neon_vld2_4_regs): Likewise.
1119 (cortex_a8_neon_vld3_vld4): Likewise.
1120 (cortex_a8_neon_vld1_vld2_lane): Likewise.
1121 (cortex_a8_neon_vld3_vld4_lane): Likewise.
1122 (cortex_a8_neon_vst1_1_2_regs_vst2_2_regs): Likewise.
1123 (cortex_a8_neon_vst1_3_4_regs): Likewise.
1124 (cortex_a8_neon_vst2_4_regs_vst3_vst4): Likewise.
1125 (cortex_a8_neon_vst1_vst2_lane): Likewise.
1126 (cortex_a8_neon_vst3_vst4_lane): Likewise.
1127 (cortex_a8_neon_mcr): Likewise.
1128 (cortex_a8_neon_mcr_2_mcrr): Likewise.
1129 (cortex_a8_neon_mrc): Likewise.
1130 (cortex_a8_neon_mrrc): Likewise.
1131
0f686aa9
JG
11322013-10-15 James Greenhalgh <james.greenhalgh@arm.com>
1133
1134 * config/aarch64/iterators.md (Vetype): Add SF and DF modes.
1135 (fp): New.
1136 * config/aarch64/aarch64-simd.md (neon_type): Remove.
1137 (aarch64_simd_dup<mode>): Add "type" attribute.
1138 (aarch64_dup_lane<mode>): Likewise.
1139 (aarch64_dup_lane_<vswap_width_name><mode>): Likewise.
1140 (*aarch64_simd_mov<mode>): Likewise.
1141 (aarch64_simd_mov_from_<mode>low): Likewise.
1142 (aarch64_simd_mov_from_<mode>high): Likewise.
1143 (orn<mode>3): Likewise.
1144 (bic<mode>3): Likewise.
1145 (add<mode>3): Likewise.
1146 (sub<mode>3): Likewise.
1147 (mul<mode>3): Likewise.
1148 (*aarch64_mul3_elt<mode>): Likewise.
1149 (*aarch64_mul3_elt_<vswap_width_name><mode>): Likewise.
1150 (*aarch64_mul3_elt_to_128df): Likewise.
1151 (*aarch64_mul3_elt_to_64v2df): Likewise.
1152 (neg<mode>2): Likewise.
1153 (abs<mode>2): Likewise.
1154 (abd<mode>_3): Likewise.
1155 (aba<mode>_3): Likewise.
1156 (fabd<mode>_3): Likewise.
1157 (*fabd_scalar<mode>3): Likewise.
1158 (and<mode>3): Likewise.
1159 (ior<mode>3): Likewise.
1160 (xor<mode>3): Likewise.
1161 (one_cmpl<mode>2): Likewise.
1162 (aarch64_simd_vec_set<mode>): Likewise.
1163 (aarch64_simd_lshr<mode>): Likewise.
1164 (aarch64_simd_ashr<mode>): Likewise.
1165 (aarch64_simd_imm_shl<mode>): Likewise.
1166 (aarch64_simd_reg_sshl<mode): Likewise.
1167 (aarch64_simd_reg_shl<mode>_unsigned): Likewise.
1168 (aarch64_simd_reg_shl<mode>_signed): Likewise.
1169 (aarch64_simd_vec_setv2di): Likewise.
1170 (aarch64_simd_vec_set<mode>): Likewise.
1171 (aarch64_mla<mode>): Likewise.
1172 (*aarch64_mla_elt<mode>): Likewise.
1173 (*aarch64_mla_elt_<vswap_width_name><mode>): Likewise.
1174 (aarch64_mls<mode>): Likewise.
1175 (*aarch64_mls_elt<mode>): Likewise.
1176 (*aarch64_mls_elt_<vswap_width_name><mode>): Likewise.
1177 (<su><maxmin><mode>3): Likewise.
1178 (move_lo_quad_<mode>): Likewise.
1179 (aarch64_simd_move_hi_quad_<mode>): Likewise.
1180 (aarch64_simd_vec_pack_trunc_<mode>): Likewise.
1181 (vec_pack_trunc_<mode>): Likewise.
1182 (aarch64_simd_vec_unpack<su>_lo_<mode>): Likewise.
1183 (aarch64_simd_vec_unpack<su>_hi_<mode>): Likewise.
1184 (*aarch64_<su>mlal_lo<mode>): Likewise.
1185 (*aarch64_<su>mlal_hi<mode>): Likewise.
1186 (*aarch64_<su>mlsl_lo<mode>): Likewise.
1187 (*aarch64_<su>mlsl_hi<mode>): Likewise.
1188 (*aarch64_<su>mlal<mode>): Likewise.
1189 (*aarch64_<su>mlsl<mode>): Likewise.
1190 (aarch64_simd_vec_<su>mult_lo_<mode>): Likewise.
1191 (aarch64_simd_vec_<su>mult_hi_<mode>): Likewise.
1192 (add<mode>3): Likewise.
1193 (sub<mode>3): Likewise.
1194 (mul<mode>3): Likewise.
1195 (div<mode>3): Likewise.
1196 (neg<mode>2): Likewise.
1197 (abs<mode>2): Likewise.
1198 (fma<mode>4): Likewise.
1199 (*aarch64_fma4_elt<mode>): Likewise.
1200 (*aarch64_fma4_elt_<vswap_width_name><mode>): Likewise.
1201 (*aarch64_fma4_elt_to_128df): Likewise.
1202 (*aarch64_fma4_elt_to_64v2df): Likewise.
1203 (fnma<mode>4): Likewise.
1204 (*aarch64_fnma4_elt<mode>): Likewise.
1205 (*aarch64_fnma4_elt_<vswap_width_name><mode>
1206 (*aarch64_fnma4_elt_to_128df): Likewise.
1207 (*aarch64_fnma4_elt_to_64v2df): Likewise.
1208 (<frint_pattern><mode>2): Likewise.
1209 (l<fcvt_pattern><su_optab><VDQF:mode><fcvt_target>2): Likewise.
1210 (<optab><fcvt_target><VDQF:VDQF:mode>2): Likewise.
1211 (vec_unpacks_lo_v4sf): Likewise.
1212 (aarch64_float_extend_lo_v2df): Likewise.
1213 (vec_unpacks_hi_v4sf): Likewise.
1214 (aarch64_float_truncate_lo_v2sf): Likewise.
1215 (aarch64_float_truncate_hi_v4sf): Likewise.
1216 (aarch64_vmls<mode>): Likewise.
1217 (<su><maxmin><mode>3): Likewise.
1218 (<maxmin_uns><mode>3): Likewise.
1219 (reduc_<sur>plus_<mode>): Likewise.
1220 (reduc_<sur>plus_v2di): Likewise.
1221 (reduc_<sur>plus_v2si): Likewise.
1222 (reduc_<sur>plus_<mode>): Likewise.
1223 (aarch64_addpv4sf): Likewise.
1224 (clz<mode>2): Likewise.
1225 (reduc_<maxmin_uns>_<mode>): Likewise.
1226 (reduc_<maxmin_uns>_v2di): Likewise.
1227 (reduc_<maxmin_uns>_v2si): Likewise.
1228 (reduc_<maxmin_uns>_<mode>): Likewise.
1229 (reduc_<maxmin_uns>_v4sf): Likewise.
1230 (aarch64_simd_bsl<mode>_internal): Likewise.
1231 (*aarch64_get_lane_extend<GPI:mode><VDQQH:mode>): Likewise.
1232 (*aarch64_get_lane_zero_extendsi<mode>): Likewise.
1233 (aarch64_get_lane<mode>): Likewise.
1234 (*aarch64_combinez<mode>): Likewise.
1235 (aarch64_combine<mode>): Likewise.
1236 (aarch64_simd_combine<mode>): Likewise.
1237 (aarch64_<ANY_EXTEND:su><ADDSUB:optab>l<mode>_hi_internal): Likewise.
1238 (aarch64_<ANY_EXTEND:su><ADDSUB:optab>l<mode>_lo_internal): Likewise.
1239 (aarch64_<ANY_EXTEND:su><ADDSUB:optab>l<mode>): Likewise.
1240 (aarch64_<ANY_EXTEND:su><ADDSUB:optab>w<mode>): Likewise.
1241 (aarch64_<ANY_EXTEND:su><ADDSUB:optab>w2<mode>_internal): Likewise.
1242 (aarch64_<sur>h<addsub><mode>): Likewise.
1243 (aarch64_<sur><addsub>hn<mode>): Likewise.
1244 (aarch64_<sur><addsub>hn2<mode>): Likewise.
1245 (aarch64_pmul<mode>): Likewise.
1246 (aarch64_<su_optab><optab><mode>): Likewise.
1247 (aarch64_<sur>qadd<mode>): Likewise.
1248 (aarch64_sqmovun<mode>): Likewise.
1249 (aarch64_<sur>qmovn<mode>): Likewise.
1250 (aarch64_s<optab><mode>): Likewise.
1251 (aarch64_sq<r>dmulh<mode>): Likewise.
1252 (aarch64_sq<r>dmulh_lane<mode>): Likewise.
1253 (aarch64_sq<r>dmulh_laneq<mode>): Likewise.
1254 (aarch64_sq<r>dmulh_lane<mode>): Likewise.
1255 (aarch64_sqdml<SBINQOPS:as>l<mode>): Likewise.
1256 (aarch64_sqdml<SBINQOPS:as>l_lane<mode>_internal): Likewise.
1257 (aarch64_sqdml<SBINQOPS:as>l_lane<mode>_internal): Likewise.
1258 (aarch64_sqdml<SBINQOPS:as>l_n<mode>): Likewise.
1259 (aarch64_sqdml<SBINQOPS:as>l2<mode>_internal): Likewise.
1260 (aarch64_sqdml<SBINQOPS:as>l2_lane<mode>_internal): Likewise.
1261 (aarch64_sqdml<SBINQOPS:as>l2_n<mode>_internal): Likewise.
1262 (aarch64_sqdmull<mode>): Likewise.
1263 (aarch64_sqdmull_lane<mode>_internal): Likewise.
1264 (aarch64_sqdmull_n<mode>): Likewise.
1265 (aarch64_sqdmull2<mode>_internal): Likewise.
1266 (aarch64_sqdmull2_lane<mode>_internal): Likewise.
1267 (aarch64_sqdmull2_n<mode>_internal): Likewise.
1268 (aarch64_<sur>shl<mode>): Likewise.
1269 (aarch64_<sur>q<r>shl<mode>
1270 (aarch64_<sur>shll_n<mode>): Likewise.
1271 (aarch64_<sur>shll2_n<mode>): Likewise.
1272 (aarch64_<sur>shr_n<mode>): Likewise.
1273 (aarch64_<sur>sra_n<mode>): Likewise.
1274 (aarch64_<sur>s<lr>i_n<mode>): Likewise.
1275 (aarch64_<sur>qshl<u>_n<mode>): Likewise.
1276 (aarch64_<sur>q<r>shr<u>n_n<mode>): Likewise.
1277 (aarch64_cm<optab><mode>): Likewise.
1278 (aarch64_cm<optab>di): Likewise.
1279 (aarch64_cm<optab><mode>): Likewise.
1280 (aarch64_cm<optab>di): Likewise.
1281 (aarch64_cmtst<mode>): Likewise.
1282 (aarch64_cmtstdi): Likewise.
1283 (aarch64_cm<optab><mode>): Likewise.
1284 (*aarch64_fac<optab><mode>): Likewise.
1285 (aarch64_addp<mode>): Likewise.
1286 (aarch64_addpdi): Likewise.
1287 (sqrt<mode>2): Likewise.
1288 (vec_load_lanesoi<mode>): Likewise.
1289 (vec_store_lanesoi<mode>): Likewise.
1290 (vec_load_lanesci<mode>): Likewise.
1291 (vec_store_lanesci<mode>): Likewise.
1292 (vec_load_lanesxi<mode>): Likewise.
1293 (vec_store_lanesxi<mode>): Likewise.
1294 (*aarch64_mov<mode>): Likewise.
1295 (aarch64_ld2<mode>_dreg): Likewise.
1296 (aarch64_ld2<mode>_dreg): Likewise.
1297 (aarch64_ld3<mode>_dreg): Likewise.
1298 (aarch64_ld3<mode>_dreg): Likewise.
1299 (aarch64_ld4<mode>_dreg): Likewise.
1300 (aarch64_ld4<mode>_dreg): Likewise.
1301 (aarch64_tbl1<mode>): Likewise.
1302 (aarch64_tbl2v16qi): Likewise.
1303 (aarch64_combinev16qi): Likewise.
1304 (aarch64_<PERMUTE:perm_insn><PERMUTE:perm_hilo><mode>): Likewise.
1305 (aarch64_st2<mode>_dreg): Likewise.
1306 (aarch64_st2<mode>_dreg): Likewise.
1307 (aarch64_st3<mode>_dreg): Likewise.
1308 (aarch64_st3<mode>_dreg): Likewise.
1309 (aarch64_st4<mode>_dreg): Likewise.
1310 (aarch64_st4<mode>_dreg): Likewise.
1311 (*aarch64_simd_ld1r<mode>): Likewise.
1312 (aarch64_frecpe<mode>): Likewise.
1313 (aarch64_frecp<FRECP:frecp_suffix><mode>): Likewise.
1314 (aarch64_frecps<mode>): Likewise.
1315
f7379e5e
JG
13162013-10-15 James Greenhalgh <james.greenhalgh@arm.com>
1317
1318 * config/arm/iterators.md (V_elem_ch): New.
1319 (q): Likewise.
1320 (VQH_type): Likewise.
1321 * config/arm/arm.md (is_neon_type): New.
1322 (conds): Use is_neon_type.
1323 (anddi3_insn): Update type attribute.
1324 (xordi3_insn): Likewise.
1325 (one_cmpldi2): Likewise.
1326 * gcc/config/arm/vfp.md (movhf_vfp_neon): Update type attribute.
1327 * gcc/config/arm/neon.md (neon_mov): Update type attribute.
1328 (*movmisalign<mode>_neon_store): Likewise.
1329 (*movmisalign<mode>_neon_load): Likewise.
1330 (vec_set<mode>_internal): Likewise.
1331 (vec_set<mode>_internal): Likewise.
1332 (vec_setv2di_internal): Likewise.
1333 (vec_extract<mode>): Likewise.
1334 (vec_extract<mode>): Likewise.
1335 (vec_extractv2di): Likewise.
1336 (*add<mode>3_neon): Likewise.
1337 (adddi3_neon): Likewise.
1338 (*sub<mode>3_neon): Likewise.
1339 (subdi3_neon): Likewise.
1340 (fma<VCVTF:mode>4): Likewise.
1341 (fma<VCVTF:mode>4_intrinsic): Likewise.
1342 (*fmsub<VCVTF:mode>4): Likewise.
1343 (fmsub<VCVTF:mode>4_intrinsic): Likewise.
1344 (neon_vrint<NEON_VRINT:nvrint_variant><VCVTF:mode>): Likewise.
1345 (ior<mode>3): Likewise.
1346 (and<mode>3): Likewise.
1347 (orn<mode>3_neon): Likewise.
1348 (orndi3_neon): Likewise.
1349 (bic<mode>3_neon): Likewise.
1350 (bicdi3_neon): Likewise.
1351 (xor<mode>3): Likewise.
1352 (one_cmpl<mode>2): Likewise.
1353 (abs<mode>2): Likewise.
1354 (neg<mode>2): Likewise.
1355 (negdi2_neon): Likewise.
1356 (*umin<mode>3_neon): Likewise.
1357 (*umax<mode>3_neon): Likewise.
1358 (*smin<mode>3_neon): Likewise.
1359 (*smax<mode>3_neon): Likewise.
1360 (vashl<mode>3): Likewise.
1361 (vashr<mode>3_imm): Likewise.
1362 (vlshr<mode>3_imm): Likewise.
1363 (ashl<mode>3_signed): Likewise.
1364 (ashl<mode>3_unsigned): Likewise.
1365 (neon_load_count): Likewise.
1366 (ashldi3_neon_noclobber): Likewise.
1367 (ashldi3_neon): Likewise.
1368 (signed_shift_di3_neon): Likewise.
1369 (unsigned_shift_di3_neon): Likewise.
1370 (ashrdi3_neon_imm_noclobber): Likewise.
1371 (lshrdi3_neon_imm_noclobber): Likewise.
1372 (<shift>di3_neon): Likewise.
1373 (widen_ssum<mode>3): Likewise.
1374 (widen_usum<mode>3): Likewise.
1375 (quad_halves_<code>v4si): Likewise.
1376 (quad_halves_<code>v4sf): Likewise.
1377 (quad_halves_<code>v8hi): Likewise.
1378 (quad_halves_<code>v16qi): Likewise.
1379 (reduc_splus_v2di): Likewise.
1380 (neon_vpadd_internal<mode>): Likewise.
1381 (neon_vpsmin<mode>): Likewise.
1382 (neon_vpsmax<mode>): Likewise.
1383 (neon_vpumin<mode>): Likewise.
1384 (neon_vpumax<mode>): Likewise.
1385 (*ss_add<mode>_neon): Likewise.
1386 (*us_add<mode>_neon): Likewise.
1387 (*ss_sub<mode>_neon): Likewise.
1388 (*us_sub<mode>_neon): Likewise.
1389 (neon_vadd<mode>_unspec): Likewise.
1390 (neon_vaddl<mode>): Likewise.
1391 (neon_vaddw<mode>): Likewise.
1392 (neon_vhadd<mode>): Likewise.
1393 (neon_vqadd<mode>): Likewise.
1394 (neon_vaddhn<mode>): Likewise.
1395 (neon_vmul<mode>): Likewise.
1396 (neon_vfms<VCVTF:mode>): Likewise.
1397 (neon_vmlal<mode>): Likewise.
1398 (neon_vmls<mode>): Likewise.
1399 (neon_vmlsl<mode>): Likewise.
1400 (neon_vqdmulh<mode>): Likewise.
1401 (neon_vqdmlal<mode>): Likewise.
1402 (neon_vqdmlsl<mode>): Likewise.
1403 (neon_vmull<mode>): Likewise.
1404 (neon_vqdmull<mode>): Likewise.
1405 (neon_vsub<mode>_unspec): Likewise.
1406 (neon_vsubl<mode>): Likewise.
1407 (neon_vsubw<mode>): Likewise.
1408 (neon_vqsub<mode>): Likewise.
1409 (neon_vhsub<mode>): Likewise.
1410 (neon_vsubhn<mode>): Likewise.
1411 (neon_vceq<mode>): Likewise.
1412 (neon_vcge<mode>): Likewise.
1413 (neon_vcgeu<mode>): Likewise.
1414 (neon_vcgt<mode>): Likewise.
1415 (neon_vcgtu<mode>): Likewise.
1416 (neon_vcle<mode>): Likewise.
1417 (neon_vclt<mode>): Likewise.
1418 (neon_vcage<mode>): Likewise.
1419 (neon_vcagt<mode>): Likewise.
1420 (neon_vtst<mode>): Likewise.
1421 (neon_vabd<mode>): Likewise.
1422 (neon_vabdl<mode>): Likewise.
1423 (neon_vaba<mode>): Likewise.
1424 (neon_vabal<mode>): Likewise.
1425 (neon_vmax<mode>): Likewise.
1426 (neon_vmin<mode>): Likewise.
1427 (neon_vpaddl<mode>): Likewise.
1428 (neon_vpadal<mode>): Likewise.
1429 (neon_vpmax<mode>): Likewise.
1430 (neon_vpmin<mode>): Likewise.
1431 (neon_vrecps<mode>): Likewise.
1432 (neon_vrsqrts<mode>): Likewise.
1433 (neon_vqabs<mode>): Likewise.
1434 (neon_vqneg<mode>): Likewise.
1435 (neon_vcls<mode>): Likewise.
1436 (clz<mode>2): Likewise.
1437 (popcount<mode>2): Likewise.
1438 (neon_vrecpe<mode>): Likewise.
1439 (neon_vrsqrte<mode>): Likewise.
1440 (neon_vget_lane<mode>_sext_internal): Likewise.
1441 (neon_vget_lane<mode>_zext_internal): Likewise.
1442 (neon_vdup_n<mode>): Likewise.
1443 (neon_vdup_n<mode>): Likewise.
1444 (neon_vdup_nv2di): Likewise.
1445 (neon_vdup_lane<mode>_interal): Likewise.
1446 (*neon_vswp<mode>): Likewise.
1447 (neon_vcombine<mode>): Likewise.
1448 (float<mode><V_cvtto>2): Likewise.
1449 (floatuns<mode><V_cvtto>2): Likewise.
1450 (fix_trunc<mode><V_cvtto>2): Likewise.
1451 (fixuns_trunc<mode><V_cvtto>2
1452 (neon_vcvt<mode>): Likewise.
1453 (neon_vcvt<mode>): Likewise.
1454 (neon_vcvtv4sfv4hf): Likewise.
1455 (neon_vcvtv4hfv4sf): Likewise.
1456 (neon_vcvt_n<mode>): Likewise.
1457 (neon_vcvt_n<mode>): Likewise.
1458 (neon_vmovn<mode>): Likewise.
1459 (neon_vqmovn<mode>): Likewise.
1460 (neon_vqmovun<mode>): Likewise.
1461 (neon_vmovl<mode>): Likewise.
1462 (neon_vmul_lane<mode>): Likewise.
1463 (neon_vmul_lane<mode>): Likewise.
1464 (neon_vmull_lane<mode>): Likewise.
1465 (neon_vqdmull_lane<mode>): Likewise.
1466 (neon_vqdmulh_lane<mode>): Likewise.
1467 (neon_vqdmulh_lane<mode>): Likewise.
1468 (neon_vmla_lane<mode>): Likewise.
1469 (neon_vmla_lane<mode>): Likewise.
1470 (neon_vmlal_lane<mode>): Likewise.
1471 (neon_vqdmlal_lane<mode>): Likewise.
1472 (neon_vmls_lane<mode>): Likewise.
1473 (neon_vmls_lane<mode>): Likewise.
1474 (neon_vmlsl_lane<mode>): Likewise.
1475 (neon_vqdmlsl_lane<mode>): Likewise.
1476 (neon_vext<mode>): Likewise.
1477 (neon_vrev64<mode>): Likewise.
1478 (neon_vrev32<mode>): Likewise.
1479 (neon_vrev16<mode>): Likewise.
1480 (neon_vbsl<mode>_internal): Likewise.
1481 (neon_vshl<mode>): Likewise.
1482 (neon_vqshl<mode>): Likewise.
1483 (neon_vshr_n<mode>): Likewise.
1484 (neon_vshrn_n<mode>): Likewise.
1485 (neon_vqshrn_n<mode>): Likewise.
1486 (neon_vqshrun_n<mode>): Likewise.
1487 (neon_vshl_n<mode>): Likewise.
1488 (neon_vqshl_n<mode>): Likewise.
1489 (neon_vqshlu_n<mode>): Likewise.
1490 (neon_vshll_n<mode>): Likewise.
1491 (neon_vsra_n<mode>): Likewise.
1492 (neon_vsri_n<mode>): Likewise.
1493 (neon_vsli_n<mode>): Likewise.
1494 (neon_vtbl1v8qi): Likewise.
1495 (neon_vtbl2v8qi): Likewise.
1496 (neon_vtbl3v8qi): Likewise.
1497 (neon_vtbl4v8qi): Likewise.
1498 (neon_vtbl1v16qi): Likewise.
1499 (neon_vtbl2v16qi): Likewise.
1500 (neon_vcombinev16qi): Likewise.
1501 (neon_vtbx1v8qi): Likewise.
1502 (neon_vtbx2v8qi): Likewise.
1503 (neon_vtbx3v8qi): Likewise.
1504 (neon_vtbx4v8qi): Likewise.
1505 (*neon_vtrn<mode>_insn): Likewise.
1506 (*neon_vzip<mode>_insn): Likewise.
1507 (*neon_vuzp<mode>_insn): Likewise.
1508 (neon_vld1<mode>): Likewise.
1509 (neon_vld1_lane<mode>): Likewise.
1510 (neon_vld1_lane<mode>): Likewise.
1511 (neon_vld1_dup<mode>): Likewise.
1512 (neon_vld1_dup<mode>): Likewise.
1513 (neon_vld1_dupv2di): Likewise.
1514 (neon_vst1<mode>): Likewise.
1515 (neon_vst1_lane<mode>): Likewise.
1516 (neon_vst1_lane<mode>): Likewise.
1517 (neon_vld2<mode>): Likewise.
1518 (neon_vld2<mode>): Likewise.
1519 (neon_vld2_lane<mode>): Likewise.
1520 (neon_vld2_lane<mode>): Likewise.
1521 (neon_vld2_dup<mode>): Likewise.
1522 (neon_vst2<mode>): Likewise.
1523 (neon_vst2<mode>): Likewise.
1524 (neon_vst2_lane<mode>): Likewise.
1525 (neon_vst2_lane<mode>): Likewise.
1526 (neon_vld3<mode>): Likewise.
1527 (neon_vld3qa<mode>): Likewise.
1528 (neon_vld3qb<mode>): Likewise.
1529 (neon_vld3_lane<mode>): Likewise.
1530 (neon_vld3_lane<mode>): Likewise.
1531 (neon_vld3_dup<mode>): Likewise.
1532 (neon_vst3<mode>): Likewise.
1533 (neon_vst3qa<mode>): Likewise.
1534 (neon_vst3qb<mode>): Likewise.
1535 (neon_vst3_lane<mode>): Likewise.
1536 (neon_vst3_lane<mode>): Likewise.
1537 (neon_vld4<mode>): Likewise.
1538 (neon_vld4qa<mode>): Likewise.
1539 (neon_vld4qb<mode>): Likewise.
1540 (neon_vld4_lane<mode>): Likewise.
1541 (neon_vld4_lane<mode>): Likewise.
1542 (neon_vld4_dup<mode>): Likewise.
1543 (neon_vst4<mode>): Likewise.
1544 (neon_vst4qa<mode>): Likewise.
1545 (neon_vst4qb<mode>): Likewise.
1546 (neon_vst4_lane<mode>): Likewise.
1547 (neon_vst4_lane<mode>): Likewise.
1548 (neon_vec_unpack<US>_lo_<mode>): Likewise.
1549 (neon_vec_unpack<US>_hi_<mode>): Likewise.
1550 (neon_vec_<US>mult_lo_<mode>): Likewise.
1551 (neon_vec_<US>mult_hi_<mode>): Likewise.
1552 (neon_vec_<US>shiftl_<mode>): Likewise.
1553 (neon_unpack<US>_<mode>): Likewise.
1554 (neon_vec_<US>mult_<mode>): Likewise.
1555 (vec_pack_trunc_<mode>): Likewise.
1556 (neon_vec_pack_trunc_<mode>): Likewise.
1557 (neon_vabd<mode>_2): Likewise.
1558 (neon_vabd<mode>_3): Likewise.
1559
a9e66678
JG
15602013-10-15 James Greenhalgh <james.greenhalgh@arm.com>
1561
1562 * config/aarch64/aarch64.md (movtf_aarch64): Update type attribute.
1563 (load_pair): Update type attribute.
1564 (store_pair): Update type attribute.
1565 * config/aarch64/iterators.md (q): New.
1566
20445ed6
JG
15672013-10-15 James Greenhalgh <james.greenhalgh@arm.com>
1568
1569 * config/arm/types.md: Add new types for Neon insns.
1570
afb4ac68
AI
15712013-10-15 Alexander Ivchenko <alexander.ivchenko@intel.com>
1572 Maxim Kuznetsov <maxim.kuznetsov@intel.com>
1573 Sergey Lega <sergey.s.lega@intel.com>
1574 Anna Tikhonova <anna.tikhonova@intel.com>
1575 Ilya Tocar <ilya.tocar@intel.com>
1576 Andrey Turetskiy <andrey.turetskiy@intel.com>
1577 Ilya Verbin <ilya.verbin@intel.com>
1578 Kirill Yukhin <kirill.yukhin@intel.com>
1579 Michael Zolotukhin <michael.v.zolotukhin@intel.com>
1580
1581 * config/i386/sse.md (unspec): Add UNSPEC_RCP14, UNSPEC_RSQRT14,
1582 UNSPEC_FIXUPIMM, UNSPEC_SCALEF, UNSPEC_GETEXP, UNSPEC_GETMANT,
1583 UNSPEC_EXP2, UNSPEC_RCP28, UNSPEC_RSQRT28.
1584 (rcp14<mode>): New.
1585 (srcp14<mode>): Ditto.
1586 (rsqrt14<mode>): Ditto.
1587 (rsqrt14<mode>): Ditto.
1588 (avx512f_vmscalef<mode>): Ditto.
1589 (avx512f_scalef<mode>): Ditto.
1590 (avx512f_getexp<mode>): Ditto.
1591 (avx512f_sgetexp<mode>): Ditto.
1592 (avx512f_fixupimm<mode>): Ditto.
1593 (avx512f_sfixupimm<mode>): Ditto.
1594 (avx512f_rndscale<mode>): Ditto.
1595 (*avx512er_exp2<mode>): Ditto.
1596 (*avx512er_rcp28<mode>): Ditto.
1597 (avx512er_rsqrt28<mode>): Ditto.
1598 (avx512f_getmant<mode>): Ditto.
1599 (avx512f_getmant<mode>): Ditto.
1600 (avx512f_rndscale<mode>): Fix formatting.
1601
a1b499a3
JL
16022013-10-15 Martin Jambor <mjambor@suse.cz>
1603
1604 * ipa-utils.h (ipa_edge_within_scc): Declare.
1605 * ipa-cp.c (edge_within_scc): Moved...
1606 * ipa-utils.c (ipa_edge_within_scc): ...here. Updated all callers.
afb4ac68 1607
2e2206fa
AI
16082013-10-15 Alexander Ivchenko <alexander.ivchenko@intel.com>
1609 Maxim Kuznetsov <maxim.kuznetsov@intel.com>
1610 Sergey Lega <sergey.s.lega@intel.com>
1611 Anna Tikhonova <anna.tikhonova@intel.com>
1612 Ilya Tocar <ilya.tocar@intel.com>
1613 Andrey Turetskiy <andrey.turetskiy@intel.com>
1614 Ilya Verbin <ilya.verbin@intel.com>
1615 Kirill Yukhin <kirill.yukhin@intel.com>
1616 Michael Zolotukhin <michael.v.zolotukhin@intel.com>
1617
1618 * config/i386/predicates.md (const_8_to_15_operand): New.
1619 (const_16_to_31_operand): Ditto.
1620 * config/i386/sse.md (V8FI): New.
1621 (V16FI): Ditto.
1622 (reduc_splus_v8df): Ditto.
1623 (reduc_splus_v16sf): Ditto.
1624 (avx512f_vextract<shuffletype>32x4_1): Ditto.
1625 (vec_extract_hi_<mode>): Ditto.
1626 (avx512f_vinsert<shuffletype>32x4_1): Ditto.
1627 (vec_set_lo_<mode>): Ditto.
1628 (vec_set_hi_<mode>): Ditto.
1629 (avx512f_shuf_<shuffletype>64x2_1): Ditto.
1630 (avx512f_shuf_<shuffletype>32x4_1): Ditto.
1631 (avx512f_pshufd_1): Ditto.
1632 (avx512f_broadcast<mode>): Ditto.
1633 (avx512f_broadcast<mode>): Ditto.
1634 (define_split): Split vec_extract_lo into move.
1635 (ssequartermode): Ditto.
1636 (ssedoublemode): Extened with wider modes.
1637 (vec_extract_lo_<mode>): Ditto.
1638
0fe65b75
AI
16392013-10-15 Alexander Ivchenko <alexander.ivchenko@intel.com>
1640 Maxim Kuznetsov <maxim.kuznetsov@intel.com>
1641 Sergey Lega <sergey.s.lega@intel.com>
1642 Anna Tikhonova <anna.tikhonova@intel.com>
1643 Ilya Tocar <ilya.tocar@intel.com>
1644 Andrey Turetskiy <andrey.turetskiy@intel.com>
1645 Ilya Verbin <ilya.verbin@intel.com>
1646 Kirill Yukhin <kirill.yukhin@intel.com>
1647 Michael Zolotukhin <michael.v.zolotukhin@intel.com>
1648
1649 * config/i386/predicates.md (register_or_constm1_operand): New.
1650 * config/i386/sse.md (unspec): Add UNSPEC_UNSIGNED_PCMP, UNSPEC_TESTM,
1651 UNSPEC_TESTNM, UNSPEC_VTERNLOG, UNSPEC_ALIGN, UNSPEC_CONFLICT,
1652 UNSPEC_MASKED_EQ, UNSPEC_MASKED_GT, UNSPEC_GATHER_PREFETCH,
1653 UNSPEC_SCATTER_PREFETCH
1654 (VI48_512): New.
1655 (avx512f_ucmp<mode>3): Ditto.
1656 (avx512f_vternlog<mode>): Ditto.
1657 (avx512f_align<mode>): Ditto.
1658 (<shift_insn><mode>3): Ditto.
1659 (avx512f_<rotate>v<mode>): Ditto.
1660 (avx512f_<rotate><mode>): Ditto.
1661 (avx512f_eq<mode>3): Ditto.
1662 (avx512f_eq<mode>3_1): Ditto.
1663 (avx512f_gt<mode>3): Ditto.
1664 (avx512f_testm<mode>3): Ditto.
1665 (avx512f_testnm<mode>3): Ditto.
1666 (avx512pf_gatherpf<mode>): Ditto.
1667 (*avx512pf_gatherpf<mode>_mask): Ditto.
1668 (*avx512pf_gatherpf<mode>): Ditto.
1669 (avx512pf_scatterpf<mode>): Ditto.
1670 (*avx512pf_scatterpf<mode>_mask): Ditto.
1671 (*avx512pf_scatterpf<mode>): Ditto.
1672 (avx512f_vec_dup_gpr<mode>): Ditto.
1673 (clz<mode>2): Ditto.
1674 (conflict<mode>): Ditto.
1675 (REDUC_SMINMAX_MODE): Extened with wider modes.
1676 (reduc_<code>_<mode>): Ditto.
1677 (vlshr<mode>3): Ditto.
1678 (vashl<mode>3): Ditto.
1679
ab931c71
AI
16802013-10-15 Alexander Ivchenko <alexander.ivchenko@intel.com>
1681 Maxim Kuznetsov <maxim.kuznetsov@intel.com>
1682 Sergey Lega <sergey.s.lega@intel.com>
1683 Anna Tikhonova <anna.tikhonova@intel.com>
1684 Ilya Tocar <ilya.tocar@intel.com>
1685 Andrey Turetskiy <andrey.turetskiy@intel.com>
1686 Ilya Verbin <ilya.verbin@intel.com>
1687 Kirill Yukhin <kirill.yukhin@intel.com>
1688 Michael Zolotukhin <michael.v.zolotukhin@intel.com>
1689
1690 * config/i386/sse.md (unspec): Added UNSPEC_VPERMI2, UNSPEC_VPERMT2,
1691 UNSPEC_SCATTER.
1692 (VI48F_512): New.
1693 (avx512fmaskmode): Ditto.
1694 (bcstscalarsuff): Ditto.
1695 (avx512f_blendm<mode>): Ditto.
1696 (cmp_imm_predicate): Ditto.
1697 (avx512f_cmp<mode>3): Ditto.
1698 (avx512f_vec_dup<mode>): Ditto.
1699 (avx512f_vec_dup_mem<mode>): Ditto.
1700 (avx512f_vpermi2var<mode>3): Ditto.
1701 (avx512f_vpermt2var<mode>3): Ditto.
1702 (vec_init<mode>): Ditto.
1703 (avx512f_gathersi<mode>): Ditto.
1704 (*avx512f_gathersi<mode>): Ditto.
1705 (*avx512f_gathersi<mode>_2): Ditto.
1706 (avx512f_gatherdi<mode>): Ditto.
1707 (*avx512f_gatherdi<mode>): Ditto.
1708 (*avx512f_gatherdi<mode>_2): Ditto.
1709 (avx512f_scattersi<mode>): Ditto.
1710 (*avx512f_scattersi<mode>): Ditto.
1711 (avx512f_scatterdi<mode>): Ditto.
1712 (*avx512f_scatterdi<mode>): Ditto.
1713 (sseintprefix): Extened with wider modes.
1714 (VEC_GATHER_IDXSI): Ditto.
1715 (VEC_GATHER_IDXDI): Ditto.
1716 (VEC_GATHER_SRCDI): Ditto.
1717
c9acb877
MGD
17182013-10-15 Matthew Gretton-Dann <matthew.gretton-dann@arm.com>
1719 Ramana Radhakrishnan <ramana.radhakrishnan@arm.com>
1720
1721 * config/arm/t-aprofile: New file.
1722 * config.gcc: Handle --with-multilib-list option.
1723
ecbda484
BS
17242013-10-15 Bernd Schmidt <bernds@codesourcery.com>
1725
6f1abb55
UB
1726 * reload1.c (reloads_unique_chain_p): Ensure that r1 is
1727 the input for r2.
ecbda484 1728
24f161fd
RB
17292013-10-15 Richard Biener <rguenther@suse.de>
1730
1731 * tree-loop-distribution.c (build_empty_rdg): Inline into
1732 single user.
1733 (rdg_flag_vertex): Inline into single user.
1734 (rdg_flag_vertex_and_dependent): Likewise.
1735 (build_rdg_partition_for_vertex): Remove processed bitmap.
1736 (rdg_build_partitions): Simplify.
1737
5de989ed
RB
17382013-10-15 Richard Biener <rguenther@suse.de>
1739
1740 * tree-ssa-forwprop.c (forward_propagate_addr_expr_1):
1741 Restructure forwarding through conversions and copies to
1742 avoid performing copy-propagation the wrong way. Adjust
1743 recursion invocations.
1744 (forward_propagate_addr_expr): Add argument stating if we
1745 are recursing from a single-use.
1746 (ssa_forward_propagate_and_combine): Adjust.
1747
47e0da37
DM
17482013-10-14 David Malcolm <dmalcolm@redhat.com>
1749
1750 * dumpfile.h (gcc::dump_manager): New class, to hold state
1751 relating to dumpfile management.
1752 (get_dump_file_name): Remove in favor of method of dump_manager.
1753 (dump_initialized_p): Likewise.
1754 (dump_start): Likewise.
1755 (dump_finish): Likewise.
1756 (dump_switch_p): Likewise.
1757 (dump_register): Likewise.
1758 (get_dump_file_info): Likewise.
1759 * context.c (gcc::context::context): Construct the dump_manager
1760 instance.
1761 * context.h (gcc::context::get_dumps): New.
1762 (gcc::context::m_dumps): New.
1763 * coverage.c (coverage_init): Port to dump_manager API.
6f1abb55 1764 * dumpfile.c (extra_dump_files): Convert to field of gcc::dump_manager.
47e0da37
DM
1765 (extra_dump_files_in_use): Likewise.
1766 (extra_dump_files_alloced): Likewise.
1767 (gcc::dump_manager::dump_manager): New.
1768 (dump_register): Convert to...
1769 (gcc::dump_manager::dump_register): ...method, replacing
1770 function-static next_dump with m_next_dump field.
1771 (get_dump_file_info): Convert to...
1772 (gcc::dump_manager::get_dump_file_info): ...method.
1773 (get_dump_file_name): Convert to...
1774 (gcc::dump_manager::get_dump_file_name): ...method.
1775 (dump_start): Convert to...
1776 (gcc::dump_manager::dump_start): ...method.
1777 (dump_finish): Convert to...
1778 (gcc::dump_manager::dump_finish): ...method.
1779 (dump_begin): Replace body with...
1780 (gcc::dump_manager::dump_begin): ...new method.
1781 (dump_phase_enabled_p): Convert to...
1782 (gcc::dump_manager::dump_phase_enabled_p): ...method.
1783 (dump_phase_enabled_p): Convert to...
1784 (gcc::dump_manager::dump_phase_enabled_p): ...method.
1785 (dump_initialized_p): Convert to...
1786 (gcc::dump_manager::dump_initialized_p): ...method.
1787 (dump_flag_name): Replace body with...
1788 (gcc::dump_manager::dump_flag_name): ...new method.
1789 (dump_enable_all): Convert to...
1790 (gcc::dump_manager::dump_enable_all): ...new method.
1791 (opt_info_enable_passes): Convert to...
1792 (gcc::dump_manager::opt_info_enable_passes): ...new method.
1793 (dump_switch_p_1): Convert to...
1794 (gcc::dump_manager::dump_switch_p_1): ...new method.
1795 (dump_switch_p): Convert to...
1796 (gcc::dump_manager::dump_switch_p): ...new method.
1797 (opt_info_switch_p): Port to dump_manager API.
1798 (enable_rtl_dump_file): Likewise.
1799 * opts-global.c (handle_common_deferred_options): Port to new
1800 dump_manager API.
1801 * passes.c (pass_manager::finish_optimization_passes): Likewise.
1802 (pass_manager::register_one_dump_file): Likewise.
1803 (pass_manager::register_pass): Likewise.
1804 (pass_init_dump_file): Likewise.
1805 (pass_fini_dump_file): Likewise.
1806 * statistics.c (statistics_early_init): Likewise.
1807
83fd5d11
RB
18082013-10-14 Richard Biener <rguenther@suse.de>
1809
1810 * gimple.c (gimple_canonical_types, canonical_type_hash_cache,
1811 iterative_hash_canonical_type, gimple_canonical_type_hash,
1812 gimple_canonical_types_compatible_p, gimple_canonical_type_eq,
1813 gimple_register_canonical_type, print_gimple_types_stats,
1814 free_gimple_type_tables): Move to lto/lto.c
1815 (gt-gimple.h): Do not include.
1816 * gimple.h (gimple_register_canonical_type,
1817 print_gimple_types_stats, free_gimple_type_tables): Remove.
1818 * Makefile.in (GTFILES): Remove gimple.c.
1819
a6dbd0f7
TS
18202013-10-14 Travis Snoozy <quandary@remstate.com>
1821
1822 PR target/58716
1823 * config/msp430/msp430.c (msp430_option_override): Correct thinko
1824 scanning for msp430x targets.
1825
b5ccb9ed
EB
18262013-10-14 Eric Botcazou <ebotcazou@adacore.com>
1827
1828 PR bootstrap/58509
1829 * config/sparc/sparc-protos.h (widen_mem_for_ldd_peep): Declare.
1830 (registers_ok_for_ldd_peep): Move around.
1831 * config/sparc/sparc.c (widen_mem_for_ldd_peep): New.
1832 * config/sparc/sparc.md (widening peepholes): Use it.
1833
fd7e5084
RB
18342013-10-14 Richard Biener <rguenther@suse.de>
1835
1836 PR middle-end/58712
1837 PR middle-end/55358
1838 * gimple.c (iterative_hash_canonical_type): Make sure to
1839 record the hash into the correct hashtable slot.
1840
3246afaa
EB
18412013-10-13 Eric Botcazou <ebotcazou@adacore.com>
1842
1843 PR rtl-optimization/58662
1844 * combine.c (try_combine): Take into account death nodes on I2 when
1845 splitting a PARALLEL of two independent SETs. Fix dump message.
1846
5d30dc5b
OE
18472013-10-12 Oleg Endo <olegendo@gcc.gnu.org>
1848
1849 PR target/51244
1850 * config/sh/sh_treg_combine.cc: New SH specific RTL pass.
1851 * config.gcc (SH extra_objs): Add sh_ifcvt.o.
1852 * config/sh/t-sh (sh_treg_combine.o): New entry.
1853 * config/sh/sh.c (sh_fixed_condition_code_regs): New function that
1854 implements the target hook TARGET_FIXED_CONDITION_CODE_REGS.
1855 (register_sh_passes): New function. Register sh_treg_combine pass.
1856 (sh_option_override): Invoke it.
1857 (sh_canonicalize_comparison): Handle op0_preserve_value.
1858 * sh.md (*cbranch_t"): Do not try to optimize missed test and branch
1859 opportunities. Canonicalize branch condition.
1860 (nott): Allow only if pseudos can be created for non-SH2A.
1861
37d1c5db
L
18622013-10-12 H.J. Lu <hongjiu.lu@intel.com>
1863
1864 PR target/58690
1865 * config/i386/i386.c (ix86_copy_addr_to_reg): New function.
1866 (ix86_expand_movmem): Replace copy_addr_to_reg with
1867 ix86_copy_addr_to_reg.
1868 (ix86_expand_setmem): Likewise.
1869
b86a70ff
AM
18702013-10-12 Alexander Monakov <amonakov@ispras.ru>
1871
1872 * config/i386/i386.c (ix86_expand_sse_compare_and_jump): Use mode
1873 provided by ix86_fp_compare_mode instead of CCFPUmode.
1874
6e1d5961
JG
18752013-10-12 James Greenhalgh <james.greenhalgh@arm.com>
1876
1877 * config/aarch64/arm_neon.h
1878 (vtbx<1,3>_<psu>8): Fix register constriants.
1879
ef3cfba2
JL
18802013-10-11 Jeff Law <law@redhat.com>
1881
1882 PR tree-optimization/58640
adcfd489
UB
1883 * tree-ssa-threadupdate.c (mark_threaded_blocks): Truncate jump
1884 threading paths that cross over two loop entry points.
ef3cfba2 1885
6e8b7d9c
BS
18862013-10-11 Bill Schmidt <wschmidt@linux.vnet.ibm.com>
1887
1888 * config/rs6000/vsx.md (*vsx_le_perm_load_v2di): Generalize to
1889 handle vector float as well.
1890 (*vsx_le_perm_load_v4si): Likewise.
1891 (*vsx_le_perm_store_v2di): Likewise.
1892 (*vsx_le_perm_store_v4si): Likewise.
1893
f200869a
BS
18942013-10-11 Bill Schmidt <wschmidt@linux.vnet.ibm.com>
1895
1896 * config/rs6000/vector.md (vec_realign_load<mode>): Generate vperm
1897 directly to circumvent subtract from splat{31} workaround.
1898 * config/rs6000/rs6000-protos.h (altivec_expand_vec_perm_le): New
1899 prototype.
1900 * config/rs6000/rs6000.c (altivec_expand_vec_perm_le): New.
1901 * config/rs6000/altivec.md (define_c_enum "unspec"): Add
1902 UNSPEC_VPERM_X and UNSPEC_VPERM_UNS_X.
1903 (altivec_vperm_<mode>): Convert to define_insn_and_split to
1904 separate big and little endian logic.
1905 (*altivec_vperm_<mode>_internal): New define_insn.
1906 (altivec_vperm_<mode>_uns): Convert to define_insn_and_split to
1907 separate big and little endian logic.
1908 (*altivec_vperm_<mode>_uns_internal): New define_insn.
1909 (vec_permv16qi): Add little endian logic.
1910
8d28e3fc
MG
19112013-10-11 Marc Glisse <marc.glisse@inria.fr>
1912
1913 * doc/extend.texi (returns_nonnull): Remove arguments.
1914
3c87b77b
AI
19152013-10-11 Alexander Ivchenko <alexander.ivchenko@intel.com>
1916 Maxim Kuznetsov <maxim.kuznetsov@intel.com>
1917 Sergey Lega <sergey.s.lega@intel.com>
1918 Anna Tikhonova <anna.tikhonova@intel.com>
1919 Ilya Tocar <ilya.tocar@intel.com>
1920 Andrey Turetskiy <andrey.turetskiy@intel.com>
1921 Ilya Verbin <ilya.verbin@intel.com>
1922 Kirill Yukhin <kirill.yukhin@intel.com>
1923 Michael Zolotukhin <michael.v.zolotukhin@intel.com>
1924
1925 * config/i386/sse.md (VI48F_256_512): New.
1926 (avx2_permvar<mode>): Change to ...
1927 (<avx2_avx512f>_permvar<mode>): This.
1928
50e60d7d
AI
19292013-10-11 Alexander Ivchenko <alexander.ivchenko@intel.com>
1930 Maxim Kuznetsov <maxim.kuznetsov@intel.com>
1931 Sergey Lega <sergey.s.lega@intel.com>
1932 Anna Tikhonova <anna.tikhonova@intel.com>
1933 Ilya Tocar <ilya.tocar@intel.com>
1934 Andrey Turetskiy <andrey.turetskiy@intel.com>
1935 Ilya Verbin <ilya.verbin@intel.com>
1936 Kirill Yukhin <kirill.yukhin@intel.com>
1937 Michael Zolotukhin <michael.v.zolotukhin@intel.com>
1938
6f1abb55 1939 * config/i386/i386.c (bdesc_args): Change corresponding pattern for
50e60d7d
AI
1940 __builtin_ia32_cvtps2dq, __builtin_ia32_cvtps2dq256.
1941 * config/i386/sse.md (VI4_AVX): New.
1942 (sf2simodelower): Ditto.
1943 (sse2_cvtps2dq): Change to ...
1944 (<sse2_avx_avx512f>_fix_notrunc<sf2simodelower><mode>): This.
1945
f62ce24f
AI
19462013-10-11 Alexander Ivchenko <alexander.ivchenko@intel.com>
1947 Maxim Kuznetsov <maxim.kuznetsov@intel.com>
1948 Sergey Lega <sergey.s.lega@intel.com>
1949 Anna Tikhonova <anna.tikhonova@intel.com>
1950 Ilya Tocar <ilya.tocar@intel.com>
1951 Andrey Turetskiy <andrey.turetskiy@intel.com>
1952 Ilya Verbin <ilya.verbin@intel.com>
1953 Kirill Yukhin <kirill.yukhin@intel.com>
1954 Michael Zolotukhin <michael.v.zolotukhin@intel.com>
1955
1956 * config/i386/sse.md (V_512): New.
1957 (VI_512): Ditto.
1958 (vcond<V_512:mode><VF_512:mode>): Ditto.
1959 (vcond<V_512:mode><VI_512:mode>): Ditto.
1960 (vcondu<V_512:mode><VI_512:mode>): Ditto.
1961
558d9f79
AI
19622013-10-11 Alexander Ivchenko <alexander.ivchenko@intel.com>
1963 Maxim Kuznetsov <maxim.kuznetsov@intel.com>
1964 Sergey Lega <sergey.s.lega@intel.com>
1965 Anna Tikhonova <anna.tikhonova@intel.com>
1966 Ilya Tocar <ilya.tocar@intel.com>
1967 Andrey Turetskiy <andrey.turetskiy@intel.com>
1968 Ilya Verbin <ilya.verbin@intel.com>
1969 Kirill Yukhin <kirill.yukhin@intel.com>
1970 Michael Zolotukhin <michael.v.zolotukhin@intel.com>
1971
1972 * config/i386/i386.c (ix86_rtx_costs): Enable fma for TARGET_AVX512F.
1973 * config/i386/sse.md (FMAMODEM): Changed modes and conditions.
1974 (FMAMODE): Ditto.
1975 (fma<mode>4): Removed condition.
1976 (fms<mode>4): Ditto.
1977 (fnma<mode>4): Ditto.
1978 (fnms<mode>4): Ditto.
1979 (fma4i_fmadd_<mode>): Ditto.
1980 (*fma_fmadd_<mode>): Ditto.
1981 (*fma_fmsub_<mode>): Ditto.
1982 (*fma_fnmadd_<mode>): Ditto.
1983 (*fma_fnmsub_<mode>): Ditto.
1984 (fmaddsub_<mode>): Allow for TARGET_AVX512F.
1985 (*fma_fmaddsub_<mode>): Ditto.
1986 (*fma_fmsubadd_<mode>): Ditto.
1987 (*fmai_fmadd_<mode>): Ditto.
1988 (*fmai_fmsub_<mode>): Ditto.
1989 (*fmai_fnmadd_<mode>): Ditto.
1990 (*fmai_fnmsub_<mode>): Ditto.
1991
e8d08206
AI
19922013-10-11 Alexander Ivchenko <alexander.ivchenko@intel.com>
1993 Maxim Kuznetsov <maxim.kuznetsov@intel.com>
1994 Sergey Lega <sergey.s.lega@intel.com>
1995 Anna Tikhonova <anna.tikhonova@intel.com>
1996 Ilya Tocar <ilya.tocar@intel.com>
1997 Andrey Turetskiy <andrey.turetskiy@intel.com>
1998 Ilya Verbin <ilya.verbin@intel.com>
1999 Kirill Yukhin <kirill.yukhin@intel.com>
2000 Michael Zolotukhin <michael.v.zolotukhin@intel.com>
2001
2002 * config/i386/sse.md (VI248_AVX2_8_AVX512F): New.
2003 (VI124_256): Changed to ...
2004 (VI124_256_48_512): This.
2005 (ssepackmode): Extended with wider modes.
2006 (<code><mode>3): Changed iterator.
2007 (*avx2_<code><mode>3): Ditto.
2008 (vec_pack_trunc_<mode>): Ditto.
2009
16821545
AI
20102013-10-11 Alexander Ivchenko <alexander.ivchenko@intel.com>
2011 Maxim Kuznetsov <maxim.kuznetsov@intel.com>
2012 Sergey Lega <sergey.s.lega@intel.com>
2013 Anna Tikhonova <anna.tikhonova@intel.com>
2014 Ilya Tocar <ilya.tocar@intel.com>
2015 Andrey Turetskiy <andrey.turetskiy@intel.com>
2016 Ilya Verbin <ilya.verbin@intel.com>
2017 Kirill Yukhin <kirill.yukhin@intel.com>
2018 Michael Zolotukhin <michael.v.zolotukhin@intel.com>
2019
2020 * config/i386/sse.md (VI124_AVX2_48_AVX512F): New.
2021 (VI8F_256_512): Ditto.
2022 (abs<mode>2): Changed iterator.
2023 (avx2_perm<mode>): Changed to ...
2024 (<avx2_avx512f>_perm<mode>): This.
2025 (avx2_perm<mode>_1): Changed to ...
2026 (<avx2_avx512f>_perm<mode>_1): This.
2027
5348cff8
AI
20282013-10-11 Alexander Ivchenko <alexander.ivchenko@intel.com>
2029 Maxim Kuznetsov <maxim.kuznetsov@intel.com>
2030 Sergey Lega <sergey.s.lega@intel.com>
2031 Anna Tikhonova <anna.tikhonova@intel.com>
2032 Ilya Tocar <ilya.tocar@intel.com>
2033 Andrey Turetskiy <andrey.turetskiy@intel.com>
2034 Ilya Verbin <ilya.verbin@intel.com>
2035 Kirill Yukhin <kirill.yukhin@intel.com>
2036 Michael Zolotukhin <michael.v.zolotukhin@intel.com>
2037
2038 * config/i386/sse.md (VI48_AVX512F): New.
2039 (VI48_AVX2): Changed to ...
2040 (VI48_AVX2_48_AVX512F): This.
2041 (avx2_ashrv<mode>): Changed to ...
2042 (<avx2_avx512f>_ashrv<mode>): This.
2043 (avx2_<shift_insn>v<mode>): Changed to ...
2044 (<avx2_avx512f>_<shift_insn>v<mode>): This.
2045
f5f41d88
AI
20462013-10-11 Alexander Ivchenko <alexander.ivchenko@intel.com>
2047 Maxim Kuznetsov <maxim.kuznetsov@intel.com>
2048 Sergey Lega <sergey.s.lega@intel.com>
2049 Anna Tikhonova <anna.tikhonova@intel.com>
2050 Ilya Tocar <ilya.tocar@intel.com>
2051 Andrey Turetskiy <andrey.turetskiy@intel.com>
2052 Ilya Verbin <ilya.verbin@intel.com>
2053 Kirill Yukhin <kirill.yukhin@intel.com>
2054 Michael Zolotukhin <michael.v.zolotukhin@intel.com>
2055
2056 * config/i386/sse.md (VI4_AVX512F): New.
2057 (VI8_AVX2_AVX512F): Ditto.
2058 (mul<mode>3): Extended with wider modes.
2059 (*<sse4_1_avx2>_mul<mode>3): Ditto.
2060 (mul<mode>3): Ditto.
2061 (vec_widen_<s>mult_odd_<mode>): Ditto.
2062
3bdf6340
AI
20632013-10-11 Alexander Ivchenko <alexander.ivchenko@intel.com>
2064 Maxim Kuznetsov <maxim.kuznetsov@intel.com>
2065 Sergey Lega <sergey.s.lega@intel.com>
2066 Anna Tikhonova <anna.tikhonova@intel.com>
2067 Ilya Tocar <ilya.tocar@intel.com>
2068 Andrey Turetskiy <andrey.turetskiy@intel.com>
2069 Ilya Verbin <ilya.verbin@intel.com>
2070 Kirill Yukhin <kirill.yukhin@intel.com>
2071 Michael Zolotukhin <michael.v.zolotukhin@intel.com>
2072
2073 * config/i386/sse.md (VI2_AVX512F): New.
2074 (VI124_AVX512F): Ditto.
2075 (sseunpackmode): Extended with wider modes.
2076 (sseunpackfltmode): Ditto.
2077 (vec_unpacks_float_hi_<mode>): Ditto.
2078 (vec_unpacks_float_lo_<mode>): Ditto.
2079 (vec_unpacku_float_hi_<mode>): Ditto.
2080 (vec_unpacku_float_lo_<mode>): Ditto.
2081 (vec_unpacks_lo_<mode>): Ditto.
2082 (vec_unpacks_hi_<mode>): Ditto.
2083 (vec_unpacku_lo_<mode>): Ditto.
2084 (vec_unpacku_hi_<mode>): Ditto.
2085
67f783cb
AI
20862013-10-11 Alexander Ivchenko <alexander.ivchenko@intel.com>
2087 Maxim Kuznetsov <maxim.kuznetsov@intel.com>
2088 Sergey Lega <sergey.s.lega@intel.com>
2089 Anna Tikhonova <anna.tikhonova@intel.com>
2090 Ilya Tocar <ilya.tocar@intel.com>
2091 Andrey Turetskiy <andrey.turetskiy@intel.com>
2092 Ilya Verbin <ilya.verbin@intel.com>
2093 Kirill Yukhin <kirill.yukhin@intel.com>
2094 Michael Zolotukhin <michael.v.zolotukhin@intel.com>
2095
2096 * config/i386/i386.md (multdiv): New.
2097 (multdiv_mnemonic): Ditto.
2098 * config/i386/sse.md (<sse>_vmmul<mode>3): Changed to...
2099 (<sse>_vm<multdiv_mnemonic><mode>3): This.
2100 (<sse>_vmdiv<mode>3): Removed.
2101
ec5e777c
AI
21022013-10-11 Alexander Ivchenko <alexander.ivchenko@intel.com>
2103 Maxim Kuznetsov <maxim.kuznetsov@intel.com>
2104 Sergey Lega <sergey.s.lega@intel.com>
2105 Anna Tikhonova <anna.tikhonova@intel.com>
2106 Ilya Tocar <ilya.tocar@intel.com>
2107 Andrey Turetskiy <andrey.turetskiy@intel.com>
2108 Ilya Verbin <ilya.verbin@intel.com>
2109 Kirill Yukhin <kirill.yukhin@intel.com>
2110 Michael Zolotukhin <michael.v.zolotukhin@intel.com>
2111
2112 * config/i386/sse.md (V): Extended with wider modes.
2113 (VF2): Ditto.
2114 (ssehalfvecmode): Ditto.
2115 (i128): Ditto.
2116 (ssepackfltmode): Ditto.
2117 (avx_vec_concat<mode>): Ditto.
2118 (V_256_512): New iterator.
2119 (VF2_512_256): Ditto.
2120 (si2dfmode): New attribute.
2121 (si2dfmodelower): Ditto.
2122 (sf2dfmode): Ditto.
2123 (concat_tg_mode): Ditto.
2124 (floatv4siv4df2): Changed to ...
2125 (float<si2dfmodelower><mode>2): This.
2126 (avx_cvtps2pd256): Changed to ...
2127 (<sse2_avx_avx512f>_cvtps2pd<avxsizesuffix>): This.
2128 (vec_pack_trunc_v4df): Changed to ...
2129 (vec_pack_trunc_<mode>): This.
2130 (avx_vpermil<mode>): Changed to ...
2131 (<sse2_avx_avx512f>_vpermil<mode>): This.
2132 (<fixsuffix>fix_truncv8dfv8si2): New.
2133 (vec_pack_sfix_trunc_v8df): Ditto.
2134 (avx512f_rndscale<mode>): Ditto.
2135 (avx512f_roundpd512): Ditto.
2136 (vec_pack_ufix_trunc_<mode>): Updated iterator.
2137
a9ccbba2
AI
21382013-10-11 Alexander Ivchenko <alexander.ivchenko@intel.com>
2139 Maxim Kuznetsov <maxim.kuznetsov@intel.com>
2140 Sergey Lega <sergey.s.lega@intel.com>
2141 Anna Tikhonova <anna.tikhonova@intel.com>
2142 Ilya Tocar <ilya.tocar@intel.com>
2143 Andrey Turetskiy <andrey.turetskiy@intel.com>
2144 Ilya Verbin <ilya.verbin@intel.com>
2145 Kirill Yukhin <kirill.yukhin@intel.com>
2146 Michael Zolotukhin <michael.v.zolotukhin@intel.com>
2147
2148 * config/i386/i386.md (any_fix): New iterator.
2149 (fixsuffix): New attribute.
2150 * config/i386/sse.md (VF1): Extened with wider modes.
2151 (VI): Ditto.
2152 (VI_AVX2): Ditto.
2153 (VI8): Ditto.
2154 (sseintvecmodelower): Ditto.
2155 (ssescalarmode): Ditto.
2156 (ssescalarnum): Ditto.
2157 (VF1_128_256): New.
2158 (ssexmmmode): Ditto.
2159 (<fixsuffix>fix_truncv16sfv16si2): Ditto.
2160 (<sse>_rcp<mode>2): Change iterator.
2161 (rsqrt<mode>2): Ditto.
2162 (<sse>_rsqrt<mode>2): Ditto.
2163 (avx2_vec_dup<mode>): Ditto.
2164 (<sse4_1>_round<ssemodesuffix>_sfix<avxsizesuffix>): Ditto.
2165 (round<mode>2_sfix): Ditto.
2166 (avx2_pbroadcast<mode>): Ditto.
2167 (*andnot<mode>3): Handle XI mode.
2168 (*<code><mode>3): Ditto.
2169 (AVXTOSSEMODE): Removed.
2170 (avx_vpermil<mode>): Changed to ...
2171 (<sse2_avx_avx512f>_vpermil<mode>): This.
2172
2b1ebb0c
AI
21732013-10-11 Alexander Ivchenko <alexander.ivchenko@intel.com>
2174 Maxim Kuznetsov <maxim.kuznetsov@intel.com>
2175 Sergey Lega <sergey.s.lega@intel.com>
2176 Anna Tikhonova <anna.tikhonova@intel.com>
2177 Ilya Tocar <ilya.tocar@intel.com>
2178 Andrey Turetskiy <andrey.turetskiy@intel.com>
2179 Ilya Verbin <ilya.verbin@intel.com>
2180 Kirill Yukhin <kirill.yukhin@intel.com>
2181 Michael Zolotukhin <michael.v.zolotukhin@intel.com>
2182
2183 * config/i386/sse.md (<sse2>_movnt<mode>): Update constraint to "v".
2184 (<sse>_comi): Ditto.
2185 (<sse>_ucomi): Ditto.
2186 (sse_cvtss2siq_2): Ditto.
2187 (sse2_cvtsd2si): Ditto.
2188 (sse2_cvtsd2siq): Ditto.
2189 (sse2_cvttsd2si): Ditto.
2190 (sse2_cvttsd2siq): Ditto.
2191 (<shift_insn><mode>3): Ditto.
2192 (sse2_cvtsi2sdq): Update constraint and prefix.
2193 (sse_cvtsi2ss): Update prefix.
2194 (sse_cvtsi2ssq): Ditto.
2195
d89124ed
JJ
21962013-10-11 Jakub Jelinek <jakub@redhat.com>
2197
2198 * tree-vrp.c (infer_nonnull_range): Use is_gimple_call,
2199 ignore internal calls.
2200
e6ad28c3
RB
22012013-10-11 Richard Biener <rguenther@suse.de>
2202
6f1abb55
UB
2203 * tree-pretty-print.c (dump_generic_node): Allow to dump both (D)
2204 and (ab) for SSA_NAMEs. Mark INTEGER_CSTs with (OVF) if
2205 TREE_OVERFLOW is set.
e6ad28c3 2206
d1227447
TS
22072013-10-11 Thomas Schwinge <thomas@codesourcery.com>
2208
1ac4b8e4
TS
2209 * tree.h (OMP_CLAUSE_CODE): Remove duplicate definition.
2210
7d05cebb
TS
2211 * gimple.c: GIMPLE statements have subcodes, not sub-codes.
2212 * gimple.h: Likewise.
2213
adcfd489 2214 * doc/generic.texi (OpenMP): OMP_CLAUSE_* are subcodes, not sub-codes.
6545f0b3 2215
1a79fb8a
TS
2216 * doc/generic.texi (Adding new DECL node types): Explain *_CHECK
2217 macros.
2218
3afd2873
TS
2219 * doc/gimple.texi (is_gimple_omp): Move into the correct section.
2220
adcfd489 2221 * acinclude.m4 (gcc_GAS_FLAGS): Add more gcc_cv_as_flags overrides.
d1227447
TS
2222 * configure: Regenerate.
2223
acf0174b
JJ
22242013-10-11 Jakub Jelinek <jakub@redhat.com>
2225
2226 * tree-pretty-print.c (dump_omp_clause): Handle OMP_CLAUSE__LOOPTEMP_
2227 and new OpenMP 4.0 clauses, handle UDR OMP_CLAUSE_REDUCTION,
2228 formatting fixes, use pp_colon instead of pp_character (..., ':'),
2229 similarly pp_right_paren.
2230 (dump_generic_node): Handle OMP_DISTRIBUTE, OMP_TEAMS,
2231 OMP_TARGET_DATA, OMP_TARGET, OMP_TARGET_UPDATE, OMP_TASKGROUP,
2232 allow OMP_FOR_INIT to be NULL, handle OMP_ATOMIC_SEQ_CST.
2233 * tree.c (omp_clause_num_ops, omp_clause_code_name): Add OpenMP 4.0
2234 clauses.
2235 (omp_declare_simd_clauses_equal,
2236 omp_remove_redundant_declare_simd_attrs): New functions.
2237 (attribute_value_equal): Use omp_declare_simd_clauses_equal.
2238 (walk_tree_1): Handle new OpenMP 4.0 clauses.
2239 * tree.h (OMP_LOOP_CHECK): Define.
2240 (OMP_FOR_BODY, OMP_FOR_CLAUSES, OMP_FOR_INIT, OMP_FOR_COND,
2241 OMP_FOR_INCR, OMP_FOR_PRE_BODY): Use it.
2242 (OMP_TASKGROUP_BODY, OMP_TEAMS_BODY, OMP_TEAMS_CLAUSES,
2243 OMP_TARGET_DATA_BODY, OMP_TARGET_DATA_CLAUSES, OMP_TARGET_BODY,
2244 OMP_TARGET_CLAUSES, OMP_TARGET_UPDATE_CLAUSES, OMP_CLAUSE_SIZE,
2245 OMP_ATOMIC_SEQ_CST, OMP_CLAUSE_DEPEND_KIND, OMP_CLAUSE_MAP_KIND,
2246 OMP_CLAUSE_MAP_ZERO_BIAS_ARRAY_SECTION, OMP_CLAUSE_PROC_BIND_KIND,
2247 OMP_CLAUSE_REDUCTION_OMP_ORIG_REF, OMP_CLAUSE_ALIGNED_ALIGNMENT,
2248 OMP_CLAUSE_NUM_TEAMS_EXPR, OMP_CLAUSE_THREAD_LIMIT_EXPR,
2249 OMP_CLAUSE_DEVICE_ID, OMP_CLAUSE_DIST_SCHEDULE_CHUNK_EXPR,
2250 OMP_CLAUSE_SIMDLEN_EXPR): Define.
2251 (OMP_CLAUSE_DECL): Change range up to OMP_CLAUSE__LOOPTEMP_.
2252 (omp_remove_redundant_declare_simd_attrs): New prototype.
2253 * gimple.def (GIMPLE_OMP_TASKGROUP, GIMPLE_OMP_TARGET,
2254 GIMPLE_OMP_TEAMS): New codes.
2255 (GIMPLE_OMP_RETURN): Use GSS_OMP_ATOMIC_STORE instead of GSS_BASE.
2256 * omp-low.c (struct omp_context): Add cancel_label and cancellable
2257 fields.
2258 (target_nesting_level): New variable.
2259 (extract_omp_for_data): Handle GF_OMP_FOR_KIND_DISTRIBUTE and
2260 OMP_CLAUSE_DIST_SCHEDULE. Don't fallback to library implementation
2261 for collapse > 1 static schedule unless ordered.
2262 (get_ws_args_for): Add par_stmt argument. Handle combined loops.
2263 (determine_parallel_type): Adjust get_ws_args_for caller.
2264 (install_var_field): Handle mask & 4 for double indirection.
2265 (scan_sharing_clauses): Ignore shared clause on teams construct.
2266 Handle OMP_CLAUSE__LOOPTEMP_ and new OpenMP 4.0 clauses.
2267 (create_omp_child_function): If inside target or declare target
adcfd489 2268 constructs, set "omp declare target" attribute on the child function.
acf0174b
JJ
2269 (find_combined_for): New function.
2270 (scan_omp_parallel): Handle combined loops.
2271 (scan_omp_target, scan_omp_teams): New functions.
2272 (check_omp_nesting_restrictions): Check new OpenMP 4.0 nesting
2273 restrictions and set ctx->cancellable for cancellable constructs.
2274 (scan_omp_1_stmt): Call check_omp_nesting_restrictions also on
2275 selected builtin calls. Handle GIMPLE_OMP_TASKGROUP,
2276 GIMPLE_OMP_TARGET, GIMPLE_OMP_TEAMS.
adcfd489 2277 (build_omp_barrier): Add lhs argument, return gimple rather than tree.
acf0174b 2278 (omp_clause_aligned_alignment): New function.
6f1abb55 2279 (lower_rec_simd_input_clauses): Only call SET_DECL_VALUE_EXPR on decls.
acf0174b
JJ
2280 (lower_rec_input_clauses): Add FD argument. Ignore shared clauses
2281 on teams constructs. Handle user defined reductions and new
2282 OpenMP 4.0 clauses.
2283 (lower_reduction_clauses): Don't set placeholder to address of ref
2284 if it has already the right type.
2285 (lower_send_clauses): Handle OMP_CLAUSE__LOOPTEMP_.
2286 (expand_parallel_call): Use the new non-_start suffixed builtins,
2287 handle OMP_CLAUSE_PROC_BIND, don't call the outlined function
2288 and GOMP_parallel_end after the call.
2289 (expand_task_call): Handle OMP_CLAUSE_DEPEND.
2290 (expand_omp_for_init_counts): Handle combined loops.
2291 (expand_omp_for_init_vars): Add inner_stmt argument, handle combined
2292 loops.
2293 (expand_omp_for_generic): Likewise. Use GOMP_loop_end_cancel at the
2294 end of cancellable loops.
2295 (expand_omp_for_static_nochunk, expand_omp_for_static_chunk):
2296 Likewise. Handle collapse > 1 loops.
2297 (expand_omp_simd): Handle combined loops.
2298 (expand_omp_for): Add inner_stmt argument, adjust callers of
2299 expand_omp_for* functions, use expand_omp_for_static*chunk even
2300 for collapse > 1 unless ordered.
2301 (expand_omp_sections): Use GOMP_sections_end_cancel at the end
2302 of cancellable sections.
2303 (expand_omp_single): Remove need_barrier variable, just rely on
2304 gimple_omp_return_nowait_p. Adjust build_omp_barrier caller.
2305 (expand_omp_synch): Allow GIMPLE_OMP_TASKGROUP and GIMPLE_OMP_TEAMS.
2306 (expand_omp_atomic_load, expand_omp_atomic_store,
2307 expand_omp_atomic_fetch_op): Handle gimple_omp_atomic_seq_cst_p.
2308 (expand_omp_target): New function.
2309 (expand_omp): Handle combined loops. Handle GIMPLE_OMP_TASKGROUP,
2310 GIMPLE_OMP_TEAMS, GIMPLE_OMP_TARGET.
2311 (build_omp_regions_1): Immediately close region for
2312 GF_OMP_TARGET_KIND_UPDATE.
2313 (maybe_add_implicit_barrier_cancel): New function.
2314 (lower_omp_sections): Adjust lower_rec_input_clauses caller. Handle
2315 cancellation.
2316 (lower_omp_single): Likewise. Add clobber after the barrier.
2317 (lower_omp_taskgroup): New function.
2318 (lower_omp_for): Handle combined loops. Adjust
2319 lower_rec_input_clauses caller. Handle cancellation.
2320 (lower_depend_clauses): New function.
2321 (lower_omp_taskreg): Lower depend clauses. Adjust
2322 lower_rec_input_clauses caller. Add clobber after the call. Handle
2323 cancellation.
2324 (lower_omp_target, lower_omp_teams): New functions.
2325 (lower_omp_1): Handle cancellation. Handle GIMPLE_OMP_TASKGROUP,
2326 GIMPLE_OMP_TARGET, GIMPLE_OMP_TEAMS and GOMP_barrier, GOMP_cancel
2327 and GOMP_cancellation_point calls.
2328 (lower_omp): Fold stmts inside of target region.
2329 (diagnose_sb_1, diagnose_sb_2): Handle GIMPLE_OMP_TASKGROUP,
2330 GIMPLE_OMP_TARGET and GIMPLE_OMP_TEAMS.
2331 * builtin-types.def (DEF_FUNCTION_TYPE_8): Document.
2332 (BT_FN_VOID_OMPFN_PTR_UINT,
2333 BT_FN_VOID_OMPFN_PTR_UINT_LONG_LONG_LONG,
2334 BT_FN_VOID_OMPFN_PTR_UINT_LONG_LONG_LONG_LONG,
2335 BT_FN_VOID_OMPFN_PTR_OMPCPYFN_LONG_LONG_BOOL_UINT): Remove.
2336 (BT_FN_VOID_OMPFN_PTR_UINT_UINT_UINT,
2337 BT_FN_VOID_OMPFN_PTR_UINT_LONG_LONG_LONG_UINT,
2338 BT_FN_VOID_OMPFN_PTR_UINT_LONG_LONG_LONG_LONG_UINT,
2339 BT_FN_BOOL_INT, BT_FN_BOOL_INT_BOOL, BT_FN_VOID_UINT_UINT,
2340 BT_FN_VOID_INT_PTR_SIZE_PTR_PTR_PTR,
2341 BT_FN_VOID_INT_OMPFN_PTR_SIZE_PTR_PTR_PTR,
2342 BT_FN_VOID_OMPFN_PTR_OMPCPYFN_LONG_LONG_BOOL_UINT_PTR): New.
2343 * tree-ssa-alias.c (ref_maybe_used_by_call_p_1,
2344 call_may_clobber_ref_p_1): Handle BUILT_IN_GOMP_BARRIER_CANCEL,
2345 BUILT_IN_GOMP_TASKGROUP_END, BUILT_IN_GOMP_LOOP_END_CANCEL,
2346 BUILT_IN_GOMP_SECTIONS_END_CANCEL. Don't handle
2347 BUILT_IN_GOMP_PARALLEL_END.
2348 * gimple-low.c (lower_stmt): Handle GIMPLE_OMP_TASKGROUP,
2349 GIMPLE_OMP_TARGET and GIMPLE_OMP_TEAMS.
2350 * gimple-pretty-print.c (dump_gimple_omp_for): Handle
2351 GF_OMP_FOR_KIND_DISTRIBUTE.
2352 (dump_gimple_omp_target, dump_gimple_omp_teams): New functions.
2353 (dump_gimple_omp_block): Handle GIMPLE_OMP_TASKGROUP.
2354 (dump_gimple_omp_return): Print lhs if it has any.
2355 (dump_gimple_omp_atomic_load, dump_gimple_omp_atomic_store): Handle
2356 gimple_omp_atomic_seq_cst_p.
2357 (pp_gimple_stmt_1): Handle GIMPLE_OMP_TASKGROUP, GIMPLE_OMP_TARGET
2358 and GIMPLE_OMP_TEAMS.
2359 * langhooks.c (lhd_omp_mappable_type): New function.
2360 * tree-vectorizer.c (struct simd_array_to_simduid): Fix up comment.
2361 * langhooks.h (struct lang_hooks_for_types): Add omp_mappable_type
2362 hook.
2363 * gimplify.c (enum gimplify_omp_var_data): Add GOVD_MAP,
2364 GOVD_ALIGNED and GOVD_MAP_TO_ONLY.
6f1abb55 2365 (enum omp_region_type): Add ORT_TEAMS, ORT_TARGET_DATA and ORT_TARGET.
acf0174b
JJ
2366 (struct gimplify_omp_ctx): Add combined_loop field.
2367 (gimplify_call_expr, gimplify_modify_expr): Don't call fold_stmt
2368 on stmts inside of target region.
2369 (is_gimple_stmt): Return true for OMP_DISTRIBUTE and OMP_TASKGROUP.
2370 (omp_firstprivatize_variable): Handle GOVD_MAP, GOVD_ALIGNED,
2371 ORT_TARGET and ORT_TARGET_DATA.
2372 (omp_add_variable): Avoid checks on readding var for GOVD_ALIGNED.
2373 Handle GOVD_MAP.
2374 (omp_notice_threadprivate_variable): Complain about threadprivate
2375 variables in target region.
2376 (omp_notice_variable): Complain about vars with non-mappable type
2377 in target region. Handle ORT_TEAMS, ORT_TARGET and ORT_TARGET_DATA.
2378 (omp_check_private): Ignore ORT_TARGET* regions.
2379 (gimplify_scan_omp_clauses, gimplify_adjust_omp_clauses_1,
2380 gimplify_adjust_omp_clauses): Handle new OpenMP 4.0 clauses.
2381 (find_combined_omp_for): New function.
2382 (gimplify_omp_for): Handle gimplification of combined loops.
2383 (gimplify_omp_workshare): Gimplify also OMP_TARGET, OMP_TARGET_DATA,
2384 OMP_TEAMS.
2385 (gimplify_omp_target_update): New function.
2386 (gimplify_omp_atomic): Handle OMP_ATOMIC_SEQ_CST.
2387 (gimplify_expr): Handle OMP_DISTRIBUTE, OMP_TARGET, OMP_TARGET_DATA,
2388 OMP_TARGET_UPDATE, OMP_TEAMS, OMP_TASKGROUP.
2389 (gimplify_body): If fndecl has "omp declare target" attribute, add
2390 implicit ORT_TARGET context around it.
2391 * tree.def (OMP_DISTRIBUTE, OMP_TEAMS, OMP_TARGET_DATA, OMP_TARGET,
2392 OMP_TASKGROUP, OMP_TARGET_UPDATE): New tree codes.
2393 * tree-nested.c (convert_nonlocal_reference_stmt,
2394 convert_local_reference_stmt, convert_gimple_call): Handle
2395 GIMPLE_OMP_TARGET, GIMPLE_OMP_TEAMS and GIMPLE_OMP_TASKGROUP.
2396 * omp-builtins.def (BUILT_IN_GOMP_TASK): Use
2397 BT_FN_VOID_OMPFN_PTR_OMPCPYFN_LONG_LONG_BOOL_UINT_PTR
2398 instead of BT_FN_VOID_OMPFN_PTR_OMPCPYFN_LONG_LONG_BOOL_UINT.
2399 (BUILT_IN_GOMP_TARGET, BUILT_IN_GOMP_TARGET_DATA,
2400 BUILT_IN_GOMP_TARGET_END_DATA, BUILT_IN_GOMP_TARGET_UPDATE,
2401 BUILT_IN_GOMP_TEAMS, BUILT_IN_BARRIER_CANCEL,
2402 BUILT_IN_GOMP_LOOP_END_CANCEL,
2403 BUILT_IN_GOMP_SECTIONS_END_CANCEL, BUILT_IN_OMP_GET_TEAM_NUM,
2404 BUILT_IN_OMP_GET_NUM_TEAMS, BUILT_IN_GOMP_TASKGROUP_START,
2405 BUILT_IN_GOMP_TASKGROUP_END, BUILT_IN_GOMP_PARALLEL_LOOP_STATIC,
2406 BUILT_IN_GOMP_PARALLEL_LOOP_DYNAMIC,
2407 BUILT_IN_GOMP_PARALLEL_LOOP_GUIDED,
2408 BUILT_IN_GOMP_PARALLEL_LOOP_RUNTIME, BUILT_IN_GOMP_PARALLEL,
2409 BUILT_IN_GOMP_PARALLEL_SECTIONS, BUILT_IN_GOMP_CANCEL,
2410 BUILT_IN_GOMP_CANCELLATION_POINT): New built-ins.
2411 (BUILT_IN_GOMP_PARALLEL_LOOP_STATIC_START,
2412 BUILT_IN_GOMP_PARALLEL_LOOP_DYNAMIC_START,
2413 BUILT_IN_GOMP_PARALLEL_LOOP_GUIDED_START,
2414 BUILT_IN_GOMP_PARALLEL_LOOP_RUNTIME_START,
2415 BUILT_IN_GOMP_PARALLEL_START, BUILT_IN_GOMP_PARALLEL_END,
2416 BUILT_IN_GOMP_PARALLEL_SECTIONS_START): Remove.
2417 * tree-inline.c (remap_gimple_stmt, estimate_num_insns):
2418 Handle GIMPLE_OMP_TARGET, GIMPLE_OMP_TEAMS and GIMPLE_OMP_TASKGROUP.
2419 * gimple.c (gimple_build_omp_taskgroup, gimple_build_omp_target,
2420 gimple_build_omp_teams): New functions.
2421 (walk_gimple_op): Handle GIMPLE_OMP_TARGET, GIMPLE_OMP_TEAMS and
2422 GIMPLE_OMP_TASKGROUP. Walk optional lhs on GIMPLE_OMP_RETURN.
2423 (walk_gimple_stmt, gimple_copy): Handle GIMPLE_OMP_TARGET,
2424 GIMPLE_OMP_TEAMS and GIMPLE_OMP_TASKGROUP.
2425 * gimple.h (enum gf_mask): GF_OMP_FOR_KIND_DISTRIBUTE,
2426 GF_OMP_FOR_COMBINED, GF_OMP_FOR_COMBINED_INTO,
2427 GF_OMP_TARGET_KIND_MASK, GF_OMP_TARGET_KIND_REGION,
2428 GF_OMP_TARGET_KIND_DATA, GF_OMP_TARGET_KIND_UPDATE,
2429 GF_OMP_ATOMIC_SEQ_CST): New.
2430 (gimple_build_omp_taskgroup, gimple_build_omp_target,
2431 gimple_build_omp_teams): New prototypes.
2432 (gimple_has_substatements): Handle GIMPLE_OMP_TARGET,
2433 GIMPLE_OMP_TEAMS and GIMPLE_OMP_TASKGROUP.
2434 (gimple_omp_subcode): Use GIMPLE_OMP_TEAMS instead of
2435 GIMPLE_OMP_SINGLE as end of range.
2436 (gimple_omp_return_set_lhs, gimple_omp_return_lhs,
2437 gimple_omp_return_lhs_ptr, gimple_omp_atomic_seq_cst_p,
2438 gimple_omp_atomic_set_seq_cst, gimple_omp_for_combined_p,
2439 gimple_omp_for_set_combined_p, gimple_omp_for_combined_into_p,
2440 gimple_omp_for_set_combined_into_p, gimple_omp_target_clauses,
2441 gimple_omp_target_clauses_ptr, gimple_omp_target_set_clauses,
2442 gimple_omp_target_kind, gimple_omp_target_set_kind,
2443 gimple_omp_target_child_fn, gimple_omp_target_child_fn_ptr,
2444 gimple_omp_target_set_child_fn, gimple_omp_target_data_arg,
2445 gimple_omp_target_data_arg_ptr, gimple_omp_target_set_data_arg,
2446 gimple_omp_teams_clauses, gimple_omp_teams_clauses_ptr,
2447 gimple_omp_teams_set_clauses): New inlines.
2448 (CASE_GIMPLE_OMP): Add GIMPLE_OMP_TARGET, GIMPLE_OMP_TEAMS
2449 and GIMPLE_OMP_TASKGROUP.
adcfd489 2450 * tree-core.h (enum omp_clause_code): Add new OpenMP 4.0 clause codes.
acf0174b
JJ
2451 (enum omp_clause_depend_kind, enum omp_clause_map_kind,
2452 enum omp_clause_proc_bind_kind): New.
2453 (union omp_clause_subcode): Add depend_kind, map_kind and
2454 proc_bind_kind fields.
2455 * tree-cfg.c (make_edges): Handle GIMPLE_OMP_TARGET,
2456 GIMPLE_OMP_TEAMS and GIMPLE_OMP_TASKGROUP.
2457 * langhooks-def.h (lhd_omp_mappable_type): New prototype.
2458 (LANG_HOOKS_OMP_MAPPABLE_TYPE): Define.
2459 (LANG_HOOKS_FOR_TYPES_INITIALIZER): Add it.
2460
67fa7880
TJ
24612013-10-10 Teresa Johnson <tejohnson@google.com>
2462
2463 * predict.c (tree_estimate_probability): Add new parameter
2464 for estimate_bb_frequencies.
2465 (estimate_bb_frequencies): Add new parameter to force estimation.
2466 (rebuild_frequencies): When max frequency in function is small,
2467 recompute counts from frequencies.
2468 * predict.h (estimate_bb_frequencies): New parameter.
2469
b591a8b7
DM
24702013-10-10 David Malcolm <dmalcolm@redhat.com>
2471
2472 * ipa-inline.c (ipa_inline): Fix leak of "order" when
2473 optimizations are disabled.
2474
782f0db2
DM
24752013-10-10 David Malcolm <dmalcolm@redhat.com>
2476
2477 * coverage.c (coverage_finish): Fix leak of da_file_name.
2478
13ef00fa
JH
24792013-10-10 Jan Hubicka <jh@suse.cz>
2480
2481 * config/i386/x86-tune.def: Enable X86_TUNE_SSE_TYPELESS_STORES
2482 for generic, enable X86_TUNE_SSE_LOAD0_BY_PXOR for Bulldozer,
2483 Bobcat and generic.
2484
a1d50386
JJ
24852013-10-10 Jakub Jelinek <jakub@redhat.com>
2486
2487 PR middle-end/58670
2488 * stmt.c (expand_asm_operands): Add FALLTHRU_BB argument,
2489 if any labels are in FALLTHRU_BB, use a special label emitted
2490 immediately after the asm goto insn rather than label_rtx
2491 of the LABEL_DECL.
2492 (expand_asm_stmt): Adjust caller.
2493 * cfgrtl.c (commit_one_edge_insertion): Force splitting of
2494 edge if the last insn in predecessor is a jump with single successor,
2495 but it isn't simplejump_p.
2496
e2c2fde2
RB
24972013-10-10 Richard Biener <rguenther@suse.de>
2498
2499 PR tree-optimization/58656
6f1abb55 2500 * tree-ssa-pre.c (phi_translate): Do not cache failed translations.
e2c2fde2 2501
de6a5608
AM
25022013-10-10 Andrew MacLeod <amacleod@redhat.com>
2503
adcfd489
UB
2504 * gimplify.c: Include expr.h and tm_p.h for targets with special
2505 va-arg padding requirements.
de6a5608 2506
b184c8f1
AM
25072013-10-10 Andrew MacLeod <amacleod@redhat.com>
2508
2509 * tree-flow.h: Move some prototypes to gimple.h.
2510 (gimple_fold_indirect_ref): Move prototype to gimple-fold.h.
2511 * gimple.h: Relocate some prototypes from tree-flow.h
2512 * builtins.c (std_gimplify_va_arg_expr, build_va_arg_indirect_ref):
2513 Move to gimplify.c.
2514 * gimplify.c (gimple_fold_indirect_ref): Move to gimple-fold.c.
2515 (build_va_arg_indirect_ref): Relocate and make static.
2516 (std_gimplify_va_arg_expr): Relocate here.
2517 * gimple-fold.c (gimple_fold_indirect_ref): Relocate here.
2518 * gimple-fold.h (gimple_fold_indirect_ref): Add prototype.
2519
13b72c22
AK
25202013-10-10 Andreas Krebbel <Andreas.Krebbel@de.ibm.com>
2521
2522 * doc/md.texi: Document the mnemonic attribute.
2523
5db40447
AK
25242013-10-10 Andreas Krebbel <Andreas.Krebbel@de.ibm.com>
2525
2526 PR target/57377
2527 * gensupport.c (gen_mnemonic_attr): Handle (set (attr x) y) and
2528 (set_attr_alternative x ...) when searching for user defined
2529 mnemonic attribute.
2530
ed018d4b
AM
25312013-10-10 Andrew MacLeod <amacleod@redhat.com>
2532
2533 * config/aplha/alpha.c: Add gimple-ssa.h to include list.
2534
4d5b5e9f
ER
25352013-10-09 Easwaran Raman <eraman@google.com>
2536
2537 * params.def (PARAM_MIN_SIZE_FOR_STACK_SHARING): New param...
2538 * cfgexpand.c (defer_stack_allocation): ...use here
2539 * doc/invoke.texi: Add documentation for min-size-for-stack-sharing.
2540
c9ef86a1
ZC
25412013-10-09 Zhenqiang Chen <zhenqiang.chen@arm.com>
2542
2543 * tree-ssa-phiopts.c (rhs_is_fed_for_value_replacement): New function.
2544 (operand_equal_for_value_replacement): New function, extracted from
2545 value_replacement and enhanced to catch more cases.
2546 (value_replacement): Use operand_equal_for_value_replacement.
2547
2431114f
AM
25482013-10-09 Andrew MacLeod <amacleod@redhat.com>
2549
adcfd489 2550 * loop-doloop.c (doloop_modify, doloop_optimize): Use
2431114f
AM
2551 get_max_loop_iterations.
2552
1dd2a9a0
KT
25532013-10-09 Kyrylo Tkachov <kyrylo.tkachov@arm.com>
2554
2555 * config/arm/aarch-common.c (arm_early_load_addr_dep):
2556 Place comment above function.
2557
c1bf2a39
AM
25582013-10-09 Andrew MacLeod <amacleod@redhat.com>
2559
2560 * tree-flow.h: Remove all remaining prototypes, enums and structs that
2561 are not related to tree-cfg.c.
2562 * tree-ssa-address.h: New file. Relocate prototypes.
2563 * tree-ssa-address.c: (struct mem_address): Relocate from tree-flow.h.
2564 (addr_for_mem_ref): New. Combine call to get_address_description and
2565 return addr_for_mem_ref.
2566 * expr.c (expand_expr_real_1): Use new addr_for_mem_ref routine.
2567 * tree-ssa-live.h: Adjust prototypes.
2568 * passes.c: Include tree-ssa-live.h.
2569 * gimple-pretty-print.h (gimple_dump_bb): Add prototype.
2570 * graphite.c (graphite_transform_loops): Make static.
2571 (graphite_transforms, gate_graphite_transforms, pass_data_graphite,
adcfd489 2572 make_pass_graphite, pass_data_graphite_transforms,
c1bf2a39
AM
2573 make_pass_graphite_transforms): Relocate here from tree-ssa-loop.c.
2574 * ipa-pure-const.c (warn_function_noreturn): Make static.
2575 (execute_warn_function_noreturn, gate_warn_function_noreturn,
2576 class pass_warn_function_noreturn, make_pass_warn_function_noreturn):
2577 Relocate from tree-cfg.c
2578 * tree-cfg.c (tree_node_can_be_shared, gimple_empty_block_p): Make
2579 static.
2580 (execute_warn_function_noreturn, gate_warn_function_noreturn,
2581 class pass_warn_function_noreturn, make_pass_warn_function_noreturn):
2582 Move to ipa-pure-const.c.
adcfd489
UB
2583 (execute_fixup_cfg, class pass_fixup_cfg, make_pass_fixup_cfg):
2584 Relocate from tree-optimize.c.
c1bf2a39
AM
2585 * tree-optimize.c (execute_fixup_cfg, class pass_fixup_cfg,
2586 make_pass_fixup_cfg): Move to tree-cfg.c.
2587 * tree-chrec.h: (enum ev_direction): Relocate here from tree-flow.h.
2588 Relocate some prototypes.
2589 * tree-data-ref.h (tree_check_data_deps) Add prototype.
2590 * tree-dump.c (dump_function_to_file): Remove prototype.
2591 Add tree-flow.h to the include file.
2592 * tree-dump.h: Remove prototype.
2593 * tree-parloops.h: New File. Add prototypes.
2594 * tree-parloops.c (gate_tree_parallelize_loops, tree_parallelize_loops,
2595 pass_data_parallelize_loops, make_pass_parallelize_loops): Relocate
2596 from tree-ssa-loop.c.
2597 * tree-predcom.c (run_tree_predictive_commoning,
2598 gate_tree_predictive_commoning, pass_data_predcom, make_pass_predcom):
2599 Relocate here from tree-ssa-loop.c.
adcfd489 2600 * tree-ssa-dom.c (tree_ssa_dominator_optimize) Don't call
c1bf2a39
AM
2601 ssa_name_values.release ().
2602 * tree-ssa-threadedge.h: New File. Relocate prototypes here.
2603 (ssa_name_values): Relocate from tree-flow.h.
2604 * tree-ssa.h: Include tree-ssa-threadedge.h and tree-ssa-address.h.
2605 * tree-ssa-loop.c (run_tree_predictive_commoning,
2606 gate_tree_predictive_commoning, pass_data_predcom, make_pass_predcom,
2607 graphite_transforms, gate_graphite_transforms, pass_data_graphite,
2608 make_pass_graphite, pass_data_graphite_transforms,
2609 make_pass_graphite_transforms, gate_tree_parallelize_loops,
2610 tree_parallelize_loops, pass_data_parallelize_loops,
2611 make_pass_parallelize_loops): Move to other files.
2612 * tree-vectorizer.h (lpeel_tree_duplicate_loop_to_edge_cfg): Prototype
2613 moved here.
2614 * tree.h: Remove prototypes from tree-address.c.
2615
cc524fc7
AM
26162013-10-09 Andrew MacLeod <amacleod@redhat.com>
2617
2618 * tree-flow.h (tm_restart_node, gimple_df): Move to gimple-ssa.h.
2619 (struct int_tree_map): Move to tree-hasher.h
2620 (SCALE, LABEL, PERCENT): Move to gimple.h
2621 * tree-flow-inline.h: Delete. Move functions to other files.
2622 (unmodifiable_var_p, ref_contains_array_ref): Unused, so delete.
2623 * gimple-ssa.h (tm_restart_node, gimple_df): Relocate from tree-flow.h.
2624 (gimple_in_ssa_p, gimple_vop): Relocate from tree-flow-inline.h
2625 * gimple.h (imple_stmt_max_uid, set_gimple_stmt_max_uid,
2626 inc_gimple_stmt_max_uid, get_lineno): Relocate from tree-flow-inline.h.
2627 (SCALE, LABEL, PERCENT): Relocate from tree-flow.h
2628 * tree-hasher.h: Don't include tree-flow.h.
2629 (struct int_tree_map): Relocate from tree-flow.h.
2630 * tree-sra.c (contains_view_convert_expr_p): Relocate from
2631 tree-flow-inline.h and make static.
adcfd489
UB
2632 * tree-ssa-alias.h (ranges_overlap_p): Relocate from
2633 tree-flow-inline.h.
cc524fc7
AM
2634 * tree-ssa-operands.c (gimple_ssa_operands): Relocate from
2635 tree-flow-inline.h and make static.
2636 * tree.h (is_global_var, may_be_aliased): Relocate from
2637 tree-flow-inline.h.
2638 * Makefile.in (GTFILES): Remove tree-flow.h and add gimple-ssa.h.
2639 * value-prof.c: No longer include tree-flow-inline.h.
2640 * tree-switch-conversion.c: No longer include tree-flow-inline.h.
2641
71343877
AM
26422013-10-09 Andrew MacLeod <amacleod@redhat.com>
2643
2644 * tree-flow.h: Move some protoypes. Include new tree-ssa-loop.h.
2645 (struct affine_iv, struct tree_niter_desc): Move to tree-ssa-loop.h.
2646 (enum move_pos): Move to tree-ssa-loop-im.h
2647 * cfgloop.h: Move some prototypes.
2648 (gcov_type_to_double_int): relocate from tree-ssa-loop.niter.c.
2649 * tree-flow-inline.h (loop_containing_stmt): Move to tree-ssa-loop.h.
2650 * tree-ssa-loop.h: New File. Include other tree-ssa-loop-*.h files.
2651 (struct affine_iv, struct tree_niter_desc): Relocate from tree-flow.h.
2652 (loop_containing_stmt): Relocate from tree-flow-inline.h.
2653 * tree-ssa-loop-ch.c: (do_while_loop_p): Make static.
2654 * tree-ssa-loop-im.c (for_each_index): Move to tree-ssa-loop.c.
2655 (enum move_pos): Relocate here.
2656 (lsm_tmp_name_add, gen_lsm_tmp_name, get_lsm_tmp_name): Move to
2657 tree-ssa-loop.c.
2658 (execute_sm_if_changed_flag_set): Change get_lsm_tmp_name call.
adcfd489
UB
2659 (tree_ssa_loop_im, gate_tree_ssa_loop_im, pass_data_lim,
2660 make_pass_lim): Relocate here from tree-ssa-loop.c.
2661 * tree-ssa-loop-ivcanon.c (tree_num_loop_insns): Move to
71343877
AM
2662 tree-ssa-loop.c.
2663 (loop_edge_to_cancel, unloop_loops): Make static.
2664 (tree_ssa_loop_ivcanon, gate_tree_ssa_loop_ivcanon, pass_data_iv_canon,
2665 make_pass_iv_canon): Relocate from tree-ssa-loop.c.
2666 (tree_complete_unroll, gate_tree_complete_unroll,
2667 pass_data_complete_unroll, make_pass_complete_unroll): Relocate here.
2668 (tree_complete_unroll_inner, gate_tree_complete_unroll_inner,
2669 pass_data_complete_unrolli, make_pass_complete_unrolli): Relocate here.
2670 * tree-ssa-loop-ivopts.c: Remove local prototypes.
2671 (stmt_invariant_in_loop_p): Remove unused function.
2672 * tree-ssa-loop-ivopts.h: New file. Add prototypes.
2673 * tree-ssa-loop-manip.h: New file. Add prototypes.
2674 * tree-ssa-loop-niter.c (record_niter_bound): Move to cfgloop.c.
2675 (gcov_type_to_double_int): Move to cfgloop.h.
2676 (double_int_cmp, bound_index,
2677 estimate_numbers_of_iterations_loop): Make static.
2678 (estimated_loop_iterations): Factor out get_estimated_loop_iterations.
2679 (max_loop_iterations): Factor out get_max_loop_iterations.
2680 (estimated_loop_iterations_int, max_stmt_executions_int): Move to
2681 cfgloop.c.
2682 * tree-ssa-loop-niter.h: New file. Add prototypes.
2683 * tree-ssa-loop-prefetch.c (tree_ssa_loop_prefetch,
2684 gate_tree_ssa_loop_prefetch, pass_data_loop_prefetch,
2685 make_pass_loop_prefetch): Relocate from tree-ssa-loop.c.
2686 * tree-ssa-loop-unswitch.c (tree_ssa_loop_unswitch,
2687 gate_tree_ssa_loop_unswitch, pass_data_tree_unswitch,
2688 make_pass_tree_unswitch): Relocate from tree-ssa-loop.c.
2689 * tree-ssa-loop.c (tree_ssa_loop_im, gate_tree_ssa_loop_im,
2690 pass_data_lim, make_pass_lim): Move to tree-ssa-loop-im.c.
2691 (tree_ssa_loop_unswitch, gate_tree_ssa_loop_unswitch,
2692 pass_data_tree_unswitch, make_pass_tree_unswitch): Move.
2693 (tree_ssa_loop_ivcanon, gate_tree_ssa_loop_ivcanon, pass_data_iv_canon,
2694 make_pass_iv_canon, tree_complete_unroll, gate_tree_complete_unroll,
2695 pass_data_complete_unroll, make_pass_complete_unroll,
2696 tree_complete_unroll_inner, gate_tree_complete_unroll_inner,
2697 pass_data_complete_unrolli, make_pass_complete_unrolli): Move to
2698 tree-ssa-loop-ivcanon.c.
2699 (tree_ssa_loop_prefetch, gate_tree_ssa_loop_prefetch,
2700 pass_data_loop_prefetch, make_pass_loop_prefetch): Move to
2701 tree-ssa-loop-prefetch.c.
2702 (for_each_index, lsm_tmp_name_add, gen_lsm_tmp_name): Relocate from
2703 tree-ssa-loop-im.c.
2704 (get_lsm_tmp_name): Relocate and add suffix parameter.
2705 (tree_num_loop_insns): Relocate from tree-ssa-ivcanon.c.
2706 * tree-scalar-evolution.h (simple_iv): Don't use affive_iv typedef.
2707 * cfgloop.c (record_niter_bound, estimated_loop_iterations_int,
2708 max_stmt_executions_int): Move from tree-ssa-loop-niter.c.
adcfd489 2709 (get_estimated_loop_iterations): Factor out accessor from
71343877 2710 estimated_loop_iterations in tree-ssa-loop-niter.c.
adcfd489
UB
2711 (get_max_loop_iterations): Factor out accessor from
2712 _max_loop_iterations in tree-ssa-niter.c.
71343877
AM
2713 * loop-unroll.c (decide_unroll_constant_iterations,
2714 decide_unroll_runtime_iterations, decide_peel_simple,
2715 decide_unroll_stupid): Use new get_* accessors.
2716
826cacfe
MG
27172013-10-09 Marc Glisse <marc.glisse@inria.fr>
2718
2719 PR tree-optimization/20318
2720 * doc/extend.texi (returns_nonnull): New function attribute.
2721 * fold-const.c (tree_expr_nonzero_warnv_p): Look for returns_nonnull
2722 attribute.
2723 * tree-vrp.c (gimple_stmt_nonzero_warnv_p): Likewise.
2724 (stmt_interesting_for_vrp): Accept all GIMPLE_CALL.
2725
378f8976
EB
27262013-10-09 Eric Botcazou <ebotcazou@adacore.com>
2727
2728 PR middle-end/58570
2729 * tree-ssa-alias.c (nonoverlapping_component_refs_of_decl_p): Return
2730 false if both components are bitfields.
2731
0fe04f5c
AV
27322013-10-09 Alex Velenko <Alex.Velenko@arm.com>
2733
2734 * config/aarch64/arm_neon.h (vclz_s8, vclz_s16, vclz_s32)
2735 (vclzq_s8, vclzq_s16, vclzq_s32, vclz_u8, vclz_u16, vclz_u32)
2736 (vclzq_u8, vclzq_u16, vclzq_u32): Replace ASM with C.
2737 * config/aarch64/aarch64.h
2738 (CLZ_DEFINED_VALUE_AT_ZERO): Macro fixed for clz.
2739 * config/aarch64/aarch64-simd-builtins.def
2740 (VAR1 (UNOP, clz, 0, v4si)): Replaced with iterator.
2741
bed9bae4
AV
27422013-10-09 Alex Velenko <Alex.Velenko@arm.com>
2743
2744 * config/aarch64/arm_neon.h (vadd_f64, vsub_f64): Implementation added.
2745
7f3d8b19
AV
27462013-10-09 Alex Velenko <Alex.Velenko@arm.com>
2747
2748 * config/aarch64/arm_neon.h (vdiv_f64): Added.
2749
7df625a6
AV
27502013-10-09 Alex Velenko <Alex.Velenko@arm.com>
2751
2752 * config/aarch64/arm_neon.h (vneg_f32): Asm replaced with C.
2753 (vneg_f64): New intrinsic.
2754 (vneg_s8): Asm replaced with C.
2755 (vneg_s16): Likewise.
2756 (vneg_s32): Likewise.
2757 (vneg_s64): New intrinsic.
2758 (vnegq_f32): Asm replaced with C.
2759 (vnegq_f64): Likewise.
2760 (vnegq_s8): Likewise.
2761 (vnegq_s16): Likewise.
2762 (vnegq_s32): Likewise.
2763 (vnegq_s64): Likewise.
2764
2b86fca7
RL
27652013-10-09 Renlin Li <Renlin.Li@arm.com>
2766
2767 * config/arm/arm.c (arm_output_mi_thunk): Use plus_constant.
2768
e179df83
AK
27692013-10-09 Andreas Krebbel <Andreas.Krebbel@de.ibm.com>
2770
2771 * config/s390/s390.c (s390_register_info_stdarg_fpr): Remove
2772 packed stack special handling.
2773 (s390_frame_info, s390_emit_prologue, s390_emit_epilogue): Switch
2774 back to fixed stack slots for FPRs saved due to stdarg.
2775
74129172
AK
27762013-10-09 Andreas Krebbel <Andreas.Krebbel@de.ibm.com>
2777
2778 * config/s390/s390.c (s390_frame_info): Restructure function.
2779
6455a49e
AK
27802013-10-09 Andreas Krebbel <Andreas.Krebbel@de.ibm.com>
2781
2782 * config/s390/s390.c (struct s390_frame_layout): New field
2783 gpr_save_slots.
2784 (cfun_save_arg_fprs_p, cfun_gpr_save_slot): New macros.
2785 (s390_reg_clobbered_rtx, s390_regs_ever_clobbered): Change type of
2786 regs_ever_clobbered to char*.
adcfd489
UB
2787 (s390_regs_ever_clobbered): Check crtl->saves_all_registers instead
2788 of cfun->has_nonlocal_label. Ignore frame related restore INSNs.
6455a49e
AK
2789 (s390_register_info): Enable FPR save slots. Move/Copy some
2790 functionality into ...
2791 (s390_register_info_gprtofpr, s390_register_info_stdarg_fpr)
2792 (s390_register_info_stdarg_gpr, s390_optimize_register_info): New
2793 function.
2794 (s390_frame_info): Do gpr slot allocation here now. stdarg does
2795 not imply a stack frame.
2796 (s390_init_frame_layout): Remove variable clobbered_regs.
2797 (s390_update_register_info): Remove function.
2798 (s390_hard_regno_rename_ok): Call-saved regs without a save slot
2799 cannot be used for register renaming.
2800 (s390_hard_regno_scratch_ok): New function.
2801 (TARGET_HARD_REGNO_SCRATCH_OK): Define target hook.
2802 (s390_initial_elimination_offset): Change offset calculation of
2803 the return address pointer.
adcfd489 2804 (save_gprs): Deal with only r6 being saved from the call-saved regs.
6455a49e 2805 (restore_gprs): Set frame related flag.
adcfd489 2806 (s390_save_gprs_to_fprs, s390_restore_gprs_from_fprs): New functions.
6455a49e
AK
2807 (s390_emit_prologue): Call s390_register_info instead of
2808 s390_update_frame_layout. Call s390_save_gprs_to_fprs.
2809 (s390_emit_epilogue): Call s390_restore_gprs_from_fprs.
2810 (s390_optimize_prologue): Call s390_optimize_register_info.
2811 Try to remove also FPR slot save/restore INSNs. Remove frame
2812 related flags from restore INSNs.
2813
59d96342
DD
28142013-10-08 DJ Delorie <dj@redhat.com>
2815
0dc7adad
DD
2816 * config/rl78/rl78-expand.md (movqi): use operands[] not operandN.
2817 (movhi): Likewise.
2818
59d96342
DD
2819 * config/rl78/rl78.c (rl78_print_operand_1): Change %c to %C to
2820 avoid conflict with the MI use of %c.
2821 * config/rl78/rl78-real.md: change %c to %C throughout.
2822 * config/rl78/rl78-virt.md: Likewise.
2823
a81169d8
JH
28242013-10-08 Jan Hubicka <jh@suse.cz>
2825
2826 * config/i386/i386.c (ix86_option_override_internal): Switch
2827 to SSE math for -ffast-math when target ISA supports SSE2.
2828
481d1b81
AM
28292013-10-08 Andrew MacLeod <amacleod@redhat.com>
2830
2831 * tree-flow.h: Remove some prototypes.
2832 * tree.h: Remove some protypes, add a couple.
2833 * tree.c (using_eh_for_cleanups_flag, using_eh_for_cleanups,
2834 using_eh_for_cleanups_p): Add interface routines for front ends.
2835 * tree-eh.h: New file. Add protoptyes.
2836 * tree-eh.c (using_eh_for_cleanups_p, using_eh_for_cleanups): Delete.
2837 (add_stmt_to_eh_lp_fn): Make static.
2838 (lower_try_finally): Use new using_eh_for_cleanups_p.
2839 * emit-rtl.c: Include tree-eh.h.
2840 * gimple.h: Include tree-eh.h.
2841
0e6a0e48
MG
28422013-10-08 Marc Glisse <marc.glisse@inria.fr>
2843
2844 PR tree-optimization/58480
2845 * tree-vrp.c (infer_nonnull_range): New function.
2846 (infer_value_range): Call infer_nonnull_range.
2847
ef6179d1
DC
28482013-10-08 Dehao Chen <dehao@google.com>
2849
2850 PR tree-optimization/58619
2851 * tree-inline.c (copy_phis_for_bb): Combine location data
2852 only if non-null.
2853
07d964d5
ZC
28542013-10-08 Zhenqiang Chen <zhenqiang.chen@linaro.org>
2855
2856 PR target/58423
2857 * config/arm/arm.c (arm_emit_ldrd_pop): Attach
2858 RTX_FRAME_RELATED_P on INSN.
2859
19e9b2a3
BS
28602013-10-07 Bill Schmidt <wschmidt@linux.vnet.ibm.com>
2861
2862 * config/rs6000/rs6000.c (altivec_expand_vec_perm_const_le): New.
2863 (altivec_expand_vec_perm_const): Call it.
2864
0cf68694
BS
28652013-10-07 Bill Schmidt <wschmidt@linux.vnet.ibm.com>
2866
2867 * config/rs6000/vector.md (mov<mode>): Emit permuted move
2868 sequences for LE VSX loads and stores at expand time.
2869 * config/rs6000/rs6000-protos.h (rs6000_emit_le_vsx_move): New
2870 prototype.
2871 * config/rs6000/rs6000.c (rs6000_const_vec): New.
2872 (rs6000_gen_le_vsx_permute): New.
2873 (rs6000_gen_le_vsx_load): New.
2874 (rs6000_gen_le_vsx_store): New.
2875 (rs6000_gen_le_vsx_move): New.
2876 * config/rs6000/vsx.md (*vsx_le_perm_load_v2di): New.
2877 (*vsx_le_perm_load_v4si): New.
2878 (*vsx_le_perm_load_v8hi): New.
2879 (*vsx_le_perm_load_v16qi): New.
2880 (*vsx_le_perm_store_v2di): New.
2881 (*vsx_le_perm_store_v4si): New.
2882 (*vsx_le_perm_store_v8hi): New.
2883 (*vsx_le_perm_store_v16qi): New.
2884 (*vsx_xxpermdi2_le_<mode>): New.
2885 (*vsx_xxpermdi4_le_<mode>): New.
2886 (*vsx_xxpermdi8_le_V8HI): New.
2887 (*vsx_xxpermdi16_le_V16QI): New.
2888 (*vsx_lxvd2x2_le_<mode>): New.
2889 (*vsx_lxvd2x4_le_<mode>): New.
2890 (*vsx_lxvd2x8_le_V8HI): New.
2891 (*vsx_lxvd2x16_le_V16QI): New.
2892 (*vsx_stxvd2x2_le_<mode>): New.
2893 (*vsx_stxvd2x4_le_<mode>): New.
2894 (*vsx_stxvd2x8_le_V8HI): New.
2895 (*vsx_stxvd2x16_le_V16QI): New.
2896
9520e1eb
RL
28972013-10-07 Renlin Li <Renlin.Li@arm.com>
2898
2899 * config/arm/arm-cores.def (cortex-a53): Use cortex tuning.
2900
36c0bd4f
AK
29012013-10-07 Andreas Krebbel <Andreas.Krebbel@de.ibm.com>
2902
2903 * config/s390/s390.c (s390_register_info): Make the call-saved FPR
2904 loop to work also for 31bit ABI.
2905 Save the stack pointer for frame_size > 0.
2906
ee163e72
AK
29072013-10-07 Andreas Krebbel <Andreas.Krebbel@de.ibm.com>
2908
2909 * config/s390/s390.md ("tbegin", "tbegin_nofloat", "tbegin_retry")
2910 ("tbegin_retry_nofloat", "tend", "tabort", "tx_assist"): Remove
2911 constraint letters from expanders.
2912 ("tbegin_retry", "tbegin_retry_nofloat"): Change predicate of the
2913 retry count to general_operand.
2914 ("tabort"): Give operand 0 a mode.
2915 ("tabort_1"): Add mode and constraint letter for operand 0.
2916 * doc/extend.texi: Fix protoype of __builtin_non_tx_store.
2917
45d99234
JL
29182013-10-04 Jeff Law <law@redhat.com>
2919
2920 * tree-ssa-threadedge.c: Fix some trailing whitespace problems.
2921
adcfd489
UB
2922 * tree-ssa-threadedge.c (thread_through_normal_block): Broken
2923 out of ...
45d99234
JL
2924 (thread_across_edge): Here. Call it.
2925
7c327f7b
CC
29262013-10-04 Cary Coutant <ccoutant@google.com>
2927
2928 * dwarf2out.c (dw_sra_loc_expr): Release addr_table entries when
adcfd489 2929 discarding a location list expression (or a piece of one).
7c327f7b 2930
6867e128
JH
29312013-10-03 Jan Hubicka <jh@suse.cz>
2932
65efa7e7
UB
2933 * config/i386/i386.c (ix86_issue_rate): Pentium4, Nocona has issue
2934 rate of 2. Core2, Corei7 and Haswell has issue rate of 4.
3700bce3 2935 (ix86_adjust_cost): Remove Atom case; fix core2/corei7/Haswell case.
6867e128 2936
c7f36d55
JH
29372013-10-03 Jan Hubicka <jh@suse.cz>
2938
65efa7e7 2939 * config/i386/i386.c (ix86_option_override_internal): Do not enable
c7f36d55
JH
2940 accumulate-outgoing-args when producing unwind info.
2941
4f0bee4c
WM
29422013-10-03 Wei Mi <wmi@google.com>
2943
2944 * lra-constraints.c (insert_move_for_subreg): New function
2945 extracted from simplify_operand_subreg.
2946 (simplify_operand_subreg): Add reload for paradoxical subreg.
2947
111c3f39
RX
29482013-10-03 Rong Xu <xur@google.com>
2949
65efa7e7
UB
2950 * ipa-inline-analysis.c (find_foldable_builtin_expect): Find
2951 the candidate of builtin_expect such that we should fix the
2952 size/time estimation.
2953 (estimate_function_body_sizes): Do the acutally size/time fix-up
2954 for builtin_expect.
111c3f39 2955
942df739
RX
29562013-10-03 Rong Xu <xur@google.com>
2957
65efa7e7
UB
2958 * predict.c (tree_predict_by_opcode): Get the probability
2959 for builtin_expect from param builtin_expect_probability.
2960 * params.def (BUILTIN_EXPECT_PROBABILITY): New parameter.
2961 * predict.def (PRED_BUILTIN_EXPECT_RELAXED): Fix comments.
2962 * doc/invoke.texi: Add documentation for builtin-expect-probability.
942df739 2963
2284b034
MG
29642013-10-03 Marc Glisse <marc.glisse@inria.fr>
2965
2966 PR c++/19476
2967 * common.opt (fcheck-new): Moved from c.opt. Make it 'Common'.
2968 * calls.c (alloca_call_p): Use get_callee_fndecl.
2969 * fold-const.c (tree_expr_nonzero_warnv_p): Handle operator new.
2970 * tree-vrp.c (gimple_stmt_nonzero_warnv_p, stmt_interesting_for_vrp):
2971 Likewise.
2972 (vrp_visit_stmt): Remove duplicated code.
2973
0609bdf2
MM
29742013-10-03 Michael Meissner <meissner@linux.vnet.ibm.com>
2975
2976 * config/rs6000/rs6000-builtin.def (XSRDPIM): Use floatdf2,
2977 ceildf2, btruncdf2, instead of vsx_* name.
2978
2979 * config/rs6000/vsx.md (vsx_add<mode>3): Change arithmetic
2980 iterators to only do V2DF and V4SF here. Move the DF code to
2981 rs6000.md where it is combined with SF mode. Replace <VSv> with
2982 just 'v' since only vector operations are handled with these insns
2983 after moving the DF support to rs6000.md.
2984 (vsx_sub<mode>3): Likewise.
2985 (vsx_mul<mode>3): Likewise.
2986 (vsx_div<mode>3): Likewise.
2987 (vsx_fre<mode>2): Likewise.
2988 (vsx_neg<mode>2): Likewise.
2989 (vsx_abs<mode>2): Likewise.
2990 (vsx_nabs<mode>2): Likewise.
2991 (vsx_smax<mode>3): Likewise.
2992 (vsx_smin<mode>3): Likewise.
2993 (vsx_sqrt<mode>2): Likewise.
2994 (vsx_rsqrte<mode>2): Likewise.
2995 (vsx_fms<mode>4): Likewise.
2996 (vsx_nfma<mode>4): Likewise.
2997 (vsx_copysign<mode>3): Likewise.
2998 (vsx_btrunc<mode>2): Likewise.
2999 (vsx_floor<mode>2): Likewise.
3000 (vsx_ceil<mode>2): Likewise.
3001 (vsx_smaxsf3): Delete scalar ops that were moved to rs6000.md.
3002 (vsx_sminsf3): Likewise.
3003 (vsx_fmadf4): Likewise.
3004 (vsx_fmsdf4): Likewise.
3005 (vsx_nfmadf4): Likewise.
3006 (vsx_nfmsdf4): Likewise.
3007 (vsx_cmpdf_internal1): Likewise.
3008
3009 * config/rs6000/rs6000.h (TARGET_SF_SPE): Define macros to make it
3010 simpler to select whether a target has SPE or traditional floating
3011 point support in iterators.
3012 (TARGET_DF_SPE): Likewise.
3013 (TARGET_SF_FPR): Likewise.
3014 (TARGET_DF_FPR): Likewise.
3015 (TARGET_SF_INSN): Macros to say whether floating point support
3016 exists for a given operation for expanders.
3017 (TARGET_DF_INSN): Likewise.
3018
3019 * config/rs6000/rs6000.c (Ftrad): New mode attributes to allow
3020 combining of SF/DF mode operations, using both traditional and VSX
3021 registers.
3022 (Fvsx): Likewise.
3023 (Ff): Likewise.
3024 (Fv): Likewise.
3025 (Fs): Likewise.
3026 (Ffre): Likewise.
3027 (FFRE): Likewise.
3028 (abs<mode>2): Combine SF/DF modes using traditional floating point
3029 instructions. Add support for using the upper DF registers with
3030 VSX support, and SF registers with power8-vector support. Update
3031 expanders for operations supported by both the SPE and traditional
3032 floating point units.
3033 (abs<mode>2_fpr): Likewise.
3034 (nabs<mode>2): Likewise.
3035 (nabs<mode>2_fpr): Likewise.
3036 (neg<mode>2): Likewise.
3037 (neg<mode>2_fpr): Likewise.
3038 (add<mode>3): Likewise.
3039 (add<mode>3_fpr): Likewise.
3040 (sub<mode>3): Likewise.
3041 (sub<mode>3_fpr): Likewise.
3042 (mul<mode>3): Likewise.
3043 (mul<mode>3_fpr): Likewise.
3044 (div<mode>3): Likewise.
3045 (div<mode>3_fpr): Likewise.
3046 (sqrt<mode>3): Likewise.
3047 (sqrt<mode>3_fpr): Likewise.
3048 (fre<Fs>): Likewise.
3049 (rsqrt<mode>2): Likewise.
3050 (cmp<mode>_fpr): Likewise.
3051 (smax<mode>3): Likewise.
3052 (smin<mode>3): Likewise.
3053 (smax<mode>3_vsx): Likewise.
3054 (smin<mode>3_vsx): Likewise.
3055 (negsf2): Delete SF operations that are merged with DF.
3056 (abssf2): Likewise.
3057 (addsf3): Likewise.
3058 (subsf3): Likewise.
3059 (mulsf3): Likewise.
3060 (divsf3): Likewise.
3061 (fres): Likewise.
3062 (fmasf4_fpr): Likewise.
3063 (fmssf4_fpr): Likewise.
3064 (nfmasf4_fpr): Likewise.
3065 (nfmssf4_fpr): Likewise.
3066 (sqrtsf2): Likewise.
3067 (rsqrtsf_internal1): Likewise.
3068 (smaxsf3): Likewise.
3069 (sminsf3): Likewise.
3070 (cmpsf_internal1): Likewise.
3071 (copysign<mode>3_fcpsgn): Add VSX/power8-vector support.
3072 (negdf2): Delete DF operations that are merged with SF.
3073 (absdf2): Likewise.
3074 (nabsdf2): Likewise.
3075 (adddf3): Likewise.
3076 (subdf3): Likewise.
3077 (muldf3): Likewise.
3078 (divdf3): Likewise.
3079 (fred): Likewise.
3080 (rsqrtdf_internal1): Likewise.
3081 (fmadf4_fpr): Likewise.
3082 (fmsdf4_fpr): Likewise.
3083 (nfmadf4_fpr): Likewise.
3084 (nfmsdf4_fpr): Likewise.
3085 (sqrtdf2): Likewise.
3086 (smaxdf3): Likewise.
3087 (smindf3): Likewise.
3088 (cmpdf_internal1): Likewise.
3089 (lrint<mode>di2): Use TARGET_<MODE>_FPR macro.
3090 (btrunc<mode>2): Delete separate expander, and combine with the
3091 insn and add VSX instruction support. Use TARGET_<MODE>_FPR.
3092 (btrunc<mode>2_fpr): Likewise.
3093 (ceil<mode>2): Likewise.
3094 (ceil<mode>2_fpr): Likewise.
3095 (floor<mode>2): Likewise.
3096 (floor<mode>2_fpr): Likewise.
3097 (fma<mode>4_fpr): Combine SF and DF fused multiply/add support.
3098 Add support for using the upper registers with VSX and
3099 power8-vector. Move insns to be closer to the define_expands. On
3100 VSX systems, prefer the traditional form of FMA over the VSX
3101 version, since the traditional form allows the target not to
3102 overlap with the inputs.
3103 (fms<mode>4_fpr): Likewise.
3104 (nfma<mode>4_fpr): Likewise.
3105 (nfms<mode>4_fpr): Likewise.
3106
5bea0c6c 31072013-10-03 Kyrylo Tkachov <kyrylo.tkachov@arm.com>
65efa7e7 3108 Richard Earnshaw <richard.earnshaw@arm.com>
5bea0c6c
KT
3109
3110 * config/arm/aarch-common-protos.h (struct alu_cost_table): New.
3111 (struct mult_cost_table): Likewise.
3112 (struct mem_cost_table): Likewise.
3113 (struct fp_cost_table): Likewise.
3114 (struct vector_cost_table): Likewise.
3115 (cpu_cost_table): Likewise.
3116 * config/arm/arm.opt (mold-rts-costs): New option.
3117 (mnew-generic-costs): Likewise.
3118 * config/arm/arm.c (generic_extra_costs): New table.
3119 (cortexa15_extra_costs): Likewise.
3120 (arm_slowmul_tune): Use NULL as new costs.
3121 (arm_fastmul_tune): Likewise.
3122 (arm_strongarm_tune): Likewise.
3123 (arm_xscale_tune): Likewise.
3124 (arm_9e_tune): Likewise.
3125 (arm_v6t2_tune): Likewise.
3126 (arm_cortex_a5_tune): Likewise.
3127 (arm_cortex_a9_tune): Likewise.
3128 (arm_v6m_tune): Likewise.
3129 (arm_fa726te_tune): Likewise.
3130 (arm_cortex_a15_tune): Use cortex15_extra_costs.
3131 (arm_cortex_tune): Use generict_extra_costs.
3132 (shifter_op_p): New function.
3133 (arm_unspec_cost): Likewise.
3134 (LIBCALL_COST): Define.
3135 (arm_new_rtx_costs): New function.
3136 (arm_rtx_costs): Use arm_new_rtx_costs when core-specific
3137 table is available. Use old costs otherwise unless mnew-generic-costs
3138 is specified.
3139 * config/arm/arm-protos.h (tune_params): Add insn_extra_cost field.
3140 (cpu_cost_table): Declare.
3141
24c56925
MS
31422013-10-03 Marcus Shawcroft <marcus.shawcroft@arm.com>
3143
3144 PR target/58460
3145 * config/aarch64/aarch64.md (*adds_mul_imm_<mode>)
3146 (*subs_mul_imm_<mode>)
3147 (*add_<shift>_<mode>, *add_<shift>_si_uxtw,*add_mul_imm_<mode>)
3148 (*sub_<shift>_<mode>)
3149 (*sub_<shift>_si_uxtw,*sub_mul_imm_<mode>, *sub_mul_imm_si_uxtw):
3150 Remove k constraint.
3151
2ab8f063
IB
31522013-10-03 Ian Bolton <ian.bolton@arm.com>
3153
3154 * config/aarch64/aarch64.c (aarch64_secondary_reload): Remove legacy
3155 code.
3156 * config/aarch64/aarch64.md (reload_sp_immediate): Likewise.
3157
79221839
TJ
31582013-10-02 Teresa Johnson <tejohnson@google.com>
3159
3160 * predict.c (probably_never_executed): New function.
3161 (probably_never_executed_bb_p): Invoke probably_never_executed.
3162 (probably_never_executed_edge_p): Ditto.
3163 * bb-reorder.c (find_rarely_executed_basic_blocks_and_crossing_edges):
3164 Treat profile insanities conservatively.
3165
d441afe0
JDA
31662013-10-02 John David Anglin <danglin@gcc.gnu.org>
3167
3168 * config.gcc (hppa*64*-*-linux*): Don't add pa/t-linux to tmake_file.
3169
80f466c4
VM
31702013-10-02 Vladimir Makarov <vmakarov@redhat.com>
3171
3172 * lra-constraints.c (process_alt_operand): Calculate scratch_p and
3173 use it. Use smaller increase for scratch. Don't increase reject
3174 for early clobber scratch.
3175 * lra-eliminations.c (eliminate_regs_in_insn): Remove all insns
3176 setting eliminated regs except setting fp from hfp.
3177 (lra_eliminate): Check lra_insn_recog_data on NULL.
3178
6e228b4b
MM
31792013-10-02 Michael Meissner <meissner@linux.vnet.ibm.com>
3180
3181 PR target/58587
3182 * config/rs6000/rs6000-cpus.def (ISA_2_6_MASKS_SERVER): Turn off
65efa7e7 3183 setting -mvsx-timode by default until the underlying problem is fixed.
6e228b4b
MM
3184 (RS6000_CPU, power7 defaults): Likewise.
3185
11b54a5a
UB
31862013-10-02 Uros Bizjak <ubizjak@gmail.com>
3187
3188 * config/x-linux (host-linux.o): Remove header dependencies.
3189 Use $(COMPILE) and $(POSTCOMPILE).
3190 * config/t-linux-android (linux-android.o): Ditto.
3191
bbc02b69
UB
31922013-10-02 Uros Bizjak <ubizjak@gmail.com>
3193
3194 * Makefile.in (expmed.o-warn): Remove.
3195
440917de
AM
31962013-10-02 Andrew MacLeod <amacleod@redhat.com>
3197
3198 * graphite-scop-detection.c: Include tree-ssa-propagate,h.
3199 * graphite-sese-to-poly.c: Include tree-ssa-propagate.h.
3200
99206ca9
TJ
32012013-10-02 Teresa Johnson <tejohnson@google.com>
3202
3203 * dojump.c (do_jump_1): Divide probability between
3204 both conditions of a TRUTH_ANDIF_EXPR/TRUTH_ORIF_EXPR.
3205
56e82b14
TT
32062013-10-02 Tom Tromey <tromey@redhat.com>
3207
3208 * Makefile.in (DRIVER_DEFINES): Use $(if), not $(and).
3209
3d9c733e
AM
32102013-10-02 Andrew MacLeod <amacleod@redhat.com>
3211
3212 * tree-flow.h: Remove some prototypes.
3213 * tree-ssa-dce.c (mark_virtual_operand_for_renaming,
3214 mark_virtual_phi_result_for_renaming): Move to tree-into-ssa.c.
3215 * tree-into-ssa.c (mark_virtual_operand_for_renaming,
3216 mark_virtual_phi_result_for_renaming): Relocate here.
3217 * tree-into-ssa.h: Add prototypes.
bbc02b69 3218 * tree-ssa-phiopt.c: (tree_ssa_phiopt_worker) Use
3d9c733e
AM
3219 single_pred_before_succ_order.
3220 (blocks_in_phiopt_order): Rename and move to cfganal.c.
3221 (nonfreeing_call_p) Move to gimple.c.
3222 * cfganal.c (single_pred_before_succ_order): Move and renamed from
3223 tree-ssa-phiopt.c.
3224 * basic-block.h (single_pred_before_succ_order): Add prototype.
3225 * gimple.c (nonfreeing_call_p): Relocate here.
3226 * gimple.h: Add prototype.
3227 * tree-ssa-ifcombine.c: Include tree-ssa-phiopt.h.
3228 * tree-ssa-dom.h: New file. Relocate prototypes here.
3229 * tree-ssa.h: Include tree-ssa-dom.h.
3230
78cedfb1
UB
32312013-10-02 Uros Bizjak <ubizjak@gmail.com>
3232
3233 * config/i386/x-i386 (driver-i386.o): Remove header dependencies.
3234 Use $(COMPILE) and $(POSTCOMPILE).
3235
3236 * config/alpha/x-alpha (driver-alpha.o): Ditto.
3237
744730a4
AM
32382013-10-02 Andrew MacLeod <amacleod@redhat.com>
3239
3240 * tree-flow.h: Remove some prototypes.
3241 * gimple-fold.h: Add prototypes from gimple.h and tree-flow.h.
3242 * tree-ssa-propagate.h: Relocate prototypes from tree-flow.h.
3243 * tree-ssa-copy.c (may_propagate*, propagate_value, replace_exp,
3244 propagate_tree_value*): Move from here to...
3245 * tree-ssa-propagate.c (may_propagate*, propagate_value, replace_exp,
3246 propagate_tree_value*): Relocate here.
3247 * tree-ssa-propagate.h: Relocate prototypes from tree-flow.h.
3248 * gimple.h: Include gimple-fold.h, move prototypes into gimple-fold.h.
3249 * gimple-fold.c: Remove gimple-fold.h from include list.
3250 * tree-vrp.c: Remove gimple-fold.h from include list.
3251 * tree-ssa-sccvn.c: Remove gimple-fold.h from include list.
3252 * tree-ssa-ccp.c: Remove gimple-fold.h from include list.
3253 * tree-scalar-evolution.c: Add tree-ssa-propagate.h to include list.
3254 * tree-ssa-pre.c: Add tree-ssa-propagate.h to include list.
3255 * sese.c: Add tree-ssa-propagate.h to include list.
3256
826a536d
RB
32572013-10-02 Richard Biener <rguenther@suse.de>
3258
3259 * tree-loop-distribution.c: Include tree-vectorizer.h for
3260 find_loop_location.
3261 (enum partition_kind): Remove PKIND_REDUCTION.
78cedfb1 3262 (struct partition_s): Remove has_writes member, add reduction_p member.
826a536d
RB
3263 (partition_alloc): Adjust.
3264 (partition_builtin_p): Likewise.
3265 (partition_has_writes): Remove.
3266 (partition_reduction_p): New function.
3267 (partition_merge_into): Likewise.
3268 (generate_code_for_partition): Commonize builtin partition
3269 handling tail.
3270 (rdg_cannot_recompute_vertex_p): Remove.
3271 (already_processed_vertex_p): Likewise.
3272 (rdg_flag_vertex): Do not set has_writes.
3273 (classify_partition): Adjust.
3274 (rdg_build_partitions): Do not set has_writes, treat all
3275 partitions as useful.
78cedfb1 3276 (distribute_loop): Record number of library calls generated. Adjust.
826a536d
RB
3277 (tree_loop_distribution): Report number of loops and library
3278 calls generated as opt-info.
3279
4b403ece
AM
32802013-10-02 Andrew MacLeod <amacleod@redhat.com>
3281
3282 * tree-flow.h: Include new .h files. Move prototypes.
3283 * tree-cfgcleanup.h: New file. Add prototypes from tree-flow.h.
3284 * tree-dfa.h: New File. Add prototypes from tree-flow.h.
3285 (get_addr_base_and_unit_offset_1) Move from tree-flow-inline.h.
3286 * tree-pretty-print.h: Add prototypes from tree-flow.h.
3287 * tree-into-ssa.h: New File. Add prototypes from tree-flow.h.
3288 ({debug|dump}*): Move debugging prototypes out of tree-into-ssa.c.
3289 * tree-into-ssa.c ({debug|dump}*): Move prototypes to header file.
3290 * tree.h (get_ref_base_and_extent): Move prototype out.
78cedfb1 3291 * tree-flow-inline.h (get_addr_base_and_unit_offset_1): Move to
4b403ece
AM
3292 tree-dfa.h.
3293 * gimple-low.h: New File. Add prototypes from tree-flow.h.
3294 * gimple-low.c (try_catch_may_fallthru, block_may_fallthru): Move to...
3295 * tree.c (try_catch_may_fallthru, block_may_fallthru): Here.
3296 * tree-scalar-evolution.c: Include tree.h.
3297 * sese.c: Include tree.h.
3298 * dumpfile.c: Move gimple-pretty-print.h include after tree.h.
3299 * dwarf2out.c: Include tree-dfa.h.
3300 * tree-chrec.c: Include tree.h.
3301 * tree-data-ref.c: Include tree.h.
3302
1d2151c6
YZ
33032013-10-02 Yufeng Zhang <yufeng.zhang@arm.com>
3304
3305 * gimple-ssa-strength-reduction.c (backtrace_base_for_ref):
3306 Fix whitespace.
3307
441ad147
RO
33082013-10-02 Rainer Orth <ro@CeBiTec.Uni-Bielefeld.DE>
3309
3310 * config/t-sol2 (sol2-c.o): Remove header dependencies.
3311 Use $(COMPILE) and $(POSTCOMPILE).
3312 (sol2-cxx.o): Likewise.
3313 (sol2-stubs.o): Likewise.
3314 (sol2.o): Likewise.
3315 * config/x-solaris (host-solaris.o): Likewise.
3316
3317 * config/sparc/t-sparc (sparc.o): Remove.
3318 (sparc-c.o): Remove header dependencies.
3319 Use $(COMPILE) and $(POSTCOMPILE).
3320 * config/sparc/x-sparc: Likewise.
3321
2deaf8b0
JR
33222013-10-02 Joern Rennecke <joern.rennecke@embecosm.com>
3323
3324 * config/arc/arc-opts.h: Add 2013 to Copyright years.
3325 * config/arc/arc700.md: Likewise.
3326 * config/arc/arc-modes.def: Likewise.
3327 * config/arc/arc-simd.h: Likewise.
3328 * config/arc/t-arc-uClibc: Likewise.
3329 * config/arc/t-arc-newlib: Likewise.
3330
f6fe771a
RL
33312013-10-02 Renlin Li <renlin.li@arm.com>
3332
78cedfb1
UB
3333 * config/aarch64/aarch64.c (aarch64_expand_prologue): Use
3334 plus_constant.
f6fe771a
RL
3335 (aarch64_expand_epilogue): Likewise.
3336
0916f876
YZ
33372013-10-02 Bill Schmidt <wschmidt@linux.vnet.ibm.com>
3338 Yufeng Zhang <yufeng.zhang@arm.com>
3339
3340 * gimple-ssa-strength-reduction.c (legal_cast_p_1): Forward
3341 declaration.
3342 (backtrace_base_for_ref): Call get_unwidened with 'base_in' if
3343 'base_in' represent a conversion and legal_cast_p_1 holds; set
3344 'base_in' with the returned value from get_unwidened.
3345
ebfcd719
KT
33462013-10-02 Kyrylo Tkachov <kyrylo.tkachov@arm.com>
3347
3348 * config/arm/arm.c (arm_legitimize_reload_address): Explain why
3349 plus_constant is not used.
3350
157ca3e9
WM
33512013-10-01 Wei Mi <wmi@google.com>
3352
78cedfb1 3353 * config/i386/x86-tune.def (DEF_TUNE): Remove m_CORE_ALL.
157ca3e9
WM
3354 * config/i386/i386.md: Add define_peephole2 to
3355 break partial reg stall for cvtss2sd/cvtsd2ss.
3356
cd4dd8f0
JR
33572013-10-01 Joern Rennecke <joern.rennecke@embecosm.com>
3358
3359 * config/arc/arc.c (pass_arc_ifcvt::clone):
78cedfb1 3360 Update for ctxt_ -> m_ctxt change.
cd4dd8f0 3361
1465e5cf
JL
33622013-10-01 Jeff Law <law@redhat.com>
3363
3364 * tree-ssa-threadupdate.c (struct redirection_data): Delete
3365 outgoing_edge and intermediate_edge fields. Instead store the path.
3366 (redirection_data::hash): Hash on the last edge's destination index.
3367 (redirection_data::equal): Check the entire thread path.
3368 (lookup_redirectio_data): Corresponding changes.
3369 (create_edge_and_update_destination_phis): Likewise.
3370 (thread_single_edge): Likewise.
3371
25c606cb 33722013-10-01 Joern Rennecke <joern.rennecke@embecosm.com>
0ccbc132 3373 Diego Novillo <dnovillo@google.com>
25c606cb
JR
3374
3375 * config/arc/simdext.md (UNSPEC_ARC_SIMD_VLD32WH): Delete.
3376 (UNSPEC_ARC_SIMD_VLD32WL): Likewise.
3377 (vld32wh_insn, vld32wl_insn): Delete commented-out old
3378 versions of these patterns.
3379
0ccbc132
JR
3380 * doc/extend.texi (long_call/medium_call/short_call): Typo fix.
3381 (__builtin_arc_aligned): Likewise.
3382
f55d4a20
JR
3383 * config/arc/arc.md: Expand adc_0 comment stating the intended
3384 purpose and why it isn't ready.
3385 Replace commented out call_value_via_label_mixed with a
3386 plain comment about bl_s.
3387
5719867d 3388 * config/arc/arc.c (stdio.h): Don't include directly.
5719867d
JR
3389 (arc_expand_epilogue): Remove [0]: Remove fp_restored_p.
3390 Remove if (1) condition.
3391 (arc_encode_section_info): Fix comment.
3392
6462fab0
JR
33932013-10-01 Joern Rennecke <joern.rennecke@embecosm.com>
3394
3395 * config/arc/arc.c (arc_conditional_register_usage):
3396 Use ARC_FIRST_SIMD_VR_REG / ARC_LAST_SIMD_VR_REG.
3397 Also set reg_alloc_order for DMA config regs.
3398
5d5f6720
JR
33992013-10-01 Joern Rennecke <joern.rennecke@embecosm.com>
3400 Jeremy Bennett <jeremy.bennett@embecosm.com>
3401
3402 * doc/install.texi (--with-cpu): Mention ARC.
3403 (arc-*-elf32): New paragraph.
3404 (arc-linux-uclibc): Likewise.
3405 * doc/md.texi (Machine Constraints): Add ARC part.
3406 * doc/invoke.texi: (menu): Add ARC Options.
3407 (Machine Dependent Options) <ARC Options>: Add synopsis.
3408 (node ARC Options): Add.
3409 * doc/extend.texi (long_call / short_call attribute): Add ARC.
3410 (ARC Built-in Functions): New section defining
3411 generic ARC built-in functions.
3412 (ARC SIMD Built-in Functions): New section defining SIMD specific
3413 built-in functions.
3414 (Declaring Attributes of Functions): Extended
3415 description of short_call and long_call attributes for ARC and
3416 added index entries.
3417
526b7aee
SV
34182013-10-01 Saurabh Verma <saurabh.verma@codito.com>
3419 Ramana Radhakrishnan <ramana.radhakrishnan@codito.com>
3420 Joern Rennecke <joern.rennecke@embecosm.com>
3421 Muhammad Khurram Riaz <khurram.riaz@arc.com>
3422 Brendan Kehoe <brendan@zen.org>
3423 Michael Eager <eager@eagercon.com>
3424 Simon Cook <simon.cook@embecosm.com>
3425 Jeremy Bennett <jeremy.bennett@embecosm.com>
3426
3427 * config/arc, common/config/arc: New directories.
3428
53426f6c
JR
34292013-10-01 Joern Rennecke <joern.rennecke@embecosm.com>
3430 Brendan Kehoe <brendan@zen.org>
3431 Simon Cook <simon.cook@embecosm.com>
3432
3433 * config.gcc (arc*-*-elf*, arc*-*-linux-uclibc*): New configurations.
3434
1df9f5a9
AM
34352013-10-01 Andrew MacLeod <amacleod@redhat.com>
3436
3437 * tree-ssa-live.h (coalesce_ssa_name): Move Prototype to...
3438 * tree-ssa-coalesce.h: New. Move prototype to here.
3439 * tree-outof-ssa.h: Include tree-ssa-coalesce.h.
3440 * tree-ssa-coalesce.c: Include tree-outof-ssa.h.
3441 (gimple_can_coalesce_p): Move to...
3442 * gimple.c (gimple_can_coalesce_p): Here.
3443
80560f95
AM
34442013-10-01 Andrew MacLeod <amacleod@redhat.com>
3445
3446 * tree-into-ssa.c (enum need_phi_state): Relocate from tree-flow.h.
3447 (dump_decl_set): Move to gimple.c.
3448 * gimple.h: Don't include tree-ssa-operands.h.
3449 (dump_decl_set): Add prototype.
3450 (gimple_vuse_op, gimple_vdef_op, update_stmt, update_stmt_if_modified):
3451 Move to gimple-ssa.h.
3452 (phi_ssa_name_p, phi_nodes, phi_nodes_ptr, gimple_phi_arg_def,
3453 gimple_phi_arg_def_ptr, gimple_phi_arg_edge, gimple_phi_arg_location,
3454 gimple_phi_arg_location_from_edge, gimple_phi_arg_set_location,
3455 gimple_phi_arg_has_location): Relocate from tree-flow-inline.h
3456 * gimple.c (walk_stmt_load_store_ops): Use gimple_phi_arg_def rather
3457 than PHI_ARG_DEF.
3458 (dump_decl_set): Relocate here.
3459 * gimple-ssa.h: New file.
3460 (gimple_vuse_op, gimple_vdef_op, update_stmt, update_stmt_if_modified):
3461 Relocate from gimple.h.
3462 * tree-cfg.c (has_zero_uses_1, single_imm_use_1): Move to...
3463 * tree-ssa-operands.c (swap_ssa_operands): Rename from
3464 swap_tree_operands and remove non-ssa path.
3465 (has_zero_uses_1, single_imm_use_1): Relocate from tree-cfg.c.
3466 * tree-ssa-reassoc.c (linearize_expr_tree, repropagate_negates): Use
3467 swap_ssa_operands.
3468 * tree-vect-loop.c (destroy_loop_vec_info, vect_is_slp_reduction,
3469 vect_is_simple_reduction_1): Use swap_ssa_operands.
3470 * tree-flow.h: Move various prototypes to tree-phinodes.h.
3471 (enum need_phi_state): Move to tree-into-ssa.c.
3472 (struct immediate_use_iterator_d, FOR_EACH_IMM_*,
3473 BREAK_FROM_IMM_USE_STMT): Move to ssa-iterators.h.
3474 (swap_tree_operands): Rename and move prototype to tree-ssa-operands.h.
3475 * tree-flow-inline.h (delink_imm_use, link_imm_use_to_list,
3476 link_imm_use, set_ssa_use_from_ptr, link_imm_use_stmt, relink_imm_use,
3477 relink_imm_use_stmt, end_readonly_imm_use_p, first_readonly_imm_use,
3478 next_readonly_imm_use, has_zero_uses, has_single_use, single_imm_use,
3479 num_imm_uses): Move to ssa-iterators.h.
3480 (get_use_from_ptr, get_def_from_ptr): Move to tree-ssa-operands.h
7ab4168e 3481 (gimple_phi_arg_imm_use_ptr, phi_arg_index_from_use): Move to
80560f95
AM
3482 tree-phinodes.h.
3483 (op_iter_done, op_iter_next_def, op_iter_next_tree,
3484 clear_and_done_ssa_iter, op_iter_init, op_iter_init_use,
3485 op_iter_init_def, op_iter_init_tree, single_ssa_tree_operand,
3486 single_ssa_use_operand, single_ssa_def_operand, zero_ssa_operands,
3487 num_ssa_operands, delink_stmt_imm_use, single_phi_def,
3488 op_iter_init_phiuse, op_iter_init_phidef, end_imm_use_stmt_p,
3489 end_imm_use_stmt_traverse, move_use_after_head, link_use_stmts_after,
3490 first_imm_use_stmt, next_imm_use_stmt, first_imm_use_on_stmt,
3491 end_imm_use_on_stmt_p, next_imm_use_on_stmt): Move to ssa-iterators.h.
3492 (gimple_phi_arg_def, gimple_phi_arg_def_ptr, gimple_phi_arg_edge,
3493 gimple_phi_arg_location, gimple_phi_arg_location_from_edge,
3494 gimple_phi_arg_set_location, gimple_phi_arg_has_location, phi_nodes,
3495 phi_nodes_ptr, phi_ssa_name_p): Move to gimple.h.
3496 (set_phi_nodes): Move to tree-phinodes.h.
3497 * tree-ssa-operands.h (enum ssa_op_iter_type,
3498 struct ssa_operand_iterator_d, SSA_OP*, FOR_EACH_SSA*, SINGLE_SSA*,
3499 ZERO_SSA_OPERANDS, NUM_SSA_OPERANDS): Move to ssa-iterators.h.
3500 (dump_decl_set): Remove prototype.
3501 (get_use_from_ptr, get_def_from_ptr): Relocate from tree-flow.h.
3502 * tree-phinodes.h: New file. Move some prototypes from tree-flow.h.
3503 (set_phi_nodes): Relocate from tree-flow-inline.h.
3504 (gimple_phi_arg_imm_use_ptr, phi_arg_index_from_use): Relocate from
3505 tree-flow-inline.h
3506 * tree-ssa.h: Add tree-phinodes.h, gimple-ssa.h, ssa-iterators.h to
3507 include list. Temporarily add gimple.h to include list.
3508 * ssa-iterators.h: New file.
3509 (struct immediate_use_iterator_d, FOR_EACH_IMM_*,
3510 BREAK_FROM_IMM_USE_STMT): Relocate from tree-flow.h.
3511 (enum ssa_op_iter_type, struct ssa_operand_iterator_d, SSA_OP*,
3512 FOR_EACH_SSA*, SINGLE_SSA*, ZERO_SSA_OPERANDS, NUM_SSA_OPERANDS):
3513 Relocate from tree-ssa-operands.h.
3514 (delink_imm_use, link_imm_use_to_list, link_imm_use,
3515 set_ssa_use_from_ptr, link_imm_use_stmt, relink_imm_use,
3516 relink_imm_use_stmt, end_readonly_imm_use_p, first_readonly_imm_use,
3517 next_readonly_imm_use, has_zero_uses, has_single_use, single_imm_use,
3518 num_imm_uses, get_use_from_ptr, get_def_from_ptr,
3519 phi_arg_index_from_use, op_iter_done, op_iter_next_def,
3520 op_iter_next_tree, clear_and_done_ssa_iter, op_iter_init,
3521 op_iter_init_use, op_iter_init_def, op_iter_init_tree,
3522 single_ssa_tree_operand, single_ssa_use_operand, single_ssa_def_operand,
3523 zero_ssa_operands, num_ssa_operands, delink_stmt_imm_use,
3524 single_phi_def, op_iter_init_phiuse, op_iter_init_phidef,
3525 end_imm_use_stmt_p, end_imm_use_stmt_traverse, move_use_after_head,
3526 link_use_stmts_after, first_imm_use_stmt, next_imm_use_stmt,
3527 first_imm_use_on_stmt, end_imm_use_on_stmt_p, next_imm_use_on_stmt):
3528 Relocate from tree-flow-inline.h.
3529 * tree-outof-ssa.h: Change _SSAEXPAND_H macro to GCC_TREE_OUTOF_SSA_H.
3530
8da00d65
VP
35312013-10-01 Vidya Praveen <vidyapraveen@arm.com>
3532
3533 * aarch64-simd.md
3534 (aarch64_<ANY_EXTEND:su><ADDSUB:optab>l2<mode>_internal): Rename to ...
3535 (aarch64_<ANY_EXTEND:su><ADDSUB:optab>l<mode>_hi_internal): ... this;
3536 Insert '\t' to output template.
3537 (aarch64_<ANY_EXTEND:su><ADDSUB:optab>l<mode>_lo_internal): New.
3538 (aarch64_saddl2<mode>, aarch64_uaddl2<mode>): Modify to call
3539 gen_aarch64_<ANY_EXTEND:su><ADDSUB:optab>l<mode>_hi_internal() instead.
3540 (aarch64_ssubl2<mode>, aarch64_usubl2<mode>): Ditto.
3541
e2ebe1c2
UB
35422013-10-01 Uros Bizjak <ubizjak@gmail.com>
3543
3544 * doc/install.texi (Host/target specific installation notes for GCC):
3545 Put @anchor before @heading.
3546 * doc/gcc.texi (titlepage): Use @uref and http:// prefix for website.
3547 Use @email for email addresses.
3548
aee2d611
JL
35492013-10-01 Jeff Law <law@redhat.com>
3550
3551 * tree-ssa-threadedge.c (thread_across_edge): Make path a pointer to
3552 a vec. Only delete the path if we create one without successfully
3553 registering a jump thread.
3554 * tree-ssa-threadupdate.h (register_jump_thread): Pass in path vector
3555 as a pointer.
3556 * tree-ssa-threadupdate.c (threaded_edges): Remove. No longer used
3557 (paths): New vector of jump threading paths.
3558 (THREAD_TARGET, THREAD_TARGET2): Remove accessor macros.
3559 (THREAD_PATH): New accessor macro for the entire thread path.
3560 (lookup_redirection_data): Get intermediate and final outgoing edge
3561 from the thread path.
3562 (create_edge_and_update_destination_phis): Copy the threading path.
3563 (ssa_fix_duplicate_block_edges): Get edges and block types from the
3564 jump threading path.
3565 (ssa_redirect_edges): Get edges and block types from the jump threading
3566 path. Free the path vector.
3567 (thread_block): Get edges from the jump threading path. Look at the
3568 entire path to see if we thread to a loop exit. If we cancel a jump
3569 thread request, then free the path vector.
3570 (thread_single_edge): Get edges and block types from the jump threading
3571 path. Free the path vector.
3572 (thread_through_loop_header): Get edges and block types from the jump
3573 threading path. Free the path vector.
3574 (mark_threaded_blocks): Iterate over the vector of paths and store
3575 the path on the appropriate edge. Get edges and block types from the
3576 jump threading path.
3577 (mark_threaded_blocks): Get edges and block types from the jump
3578 threading path. Free the path vector.
3579 (thread_through_all_blocks): Use the vector of paths rather than
3580 a vector of 3-edge sets.
3581 (register_jump_thread): Accept pointer to a path vector rather
3582 than the path vector itself. Store the path vector for later use.
3583 Simplify.
3584
966f97ac 35852013-10-01 Jakub Jelinek <jakub@redhat.com>
8da00d65 3586 Andreas Krebbel <Andreas.Krebbel@de.ibm.com>
966f97ac
JJ
3587
3588 PR target/58574
3589 * config/s390/s390.c (s390_split_branches): Modify check for table
3590 jump insns.
3591 (s390_chunkify_start): Rearrange table jump insn check in order to
3592 deal with compare and branch insns correctly.
3593
3a323a38
KV
35942013-10-01 Kugan Vivekanandarajah <kuganv@linaro.org>
3595
3596 PR target/58578
3597 Revert
3598 2013-04-05 Greta Yorsh <Greta.Yorsh@arm.com>
3599 * config/arm/arm.md (arm_ashldi3_1bit): define_insn into
3600 define_insn_and_split.
3601 (arm_ashrdi3_1bit,arm_lshrdi3_1bit): Likewise.
3602 (shiftsi3_compare): New pattern.
3603 (rrx): New pattern.
3604 * config/arm/unspecs.md (UNSPEC_RRX): New.
3605
90444831
AM
36062013-10-01 Alan Modra <amodra@gmail.com>
3607
3608 * stmt.c (expand_asm_operands): Revert part of 2013-09-24 special
3609 casing inout operands.
3610
818625cf
RB
36112013-10-01 Richard Biener <rguenther@suse.de>
3612
3613 PR tree-optimization/58553
3614 * tree-loop-distribution.c (struct partition_s): Add niter member.
3615 (classify_partition): Populate niter member for the partition
3616 and properly identify whether the relevant store happens before
3617 or after the loop exit.
3618 (generate_memset_builtin): Use niter member from the partition.
3619 (generate_memcpy_builtin): Likewise.
3620
30f641cd
RS
36212013-09-30 Richard Sandiford <rdsandiford@googlemail.com>
3622
3623 * vec.h (vec_prefix, vec): Prefix member names with "m_".
3624 * vec.c (vec_prefix::calculate_allocation): Update accordingly.
3625
65d3284b
RS
36262013-09-30 Richard Sandiford <rdsandiford@googlemail.com>
3627
3628 * basic-block.h (edge_list): Prefix member names with "m_".
3629 * context.h (context): Likewise.
3630 * domwalk.h (dom_walker): Likewise.
3631 * gengtype-state.c (s_expr_writer, state_writer): Likewise.
3632 * graphite-sese-to-poly.c (sese_dom_walker): Likewise.
3633 * hash-table.h (hash_table): Likewise.
3634 * machmode.h (bit_field_mode_iterator): Likewise.
3635 * pass_manager.h (pass_list): Likewise.
3636 * tree-into-ssa.c (mark_def_dom_walker): Likewise.
3637 * tree-pass.h (pass_data): Likewise.
3638 * tree-ssa-dom.c (dom_opt_dom_walker): Likewise.
3639 * tree-ssa-phiopt.c (nontrapping_dom_walker): Likewise,
3640 * tree-ssa-uncprop.c (uncprop_dom_walker): Likewise.
3641 * asan.c (pass_data_asan): Update accordingly.
3642 * cfganal.c (control_dependences::find_control_dependence): Likewise.
3643 (control_dependences::control_dependences): Likewise.
3644 (control_dependences::~control_dependences): Likewise.
3645 (control_dependences::~control_dependences): Likewise.
3646 (control_dependences::get_edges_dependent_on): Likewise.
3647 * cgraphbuild.c (pass_data_rebuild_cgraph_edges::clone): Likewise.
3648 (pass_data_remove_cgraph_callee_edges::clone): Likewise.
3649 * context.c (gcc::context::context): Likewise.
3650 * cprop.c (pass_rtl_cprop::clone): Likewise.
3651 * domwalk.c (dom_walker::walk): Likewise.
3652 * ipa-inline-analysis.c (pass_inline_parameters::clone): Likewise.
3653 * ipa-pure-const.c (pass_local_pure_const::clone): Likewise.
3654 * mode-switching.c (pass_mode_switching::clone): Likewise.
3655 * passes.c (opt_pass::opt_pass): Likewise.
3656 (pass_manager::pass_manager): Likewise.
3657 * predict.c (pass_strip_predict_hints::clone): Likewise.
3658 * recog.c (pass_data pass_data_peephole2::clone): Likewise.
3659 (pass_split_all_insns::clone): Likewise.
3660 * stor-layout.c (bit_field_mode_iterator::bit_field_mode_iterator):
3661 Likewise.
3662 (bit_field_mode_iterator::next_mode): Likewise.
3663 (bit_field_mode_iterator::prefer_smaller_modes): Likewise.
3664 * tree-cfg.c (pass_split_crit_edges::clone): Likewise.
3665 * tree-cfgcleanup.c (pass_merge_phi::clone): Likewise.
3666 * tree-complex.c (pass_lower_complex::clone): Likewise.
3667 * tree-eh.c (pass_cleanup_eh::clone): Likewise.
3668 * tree-object-size.c (pass_object_sizes::clone): Likewise.
3669 * tree-optimize.c (pass_fixup_cfg::clone): Likewise.
3670 * tree-ssa-ccp.c (pass_data_ccp::clone): Likewise.
3671 (pass_fold_builtins::clone): Likewise.
3672 * tree-ssa-copy.c (pass_data_copy_prop::clone): Likewise.
3673 * tree-ssa-copyrename.c (pass_rename_ssa_copies::clone): Likewise.
3674 * tree-ssa-dce.c (pass_dce::clone, pass_dce_loop::clone): Likewise.
3675 (pass_cd_dce::clone): Likewise.
3676 * tree-ssa-dom.c (pass_dominator::clone): Likewise.
3677 (pass_phi_only_cprop::clone): Likewise.
3678 * tree-ssa-dse.c (pass_dse::clone): Likewise.
3679 * tree-ssa-forwprop.c (pass_forwprop::clone): Likewise.
3680 * tree-ssa-loop.c (pass_lim::clone): Likewise.
3681 * tree-ssa-phiopt.c (pass_phiopt::clone): Likewise.
3682 * tree-ssa-pre.c (pass_fre::clone): Likewise.
3683 * tree-ssa-reassoc.c (pass_reassoc::clone): Likewise.
3684 * tree-ssa-uninit.c (pass_late_warn_uninitialized::clone): Likewise.
3685 * tree-tailcall.c (pass_tail_recursion::clone): Likewise.
3686 * tree-vect-generic.c (pass_lower_vector_ssa::clone): Likewise.
3687 * tree-vrp.c (pass_vrp::clone): Likewise.
3688 * tsan.c (pass_tsan::clone): Likewise.
3689
f66d0891
JJ
36902013-09-30 Jakub Jelinek <jakub@redhat.com>
3691
d2a365a8
JJ
3692 PR middle-end/58564
3693 * fold-const.c (tree_unary_nonnegative_warnv_p): Use
3694 INTEGRAL_TYPE_P (t) instead of TREE_CODE (t) == INTEGER_TYPE.
3695
f66d0891
JJ
3696 PR middle-end/58564
3697 * fold-const.c (fold_ternary_loc): For A < 0 : <sign bit of A> : 0
3698 optimization, punt if sign_bit_p looked through any zero extension.
3699
05357ac3
TJ
37002013-09-30 Teresa Johnson <tejohnson@google.com>
3701
3702 * tree-ssa-threadupdate.c (ssa_fix_duplicate_block_edges):
3703 Update redirected out edge count in joiner case.
3704 (ssa_redirect_edges): Common the joiner and non-joiner cases
3705 so that joiner case gets profile updates.
3706
ca406576
RB
37072013-09-30 Richard Biener <rguenther@suse.de>
3708
3709 PR tree-optimization/58554
e2ebe1c2
UB
3710 * tree-loop-distribution.c (classify_partition): Require
3711 unconditionally executed stores for memcpy and memset recognition.
ca406576
RB
3712 (tree_loop_distribution): Calculate dominance info.
3713
92d649c4
VK
37142013-09-30 Venkataramanan Kumar <venkataramanan.kumar@linaro.org>
3715
3716 * config/aarch64/aarch64.h (MCOUNT_NAME): Define.
3717 (NO_PROFILE_COUNTERS): Likewise.
3718 (PROFILE_HOOK): Likewise.
3719 (FUNCTION_PROFILER): Likewise.
3720 * config/aarch64/aarch64.c (aarch64_function_profiler): Remove.
3721
bd9534e2
IS
37222013-09-30 Iain Sandoe <iain@codesourcery.com>
3723
3724 * config/rs6000/darwin.md (load_macho_picbase_si): Wrap machopic
3725 calls and defines in TARGET_MACHO conditional.
3726 (load_macho_picbase_di): Likewise.
3727 (reload_macho_picbase): Likewise.
3728 (reload_macho_picbase_si): Likewise.
3729 (reload_macho_picbase_di): Likewise.
3730 (nonlocal_goto_receiver): Likewise.
3731
4494fbc9
NC
37322013-09-30 Nick Clifton <nickc@redhat.com>
3733
3734 * config/msp430/msp430.c (msp430x_names): New array. Lists MCUs
3735 that use the MSP430X ISA.
3736 (msp430_option_override): Scan -mmcu command line option for any
3737 MCU name that supports the MSP430X ISA.
3738 * config/msp430/t-msp430 (MULTILIB_MATCHES): Add matches for known
3739 -mmcu options which enable the MSP430X ISA.
3740
791d9044
RB
37412013-09-30 Richard Biener <rguenther@suse.de>
3742
3743 PR middle-end/58532
3744 * tree-cfg.c (make_abnormal_goto_edges): Skip debug statements
3745 before looking for setjmp-like calls.
3746
e1c5c877
IS
37472013-09-29 Iain Sandoe <iain@codesourcery.com>
3748
3749 PR target/10901
3750 * config/darwin-protos.h (machopic_get_function_picbase): New.
3751 * config/darwin.c (machopic_get_function_picbase): New.
3752 * config/rs6000/darwin.md (load_macho_picbase_si): Update picbase
3753 label for a new func. (load_macho_picbase_di): Likewise.
3754 (reload_macho_picbase): New expand.
3755 (reload_macho_picbase_si): New insn.
3756 (reload_macho_picbase_di): New insn.
3757 (nonlocal_goto_receiver): New define and split.
3758 * config/rs6000/rs6000.md (unspec enum): Add UNSPEC_RELD_MPIC.
3759 (unspecv enum): Add UNSPECV_NLGR.
3760
749278c5
IS
37612013-09-29 Iain Sandoe <iain@codesourcery.com>
3762
3763 * config/rs6000/rs6000.c (rs6000_init_dwarf_reg_sizes_extra): Ensure
3764 that altivec registers are correctly sized on Darwin.
3765
2c43a51c
IS
37662013-09-29 Iain Sandoe <iain@codesourcery.com>
3767
3768 * config/t-darwin (darwin.o, darwin-c.o, darwin-f.o,
3769 darwin-driver.o): Use COMPILE and POSTCOMPILE.
3770 * config/x-darwin (host-darwin.o): Likewise.
3771 * config/i386/x-darwin (host-i386-darwin.o): Likewise.
3772 * config/rs6000/x-darwin (host-ppc-darwin.o): Likewise.
3773 * config/rs6000/x-darwin64 (host-ppc64-darwin.o): Likewise.
3774
7a5911d3
UB
37752013-09-29 Uros Bizjak <ubizjak@gmail.com>
3776
3777 * doc/invoke.texi: Fix usage of @tie{} command.
3778
3f67a73d
EB
37792013-09-29 Eric Botcazou <ebotcazou@adacore.com>
3780
3781 * config/sparc/sync.md: Add peephole for consecutive memory barriers.
3782
9ac2f538
JH
37832013-09-28 Jan Hubicka <jh@suse.cz>
3784
3785 * config/i386/x86-tune.def: Add documentation for each of the options;
3786 add whitespace.
3787
b0ff06da
JH
37882013-09-28 Jan Hubicka <jh@suse.cz>
3789
7a5911d3
UB
3790 * x86-tune.def (X86_TUNE_SSE_UNALIGNED_LOAD_OPTIMAL): Enable for
3791 generic.
b0ff06da
JH
3792 (X86_TUNE_SSE_UNALIGNED_STORE_OPTIMAL): Likewise.
3793 (X86_TUNE_FOUR_JUMP_LIMIT): Drop for generic and buldozer.
3794 (X86_TUNE_PAD_RETURNS): Drop for buldozer chips.
3795 (X86_TUNE_AVOID_VECTOR_DECODE): Drop for generic.
3796 (X86_TUNE_REASSOC_FP_TO_PARALLEL): Enable for generic.
3797
c3284718
RS
37982013-09-28 Richard Sandiford <rdsandiford@googlemail.com>
3799
3800 * alloc-pool.c, asan.c, auto-inc-dec.c, basic-block.h, bb-reorder.c,
3801 bitmap.c, bitmap.h, bt-load.c, builtins.c, calls.c, cfgcleanup.c,
3802 cfgexpand.c, cfghooks.c, cfgloop.c, cfgloopmanip.c, cfgrtl.c, cgraph.c,
3803 cgraph.h, cgraphbuild.c, cgraphclones.c, cgraphunit.c, collect2.c,
3804 combine-stack-adj.c, combine.c, compare-elim.c, context.c, context.h,
3805 cprop.c, cse.c, cselib.c, dbxout.c, dce.c, defaults.h, df-core.c,
3806 df-problems.c, df-scan.c, df.h, diagnostic.c, double-int.c, dse.c,
3807 dumpfile.c, dwarf2asm.c, dwarf2cfi.c, dwarf2out.c, emit-rtl.c,
3808 errors.c, except.c, expmed.c, expr.c, file-find.c, final.c,
3809 fixed-value.c, fold-const.c, function.c, fwprop.c, gcc-ar.c, gcc.c,
3810 gcov-io.c, gcov-io.h, gcov.c, gcse.c, genattr-common.c, genattr.c,
3811 genattrtab.c, genautomata.c, genconfig.c, genemit.c, genextract.c,
3812 genflags.c, gengenrtl.c, gengtype-state.c, gengtype.c, genmodes.c,
3813 genopinit.c, genoutput.c, genpeep.c, genpreds.c, genrecog.c,
3814 gensupport.c, ggc-common.c, ggc-page.c, gimple-fold.c, gimple-low.c,
3815 gimple-pretty-print.c, gimple-ssa-strength-reduction.c, gimple.c,
3816 gimple.h, godump.c, graphite-clast-to-gimple.c,
3817 graphite-optimize-isl.c, graphite-poly.h, graphite-sese-to-poly.c,
3818 graphite.c, haifa-sched.c, hash-table.c, hash-table.h, hwint.c,
3819 hwint.h, ifcvt.c, incpath.c, init-regs.c, input.h, intl.c, intl.h,
3820 ipa-cp.c, ipa-devirt.c, ipa-inline-analysis.c, ipa-inline.c,
3821 ipa-profile.c, ipa-pure-const.c, ipa-reference.c, ipa-split.c,
3822 ipa-utils.c, ipa.c, ira-build.c, ira.c, jump.c, loop-doloop.c,
3823 loop-init.c, loop-invariant.c, loop-iv.c, lower-subreg.c, lto-cgraph.c,
3824 lto-streamer-in.c, lto-streamer-out.c, lto-wrapper.c, mcf.c,
3825 mode-switching.c, modulo-sched.c, omp-low.c, optabs.c, opts.c,
3826 pass_manager.h, passes.c, plugin.c, postreload-gcse.c, postreload.c,
3827 predict.c, prefix.c, pretty-print.c, print-rtl.c, print-tree.c,
3828 profile.c, read-md.c, real.c, real.h, recog.c, ree.c, reg-stack.c,
3829 regcprop.c, reginfo.c, regmove.c, regrename.c, regs.h, regstat.c,
3830 reload1.c, reorg.c, rtl.c, rtl.h, rtlanal.c, sbitmap.c, sched-rgn.c,
3831 sdbout.c, sel-sched-ir.c, sel-sched.c, sparseset.c, stack-ptr-mod.c,
3832 statistics.c, stmt.c, stor-layout.c, store-motion.c, streamer-hooks.h,
3833 system.h, target-hooks-macros.h, targhooks.c, targhooks.h, toplev.c,
3834 tracer.c, trans-mem.c, tree-browser.c, tree-call-cdce.c, tree-cfg.c,
3835 tree-cfgcleanup.c, tree-complex.c, tree-data-ref.c, tree-data-ref.h,
3836 tree-eh.c, tree-emutls.c, tree-flow.h, tree-if-conv.c, tree-into-ssa.c,
3837 tree-iterator.c, tree-loop-distribution.c, tree-mudflap.c,
3838 tree-nested.c, tree-nomudflap.c, tree-nrv.c, tree-object-size.c,
3839 tree-optimize.c, tree-pass.h, tree-pretty-print.c, tree-profile.c,
3840 tree-scalar-evolution.c, tree-sra.c, tree-ssa-ccp.c,
3841 tree-ssa-coalesce.c, tree-ssa-copy.c, tree-ssa-copyrename.c,
3842 tree-ssa-dce.c, tree-ssa-dom.c, tree-ssa-dse.c, tree-ssa-forwprop.c,
3843 tree-ssa-ifcombine.c, tree-ssa-live.c, tree-ssa-loop-ch.c,
3844 tree-ssa-loop-im.c, tree-ssa-loop-ivopts.c, tree-ssa-loop-prefetch.c,
3845 tree-ssa-loop.c, tree-ssa-math-opts.c, tree-ssa-operands.c,
3846 tree-ssa-phiopt.c, tree-ssa-phiprop.c, tree-ssa-pre.c,
3847 tree-ssa-reassoc.c, tree-ssa-sink.c, tree-ssa-strlen.c,
3848 tree-ssa-structalias.c, tree-ssa-threadedge.c, tree-ssa-threadupdate.c,
3849 tree-ssa-uncprop.c, tree-ssa-uninit.c, tree-ssa.c, tree-ssanames.c,
3850 tree-stdarg.c, tree-switch-conversion.c, tree-tailcall.c,
3851 tree-vect-data-refs.c, tree-vect-generic.c, tree-vect-loop-manip.c,
3852 tree-vect-stmts.c, tree-vectorizer.c, tree-vectorizer.h, tree-vrp.c,
3853 tree.c, tree.h, tsan.c, tsystem.h, value-prof.c, var-tracking.c,
3854 varasm.c, vec.h, vmsdbgout.c, vtable-verify.c, web.c: Add missing
3855 whitespace before "(".
3856
c6285bd7
SL
38572013-09-28 Sandra Loosemore <sandra@codesourcery.com>
3858
3859 * expr.h (extract_bit_field): Remove packedp parameter.
3860 * expmed.c (extract_fixed_bit_field): Remove packedp parameter
3861 from forward declaration.
3862 (store_split_bit_field): Remove packedp arg from calls to
3863 extract_fixed_bit_field.
3864 (extract_bit_field_1): Remove packedp parameter and packedp
3865 argument from recursive calls and calls to extract_fixed_bit_field.
3866 (extract_bit_field): Remove packedp parameter and corresponding
3867 arg to extract_bit_field_1.
3868 (extract_fixed_bit_field): Remove packedp parameter. Remove code
3869 to issue warnings.
3870 (extract_split_bit_field): Remove packedp arg from call to
3871 extract_fixed_bit_field.
3872 * expr.c (emit_group_load_1): Adjust calls to extract_bit_field.
3873 (copy_blkmode_from_reg): Likewise.
3874 (copy_blkmode_to_reg): Likewise.
3875 (read_complex_part): Likewise.
3876 (store_field): Likewise.
3877 (expand_expr_real_1): Likewise.
3878 * calls.c (store_unaligned_arguments_into_pseudos): Adjust call
3879 to extract_bit_field.
3880 * config/tilegx/tilegx.c (tilegx_expand_unaligned_load): Adjust
3881 call to extract_bit_field.
3882 * config/tilepro/tilepro.c (tilepro_expand_unaligned_load): Adjust
3883 call to extract_bit_field.
3884 * doc/invoke.texi (Code Gen Options): Remove mention of warnings
3885 and special packedp behavior from -fstrict-volatile-bitfields
3886 documentation.
3887
522d4efc
JBG
38882013-09-27 Jan-Benedict Glaw <jbglaw@lug-owl.de>
3889
3890 * lra-eliminations.c (init_elim_table): Guard value_p.
3891
19cdb489
MM
38922013-09-27 Michael Meissner <meissner@linux.vnet.ibm.com>
3893
3894 * config/rs6000/rs6000.c (rs6000_hard_regno_mode_ok): Allow
3895 DFmode, DImode, and SFmode in the upper VSX registers based on the
3896 -mupper-regs-{df,sf} flags. Fix wu constraint to be ALTIVEC_REGS
3897 if -mpower8-vector. Combine -mvsx-timode handling with the rest
3898 of the VSX register handling.
3899
3900 * config/rs6000/rs6000.md (f32_lv): Use %x0 for VSX regsters.
3901 (f32_sv): Likewise.
3902 (zero_extendsidi2_lfiwzx): Add support for loading into the
3903 Altivec registers with -mpower8-vector. Use wu/wv constraints to
3904 only do VSX memory options on Altivec registers.
3905 (extendsidi2_lfiwax): Likewise.
3906 (extendsfdf2_fpr): Likewise.
3907 (mov<mode>_hardfloat, SF/SD modes): Likewise.
3908 (mov<mode>_hardfloat32, DF/DD modes): Likewise.
3909 (mov<mode>_hardfloat64, DF/DD modes): Likewise.
3910 (movdi_internal64): Likewise.
3911
d6d11272
XDL
39122013-09-27 Xinliang David Li <davidxl@google.com>
3913
3914 * opts.c (finish_options): Adjust parameters
3915 according to vect cost model.
3916 (common_handle_option): Set dynamic vect cost
3917 model for FDO.
3918 targhooks.c (default_add_stmt_cost): Compute stmt cost
3919 unconditionally.
3920 * tree-vect-loop.c (vect_estimate_min_profitable_iters):
3921 Use helper function.
3922 * tree-vectorizer.h (unlimited_cost_model): New function.
3923 * tree-vect-slp.c (vect_slp_analyze_bb_1): Use helper function.
3924 * tree-vect-data-refs.c (vect_peeling_hash_insert): Use helper
3925 function.
3926 (vect_enhance_data_refs_alignment): Ditto.
3927 * flag-types.h: New enum.
3928 * common/config/i386/i386-common.c (ix86_option_init_struct):
3929 No need to initialize vect_cost_model flag.
3930 * config/i386/i386.c (ix86_add_stmt_cost): Compute stmt cost
3931 unconditionally.
3932
ac1857a3
DN
39332013-09-27 Diego Novillo <dnovillo@google.com>
3934
3935 * gimple.h (enum ssa_mode): Remove.
3936
0d0e2af6
PM
39372013-09-27 Paulo Matos <pmatos@broadcom.com>
3938
3939 * cfgloop.h (number_of_loops): Fix typo in check for null.
3940
09dc585e
JJ
39412013-09-27 Jakub Jelinek <jakub@redhat.com>
3942
3943 PR middle-end/58551
3944 * tree-cfg.c (move_sese_region_to_fn): Also move loops that
3945 are children of outermost saved_cfun's loop, and set it up to
3946 be moved to dest_cfun's outermost loop. Fix up num_nodes adjustments
3947 if loop != loop0 and SESE region contains bbs that belong to loop0.
3948
ec5a3504
RS
39492013-09-27 Richard Sandiford <rdsandiford@googlemail.com>
3950
3951 * rtlanal.c (must_be_base_p, must_be_index_p): Delete.
3952 (binary_scale_code_p, get_base_term, get_index_term): New functions.
3953 (set_address_segment, set_address_base, set_address_index)
3954 (set_address_disp): Accept the argument unconditionally.
3955 (baseness): Remove must_be_base_p and must_be_index_p checks.
3956 (decompose_normal_address): Classify as much as possible in the
3957 main loop.
3958
f91aec98
RS
39592013-09-27 Richard Sandiford <rdsandiford@googlemail.com>
3960
3961 * cse.c (count_reg_usage): Handle INT_LIST.
3962 * lra-eliminations.c (lra_eliminate_regs_1): Likewise.
3963 * reginfo.c (reg_scan_mark_refs): Likewise.
3964 * reload1.c (eliminate_regs_1): Likewise.
3965
a9195970
IS
39662013-09-27 Iain Sandoe <iain@codesourcery.com>
3967
3968 PR middle-end/58547
3969 * rtlanal.c (lsb_bitfield_op_p): Make both parts of the comparison
3970 signed.
3971
e6f1c509
RB
39722013-09-27 Richard Biener <rguenther@suse.de>
3973
3974 PR tree-optimization/58459
3975 * tree-ssa-forwprop.c (forward_propagate_addr_expr): Remove
3976 restriction not propagating into loops.
3977
84f48495
FW
39782013-09-26 Florian Weimer <fw@deneb.enyo.de>
3979
3980 * tree-ssa.h (walk_use_def_chains_fn, walk_use_def_chains): Delete.
3981 * tree-ssa.c (walk_use_def_chains_1, walk_use_def_chains): Delete.
3982 * doc/tree-ssa.texi (Walking use-def chains): Delete.
3983
74fc8b8a
RB
39842013-09-26 Richard Biener <rguenther@suse.de>
3985
7a5911d3 3986 * tree-into-ssa.c (rewrite_into_ssa): Make more SSA names to anonymous.
74fc8b8a 3987
b4ada065
RB
39882013-09-26 Richard Biener <rguenther@suse.de>
3989
3990 * alias.h (component_uses_parent_alias_set): Rename to ...
3991 (component_uses_parent_alias_set_from): ... this.
3992 * alias.c (component_uses_parent_alias_set): Rename to ...
3993 (component_uses_parent_alias_set_from): ... this and return
3994 the desired parent.
3995 (reference_alias_ptr_type_1): Use the result from
3996 component_uses_parent_alias_set_from instead of stripping
3997 components one at a time.
3998 * emit-rtl.c (set_mem_attributes_minus_bitpos): Adjust.
3999
78bca40d
AM
40002013-09-26 Andrew MacLeod <amacleod@redhat.com>
4001
7a5911d3
UB
4002 * tree-ssa-live.h (find_replaceable_exprs, dump_replaceable_exprs):
4003 Move prototypes to...
78bca40d
AM
4004 * tree-ssa-ter.h: New File. Move prototypes here.
4005 * tree-flow.h (stmt_is_replaceable_p): Remove prototype.
4006 * tree-outof-ssa.h: New. Rename ssaexpand.h, include tree-ssa-ter.h.
4007 * tree-outof-ssa.c (ssa_is_replaceable_p): New. Refactor common bits
7a5911d3 4008 from is_replaceable_p.
78bca40d
AM
4009 * tree-ssa-ter.c (is_replaceable_p, stmt_is_replaceable_p): Delete.
4010 (ter_is_replaceable_p): New. Use new refactored ssa_is_replaceable_p.
4011 (process_replaceable): Use ter_is_replaceable_p.
4012 (find_replaceable_in_bb): Use ter_is_replaceable_p.
4013 * expr.c (stmt_is_replaceable_p): Relocate from tree-ssa-ter.c. Use
4014 newly refactored ssa_is_replaceable_p.
4015 * cfgexpand.c: Include tree-outof-ssa.h.
4016 * ssaexpand.h: Delete.
4017
ff2a63a7
AM
40182013-09-26 Andrew MacLeod <amacleod@redhat.com>
4019
4020 * gimple.c (gimple_replace_lhs): Move to tree-ssa.c and rename.
4021 (struct count_ptr_d, count_ptr_derefs, count_uses_and_derefs): Move to
4022 tree-ssa.c
4023 (create_gimple_tmp): Delete.
4024 (get_expr_type, build_assign, build_type_cast): Move to...
4025 * gimple-builder.c: New File.
4026 (get_expr_type): Relocate from gimple.c.
4027 (build_assign, build_type_cast): Change to only create ssanames.
4028 * gimple.h: Move prototypes to...
4029 * gimple-builder.h: New File. Here.
4030 * tree-ssa.h: And here.
4031 * tree-ssa.c (struct count_ptr_d, count_ptr_derefs,
4032 count_uses_and_derefs): Relocate from gimple.c.
4033 (gimple_replace_ssa_lhs): Renamed gimple_replace_ssa from gimple.c
4034 * tree-ssa-reassoc.c (repropagate_negates): Use gimple_replace_ssa_lhs.
4035 * tree-ssa-math-opts (execute_cse_reciprocals): Use
4036 gimple_replace_ssa_lhs.
4037 * asan.c: Include gimple-builder.h.
4038 * Makefile.in: Add gimple-builder.o.
4039
a2544177
RB
40402013-09-26 Richard Biener <rguenther@suse.de>
4041
4042 * tree-ssa-live.c (var_map_base_init): Handle SSA names with
4043 DECL_IGNORED_P base VAR_DECLs like anonymous SSA names.
4044 (loe_visit_block): Use gcc_checking_assert.
4045 * tree-ssa-coalesce.c (create_outofssa_var_map): Use
4046 gimple_assign_ssa_name_copy_p.
4047 (gimple_can_coalesce_p): Adjust according to the var_map_base_init
4048 change.
4049
6d1ca9a0
DE
40502013-09-26 David Edelsohn <dje.gcc@gmail.com>
4051
4052 * config/rs6000/t-rs6000 (rs6000.o): Remove.
4053 (rs6000-c.o): Use COMPILE and POSTCOMPILE.
4054
46e950db
RB
40552013-09-26 Richard Biener <rguenther@suse.de>
4056
4057 PR tree-optimization/58539
4058 * tree-vect-loop.c (vect_create_epilog_for_reduction): Honor
4059 the fact that debug statements are not taking part in loop-closed
4060 SSA construction.
4061
40ada30a
NC
40622013-09-26 Nick Clifton <nickc@redhat.com>
4063
4064 * config/msp430/msp430.c (msp430_expand_epilogue): Fix compile
4065 time warning message.
4066 (msp430_print_operand_raw): Delete unused letter parameter.
4067 (TARGET_PRINT_OPERAND_ADDRESS): Define.
4068 (msp430_print_operand_address): New function.
4069 (msp430_print_operand): Move address printing code from here to
4070 new function.
4071 * config/msp430/msp430.md (movsipsi2): Add comment in generated
4072 assembler.
4073 (zero_extendpsisi2): Likewise.
4074 (extendpsisi2): New pattern.
4075 (andneghi3): New pattern.
4076
38e8f663
YR
40772013-09-26 Yvan Roux <yvan.roux@linaro.org>
4078
4079 * config/aarch64/aarch64.opt (mlra): New option.
4080 * config/aarch64/aarch64.c (aarch64_lra_p): New function.
4081 (TARGET_LRA_P): Define.
4082
4167e8d4
EB
40832013-09-26 Eric Botcazou <ebotcazou@adacore.com>
4084
4085 * expr.c (expand_assignment): Remove obsolete comment.
4086
5254eac4
JL
40872013-09-25 Jeff Law <law@redhat.com>
4088
4089 * tree-flow.h (thread_through_all_blocks): Prototype moved into
4090 tree-ssa-threadupdate.h.
4091 (register_jump_thread): Similarly.
4092 * tree-ssa-threadupdate.h: New header file.
4093 * tree-ssa-dom.c: Include tree-ssa-threadupdate.h.
4094 * tree-vrp.c: Likewise.
4095 * tree-ssa-threadedge.c: Include tree-ssa-threadupdate.h.
4096 (thread_around_empty_blocks): Change type of path vector argument to
4097 an edge,type pair from just an edge. Initialize both elements when
4098 appending to a jump threading path. Tweak references to elements
4099 appropriately.
4100 (thread_across_edge): Similarly. Release memory for the elements
4101 as needed.
4102 * tree-ssa-threadupdate.c: Include tree-ssa-threadupdate.h.
4103 (dump_jump_thread_path): New function broken out from
4104 register_jump_thread.
4105 (register_jump_thread): Use dump_jump_thread_path. Change type of
4106 path vector entries. Search the path for NULL edges and dump
4107 the path if one is found. Tweak the conversion of path to 3-edge
4108 form to use the block copy type information embedded in the path.
4109
20f114a3
YR
41102013-09-25 Yvan Roux <yvan.roux@linaro.org>
4111
4112 * lra.c (update_inc_notes): Remove all REG_DEAD and REG_UNUSED notes.
4113
3936bafc
YR
41142013-09-25 Yvan Roux <yvan.roux@linaro.org>
4115 Vladimir Makarov <vmakarov@redhat.com>
4116
4117 * rtlanal.c (lsb_bitfield_op_p): New predicate for bitfield operations
4118 from the least significant bit.
4119 (strip_address_mutations): Add bitfield operations handling.
4120 (must_be_index_p): Add shifting and rotate operations handling.
4121 (set_address_base): Use must_be_base_p predicate.
4167e8d4 4122 (set_address_index): Use must_be_index_p predicate.
3936bafc 4123
b86f6e9e
AI
41242013-09-25 Alexander Ivchenko <alexander.ivchenko@intel.com>
4125 Maxim Kuznetsov <maxim.kuznetsov@intel.com>
4126 Sergey Lega <sergey.s.lega@intel.com>
4127 Anna Tikhonova <anna.tikhonova@intel.com>
4128 Ilya Tocar <ilya.tocar@intel.com>
4129 Andrey Turetskiy <andrey.turetskiy@intel.com>
4130 Ilya Verbin <ilya.verbin@intel.com>
4131 Kirill Yukhin <kirill.yukhin@intel.com>
4132 Michael Zolotukhin <michael.v.zolotukhin@intel.com>
4133
4134 * config/i386/i386.c (ix86_avx256_split_vector_move_misalign):
4135 Use new names.
4136 (ix86_expand_vector_move_misalign): Support new unaligned load and
4137 stores and use new names.
4138 (CODE_FOR_sse2_storedqu): Rename to ...
4139 (CODE_FOR_sse2_storedquv16qi): ... this.
4140 (CODE_FOR_sse2_loaddqu): Rename to ...
4141 (CODE_FOR_sse2_loaddquv16qi): ... this.
4142 (CODE_FOR_avx_loaddqu256): Rename to ...
4143 (CODE_FOR_avx_loaddquv32qi): ... this.
4144 (CODE_FOR_avx_storedqu256): Rename to ...
4145 (CODE_FOR_avx_storedquv32qi): ... this.
4146 * config/i386/i386.md (fpint_logic): New.
4147 * config/i386/sse.md (VMOVE): Extend for AVX512.
4148 (VF): Ditto.
4149 (VF_128_256): New.
4150 (VF_512): Ditto.
4151 (VI_UNALIGNED_LOADSTORE): Ditto.
4152 (sse2_avx_avx512f): Ditto.
4153 (sse2_avx2): Extend for AVX512.
4154 (sse4_1_avx2): Ditto.
4155 (avx2_avx512f): New.
4156 (sse): Extend for AVX512.
4157 (sse2): Ditto.
4158 (sse4_1): Ditto.
4159 (avxsizesuffix): Ditto.
4160 (sseintvecmode): Ditto.
4161 (ssePSmode): Ditto.
4162 (<sse>_loadu<ssemodesuffix><avxsizesuffix>): Ditto.
4163 (<sse>_storeu<ssemodesuffix><avxsizesuffix>): Ditto.
4164 (<sse2>_loaddqu<avxsizesuffix>): Extend for AVX512 and rename to ...
4165 (<sse2_avx_avx512f>_loaddqu<mode>): ... this.
4166 (<sse2>_storedqu<avxsizesuffix>): Extend for AVX512 and rename to ...
4167 (<sse2_avx_avx512f>_storedqu<mode): ... this.
4168 (<sse>_movnt<mode>): Replace constraint "x" with "v".
4169 (STORENT_MODE): Extend for AVX512.
4170 (*absneg<mode>2): Replace constraint "x" with "v".
4171 (*mul<mode>3): Ditto.
4172 (*ieee_smin<mode>3): Ditto.
4173 (*ieee_smax<mode>3): Ditto.
4174 (avx_cmp<mode>3): Replace VF with VF_128_256.
4175 (*<sse>_maskcmp<mode>3_comm): Ditto.
4176 (<sse>_maskcmp<mode>3): Ditto.
4177 (<sse>_andnot<mode>3): Extend for AVX512.
4178 (<code><mode>3, anylogic): Replace VF with VF_128_256.
4179 (<code><mode>3, fpint_logic): New.
4180 (*<code><mode>3): Extend for AVX512.
4181 (avx512flogicsuff): New.
4182 (avx512f_<logic><mode>): Ditto.
4183 (<sse>_movmsk<ssemodesuffix><avxsizesuffix>): Replace VF with
4184 VF_128_256.
4185 (<sse4_1>_blend<ssemodesuffix><avxsizesuffix>): Ditto.
4186 (<sse4_1>_blendv<ssemodesuffix><avxsizesuffix>): Ditto.
4187 (<sse4_1>_dp<ssemodesuffix><avxsizesuffix>): Ditto.
4188 (avx_vtest<ssemodesuffix><avxsizesuffix>): Ditto.
4189 (<sse4_1>_round<ssemodesuffix><avxsizesuffix>): Ditto.
4190 (xop_vpermil2<mode>3): Ditto.
4191 (*avx_vpermilp<mode>): Extend for AVX512 and rename to ...
4192 (*<sse2_avx_avx512f>_vpermilp<mode>): ... this.
4193 (avx_vpermilvar<mode>3): Extend for AVX512 and rename to ...
4194 (<sse2_avx_avx512f>_vpermilvar<mode>3): ... this.
4195
4d44d03c
TT
41962013-09-25 Tom Tromey <tromey@redhat.com>
4197
4198 * Makefile.in (PARTITION_H, LTO_SYMTAB_H, COMMON_TARGET_DEF_H)
4199 (RTL_ERROR_H, TRANS_MEM_H, COVERAGE_H, DEMANGLE_H, ALIAS_H)
4200 (SCHED_INT_H, SEL_SCHED_IR_H, SEL_SCHED_DUMP_H, VALTRACK_H, DDG_H)
4201 (GGC_INTERNAL_H, DECNUM_H, BACKTRACE_H, MKDEPS_H, TREE_HASHER_H)
4202 (TREE_SSA_LIVE_H, SSAEXPAND_H, DWARF2OUT_H, SCEV_H, OMEGA_H)
4203 (TREE_DATA_REF_H, IRA_INT_H, LRA_INT_H, DBGCNT_H, DATA_STREAMER_H)
4204 (GIMPLE_STREAMER_H, TREE_STREAMER_H, STREAMER_HOOKS_H)
4205 (TREE_VECTORIZER_H, IPA_INLINE_H, GSTAB_H, LIBFUNCS_H)
4206 (GRAPHITE_HTAB_H): Remove.
4207
86aaf68d
TT
42082013-09-25 Tom Tromey <tromey@redhat.com>
4209
4210 * config/mcore/t-mcore (CROSS_FLOAT_H): Remove.
4211
c987deb8
TT
42122013-09-25 Tom Tromey <tromey@redhat.com>
4213
4214 * config/t-glibc (glibc-c.o): Use COMPILE and POSTCOMPILE.
4215
9439747e
TT
42162013-09-25 Tom Tromey <tromey@redhat.com>
4217
4218 * config/i386/t-i386 (i386.o): Remove.
4219 (i386-c.o): Use COMPILE and POSTCOMPILE.
4220
aefc31a1
TT
42212013-09-25 Tom Tromey <tromey@redhat.com>
4222
4223 * Makefile.in ($(out_object_file)): Use COMPILE and POSTCOMPILE.
4224
b6541edc
TT
42252013-09-25 Tom Tromey <tromey@redhat.com>
4226
4227 * Makefile.in (graph.o, sbitmap.o, sparseset.o, gcc-ar.o)
4228 (gcc-ranlib.o, gcc-nm.o, collect2.o, collect2-aix.o, tlink.o)
4229 (lto-wrapper.o, default-c.o, attribs.o, incpath.o, prefix.o)
4230 (gcc.o, options.o, options-save.o, version.o, gtype-desc.o)
4231 (trans-mem.o, ggc-common.o, ggc-page.o, ggc-none.o, stringpool.o)
4232 (convert.o, double-int.o, lto-compress.o, data-streamer-in.o)
4233 (data-streamer-out.o, data-streamer.o, gimple-streamer-in.o)
4234 (gimple-streamer-out.o, tree-streamer.o, tree-streamer-in.o)
4235 (tree-streamer-out.o, streamer-hooks.o, lto-cgraph.o)
4236 (lto-streamer-in.o, lto-streamer-out.o, lto-section-in.o)
4237 (lto-section-out.o, lto-opts.o, lto-streamer.o, langhooks.o)
4238 (test-dump.o, tree.o, tree-dump.o, tree-inline.o, print-tree.o)
4239 (stor-layout.o, asan.o, tsan.o, ubsan.o, tree-ssa-tail-merge.o)
4240 (tree-ssa-structalias.o, tree-ssa-uninit.o, tree-ssa.o)
4241 (tree-into-ssa.o, tree-ssa-ter.o, tree-ssa-coalesce.o)
4242 (tree-outof-ssa.o, tree-ssa-dse.o, tree-ssa-forwprop.o)
4243 (tree-ssa-phiprop.o, tree-ssa-ifcombine.o, tree-ssa-phiopt.o)
4244 (tree-nrv.o, tree-ssa-copy.o, tree-ssa-propagate.o)
4245 (tree-ssa-dom.o, tree-ssa-uncprop.o, tree-ssa-threadedge.o)
4246 (tree-ssa-threadupdate.o, tree-ssanames.o, tree-phinodes.o)
4247 (domwalk.o, tree-ssa-live.o, tree-ssa-copyrename.o)
4248 (tree-ssa-pre.o, tree-ssa-sccvn.o)
4249 (gimple-ssa-strength-reduction.o, tree-vrp.o, tree-cfg.o)
4250 (tree-cfgcleanup.o, tree-tailcall.o, tree-ssa-sink.o)
4251 (tree-nested.o, tree-if-conv.o, tree-iterator.o, tree-dfa.o)
4252 (tree-ssa-operands.o, tree-eh.o, tree-ssa-loop.o)
4253 (tree-ssa-loop-unswitch.o, tree-ssa-address.o)
4254 (tree-ssa-loop-niter.o, tree-ssa-loop-ivcanon.o)
4255 (tree-ssa-loop-ch.o, tree-ssa-loop-prefetch.o, tree-predcom.o)
4256 (tree-ssa-loop-ivopts.o, tree-affine.o, tree-ssa-loop-manip.o)
4257 (tree-ssa-loop-im.o, tree-ssa-math-opts.o, tree-ssa-alias.o)
4258 (tree-ssa-reassoc.o, tree-optimize.o, gimplify.o)
4259 (gimple-iterator.o, gimple-fold.o, gimple-low.o, omp-low.o)
4260 (tree-browser.o, omega.o, tree-chrec.o, tree-scalar-evolution.o)
4261 (tree-data-ref.o, sese.o, graphite.o, graphite-blocking.o)
4262 (graphite-clast-to-gimple.o, graphite-dependences.o)
4263 (graphite-interchange.o, graphite-poly.o)
4264 (graphite-scop-detection.o, graphite-sese-to-poly.o)
4265 (graphite-optimize-isl.o, tree-vect-loop.o)
4266 (tree-vect-loop-manip.o, tree-vect-patterns.o, tree-vect-slp.o)
4267 (tree-vect-stmts.o, tree-vect-data-refs.o, tree-vectorizer.o)
4268 (vtable-verify.o, tree-loop-distribution.o, tree-parloops.o)
4269 (tree-stdarg.o, tree-object-size.o, internal-fn.o, gimple.o)
4270 (gimple-pretty-print.o, tree-mudflap.o, tree-nomudflap.o)
4271 (tree-pretty-print.o, tree-diagnostic.o, fold-const.o)
4272 (diagnostic.o, diagnostic-color.o, opts.o, opts-global.o)
4273 (opts-common.o, targhooks.o, common/common-targhooks.o, input.o)
4274 (toplev.o, hwint.o, passes.o, plugin.o, main.o, host-default.o)
4275 (rtl-error.o, rtl.o, print-rtl.o, rtlanal.o, varasm.o, function.o)
4276 (statistics.o, stmt.o, except.o, expr.o, dojump.o, builtins.o)
4277 (calls.o, expmed.o, explow.o, optabs.o, dbxout.o, debug.o)
4278 (sdbout.o, dwarf2out.o, dwarf2cfi.o, dwarf2asm.o, vmsdbgout.o)
4279 (xcoffout.o, godump.o, emit-rtl.o, real.o, realmpfr.o, dfp.o)
4280 (fixed-value.o, jump.o, simplify-rtx.o, symtab.o, cgraph.o)
4281 (cgraphunit.o, cgraphclones.o, cgraphbuild.o, varpool.o, ipa.o)
4282 (ipa-profile.o, ipa-devirt.o, ipa-prop.o, ipa-ref.o, ipa-cp.o)
4283 (ipa-split.o, ipa-inline.o, ipa-inline-analysis.o)
4284 (ipa-inline-transform.o, ipa-utils.o, ipa-reference.o)
4285 (ipa-pure-const.o, coverage.o, cselib.o, cse.o, dce.o, dumpfile.o)
4286 (dse.o, fwprop.o, web.o, ree.o, cprop.o, gcse.o, store-motion.o)
4287 (resource.o, lcm.o, mode-switching.o, tree-ssa-dce.o)
4288 (tree-call-cdce.o, tree-ssa-ccp.o, tree-ssa-strlen.o, tree-sra.o)
4289 (tree-switch-conversion.o, tree-complex.o, tree-emutls.o)
4290 (tree-vect-generic.o, df-core.o, df-problems.o, df-scan.o)
4291 (regstat.o, valtrack.o, var-tracking.o, profile.o, mcf.o)
4292 (tree-profile.o, value-prof.o, loop-doloop.o, alloc-pool.o)
4293 (auto-inc-dec.o, cfg.o, cfghooks.o, cfgexpand.o, cfgrtl.o)
4294 (cfganal.o, cfgbuild.o, cfgcleanup.o, cfgloop.o, cfgloopanal.o)
4295 (graphds.o, loop-iv.o, loop-invariant.o, cfgloopmanip.o)
4296 (loop-init.o, loop-unswitch.o, loop-unroll.o, dominance.o)
4297 (et-forest.o, combine.o, reginfo.o, bitmap.o, vec.o, hash-table.o)
4298 (reload.o, reload1.o, rtlhooks.o, postreload.o, postreload-gcse.o)
4299 (caller-save.o, bt-load.o, reorg.o, alias.o, stack-ptr-mod.o)
4300 (init-regs.o, ira-build.o, ira-costs.o, ira-conflicts.o)
4301 (ira-color.o, ira-emit.o, ira-lives.o, ira.o, lra.o)
4302 (lra-assigns.o, lra-coalesce.o, lra-constraints.o)
4303 (lra-eliminations.o, lra-lives.o, lra-spills.o, regmove.o)
4304 (combine-stack-adj.o, compare-elim.o, ddg.o, modulo-sched.o)
4305 (haifa-sched.o, sched-deps.o, sched-rgn.o, sched-ebb.o)
4306 (sched-vis.o, sel-sched.o, sel-sched-dump.o, sel-sched-ir.o)
4307 (final.o, recog.o, reg-stack.o, sreal.o, predict.o, lists.o)
4308 (bb-reorder.o, tracer.o, timevar.o, regcprop.o, regrename.o)
4309 (ifcvt.o, params.o, pointer-set.o, hooks.o, pretty-print.o)
4310 (errors.o, dbgcnt.o, lower-subreg.o, target-globals.o)
4311 (hw-doloop.o, file-find.o, context.o, $(common_out_object_file))
4312 (insn-attrtab.o, insn-automata.o, insn-dfatab.o, insn-emit.o)
4313 (insn-enums.o, insn-extract.o, insn-latencytab.o, insn-modes.o)
4314 (insn-opinit.o, insn-output.o, insn-peep.o, insn-preds.o)
4315 (insn-recog.o, intl.o, cppbuiltin.o, cppdefault.o, gcov.o)
4316 (gcov-dump.o): Remove.
4317 (default-c.o): Use COMPILE and POSTCOMPILE.
4318 (CFLAGS-gcc.o): New variable.
4319 ($(common_out_object_file)): Use COMPILE and POSTCOMPILE.
4320
8c796d19
TT
43212013-09-25 Tom Tromey <tromey@redhat.com>
4322
4323 * Makefile.in (c-family/cppspec.o, c-family/c-common.o)
4324 (c-family/c-cppbuiltin.o, c-family/c-dump.o, c-family/c-format.o)
4325 (c-family/c-gimplify.o, c-family/c-lex.o, c-family/c-omp.o)
4326 (c-family/c-opts.o, c-family/c-pch.o, c-family/c-ppoutput.o)
4327 (c-family/c-pragma.o, c-family/c-pretty-print.o)
4328 (c-family/c-semantics.o, c-family/c-ada-spec.o)
4329 (c-family/array-notation-common.o, c-family/stub-objc.o)
4330 (c-family/c-ubsan.o): Remove.
4331
0a6c2227
TT
43322013-09-25 Tom Tromey <tromey@redhat.com>
4333
4334 * Makefile.in (C_TREE_H): Reference c/c-tree.h.
4335
f3bc55f0
TT
43362013-09-25 Tom Tromey <tromey@redhat.com>
4337
4338 * Makefile.in (DRIVER_DEFINES): Use $(and), not shell code,
4339 to add -DENABLE_SHARED_LIBGCC.
4340 (gcc.o): Don't use subshell.
4341
c11c795e
TT
43422013-09-25 Tom Tromey <tromey@redhat.com>
4343
4344 * Makefile.in (OUTPUT_OPTION): Define as "-o $@".
4345 * configure.ac: Don't invoke AM_PROG_CC_C_O.
4346 (NO_MINUS_C_MINUS_O, OUTPUT_OPTION): Don't subst.
4347 * configure, config.in: Rebuild.
4348
043378c3
TT
43492013-09-25 Tom Tromey <tromey@redhat.com>
4350
4351 * Makefile.in (CCDEPMODE, DEPDIR, depcomp, COMPILE.base)
4352 (COMPILE, POSTCOMPILE): New variables.
4353 (.cc.o .c.o): Use COMPILE, POSTCOMPILE.
4354 (DEPFILES): New variable.
4355 Include ".Po" files.
4356 * configure.ac: Add checks for dependency checking.
4357 * configure, aclocal.m4: Regenerate.
4358
eab34643
TT
43592013-09-25 Tom Tromey <tromey@redhat.com>
4360
4361 * Makefile.in (ALL_HOST_BACKEND_OBJS): Add lto-wrapper.o.
7a5911d3 4362 ($(ALL_HOST_OBJS)): Move order-only dependency to end of file.
eab34643 4363
d2db5e91
TT
43642013-09-25 Tom Tromey <tromey@redhat.com>
4365
4366 * Makefile.in (generated_files): Add options.h,
4367 target-hooks-def.h, insn-opinit.h,
4368 common/common-target-hooks-def.h, pass-instances.def,
4369 c-family/c-target-hooks-def.h.
4370
af33629e
JL
43712013-09-25 Jeff Law <law@redhat.com>
4372
4373 * tree-ssa-threadedge.c (thread_across_edge): Use foo.last () rather
4374 than foo[foo.length () - 1] to access last member in a vec.
4375 * tree-ssa-threadupdate.c (register_jump_thread): Similarly.
4376
ed0d3051
RB
43772013-09-25 Richard Biener <rguenther@suse.de>
4378
4379 PR middle-end/58521
4380 * tree.c (iterative_hash_expr): Remove MEM_REF special handling.
4381
123485ca
JH
43822013-09-25 Jan Hubicka <jh@suse.cz>
4383
4384 * cgraph.c (cgraph_resolve_speculation): Use semantical equivalency
4385 test.
4386
d70b0f1f
MP
43872013-09-25 Marek Polacek <polacek@redhat.com>
4388
4389 PR sanitizer/58420
4390 * ubsan.c (ubsan_type_descriptor): Handle IDENTIFIER_NODEs
4391 when determining the type name.
4392
e3f02798
OE
43932013-09-24 Oleg Endo <olegendo@gcc.gnu.org>
4394
4395 * config/sh/sh.md: Fix formatting.
4396
7a5911d3 43972013-09-24 Xinliang David Li <davidxl@google.com>
4f17aa0b 4398
7a5911d3
UB
4399 * tree-vect-data-refs.c (vect_enhance_data_refs_alignment): Check
4400 max peel iterations parameter.
4f17aa0b
XDL
4401 * param.def: New parameter.
4402 * doc/invoke.texi: Document New parameter.
4403
0498471b
CL
44042013-09-24 Christophe Lyon <christophe.lyon@linaro.org>
4405
4406 * gimple-pretty-print.c: Various whitespace tweaks.
4407 * tree-core.h: Likewise.
4408 * tree-pretty-print.c: Likewise.
4409 * tree-ssa-alias.c: Likewise.
4410 * tree-ssa-copy.c: Likewise.
4411 * tree-ssanames.c: Likewise.
4412 * tree-ssanames.h: Likewise.
4413 * tree-vrp.c: Likewise.
4414
5a8c2b57
AM
44152013-09-24 Alan Modra <amodra@gmail.com>
4416
4417 PR middle-end/57134
4418 PR middle-end/57586
7a5911d3
UB
4419 * stmt.c (expand_asm_operands): Call expand_expr with EXPAND_MEMORY
4420 for output operands that disallow regs. Don't use EXPAND_WRITE on
4421 inout operands.
5a8c2b57 4422
f40333af
RB
44232013-09-24 Richard Biener <rguenther@suse.de>
4424
4425 PR middle-end/58513
4426 * tree.c (reference_alias_ptr_type): Move ...
4427 * alias.c (reference_alias_ptr_type): ... here and implement
4428 in terms of the new reference_alias_ptr_type_1.
4429 (ref_all_alias_ptr_type_p): New helper.
4430 (get_deref_alias_set_1): Drop flag_strict_aliasing here,
4431 use ref_all_alias_ptr_type_p.
4432 (get_deref_alias_set): Add flag_strict_aliasing check here.
4433 (reference_alias_ptr_type_1): New function, split out from ...
4434 (get_alias_set): ... here.
4435 (alias_ptr_types_compatible_p): New function.
4436 * alias.h (reference_alias_ptr_type): Declare.
4437 (alias_ptr_types_compatible_p): Likewise.
4438 * tree.h (reference_alias_ptr_type): Remove.
4439 * fold-const.c (operand_equal_p): Use alias_ptr_types_compatible_p
4440 to compare MEM_REF alias types.
4441
583e8bf5
RB
44422013-09-24 Richard Biener <rguenther@suse.de>
4443
4444 * tree-vrp.c (vrp_finalize): Check for SSA name presence.
4445
5e8586d7
MM
44462013-09-23 Michael Meissner <meissner@linux.vnet.ibm.com>
4447
19c3e797
MM
4448 * config/rs6000/rs6000.c (rs6000_vector_reload): Delete, combine
4449 reload helper function arrays into a single array reg_addr.
4450 (reload_fpr_gpr): Likewise.
4451 (reload_gpr_vsx): Likewise.
4452 (reload_vsx_gpr): Likewise.
4453 (struct rs6000_reg_addr): Likewise.
4454 (reg_addr): Likewise.
4455 (rs6000_debug_reg_global): Change rs6000_vector_reload,
4456 reload_fpr_gpr, reload_gpr_vsx, reload_vsx_gpr uses to reg_addr.
4457 (rs6000_init_hard_regno_mode_ok): Likewise.
4458 (rs6000_secondary_reload_direct_move): Likewise.
4459 (rs6000_secondary_reload): Likewise.
4460
5e8586d7
MM
4461 * config/rs6000/rs6000.h (enum r6000_reg_class_enum): Add new
4462 constraints: wu, ww, and wy. Repurpose wv constraint added during
4463 power8 changes. Put wg constraint in alphabetical order.
4464
4465 * config/rs6000/rs6000.opt (-mvsx-scalar-float): New debug switch
4466 for future work to add ISA 2.07 VSX single precision support.
4467 (-mvsx-scalar-double): Change default from -1 to 1, update
4468 documentation comment.
4469 (-mvsx-scalar-memory): Rename debug switch to -mupper-regs-df.
4470 (-mupper-regs-df): New debug switch to control whether DF values
4471 can go in the traditional Altivec registers.
4472 (-mupper-regs-sf): New debug switch to control whether SF values
4473 can go in the traditional Altivec registers.
4474
4475 * config/rs6000/rs6000.c (rs6000_debug_reg_global): Print wu, ww,
4476 and wy constraints.
4477 (rs6000_init_hard_regno_mode_ok): Use ssize_t instead of int for
4478 loop variables. Rename -mvsx-scalar-memory to -mupper-regs-df.
4479 Add new constraints, wu/ww/wy. Repurpose wv constraint.
4480 (rs6000_debug_legitimate_address_p): Print if we are running
4481 before, during, or after reload.
4482 (rs6000_secondary_reload): Add a comment.
4483 (rs6000_opt_masks): Add -mupper-regs-df, -mupper-regs-sf.
4484
4485 * config/rs6000/constraints.md (wa constraint): Sort w<x>
4486 constraints. Update documentation string.
4487 (wd constraint): Likewise.
4488 (wf constraint): Likewise.
4489 (wg constraint): Likewise.
4490 (wn constraint): Likewise.
4491 (ws constraint): Likewise.
4492 (wt constraint): Likewise.
4493 (wx constraint): Likewise.
4494 (wz constraint): Likewise.
4495 (wu constraint): New constraint for ISA 2.07 SFmode scalar
4496 instructions.
4497 (ww constraint): Likewise.
4498 (wy constraint): Likewise.
358e1993
MM
4499 (wv constraint): Repurpose ISA 2.07 constraint that we did not use
4500 in the previous submissions.
5e8586d7
MM
4501 * doc/md.texi (PowerPC and IBM RS6000): Likewise.
4502
e5af9ddd
RS
45032013-09-23 Richard Sandiford <rdsandiford@googlemail.com>
4504
4505 * doc/rtl.texi (REG_NOTES): Say that int_list can also be used.
4506 (REG_BR_PROB): Say that the probability is stored in an int_list.
4507 * reg-notes.def: Update commentary to mention INT_LIST.
4508 * rtl.def (EXPR_LIST, INSN_LIST): Capitalize comments.
4509 (INT_LIST): New rtx.
4510 * rtl.h (add_int_reg_note, add_shallow_copy_of_reg_note): Declare.
4511 * rtlanal.c (int_reg_note_p): New function.
4512 (alloc_reg_note): Assert that the note does not have an int argument.
4513 (add_int_reg_note, add_shallow_copy_of_reg_note): New functions.
4514 * combine.c (distribute_notes): Use add_shallow_copy_of_rtx.
4515 * cse.c (cse_process_notes_1): Expect REG_EQUAL to be an EXPR_LIST
4516 rather than an INSN_LIST. Handle INT_LIST.
4517 * ifcvt.c (cond_exec_process_insns): Take the probability as an int
4518 rather than an rtx. Use gen_rtx_INT_LIST to create a REG_BR_PROB note.
4519 (cond_exec_process_if_block): Use XINT to extract REG_BR_PROB values.
4520 Manipulate them as ints rather than rtxes.
4521 * reg-stack.c (subst_asm_stack_regs): Only handle EXPR_LIST notes.
4522 * regmove.c (copy_src_to_dest): Likewise.
4523 * sched-vis.c (print_insn_with_notes): Handle INT_LIST.
4524
4525 * config/i386/winnt.c (i386_pe_seh_unwind_emit): Sink pat assignment
4526 into the cases that need it.
4527 * config/arm/arm.c (arm_unwind_emit): Likewise.
4528
4529 * asan.c (asan_clear_shadow): Use add_int_reg_note for REG_BR_PROB.
4530 * emit-rtl.c (try_split, emit_copy_of_insn_after): Likewise.
4531 * loop-doloop.c (add_test, doloop_modify): Likewise.
4532 * loop-unswitch.c (compare_and_jump_seq): Likewise.
4533 * optabs.c (emit_cmp_and_jump_insn_1): Likewise.
4534 * predict.c (combine_predictions_for_insn): Likewise.
4535 * print-rtl.c (print_rtx): Handle INT_LIST.
4536 * config/aarch64/aarch64.c (aarch64_emit_unlikely_jump): Likewise.
4537 * config/alpha/alpha.c (emit_unlikely_jump): Likewise.
4538 * config/arm/arm.c (emit_unlikely_jump): Likewise.
4539 * config/i386/i386.c (ix86_expand_split_stack_prologue): Likewise.
4540 (ix86_split_fp_branch, predict_jump): Likewise.
4541 * config/rs6000/rs6000.c (emit_unlikely_jump): Likewise.
4542 * config/sh/sh.c (expand_cbranchsi4): Likewise.
4543 * config/spu/spu.c (ea_load_store_inline): Likewise.
4544
4545 * cfgbuild.c (compute_outgoing_frequencies): Use XINT to access the
4546 value of a REG_BR_PROB note.
4547 * cfgrtl.c (force_nonfallthru_and_redirect): Likewise.
4548 (update_br_prob_note, rtl_verify_edges, purge_dead_edges): Likewise.
4549 * emit-rtl.c (try_split): Likewise.
4550 * predict.c (br_prob_note_reliable_p): Likewise.
4551 (invert_br_probabilities, combine_predictions_for_insn): Likewise.
4552 * reorg.c (mostly_true_jump): Likewise.
4553 * config/bfin/bfin.c (cbranch_predicted_taken_p): Likewise.
4554 * config/frv/frv.c (frv_print_operand_jump_hint): Likewise.
4555 * config/i386/i386.c (ix86_print_operand): Likewise.
4556 * config/ia64/ia64.c (ia64_print_operand): Likewise.
4557 * config/mmix/mmix.c (mmix_print_operand): Likewise.
4558 * config/rs6000/rs6000.c (output_cbranch): Likewise.
4559 * config/s390/s390.c (s390_expand_tbegin): Likewise.
4560 * config/sh/sh.c (sh_print_operand, sh_adjust_cost): Likewise.
4561 * config/sparc/sparc.c (output_cbranch): Likewise.
4562 * config/spu/spu.c (get_branch_target): Likewise.
4563 * config/tilegx/tilegx.c (cbranch_predicted_p): Likewise.
4564 * config/tilepro/tilepro.c (cbranch_predicted_p): Likewise.
4565
450ad0cd
JH
45662013-09-23 Jan Hubicka <jh@suse.cz>
4567
4568 * ipa-cp.c (ipa_get_indirect_edge_target_1): Add sanity check
4569 for ipa-devirt.
4570 * ipa-utils.h (possible_polymorphic_call_target_p): New function.
4571 * ipa-devirt.c (possible_polymorphic_call_target_p): Be tolerant
4572 of external calls
4573 * gimple-fold.c: Include ipa-utils.h and gimple-pretty-print.h
4574 (gimple_fold_call): Dump inconsistent devirtualizations; add
4575 sanity check for type based devirtualizations.
4576 * ipa-prop.c: Include ipa-utils.h
4577 (ipa_intraprocedural_devirtualization): Add sanity check.
4578 (try_make_edge_direct_virtual_call): Likewise.
4579
cefce769
EB
45802013-09-23 Eric Botcazou <ebotcazou@adacore.com>
4581
4582 * tree-ssa-ccp.c (insert_clobber_before_stack_restore): Recurse on copy
4583 assignment statements.
4584
a895a2b8
KV
45852013-09-23 Kugan Vivekanandarajah <kuganv@linaro.org>
4586
4587 * gimple-pretty-print.c (dump_ssaname_info): New function.
4588 (dump_gimple_phi): Call it.
4589 (pp_gimple_stmt_1): Likewise.
4590 * tree-core.h (tree_ssa_name): New union ssa_name_info_type field.
4591 (range_info_def): Declare.
4592 * tree-pretty-print.c (pp_double_int): New function.
4593 (dump_generic_node): Call it.
4594 * tree-pretty-print.h (pp_double_int): Declare.
4595 * tree-ssa-alias.c (dump_alias_info): Check pointer type.
4596 * tree-ssanames.h (range_info_def): New structure.
4597 (value_range_type): Move definition here.
4598 (set_range_info, value_range_type, duplicate_ssa_name_range_info):
4599 Declare.
4600 * tree-ssanames.c (make_ssa_name_fn): Check pointer type at
4601 initialization.
4602 (set_range_info): New function.
4603 (get_range_info): Likewise.
4604 (duplicate_ssa_name_range_info): Likewise.
4605 (duplicate_ssa_name_fn): Check pointer type and call
4606 duplicate_ssa_name_range_info.
4607 * tree-ssa-copy.c (fini_copy_prop): Likewise.
4608 * tree-vrp.c (value_range_type): Remove definition, now in
4609 tree-ssanames.h.
7a5911d3 4610 (vrp_finalize): Call set_range_info to update value range of SSA_NAMEs.
a895a2b8
KV
4611 * tree.h (SSA_NAME_PTR_INFO): Macro changed to access via union.
4612 (SSA_NAME_RANGE_INFO): New macro.
4613
984af6ac
RB
46142013-09-23 Richard Biener <rguenther@suse.de>
4615
4616 PR tree-optimization/58464
4617 * tree-ssa-pre.c (phi_trans_lookup): Remove.
7a5911d3 4618 (phi_trans_add): Change to add conditionally on being not present.
984af6ac
RB
4619 (phi_translate_1): Remove recursion detection here.
4620 (phi_translate): Pre-seed the cache with NULL to catch
4621 recursion here in a more generic way.
4622 (bitmap_find_leader): Adjust comment.
4623 (get_representative_for): Dump value-numbers.
4624 (create_expression_by_pieces): Likewise.
4625 (insert_into_preds_of_block): Likewise.
4626
0e26cf79
CB
46272013-09-23 Christian Bruel <christian.bruel@st.com>
4628
4629 PR target/58475
4630 * config/sh/sh.md (movsf_ie): Allow fpul_operand.
4631 * config/sh/predicate.md (arith_reg_operand): Disallow FPUL_REG.
4632
4d0e5b54
JG
46332013-09-23 James Greenhalgh <james.greenhalgh@arm.com>
4634
4635 Revert r202780:
4636 2013-09-20 Renlin Li <renlin.li@arm.com>
4637
7a5911d3
UB
4638 * config/aarch64/aarch64.c (aarch64_expand_prologue): Use
4639 plus_constant.
4d0e5b54
JG
4640 (aarch64_expand_epilogue): Likewise.
4641 (aarch64_legitimize_reload_address): Likewise.
4642
ca081cc8
EB
46432013-09-22 Eric Botcazou <ebotcazou@adacore.com>
4644
4645 * gimplify.c (gimplify_asm_expr): Reset the TREE_CHAIN of clobbers to
4646 NULL_TREE before pushing them onto the vector. Likewise for labels.
4647
8fc53a5f
EB
46482013-09-21 Eric Botcazou <ebotcazou@adacore.com>
4649
4650 * config/ia64/predicates.md (ia64_cbranch_operator): Accept unordered
4651 comparison operators when -fno-trapping-math is in effect.
4652 * config/ia64/ia64.c (ia64_expand_compare): Add support for unordered
4653 comparison operators in TFmode and assert that unsupported operators
4654 cannot reach here.
4655 (ia64_print_operand): Likewise.
4656
46572013-09-21 Jan Hubicka <jh@suse.cz>
0ca6c49f
JH
4658
4659 * x86-tune.def (partial_reg_stall): Disable for CoreI7 and newer.
4660 (sse_typeless_stores): Enable for core
4661 (sse_load0_by_pxor): Likewise.
4662 (four_jump_limit): Disable for core.
4663 (pad_returns): Likewise.
4664 (avoid_vector_decode): Likewise.
4665 (fuse_cmp_and_branch): Enable for cores.
4666 * i386.c (x86_accumulate_outgoing_args): Disable for cores.
4667
578e51e6
JDA
46682013-09-20 John David Anglin <danglin@gcc.gnu.org>
4669
96ce28a3
JDA
4670 PR middle-end/56791
4671 * config/pa/pa.c (pa_option_override): Disable auto increment and
4672 decrement instructions until reload is completed.
4673
578e51e6
JDA
4674 * config/pa/pa-linux.h (TARGET_OS_CPP_BUILTINS): Define
4675 __GCC_HAVE_SYNC_COMPARE_AND_SWAP_1, __GCC_HAVE_SYNC_COMPARE_AND_SWAP_2,
4676 and __GCC_HAVE_SYNC_COMPARE_AND_SWAP_4.
4677
6fcd3a13 46782013-09-20 DJ Delorie <dj@redhat.com>
7a5911d3 4679 Nick Clifton <nickc@redhat.com>
6fcd3a13
DD
4680
4681 * config/rl78/rl78.c: Various whitespace and comment tweaks.
4682 (need_to_save): Save bank 0 on interrupts.
4683 (characterize_address): Strip far address wrappers.
4684 (rl78_as_legitimate_address): Likewise.
4685 (transcode_memory_rtx): Likewise.
4686 (rl78_peep_movhi_p): Disable this peephole after devirt.
4687 (rl78_propogate_register_origins): Forget all origins when a
4688 CLOBBER is seen.
4689 * config/rl78/rl78-virt.md: Various whitespace tweaks.
4690 * config/rl78/rl78-real.md: Various whitespace tweaks. Additional
4691 peephole2's.
4692 * config/rl78/rl78.md (sel_rb): Disable for G10 just in case.
4693 * config/rl78/rl78-expand.md (movqi): Check for subregs of consts.
4694 * config/rl78/rl78.h (LINK_SPEC): Pass -gc-sections unless
4695 relocating.
4696 * config/rl78/constraints.md: Various whitespace and paren tweaks.
4697
4a08db33
JDA
46982013-09-20 John David Anglin <danglin@gcc.gnu.org>
4699
4700 * config/pa/pa.md: In "scc" insn patterns, change output template to
4701 handle const0_rtx in reg_or_0_operand operands.
4702
7691132c
JBG
47032013-09-20 Martin Husemann <martin@NetBSD.org>
4704
4705 PR target/56875
4706 * config/vax/vax.c (vax_output_int_move): Use D format specifier.
4707 * config/vax/vax.md (ashldi3, <unnamed>): Ditto.
4708
a3cc13cc
RB
47092013-09-20 Richard Biener <rguenther@suse.de>
4710
4711 PR middle-end/58484
4712 * tree-scalar-evolution.c (struct scev_info_str): Shrink by
4713 remembering SSA name version and block index.
4714 (new_scev_info_str): Adjust.
4715 (hash_scev_info): Likewise. Also hash the block index.
4716 (eq_scev_info): Adjust.
4717 (find_var_scev_info): Likewise.
4718 (struct instantiate_cache_entry): Remove.
4719 (struct instantiate_cache_type): Use a htab to map name, block
4720 to chrec.
4721 (instantiate_cache_type::~instantiate_cache_type): Adjust.
4722 (get_instantiated_value_entry): Likewise.
4723 (hash_idx_scev_info, eq_idx_scev_info): New functions.
4724 (instantiate_scev_name): Adjust.
4725
925f3871
JL
47262013-09-20 Jeff Law <law@redhat.com>
4727
4728 * tree-ssa-dom.c (record_temporary_equivalences): Add comment.
4729
2888c331
YZ
47302013-09-20 Yufeng Zhang <yufeng.zhang@arm.com>
4731
4732 * config/aarch64/aarch64-builtins.c (aarch64_simd_expand_args):
4733 Call aarch64_simd_expand_args to update op[argc].
4734
0a811e96
BS
47352013-09-20 Basile Starynkevitch <basile@starynkevitch.net>
4736
7a5911d3
UB
4737 * plugin.c (parse_plugin_arg_opt): Accept equal sign inside
4738 plugin argument.
0a811e96 4739
0078f462
BS
47402013-09-20 Basile Starynkevitch <basile@starynkevitch.net>
4741
7a5911d3
UB
4742 * gengtype.c (file_rules): Added rule for *.cc files.
4743 (get_output_file_with_visibility): Give fatal message when no
4744 rules found.
0078f462 4745
d20e48b0
RL
47462013-09-20 Renlin Li <renlin.li@arm.com>
4747
4748 * config/aarch64/aarch64.c (aarch64_expand_prologue): Use plus_constant.
4749 (aarch64_expand_epilogue): Likewise.
4750 (aarch64_legitimize_reload_address): Likewise.
4751
59bb1544
BE
47522013-09-20 Bernd Edlinger <bernd.edlinger@hotmail.de>
4753
4754 PR middle-end/57748
4755 * expr.c (expand_assignment): Remove misalignp code path.
4756
a5430019
MP
47572013-09-20 Marek Polacek <polacek@redhat.com>
4758
4759 PR sanitizer/58413
4760 * ubsan.c (get_ubsan_type_info_for_type): Use TYPE_SIZE instead of
4761 TYPE_PRECISION. Add asserts.
4762
9fed7f3a
RB
47632013-09-20 Richard Biener <rguenther@suse.de>
4764
4765 PR tree-optimization/58453
4766 * tree-loop-distribution.c (distribute_loop): Apply the cost
4767 model for -ftree-loop-distribute-patterns, too.
4768
a820c834
RB
47692013-09-20 Richard Biener <rguenther@suse.de>
4770
4771 PR middle-end/58473
4772 * tree-chrec.h (build_polynomial_chrec): Use gcc_checking_assert,
4773 make type comparison less strict.
4774
3cbe17f7
AM
47752013-09-20 Alan Modra <amodra@gmail.com>
4776
4777 * configure: Regenerate.
4778 * aclocal.m4: Regenerate.
4779
87fccdbb
MP
47802013-09-20 Marek Polacek <polacek@redhat.com>
4781
4782 PR other/58467
4783 * doc/extend.texi: Document that attribute used is meant to be used
4784 on variables with static storage duration.
4785
8928eff3
JJ
47862013-09-19 Jakub Jelinek <jakub@redhat.com>
4787
4788 PR tree-optimization/58472
4789 * tree-vect-stmts.c (vectorizable_store, vectorizable_load): For
4790 simd_lane_access set inv_p = false.
4791 * omp-low.c (lower_rec_input_clauses): Set TREE_NO_WARNING on
4792 the simduid magic VAR_DECL.
4793
b7b28c53
JH
47942013-09-19 Jan Hubicka <jh@suse.cz>
4795
4796 * i386.c (generic_memcpy, generic_memset): Fix 32bit template.
4797
83ae86f5
JL
47982013-09-17 Jeff Law <law@redhat.com>
4799
4800 * tree-ssa-dom.c (record_temporary_equivalences): New function
4801 split out of dom_opt_dom_walker::after_dom_children.
4802 (dom_opt_dom_walker::thread_across_edge): Move common code
4803 in here from dom_opt_dom_walker::after_dom_children.
4804 (dom_opt_dom_walker::after_dom_children): Corresponding simplifictions.
4805
9d532162
JH
48062013-09-19 Jan Hubicka <jh@suse.cz>
4807
4808 * i386.h (TARGET_GENERIC32, TARGET_GENERIC64): Remove.
4809 (TARGET_GENERIC): Use PROCESOR_GENERIC
4810 (enum processor_type): Unify generic32 and 64.
4811 * i386.md (cpu): Likewise.
4812 * x86-tune.def (use_leave): Enable for generic32.
4813 (avoid_vector_decode, slow_imul_imm32_mem, slow_imul_imm8): Likewise.
4814 * athlon.md: Change generic64 to generic in all occurences.
4815 * i386-c.c (ix86_target_macros_internal): Unify generic64 and 32.
4816 (ix86_target_macros_internal): Likewise.
4817 * driver-i386.c (host_detect_local_cpu): Likewise.
7a5911d3
UB
4818 * i386.c (generic64_memcpy, generic64_memset, generic64_cost): Rename
4819 to ..
9d532162
JH
4820 (generic_memcpy, generic_memset, generic_cost): This one.
4821 (generic32_memcpy, generic32_memset, generic32_cost): Remove.
4822 (m_GENERIC32, m_GENERIC64): Remove.
4823 (m_GENERIC): Turn into one flag.
4824 (processor_target): Unify generic tunnings.
4825 (ix86_option_override_internal): Replace generic32/64 by generic.
4826 (ix86_issue_rate): Likewise.
4827 (ix86_adjust_cost): Likewise.
4828
188c7d00
JH
48292013-09-19 Jan Hubicka <jh@suse.cz>
4830
4831 * cgraph.c (cgraph_create_edge_1): Avoid uninitialized read
4832 of speculative flag.
4833
4befd127
JJ
48342013-09-19 Jakub Jelinek <jakub@redhat.com>
4835
4836 * omp-low.c (expand_omp_sections): Always pass len - 1 to
4837 GOMP_sections_start, even if !exit_reachable.
4838
8a26ad39
VM
48392013-09-18 Vladimir Makarov <vmakarov@redhat.com>
4840
4841 * lra-constraints.c (need_for_all_save_p): Use macro
4842 HARD_REGNO_CALL_PART_CLOBBERED.
4843 * lra-lives.c (check_pseudos_live_through_calls): Use the macro to
4844 set up pseudo conflict hard regs.
4845
09487185
MM
48462013-09-18 Michael Meissner <meissner@linux.vnet.ibm.com>
4847
4848 PR target/58452
4849 * config/rs6000/paired.md (movmisalignv2sf): Fix to allow memory
4850 operands.
4851
080cbf9e
VM
48522013-09-18 Vladimir Makarov <vmakarov@redhat.com>
4853
4854 PR rtl-optimization/58438
4855 * lra.c (lra): Clear lra_optional_reload_pseudos in upper loop.
4856 * lra-constraints.c (undo_optional_reloads): Keep optional reloads
4857 from previous subpasses.
4858
a0f70fcb
RE
48592013-09-18 Richard Earnshaw <rearnsha@arm.com>
4860
4861 * arm.c (arm_get_frame_offsets): Validate architecture supports
4862 LDRD/STRD before accepting the tuning preference.
4863 (arm_expand_prologue): Likewise.
4864 (arm_expand_epilogue): Likewise.
4865
0547c9b6
RB
48662013-09-18 Richard Biener <rguenther@suse.de>
4867
4868 PR tree-optimization/58417
4869 * tree-chrec.c (chrec_fold_plus_1): Assert that we do not
4870 have chrecs with symbols defined in the loop as operands.
4871 (chrec_fold_multiply): Likewise.
4872 * tree-scalar-evolution.c (interpret_rhs_expr): Instantiate
4873 parameters before folding binary operations.
4874 (struct instantiate_cache_entry_hasher): Remove.
4875 (struct instantiate_cache_type): Use a pointer-map.
4876 (instantiate_cache_type::instantiate_cache_type): New function.
4877 (instantiate_cache_type::get): Likewise.
4878 (instantiate_cache_type::set): Likewise.
4879 (instantiate_cache_type::~instantiate_cache_type): Adjust.
4880 (get_instantiated_value_entry): Likewise.
4881 (global_cache): New global.
4882 (instantiate_scev_r, instantiate_scev_poly, instantiate_scev_binary,
4883 instantiate_array_ref, instantiate_scev_convert, instantiate_scev_3,
4884 instantiate_scev_2, instantiate_scev_1): Do not pass along cache.
4885 (instantiate_scev_name): Adjust.
4886 (instantiate_scev): Construct global instead of local cache.
4887 (resolve_mixers): Likewise.
4888
309dc1aa
DM
48892013-09-18 Daniel Morris <danielm@ecoscentric.com>
4890 Paolo Carlini <paolo.carlini@oracle.com>
4891
4892 PR c++/58458
4893 * doc/implement-cxx.texi: Fix references to the C++ standards.
4894
995a1b4a
JJ
48952013-09-18 Jakub Jelinek <jakub@redhat.com>
4896
4897 * omp-low.c (copy_var_decl): Copy DECL_ATTRIBUTES.
4898 * tree-vect-data-refs.c (vect_analyze_data_refs): For
4899 simd_lane_access drs, update also DR_ALIGNED_TO.
4900
ce6923c5
MP
49012013-09-18 Marek Polacek <polacek@redhat.com>
4902
4903 PR sanitizer/58411
4904 * doc/extend.texi: Document no_sanitize_undefined attribute.
4905 * builtins.c (fold_builtin_0): Don't sanitize function if it has the
4906 no_sanitize_undefined attribute.
4907
d30d00a2
NC
49082013-09-18 Nick Clifton <nickc@redhat.com>
4909
4910 * config/msp430/msp430.h (ASM_SPEC): Pass -md on to the assembler.
4911 (ASM_DECLARE_FUNCTION_NAME): Define.
4912
4d9192b5
TS
49132013-09-17 Trevor Saunders <tsaunders@mozilla.com>
4914
4915 * compare-elim.c (find_comparison_dom_walker): New class
4916 (find_comparisons_in_bb): Rename to
4917 find_comparison_dom_walker::before_dom_children
4918 (find_comparisons): Adjust
4919 * domwalk.c (walk_dominator_tree): Rename to dom_walker::walk, and
4920 adjust.
4921 (init_walk_dominator_tree, fini_walk_dominator_tree): Remove
4922 * domwalk.h (dom_walk_data): Convert it To a class dom_walker.
4923 (init_walk_dominator_tree): Remove declaration.
4924 (fini_walk_dominator_tree): Remove declaration.
4925 * fwprop.c (single_def_use_dom_walker): New class
4926 (single_def_use_enter_block): Convert to
4927 single_def_use_dom_walker::before_dom_children.
4928 (single_def_use_leave_block): Convert to
4929 single_def_use_dom_walker::after_dom_children.
4930 (build_single_def_use_links): Adjust.
4931 * gimple-ssa-strength-reduction.c (find_candidates_dom_walker): New
4932 class.
4933 (find_candidates_in_block): Convert to
4934 find_candidates_dom_walker::before_dom_children.
4935 (execute_strength_reduction): Adjust.
4936 * graphite-sese-to-poly.c (struct bsc, build_sese_conditions): Remove.
4937 (sese_dom_walker): New class.
4938 (sese_dom_walker::sese_dom_walker): New constructor.
4939 (sese_dom_walker::~sese_dom_walker): New destructor.
4940 (build_sese_conditions_before): Convert to
4941 sese_dom_walker::before_dom_children.
4942 (build_sese_conditions_after): Convert to
4943 sese_dom_walker::after_dom_children.
4944 (build_poly_scop): Adjust
4945 * tree-into-ssa.c (rewrite_dom_walker): New class
4946 (rewrite_enter_block): Convert to
4947 rewrite_dom_walker::before_dom_children.
4948 (rewrite_leave_block): Convert to
4949 rewrite_dom_walker::after_dom_children.
4950 (rewrite_update_dom_walker): New class.
4951 (rewrite_update_enter_block): Convert to
4952 rewrite_update_dom_walker::before_dom_children.
4953 (rewrite_update_leave_block): Convert to
4954 rewrite_update_dom_walker::after_dom_children.
4955 (rewrite_blocks, rewrite_into_ssa): Adjust.
4956 (mark_def_dom_walker): New class.
4957 (mark_def_dom_walker::mark_def_dom_walker): New constructor.
4958 (mark_def_dom_walker::~mark_def_dom_walker): New destructor.
4959 (mark_def_sites_blocks): Convert to
4960 mark_def_dom_walker::before_dom_children.
4961 (mark_def_site_blocks): Remove.
4962 * tree-ssa-dom.c (dom_opt_dom_walker): New class.
4963 (tree_ssa_dominator_optimize): Adjust.
4964 (dom_thread_across_edge): Convert to method
4965 dom_opt_dom_walker::thread_across_edge.
4966 (dom_opt_enter_block): Convert to member function
4967 dom_opt_dom_walker::before_dom_children.
4968 (dom_opt_leave_block): Convert to member function
4969 dom_opt_dom_walker::after_dom_children.
4970 * tree-ssa-dse.c (dse_dom_walker): New class.
4971 (dse_enter_block): Convert to member function
4972 dse_dom_walker::before_dom_children.
4973 (tree_ssa_dse): Adjust.
4974 * tree-ssa-loop-im.c (invariantness_dom_walker): New class.
4975 (determine_invariantness_stmt): Convert to method
4976 invariantness_dom_walker::before_dom_children.
4977 (determine_invariantness): Remove
4978 (move_computations_dom_walker): New class.
4979 (move_computations_stmt): Convert to method
4980 move_computations_dom_walker::before_dom_children.
4981 (move_computations, tree_ssa_lim): Adjust.
4982 * tree-ssa-phiopt.c (nontrapping_dom_walker): new class
4983 (nt_init_block): Make method
4984 notrappping_dom_walker::before_dom_children.
4985 (nt_fini_block): Make
4986 method nontrapping_dom_walker::after_dom_children.
4987 (get_non_trapping): Adjust.
4988 * tree-ssa-pre.c (eliminate_dom_walker): New class.
4989 (eliminate_bb): Make method eliminate_dom_walker::before_dom_children.
4990 (eliminate_leave_block): Make method.
4991 eliminate_dom_walker::after_dom_children.
4992 (eliminate): Adjust
4993 * tree-ssa-strlen.c (strlen_dom_walker): New class.
4994 (strlen_enter_block): Make method
4995 strlen_dom_walker::before_dom_children.
4996 (strlen_leave_block): Make
4997 method strlen_dom_walker::after_dom_children.
4998 (tree_ssa_strlen): Adjust.
4999 * tree-ssa-uncprop.c (uncprop_dom_walker): New class.
5000 (tree_ssa_uncprop): Adjust.
5001 (uncprop_leave_block): Make method
5002 uncprop_dom_walker::after_dom_children.
5003 (uncprop_leave_block): Make method
5004 uncprop_dom_walker::before_dom_children.
5005
3b3cc26b
BC
50062013-09-18 Bin Cheng <bin.cheng@arm.com>
5007
7a5911d3
UB
5008 * config/arm/arm.c (thumb1_reorg): Search for flag setting insn before
5009 branch in same basic block. Check both src and dest of the move insn.
3b3cc26b 5010
b5ba3a9b
NC
50112013-09-17 Nick Clifton <nickc@redhat.com>
5012
5013 * config/rl78/rl78-real.md (bf): New pattern.
5014 (bt): New pattern.
5015 * config/rl78/rl78.c (rl78_print_operand_1): Handle %B.
5016 (rl78_print_operand): Do not put a # before a %B.
5017 * config/rl78/rl78.opt: Tweak doc strings.
5018
09d5094b
DD
50192013-09-17 DJ Delorie <dj@redhat.com>
5020
5021 * config/rl78/constraints.md (Wcv): Allow up to $r31.
5022 * config/rl78/rl78.c (rl78_asm_file_start: Likewise.
5023 (rl78_option_override): Likewise, if -mallregs.
5024 (is_virtual_register): Likewise.
5025 * config/rl78/rl78.h (reg_class): Extend VREGS to $r31.
5026 (REGNO_OK_FOR_BASE_P): Likewise.
5027 * config/rl78/rl78.opt (-mallregs): New.
5028
2ba1dca1
NC
50292013-09-17 Nick Clifton <nickc@redhat.com>
5030
5031 * config/rl78/rl78.c (need_to_save): Change return type to bool.
5032 For interrupt functions: save all call clobbered registers if the
5033 interrupt handler is not a leaf function.
5034 (rl78_expand_prologue): Always recompute the frame information.
5035 For interrupt functions: only select bank 0 if one of the bank 0
5036 registers is going to be psuhed.
5037
2e7c3f21
DD
50382013-09-17 DJ Delorie <dj@redhat.com>
5039
5040 * config/rl78/constraints.md: For each W* constraint, rename to C*
5041 and create a W* constraint that checks for an optional ES: prefix
5042 pattern also.
5043 * config/rl78/rl78.md (UNS_ES_ADDR): New.
5044 (es_addr): New. Used to wrap far addresses.
5045 * config/rl78/rl78-protos.h (rl78_es_addr): New.
5046 (rl78_es_base): New.
5047 * config/rl78/rl78.c (rl78_as_legitimate_address): Accept "unspec"
5048 wrapped far addresses.
5049 (rl78_print_operand_1): Unwrap far addresses before processing.
5050 (rl78_lo16): Wrap far addresses in unspecs.
5051 (rl78_es_addr): New.
5052 (rl78_es_base): New.
5053 (insn_ok_now): Check for not-yet-wrapped far addresses.
5054 (transcode_memory_rtx): Properly re-wrap far addresses.
5055
5c26a69a
SH
50562013-09-17 Sebastian Huber <sebastian.huber@embedded-brains.de>
5057
5058 * config/sparc/t-rtems: Add leon3 multilibs.
5059
181f5f3e
CH
50602013-09-17 Cong Hou <congh@google.com>
5061
5062 * tree-vect-patterns.c (vect_recog_dot_prod_pattern): Fix a bug
5063 when checking the dot production pattern. The type of rhs operand
5064 of multiply is now checked correctly.
5065
8d34e421
JL
50662013-09-17 Jeff Law <law@redhat.com>
5067
5068 * tree-ssa-dom.c (cprop_into_successor_phis): Also propagate
5069 edge implied equivalences into successor phis.
5070 * tree-ssa-threadupdate.c (phi_args_equal_on_edges): Moved into
5071 here from tree-ssa-threadedge.c.
5072 (mark_threaded_blocks): When threading through a joiner, if both
5073 successors of the joiner's clone reach the same block, verify the
5074 PHI arguments are equal. If not, cancel the jump threading request.
5075 * tree-ssa-threadedge.c (phi_args_equal_on_edges): Moved into
5076 tree-ssa-threadupdate.c
5077 (thread_across_edge): Don't check PHI argument equality when
5078 threading through joiner block here.
7a5911d3 5079
c152901f
AM
50802013-09-17 Andrew MacLeod <amacleod@redhat.com>
5081
5082 * tree-flow.h (ssa_undefined_value_p): Remove prototype.
5083 * tree-ssa.c (ssa_undefined_value_p): Move pass independent parts here.
7a5911d3
UB
5084 (warn_uninit, warn_uninitialized_vars,
5085 execute_early_warn_uninitialized, make_pass_early_warn_uninitialized):
5086 Move to tree-ssa-uninit.c.
78cedfb1 5087 * tree-ssa-uninit.c (ssa_undefined_value_p): Move to tree-ssa.c.
7a5911d3 5088 (has_undefined_value_p): New. Pass dependant parts of
c152901f
AM
5089 ssa_undefined_value_p.
5090 (uninit_undefined_value_p): Use has_undefined_value_p.
7a5911d3
UB
5091 (warn_uninit, warn_uninitialized_vars,
5092 execute_early_warn_uninitialized, make_pass_early_warn_uninitialized):
5093 Move from tree-ssa.c.
5094 * tree-ssa.h: Adjust prototypes.
c152901f 5095
b631d45a
JH
50962013-09-17 Jan Hubicka <jh@suse.cz>
5097
5098 PR middle-end/58332
5099 * cif-code.def (FUNCTION_NOT_OPTIMIZED): New CIF code.
5100 * ipa-inline.c (can_inline_edge_p): Do not downgrade
5101 FUNCTION_NOT_OPTIMIZED.
5102 * ipa-inline-analysis.c (compute_inline_parameters): Function
5103 not optimized is not inlinable unless it is alwaysinline.
5104 (inline_analyze_function): Force calls in not optimized
5105 function not inlinable.
5106
5b79657a
JH
51072013-09-17 Jan Hubicka <jh@suse.cz>
5108
5109 PR middle-end/58329
5110 * ipa-devirt.c (ipa_devirt): Be ready for symtab_nonoverwritable_alias
5111 to return NULL.
5112 * ipa.c (function_and_variable_visibility): Likewise.
5113 * ipa-profile.c (ipa_profile): Likewise.
5114
27442c24
BE
51152013-09-17 Bernd Edlinger <bernd.edlinger@hotmail.de>
5116
5117 PR ipa/58398
5118 * cgraph.c (cgraph_function_body_availability): Check for ifunc
5119 attribute, and don't inline the resolver in this case.
5120
753b9bc3
TJ
51212013-09-17 Teresa Johnson <tejohnson@google.com>
5122
5123 * coverage.c (get_coverage_counts): Add missing newline.
5124
e3d3cfb4
KT
51252013-09-17 Kyrylo Tkachov <kyrylo.tkachov@arm.com>
5126
5127 PR tree-optimization/58088
5128 * fold-const.c (mask_with_trailing_zeros): New function.
5129 (fold_binary_loc): Make sure we don't recurse infinitely
5130 when the X in (X & C1) | C2 is a tree of the form (Y * K1) & K2.
5131 Use mask_with_trailing_zeros where appropriate.
5132
9a0ac98f
YR
51332013-09-17 Yuri Rumyantsev <ysrumyan@gmail.com>
5134
5135 * config/i386/i386.c (distance_agu_use_in_bb) : Proper initialization
5136 of 'prev' var to get better distance estimation.
5137
78bbd765
EB
51382013-09-17 Eric Botcazou <ebotcazou@adacore.com>
5139
5140 * tree-inline.h (struct copy_body_data): Add transform_parameter.
5141 * tree-inline.c (is_parameter_of): New predicate.
5142 (remap_gimple_op_r): Do not propagate TREE_THIS_NOTRAP on MEM_REF if
5143 a parameter has been remapped.
5144 (copy_tree_body_r): Likewise on INDIRECT_REF and MEM_REF.
5145 (optimize_inline_calls): Initialize transform_parameter.
5146 (copy_gimple_seq_and_replace_locals): Likewise.
5147 (tree_function_versioning): Likewise.
5148 (maybe_inline_call_in_expr): Likewise.
5149
cad055a4
NC
51502013-09-17 Nick Clifton <nickc@redhat.com>
5151
5152 * config/msp430/msp430-protos.h: Add prototypes for new functions.
5153 * config/msp430/msp430.c (msp430_preserve_reg_p): Add support for
5154 interrupt handlers.
5155 (is_attr_func): New function.
5156 (msp430_is_interrupt_func): New function.
5157 (is_naked_func): New function.
5158 (is_reentrant_func): New function.
5159 (is_critical_func): New function.
5160 (msp430_start_function): Add annotations for function attributes.
5161 (msp430_attr): New function.
5162 (msp430_attribute_table): New.
5163 (msp430_function_section): New function.
5164 (TARGET_ASM_FUNCTION_SECTION): Define.
5165 (msp430_builtin): New enum.
5166 (msp430_init_builtins): New function.
5167 (msp430_builtin_devl): New function.
5168 (msp430_expand_builtin): New function.
5169 (TARGET_INIT_BUILTINS): Define.
5170 (TARGET_EXPAND_BUILTINS): Define.
5171 (TARGET_BUILTIN_DECL): Define.
5172 (msp430_expand_prologue): Add support for naked, interrupt,
5173 critical and reentrant functions.
5174 (msp430_expand_epilogue): Likewise.
5175 (msp430_print_operand): Handle 'O' character.
5176 * config/msp430/msp430.h (TARGET_CPU_CPP_BUILTINS): Define
5177 NO_TRAMPOLINES.
5178 * config/msp430/msp430.md (unspec): Add UNS_DINT, UNS_EINT,
5179 UNS_PUSH_INTR, UNS_POP_INTR, UNS_BIC_SR, UNS_BIS_SR.
5180 (pushm): Use a 'n' rather than an 'i' constraint.
5181 (msp_return): Add generation of the interrupt return instruction.
5182 (disable_interrupts): New pattern.
5183 (enable_interrupts): New pattern.
5184 (push_intr_state): New pattern.
5185 (pop_intr_state): New pattern.
5186 (bic_SR): New pattern.
5187 (bis_SR): New pattern.
5188 * doc/extend.texi: Document MSP430 function attributes and builtin
5189 functions.
5190
deb6c11a
RB
51912013-09-17 Richard Biener <rguenther@suse.de>
5192
5193 PR tree-optimization/58432
5194 * tree-loop-distribution.c (tree_loop_distribution): Also
5195 scan PHIs for outside loop uses and seed a partition from them.
5196
c068654b
BC
51972013-09-17 Bin Cheng <bin.cheng@arm.com>
5198
5199 * gimple-ssa-strength-reduction.c (backtrace_base_for_ref): New.
5200 (restructure_reference): Call backtrace_base_for_ref.
5201
567969e4
AM
52022013-09-17 Alan Modra <amodra@gmail.com>
5203
5204 PR target/57589
7a5911d3
UB
5205 * config/rs6000/driver-rs6000.c (elf_platform): Revert 2013-06-11
5206 patch.
567969e4 5207
5c0029de
DD
52082013-09-16 DJ Delorie <dj@redhat.com>
5209
5210 * config/rl78/rl78.c (rl78_asm_file_start): Specify alternate
5211 vregs location for RL78/G10.
5212 (rl78_expand_prologue): Avoid SEL on G10.
5213 (rl78_expand_epilogue): Likewise.
5214 (rl78_peep_movhi_p): Can't move a constant to memory in HImode.
5215 * config/rl78/rl78.h (TARGET_CPU_CPP_BUILTINS): Define
5216 __RL78_G10__ when appropriate.
5217 (ASM_SPEC): Pass -mg10 along to the assembler.
5218 * config/rl78/rl78.md (sel_rb): Disable for G10.
5219 * config/rl78/rl78.opt: Add -mg10 option.
5220 * config/rl78/t-rl78: Add -mg10 multilib.
5221
ea0f3e87
XDL
52222013-09-16 Xinliang David Li <davidxl@google.com>
5223
5224 * tree-if-conv.c (main_tree_if_conversion): Check new flag.
5225 * omp-low.c (omp_max_vf): Ditto.
5226 (expand_omp_simd): Ditto.
5227 * tree-vectorizer.c (vectorize_loops): Ditto.
5228 (gate_vect_slp): Ditto.
5229 (gate_increase_alignment): Ditto.
5230 * tree-ssa-pre.c (inhibit_phi_insertion): Ditto.
5231 * tree-ssa-loop.c (gate_tree_vectorize): Ditto.
5232 (gate_tree_vectorize): Name change.
5233 (tree_vectorize): Ditto.
5234 (pass_vectorize::gate): Call new function.
5235 (pass_vectorize::execute): Ditto.
7a5911d3 5236 * opts.c: O3 default setting change.
ea0f3e87
XDL
5237 (finish_options): Check new flag.
5238 * doc/invoke.texi: Document new flags.
5239 * common.opt: New flags.
5240
4f3da2e5
AS
52412013-09-16 Andreas Schwab <schwab@linux-m68k.org>
5242
5243 * doc/tm.texi.in (Cond Exec Macros): Remove node.
5244 (Condition Code): Don't reference it.
5245 * doc/tm.texi: Regenerate.
5246
aa940f7c
VM
52472013-09-16 Vladimir Makarov <vmakarov@redhat.com>
5248
5249 PR middle-end/58418
5250 * lra-constraints.c (undo_optional_reloads): Consider all optional
5251 reload even if it did not get a hard reg.
5252
e645e942
TJ
52532013-09-16 Teresa Johnson <tejohnson@google.com>
5254
5255 * dumpfile.c (dump_loc): Remove newline emission.
5256 * tree-vect-data-refs.c (vect_lanes_optab_supported_p): Add newline
5257 emission to dump_printf_loc calls where missing.
5258 (vect_mark_for_runtime_alias_test): Ditto.
5259 (vect_analyze_data_ref_dependence): Ditto.
5260 (vect_analyze_data_ref_dependences): Ditto.
5261 (vect_slp_analyze_data_ref_dependence): Ditto.
5262 (vect_slp_analyze_data_ref_dependences): Ditto.
5263 (vect_compute_data_ref_alignment): Ditto.
5264 (vect_update_misalignment_for_peel): Ditto.
5265 (vect_verify_datarefs_alignment): Ditto.
5266 (vector_alignment_reachable_p): Ditto.
5267 (vect_get_data_access_cost): Ditto.
5268 (vect_enhance_data_refs_alignment): Ditto.
5269 (vect_find_same_alignment_drs): Ditto.
5270 (vect_analyze_data_refs_alignment): Ditto.
5271 (vect_analyze_group_access): Ditto.
5272 (vect_analyze_data_ref_access): Ditto.
5273 (vect_analyze_data_ref_accesses): Ditto.
5274 (vect_prune_runtime_alias_test_list): Ditto.
5275 (vect_analyze_data_refs): Ditto.
5276 (vect_create_addr_base_for_vector_ref): Ditto.
5277 (vect_create_data_ref_ptr): Ditto.
5278 (vect_grouped_store_supported): Ditto.
5279 (vect_grouped_load_supported): Ditto.
5280 * value-prof.c (check_counter): Ditto.
5281 (check_ic_target): Ditto.
5282 * tree-vect-patterns.c (vect_recog_dot_prod_pattern): Ditto.
5283 (vect_recog_widen_mult_pattern): Ditto.
5284 (vect_recog_widen_sum_pattern): Ditto.
5285 (vect_recog_over_widening_pattern): Ditto.
5286 (vect_recog_widen_shift_pattern): Ditto.
5287 (vect_recog_rotate_pattern): Ditto.
5288 (vect_recog_vector_vector_shift_pattern): Ditto.
5289 (vect_recog_divmod_pattern): Ditto.
5290 (vect_recog_mixed_size_cond_pattern): Ditto.
5291 (vect_recog_bool_pattern): Ditto.
5292 (vect_pattern_recog_1): Ditto.
5293 (vect_pattern_recog): Ditto.
5294 * tree-vect-loop.c (vect_determine_vectorization_factor): Ditto.
5295 (vect_is_simple_iv_evolution): Ditto.
5296 (vect_analyze_scalar_cycles_1): Ditto.
5297 (vect_get_loop_niters): Ditto.
5298 (vect_analyze_loop_1): Ditto.
5299 (vect_analyze_loop_form): Ditto.
5300 (vect_analyze_loop_operations): Ditto.
5301 (vect_analyze_loop_2): Ditto.
5302 (vect_analyze_loop): Ditto.
5303 (report_vect_op): Ditto.
5304 (vect_is_slp_reduction): Ditto.
5305 (vect_is_simple_reduction_1): Ditto.
5306 (vect_get_known_peeling_cost): Ditto.
5307 (vect_estimate_min_profitable_iters): Ditto.
5308 (vect_model_reduction_cost): Ditto.
5309 (vect_model_induction_cost): Ditto.
5310 (get_initial_def_for_induction): Ditto.
5311 (vect_create_epilog_for_reduction): Ditto.
5312 (vectorizable_reduction): Ditto.
5313 (vectorizable_induction): Ditto.
5314 (vectorizable_live_operation): Ditto.
5315 (vect_loop_kill_debug_uses): Ditto.
5316 (vect_transform_loop): Ditto.
5317 * tree-vect-stmts.c (vect_mark_relevant): Ditto.
5318 (vect_stmt_relevant_p): Ditto.
5319 (process_use): Ditto.
5320 (vect_mark_stmts_to_be_vectorized): Ditto.
5321 (vect_model_simple_cost): Ditto.
5322 (vect_model_promotion_demotion_cost): Ditto.
5323 (vect_model_store_cost): Ditto.
5324 (vect_get_store_cost): Ditto.
5325 (vect_model_load_cost): Ditto.
5326 (vect_get_load_cost): Ditto.
5327 (vect_init_vector_1): Ditto.
5328 (vect_get_vec_def_for_operand): Ditto.
5329 (vect_finish_stmt_generation): Ditto.
5330 (vectorizable_call): Ditto.
5331 (vectorizable_conversion): Ditto.
5332 (vectorizable_assignment): Ditto.
5333 (vectorizable_shift): Ditto.
5334 (vectorizable_operation): Ditto.
5335 (vectorizable_store): Ditto.
5336 (vectorizable_load): Ditto.
5337 (vectorizable_condition): Ditto.
5338 (vect_analyze_stmt): Ditto.
5339 (vect_transform_stmt): Ditto.
5340 (vect_is_simple_use): Ditto.
5341 * tree-vect-loop-manip.c (slpeel_make_loop_iterate_ntimes): Ditto.
5342 (vect_can_advance_ivs_p): Ditto.
5343 (vect_update_ivs_after_vectorizer): Ditto.
5344 (vect_do_peeling_for_loop_bound): Ditto.
5345 (vect_gen_niters_for_prolog_loop): Ditto.
5346 (vect_update_inits_of_drs): Ditto.
5347 (vect_create_cond_for_alias_checks): Ditto.
5348 * tree-vect-slp.c (vect_get_and_check_slp_defs): Ditto.
5349 (vect_build_slp_tree_1): Ditto.
5350 (vect_supported_load_permutation_p): Ditto.
5351 (vect_analyze_slp_instance): Ditto.
5352 (vect_analyze_slp): Ditto.
5353 (vect_make_slp_decision): Ditto.
5354 (vect_detect_hybrid_slp): Ditto.
5355 (vect_bb_vectorization_profitable_p): Ditto.
5356 (vect_slp_analyze_bb_1): Ditto.
5357 (vect_update_slp_costs_according_to_vf): Ditto.
5358 (vect_get_mask_element): Ditto.
5359 (vect_transform_slp_perm_load): Ditto.
5360 (vect_schedule_slp_instance): Ditto.
5361 (vect_schedule_slp): Ditto.
5362 (vect_slp_transform_bb): Ditto.
5363 * profile.c (read_profile_edge_counts): Ditto.
5364 (compute_branch_probabilities): Ditto.
5365 * coverage.c (get_coverage_counts): Ditto.
5366
46f851f3
DN
53672013-09-16 Diego Novillo <dnovillo@google.com>
5368
7a5911d3 5369 * tree-core.h: Add missing comment lines from refactoring of tree.h.
46f851f3 5370
1632a686
JH
53712013-09-16 Jan Hubicka <jh@suse.cz>
5372
5373 * gimple-fold.c (can_refer_decl_in_current_unit_p): Do not accept
7a5911d3 5374 abstract functions; for static functions check the presence of body.
1632a686 5375
828e70c1
JG
53762013-09-16 James Greenhalgh <james.greenhalgh@arm.com>
5377
5378 * config/aarch64/aarch64-simd-builtins.def (fma): New.
5379 * config/aarch64/aarch64-simd.md
5380 (aarch64_mla_elt<mode>): New.
5381 (aarch64_mla_elt_<vswap_width_name><mode>): Likewise.
5382 (aarch64_mls_elt<mode>): Likewise.
5383 (aarch64_mls_elt_<vswap_width_name><mode>): Likewise.
5384 (aarch64_fma4_elt<mode>): Likewise.
5385 (aarch64_fma4_elt_<vswap_width_name><mode>): Likewise.
5386 (aarch64_fma4_elt_to_128v2df): Likewise.
5387 (aarch64_fma4_elt_to_64df): Likewise.
5388 (fnma<mode>4): Likewise.
5389 (aarch64_fnma4_elt<mode>): Likewise.
5390 (aarch64_fnma4_elt_<vswap_width_name><mode>): Likewise.
5391 (aarch64_fnma4_elt_to_128v2df): Likewise.
5392 (aarch64_fnma4_elt_to_64df): Likewise.
5393 * config/aarch64/iterators.md (VDQSF): New.
5394 * config/aarch64/arm_neon.h
5395 (vfm<as><sdq>_lane<q>_f<32, 64>): Convert to C implementation.
5396 (vml<sa><q>_lane<q>_<fsu><16, 32, 64>): Likewise.
5397
779aea46
JG
53982013-09-16 James Greenhalgh <james.greenhalgh@arm.com>
5399
5400 * config/aarch64/aarch64-simd.md (aarch64_mul3_elt<mode>): New.
5401 (aarch64_mul3_elt_<vswap_width_name><mode>): Likewise.
5402 (aarch64_mul3_elt_to_128df): Likewise.
5403 (aarch64_mul3_elt_to_64v2df): Likewise.
5404 * config/aarch64/iterators.md (VEL): Also handle DFmode.
5405 (VMUL): New.
5406 (VMUL_CHANGE_NLANES) Likewise.
5407 (h_con): Likewise.
5408 (f): Likewise.
5409 * config/aarch64/arm_neon.h
5410 (vmul<q>_lane<q>_<suf><16,32,64>): Convert to C implementation.
5411
a407a750
JG
54122013-09-16 James Greenhalgh <james.greenhalgh@arm.com>
5413
5414 * config/aarch64/arm_neon.h
5415 (vcvtx_high_f32_f64): Fix parameters.
5416
f77e2d2d 54172013-09-16 Jan-Benedict Glaw <jbglaw@lug-owl.de>
a5bd71fa 5418 Uros Bizjak <ubizjak@gmail.com>
f77e2d2d 5419
a5bd71fa 5420 * config/alpha.c: Include tree-ssanames.h.
f77e2d2d 5421
36875e8f
RB
54222013-09-16 Richard Biener <rguenther@suse.de>
5423
5424 * tree-loop-distribution.c (enum rdg_dep_type): Add control_dd.
5425 (dot_rdg_1): Handle control_dd.
5426 (create_edge_for_control_dependence): New function.
5427 (create_rdg_edges): Add control dependences if asked for.
5428 (build_rdg): Likewise.
5429 (generate_loops_for_partition): If there are not necessary
5430 control stmts remove all their dependencies.
5431 (collect_condition_stmts, rdg_flag_loop_exits): Remove.
5432 (distribute_loop): Pass on control dependences.
5433 (tree_loop_distribution): Compute control dependences and remove
5434 restriction on number of loop nodes.
5435
96e24d49
JJ
54362013-09-16 Jakub Jelinek <jakub@redhat.com>
5437
5438 * ipa-prop.c (ipa_compute_jump_functions_for_edge): Return early
5439 for internal calls.
5440
40dbb05c
RS
54412013-09-16 Richard Sandiford <rdsandiford@googlemail.com>
5442
5443 * cse.c (try_const_anchors): Punt on CC modes.
5444
97191ad0
JBG
54452013-09-15 Jan-Benedict Glaw <jbglaw@lug-owl.de>
5446
5447 * config/vax/constraints.md (T): Add missing CONSTANT_P check.
5448
403a3fb7
JDA
54492013-09-14 John David Anglin <danglin@gcc.gnu.org>
5450
5451 PR target/58382
5452 * config/pa/pa.c (pa_expand_prologue): Change mode in gen_rtx_POST_INC
5453 calls to word_mode.
5454
fb5610fb
IS
54552013-09-14 Iain Sandoe <iain@codesourcery.com>
5456
5457 PR target/48094
5458 * config/darwin.c (darwin_objc2_section): Note if ObjC Metadata is
f77e2d2d
JBG
5459 seen.
5460 (darwin_objc1_section): Likewise.
fb5610fb
IS
5461 (darwin_file_end): Emit Image Info section when required.
5462
afdec9bd
JH
54632013-09-14 Jan Hubicka <jh@suse.cz>
5464
5465 * tree-into-ssa.c (gate_into_ssa): New.
5466 (pass_data_build_ssa): Use it.
5467 * cgraph.h (expand_thunk): Update prototype.
5468 * cgraphunit.c (analyze_function): Expand thunks early.
5469 (expand_thunk): Fix DECL_CONTEXT of reust_decl;
5470 build proper cgraph; set in_ssa_p; clear bogus TREE_ASM_WRITTEN;
5471 set lowered flag; do not add new function.
5472 (assemble_thunks_and_aliases): Update.
5473 * tree-ssa.c (gate_init_datastructures): New gate.
5474 (pass_data_init_datastructures): Use it.
5475
492d1e0a
IS
54762013-09-14 Iain Sandoe <iain@codesourcery.com>
5477
5478 PR target/58269
f77e2d2d 5479 * config/i386/i386.c (ix86_function_arg_regno_p): Make Darwin use the
492d1e0a
IS
5480 xmm register set described in the psABI.
5481
d13c6b70
EG
54822013-09-13 Evgeny Gavrin <e.gavrin@samsung.com>
5483
f77e2d2d
JBG
5484 * dwarf2out.c (should_emit_struct_debug): Add check
5485 for type_decl variable is not NULL.
d13c6b70 5486
88b0e79e
JC
54872013-09-13 Jacek Caban <jacek@codeweavers.com>
5488
5489 * config.gcc: Use new winnt-c.c target hooks
5490 * config/t-winnt: New file
5491 * config/winnt-c.c: New file
5492 * doc/tm.texi.in: Document new hook
5493 * doc/tm.texi: Regenerated
5494
19ba6aab
JH
54952013-09-13 Jan Hubicka <jh@suse.cz>
5496
5497 PR middle-end/58094
5498 * ipa-inline.c (check_callers): New function.
5499 (check_caller_edge): Remove.
5500 (want_inline_function_to_all_callers_p): Also permit alises that are
5501 called dirrectly.
5502 (inline_to_all_callers): Terminate the walk when devirtualization
5503 introduce new calls.
5504
a93c18c8
JH
55052013-09-13 Jan Hubicka <jh@suse.cz>
5506
5507 * ipa-inline-analysis.c (struct growth_data): Add node.
5508 (do_estimate_growth_1): Fix detection of recursion.
5509
f3b331d1
JJ
55102013-09-13 Jakub Jelinek <jakub@redhat.com>
5511
5512 PR tree-optimization/58392
5513 * tree-cfg.c (move_sese_region_to_fn): Rename loop variable
5514 to avoid shadowing of outer loop variable. If
5515 saved_cfun->has_simduid_loops or saved_cfun->has_force_vect_loops,
5516 replace_by_duplicate_decl simduid of loops that have it set and
5517 set dest_cfun->has_simduid_loops and/or
5518 dest_cfun->has_force_vect_loops.
5519 * omp-low.c (build_outer_var_ref): Call maybe_lookup_decl_in_outer_ctx
5520 instead of maybe_lookup_decl.
5521 * tree-inline.c (copy_loops): Change blocks_to_copy argument to id.
5522 Use id->blocks_to_copy instead of blocks_to_copy. Adjust recursive
5523 call. Copy over force_vect and copy and remap simduid. Set
5524 cfun->has_simduid_loops and/or cfun->has_force_vect_loops.
5525 (copy_cfg_body): Remove blocks_to_copy argument. Use
5526 id->blocks_to_copy instead of blocks_to_copy. Adjust copy_loops
5527 caller. Don't set cfun->has_simduid_loops and/or
5528 cfun->has_force_vect_loops here.
5529 (copy_body): Remove blocks_to_copy argument. Adjust copy_cfg_body
5530 caller.
5531 (expand_call_inline, tree_function_versioning): Adjust copy_body
5532 callers.
5533
ae6d0907
MJ
55342013-09-13 Martin Jambor <mjambor@suse.cz>
5535
5536 PR bootstrap/58388
5537 * ipa-prop.c (try_make_edge_direct_simple_call): Be less strict in
5538 the assert if the edge was a speculative one.
5539
2fd5894f
RB
55402013-09-13 Richard Biener <rguenther@suse.de>
5541
5542 * tree-data-ref.h (known_dependences_p): Move here ...
5543 * tree-loop-distribution.c (known_dependences_p): ... from here.
5544 (dump_rdg_component, debug_rdg_component): Remove.
5545 (dump_rdg): Adjust.
5546 (generate_loops_for_partition): Use gimple_uid instead of
5547 relying on matching stmt visit order.
5548 (rdg_build_partitions): Take starting stmt vector.
5549 (ldist_gen): Merge into ...
5550 (distribute_loop): ... this function. Do not compute starting
5551 vertices vector.
5552 * tree-cfg.c (gimple_duplicate_bb): Copy UID for PHIs.
5553
86eb4bd7
KT
55542013-09-13 Kyrylo Tkachov <kyrylo.tkachov@arm.com>
5555
5556 * config/arm/arm.md (arm_cmpsi_insn): Split rI alternative.
5557 Set type attribute correctly. Set predicable_short_it attribute.
5558 (cmpsi_shiftsi): Remove %? from output template.
5559
83a95546
RB
55602013-09-13 Richard Biener <rguenther@suse.de>
5561
5562 * tree-loop-distribution.c (struct rdg_component,
5563 rdg_defs_used_in_other_loops_p, free_rdg_components,
5564 rdg_build_components): Remove.
5565 (stmts_from_loop): Do not record virtual PHIs.
5566 (generate_loops_for_partition): Skip virtual PHIs.
5567 (build_rdg_partition_for_component): Rename to ...
5568 (build_rdg_partition_for_vertex): ... this and adjust.
5569 (rdg_build_partitions): Take a vector of starting vertices
5570 instead of components. Remove unnecessary leftover handling.
5571 (ldist_gen): Do not build components or record other stores.
5572 (distribute_loop): Do not distribute loops containing stmts
5573 with side-effects.
5574
c0ecf7f1
CB
55752013-09-13 Christian Bruel <christian.bruel@st.com>
5576
5577 PR target/58314
5578 * config/sh/sh.md (mov<mode>_reg_reg): Allow memory reloads.
5579
34f56efc
KT
55802013-09-13 Kai Tietz <ktietz@redhat.com>
5581
740f2bf3
KT
5582 * config.gcc: Separate cases for mingw and cygwin targets,
5583 and add 64-bit cygwin target case.
5584
34f56efc
KT
5585 * config/i386/winnt-cxx.c (i386_pe_type_dllexport_p): Don't
5586 dll-export inline-functions.
5587 * config/i386/winnt.c (i386_pe_determine_dllexport_p): Likewise.
5588
5589
55902013-09-13 Jeff Law <law@redhat.com>
823b5d6b
JL
5591
5592 PR middle-end/58387
5593 Revert:
5594 2013-09-06 Jeff Law <law@redhat.com>
5595
5596 * tree-ssa-dom.c (cprop_into_successor_phis): Also propagate
5597 edge implied equivalences into successor phis.
5598
aadb52ad
DD
55992013-09-12 DJ Delorie <dj@redhat.com>
5600
5601 * config/rl78/rl78-virt.md: Change from | to \; for asm line
5602 separators.
5603
90ae7019
BM
56042013-09-12 Brooks Moses <bmoses@google.com>
5605
5606 PR driver/42955
5607 * Makefile.in: Do not install driver binaries in $(target)/bin.
5608
17b2557c
DD
56092013-09-12 DJ Delorie <dj@redhat.com>
5610
c5215a9b
DD
5611 * config/rl78/rl78.opt (mrelax): New.
5612 * config/rl78/rl78.h (ASM_SPEC): New, pass on -mrelax to gas.
5613 * config/rl78/rl78.h (LINK_SPEC): New, pass on -mrelax to ld.
5614
17b2557c
DD
5615 * config/rl78/rl78.c (rl78_expand_prologue): Use AX to copy
5616 between SP and FP.
5617 (rl78_expand_epilogue): Likewise.
5618
ecafcf05
VM
56192013-09-12 Vladimir Makarov <vmakarov@redhat.com>
5620
5621 PR middle-end/58335
84f6d9fa 5622 * lra-eliminations.c (remove_reg_equal_offset_note): New.
ecafcf05
VM
5623 (eliminate_regs_in_insn): Rewrite frame pointer to hard frame
5624 pointer elimination with using remove_reg_equal_offset_note.
5625
8369f38a
DD
56262013-09-12 DJ Delorie <dj@redhat.com>
5627
f6a83b4a
DD
5628 * config/msp430/: New port.
5629 * config.gcc (msp430): Added.
5630 * doc/invoke.texi: Document MSP430 options.
5631 * doc/install.texi: Document msp430-elf
5632 * doc/md.texi: Document msp430-elf
5633 * doc/contrib.texi: Document msp430-elf
5634
8369f38a
DD
5635 * cfgexpand.c (expand_debug_expr): Avoid sign-extending SImode to
5636 PSImode.
5637
a854f856
MJ
56382013-09-12 Martin Jambor <mjambor@suse.cz>
5639
5640 PR ipa/58389
5641 * ipa-prop.c (remove_described_reference): Give up if the edge in the
5642 reference descriptor is NULL.
5643 (ipa_edge_removal_hook): If owning a reference descriptor, set its
5644 edge to NULL.
5645
b12ebd96
AM
56462013-09-12 Andrew MacLeod <amacleod@redhat.com>
5647
5648 * tree-flow.h (FREE_SSANAMES): Move to tree-ssanames.c
5649 (SSANAMES, MODIFIED_NORETURN_CALLS, DEFAULT_DEFS, ptr_info_def,
5650 num_ssa_names, ssa_name): Move to tree-ssanames.h + prototypes.
5651 * tree-flow-inline.h (make_ssa_name, copy_ssa_name, duplicate_ssa_name,
5652 make_temp_ssa_name): move to tree-ssanames.h
5653 * tree-ssa-alias.h: Move prototype.
5654 * tree-ssa.h: Include tree-ssanames.h.
5655 * tree-ssanames.c (FREE_SSANAMES): Move to here.
5656 * tree-ssanames.h: New. Move items from tree-flow*.h
5657 * Makefile.in (tree-ssanames.h): Add to tree-ssanames.o and GTFILES.
5658
e1fb4ad3
RB
56592013-09-12 Richard Biener <rguenther@suse.de>
5660
5661 PR tree-optimization/58404
5662 * tree-ssa-forwprop.c (forward_propagate_addr_expr_1): Also
5663 propagate non-invariant addresses into dereferences wrapped
5664 in component references.
5665
154c9188
RB
56662013-09-12 Richard Biener <rguenther@suse.de>
5667
5668 PR tree-optimization/58402
5669 * passes.def: Move pass_late_warn_uninitialized later.
5670
7a300452
AM
56712013-09-12 Andrew MacLeod <amacleod@redhat.com>
5672
5673 * tree-ssa.h: New. Move content from tree-flow.h and
5674 tree-flow-inline.h.
5675 * tree-flow.h (_edge_var_map, edge_var_map_vector): Move to tree-ssa.h.
5676 Move prototypes belonging to tree-ssa.c.
f77e2d2d 5677 * tree-flow-inline.h (redirect_edge_var_map_def,
7a300452
AM
5678 redirect_edge_var_map_result, redirect_edge_var_map_location): Move to
5679 tree-ssa.h.
5680 * gimple.h: Adjust prototypes.
f77e2d2d
JBG
5681 * tree-ssa.c (useless_type_conversion_p, types_compatible_p): Move
5682 to...
5683 * gimple.c (useless_type_conversion_p, types_compatible_p): Here.
7a300452
AM
5684 * tree.h: Move prototype to tree-ssa.h.
5685 * gengtype.c (open_base_files): Replace tree-flow.h with tree-ssa.h.
5686 * Makefile.in: (TREE_SSA_H, TREE_FLOW_H): Adjust dependencies.
5687 * alias.c, asan.c, builtins.c, calls.c, cfgexpand.c, cfghooks.c,
5688 cfgloop.c, cfgloopmanip.c, cgraph.c, cgraphbuild.c, cgraphclones.c,
5689 cgraphunit.c, dse.c, except.c, expr.c, final.c, fold-const.c,
5690 ggc-page.c, gimple-fold.c, gimple-iterator.c, gimple-low.c,
5691 gimple-pretty-print.c, gimple-ssa-strength-reduction.c,
5692 gimple-streamer-in.c, gimple-streamer-out.c, gimple.c, gimplify.c,
f77e2d2d
JBG
5693 graphite-blocking.c, graphite-clast-to-gimple.c,
5694 graphite-dependences.c, graphite-interchange.c,
5695 graphite-optimize-isl.c, graphite-poly.c, graphite-scop-detection.c,
5696 graphite-sese-to-poly.c, graphite.c, ipa-cp.c, ipa-inline-analysis.c,
5697 ipa-inline-transform.c, ipa-inline.c, ipa-prop.c, ipa-pure-const.c,
5698 ipa-reference.c, ipa-split.c, ipa-utils.c,
7a300452
AM
5699 loop-init.c, lto-cgraph.c, lto-section-in.c, lto-section-out.c,
5700 lto-streamer-in.c, lto-streamer-out.c, lto-streamer.c, omp-low.c,
5701 passes.c, predict.c, print-tree.c, profile.c, sese.c, targhooks.c,
5702 tracer.c, trans-mem.c, tree-call-cdce.c, tree-cfg.c, tree-cfgcleanup.c,
5703 tree-chrec.c, tree-complex.c, tree-data-ref.c, tree-dfa.c, tree-eh.c,
5704 tree-emutls.c, tree-if-conv.c, tree-inline.c, tree-into-ssa.c,
5705 tree-loop-distribution.c, tree-mudflap.c, tree-nested.c, tree-nrv.c,
5706 tree-object-size.c, tree-optimize.c, tree-outof-ssa.c, tree-parloops.c,
5707 tree-phinodes.c, tree-predcom.c, tree-pretty-print.c, tree-profile.c,
5708 tree-scalar-evolution.c, tree-sra.c, tree-ssa*.c, tree-stdarg.c,
5709 tree-streamer-in.c, tree-switch-conversion.c, tree-tailcall.c,
5710 tree-vect-data-refs.c, tree-vect-generic.c, tree-vect-loop-manip.c,
5711 tree-vect-loop.c, tree-vect-patterns.c, tree-vect-slp.c,
5712 tree-vect-stmts.c, tree-vectorizer.c, tree-vrp.c, tsan.c,
5713 value-prof.c, var-tracking.c,
5714 varpool.c, vtable-verify.c: Replace tree-flow.h with tree-ssa.h
5715
97463b2b
RB
57162013-09-12 Richard Biener <rguenther@suse.de>
5717
5718 PR tree-optimization/58396
5719 * tree-loop-distribution.c (create_rdg_edges): Free unused DDRs.
5720 (build_rdg): Take a loop-nest parameter, fix memleaks.
5721 (distribute_loop): Compute loop-nest here and pass it to build_rdg.
5722
80ecd89d
YR
57232013-09-12 Yuri Rumyantsev <ysrumyan@gmail.com>
5724
5725 * config/i386/x86-tune.def: Turn on X86_TUNE_AVOID_MEM_OPND_FOR_CMOVE
5726 for SLM.
5727
893e85fa
CM
57282013-09-12 Cameron McInally <cameron.mcinally@nyu.edu>
5729
5730 * doc/extend.texi: Fix errors in x86 FMA builtin naming.
5731 The FMA instruction names should have a 'v' prefix.
5732
174ec470
RB
57332013-09-12 Richard Biener <rguenther@suse.de>
5734
5735 * tree-loop-distribution.c (dot_rdg_1): Make graph prettier.
5736 (dot_rdg): Use popen instead of system in optional code.
5737 (remaining_stmts, upstream_mem_writes): Remove global bitmaps.
5738 (already_processed_vertex_p): Adjust.
5739 (has_anti_or_output_dependence, predecessor_has_mem_write,
5740 mark_nodes_having_upstream_mem_writes, has_upstream_mem_writes,
5741 rdg_flag_uses): Remove.
5742 (rdg_flag_vertex): Simplify.
5743 (rdg_flag_vertex_and_dependent): Rely on a correct RDG and
5744 remove recursion.
5745 (build_rdg_partition_for_component): Process the first vertex
5746 of a component only.
5747 (ldist_gen): Do not compute remaining_stmts or upstream_mem_writes.
5748
38c89763
AM
57492013-09-12 Alan Modra <amodra@gmail.com>
5750
5751 * config/rs6000/rs6000.c (toc_relative_expr_p): Use add_cint_operand.
5752
3e3d9d17
DD
57532013-09-11 DJ Delorie <dj@redhat.com>
5754 Nick Clifton <nickc@redhat.com>
5755
5756 * config/rl78/predicates.md (rl78_cmp_operator_signed): New.
5757 (rl78_stack_based_mem): New.
5758 * config/rl78/constraints.md (Iv08): New.
5759 (Iv16): New.
5760 (Iv24): New.
5761 (Is09): New.
5762 (Is17): New.
5763 (Is25): New.
5764 (ISsi): New.
5765 (IShi): New.
5766 (ISqi): New.
5767 * config/rl78/rl78-expand.md (movqi): Reject more SUBREG operands.
5768 (movhi): Likewise.
5769 (movsi): Change from expand to insn-and-split.
5770 (ashrsi3): Clobber AX.
5771 (lshrsi3): New.
5772 (ashlsi3): New.
5773 (cbranchsi4): New.
5774 * config/rl78/rl78.md (CC_REG): Fix.
5775 (addsi3): Allow memory and immediate operands.
5776 (addsi3_internal): Split into...
5777 (addsi3_internal_virt): ...new, and ...
5778 (addsi3_internal_real): ...new.
5779 (subsi): New.
5780 (subsi3_internal_virt): New.
5781 (subsi3_internal_real): New.
5782 (mulsi3): Add memory operand.
5783 (mulsi3_rl78): Likewise.
5784 (mulsi3_g13): Likewise.
5785 * config/rl78/rl78-real.md (cbranchqi4_real_signed): New.
5786 (cbranchqi4_real): Add more constraint options.
5787 (cbranchhi4_real): Expand pattern.
5788 (cbranchhi4_real_signed): New.
5789 (cbranchhi4_real_inverted): New.
5790 (cbranchsi4_real_lt): New.
5791 (cbranchsi4_real_ge): New.
5792 (cbranchsi4_real_signed): New.
5793 (cbranchsi4_real): New.
5794 (peephole2): New.
893e85fa
CM
5795 * config/rl78/rl78-virt.md (ashrsi3_virt): Add custom cases for
5796 constant shifts.
3e3d9d17
DD
5797 (lshrsi3_virt): Likewise.
5798 (ashlsi3_virt): Likewise.
5799 (cbranchqi4_virt_signed): New.
5800 (cbranchhi4_virt_signed): New.
5801 (cbranchsi4_virt): New.
5802 * config/rl78/rl78.c: Whitespace fixes throughout.
5803 (move_elim_pass): New.
5804 (pass_data_rl78_move_elim): New.
5805 (pass_rl78_move_elim): New.
5806 (make_pass_rl78_move_elim): New.
5807 (rl78_devirt_info): Run devirt earlier.
5808 (rl78_move_elim_info): New.
5809 (rl78_asm_file_start): Register it.
5810 (rl78_split_movsi): New.
5811 (rl78_as_legitimate_address): Allow virtual base registers when
5812 appropriate.
5813 (rl78_addr_space_convert): Remove spurious debug stuff.
5814 (rl78_print_operand_1): Add z,s,S,r,E modifiers.
5815 (rl78_print_operand): More cases for not printing '#'.
5816 (rl78_expand_compare): Remove most of the logic.
5817 (content_memory): New.
5818 (clear_content_memory): New.
5819 (get_content_index): New.
5820 (get_content_name): New.
5821 (display_content_memory): New.
5822 (update_content): New.
5823 (record_content): New.
5824 (already_contains): New.
5825 (insn_ok_now): Re-recog insns with virtual registers.
5826 (add_postponed_content_update): New.
5827 (process_postponed_content_update): New.
5828 (gen_and_emit_move): New.
893e85fa
CM
5829 (transcode_memory_rtx): Record new location content.
5830 Use gen_and_emit_move.
3e3d9d17
DD
5831 (force_into_acc): New.
5832 (move_to_acc): Use gen_and_emit_move.
5833 (move_from_acc): Likewise.
5834 (move_acc_to_reg): Likewise.
5835 (move_to_x): Likewise.
5836 (move_to_hl): Likewise.
5837 (move_to_de): Likewise.
5838 (rl78_alloc_physical_registers_op1): Record location content.
5839 (has_constraint): New.
5840 (rl78_alloc_physical_registers_op2): Record location content.
5841 Optimize use of HL.
5842 (rl78_alloc_physical_registers_ro1): Likewise.
5843 (rl78_alloc_physical_registers_cmp): Likewise.
5844 (rl78_alloc_physical_registers_umul): Likewise.
5845 (rl78_alloc_address_registers_macax): New.
5846 (rl78_alloc_physical_registers): Initialize and set location
5847 content memory as needed.
5848 (rl78_reorg): Make sure split2 is called.
5849 (rl78_rtx_costs): New.
5850
088845a5
RS
58512013-09-11 Richard Sandiford <rdsandiford@googlemail.com>
5852
5853 * simplify-rtx.c (simplify_unary_operation_1): Use simplify_gen_binary
5854 for (not (neg ...)) and (neg (not ...)) cases.
5855
5c2961cf
RB
58562013-09-11 Richard Biener <rguenther@suse.de>
5857
5858 PR middle-end/58377
5859 * passes.def: Split critical edges before late uninit warning passes.
5860 * tree-cfg.c (pass_split_crit_edges): Implement clone method.
5861
427e6a14
JJ
58622013-09-11 Jakub Jelinek <jakub@redhat.com>
5863
5864 PR tree-optimization/58385
5865 * fold-const.c (build_range_check): If both low and high are NULL,
5866 use omit_one_operand_loc to preserve exp side-effects.
5867
df45c0cc
KT
58682013-09-11 Kyrylo Tkachov <kyrylo.tkachov@arm.com>
5869
5870 * config/arm/arm.md (arm_shiftsi3): New alternative l/l/M.
5871
80ab0b19
RB
58722013-09-11 Richard Biener <rguenther@suse.de>
5873
5874 * tree-data-ref.c (dump_rdg_vertex, debug_rdg_vertex,
5875 dump_rdg_component, debug_rdg_component, dump_rdg, debug_rdg,
5876 dot_rdg_1, dot_rdg, rdg_vertex_for_stmt, create_rdg_edge_for_ddr,
5877 create_rdg_edges_for_scalar, create_rdg_edges, create_rdg_vertices,
5878 stmts_from_loop, known_dependences_p, build_empty_rdg,
5879 build_rdg, free_rdg, rdg_defs_used_in_other_loops_p): Move ...
5880 * tree-loop-distribution.c: ... here.
5881 * tree-data-ref.h (struct rdg_vertex, RDGV_STMT, RDGV_DATAREFS,
5882 RDGV_HAS_MEM_WRITE, RDGV_HAS_MEM_READS, RDG_STMT, RDG_DATAREFS,
5883 RDG_MEM_WRITE_STMT, RDG_MEM_READS_STMT, enum rdg_dep_type,
5884 struct rdg_edge, RDGE_TYPE, RDGE_LEVEL, RDGE_RELATION): Move ...
5885 * tree-loop-distribution.c: ... here.
5886 * tree-loop-distribution.c: Include gimple-pretty-print.h.
5887 (struct partition_s): Add loops member.
5888 (partition_alloc, partition_free, rdg_flag_uses, rdg_flag_vertex,
5889 rdg_flag_vertex_and_dependent, rdg_flag_loop_exits,
5890 build_rdg_partition_for_component, rdg_build_partitions): Adjust.
5891
85a77221
AI
58922013-09-11 Alexander Ivchenko <alexander.ivchenko@intel.com>
5893 Maxim Kuznetsov <maxim.kuznetsov@intel.com>
5894 Sergey Lega <sergey.s.lega@intel.com>
5895 Anna Tikhonova <anna.tikhonova@intel.com>
5896 Ilya Tocar <ilya.tocar@intel.com>
5897 Andrey Turetskiy <andrey.turetskiy@intel.com>
5898 Ilya Verbin <ilya.verbin@intel.com>
5899 Kirill Yukhin <kirill.yukhin@intel.com>
5900 Michael Zolotukhin <michael.v.zolotukhin@intel.com>
5901
5902 * config/i386/constraints.md (k): New.
5903 (Yk): Ditto.
5904 * config/i386/i386.c (const regclass_map): Add new mask registers.
5905 (dbx_register_map): Ditto.
5906 (dbx64_register_map): Ditto.
5907 (svr4_dbx_register_map): Ditto.
5908 (ix86_conditional_register_usage): Squash mask registers if AVX512F is
5909 disabled.
5910 (ix86_preferred_reload_class): Disable constants for mask registers.
5911 (ix86_secondary_reload): Do spill of mask register using 32-bit insn.
5912 (ix86_hard_regno_mode_ok): Support new mask registers.
5913 (x86_order_regs_for_local_alloc): Ditto.
5914 * config/i386/i386.h (FIRST_PSEUDO_REGISTER): Update.
5915 (FIXED_REGISTERS): Add new mask registers.
5916 (CALL_USED_REGISTERS): Ditto.
5917 (REG_ALLOC_ORDER): Ditto.
5918 (VALID_MASK_REG_MODE): New.
5919 (FIRST_MASK_REG): Ditto.
5920 (LAST_MASK_REG): Ditto.
5921 (reg_class): Add MASK_EVEX_REGS, MASK_REGS.
5922 (MAYBE_MASK_CLASS_P): New.
5923 (REG_CLASS_NAMES): Add MASK_EVEX_REGS, MASK_REGS.
5924 (REG_CLASS_CONTENTS): Ditto.
5925 (MASK_REGNO_P): New.
5926 (ANY_MASK_REG_P): Ditto.
5927 (HI_REGISTER_NAMES): Add new mask registers.
893e85fa
CM
5928 * config/i386/i386.md (MASK0_REG, MASK1_REG, MASK2_REG, MASK3_REG,
5929 MASK4_REG, MASK5_REG, MASK6_REG, MASK7_REG): Constants for new
5930 mask registers.
85a77221
AI
5931 (attribute "type"): Add mskmov, msklog.
5932 (attribute "length_immediate"): Support them.
5933 (attribute "memory"): Ditto.
5934 (attribute "prefix_0f"): Ditto.
5935 (*movhi_internal): Support new mask registers.
5936 (*movqi_internal): Ditto.
5937 (define_split): Split out clobber pattern is a logic
5938 insn on mask registers.
5939 (*k<logic><mode>): New.
5940 (*andhi_1): Extend to support mask regs.
5941 (*andqi_1): Extend to support mask regs.
5942 (kandn<mode>): New.
5943 (define_split): Split and-not to and and not if operands
5944 are not mask regs.
5945 (*<code><mode>_1): Separate HI mode to new pattern...
5946 (*<code>hi_1): This.
5947 (*<code>qi_1): Extend to support mask regs.
5948 (kxnor<mode>): New.
5949 (kortestzhi): Ditto.
5950 (kortestchi): Ditto.
5951 (kunpckhi): Ditto.
5952 (*one_cmpl<mode>2_1): Remove HImode and handle it...
5953 (*one_cmplhi2_1): ...Here, now with mask registers support.
5954 (*one_cmplqi2_1): Support new mask registers.
893e85fa
CM
5955 (HI/QImode arithmetics splitter): Don't split if mask registers
5956 are used.
85a77221
AI
5957 (HI/QImode not splitter): Ditto.
5958 * config/i386/predicated.md (mask_reg_operand): New.
5959 (general_reg_operand): Ditto.
5960
7e570821
AI
59612013-09-11 Alexander Ivchenko <alexander.ivchenko@intel.com>
5962
5963 * doc/invoke.texi: Document fxsr, xsave and xsaveopt options.
5964 * doc/extend.texi: Document fxsr, xsave and xsaveopt builtins.
5965
9e1376e9
JL
59662013-09-10 Jeff Law <law@redhat.com>
5967
5968 PR tree-optimization/58380
5969 * tree-ssa-threadupdate.c (thread_block): Recognize another case
5970 of threading through a buried loop header.
5971
5972 * tree-ssa-threadedge.c (thread_around_empty_blocks): Correct
5973 return value for single successor case.
5974
5bbcb888
JH
59752013-09-10 Jan Hubicka <jh@suse.cz>
5976
5977 * ipa-devirt.c (ipa_devirt): Enable with LTO.
5978
d742ff4b
RE
59792013-09-10 Richard Earnshaw <rearnsha@arm.com>
5980
5981 PR target/58361
5982 * arm/vfp.md (combine_vcvt_f32_<FCVTI32typename>): Fix pattern to
5983 support conditional execution.
5984 (combine_vcvt_f64_<FCVTI32typename>): Likewise.
5985
b0681c9e
VM
59862013-09-10 Vladimir Makarov <vmakarov@redhat.com>
5987
5988 * lra.c (lra): Clear lra_optional_reload_pseudos before every
5989 constraint pass.
f77e2d2d
JBG
5990 * lra-constraints.c (curr_insn_transform): Switch on optional reloads.
5991 Check destination too to check move insn.
b0681c9e 5992 (undo_optional_reloads): Add check that the original peudo did not
f77e2d2d
JBG
5993 changed its allocation and the optional reload was inherited on last
5994 inheritance pass. Break loop after deciding to keep optional reload.
b0681c9e
VM
5995 (lra_undo_inheritance): Add check that inherited pseudo still in
5996 memory.
5997
c9b1c957
JG
59982013-09-10 James Greenhalgh <james.greenhalgh@arm.com>
5999
6000 * config/aarch64/aarch64.md (generic_sched): New.
6001 * config/aarch64/aarch64-generic.md (load): Make conditional
6002 on generic_sched attribute.
6003 (nonload): Likewise.
6004
82d618d3
JH
60052013-09-10 Jan Hubicka <jh@suse.cz>
6006
6007 * lto-cgraph.c: Include ipa-utils.h.
6008 (compute_ltrans_boundary): Also add possible targets into the boundary.
6009
668b6894
JH
60102013-09-10 Jan Hubicka <jh@suse.cz>
6011
6012 * gimple-fold.c (gimple_get_virt_method_for_binfo): Pass real
6013 VAR_DECL of vtable rather than full expression.
6014
4f90d3e0
JH
60152013-09-10 Jan Hubicka <jh@suse.cz>
6016 Paolo Carlini <paolo.carlini@oracle.com>
6017
6018 * cgraphunit.c (analyze_functions): Save input_location, set it
6019 to UNKNOWN_LOCATION and restore it at the end.
6020
39e87baf
MJ
60212013-09-10 Martin Jambor <mjambor@suse.cz>
6022
6023 * ipa-cp.c (propagate_constants_topo): Do not ignore SCC
6024 represented by a thunk.
6025
581aedec
JL
60262013-09-10 Jeff Law <law@redhat.com>
6027
6028 PR tree-optimization/58343
6029 * tree-ssa-threadupdate.c (thread_block): Identify and disable
6030 jump threading requests through loop headers buried in the middle
6031 of a jump threading path.
6032
6033 * tree-ssa-threadedge.c (thread_around_empty_blocks): Fix thinko
6034 in return value/type.
6035
84cf4ab6
JJ
60362013-09-10 Jakub Jelinek <jakub@redhat.com>
6037
6038 PR rtl-optimization/58365
6039 * cfgcleanup.c (merge_memattrs): Also clear MEM_READONLY_P
6040 resp. MEM_NOTRAP_P if they differ, or set MEM_VOLATILE_P if
6041 it differs.
6042
447dd906
RB
60432013-09-10 Richard Biener <rguenther@suse.de>
6044
6045 * tree-data-ref.h (build_rdg): Drop all parameters but loop.
6046 * tree-data-ref.c (create_rdg_vertices): Collect all data
6047 references, signal failure to the caller, use data-ref API.
893e85fa
CM
6048 (build_rdg): Compute data references only once. Maintain lifetime
6049 of data references and data dependences from within RDG.
447dd906
RB
6050 (free_rdg): Free dependence relations.
6051 * tree-loop-distribution.c (rdg_flag_uses): Drop weird code
6052 inventing extra dependences.
6053 (distribute_loop): Update for RDG API changes.
6054
f14a7324
KT
60552013-09-10 Kai Tietz <ktietz@redhat.com>
6056
6057 * doc/invoke.texi (fms-extensions): Document changed
6058 behavior for ms-abi targets.
6059 * config/i386/i386.c (ix86_option_override_internal):
6060 Set default value of option -fms-extension for ms-abi targets.
6061
ff131f26
MZ
60622013-09-10 Michael Zolotukhin <michael.v.zolotukhin@gmail.com>
6063
6064 * config/i386/i386.c (ix86_expand_movmem): Fix epilogue generation.
6065
33115a2c
AM
60662013-09-10 Alan Modra <amodra@gmail.com>
6067
6068 PR target/58330
6069 * config/rs6000/rs6000.md (bswapdi2_64bit): Disable for volatile mems.
6070
b6872a13
AM
60712013-09-10 Alan Modra <amodra@gmail.com>
6072
6073 * config/rs6000/predicates.md (add_cint_operand): New.
6074 (reg_or_add_cint_operand, small_toc_ref): Use add_cint_operand.
6075 * config/rs6000/rs6000.md (largetoc_high_plus): Restrict offset
6076 using add_cint_operand.
6077 (largetoc_high_plus_aix): Likewise.
6078
28fd0ba2
JJ
60792013-09-09 Jakub Jelinek <jakub@redhat.com>
6080
6081 PR tree-optimization/58364
6082 * tree-ssa-reassoc.c (init_range_entry): For BIT_NOT_EXPR on
6083 BOOLEAN_TYPE, only invert in_p and continue with arg0 if
6084 the current range can't be an unconditional true or false.
6085
552d99fe
JG
60862013-09-09 James Greenhalgh <james.greenhalgh@arm.com>
6087
6088 * config/aarch64/arm_neon.h (vrsqrte_f64): Fix parameter type.
6089
f43245d1
UB
60902013-09-09 Uros Bizjak <ubizjak@gmail.com>
6091
6092 * ipa-prop.c (ipa_modify_call_arguments): Initialize deref_align.
6093
c9b0866a
PC
60942013-09-09 Paolo Carlini <paolo.carlini@oracle.com>
6095
6096 PR c++/43452
6097 * doc/invoke.texi (-Wdelete-incomplete): Document it.
6098
78d8b9f0
IB
60992013-09-09 Ian Bolton <ian.bolton@arm.com>
6100
6101 * config/aarch64/aarch64.c (aarch64_preferred_reload_class): Return
6102 NO_REGS for immediate that can't be moved directly into FP_REGS.
6103
274b2532
KT
61042013-09-09 Kyrylo Tkachov <kyrylo.tkachov@arm.com>
6105
6106 * config/aarch64/aarch64.c (aarch64_select_cc_mode): Return CC_SWP for
6107 comparison with negated operand.
c9b0866a
PC
6108 * config/aarch64/aarch64.md (compare_neg<mode>): Match canonical
6109 RTL form.
274b2532 6110
467a3558
RB
61112013-09-09 Richard Biener <rguenther@suse.de>
6112
6113 PR middle-end/58326
6114 * cfgloopmanip.c (fix_bb_placements): When fixing the placement
6115 of a subloop record all its block as affecting loop-closed SSA form.
6116
088c5368
RS
61172013-09-09 Richard Sandiford <rdsandiford@googlemail.com>
6118
6119 * expmed.c (lshift_value): Take an unsigned HOST_WIDE_INT instead
6120 of an rtx/bitpos pair.
6121 (store_fixed_bit_field): Update accordingly.
6122
69db2d57
RS
61232013-09-09 Richard Sandiford <rdsandiford@googlemail.com>
6124
6125 * asan.c (asan_emit_stack_protection): Use gen_int_mode instead of
6126 GEN_INT.
6127 * builtins.c (expand_errno_check): Likewise.
6128 * dwarf2cfi.c (init_return_column_size): Likewise.
6129 * except.c (sjlj_mark_call_sites): Likewise.
6130 * expr.c (move_by_pieces_1, store_by_pieces_2): Likewise.
6131 * lra-constraints.c (emit_inc): Likewise.
6132 * ree.c (combine_set_extension): Likewise.
6133 * regmove.c (fixup_match_2): Likewise.
6134 * reload1.c (inc_for_reload): Likewise.
6135
69a59f0f
RS
61362013-09-09 Richard Sandiford <rdsandiford@googlemail.com>
6137
6138 * combine.c (simplify_set, expand_field_assignment, extract_left_shift)
6139 (force_to_mode, simplify_shift_const_1, simplify_comparison):
6140 Use gen_int_mode with the mode of the associated simplify_* call.
6141 * explow.c (probe_stack_range, anti_adjust_stack_and_probe): Likewise.
6142 * expmed.c (expand_shift_1): Likewise.
6143 * function.c (instantiate_virtual_regs_in_insn): Likewise.
6144 * loop-iv.c (iv_number_of_iterations): Likewise.
6145 * loop-unroll.c (unroll_loop_runtime_iterations): Likewise.
6146 * simplify-rtx.c (simplify_binary_operation_1): Likewise.
6147
2f1cd2eb
RS
61482013-09-09 Richard Sandiford <rdsandiford@googlemail.com>
6149
6150 * asan.c (asan_clear_shadow): Use gen_int_mode with the mode
6151 of the associated expand_* call.
6152 (asan_emit_stack_protection): Likewise.
6153 * builtins.c (round_trampoline_addr): Likewise.
6154 * explow.c (allocate_dynamic_stack_space, probe_stack_range): Likewise.
6155 * expmed.c (expand_smod_pow2, expand_sdiv_pow2, expand_divmod)
6156 (emit_store_flag): Likewise.
6157 * expr.c (emit_move_resolve_push, push_block, emit_single_push_insn_1)
6158 (emit_push_insn, optimize_bitfield_assignment_op, expand_expr_real_1):
6159 Likewise.
6160 * function.c (instantiate_virtual_regs_in_insn): Likewise.
6161 * ifcvt.c (noce_try_store_flag_constants): Likewise.
6162 * loop-unroll.c (unroll_loop_runtime_iterations): Likewise.
6163 * modulo-sched.c (generate_prolog_epilog): Likewise.
6164 * optabs.c (expand_binop, widen_leading, expand_doubleword_clz)
6165 (expand_ctz, expand_ffs, expand_unop): Likewise.
6166
4789c0ce
RS
61672013-09-09 Richard Sandiford <rdsandiford@googlemail.com>
6168
6169 * alias.c (addr_side_effect_eval): Use gen_int_mode with the mode
6170 of the associated gen_rtx_* call.
6171 * caller-save.c (init_caller_save): Likewise.
6172 * combine.c (find_split_point, make_extraction): Likewise.
6173 (make_compound_operation): Likewise.
6174 * dwarf2out.c (mem_loc_descriptor): Likewise.
6175 * explow.c (plus_constant, probe_stack_range): Likewise.
6176 * expmed.c (expand_mult_const): Likewise.
6177 * expr.c (emit_single_push_insn_1, do_tablejump): Likewise.
6178 * reload1.c (init_reload): Likewise.
6179 * valtrack.c (cleanup_auto_inc_dec): Likewise.
6180 * var-tracking.c (adjust_mems): Likewise.
6181 * modulo-sched.c (sms_schedule): Likewise, but use gen_rtx_GT
6182 rather than gen_rtx_fmt_ee.
6183
a811f7cb
JH
61842013-09-09 Jan Hubicka <jh@suse.cz>
6185
6186 PR middle-end/58294
6187 * value-prof.c (gimple_ic): Copy also abnormal edges.
6188
dcad1dd3
RS
61892013-09-09 Richard Sandiford <rdsandiford@googlemail.com>
6190
6191 * asan.c (asan_shadow_cst): Use gen_int_mode.
6192
daf5c770
JH
61932013-09-08 Jan Hubicka <jh@suse.cz>
6194
6195 * ipa-profile.c: Add toplevel comment.
6196 (ipa_propagate_frequency_1): Be more conservative when profile is read.
6197 (contains_hot_call_p): New function.
6198 (ipa_propagate_frequency): Set frequencies based on counts when
6199 profile is read.
6200 * predict.c (compute_function_frequency): Use PROFILE_READ gueard for
6201 profile; do not tamper with profile after inlining if it is read.
6202
a12cd2db
JH
62032013-09-08 Jan Hubicka <jh@suse.cz>
6204
6205 * ipa-prop.c (try_make_edge_direct_simple_call): Do not special case
6206 speculative edges.
6207
856a1588
JH
62082013-09-08 Jan Hubicka <jh@suse.cz>
6209
6210 * ipa.c (walk_polymorphic_call_targets): Fix redirection before IPA
6211 summary generation.
6212
4f4b0b73
JL
62132013-09-08 Jeff Law <law@redhat.com>
6214
6215 PR bootstrap/58340
6216 * tree-ssa-threadedge.c (thread_across_edge): Fix initialization
6217 of 'found'.
6218
cc4029ee
AK
62192013-09-08 Andi Kleen <ak@linux.intel.com>
6220
6221 * tree-inline.c (estimate_num_insns): Limit asm cost to 1000.
6222
7b395ddd
JH
62232013-09-08 Jan Hubicka <jh@suse.cz>
6224
6225 * ipa.c (walk_polymorphic_call_targets): Fix inliner summary update.
6226
c34c46dd
RS
62272013-09-08 Richard Sandiford <rdsandiford@googlemail.com>
6228
6229 * ira.c (update_equiv_regs): Only call set_paradoxical_subreg
6230 for non-debug insns.
6231 * lra.c (new_insn_reg): Take the containing insn as a parameter.
6232 Only modify lra_reg_info[].biggest_mode if it's non-debug insn.
6233 (collect_non_operand_hard_regs, add_regs_to_insn_regno_info): Update
6234 accordingly.
6235
3462aa02
JH
62362013-09-08 Jan Hubicka <jh@suse.cz>
6237
6238 * cgraphunit.c (walk_polymorphic_call_targets): Permit 0 possible
6239 targets and devirtualize to BUILT_IN_UNREACHABLE.
6240 * timevar.def (TV_IPA_UNREACHABLE): New timevar.
6241 * ipa.c (walk_polymorphic_call_targets): New function.
6242 (symtab_remove_unreachable_nodes): Use it; do not keep all virtual
6243 functions; use the new timevar.
6244 * ipa-devirt.c (maybe_record_node): Do not insert static nodes that
6245 was removed from the program.
6246 (record_binfo): If BINFO corresponds to an anonymous namespace, we may
6247 not consider it in the walk when its vtable is dead.
6248 (possible_polymorphic_call_targets_1): Pass anonymous flag to
6249 record_binfo.
6250 (devirt_variable_node_removal_hook): New function.
6251 (possible_polymorphic_call_targets): Also register
6252 devirt_variable_node_removal_hook.
6253 (ipa_devirt): Do not do non-speculative devirtualization.
6254 (gate_ipa_devirt): One execute if devirtualizing speculatively.
6255
26e5b0fd
JH
62562013-09-08 Jan Hubicka <jh@suse.cz>
6257
6258 * cgraph.h (varpool_node_hook, varpool_node_hook_list,
6259 varpool_add_node_removal_hook, varpool_add_variable_insertion_hook,
6260 varpool_remove_variable_insertion_hook): Declare.
6261 * varpool.c (varpool_node_hook_list): New structure.
6262 (first_varpool_node_removal_hook,
6263 first_varpool_variable_insertion_hook): New variables.
6264 (varpool_add_node_removal_hook, varpool_remove_node_removal_hook,
6265 varpool_call_node_removal_hooks, varpool_add_variable_insertion_hook,
6266 varpool_remove_variable_insertion_hook,
6267 varpool_call_variable_insertion_hooks): New functions.
6268 (varpool_remove_node): Use it.
6269
97f1f314 62702013-09-08 Paolo Carlini <paolo.carlini@oracle.com>
6df2ca09 6271
97f1f314
PC
6272 PR c++/54941
6273 * diagnostic.c (diagnostic_build_prefix): When s.file is
6274 "<built-in>" don't output line and column numbers.
6df2ca09 6275
bcb650cb
JH
62762013-09-06 Jan Hubicka <jh@suse.cz>
6277
6278 * cgraphunit.c (expand_thunk): Get body before touching arguments.
6279 * lto-streamer-out.c: Stream thunks, too.
6280 * lto-streamer-in.c (input_function): Pop cfun here
6281 (lto_read_body): Instead of here.
6282
87f3fea8
CT
62832013-09-06 Caroline Tice <cmtice@google.com>
6284
6285 * doc/install.texi: Add documentation for the --enable-vtable-verify
6286 and the --disable-libvtv configure options.
6287
5e8d9b45
JL
62882013-09-06 Jeff Law <law@redhat.com>
6289
6290 * tree-ssa-dom.c (cprop_into_successor_phis): Also propagate
6291 edge implied equivalences into successor phis.
6292
3a001aff
JR
62932013-09-06 Joern Rennecke <joern.rennecke@embecosm.com>
6294
6295 * resource.c (mark_referenced_resources): Handle COND_EXEC.
6296
62972013-09-06 Claudiu Zissulescu <claziss@synopsys.com>
c42091c1
CZ
6298
6299 * resource.c (mark_target_live_regs): Compute resources taking
6300 into account if a call is predicated or not.
6301
2c641f8b
EB
63022013-09-06 Eric Botcazou <ebotcazou@adacore.com>
6303
6304 * toplev.c (output_stack_usage): Be prepared for suffixes created by
6305 the compiler in the function names.
6306
a81b0a3d
JH
63072013-09-06 Jan Hubicka <jh@suse.cz>
6308
6309 PR middle-end/58094
6310 * ipa-inline.c (has_caller_p): New function.
6311 (want_inline_function_to_all_callers_p): Use it.
6312 (sum_callers, inline_to_all_callers): Break out from ...
6313 (ipa_inline): ... here.
6314
68e4ca7e
JH
63152013-09-06 Jan Hubicka <jh@suse.cz>
6316
2c641f8b
EB
6317 * config/i386/i386.c (ix86_hard_regno_mode_ok): AVX modes are valid
6318 only when AVX is enabled.
68e4ca7e 6319
3c73e2fe
JG
63202013-09-06 James Greenhalgh <james.greenhalgh@arm.com>
6321
6322 * config/aarch64/aarch64.md
6323 (*movtf_aarch64): Use neon_<ls>dm_2 as type where v8type
6324 is fpsimd_<load/store>2.
6325 (load_pair<mode>): Likewise.
6326 (store_pair<mode>): Likewise.
6327
8c48eecd
JG
63282013-09-06 James Greenhalgh <james.greenhalgh@arm.com>
6329
6330 * config/arm/types.md (type): Add "mrs" type.
6331 * config/aarch64/aarch64.md
6332 (aarch64_load_tp_hard): Make type "mrs".
6333 * config/arm/arm.md
6334 (load_tp_hard): Make type "mrs".
6335 * config/arm/cortex-a15.md: Update with new attributes.
6336 * config/arm/cortex-a5.md: Update with new attributes.
6337 * config/arm/cortex-a53.md: Update with new attributes.
6338 * config/arm/cortex-a7.md: Update with new attributes.
6339 * config/arm/cortex-a8.md: Update with new attributes.
6340 * config/arm/cortex-a9.md: Update with new attributes.
6341 * config/arm/cortex-m4.md: Update with new attributes.
6342 * config/arm/cortex-r4.md: Update with new attributes.
6343 * config/arm/fa526.md: Update with new attributes.
6344 * config/arm/fa606te.md: Update with new attributes.
6345 * config/arm/fa626te.md: Update with new attributes.
6346 * config/arm/fa726te.md: Update with new attributes.
6347
9ea187a5
JG
63482013-09-06 James Greenhalgh <james.greenhalgh@arm.com>
6349
6350 * config/aarch64/aarch64.md
6351 (*movti_aarch64): Use "multiple" for type where v8type is "move2".
6352 (*movtf_aarch64): Likewise.
6353 * config/arm/arm.md
6354 (thumb1_movdi_insn): Use "multiple" for type where more than one
6355 instruction is used for a move.
6356 (*arm32_movhf): Likewise.
6357 (*thumb_movdf_insn): Likewise.
6358
292b89b3
JG
63592013-09-06 James Greenhalgh <james.greenhalgh@arm.com>
6360
6361 * config/arm/types.md (type): Rename fcpys to fmov.
6362 * config/arm/vfp.md
6363 (*arm_movsi_vfp): Rename type fcpys as fmov.
6364 (*thumb2_movsi_vfp): Likewise
6365 (*movhf_vfp_neon): Likewise
6366 (*movhf_vfp): Likewise
6367 (*movsf_vfp): Likewise
6368 (*thumb2_movsf_vfp): Likewise
6369 (*movsfcc_vfp): Likewise
6370 (*thumb2_movsfcc_vfp): Likewise
6371 * config/aarch64/aarch64-simd.md
6372 (move_lo_quad_<mode>): Replace type mov_reg with fmovs.
6373 * config/aarch64/aarch64.md
6374 (*movsi_aarch64): Replace type mov_reg with fmovs.
6375 (*movdi_aarch64): Likewise
6376 (*movsf_aarch64): Likewise
6377 (*movdf_aarch64): Likewise
6378 * config/arm/arm.c
6379 (cortexa7_older_only): Rename TYPE_FCPYS to TYPE_FMOV.
6380 * config/arm/iwmmxt.md
6381 (*iwmmxt_movsi_insn): Rename type fcpys as fmov.
6382 * config/arm/arm1020e.md: Update with new attributes.
6383 * config/arm/cortex-a15-neon.md: Update with new attributes.
6384 * config/arm/cortex-a5.md: Update with new attributes.
6385 * config/arm/cortex-a53.md: Update with new attributes.
6386 * config/arm/cortex-a7.md: Update with new attributes.
6387 * config/arm/cortex-a8-neon.md: Update with new attributes.
6388 * config/arm/cortex-a9.md: Update with new attributes.
6389 * config/arm/cortex-m4-fpu.md: Update with new attributes.
6390 * config/arm/cortex-r4f.md: Update with new attributes.
6391 * config/arm/marvell-pj4.md: Update with new attributes.
6392 * config/arm/vfp11.md: Update with new attributes.
6393
6a24a83e
JG
63942013-09-06 James Greenhalgh <james.greenhalgh@arm.com>
6395
6396 * config/aarch64/aarch64.md
6397 (*madd<mode>): Fix type attribute.
6398 (*maddsi_uxtw): Likewise.
6399 (*msub<mode>): Likewise.
6400 (*msubsi_uxtw): Likewise.
6401 (<su_optab>maddsidi4): Likewise.
6402 (<su_optab>msubsidi4): Likewise.
6403
b86923f0
JG
64042013-09-06 James Greenhalgh <james.greenhalgh@arm.com>
6405
6406 * config/arm/types.md: Split fdiv<sd> as fsqrt<sd>, fdiv<sd>.
6407 * config/arm/arm.md (core_cycles): Remove fdiv.
6408 * config/arm/vfp.md:
6409 (*sqrtsf2_vfp): Update for attribute changes.
6410 (*sqrtdf2_vfp): Likewise.
6411 * config/aarch64/aarch64.md:
6412 (sqrt<mode>2): Update for attribute changes.
6413 * config/arm/arm1020e.md: Update with new attributes.
6414 * config/arm/cortex-a15-neon.md: Update with new attributes.
6415 * config/arm/cortex-a5.md: Update with new attributes.
6416 * config/arm/cortex-a53.md: Update with new attributes.
6417 * config/arm/cortex-a7.md: Update with new attributes.
6418 * config/arm/cortex-a8-neon.md: Update with new attributes.
6419 * config/arm/cortex-a9.md: Update with new attributes.
6420 * config/arm/cortex-m4-fpu.md: Update with new attributes.
6421 * config/arm/cortex-r4f.md: Update with new attributes.
6422 * config/arm/marvell-pj4.md: Update with new attributes.
6423 * config/arm/vfp11.md: Update with new attributes.
6424
7b49c9e1
JG
64252013-09-06 James Greenhalgh <james.greenhalgh@arm.com>
6426
6427 * config/arm/types.md
6428 (type): Split f_cvt as f_cvt, f_cvtf2i, f_cvti2f.
6429 * config/aarch64/aarch64.md
6430 (l<fcvt_pattern><su_optab><GPF:mode><GPI:mode>2): Update with
6431 new attributes.
6432 (fix_trunc<GPF:mode><GPI:mode>2): Likewise.
6433 (fixuns_trunc<GPF:mode><GPI:mode>2): Likewise.
6434 (float<GPI:mode><GPF:mode>2): Likewise.
6435 * config/arm/vfp.md
6436 (*truncsisf2_vfp): Update with new attributes.
6437 (*truncsidf2_vfp): Likewise.
6438 (fixuns_truncsfsi2): Likewise.
6439 (fixuns_truncdfsi2): Likewise.
6440 (*floatsisf2_vfp): Likewise.
6441 (*floatsidf2_vfp): Likewise.
6442 (floatunssisf2): Likewise.
6443 (floatunssidf2): Likewise.
6444 (*combine_vcvt_f32_<FCVTI32typename>): Likewise.
6445 (*combine_vcvt_f64_<FCVTI32typename>): Likewise.
6446 * config/arm/arm1020e.md: Update with new attributes.
6447 * config/arm/cortex-a15-neon.md: Update with new attributes.
6448 * config/arm/cortex-a5.md: Update with new attributes.
6449 * config/arm/cortex-a53.md: Update with new attributes.
6450 * config/arm/cortex-a7.md: Update with new attributes.
6451 * config/arm/cortex-a8-neon.md: Update with new attributes.
6452 * config/arm/cortex-a9.md: Update with new attributes.
6453 * config/arm/cortex-m4-fpu.md: Update with new attributes.
6454 * config/arm/cortex-r4f.md: Update with new attributes.
6455 * config/arm/marvell-pj4.md: Update with new attributes.
6456 * config/arm/vfp11.md: Update with new attributes.
6457
e7df8af8
JG
64582013-09-06 James Greenhalgh <james.greenhalgh@arm.com>
6459
6460 * config/aarch64/arm_neon.h
6461 (vqtbl<1,2,3,4><q>_s8): Fix control vector parameter type.
6462 (vqtbx<1,2,3,4><q>_s8): Likewise.
6463
594726e4
JG
64642013-09-06 James Greenhalgh <james.greenhalgh@arm.com>
6465
f43245d1
UB
6466 * config/arm/types.md: Add "no_insn", "multiple" and "untyped" types.
6467 * config/arm/arm-fixed.md: Add type attribute to all insn patterns.
594726e4
JG
6468 (add<mode>3): Add type attribute.
6469 (add<mode>3): Likewise.
6470 (usadd<mode>3): Likewise.
6471 (ssadd<mode>3): Likewise.
6472 (sub<mode>3): Likewise.
6473 (sub<mode>3): Likewise.
6474 (ussub<mode>3): Likewise.
6475 (sssub<mode>3): Likewise.
6476 (ssmulsa3): Likewise.
6477 (usmulusa3): Likewise.
6478 (arm_usatsihi): Likewise.
6479 * config/arm/vfp.md
6480 (*movdi_vfp): Add types for all instructions.
6481 (*movdi_vfp_cortexa8): Likewise.
6482 (*movhf_vfp_neon): Likewise.
6483 (*movhf_vfp): Likewise.
6484 (*movdf_vfp): Likewise.
6485 (*thumb2_movdf_vfp): Likewise.
6486 (*thumb2_movdfcc_vfp): Likewise.
6487 * config/arm/arm.md: Add type attribute to all insn patterns.
6488 (*thumb1_adddi3): Add type attribute.
6489 (*arm_adddi3): Likewise.
6490 (*adddi_sesidi_di): Likewise.
6491 (*adddi_zesidi_di): Likewise.
6492 (*thumb1_addsi3): Likewise.
6493 (addsi3_compare0): Likewise.
6494 (*addsi3_compare0_scratch): Likewise.
6495 (*compare_negsi_si): Likewise.
6496 (cmpsi2_addneg): Likewise.
6497 (*addsi3_carryin_<optab>): Likewise.
6498 (*addsi3_carryin_alt2_<optab>): Likewise.
6499 (*addsi3_carryin_clobercc_<optab>): Likewise.
6500 (*subsi3_carryin): Likewise.
6501 (*subsi3_carryin_const): Likewise.
6502 (*subsi3_carryin_compare): Likewise.
6503 (*subsi3_carryin_compare_const): Likewise.
6504 (*arm_subdi3): Likewise.
6505 (*thumb_subdi3): Likewise.
6506 (*subdi_di_zesidi): Likewise.
6507 (*subdi_di_sesidi): Likewise.
6508 (*subdi_zesidi_di): Likewise.
6509 (*subdi_sesidi_di): Likewise.
6510 (*subdi_zesidi_ze): Likewise.
6511 (thumb1_subsi3_insn): Likewise.
6512 (*arm_subsi3_insn): Likewise.
6513 (*anddi3_insn): Likewise.
6514 (*anddi_zesidi_di): Likewise.
6515 (*anddi_sesdi_di): Likewise.
6516 (*ne_zeroextracts): Likewise.
6517 (*ne_zeroextracts): Likewise.
6518 (*ite_ne_zeroextr): Likewise.
6519 (*ite_ne_zeroextr): Likewise.
6520 (*anddi_notdi_di): Likewise.
6521 (*anddi_notzesidi): Likewise.
6522 (*anddi_notsesidi): Likewise.
6523 (andsi_notsi_si): Likewise.
6524 (thumb1_bicsi3): Likewise.
6525 (*iordi3_insn): Likewise.
6526 (*iordi_zesidi_di): Likewise.
6527 (*iordi_sesidi_di): Likewise.
6528 (*thumb1_iorsi3_insn): Likewise.
6529 (*xordi3_insn): Likewise.
6530 (*xordi_zesidi_di): Likewise.
6531 (*xordi_sesidi_di): Likewise.
6532 (*arm_xorsi3): Likewise.
6533 (*andsi_iorsi3_no): Likewise.
6534 (*smax_0): Likewise.
6535 (*smax_m1): Likewise.
6536 (*arm_smax_insn): Likewise.
6537 (*smin_0): Likewise.
6538 (*arm_smin_insn): Likewise.
6539 (*arm_umaxsi3): Likewise.
6540 (*arm_uminsi3): Likewise.
6541 (*minmax_arithsi): Likewise.
6542 (*minmax_arithsi_): Likewise.
6543 (*satsi_<SAT:code>): Likewise.
6544 (arm_ashldi3_1bit): Likewise.
6545 (arm_ashrdi3_1bit): Likewise.
6546 (arm_lshrdi3_1bit): Likewise.
6547 (*arm_negdi2): Likewise.
6548 (*thumb1_negdi2): Likewise.
6549 (*arm_negsi2): Likewise.
6550 (*thumb1_negsi2): Likewise.
6551 (*negdi_extendsid): Likewise.
6552 (*negdi_zero_extend): Likewise.
6553 (*arm_abssi2): Likewise.
6554 (*thumb1_abssi2): Likewise.
6555 (*arm_neg_abssi2): Likewise.
6556 (*thumb1_neg_abss): Likewise.
6557 (one_cmpldi2): Likewise.
6558 (extend<mode>di2): Likewise.
6559 (*compareqi_eq0): Likewise.
6560 (*arm_extendhisi2addsi): Likewise.
6561 (*arm_movdi): Likewise.
6562 (*thumb1_movdi_insn): Likewise.
6563 (*arm_movt): Likewise.
6564 (*thumb1_movsi_insn): Likewise.
6565 (pic_add_dot_plus_four): Likewise.
6566 (pic_add_dot_plus_eight): Likewise.
6567 (tls_load_dot_plus_eight): Likewise.
6568 (*thumb1_movhi_insn): Likewise.
6569 (*thumb1_movsf_insn): Likewise.
6570 (*movdf_soft_insn): Likewise.
6571 (*thumb_movdf_insn): Likewise.
6572 (cbranchsi4_insn): Likewise.
6573 (cbranchsi4_scratch): Likewise.
6574 (*negated_cbranchsi4): Likewise.
6575 (*tbit_cbranch): Likewise.
6576 (*tlobits_cbranch): Likewise.
6577 (*tstsi3_cbranch): Likewise.
6578 (*cbranchne_decr1): Likewise.
6579 (*addsi3_cbranch): Likewise.
6580 (*addsi3_cbranch_scratch): Likewise.
6581 (*arm_cmpdi_insn): Likewise.
6582 (*arm_cmpdi_unsig): Likewise.
6583 (*arm_cmpdi_zero): Likewise.
6584 (*thumb_cmpdi_zero): Likewise.
6585 (*deleted_compare): Likewise.
6586 (*mov_scc): Likewise.
6587 (*mov_negscc): Likewise.
6588 (*mov_notscc): Likewise.
6589 (*cstoresi_eq0_thumb1_insn): Likewise.
6590 (cstoresi_nltu_thumb1): Likewise.
6591 (cstoresi_ltu_thu): Likewise.
6592 (thumb1_addsi3_addgeu): Likewise.
6593 (*arm_jump): Likewise.
6594 (*thumb_jump): Likewise.
6595 (*check_arch2): Likewise.
6596 (arm_casesi_internal): Likewise.
6597 (thumb1_casesi_dispatch): Likewise.
6598 (*arm_indirect_jump): Likewise.
6599 (*thumb1_indirect_jump): Likewise.
6600 (nop): Likewise.
6601 (*and_scc): Likewise.
6602 (*ior_scc): Likewise.
6603 (*compare_scc): Likewise.
6604 (*cond_move): Likewise.
6605 (*cond_arith): Likewise.
6606 (*cond_sub): Likewise.
6607 (*cmp_ite0): Likewise.
6608 (*cmp_ite1): Likewise.
6609 (*cmp_and): Likewise.
6610 (*cmp_ior): Likewise.
6611 (*ior_scc_scc): Likewise.
6612 (*ior_scc_scc_cmp): Likewise.
6613 (*and_scc_scc): Likewise.
6614 (*and_scc_scc_cmp): Likewise.
6615 (*and_scc_scc_nod): Likewise.
6616 (*negscc): Likewise.
6617 (movcond_addsi): Likewise.
6618 (movcond): Likewise.
6619 (*ifcompare_plus_move): Likewise.
6620 (*if_plus_move): Likewise.
6621 (*ifcompare_move_plus): Likewise.
6622 (*if_move_plus): Likewise.
6623 (*ifcompare_arith_arith): Likewise.
6624 (*if_arith_arith): Likewise.
6625 (*ifcompare_arith_move): Likewise.
6626 (*if_arith_move): Likewise.
6627 (*ifcompare_move_arith): Likewise.
6628 (*if_move_arith): Likewise.
6629 (*ifcompare_move_not): Likewise.
6630 (*if_move_not): Likewise.
6631 (*ifcompare_not_move): Likewise.
6632 (*if_not_move): Likewise.
6633 (*ifcompare_shift_move): Likewise.
6634 (*if_shift_move): Likewise.
6635 (*ifcompare_move_shift): Likewise.
6636 (*if_move_shift): Likewise.
6637 (*ifcompare_shift_shift): Likewise.
6638 (*ifcompare_not_arith): Likewise.
6639 (*ifcompare_arith_not): Likewise.
6640 (*if_arith_not): Likewise.
6641 (*ifcompare_neg_move): Likewise.
6642 (*if_neg_move): Likewise.
6643 (*ifcompare_move_neg): Likewise.
6644 (*if_move_neg): Likewise.
6645 (prologue_thumb1_interwork): Likewise.
6646 (*cond_move_not): Likewise.
6647 (*sign_extract_onebit): Likewise.
6648 (*not_signextract_onebit): Likewise.
6649 (stack_tie): Likewise.
6650 (align_4): Likewise.
6651 (align_8): Likewise.
6652 (consttable_end): Likewise.
6653 (consttable_1): Likewise.
6654 (consttable_2): Likewise.
6655 (consttable_4): Likewise.
6656 (consttable_8): Likewise.
6657 (consttable_16): Likewise.
6658 (*thumb1_tablejump): Likewise.
6659 (prefetch): Likewise.
6660 (force_register_use): Likewise.
6661 (thumb_eh_return): Likewise.
6662 (load_tp_hard): Likewise.
6663 (load_tp_soft): Likewise.
6664 (tlscall): Likewise.
6665 (*arm_movtas_ze): Likewise.
6666 (*arm_rev): Likewise.
6667 (*arm_revsh): Likewise.
6668 (*arm_rev16): Likewise.
6669 * config/arm/thumb2.md
6670 (*thumb2_smaxsi3): Likewise.
6671 (*thumb2_sminsi3): Likewise.
6672 (*thumb32_umaxsi3): Likewise.
6673 (*thumb2_uminsi3): Likewise.
6674 (*thumb2_negdi2): Likewise.
6675 (*thumb2_abssi2): Likewise.
6676 (*thumb2_neg_abss): Likewise.
6677 (*thumb2_movsi_insn): Likewise.
6678 (tls_load_dot_plus_four): Likewise.
6679 (*thumb2_movhi_insn): Likewise.
6680 (*thumb2_mov_scc): Likewise.
6681 (*thumb2_mov_negs): Likewise.
6682 (*thumb2_mov_negs): Likewise.
6683 (*thumb2_mov_nots): Likewise.
6684 (*thumb2_mov_nots): Likewise.
6685 (*thumb2_movsicc_): Likewise.
6686 (*thumb2_movsfcc_soft_insn): Likewise.
6687 (*thumb2_indirect_jump): Likewise.
6688 (*thumb2_and_scc): Likewise.
6689 (*thumb2_ior_scc): Likewise.
6690 (*thumb2_ior_scc_strict_it): Likewise.
6691 (*thumb2_cond_move): Likewise.
6692 (*thumb2_cond_arith): Likewise.
6693 (*thumb2_cond_ari): Likewise.
6694 (*thumb2_cond_sub): Likewise.
6695 (*thumb2_negscc): Likewise.
6696 (*thumb2_movcond): Likewise.
6697 (thumb2_casesi_internal): Likewise.
6698 (thumb2_casesi_internal_pic): Likewise.
6699 (*thumb2_alusi3_short): Likewise.
6700 (*thumb2_mov<mode>_shortim): Likewise.
6701 (*thumb2_addsi_short): Likewise.
6702 (*thumb2_subsi_short): Likewise.
6703 (thumb2_addsi3_compare0): Likewise.
6704 (*thumb2_cbz): Likewise.
6705 (*thumb2_cbnz): Likewise.
6706 (*thumb2_one_cmplsi2_short): Likewise.
6707 (*thumb2_negsi2_short): Likewise.
6708 (*orsi_notsi_si): Likewise.
6709 * config/arm/arm1020e.md: Update with new attributes.
6710 * config/arm/arm1026ejs.md: Update with new attributes.
6711 * config/arm/arm1136jfs.md: Update with new attributes.
6712 * config/arm/arm926ejs.md: Update with new attributes.
6713 * config/arm/cortex-a15.md: Update with new attributes.
6714 * config/arm/cortex-a5.md: Update with new attributes.
6715 * config/arm/cortex-a53.md: Update with new attributes.
6716 * config/arm/cortex-a7.md: Update with new attributes.
6717 * config/arm/cortex-a8.md: Update with new attributes.
6718 * config/arm/cortex-a9.md: Update with new attributes.
6719 * config/arm/cortex-m4.md: Update with new attributes.
6720 * config/arm/cortex-r4.md: Update with new attributes.
6721 * config/arm/fa526.md: Update with new attributes.
6722 * config/arm/fa606te.md: Update with new attributes.
6723 * config/arm/fa626te.md: Update with new attributes.
6724 * config/arm/fa726te.md: Update with new attributes.
6725
1c83b673
JG
67262013-09-06 James Greenhalgh <james.greenhalgh@arm.com>
6727
6728 * config/aarch64/aarch64-simd.md
6729 (aarch64_sqdml<SBINQOPS:as>l_n<mode>_internal): Use
6730 <vwx> iterator to ensure correct register choice.
6731 (aarch64_sqdml<SBINQOPS:as>l2_n<mode>_internal): Likewise.
6732 (aarch64_sqdmull_n<mode>): Likewise.
6733 (aarch64_sqdmull2_n<mode>_internal): Likewise.
6734 * config/aarch64/arm_neon.h
6735 (vml<as><q>_lane<q>_<su>16): Use 'x' constraint for element vector.
6736 (vml<as><q>_n_<su>16): Likewise.
6737 (vml<as>l_high_lane<q>_<su>16): Likewise.
6738 (vml<as>l_high_n_<su>16): Likewise.
6739 (vml<as>l_lane<q>_<su>16): Likewise.
6740 (vml<as>l_n_<su>16): Likewise.
6741 (vmul<q>_lane<q>_<su>16): Likewise.
6742 (vmul<q>_n_<su>16): Likewise.
6743 (vmull_lane<q>_<su>16): Likewise.
6744 (vmull_n_<su>16): Likewise.
6745 (vmull_high_lane<q>_<su>16): Likewise.
6746 (vmull_high_n_<su>16): Likewise.
6747 (vqrdmulh<q>_n_s16): Likewise.
6748
bb1ae543
TB
67492013-09-06 Tejas Belagod <tejas.belagod@arm.com>
6750
6751 * config/aarch64/arm_neon.h: Fix all vdup<bhsd_lane<q> intrinsics to
6752 have the correct lane parameter.
6753
f23c0742
RB
67542013-09-06 Richard Biener <rguenther@suse.de>
6755
6756 * cfganal.c (control_dependences::~control_dependences):
6757 Properly free all of the vector.
6758
c1d49770
KY
67592013-09-06 Kirill Yukhin <kirill.yukhin@intel.com>
6760
6761 PR target/58269
6762 * config/i386/i386.c (ix86_conditional_register_usage):
6763 Proper initialize extended SSE registers.
6764
6da47f52
JH
67652013-09-06 Jan Hubicka <jh@suse.cz>
6766
6767 PR tree-optimization/58311
6768 * ipa-devirt.c (gate_ipa_devirt): Only execute when optimizing.
6769
67702013-09-06 Jan Hubicka <jh@suse.cz>
6771
6772 * Makefile.in (tree-sra.o): Update dependencies.
6773 * tree-sra.c: Include ipa-utils.h
6774 (scan_function): Use recursive_call_p.
6775 (has_caller_p): New function.
6776 (cgraph_for_node_and_aliases): Count also callers of aliases.
6777
fc11f321
JH
67782013-09-06 Jan Hubicka <jh@suse.cz>
6779
6780 PR middle-end/58094
6781 * cgraph.h (symtab_semantically_equivalent_p): Declare.
6782 * tree-tailcall.c: Include ipa-utils.h.
6783 (find_tail_calls): Use it.
6784 * ipa-pure-const.c (check_call): Likewise.
6785 * ipa-utils.c (recursive_call_p): New function.
6786 * ipa-utils.h (recursive_call_p): Dclare.
6787 * symtab.c (symtab_nonoverwritable_alias): Fix formatting.
6788 (symtab_semantically_equivalent_p): New function.
6789 * Makefile.in (tree-tailcall.o): Update dependencies.
6790
9a6af450
EB
67912013-09-06 Eric Botcazou <ebotcazou@adacore.com>
6792
6793 * ipa-split.c (split_function): Set DECL_NO_INLINE_WARNING_P on the
6794 non-inlinable part.
6795
35f32ad4
RB
67962013-09-06 Richard Biener <rguenther@suse.de>
6797
6798 * lto-streamer.h (lto_global_var_decls): Remove.
6799 * Makefile.in (OBJS): Remove lto-symtab.o.
6800 (lto-symtab.o): Remove.
6801 (GTFILES): Remove lto-symtab.c
6802 * lto-symtab.c: Move to lto/
6803
d12a76f3
AK
68042013-09-06 Andreas Krebbel <Andreas.Krebbel@de.ibm.com>
6805
6806 * config/s390/s390.md (UNSPEC_FPINT_FLOOR, UNSPEC_FPINT_BTRUNC)
6807 (UNSPEC_FPINT_ROUND, UNSPEC_FPINT_CEIL, UNSPEC_FPINT_NEARBYINT)
6808 (UNSPEC_FPINT_RINT): New constant definitions.
6809 (FPINT, fpint_name, fpint_roundingmode): New integer iterator
6810 definition with 2 attributes.
6811 ("<FPINT:fpint_name><BFP:mode>2", "rint<BFP:mode>2")
6812 ("<FPINT:fpint_name><DFP:mode>2", "rint<DFP:mode>2"): New pattern
6813 definitions.
6814
68152013-09-06 Andreas Krebbel <Andreas.Krebbel@de.ibm.com>
6816
f43245d1 6817 * config/s390/s390.md: Add "bcr_flush" value to mnemonic attribute.
d12a76f3
AK
6818 ("mem_thread_fence_1"): Use bcr 14,0 for z196 and later.
6819 Set the mnemonic attribute to "bcr_flush". Set the "z196prop"
6820 attribute to "z196_alone".
6821 * config/s390/2827.md: Add "bcr_flush" to "ooo_groupalone" and
6822 "zEC12_simple".
6823
c8e9d8c3
RB
68242013-09-06 Richard Biener <rguenther@suse.de>
6825
6826 * basic-block.h (class control_dependences): New.
6827 * tree-ssa-dce.c (control_dependence_map): Remove.
6828 (cd): New global.
6829 (EXECUTE_IF_CONTROL_DEPENDENT): Remove.
6830 (set_control_dependence_map_bit, clear_control_dependence_bitmap,
6831 find_pdom, find_control_dependence, find_all_control_dependences):
6832 Move to cfganal.c.
f43245d1
UB
6833 (mark_control_dependent_edges_necessary,
6834 find_obviously_necessary_stmts, propagate_necessity, tree_dce_init,
6835 tree_dce_done, perform_tree_ssa_dce): Adjust.
c8e9d8c3
RB
6836 * cfganal.c (set_control_dependence_map_bit,
6837 clear_control_dependence_bitmap, find_pdom, find_control_dependence,
6838 find_all_control_dependences): Move from tree-ssa-dce.c and
6839 implement as methods of control_dependences class.
6840 (control_dependences::control_dependences): New.
6841 (control_dependences::~control_dependences): Likewise.
6842 (control_dependences::get_edges_dependent_on): Likewise.
6843 (control_dependences::get_edge): Likewise.
6844
7d2ba471
JH
68452013-09-04 Jan Hubicka <jh@suse.cz>
6846
6847 * tree.c (types_same_for_odr): Drop overactive check.
6848 * ipa-devirt.c (hash_type_name): Likewise.
6849
e18412fc
JH
68502013-09-04 Jan Hubicka <jh@suse.cz>
6851
6852 * cgraphunit.c (walk_polymorphic_call_targets): Break out from ...
6853 (analyze_functions): ... here.
6854
d352b245
JH
68552013-09-04 Jan Hubicka <jh@suse.cz>
6856
6857 PR middle-end/58201
6858 * cgraphunit.c (analyze_functions): Clear AUX fields
6859 after processing; initialize assembler name has.
6860
770da076
JL
68612013-09-05 Jeff Law <law@redhat.com>
6862
6863 * tree-ssa-threadedge.c (thread_around_empty_blocks): Renamed
6864 from thread_around_empty_block. Record threading path into PATH.
6865 Recurse if threading through the initial block is successful.
6866 (thread_across_edge): Corresponding changes to slightly simplify.
6867
fe6f68e2
JG
68682013-09-05 James Greenhalgh <james.greenhalgh@arm.com>
6869
6870 * config/aarch64/aarch64.md
6871 (type): Remove frecpe, frecps, frecpx.
6872 (aarch64_frecp<FRECP:frecp_suffix><mode>): Move to aarch64-simd.md,
6873 fix to be a TARGET_SIMD instruction.
6874 (aarch64_frecps): Remove.
6875 * config/aarch64/aarch64-simd.md
6876 (aarch64_frecp<FRECP:frecp_suffix><mode>): New, moved from aarch64.md
6877 (aarch64_frecps<mode>): Handle all float/vector of float modes.
6878
6e4150e1
JG
68792013-09-05 James Greenhalgh <james.greenhalgh@arm.com>
6880 Sofiane Naci <sofiane.naci@arm.com>
6881
f43245d1 6882 * config/arm/types.md (define_attr "type"): Expand "arlo_imm"
6e4150e1 6883 into "adr", "alu_imm", "alus_imm", "logic_imm", "logics_imm".
f43245d1
UB
6884 Expand "arlo_reg" into "adc_reg", "adc_imm", "adcs_reg", "adcs_imm",
6885 "alu_ext", "alu_reg", "alus_ext", "alus_reg", "bfm", "csel",
6886 "logic_reg", "logics_reg", "rev". Expand "arlo_shift" into
6887 "alu_shift_imm", "alus_shift_imm", "logic_shift_imm",
6888 "logics_shift_imm". Expand "arlo_shift_reg" into "alu_shift_reg",
6889 "alus_shift_reg", "logic_shift_reg", "logics_shift_reg". Expand "clz"
6890 into "clz, "rbit". Rename "shift" to "shift_imm".
6e4150e1 6891 * config/arm/arm.md (define_attr "core_cycles"): Update for attribute
f43245d1 6892 changes. Update for attribute changes all occurrences of arlo_* and
6e4150e1
JG
6893 shift* types.
6894 * config/arm/arm-fixed.md: Update for attribute changes
6895 all occurrences of arlo_* types.
6896 * config/arm/thumb2.md: Update for attribute changes all occurrences
6897 of arlo_* types.
6898 * config/arm/arm.c (xscale_sched_adjust_cost): (rtx insn, rtx
6899 (cortexa7_older_only): Likewise.
6900 (cortexa7_younger): Likewise.
6901 * config/arm/arm1020e.md (1020alu_op): Update for attribute changes.
6902 (1020alu_shift_op): Likewise.
6903 (1020alu_shift_reg_op): Likewise.
6904 * config/arm/arm1026ejs.md (alu_op): Update for attribute changes.
6905 (alu_shift_op): Likewise.
6906 (alu_shift_reg_op): Likewise.
f43245d1 6907 * config/arm/arm1136jfs.md (11_alu_op): Update for attribute changes.
6e4150e1
JG
6908 (11_alu_shift_op): Likewise.
6909 (11_alu_shift_reg_op): Likewise.
6910 * config/arm/arm926ejs.md (9_alu_op): Update for attribute changes.
6911 (9_alu_shift_reg_op): Likewise.
6912 * config/arm/cortex-a15.md (cortex_a15_alu): Update for
6913 attribute changes.
6914 (cortex_a15_alu_shift): Likewise.
6915 (cortex_a15_alu_shift_reg): Likewise.
6916 * config/arm/cortex-a5.md (cortex_a5_alu): Update for
6917 attribute changes.
6918 (cortex_a5_alu_shift): Likewise.
f43245d1
UB
6919 * config/arm/cortex-a53.md (cortex_a53_alu): Update for
6920 attribute changes.
6e4150e1 6921 (cortex_a53_alu_shift): Likewise.
f43245d1
UB
6922 * config/arm/cortex-a7.md (cortex_a7_alu_imm): Update for
6923 attribute changes.
6e4150e1
JG
6924 (cortex_a7_alu_reg): Likewise.
6925 (cortex_a7_alu_shift): Likewise.
f43245d1
UB
6926 * config/arm/cortex-a8.md (cortex_a8_alu): Update for
6927 attribute changes.
6e4150e1
JG
6928 (cortex_a8_alu_shift): Likewise.
6929 (cortex_a8_alu_shift_reg): Likewise.
f43245d1 6930 * config/arm/cortex-a9.md (cortex_a9_dp): Update for attribute changes.
6e4150e1 6931 (cortex_a9_dp_shift): Likewise.
f43245d1
UB
6932 * config/arm/cortex-m4.md (cortex_m4_alu): Update for
6933 attribute changes.
6e4150e1
JG
6934 * config/arm/cortex-r4.md
6935 (cortex_r4_alu): Update for attribute changes.
6936 (cortex_r4_mov): Likewise.
6937 (cortex_r4_alu_shift_reg): Likewise.
f43245d1 6938 * config/arm/fa526.md (526_alu_op): Update for attribute changes.
6e4150e1 6939 (526_alu_shift_op): Likewise.
f43245d1
UB
6940 * config/arm/fa606te.md (606te_alu_op): Update for attribute changes.
6941 * config/arm/fa626te.md (626te_alu_op): Update for attribute changes.
6e4150e1 6942 (626te_alu_shift_op): Likewise.
f43245d1 6943 * config/arm/fa726te.md (726te_alu_op): Update for attribute changes.
6e4150e1
JG
6944 (726te_alu_shift_op): Likewise.
6945 (726te_alu_shift_reg_op): Likewise.
6946 * config/arm/fmp626.md (mp626_alu_op): Update for attribute changes.
6947 (mp626_alu_shift_op): Likewise.
6948 * config/arm/marvell-pj4.md (pj4_alu): Update for attribute changes.
6949 (pj4_alu_conds): Likewise.
6950 (pj4_shift): Likewise.
6951 (pj4_shift_conds): Likewise.
6952 (pj4_alu_shift): Likewise.
6953 (pj4_alu_shift_conds): Likewise.
6954 * config/aarch64/aarch64.md: Update for attribute change
6955 all occurrences of arlo_* and shift* types.
6956
e6523306
MS
69572013-09-05 Mike Stump <mikestump@comcast.net>
6958
6959 * tree.h: Move documentation for tree_function_decl to tree-core.h
6960 with the declaration.
6961
80a18161
PB
69622013-09-05 Peter Bergner <bergner@vnet.ibm.com>
6963
6964 PR target/58139
6965 * reginfo.c (choose_hard_reg_mode): Scan through all mode classes
6966 looking for widest mode.
6967
6994430a
EB
69682013-09-05 Eric Botcazou <ebotcazou@adacore.com>
6969
6970 * config.gcc (*-*-vxworks*): Do not override an existing extra_objs.
6971
e67f39f7
RB
69722013-09-05 Richard Biener <rguenther@suse.de>
6973
6974 PR tree-optimization/58137
6975 * tree-vect-stmts.c (get_vectype_for_scalar_type_and_size):
6976 Do not create vectors of pointers.
6977 * tree-vect-loop.c (get_initial_def_for_induction): Use proper
6978 types for the components of the vector initializer.
6979 * tree-cfg.c (verify_gimple_assign_binary): Remove special-casing
6980 allowing pointer vectors with PLUS_EXPR/MINUS_EXPR.
6981
568cda29
MJ
69822013-09-05 Martin Jambor <mjambor@suse.cz>
6983
6984 * ipa-prop.c (remove_described_reference): Accept missing references,
6985 return false if that hppens, otherwise return true.
6986 (cgraph_node_for_jfunc): New function.
6987 (try_decrement_rdesc_refcount): Likewise.
6988 (try_make_edge_direct_simple_call): Use them.
6989 (ipa_edge_removal_hook): Remove references from rdescs.
6990 (ipa_edge_duplication_hook): Clone rdescs and their references
6991 when the new edge has the same caller as the old one.
6992 * cgraph.c (cgraph_resolve_speculation): Remove speculative
6993 reference before removing any edges.
6994
5a200acb
RE
69952013-09-05 Richard Earnshaw <rearnsha@arm.com>
6996
6997 * arm.c (thumb2_emit_strd_push): Rewrite to use pre-decrement on
6998 initial store.
6999 * thumb2.md (thumb2_storewb_parisi): New pattern.
7000
5922847b
YZ
70012013-09-05 Yufeng Zhang <yufeng.zhang@arm.com>
7002
7003 * config/aarch64/aarch64-option-extensions.def: Add
7004 AARCH64_OPT_EXTENSION of 'crc'.
7005 * config/aarch64/aarch64.h (AARCH64_FL_CRC): New define.
7006 (AARCH64_ISA_CRC): Ditto.
7007 * doc/invoke.texi (-march and -mcpu feature modifiers): Add
7008 description of the CRC extension.
7009
996746aa
AI
70102013-09-05 Alexander Ivchenko <alexander.ivchenko@intel.com>
7011
7012 * config/rs6000/linux64.h: Define OPTION_BIONIC and OPTION_UCLIBC.
7013 * config/rs6000/linux.h: Ditto.
7014 * alpha/linux.h: Ditto.
7015 * config/bfin/uclinux.h: Define TARGET_LIBC_HAS_FUNCTION as
7016 no_c99_libc_has_function.
7017 * config/c6x/uclinux-elf.h: Ditto.
7018 * config/lm32/uclinux-elf.h: Ditto.
7019 * config/m68k/uclinux.h: Ditto.
7020 * config/moxie/uclinux.h: Ditto.
7021 * config.gcc (bfin*-linux-uclibc*): Add t-linux-android to tmake_file.
7022 (crisv32-*-linux*, cris-*-linux*): Ditto.
7023 * config/bfin/bfin.c: Include "tm_p.h".
7024
c370d99f
RB
70252013-09-05 Richard Biener <rguenther@suse.de>
7026
7027 * tree-vect-loop.c (vect_analyze_loop_operations): Properly
7028 check for a definition without a basic-block.
7029
003bb7f3
JG
70302013-09-05 James Greenhalgh <james.greenhalgh@arm.com>
7031 Sofiane Naci <sofiane.naci@arm.com>
7032
7033 * config/aarch64/aarch64.md
7034 (*movti_aarch64): Rename r_2_f and f_2_r.
7035 (*movsf_aarch64): Likewise.
7036 (*movdf_aarch64): Likewise.
7037 (*movtf_aarch64): Likewise.
7038 (aarch64_movdi_<mode>low): Likewise.
7039 (aarch64_movdi_<mode>high): Likewise.
7040 (aarch64_mov<mode>high_di): Likewise.
7041 (aarch64_mov<mode>low_di): Likewise.
7042 (aarch64_movtilow_tilow): Likewise.
7043 * config/arm/arm.md (attribute "neon_type"): Delete. Move attribute
7044 values to config/arm/types.md
7045 (attribute "conds"): Update for attribute change.
7046 (anddi3_insn): Likewise.
7047 (iordi3_insn): Likewise.
7048 (xordi3_insn): Likewise.
7049 (one_cmpldi2): Likewise.
7050 * config/arm/types.md (type): Add Neon types.
7051 * config/arm/neon.md (neon_mov<mode>): Remove "neon_type" attribute,
7052 use "type" attribute.
7053 (movmisalign<mode>_neon_store): Likewise.
7054 (movmisalign<mode>_neon_load): Likewise.
7055 (vec_set<mode>_internal): Likewise.
7056 (vec_setv2di_internal): Likewise.
7057 (vec_extract<mode>): Likewise.
7058 (vec_extractv2di): Likewise.
7059 (add<mode>3_neon): Likewise.
7060 (adddi3_neon): Likewise.
7061 (sub<mode>3_neon): Likewise.
7062 (subdi3_neon): Likewise.
7063 (mul<mode>3_neon): Likewise.
7064 (mul<mode>3add<mode>_neon): Likewise.
7065 (mul<mode>3neg<mode>add<mode>_neon): Likewise.
7066 (fma<VCVTF:mode>4)): Likewise.
7067 (fma<VCVTF:mode>4_intrinsic): Likewise.
7068 (fmsub<VCVTF:mode>4)): Likewise.
7069 (fmsub<VCVTF:mode>4_intrinsic): Likewise.
7070 (neon_vrint<NEON_VRINT:nvrint_variant><VCVTF:mode>): Likewise.
7071 (ior<mode>3): Likewise.
7072 (and<mode>3): Likewise.
7073 (anddi3_neon): Likewise.
7074 (orn<mode>3_neon): Likewise.
7075 (orndi3_neon): Likewise.
7076 (bic<mode>3_neon): Likewise.
7077 (bicdi3_neon): Likewise.
7078 (xor<mode>3): Likewise.
7079 (one_cmpl<mode>2): Likewise.
7080 (abs<mode>2): Likewise.
7081 (neg<mode>2): Likewise.
7082 (umin<mode>3_neon): Likewise.
7083 (umax<mode>3_neon): Likewise.
7084 (smin<mode>3_neon): Likewise.
7085 (smax<mode>3_neon): Likewise.
7086 (vashl<mode>3): Likewise.
7087 (vashr<mode>3_imm): Likewise.
7088 (vlshr<mode>3_imm): Likewise.
7089 (ashl<mode>3_signed): Likewise.
7090 (ashl<mode>3_unsigned): Likewise.
7091 (neon_load_count): Likewise.
7092 (ashldi3_neon_noclobber): Likewise.
7093 (signed_shift_di3_neon): Likewise.
7094 (unsigned_shift_di3_neon): Likewise.
7095 (ashrdi3_neon_imm_noclobber): Likewise.
7096 (lshrdi3_neon_imm_noclobber): Likewise.
7097 (widen_ssum<mode>3): Likewise.
7098 (widen_usum<mode>3): Likewise.
7099 (quad_halves_<code>v4si): Likewise.
7100 (quad_halves_<code>v4sf): Likewise.
7101 (quad_halves_<code>v8hi): Likewise.
7102 (quad_halves_<code>v16qi): Likewise.
7103 (reduc_splus_v2di): Likewise.
7104 (neon_vpadd_internal<mode>): Likewise.
7105 (neon_vpsmin<mode>): Likewise.
7106 (neon_vpsmax<mode>): Likewise.
7107 (neon_vpumin<mode>): Likewise.
7108 (neon_vpumax<mode>): Likewise.
7109 (ss_add<mode>_neon): Likewise.
7110 (us_add<mode>_neon): Likewise.
7111 (ss_sub<mode>_neon): Likewise.
7112 (us_sub<mode>_neon): Likewise.
7113 (neon_vadd<mode>_unspec): Likewise.
7114 (neon_vaddl<mode>): Likewise.
7115 (neon_vaddw<mode>): Likewise.
7116 (neon_vhadd<mode>): Likewise.
7117 (neon_vqadd<mode>): Likewise.
7118 (neon_vaddhn<mode>): Likewise.
7119 (neon_vmul<mode>): Likewise.
7120 (neon_vmla<mode>): Likewise.
7121 (neon_vmlal<mode>): Likewise.
7122 (neon_vmls<mode>): Likewise.
7123 (neon_vmlsl<mode>): Likewise.
7124 (neon_vqdmulh<mode>): Likewise.
7125 (neon_vqdmlal<mode>): Likewise.
7126 (neon_vqdmlsl<mode>): Likewise.
7127 (neon_vmull<mode>): Likewise.
7128 (neon_vqdmull<mode>): Likewise.
7129 (neon_vsub<mode>_unspec): Likewise.
7130 (neon_vsubl<mode>): Likewise.
7131 (neon_vsubw<mode>): Likewise.
7132 (neon_vqsub<mode>): Likewise.
7133 (neon_vhsub<mode>): Likewise.
7134 (neon_vsubhn<mode>): Likewise.
7135 (neon_vceq<mode>): Likewise.
7136 (neon_vcge<mode>): Likewise.
7137 (neon_vcgeu<mode>): Likewise.
7138 (neon_vcgt<mode>): Likewise.
7139 (neon_vcgtu<mode>): Likewise.
7140 (neon_vcle<mode>): Likewise.
7141 (neon_vclt<mode>): Likewise.
7142 (neon_vcage<mode>): Likewise.
7143 (neon_vcagt<mode>): Likewise.
7144 (neon_vtst<mode>): Likewise.
7145 (neon_vabd<mode>): Likewise.
7146 (neon_vabdl<mode>): Likewise.
7147 (neon_vaba<mode>): Likewise.
7148 (neon_vabal<mode>): Likewise.
7149 (neon_vmax<mode>): Likewise.
7150 (neon_vmin<mode>): Likewise.
7151 (neon_vpaddl<mode>): Likewise.
7152 (neon_vpadal<mode>): Likewise.
7153 (neon_vpmax<mode>): Likewise.
7154 (neon_vpmin<mode>): Likewise.
7155 (neon_vrecps<mode>): Likewise.
7156 (neon_vrsqrts<mode>): Likewise.
7157 (neon_vqabs<mode>): Likewise.
7158 (neon_vqneg<mode>): Likewise.
7159 (neon_vcls<mode>): Likewise.
7160 (clz<mode>2): Likewise.
7161 (popcount<mode>2): Likewise.
7162 (neon_vrecpe): Likewise.
7163 (neon_vrsqrte): Likewise.
7164 (neon_vget_lane<mode>_sext_internal): Likewise.
7165 (neon_vget_lane<mode>_zext_internal): Likewise.
7166 (neon_vdup_n<mode>): Likewise.
7167 (neon_vdup_nv2di): Likewise.
7168 (neon_vdpu_lane<mode>_internal): Likewise.
7169 (neon_vswp<mode>): Likewise.
7170 (float<mode><V_cvtto>2): Likewise.
7171 (floatuns<mode><V_cvtto>2): Likewise.
7172 (fix_trunc<mode><V_cvtto>)2): Likewise
7173 (fixuns_trunc<mode><V_cvtto)2): Likewise.
7174 (neon_vcvt<mode>): Likewise.
7175 (neon_vcvtv4sfv4hf): Likewise.
7176 (neon_vcvtv4hfv4sf): Likewise.
7177 (neon_vcvt_n<mode>): Likewise.
7178 (neon_vmovn<mode>): Likewise.
7179 (neon_vqmovn<mode>): Likewise.
7180 (neon_vqmovun<mode>): Likewise.
7181 (neon_vmovl<mode>): Likewise.
7182 (neon_vmul_lane<mode>): Likewise.
7183 (neon_vmull_lane<mode>): Likewise.
7184 (neon_vqdmull_lane<mode>): Likewise.
7185 (neon_vqdmulh_lane<mode>): Likewise.
7186 (neon_vmla_lane<mode>): Likewise.
7187 (neon_vmlal_lane<mode>): Likewise.
7188 (neon_vqdmlal_lane<mode>): Likewise.
7189 (neon_vmls_lane<mode>): Likewise.
7190 (neon_vmlsl_lane<mode>): Likewise.
7191 (neon_vqdmlsl_lane<mode>): Likewise.
7192 (neon_vext<mode>): Likewise.
7193 (neon_vrev64<mode>): Likewise.
7194 (neon_vrev32<mode>): Likewise.
7195 (neon_vrev16<mode>): Likewise.
7196 (neon_vbsl<mode>_internal): Likewise.
7197 (neon_vshl<mode>): Likewise.
7198 (neon_vqshl<mode>): Likewise.
7199 (neon_vshr_n<mode>): Likewise.
7200 (neon_vshrn_n<mode>): Likewise.
7201 (neon_vqshrn_n<mode>): Likewise.
7202 (neon_vqshrun_n<mode>): Likewise.
7203 (neon_vshl_n<mode>): Likewise.
7204 (neon_vqshl_n<mode>): Likewise.
7205 (neon_vqshlu_n<mode>): Likewise.
7206 (neon_vshll_n<mode>): Likewise.
7207 (neon_vsra_n<mode>): Likewise.
7208 (neon_vsri_n<mode>): Likewise.
7209 (neon_vsli_n<mode>): Likewise.
7210 (neon_vtbl1v8qi): Likewise.
7211 (neon_vtbl2v8qi): Likewise.
7212 (neon_vtbl3v8qi): Likewise.
7213 (neon_vtbl4v8qi): Likewise.
7214 (neon_vtbx1v8qi): Likewise.
7215 (neon_vtbx2v8qi): Likewise.
7216 (neon_vtbx3v8qi): Likewise.
7217 (neon_vtbx4v8qi): Likewise.
7218 (neon_vtrn<mode>_internal): Likewise.
7219 (neon_vzip<mode>_internal): Likewise.
7220 (neon_vuzp<mode>_internal): Likewise.
7221 (neon_vld1<mode>): Likewise.
7222 (neon_vld1_lane<mode>): Likewise.
7223 (neon_vld1_dup<mode>): Likewise.
7224 (neon_vld1_dupv2di): Likewise.
7225 (neon_vst1<mode>): Likewise.
7226 (neon_vst1_lane<mode>): Likewise.
7227 (neon_vld2<mode>): Likewise.
7228 (neon_vld2_lane<mode>): Likewise.
7229 (neon_vld2_dup<mode>): Likewise.
7230 (neon_vst2<mode>): Likewise.
7231 (neon_vst2_lane<mode>): Likewise.
7232 (neon_vld3<mode>): Likewise.
7233 (neon_vld3qa<mode>): Likewise.
7234 (neon_vld3qb<mode>): Likewise.
7235 (neon_vld3_lane<mode>): Likewise.
7236 (neon_vld3_dup<mode>): Likewise.
7237 (neon_vst3<mode>): Likewise.
7238 (neon_vst3qa<mode>): Likewise.
7239 (neon_vst3qb<mode>): Likewise.
7240 (neon_vst3_lane<mode>): Likewise.
7241 (neon_vld4<mode>): Likewise.
7242 (neon_vld4qa<mode>): Likewise.
7243 (neon_vld4qb<mode>): Likewise.
7244 (neon_vld4_lane<mode>): Likewise.
7245 (neon_vld4_dup<mode>): Likewise.
7246 (neon_vst4<mode>): Likewise.
7247 (neon_vst4qa<mode>): Likewise.
7248 (neon_vst4qb<mode>): Likewise.
7249 (neon_vst4_lane<mode>): Likewise.
7250 (neon_vec_unpack<US>_lo_<mode>): Likewise.
7251 (neon_vec_unpack<US>_hi_<mode>): Likewise.
7252 (neon_vec_<US>mult_lo_<mode>): Likewise.
7253 (neon_vec_<US>mult_hi_<mode>): Likewise.
7254 (neon_vec_<US>shiftl_<mode>): Likewise.
7255 (neon_unpack<US>_<mode>): Likewise.
7256 (neon_vec_<US>mult_<mode>): Likewise.
7257 (vec_pack_trunc_<mode>): Likewise.
7258 (neon_vec_pack_trunk_<mode>): Likewise.
7259 (neon_vabd<mode>_2): Likewise.
7260 (neon_vabd<mode>_3): Likewise.
7261 * config/arm/vfp.md (arm_movsi_vfp): Update for attribute changes.
7262 (thumb2_movsi_vfp): Likewise.
7263 (movdi_vfp): Likewise.
7264 (movdi_vfp_cortexa8): Likewise.
7265 (movhf_vfp_neon): Likewise.
7266 (movhf_vfp): Likewiwse.
7267 (movsf_vfp): Likewiwse.
7268 (thumb2_movsf_vfp): Likewiwse.
7269 (movdf_vfp): Likewise.
7270 (thumb2_movdf_vfp): Likewise.
7271 (movsfcc_vfp): Likewise.
7272 (thumb2_movsfcc_vfp): Likewise.
7273 (movdfcc_vfp): Likewise.
7274 (thumb2_movdfcc_vfp): Likewise.
7275 * config/arm/arm.c (cortexa7_older_only): Update for attribute change.
7276 * config/arm/arm1020e.md (v10_c2v): Update for attribute change.
7277 (v10_v2c): Likewise.
7278 * config/arm/cortex-a15-neon.md (cortex_a15_neon_int_1): Update for
7279 attribute change.
7280 (cortex_a15_neon_int_2): Likewise.
7281 (cortex_a15_neon_int_3): Likewise.
7282 (cortex_a15_neon_int_4): Likewise.
7283 (cortex_a15_neon_int_5): Likewise.
7284 (cortex_a15_neon_vqneg_vqabs): Likewise.
7285 (cortex_a15_neon_vmov): Likewise.
7286 (cortex_a15_neon_vaba): Likewise.
7287 (cortex_a15_neon_vaba_qqq): Likewise.
7288 (cortex_a15_neon_mul_ddd_8_16_qdd_16_8_long_32_16_long): Likewise.
7289 (cortex_a15_neon_mul_qqq_8_16_32_ddd_32): Likewise.
7290 (cortex_a15_neon_mul_qdd_64_32_long_qqd_16_ddd_32_\
7291 scalar_64_32_long_scalar): Likewise.
7292 (cortex_a15_neon_mla_ddd_8_16_qdd_16_8_long_32_16_long): Likewise.
7293 (cortex_a15_neon_mla_qqq_8_16): Likewise.
7294 (cortex_a15_neon_mla_ddd_32_qqd_16_ddd_32_scalar_qdd_64_32_\
7295 lotype_qdd_64_32_long): Likewise.
7296 (cortex_a15_neon_mla_qqq_32_qqd_32_scalar): Likewise.
7297 (cortex_a15_neon_mul_ddd_16_scalar_32_16_long_scalar): Likewise.
7298 (cortex_a15_neon_mul_qqd_32_scalar): Likewise.
7299 (cortex_a15_neon_mla_ddd_16_scalar_qdd_32_16_long_scalar): Likewise.
7300 (cortex_a15_neon_shift_1): Likewise.
7301 (cortex_a15_neon_shift_2): Likewise.
7302 (cortex_a15_neon_shift_3): Likewise.
7303 (cortex_a15_neon_vshl_ddd): Likewise.
7304 (cortex_a15_neon_vqshl_vrshl_vqrshl_qqq): Likewise.
7305 (cortex_a15_neon_vsra_vrsra): Likewise.
7306 (cortex_a15_neon_fp_vadd_ddd_vabs_dd): Likewise.
7307 (cortex_a15_neon_fp_vadd_qqq_vabs_qq): Likewise.
7308 (cortex_a15_neon_fp_vmul_ddd): Likewise.
7309 (cortex_a15_neon_fp_vmul_qqd): Likewise.
7310 (cortex_a15_neon_fp_vmla_ddd): Likewise.
7311 (cortex_a15_neon_fp_vmla_qqq): Likewise.
7312 (cortex_a15_neon_fp_vmla_ddd_scalar): Likewise.
7313 (cortex_a15_neon_fp_vmla_qqq_scalar): Likewise.
7314 (cortex_a15_neon_fp_vrecps_vrsqrts_ddd): Likewise.
7315 (cortex_a15_neon_fp_vrecps_vrsqrts_qqq): Likewise.
7316 (cortex_a15_neon_bp_simple): Likewise.
7317 (cortex_a15_neon_bp_2cycle): Likewise.
7318 (cortex_a15_neon_bp_3cycle): Likewise.
7319 (cortex_a15_neon_vld1_1_2_regs): Likewise.
7320 (cortex_a15_neon_vld1_3_4_regs): Likewise.
7321 (cortex_a15_neon_vld2_2_regs_vld1_vld2_all_lanes): Likewise.
7322 (cortex_a15_neon_vld2_4_regs): Likewise.
7323 (cortex_a15_neon_vld3_vld4): Likewise.
7324 (cortex_a15_neon_vst1_1_2_regs_vst2_2_regs): Likewise.
7325 (cortex_a15_neon_vst1_3_4_regs): Likewise.
7326 (cortex_a15_neon_vst2_4_regs_vst3_vst4): Likewise.
7327 (cortex_a15_neon_vst3_vst4): Likewise.
7328 (cortex_a15_neon_vld1_vld2_lane): Likewise.
7329 (cortex_a15_neon_vld3_vld4_lane" 10
7330 (cortex_a15_neon_vst1_vst2_lane): Likewise.
7331 (cortex_a15_neon_vst3_vst4_lane): Likewise.
7332 (cortex_a15_neon_vld3_vld4_all_lanes): Likewise.
7333 (cortex_a15_neon_ldm_2): Likewise.0
7334 (cortex_a15_neon_stm_2): Likewise.
7335 (cortex_a15_neon_mcr): Likewise.
7336 (cortex_a15_neon_mcr_2_mcrr): Likewise.
7337 (cortex_a15_neon_mrc): Likewise.
7338 (cortex_a15_neon_mrrc): Likewise.
7339 * config/arm/cortex-a15.md (cortex_a15_alu): Update for attribute
7340 change.
7341 (cortex_a15_alu_shift): Likewise.
7342 (cortex_a15_alu_shift_reg): Likewise.
7343 (cortex_a15_mult32): Likewise.
7344 (cortex_a15_mult64): Likewise.
7345 (cortex_a15_block): Likewise.
7346 (cortex_a15_branch): Likewise.
7347 (cortex_a15_load1): Likewise.
7348 (cortex_a15_load3): Likewise.
7349 (cortex_a15_store1): Likewise.
7350 (cortex_a15_store3): Likewise.
7351 (cortex_a15_call): Likewise.
f43245d1 7352 * config/arm/cortex-a5.md (cortex_a5_r2f): Update for attribute change.
003bb7f3
JG
7353 (cortex_a5_f2r): Likewise.
7354 * config/arm/cortex-a53.md (cortex_a53_r2f): Update for attribute
7355 change.
7356 (cortex_a53_f2r): Likewise.
7357 * config/arm/cortex-a7.md
7358 (cortex_a7_branch): Update for attribute change.
7359 (cortex_a7_call): Likewise.
7360 (cortex_a7_alu_imm): Likewise.
7361 (cortex_a7_alu_reg): Likewise.
7362 (cortex_a7_alu_shift): Likewise.
7363 (cortex_a7_mul): Likewise.
7364 (cortex_a7_load1): Likewise.
7365 (cortex_a7_store1): Likewise.
7366 (cortex_a7_load2): Likewise.
7367 (cortex_a7_store2): Likewise.
7368 (cortex_a7_load3): Likewise.
7369 (cortex_a7_store3): Likewise.
7370 (cortex_a7_load4): Likewise.
7371 (cortex_a7_store4): Likewise.
7372 (cortex_a7_fpalu): Likewise.
7373 (cortex_a7_fconst): Likewise.
7374 (cortex_a7_fpmuls): Likewise.
7375 (cortex_a7_neon_mul): Likewise.
7376 (cortex_a7_fpmacs): Likewise.
7377 (cortex_a7_neon_mla: Likewise.
7378 (cortex_a7_fpmuld: Likewise.
7379 (cortex_a7_fpmacd: Likewise.
7380 (cortex_a7_fpfmad: Likewise.
7381 (cortex_a7_fdivs: Likewise.
7382 (cortex_a7_fdivd: Likewise.
7383 (cortex_a7_r2f: Likewise.
7384 (cortex_a7_f2r: Likewise.
7385 (cortex_a7_f_flags: Likewise.
7386 (cortex_a7_f_loads: Likewise.
7387 (cortex_a7_f_loadd: Likewise.
7388 (cortex_a7_f_stores: Likewise.
7389 (cortex_a7_f_stored: Likewise.
7390 (cortex_a7_neon): Likewise.
7391 * config/arm/cortex-a8-neon.md
7392 (cortex_a8_neon_mrc): Update for attribute change.
7393 (cortex_a8_neon_mrrc): Likewise.
7394 (cortex_a8_neon_int_1): Likewise.
7395 (cortex_a8_neon_int_2): Likewise.
7396 (cortex_a8_neon_int_3): Likewise.
7397 (cortex_a8_neon_int_4): Likewise.
7398 (cortex_a8_neon_int_5): Likewise.
7399 (cortex_a8_neon_vqneg_vqabs): Likewise.
7400 (cortex_a8_neon_vmov): Likewise.
7401 (cortex_a8_neon_vaba): Likewise.
7402 (cortex_a8_neon_vaba_qqq): Likewise.
7403 (cortex_a8_neon_vsma): Likewise.
7404 (cortex_a8_neon_mul_ddd_8_16_qdd_16_8_long_32_16_long): Likewise.
7405 (cortex_a8_neon_mul_qqq_8_16_32_ddd_32): Likewise.
f43245d1
UB
7406 (cortex_a8_neon_mul_qdd_64_32_long_qqd_16_ddd_32_scalar_64_32_\
7407 long_scalar): Likewise.
003bb7f3
JG
7408 (cortex_a8_neon_mla_ddd_8_16_qdd_16_8_long_32_16_long): Likewise.
7409 (cortex_a8_neon_mla_qqq_8_16): Likewise.
7410 (cortex_a8_neon_mla_ddd_32_qqd_16_ddd_32_scalar_qdd_64_32_\
7411 long_scalar_qdd_64_32_long): Likewise.
7412 (cortex_a8_neon_mla_qqq_32_qqd_32_scalar): Likewise.
7413 (cortex_a8_neon_mul_ddd_16_scalar_32_16_long_scalar): Likewise.
7414 (cortex_a8_neon_mul_qqd_32_scalar): Likewise.
7415 (cortex_a8_neon_mla_ddd_16_scalar_qdd_32_16_long_scalar): Likewise.
7416 (cortex_a8_neon_shift_1): Likewise.
7417 (cortex_a8_neon_shift_2): Likewise.
7418 (cortex_a8_neon_shift_3): Likewise.
7419 (cortex_a8_neon_vshl_ddd): Likewise.
7420 (cortex_a8_neon_vqshl_vrshl_vqrshl_qqq): Likewise.
7421 (cortex_a8_neon_vsra_vrsra): Likewise.
7422 (cortex_a8_neon_fp_vadd_ddd_vabs_dd): Likewise.
7423 (cortex_a8_neon_fp_vadd_qqq_vabs_qq): Likewise.
7424 (cortex_a8_neon_fp_vsum): Likewise.
7425 (cortex_a8_neon_fp_vmul_ddd): Likewise.
7426 (cortex_a8_neon_fp_vmul_qqd): Likewise.
7427 (cortex_a8_neon_fp_vmla_ddd): Likewise.
7428 (cortex_a8_neon_fp_vmla_qqq): Likewise.
7429 (cortex_a8_neon_fp_vmla_ddd_scalar): Likewise.
7430 (cortex_a8_neon_fp_vmla_qqq_scalar): Likewise.
7431 (cortex_a8_neon_fp_vrecps_vrsqrts_ddd): Likewise.
7432 (cortex_a8_neon_fp_vrecps_vrsqrts_qqq): Likewise.
7433 (cortex_a8_neon_bp_simple): Likewise.
7434 (cortex_a8_neon_bp_2cycle): Likewise.
7435 (cortex_a8_neon_bp_3cycle): Likewise.
7436 (cortex_a8_neon_ldr): Likewise.
7437 (cortex_a8_neon_str): Likewise.
7438 (cortex_a8_neon_vld1_1_2_regs): Likewise.
7439 (cortex_a8_neon_vld1_3_4_regs): Likewise.
7440 (cortex_a8_neon_vld2_2_regs_vld1_vld2_all_lanes): Likewise.
7441 (cortex_a8_neon_vld2_4_regs): Likewise.
7442 (cortex_a8_neon_vld3_vld4): Likewise.
7443 (cortex_a8_neon_vst1_1_2_regs_vst2_2_regs): Likewise.
7444 (cortex_a8_neon_vst1_3_4_regs): Likewise.
7445 (cortex_a8_neon_vst2_4_regs_vst3_vst4): Likewise.
7446 (cortex_a8_neon_vst3_vst4): Likewise.
7447 (cortex_a8_neon_vld1_vld2_lane): Likewise.
7448 (cortex_a8_neon_vld3_vld4_lane): Likewise.
7449 (cortex_a8_neon_vst1_vst2_lane): Likewise.
7450 (cortex_a8_neon_vst3_vst4_lane): Likewise.
7451 (cortex_a8_neon_vld3_vld4_all_lanes): Likewise.
7452 (cortex_a8_neon_mcr): Likewise.
7453 (cortex_a8_neon_mcr_2_mcrr): Likewise.
f43245d1 7454 * config/arm/cortex-a8.md (cortex_a8_alu): Update for attribute change.
003bb7f3
JG
7455 * config/arm/cortex-a9-neon.md (ca9_neon_mrc): Update for attribute
7456 change.
7457 (ca9_neon_mrrc): Likewise.
7458 (cortex_a9_neon_int_1): Likewise.
7459 (cortex_a9_neon_int_2): Likewise.
7460 (cortex_a9_neon_int_3): Likewise.
7461 (cortex_a9_neon_int_4): Likewise.
7462 (cortex_a9_neon_int_5): Likewise.
7463 (cortex_a9_neon_vqneg_vqabs): Likewise.
7464 (cortex_a9_neon_vmov): Likewise.
7465 (cortex_a9_neon_vaba): Likewise.
7466 (cortex_a9_neon_vaba_qqq): Likewise.
7467 (cortex_a9_neon_vsma): Likewise.
7468 (cortex_a9_neon_mul_ddd_8_16_qdd_16_8_long_32_16_long): Likewise.
7469 (cortex_a9_neon_mul_qqq_8_16_32_ddd_32): Likewise.
f43245d1
UB
7470 (cortex_a9_neon_mul_qdd_64_32_long_qqd_16_ddd_32_scalar_64_32_\
7471 long_scalar): Likewise.
003bb7f3
JG
7472 (cortex_a9_neon_mla_ddd_8_16_qdd_16_8_long_32_16_long): Likewise.
7473 (cortex_a9_neon_mla_qqq_8_16): Likewise.
7474 (cortex_a9_neon_mla_ddd_32_qqd_16_ddd_32_scalar_qdd_64_32_\
7475 long_scalar_qdd_64_32_long): Likewise.
7476 (cortex_a9_neon_mla_qqq_32_qqd_32_scalar): Likewise.
7477 (cortex_a9_neon_mul_ddd_16_scalar_32_16_long_scalar): Likewise.
7478 (cortex_a9_neon_mul_qqd_32_scalar): Likewise.
7479 (cortex_a9_neon_mla_ddd_16_scalar_qdd_32_16_long_scalar): Likewise.
7480 (cortex_a9_neon_shift_1): Likewise.
7481 (cortex_a9_neon_shift_2): Likewise.
7482 (cortex_a9_neon_shift_3): Likewise.
7483 (cortex_a9_neon_vshl_ddd): Likewise.
7484 (cortex_a9_neon_vqshl_vrshl_vqrshl_qqq): Likewise.
7485 (cortex_a9_neon_vsra_vrsra): Likewise.
7486 (cortex_a9_neon_fp_vadd_ddd_vabs_dd): Likewise.
7487 (cortex_a9_neon_fp_vadd_qqq_vabs_qq): Likewise.
7488 (cortex_a9_neon_fp_vsum): Likewise.
7489 (cortex_a9_neon_fp_vmul_ddd): Likewise.
7490 (cortex_a9_neon_fp_vmul_qqd): Likewise.
7491 (cortex_a9_neon_fp_vmla_ddd): Likewise.
7492 (cortex_a9_neon_fp_vmla_qqq): Likewise.
7493 (cortex_a9_neon_fp_vmla_ddd_scalar): Likewise.
7494 (cortex_a9_neon_fp_vmla_qqq_scalar): Likewise.
7495 (cortex_a9_neon_fp_vrecps_vrsqrts_ddd): Likewise.
7496 (cortex_a9_neon_fp_vrecps_vrsqrts_qqq): Likewise.
7497 (cortex_a9_neon_bp_simple): Likewise.
7498 (cortex_a9_neon_bp_2cycle): Likewise.
7499 (cortex_a9_neon_bp_3cycle): Likewise.
7500 (cortex_a9_neon_ldr): Likewise.
7501 (cortex_a9_neon_str): Likewise.
7502 (cortex_a9_neon_vld1_1_2_regs): Likewise.
7503 (cortex_a9_neon_vld1_3_4_regs): Likewise.
7504 (cortex_a9_neon_vld2_2_regs_vld1_vld2_all_lanes): Likewise.
7505 (cortex_a9_neon_vld2_4_regs): Likewise.
7506 (cortex_a9_neon_vld3_vld4): Likewise.
7507 (cortex_a9_neon_vst1_1_2_regs_vst2_2_regs): Likewise.
7508 (cortex_a9_neon_vst1_3_4_regs): Likewise.
7509 (cortex_a9_neon_vst2_4_regs_vst3_vst4): Likewise.
7510 (cortex_a9_neon_vst3_vst4): Likewise.
7511 (cortex_a9_neon_vld1_vld2_lane): Likewise.
7512 (cortex_a9_neon_vld3_vld4_lane): Likewise.
7513 (cortex_a9_neon_vst1_vst2_lane): Likewise.
7514 (cortex_a9_neon_vst3_vst4_lane): Likewise.
7515 (cortex_a9_neon_vld3_vld4_all_lanes): Likewise.
7516 (cortex_a9_neon_mcr): Likewise.
7517 (cortex_a9_neon_mcr_2_mcrr): Likewise.
7518 * config/arm/cortex-a9.md (cortex_a9_dp): Update for attribute change.
7519 (cortex_a9_fps): Likewise.
7520 * config/arm/cortex-m4-fpu.md (cortex_m4_vmov_2): Update for attribute
7521 change.
7522 (cortex_m4_fmuls): Likewise.
7523 * config/arm/cortex-r4f.md (cortex_r4_mcr): Update for attribute
7524 change.
7525 (cortex_r4_mrc): Likewise.
7526 * config/arm/iterators.md: Update comment referring to neon_type.
f43245d1 7527 * config/arm/iwmmxt.md (iwmmxt_arm_movdi): Update for attribute change.
003bb7f3 7528 (iwmmxt_movsi_insn): Likewise.
f43245d1
UB
7529 * config/arm/marvell-pj4.md (pj4_vfp_to_core): Update for
7530 attribute change.
003bb7f3
JG
7531 (pj4_core_to_vfp): Likewise.
7532 * config/arm/neon-schedgen.ml (emit_insn_reservations): Update for
7533 attribute change.
7534 * config/arm/vfp11.md (vfp_fload): Update for attribute change.
7535 (vfp_fstore): Likewise.
7536 * doc/md.texi: Change references to neon_type to refer to type.
7537
fbe0dc78
DS
75382013-09-04 Dodji Seketeli <dodji@redhat.com>
7539
7540 * tree.h (DECL_BUILT_IN): Fix typo in comment.
7541
c167bc5b
DE
75422013-09-04 David Edelsohn <dje.gcc@gmail.com>
7543
7544 * config/rs6000/rs6000.h (ASM_OUTPUT_DEF_FROM_DECLS): Only emit
7545 lglobl if not weak.
7546
035cb59f
ER
75472013-09-04 Easwaran Raman <eraman@google.com>
7548
7549 PR middle-end/57370
7550 * tree-ssa-reassoc.c (get_stmt_uid_with_default): New function,
7551 (build_and_add_sum): Use it.
7552 (appears_later_in_bb): Simplify code.
7553
71191083
TJ
75542013-09-04 Teresa Johnson <tejohnson@google.com>
7555
7556 * dumpfile.c (dump_finish): Don't close stderr/stdout.
7557
0c488213
JG
75582013-09-04 James Greenhalgh <james.greenhalgh@arm.com>
7559
7560 * config/aarch64/arm_neon.h (vaddvq_<su>64): Fix return types.
7561
61a74079
JH
75622013-09-04 Jan Hubicka <jh@suse.cz>
7563
7564 * Makefile.in (ipa-devirt.o): Add dependency on diagnostic.h
7565 * ipa-devirt.c: Include diganostic.h
7566 (odr_type_d): Add types and types_set.
7567 (hash_type_name): Work for types with vtables during LTO.
7568 (odr_hasher::remove): Fix comment; destroy types_set.
7569 (add_type_duplicate): New function,
7570 (get_odr_type): Use it.
7571 (dump_type_inheritance_graph): Dump type duplicates.
7572 * ipa.c (symtab_remove_unreachable_nodes): Build type inheritance
7573 graph.
7574 * tree.c (types_same_for_odr): Give exact answers on types with
7575 virtual tables.
7576
24ba8195
DS
75772013-09-04 Dodji Seketeli <dodji@redhat.com>
7578
7579 * tree.h (DECL_BUILT_IN, DECL_IS_BUILTIN): Add more comments
7a5911d3 7580 explaining their differences.
24ba8195 7581
0a04869e
SKS
75822013-09-04 Sandeep Kumar Singh<Sandeep.Singh2@kpitcummins.com>
7583
7584 * config/rx/rx.h: Add option -mcpu for target variants RX100 and RX200.
7585
3b18bc42
JL
75862013-09-03 Jeff Law <law@redhat.com>
7587
7588 * tree-ssa-threadedge.c (thread_across_edge): Record entire path
7589 when not threading through a joiner block. Pass joiner/no joiner
7590 state to register_jump_thread.
7591 * tree-ssa-threadupdate.c (register_jump_thread): Get joiner/no joiner
7592 state from argument rather than implying on path length.
7593 Dump the entire jump thread path into debugging dump.
7594 * tree-flow.h (register_jump_thread): Update prototype.
7595
ae84e151
XDL
75962013-08-29 Xinliang David Li <davidxl@google.com>
7597
7598 * tree-vect-data-refs.c (vect_compute_data_ref_alignment):
7599 Remove a trivial gcc_assert.
7600
c716e67f
XDL
76012013-08-29 Xinliang David Li <davidxl@google.com>
7602
7603 * tree-vect-slp.c (destroy_bb_vec_info): Data ref cleanup.
7604 * tree-vect-loop.c (destroy_bb_vec_info): Ditto.
7605 * tree-vect-data-refs.c (vect_compute_data_ref_alignment):
7606 Delay base decl alignment adjustment.
7607 * tree-vectorizer.c (vect_destroy_datarefs): New function.
7608 * tree-vectorizer.h: New data structure.
7609 (set_dr_misalignment): New function.
7610 (dr_misalignment): Ditto.
7611 * tree-vect-stmts.c (vectorizable_store): Ensure alignment.
7612 (vectorizable_load): Ditto.
7613 (ensure_base_align): New function.
7614 (vectorize_loops): Add dbg_cnt support.
7615 (execute_vect_slp): Ditto.
7616 * dbgcnt.def: New debug counter.
7617 * Makefile: New dependency.
7618
db8800bc
MI
76192013-09-03 Meador Inge <meadori@codesourcery.com>
7620
7621 Revert:
7622
7623 2013-08-30 Meador Inge <meadori@codesourcery.com>
7624
7625 * tree-vrp.c (check_array_ref): Bail out on zero-length arrays.
7626
0d9b4f5c
DE
76272013-09-03 David Edelsohn <dje.gcc@gmail.com>
7628
7629 * config/rs6000/rs6000.h (ASM_OUTPUT_DEF_FROM_DECLS): Emit lglobl for
7630 function descriptor.
7631
29e810b9
RB
76322013-09-03 Richard Biener <rguenther@suse.de>
7633
7634 * tree-affine.c (add_elt_to_tree): Fix association issue,
7635 avoid useless converts and make sure to always return a
7636 properly typed result.
7637
dbfc2894
RB
76382013-09-03 Richard Biener <rguenther@suse.de>
7639
7640 PR middle-end/57656
7641 * fold-const.c (negate_expr_p): Fix division case.
7642 (negate_expr): Likewise.
7643
fa3bf4ce
RB
76442013-09-03 Richard Biener <rguenther@suse.de>
7645
7646 PR lto/58285
7647 * tree-streamer-out.c: Include tm.h.
7648 * Makefile.in (tree-streamer-out.o): Depend on $(TM_H).
7649
a64bbb3f
JH
76502013-09-03 Jan Hubicka <jh@suse.cz>
7651
7652 * tree-profile.c (tree_profiling): Cleanup CFG when done.
7653
93ed07e2
AM
76542013-09-03 Alan Modra <amodra@gmail.com>
7655
7656 * config.gcc (powerpc*-*-linux*): Add support for little-endian
7657 multilibs to big-endian target and vice versa.
7658 * config/rs6000/t-linux64: Use := assignment on all vars.
7659 (MULTILIB_EXTRA_OPTS): Remove fPIC.
7660 (MULTILIB_OSDIRNAMES): Specify using mapping from multilib_options.
7661 * config/rs6000/t-linux64le: New file.
7662 * config/rs6000/t-linux64bele: New file.
7663 * config/rs6000/t-linux64lebe: New file.
7664
07d6cd64
JH
76652013-09-02 Jan Hubicka <jh@suse.cz>
7666
7667 * ipa-inline-transform.c (inline_transform): Do not
7668 optimize_inline_calls when not optimizing.
7669
8e4c9a10
JH
76702013-09-02 Jan Hubicka <jh@suse.cz>
7671
7672 * lto-symtab.c (lto_symtab_merge_symbols): Add comments; merge
7673 duplicated nodes for assembler names.
7674 * symtab.c (symtab_unregister_node): Do not attempt to unlink
7675 hard registers from assembler name hash.
7676
c91061e6
JH
76772013-09-02 Jan Hubicka <jh@suse.cz>
7678
7679 * ipa-split.c (execute_split_functions): Split externally visible
7680 functions called once.
7681
2fd0985c
MJ
76822013-09-02 Martin Jambor <mjambor@suse.cz>
7683
7684 PR ipa/58106
7685 * ipa-prop.c (ipa_edge_duplication_hook): Always put new rdesc to the
7686 linked list. When finding the correct duplicate, also consider also
7687 the caller in additon to its inlined_to node.
7688
91bd4114
JG
76892013-09-02 James Greenhalgh <james.greenhalgh@arm.com>
7690
7691 * config/aarch64/aarch64-simd-builtins.def
7692 (dup_lane_scalar): Remove.
7693 * config/aarch64/aarch64-simd.md
7694 (aarch64_simd_dup): Add 'w->w' alternative.
7695 (aarch64_dup_lane<mode>): Allow for VALL.
7696 (aarch64_dup_lane_scalar<mode>): Remove.
7697 (aarch64_dup_lane_<vswap_width_name><mode>): New.
7698 (aarch64_get_lane_signed<mode>): Add w->w altenative.
7699 (aarch64_get_lane_unsigned<mode>): Likewise.
7700 (aarch64_get_lane<mode>): Likewise.
7701 * config/aarch64/aarch64.c (aarch64_evpc_dup): New.
7702 (aarch64_expand_vec_perm_const_1): Use aarch64_evpc_dup.
7703 * config/aarch64/iterators.md (VSWAP_WIDTH): New.
7704 (VCON): Change container of V2SF.
7705 (vswap_width_name): Likewise.
7706 * config/aarch64/arm_neon.h
7707 (__aarch64_vdup_lane_any): New.
7708 (__aarch64_vdup<q>_lane<q>_<fpsu><8,16,32,64>): Likewise.
7709 (vdup<q>_n_<psuf><8,16,32,64>): Convert to C implementation.
7710 (vdup<q>_lane<q>_<fpsu><8,16,32,64>): Likewise.
7711
d617d2d8
EB
77122013-09-02 Eric Botcazou <ebotcazou@adacore.com>
7713
7714 PR middle-end/56382
7715 * expr.c (emit_move_complex): Do not move complex FP values as parts if
7716 the source or the destination is a single hard register.
7717
63fdb7be
RB
77182013-09-02 Richard Biener <rguenther@suse.de>
7719
7720 PR middle-end/57511
7721 * tree-scalar-evolution.c (instantiate_scev_name): Allow
7722 non-linear SCEVs.
7723
78de2333
RB
77242013-09-02 Richard Biener <rguenther@suse.de>
7725
7726 * tree-affine.c (add_elt_to_tree): Avoid converting all pointer
7727 arithmetic to sizetype.
7728
85ff4ec6
BC
77292013-09-02 Bin Cheng <bin.cheng@arm.com>
7730
7731 * tree-ssa-loop-ivopts.c (set_autoinc_for_original_candidates):
7732 Find auto-increment use both before and after candidate.
7733
576016fe
MP
77342013-09-02 Marek Polacek <polacek@redhat.com>
7735
7736 * Makefile.in (ubsan.o): Add $(TM_P_H) dependency.
7737
08f835dc
JH
77382013-09-01 Jan Hubicka <jh@suse.cz>
7739
7740 * Makefile.in: Add ipa-profile.o
7741 (ipa.o, ipa-devrit.o, ipa-inline-analysis.o): Adjust dependencies.
7742 * cgraph.c (struct cgraph_propagate_frequency_data,
7743 cgraph_propagate_frequency_1, cgraph_propagate_frequency): Move to
7744 ipa-profile.c; replace cgraph_ by ipa_ prefix.
7745 * cgraph.h (cgraph_propagate_frequency): Remove.
f43245d1
UB
7746 * ipa-inline-analysis.c: Include ipa-utils.h;
7747 drop duplicated cfgloop.h.
08f835dc
JH
7748 (inline_update_callee_summaries): Update.
7749 * ipa-profile.c: New file.
7750 * ipa-utils.h (ipa_propagate_frequency): Declare.
7751 * ipa.c: Do not include pointer-set.h, hash-table.h, lto-streamer.h,
f43245d1 7752 data-streamer.h, value-prof.h.
08f835dc
JH
7753 (symtab_remove_unreachable_nodes): Update profile.
7754 (struct histogram_entry, histogram, histogram_pool, histogram_hash,
7755 account_time_size, cmp_counts, dump_histogram,
7756 ipa_profile_generate_summary, ipa_profile_write_summary,
7757 ipa_profile_read_summary, ipa_profile, gate_ipa_profile,
7758 pass_data_ipa_profile, pass_ipa_profile, make_pass_ipa_profile):
f43245d1 7759 Move to ipa-profile.c.
08f835dc 7760
cec8583c
JDA
77612013-09-01 John David Anglin <danglin@gcc.gnu.org>
7762
7763 * config/pa/pa.md: Allow "const 0" operand 1 in "scc" insns.
7764
bbc9396b
JH
77652013-09-01 Jan Hubicka <jh@suse.cz>
7766
7767 * common.opt (fdevirtualize-speculatively): New function.
7768 * invoke.texi (fdevirtualize-speculatively): Document.
7769 * ipa-devirt.c: Include ipa-inline.h
7770 (likely_target_p): New function.
7771 (ipa_devirt): New function.
7772 (gate_ipa_devirt): New function.
7773 (pass_data_ipa_devirt): New static var.
7774 (pass_ipa_devirt): Likewise.
7775 (make_pass_ipa_devirt): New function.
7776 * opts.c (default_options): Add OPT_fdevirtualize_speculatively.
7777 (common_handle_option): Disable devirtualization when
7778 value range profiling is available.
7779 * passes.def (pass_ipa_devirt): Add.
7780 * timever.def (TV_IPA_DEVIRT): New timevar.
f43245d1 7781 * tree-pass.h (make_pass_ipa_devirt):
bbc9396b 7782
0cea211e
IS
77832013-09-01 Iain Sandoe <iain@codesourcery.com>
7784
7785 * config/darwin.h (LINK_COMMAND_SPEC_A): Revise sanitizer specs to
7786 include sanitize(undefined).
7787
5e351e96
DN
77882013-08-31 Diego Novillo <dnovillo@google.com>
7789
7790 * Makefile.in (TREE_CORE_H): Define.
7791 (TREE_H): Use.
7792 (GTFILES): Add tree-core.h.
7793 * builtins.c (built_in_class_names): Use BUILT_IN_LAST to
7794 size the array.
7795 * tree-core.h: New file.
7796 Move all data structures, enum, typedefs, global
7797 declarations and constants from ...
7798 * tree.h: ... here.
7799
63bf9a90
JH
78002013-08-31 Jan Hubicka <jh@suse.cz>
7801
7802 * bulitins.c (expand_builtin): Do not early exit for gcov
7803 instrumented functions.
7804
5c0abd6a
MP
78052013-08-31 Marek Polacek <polacek@redhat.com>
7806
7807 * ubsan.c: Include tm_p.h.
7808
4843f032
JH
78092013-08-31 Jan Hubicka <jh@suse.cz>
7810
8356c89c
JH
7811 * gimple-streamer-in.c (input_gimple_stmt): Silence parameter unused
7812 warning.
7813
4843f032
JH
7814 * cgraph.c (cgraph_get_body): Update call of lto_input_function_body.
7815 * gimple-streamer-in.c (input_gimple_stmt): Move sanity check to ...
5e581212 7816 * tree-cfg.c (verify_gimple_label): ... here.
4843f032
JH
7817 * ipa-utils.c: Include lto-streamer.h, ipa-inline.h
7818 (ipa_merge_profiles): New function.
7819 * lto-streamer-in.c (lto_read_body): Take node instead of fn_decl.
7820 (lto_input_function_body): Likewise.
5e581212
JH
7821 * ipa-utils.h (ipa_merge_profiles): Declare.
7822 * lto-streamer.h (lto_input_function_body): Update prototype.
7823 (emit_label_in_global_context_p): Remove.
7824 * lto-symtab.c: Include ipa-utils.h
7825 (lto_cgraph_replace_node): Use ipa_merge_profiles.
4843f032 7826
57292ce9
JH
78272013-08-31 Jan Hubicka <jh@suse.cz>
7828
7829 * cgraph.c (cgraph_speculative_call_info): Fix ref lookup
7830
b9cbfeeb
JH
78312013-08-31 Jan Hubicka <jh@suse.cz>
7832
7833 * basic-block.h (apply_scale): Make scale parmeter gcov_type.
7834
41dedebd
UB
78352013-08-31 Uros Bizjak <ubizjak@gmail.com>
7836
7837 * config/alpha/alpha.c (alpha_emit_conditional_move): Update
7838 "cmp" RTX before signed_comparison_operator check to account
7839 for "code" changes.
7840
06d65050
JH
78412013-08-30 Jan Hubicka <jh@suse.cz>
7842
41dedebd 7843 * ipa-prop.c (ipa_set_jf_known_type): Check that we add only records.
06d65050
JH
7844 (detect_type_change_1): Rename to ...
7845 (detect_type_change): ... this one; early return on non-polymorphic
7846 types.
41dedebd 7847 (detect_type_change_ssa): Add comp_type parameter; update
06d65050
JH
7848 use of detect_type_change.
7849 (compute_complex_assign_jump_func): Add param_type parameter;
7850 update use of detect_type_change_ssa.
7851 (compute_complex_ancestor_jump_func): Likewise.
7852 (ipa_get_callee_param_type): New function.
7853 (ipa_compute_jump_functions_for_edge): Compute parameter type;
7854 update calls to the jump function computation functions.
7855
600b5b1d
TJ
78562013-08-30 Teresa Johnson <tejohnson@google.com>
7857 Steven Bosscher <steven@gcc.gnu.org>
7858
7859 * cfgrtl.c (fixup_new_cold_bb): New routine.
7860 (commit_edge_insertions): Invoke fixup_partitions.
7861 (find_partition_fixes): New routine.
7862 (fixup_partitions): Ditto.
7863 (verify_hot_cold_block_grouping): Update comments.
7864 (rtl_verify_edges): Invoke find_partition_fixes.
7865 (rtl_verify_bb_pointers): Update comments.
7866 (rtl_verify_bb_layout): Ditto.
7867 * basic-block.h (probably_never_executed_edge_p): Declare.
7868 (fixup_partitions): Ditto.
7869 * cfgcleanup.c (try_optimize_cfg): Invoke fixup_partitions.
7870 * bb-reorder.c (sanitize_hot_paths): New function.
7871 (find_rarely_executed_basic_blocks_and_crossing_edges): Invoke
7872 sanitize_hot_paths.
7873 * predict.c (probably_never_executed_edge_p): New routine.
7874 * cfg.c (check_bb_profile): Add partition insanity warnings.
7875
0a6ea5c9
MI
78762013-08-30 Meador Inge <meadori@codesourcery.com>
7877
7878 * tree-vrp.c (check_array_ref): Bail out on zero-length arrays.
7879
de5a5fa1
MP
78802013-08-30 Marek Polacek <polacek@redhat.com>
7881
7882 * Makefile.in (ubsan.o): Add.
7883 (c-family/c-ubsan.o): Add.
7884 (builtins.o): Add ubsan.h dependency.
7885 * ubsan.h: New file.
7886 * ubsan.c: New file.
7887 * common.opt: Add -fsanitize=undefined option.
7888 (flag_sanitize): Add variable.
7889 (fsanitize=): Add option. Add Driver.
7890 (fsanitize=thread): Remove option.
7891 (fsanitize=address): Likewise.
7892 (static-libubsan): New option.
7893 * doc/invoke.texi: Document the new flag and -static-libubsan.
7894 * sanitizer.def (DEF_SANITIZER_BUILTIN): Define.
7895 (BUILT_IN_UBSAN_HANDLE_BUILTIN_UNREACHABLE): Define.
7896 * builtin-attrs.def (ATTR_COLD): Define.
7897 (ATTR_COLD_NOTHROW_LEAF_LIST): Define.
7898 * builtins.def (BUILT_IN_UBSAN_HANDLE_DIVREM_OVERFLOW,
7899 BUILT_IN_UBSAN_HANDLE_SHIFT_OUT_OF_BOUNDS): Define.
7900 * flag-types.h (sanitize_code): New enum.
7901 * opts.c (common_handle_option): Parse command line arguments
7902 of -fsanitize=. Add -fsanitize=unreachable option.
7903 * varasm.c (get_variable_section): Adjust.
7904 (assemble_noswitch_variable): Likewise.
7905 (assemble_variable): Likewise.
7906 (output_constant_def_contents): Likewise.
7907 (categorize_decl_for_section): Likewise.
7908 (place_block_symbol): Likewise.
7909 (output_object_block): Likewise.
7910 * builtins.def: Likewise.
7911 * toplev.c (compile_file): Likewise.
7912 (process_options): Likewise.
7913 * cppbuiltin.c: Likewise.
7914 * tsan.c (tsan_pass): Likewise.
7915 (tsan_gate): Likewise.
7916 (tsan_gate_O0): Likewise.
7917 * cfgexpand.c (partition_stack_vars): Likewise.
7918 (expand_stack_vars): Likewise.
7919 (defer_stack_allocation): Likewise.
7920 (expand_used_vars): Likewise.
7921 * cfgcleanup.c (old_insns_match_p): Likewise.
7922 * asan.c (asan_finish_file): Likewise.
7923 (asan_instrument): Likewise.
7924 (gate_asan): Likewise.
7925 (initialize_sanitizer_builtins): Build BT_FN_VOID_PTR_PTR_PTR.
7926 (ATTR_COLD_NOTHROW_LEAF_LIST): Define.
7927 (asan_global_struct): Use pointer_sized_int_node instead
7928 calling build_nonstandard_integer_type.
7929 (initialize_sanitizer_builtins): Likewise.
7930 (asan_finish_file): Likewise.
7931 * gcc.c: Document %{%:function(args):X}.
7932 (static_spec_functions): Add sanitize.
7933 (handle_spec_function): Add retval_nonnull argument and if non-NULL,
7934 store funcval != NULL there.
7935 (do_spec_1): Adjust handle_spec_function caller.
7936 (handle_braces): Allow %:function(args) as condition.
7937 (sanitize_spec_function): New function.
7938 (ADD_STATIC_LIBUBSAN_LIBS): Define.
7939 (LIBUBSAN_SPEC): Likewise.
7940 (LIBUBSAN_EARLY_SPEC): Likewise.
7941 (SANITIZER_SPEC): Handle libubsan.
7942 (SANITIZER_EARLY_SPEC): Likewise.
7943 * config/darwin.h (LINK_COMMAND_SPEC_A): Use %:sanitize(address)
7944 instead of fsanitize=address.
7945 * config/arm/linux-eabi.h (ASAN_CC1_SPEC): Use %:sanitize(address)
7946 instead of fsanitize=address*.
7947 * builtins.c: Include ubsan.h.
7948 (fold_builtin_0): Instrument __builtin_unreachable.
7949 * config/rs6000/rs6000.h (FRAME_GROWS_DOWNWARD): Use flag_sanitize
7950 instead of flag_asan.
7951 * tree.h (enum tree_index): Add TI_POINTER_SIZED_TYPE.
7952 (pointer_sized_int_node): Define.
41dedebd 7953 * tree.c (build_common_tree_nodes): Initialize pointer_sized_int_node.
de5a5fa1 7954
f07f30cf
MS
79552013-08-30 Mike Stump <mikestump@comcast.net>
7956
7957 * doc/install.texi (Prerequisites): Note regression in Tcl 8.6
7958 with RE patterns.
7959
8a41354f
JH
79602013-08-29 Jan Hubicka <jh@suse.cz>
7961
7962 * cgraph.c (cgraph_function_body_availability): Handle weakref
7963 correctly.
7964 * passes.def: Remove pass_fixup_cfg.
7965 * ipa-inline.c (ipa_inline): When not optimizing, do not inline;
7966 track when we need to remove functions.
7967 (gate_ipa_inline): Execute inlining always; add comment why.
7968 (pass_data_ipa_inline): Remove TODO_remove_functions.
7969 * ipa-inline-analysis.c (inline_generate_summary): When not optimizing
7970 do not produce summaries.
7971 * symtab.c (change_decl_assembler_name): Handle renaming of weakrefs.
7972 (symtab_nonoverwritable_alias): Assert we are not called on weakref.
7973 * varpool.c (cgraph_variable_initializer_availability): Fix weakrefs,
7974 constant pool and vtable.
7975
dc8d7a0f
TB
79762013-08-30 Tejas Belagod <tejas.belagod@arm.com>
7977
41dedebd
UB
7978 * config/aarch64/arm_neon.h (__AARCH64_UINT64_C, __AARCH64_INT64_C):
7979 New arm_neon.h's internal macros to specify 64-bit constants.
7980 Avoid using stdint.h's macros.
dc8d7a0f 7981
ac7eacd2
JR
79822013-08-30 Joern Rennecke <joern.rennecke@embecosm.com>
7983
7984 * recog.c (verify_changes): Verify that changes[i].old is non-zero
7985 before applying REG_P.
7986
8b29fd4e
JJ
79872013-08-30 Jakub Jelinek <jakub@redhat.com>
7988
7989 PR tree-optimization/58277
7990 * tree-ssa-strlen.c (strlen_enter_block): If do_invalidate gave up
7991 after seeing too many stmts with vdef in between dombb and current
7992 bb, invalidate everything.
7993
26d75703
RB
79942013-08-30 Richard Biener <rguenther@suse.de>
7995
7996 * fold-const.c (fold_single_bit_test): Fix overflow test.
7997
f7e088e7
EB
79982013-08-30 Eric Botcazou <ebotcazou@adacore.com>
7999
8000 * function.c (assign_parm_setup_reg): For a parameter passed by pointer
8001 and which can live in a register, always retrieve the value on entry.
8002 * var-tracking.c (add_stores): Treat the copy on entry for a parameter
8003 passed by invisible reference specially.
8004 (emit_notes_in_bb) <MO_VAL_USE>: Emit notes before the instruction.
8005 (vt_add_function_parameter): Correctly deal with a parameter passed by
8006 invisible reference.
8007
80082013-08-30 Jan Hubicka <jh@suse.cz>
47500756
JH
8009
8010 * tree.c (set_call_expr_flags): Fix handling of TM_PURE.
8011
6e8dad05
RB
80122013-08-30 Richard Biener <rguenther@suse.de>
8013
8014 PR tree-optimization/58228
8015 * tree-vect-data-refs.c (vect_analyze_data_ref_access): Do not
8016 allow invariant loads in nested loop vectorization.
8017
062ef2c8
RB
80182013-08-30 Richard Biener <rguenther@suse.de>
8019
8020 PR tree-optimization/58223
8021 * tree-loop-distribution.c (has_anti_dependence): Rename to ...
8022 (has_anti_or_output_dependence): ... this and adjust to also
8023 look for output dependences.
8024 (mark_nodes_having_upstream_mem_writes): Adjust.
8025 (rdg_flag_uses): Likewise.
8026
7a764c60
RB
80272013-08-30 Richard Biener <rguenther@suse.de>
8028
8029 PR tree-optimization/58010
8030 * tree-vect-loop.c (vect_create_epilog_for_reduction): Remove
8031 assert that we have a loop-closed PHI.
8032
82e9d642
JH
80332013-08-29 Jan Hubicka <jh@suse.cz>
8034
8035 * lto-symtab.c (lto_cgraph_replace_node): Free decl_in_state.
8036 * cgraph.c (cgraph_release_function_body): Free decl_in_state.
8037 * lto-section-in.c (lto_free_function_in_decl_state): New function.
8038 (lto_free_function_in_decl_state_for_node): New function.
8039
9cc1fb4b
XDL
80402013-08-29 Xinliang David Li <davidxl@google.com>
8041
41dedebd 8042 * loop-unroll.c (report_unroll_peel): Minor message change.
9cc1fb4b
XDL
8043 * tree-vect-loop-manip.c (vect_do_peeling_for_alignment):
8044 Emit alignment peeling message with default -fopt-info.
8045 (vect_loop_versioning): Emit loop version info message.
41dedebd 8046 * tree-vectorizer.c (vectorize_loops): Minor message change.
9cc1fb4b
XDL
8047 (execute_vect_slp): Ditto.
8048
440a5082
EB
80492013-08-29 Eric Botcazou <ebotcazou@adacore.com>
8050
8051 * cgraphclones.c (cgraph_create_virtual_clone): Compute the DECL_NAME
8052 of the clone from the DECL_NAME of the original function.
8053
3fa3690d
OE
80542013-08-29 Oleg Endo <olegendo@gcc.gnu.org>
8055
8056 * passes.c (register_pass): Add overload.
8057 * tree-pass.h (register_pass): Forward declare it. Add comment.
8058
0170f33c
JH
80592013-08-29 Jan Hubicka <jh@suse.cz>
8060
41dedebd
UB
8061 * lto-streamer-out.c (hash_tree): Stream DECL_FINAL_P,
8062 DECL_CXX_CONSTRUCTOR_P, DECL_CXX_DESTRUCTOR_P and TYPE_FINAL_P.
8063 * lto-streamer-in.c (unpack_ts_decl_with_vis_value_fields): Stream
8064 DECL_FINAL_P, DECL_CXX_CONSTRUCTOR_P and DECL_CXX_DESTRUCTOR_P.
0170f33c
JH
8065 (unpack_ts_type_common_value_fields): Stream TYPE_FINAL_P.
8066 * tree-streamer-out.c (pack_ts_decl_with_vis_value_fields):
41dedebd 8067 Add DECL_FINAL_P, DECL_CXX_CONSTRUCTOR_P and DECL_CXX_DESTRUCTOR_P.
0170f33c
JH
8068 (pack_ts_type_common_value_fields): Add TYPE_FINAL_P.
8069
103ff0d6
TJ
80702013-08-29 Teresa Johnson <tejohnson@google.com>
8071
8072 * dumpfile.c (dump_loc): Output column number.
8073 * dumpfile.h (OPTGROUP_OTHER): Add and enable under OPTGROUP_ALL.
8074 * doc/invoke.texi: Document optall -fopt-info flag.
8075 * profile.c (read_profile_edge_counts): Use new dump framework.
8076 (compute_branch_probabilities): Ditto.
8077 * passes.c (pass_manager::register_one_dump_file): Use OPTGROUP_OTHER
8078 when pass not in any opt group.
8079 * pass_manager.h (pass_manager::get_pass_profile): New method.
8080 * value-prof.c (check_counter): Use new dump framework.
8081 (check_ic_target): Ditto.
8082 * coverage.c (get_coverage_counts): Ditto.
8083 (coverage_init): Setup new dump framework.
8084
301bbc16
RB
80852013-08-29 Richard Biener <rguenther@suse.de>
8086
8087 PR tree-optimization/58246
8088 * tree-ssa-dce.c (mark_aliased_reaching_defs_necessary_1): Properly
8089 handle the dominance check inside a basic-block.
8090
1b275000
RB
80912013-08-29 Richard Biener <rguenther@suse.de>
8092
8093 PR middle-end/57287
8094 * tree-ssa-copy.c (may_propagate_copy): Allow propagating
8095 of default defs that appear in abnormal PHI nodes.
8096
6b1184ba
RB
80972013-08-29 Richard Biener <rguenther@suse.de>
8098
8099 PR tree-optimization/57685
8100 * tree-vrp.c (register_edge_assert_for_1): Recurse only for
8101 single-use operands to avoid exponential complexity.
8102
1b08b734
DC
81032013-08-28 Dehao Chen <dehao@google.com>
8104
8105 * ipa-inline.c (edge_badness): Fix integer underflow.
8106
48a3fa69
UB
81072013-08-28 Uros Bizjak <ubizjak@gmail.com>
8108
8109 * gtm-builtins.def (_ITM_free): Declare leaf.
8110
215f73e6
JJ
81112013-08-28 Jakub Jelinek <jakub@redhat.com>
8112
8113 PR target/58067
8114 * config/i386/i386.md (*tls_global_dynamic_64_largepic): New insn.
8115 (*tls_local_dynamic_base_64_largepic): Likewise.
8116 (tls_global_dynamic_64_<mode>, tls_local_dynamic_base_64_<mode>):
8117 Remove predicate from call operand.
8118 * config/i386/i386.c (ix86_tls_get_addr): For -mcmodel=large -fpic
8119 return sum of pic_offset_table_rtx and UNSPEC_PLTOFF of the symbol.
8120
34554d1a
JL
81212013-08-28 Jeff Law <law@redhat.com>
8122
8123 * tree-ssa-threadedge.c (thread_around_empty_block): Remove
8124 checks for the number of predecessors and successors allowed.
8125 * tree-ssa-threadupdate.c (mark_threaded_blocks): Ignore requests
8126 which require copying a joiner block if there is a request which
8127 is a subpath that requires no joiner block copying.
8128
c01c111b
JH
81292013-08-28 Jan Hubicka <jh@suse.cz>
8130
8131 * lto-streamer-out.c (DFS_write_tree_body): Drop
8132 BINFO_INHERITANCE_CHAIN, BINFO_SUBVTT_INDEX and BINFO_VPTR_INDEX.
8133 (hash_tree): Do not hash DECL_DEFER_OUTPUT, BINFO_INHERITANCE_CHAIN,
8134 BINFO_SUBVTT_INDEX, BINFO_VPTR_INDEX, DECL_IN_TEXT_SECTION.
8135 * tree-streamer-in.c (unpack_ts_decl_common_value_fields):
8136 Do not read DECL_ERROR_ISSUED.
8137 (unpack_ts_decl_with_vis_value_fields): Do not read
8138 DECL_DEFER_OUTPUT.
48a3fa69
UB
8139 (lto_input_ts_binfo_tree_pointers): Do not read
8140 BINFO_INHERITANCE_CHAIN, BINFO_SUBVTT_INDEX, BINFO_VPTR_INDEX
c01c111b
JH
8141 * tree-streamer-out.c (pack_ts_decl_common_value_fields): Do not
8142 write DECL_ERROR_ISSUED..
8143 (pack_ts_decl_with_vis_value_fields): Do not write
8144 DECL_DEFER_OUTPUT.
8145 (write_ts_binfo_tree_pointers): Do not read BINFO_INHERITANCE_CHAIN,
48a3fa69 8146 BINFO_SUBVTT_INDEX, BINFO_VPTR_INDEX.
c01c111b
JH
8147 * print-tree.c (print_node): Do not print DECL_ERROR_ISSUED.
8148 * tree.h (tree_decl_common): Update comment.
8149 (DECL_ERROR_ISSUED): Remove.
8150
4b128ece
JJ
81512013-08-28 Jakub Jelinek <jakub@redhat.com>
8152
8153 PR middle-end/58257
8154 * omp-low.c (copy_var_decl): Copy over TREE_NO_WARNING flag.
8155
12211b99 81562013-08-28 Jan Hubicka <jh@suse.cz>
9e57846a
JH
8157
8158 * builtins.def (free): Declare leaf.
8159
8dce4dbc
DM
81602013-08-27 David Malcolm <dmalcolm@redhat.com>
8161
8162 * gdbhooks.py: New.
8163 * configure.ac (gdbinit.in): Add import of gcc/gdbhooks.py.
8164 * configure: Regenerate.
8165
b8f6e610
MJ
81662013-08-27 Martin Jambor <mjambor@suse.cz>
8167
8168 * ipa-prop.h (ipa_pass_through_data): New field type_preserved.
8169 (ipa_ancestor_jf_data): Likewise.
8170 (ipa_get_jf_pass_through_agg_preserved): Fix comment typo.
8171 (ipa_get_jf_pass_through_type_preserved): New function.
8172 (ipa_get_jf_ancestor_agg_preserved): Fix comment typo.
8173 (ipa_get_jf_ancestor_type_preserved): New function.
12211b99 8174 * ipa-cp.c (ipa_get_jf_pass_through_result): Honor type_preserved flag.
b8f6e610
MJ
8175 (ipa_get_jf_ancestor_result): Likewise.
8176 (propagate_vals_accross_pass_through): Use
8177 ipa_get_jf_pass_through_result to do all the value mappings.
8178 * ipa-prop.c (ipa_print_node_jump_functions_for_edge): Dump the
8179 type_preserved flag.
8180 (ipa_set_jf_cst_copy): New function.
8181 (ipa_set_jf_simple_pass_through): Set the type_preserved flag.
8182 (ipa_set_jf_arith_pass_through): Likewise.
8183 (ipa_set_ancestor_jf): Likewise.
8184 (compute_complex_assign_jump_func): Set type_preserved instead of
8185 punting.
8186 (ipa_compute_jump_functions_for_edge): Likewise.
8187 (combine_known_type_and_ancestor_jfs): Honor type_preserved.
8188 (update_jump_functions_after_inlining): Update type_preserved.
12211b99 8189 Explicitely create jump functions when combining one with pass_through.
b8f6e610
MJ
8190 (ipa_write_jump_function): Stream the type_preserved flags.
8191 (ipa_read_jump_function): Likewise.
8192
74bf76ed
JJ
81932013-08-27 Jakub Jelinek <jakub@redhat.com>
8194 Aldy Hernandez <aldyh@redhat.com>
8195
8196 * Makefile.in (omp-low.o): Depend on $(TARGET_H).
8197 * cfgloop.h (struct loop): Add safelen, force_vect, simduid.
8198 * function.h (struct function): Add has_force_vect_loops and
8199 has_simduid_loops.
12211b99 8200 * gimple-pretty-print.c (dump_gimple_omp_for): Handle GF_OMP_FOR_KIND*.
74bf76ed
JJ
8201 * gimple.c (gimple_build_omp_critical): Add KIND argument and
8202 handle it.
8203 * gimple.def: Update CLAUSES comments.
8204 * gimple.h (enum gf_mask): Add GF_OMP_FOR_KIND_{FOR,SIMD}.
8205 (gimple_build_omp_for): Add argument to prototype.
8206 (gimple_omp_for_kind): New.
8207 (gimple_omp_for_set_kind): New.
8208 * gimplify.c (enum gimplify_omp_var_data): Add GOVD_LINEAR to
8209 GOVD_DATA_SHARE_CLASS.
8210 (enum omp_region_type): Add ORT_SIMD.
8211 (gimple_add_tmp_var): Handle ORT_SIMD.
8212 (gimplify_var_or_parm_decl): Same.
8213 (is_gimple_stmt): Same.
8214 (omp_firstprivatize_variable): Same.
8215 (omp_add_variable): Only use splay_tree_insert if lookup failed.
8216 (omp_notice_variable): Handle ORT_SIMD.
12211b99 8217 (omp_is_private): Add SIMD argument and handle it as well as ORT_SIMD.
74bf76ed
JJ
8218 (omp_check_private): Handle ORT_SIMD.
8219 (gimplify_scan_omp_clauses): Handle OMP_CLAUSE_LINEAR and
8220 OMP_CLAUSE_SAFELEN.
8221 (gimplify_adjust_omp_clauses_1): Handle GOVD_LINEAR.
8222 Handle OMP_CLAUSE_LASTPRIVATE.
8223 (gimplify_adjust_omp_clauses): Handle OMP_CLAUSE_LINEAR and
8224 OMP_CLAUSE_SAFELEN.
8225 (gimplify_omp_for): Handle OMP_SIMD and OMP_CLAUSE_LINEAR.
8226 (gimplify_expr): Handle OMP_SIMD.
8227 * internal-fn.c (expand_GOMP_SIMD_LANE): New.
8228 (expand_GOMP_SIMD_VF): New.
8229 (expand_GOMP_SIMD_LAST_LANE): New.
8230 * internal-fn.def (GOMP_SIMD_LANE): New.
8231 (GOMP_SIMD_VF): New.
8232 (GOMP_SIMD_LAST_LANE): New.
8233 * omp-low.c: Include target.h.
8234 (extract_omp_for_data): Handle OMP_SIMD, OMP_CLAUSE_LINEAR,
8235 OMP_CLAUSE_SAFELEN.
8236 (check_omp_nesting_restrictions): Same.
8237 (omp_max_vf): New.
8238 (lower_rec_simd_input_clauses): New.
8239 (lower_rec_input_clauses): Handle OMP_SIMD, GF_OMP_FOR_KIND_SIMD,
8240 OMP_CLAUSE_LINEAR.
8241 (lower_lastprivate_clauses): Handle OMP_CLAUSE_LINEAR,
8242 GF_OMP_FOR_KIND_SIMD, OMP_SIMD.
8243 (expand_omp_build_assign): New.
8244 (expand_omp_for_init_counts): New.
8245 (expand_omp_for_init_vars): New.
8246 (extract_omp_for_update_vars): New.
8247 (expand_omp_for_generic): Use expand_omp_for_{init,update}_vars
8248 and rewrite accordingly.
8249 (expand_omp_simd): New.
8250 (expand_omp_for): Use expand_omp_simd.
8251 (lower_omp_for_lastprivate): Unshare vinit when appropriate.
8252 (lower_omp_for): Do not lower the body.
8253 * tree-data-ref (get_references_in_stmt): Allow IFN_GOMP_SIMD_LANE
8254 in their own loops.
8255 * tree-flow.h (find_omp_clause): Remove prototype.
12211b99
UB
8256 * tree-if-conv.c (main_tree_if_conversion): Run if doing if conversion,
8257 forcing vectorization of the loop, or if flag_tree_vectorize.
74bf76ed
JJ
8258 (gate_tree_if_conversion): Similarly.
8259 * tree-inline.c (remap_gimple_stmt): Pass for kind argument to
8260 gimple_build_omp_for.
8261 (copy_cfg_body): set has_force_vect_loops and has_simduid_loops.
8262 * tree-parloops (create_parallel_loop): Pass kind argument to
8263 gimple_build_omp_for.
8264 * tree-pretty-print.c (dump_omp_clause): Add cases for
8265 OMP_CLAUSE_UNIFORM, OMP_CLAUSE_LINEAR, OMP_CLAUSE_SAFELEN,
8266 OMP_CLAUSE__SIMDUID_.
8267 (dump_generic_node): Handle OMP_SIMD.
8268 * tree-ssa-ccp.c (likely_value): Handle IFN_GOMP_SIMD*.
8269 * tree-ssa-loop-ivcanon.c (tree_unroll_loops_completely_1): Do not
8270 unroll OMP_SIMD loops here.
12211b99 8271 * tree-ssa-loop.c (gate_tree_vectorize): Run if has_force_vect_loops.
74bf76ed 8272 * tree-vect-data-refs.c (vect_analyze_data_ref_dependence): Handle
12211b99 8273 loop->safelen.
74bf76ed
JJ
8274 (vect_analyze_data_refs): Handle simd loops.
8275 * tree-vect-loop.c (vectorizable_live_operation): Handle
8276 IFN_GOMP_SIMD*.
12211b99 8277 * tree-vect-stmts.c (vectorizable_call): Handle IFN_GOMP_SIMD_LANE.
74bf76ed
JJ
8278 (vectorizable_store): Handle STMT_VINFO_SIMD_LANE_ACCESS_P.
8279 (vectorizable_load): Same.
12211b99 8280 * tree-vectorizer.c: Include hash-table.h and tree-ssa-propagate.h.
74bf76ed
JJ
8281 (struct simduid_to_vf): New.
8282 (simduid_to_vf::hash): New.
8283 (simduid_to-vf::equal): New.
8284 (struct simd_array_to_simduid): New.
8285 (simd_array_to_simduid::hash): New.
8286 (simd_array_to_simduid::equal): New.
8287 (adjust_simduid_builtins): New.
8288 (struct note_simd_array_uses_struct): New.
8289 (note_simd_array_uses_cb): New.
8290 (note_simd_array_uses): New.
8291 (vectorize_loops): Handle simd hints and adjust simd builtins
8292 accordingly.
8293 * tree-vectorizer.h (struct _stmt_vec_info): Add
8294 simd_lane_access_p field.
8295 (STMT_VINFO_SIMD_LANE_ACCESS_P): New macro.
8296 * tree.c (omp_clause_num_ops): Add entries for OMP_CLAUSE_LINEAR,
8297 OMP_CLAUSE_SAFELEN, OMP_CLAUSE__SIMDUID_, OMP_CLAUSE_UNIFORM.
8298 (omp_clause_code_name): Same.
8299 (walk_tree_1): Handle OMP_CLAUSE_UNIFORM, OMP_CLAUSE_SAFELEN,
8300 OMP_CLAUSE__SIMDUID_, OMP_CLAUSE_LINEAR.
8301 * tree.def (OMP_SIMD): New entry.
12211b99
UB
8302 * tree.h (enum omp_clause_code): Add entries for OMP_CLAUSE_LINEAR,
8303 OMP_CLAUSE_UNIFORM, OMP_CLAUSE_SAFELEN, OMP_CLAUSE__SIMDUID_.
74bf76ed
JJ
8304 (OMP_CLAUSE_DECL): Adjust range for new clauses.
8305 (OMP_CLAUSE_LINEAR_NO_COPYIN): New.
8306 (OMP_CLAUSE_LINEAR_NO_COPYOUT): New.
8307 (OMP_CLAUSE_LINEAR_STEP): New.
8308 (OMP_CLAUSE_SAFELEN_EXPR): New.
8309 (OMP_CLAUSE__SIMDUID__DECL): New.
8310 (find_omp_clause): New prototype.
8311
d0cf4e84
L
83122013-08-27 H.J. Lu <hongjiu.lu@intel.com>
8313
8314 * config/i386/driver-i386.c (host_detect_local_cpu): Update
8315 Haswell processor detection.
8316
992592ec
CW
83172013-08-27 Christian Widmer <shadow@umbrox.de>
8318
8319 PR target/57927
8320 * config/i386/driver-i386.c (host_detect_local_cpu): Add detection
8321 of Ivy Bridge and Haswell processors. Assume core-avx2 for unknown
8322 AVX2 capable processors.
8323
9116eb22
TB
83242013-08-27 Tejas Belagod <tejas.belagod@arm.com>
8325
8326 * config/aarch64/arm_neon.h: Replace all inline asm implementations
8327 of vget_low_* with implementations in terms of other intrinsics.
8328
1c5abb53
MG
83292013-08-27 Marc Glisse <marc.glisse@inria.fr>
8330
8331 PR middle-end/57219
8332 * doc/extend.texi (__builtin_isinf_sign): Restrict the return
8333 values to -1, 0 and 1.
8334
2e100703
VP
83352013-08-27 Vidya Praveen <vidyapraveen@arm.com>
8336
8337 * config/aarch64/aarch64.md (unspec): Add UNSPEC_SISD_SSHL,
8338 UNSPEC_SISD_USHL, UNSPEC_USHL_2S, UNSPEC_SSHL_2S, UNSPEC_SISD_NEG.
8339 (<optab><mode>3_insn): Remove.
8340 (aarch64_ashl_sisd_or_int_<mode>3): New Pattern.
8341 (aarch64_lshr_sisd_or_int_<mode>3): Likewise.
8342 (aarch64_ashr_sisd_or_int_<mode>3): Likewise.
8343 (define_split for aarch64_lshr_sisd_or_int_di3): Likewise.
8344 (define_split for aarch64_lshr_sisd_or_int_si3): Likewise.
8345 (define_split for aarch64_ashr_sisd_or_int_di3): Likewise.
8346 (define_split for aarch64_ashr_sisd_or_int_si3): Likewise.
8347 (aarch64_sisd_ushl, aarch64_sisd_sshl): Likewise.
8348 (aarch64_ushl_2s, aarch64_sshl_2s, aarch64_sisd_neg_qi): Likewise.
8349 (ror<mode>3_insn): Likewise.
8350 * config/aarch64/predicates.md (aarch64_simd_register): New.
8351
4ded8276
RB
83522013-08-27 Richard Biener <rguenther@suse.de>
8353
8354 PR tree-optimization/57521
8355 * tree-if-conv.c (if_convertible_bb_p): Verify that at least
8356 one edge is non-critical.
8357 (find_phi_replacement_condition): Make sure to use a non-critical
8358 edge. Cleanup and remove old bug workarounds.
8359 (bb_postdominates_preds): Remove.
8360 (if_convertible_loop_p_1): Do not compute post-dominators.
8361 (combine_blocks): Do not free post-dominators.
8362 (main_tree_if_conversion): Likewise.
8363 (pass_data_if_conversion): Add TODO_verify_ssa.
8364
5aa11061
DD
83652013-08-27 DJ Delorie <dj@redhat.com>
8366
8367 * config/i386/djgpp.h (ASM_DECLARE_FUNCTION_NAME): New.
8368
fd91cfe3
YZ
83692013-08-27 Yufeng Zhang <yufeng.zhang@arm.com>
8370
8371 * function.c (assign_parm_find_data_types): Set passed_mode and
8372 nominal_mode to the TYPE_MODE of nominal_type for the built
8373 pointer type in case of the struct-pass-by-reference.
8374
907555ce
JR
83752013-08-26 Joern Rennecke <joern.rennecke@embecosm.com>
8376
8377 * config/avr/avr-stdint.h (INT16_TYPE): Change default to "int".
8378 (UINT16_TYPE): Change default to "unsigned int".
8379
67518c93
JR
8380 * config/avr/avr.opt (mfract-convert-truncate): New option.
8381 * config/avr/avr.c (avr_out_fract): Unless TARGET_FRACT_CONV_TRUNC
8382 is set, round negative fractional integers according to n1169
8383 when converting to integer types.
8384
d6d989fc
JH
83852013-08-26 Jan Hubicka <jh@suse.cz>
8386
8387 * cgraph.c (cgraph_propagate_frequency): Do not assume that virtual
8388 methods can not be called indirectly when their address is not taken.
8389
2aa3da06
JH
83902013-08-26 Jan Hubicka <jh@suse.cz>
8391
12211b99
UB
8392 * gimple-fold.c (gimple_get_virt_method_for_binfo): Use
8393 ctor_for_folding.
2aa3da06 8394
0987ffe7
JH
83952013-08-26 Jan Hubicka <jh@suse.cz>
8396
8397 * ipa.c (comdat_can_be_unshared_p_1): C++ constructors and destructors
8398 can be unshared.
8399
befe8647
JR
84002013-08-26 Joern Rennecke <joern.rennecke@embecosm.com>
8401
8402 * reload.c (find_valid_class): Allow classes that do not include
8403 FIRST_PSEUDO_REGISTER - 1.
8404
a21e735e
JH
84052013-08-26 Jan Hubicka <jh@suse.cz>
8406
8407 * cgraph.c (cgraph_redirect_edge_call_stmt_to_callee): Fix formatting;
8408 fix edge count/frequency when speculation failed; fix type check
8409 for the direct call.
8410
e067bd43
JH
84112013-08-26 Jan Hubicka <jh@suse.cz>
8412
8413 * ipa-prop.c (ipa_print_node_params): Do not ICE during WPA.
8414
e8aec975
JH
84152013-08-26 Jan Hubicka <jh@suse.cz>
8416
8417 * ipa-inline-transform.c (inline_transform): Be ready for basic block
8418 to be changed by edge redirection.
8419
d0b66480
JH
84202013-08-26 Jan Hubicka <jh@suse.cz>
8421
12211b99
UB
8422 * cgraph.c (cgraph_speculative_call_info): Fix parameter order and
8423 formating; add sanity check.
d0b66480
JH
8424 (cgraph_resolve_speculation): Add FIXME about scaling profiles.
8425 (cgraph_redirect_edge_call_stmt_to_callee): Fix ICE in debug dump.
8426 * ipa-inline.c (heap_edge_removal_hook): Reset node growth cache.
8427 (resolve_noninline_speculation): Update callee keys, too.
8428
0f9aaac7
JH
84292013-08-26 Jan Hubicka <jh@suse.cz>
8430
8431 * tree.h (tree_decl_with_vis): Add cxx_constructor, cxx_destructor.
8432 (DECL_CXX_CONSTRUCTOR_P, DECL_CXX_DESTRUCTOR_P): New macros.
8433
4b37444e
JR
84342013-08-26 Joern Rennecke <joern.rennecke@embecosm.com>
8435
8436 * config/i386/i386.c (x86_64_elf_select_section): Put ATTRIBUTE_UNUSED
8437 into proper place.
8438
5339fc0c
UB
84392013-08-26 Uros Bizjak <ubizjak@gmail.com>
8440
8441 * config/i386/i386.c (ix86_debug_options): Remove prototype.
8442 (x86_64_elf_select_section): Ditto.
8443 (ix86_handle_tm_regparm_attribute): Remove ATTRIBUTE_UNUSED on used
8444 arguments.
8445 (ix86_pass_by_reference): Ditto.
5339fc0c
UB
8446 (output_set_got): Ditto.
8447 (ix86_unary_operator_ok): Ditto.
8448 (ix86_expand_builtin): Ditto.
8449
d5c3d3ef
JH
84502013-08-23 Jan Hubicka <jh@suse.cz>
8451
5339fc0c 8452 * cgraph.c (cgraph_turn_edge_to_speculative): Fix debug output.
d5c3d3ef 8453
5b1e7435
JH
84542013-08-23 Jan Hubicka <jh@suse.cz>
8455
8456 * tree.h (TYPE_FINAL_P, DECL_FINAL_P): New macros.
8457 (tree_decl_with_vis): Add FINAL field.
8458
ead69dac
JL
84592013-08-23 Jeff Law <law@redhat.com>
8460
8461 * tree-ssa-pre.c (do_regular_insertion): Include the expression in
8462 the debugging dump when the expression is fully redundant.
8463
025311c4
GDR
84642013-08-23 Gabriel Dos Reis <gdr@integrable-solutions.net>
8465
8466 * diagnostic.c (diagnostic_set_caret_max_width): Use pp_buffer.
8467 * gimple-pretty-print.c (gimple_dump_bb_buff): Likewise.
8468 * pretty-print.c (pp_formatted_text_data): Likewise.
8469 (pp_write_text_to_stream): Likewise.
8470 (pp_write_text_as_dot_label_to_stream): Likewise.
8471 (pp_append_r): Likewise.
8472 (pp_format): Likewise.
8473 (pp_flush): Likewise.
8474 (pp_clear_output_area): Likewise.
8475 (pp_append_text): Likewise.
8476 (pp_formatted_text): Likewise.
8477 (pp_remaining_character_count_for_line): Likewise.
8478 (pp_newline): Likewise.
8479 (pp_character): Likewise.
8480 (output_buffer::~output_buffer): Define.
8481 (pretty_printer::~pretty_printer): Destruct output buffer.
8482 * pretty-print.h (output_buffer::~output_buffer): Declare.
8483 (pretty_printer::~pretty_printer): Declare virtual.
8484
0cadbfaa
MG
84852013-08-24 Marc Glisse <marc.glisse@inria.fr>
8486
8487 PR other/57324
8488 * hwint.h (HOST_WIDE_INT_UC, HOST_WIDE_INT_1U, HOST_WIDE_INT_M1,
8489 HOST_WIDE_INT_M1U): New macros.
8490 * fold-const.c (sign_bit_p, build_range_check, fold_unary_loc,
8491 fold_binary_loc, fold_ternary_loc): Use the new macros. Use an
8492 unsigned -1 for lshift.
8493 * cse.c (cse_insn): Likewise.
8494 * double-int.c (rshift_double, lshift_double): Likewise.
8495 * builtins.c (fold_builtin_bitop): Likewise.
8496 * combine.c (force_to_mode): Likewise.
8497 * tree.c (integer_pow2p, tree_log2, tree_floor_log2): Likewise.
8498 * simplify-rtx.c (simplify_const_unary_operation,
8499 simplify_const_binary_operation): Likewise.
8500 * tree-stdarg.c (va_list_counter_bump, va_list_ptr_read,
8501 check_va_list_escapes): Likewise.
8502 * rtlanal.c (nonzero_bits1): Likewise.
8503 * expmed.c (expand_smod_pow2): Likewise.
8504 * tree-ssa-structalias.c (UNKNOWN_OFFSET): Use HOST_WIDE_INT_MIN.
8505
009e5353
JH
85062013-08-23 Jan Hubicka <jh@suse.cz>
8507
8508 * cgraph.c (cgraph_turn_edge_to_speculative): Mark target node
8509 as having address taken.
8510
c4be6568
JH
85112013-08-23 Jan Hubicka <jh@suse.cz>
8512
64cbf23d
JH
8513 * ipa-utils.h (method_class_type): Declare.
8514 * ipa-devirt.c (method_class_type): Export.
8515
c4be6568
JH
8516 * cgraphunit.c (analyze_functions): Do basic devirtualization;
8517 do not walk base classes of anonymous types.
8518
97aba8e9
KK
85192013-08-23 Kaz Kojima <kkojima@gcc.gnu.org>
8520
8521 PR rtl-optimization/58220
8522 PR regression/58221
8523 * final.c (reemit_insn_block_notes): Use NEXT_INSN to
8524 handle SEQUENCE insns properly.
8525
0fc80001
GDR
85262013-08-23 Gabriel Dos Reis <gdr@integrable-solutions.net>
8527
8528 * pretty-print.h (pp_newline_and_flush): Declare. Remove macro
8529 definition.
8530 (pp_newline_and_indent): Likewise.
8531 (pp_separate_with): Likewise.
8532 * pretty-print.c (pp_newline_and_flush): Define.
8533 (pp_newline_and_indent): Likewise.
8534 (pp_separate_with): Likewise.
8535
520a5868
JJ
85362013-08-23 Jakub Jelinek <jakub@redhat.com>
8537
8538 PR target/58218
8539 * config/i386/x86-64.h (TARGET_SECTION_TYPE_FLAGS): Define.
8540 * config/i386/i386.c (x86_64_elf_section_type_flags): New function.
8541
e500c62a
KY
85422013-08-23 Kirill Yukhin <kirill.yukhin@intel.com>
8543
f43245d1
UB
8544 * config/i386/predicates.md (ext_sse_reg_operand): New.
8545 * config/i386/i386.md (*movti_internal): Use
e500c62a
KY
8546 predicate to determine if EVEX is needed.
8547 (*movsi_internal): Ditto.
8548 (*movdf_internal): Ditto.
8549 (*movsf_internal): Ditto.
f43245d1 8550 * config/i386/mmx.md (*mov<mode>_internal): Ditto.
e500c62a 8551
6b00d7dd
JJ
85522013-08-23 Jakub Jelinek <jakub@redhat.com>
8553
8554 PR tree-optimization/58209
8555 * tree-tailcall.c (process_assignment): Handle POINTER_PLUS_EXPR.
8556 (find_tail_calls): Give up for pointer result types if m is non-NULL.
8557 (adjust_return_value_with_ops): For PLUS_EXPR and pointer result type
8558 emit POINTER_PLUS_EXPR.
8559 (create_tailcall_accumulator): For pointer result type accumulate in
8560 sizetype type.
8561
4f2a9f90
PC
85622013-08-22 Paolo Carlini <paolo.carlini@oracle.com>
8563
8564 * configure.ac: Add backslashes missing from the last change.
8565 * configure: Regenerate.
8566
7bf4274e
JH
85672013-08-22 Jan Hubicka <jh@susue.cz>
8568
4f2a9f90
PC
8569 * ipa.c (function_and_variable_visibility): First remember function
8570 was global and then make it local.
7bf4274e 8571
1bea243d
JB
85722013-08-22 Julian Brown <julian@codesourcery.com>
8573
8574 * configure.ac: Add aarch64 to list of arches which use "nop" in
8575 debug_line test.
8576 * configure: Regenerate.
8577
bcba7ecc
AK
85782013-08-22 Andreas Krebbel <Andreas.Krebbel@de.ibm.com>
8579
8580 * config/s390/linux.h (TARGET_LIBC_HAS_FUNCTION): Define as
8581 gnu_libc_has_function.
8582 * config/s390/tpf.h: Likewise.
8583
4c495b0d
JH
85842013-08-22 Jan Hubicka <jh@susue.cz>
8585
8586 * timevar.c (validate_phases): Add cast.
8587
5e302bca
JH
85882013-08-22 Jan Hubicka <jh@susue.cz>
8589
8590 * timevar.c (validate_phases): Use size_t for memory.
8591 * timevar.h (struct timevar_time_def): Use size_t for ggc_mem.
8592
da6ca2b5
GDR
85932013-08-22 Gabriel Dos Reis <gdr@integrable-solutions.net>
8594
8595 * pretty-print.h (output_buffer::output_buffer): Declare.
8596 (pretty_printer::pretty_printer): Likewise.
8597 (pp_construct): Remove.
8598 * pretty-print.c (output_buffer::output_buffer): Define.
8599 (pretty_printer::pretty_printer): Rename from pp_construct. Simplify.
8600 * gimple-pretty-print.c (print_gimple_stmt): Do not call pp_construct.
8601 (print_gimple_expr): Likewise.
8602 (print_gimple_seq): Likewise.
8603 (gimple_dump_bb): Likewise.
8604 * sched-vis.c (dump_value_slim): Likewise.
8605 (dump_insn_slim): Likewise.
8606 (dump_rtl_slim): Likewise.
8607 (str_pattern_slim): Likewise.
8608 * tree-mudflap.c (mf_varname_tree): Likewise.
8609 * graph.c (print_graph_cfg): Likewise.
8610 (start_graph_dump): Likewise.
8611 * tree-pretty-print.c (maybe_init_pretty_print): Likewise. Use
8612 placement-new.
8613 * diagnostic.c (diagnostic_initialize): Simplify early diagnostic
8614 pretty printer initialization.
8615 * coretypes.h (diagnostic_context): Remove superflous type alias
8616 declaration.
8617 (pretty_printer): Likewise. Declare directly as a class.
8618 (pretty_print_info): Remove declaration as class.
8619 * asan.c (asan_emit_stack_protection): Remove call to pp_construct
8620 and pp_clear_output_area.
8621 (asan_add_global): Likewise.
8622
0e1474e5
JH
86232013-08-22 Jan Hubicka <jh@suse.cz>
8624
8625 * cgraphunit.c (analyze_functions) Use update_type_inheritance_graph.
8626 * ipa-utils.h (update_type_inheritance_graph): Declare.
8627 (possible_polymorphic_call_target_p): Declare.
8628 (possible_polymorphic_call_target_p): New.
8629 * ipa-devirt.c: Update toplevel comments.
8630 (cached_polymorphic_call_targets): Move up.
8631 (odr_type_d): Move ID down.
8632 (polymorphic_type_binfo_p): Update comment.
8633 (odr_hasher::remove): Likewise;
8634 (get_odr_type): Set anonymous_namespace.
8635 (dump_odr_type): Dump it.
8636 (dump_type_inheritance_graph): Do not ICE when there are no ODR types.
8637 (maybe_record_node): Record node in cached_polymorphic_call_targets.
8638 (record_binfo): Add comment.
5339fc0c
UB
8639 (free_polymorphic_call_targets_hash): Do not ICE when cache is not
8640 built.
0e1474e5
JH
8641 (devirt_node_removal_hook): Do not iCE when cache is freed.
8642 (possible_polymorphic_call_target_p): New predicate.
8643 (update_type_inheritance_graph): New function.
8644
3f97cb0b
AI
86452013-08-22 Alexander Ivchenko <alexander.ivchenko@intel.com>
8646 Maxim Kuznetsov <maxim.kuznetsov@intel.com>
8647 Sergey Lega <sergey.s.lega@intel.com>
8648 Anna Tikhonova <anna.tikhonova@intel.com>
8649 Ilya Tocar <ilya.tocar@intel.com>
8650 Andrey Turetskiy <andrey.turetskiy@intel.com>
8651 Ilya Verbin <ilya.verbin@intel.com>
8652 Kirill Yukhin <kirill.yukhin@intel.com>
8653 Michael Zolotukhin <michael.v.zolotukhin@intel.com>
8654
8655 * common/config/i386/i386-common.c (OPTION_MASK_ISA_AVX512F_SET): New.
8656 (OPTION_MASK_ISA_AVX512CD_SET): Ditto.
8657 (OPTION_MASK_ISA_AVX512PF_SET): Ditto.
8658 (OPTION_MASK_ISA_AVX512ER_SET): Ditto.
8659 (OPTION_MASK_ISA_AVX2_UNSET): Update.
8660 (OPTION_MASK_ISA_AVX512F_UNSET): New.
8661 (OPTION_MASK_ISA_AVX512CD_UNSET): Ditto.
8662 (OPTION_MASK_ISA_AVX512PF_UNSET): Ditto.
8663 (OPTION_MASK_ISA_AVX512ER_UNSET): Ditto.
8664 (ix86_handle_option): Handle OPT_mavx512f, OPT_mavx512cd,
8665 OPT_mavx512pf, OPT_mavx512er cases.
8666 * config/i386/constraints.md (v): New constraint.
8667 (Yi, Yj): Replace SSE_REGS with ALL_SSE_REGS.
8668 * config/i386/cpuid.h (bit_AVX512F, bit_AVX512PF, bit_AVX512ER)
8669 (bit_AVX512CD): New.
8670 * config/i386/driver-i386.c (host_detect_local_cpu): Detect
8671 AVX512F, AVX512ER, AVX512PF, AVX512CD features.
8672 * config/i386/i386-c.c (ix86_target_macros_internal):
8673 Conditionally define __AVX512F__, __AVX512ER__, __AVX512CD__,
8674 __AVX512PF__.
8675 * config/i386/i386-modes.def (VECTOR_MODES (INT, 128))
8676 (VECTOR_MODES (FLOAT, 128), INT_MODE (XI, 64)): New modes.
8677 * config/i386/i386.c (regclass_map, dbx_register_map)
8678 (dbx64_register_map, svr4_dbx_register_map): Add new SSE registers.
8679 (gate_insert_vzeroupper): Disable vzeroupper for TARGET_AVX512F.
8680 (ix86_target_string): Define -mavx512f, -mavx512er, -mavx512cd,
8681 -mavx512pf options.
8682 (ix86_option_override_internal): Define PTA_AVX512F, PTA_AVX512ER,
8683 PTA_AVX512PF, PTA_AVX512CD. Handle -mavx512f, -mavx512er, -mavx512cd,
8684 -mavx512pf options. Fix formatting.
8685 (ix86_conditional_register_usage): Squash EXT_REX_SSE_REGs for 32-bit
8686 targets. Squash EVEX_SSE_REGS if AVX512F is disabled.
8687 (ix86_valid_target_attribute_inner_p): Handle -mavx512f, -mavx512er,
8688 -mavx512cd, -mavx512pf options.
8689 (standard_sse_constant_opcode): Add vpternlogd for 512-bit modes.
8690 (print_reg, ix86_print_operand): Handle 'g' to output 512-bit operands.
8691 (ix86_preferred_output_reload_class): Replace SSE_REGS with
8692 ALL_SSE_REGS.
8693 (ix86_hard_regno_mode_ok): Support 512-bit registers.
8694 (ix86_set_reg_reg_cost): Ditto.
8695 (x86_order_regs_for_local_alloc): Ditto.
8696 (MAX_VECT_LEN): Extend to 64-byte.
8697 (ix86_spill_class): Replace SSE_REGS with ALL_SSE_REGS.
8698 * config/i386/i386.h (TARGET_AVX512F, TARGET_AVX512PF)
8699 (TARGET_AVX512ER, TARGET_AVX512CD): New.
8700 (BIGGEST_ALIGNMENT): Extend to 512-bits.
8701 (FIRST_PSEUDO_REGISTER, FIXED_REGISTERS): Add new registers.
8702 (CALL_USED_REGISTERS, REG_ALLOC_ORDER): Likewise.
8703 (VALID_AVX512F_SCALAR_MODE, VALID_AVX512F_REG_MODE): New.
8704 (SSE_REG_MODE_P): Support new modes.
8705 (FIRST_MMX_REG, FIRST_REX_INT_REG, FIRST_REX_SSE_REG): Add comments.
8706 (FIRST_EXT_REX_SSE_REG, LAST_EXT_REX_SSE_REG): New.
8707 (reg_class, REG_CLASS_NAMES): Add EVEX_SSE_REGS, ALL_SSE_REGS.
8708 (SSE_CLASS_P, MAYBE_SSE_CLASS_P): Replace SSE_REGS with ALL_SSE_REGS.
8709 (REG_CLASS_CONTENTS): Add new registers.
8710 (SSE_REGNO_P, SSE_REGNO, HARD_REGNO_RENAME_OK): Support new registers.
8711 (EXT_REX_SSE_REGNO_P): New.
8712 (HI_REGISTER_NAMES): Add new registers.
8713 * config/i386/i386.md: Define constants for new registers.
8714 (mode): Add new 512-bit modes.
8715 (prefix): Support evex prefix.
8716 (isa): Support avx512f, noavx512f, fma_avx512f.
8717 (ssemodesuffix): Add new 512-bit modes.
8718 (movxi): New.
8719 (*movxi_internal_avx512f): Ditto.
8720 (*movdi_internal): Replace constraint "x" with the new constraint "v".
8721 Support MODE_XI.
8722 (*movsi_internal): Likewise.
8723 (*movdf_internal): Likewise.
8724 (*movsf_internal): Likewise.
8725 (*fop_<mode>_comm_sse): Replace constraint "x" with new constraint "v".
8726 (<code><mode>3): Likewise.
5339fc0c
UB
8727 * config/i386/i386.opt (mavx512f, mavx512pf, mavx512er, mavx512cd):
8728 New.
3f97cb0b
AI
8729 * config/i386/mmx.md (*mov<mode>_internal): Replace constraint "x"
8730 with the new constraint "v".
8731 * config/i386/sse.md (*mov<mode>_internal): Support new registers and
8732 modes.
8733 (<sse>_loadu<ssemodesuffix><avxsizesuffix>): Replace constraint "x"
8734 with the new constraint "v".
8735 (<sse2>_loaddqu<avxsizesuffix>): Likewise.
8736 (<sse2>_storedqu<avxsizesuffix>): Likewise.
8737 (*<plusminus_insn><mode>3): Likewise.
8738 (<sse>_vm<plusminus_insn><mode>3): Likewise.
8739 (*mul<mode>3): Likewise.
8740 (<sse>_vmmul<mode>3): Likewise.
8741 (<sse>_div<mode>3): Likewise.
8742 (<sse>_vmdiv<mode>3): Likewise.
8743 (<sse>_sqrt<mode>2): Likewise.
8744 (<sse>_vmsqrt<mode>2): Likewise.
8745 (*<code><mode>3_finite): Likewise.
8746 (*<code><mode>3) <smaxmin>: Likewise.
8747 (<sse>_vm<code><mode>3): Likewise.
8748 (*<code><mode>3) <any_logic>: Likewise.
8749 (*fma_fmadd_<mode>): Likewise.
8750 (*fma_fmsub_<mode>): Likewise.
8751 (*fma_fnmadd_<mode>): Likewise.
8752 (*fma_fnmsub_<mode>): Likewise.
8753 (*fma_fmaddsub_<mode>): Likewise.
8754 (*fma_fmsubadd_<mode>): Likewise.
8755 (*fmai_fmadd_<mode>): Likewise.
8756 (*fmai_fmsub_<mode>): Likewise.
8757 (*fmai_fnmadd_<mode>): Likewise.
8758 (*fmai_fnmsub_<mode>): Likewise.
8759 (sse_cvtsi2ss): Likewise.
8760 (sse_cvtsi2ssq): Likewise.
8761 (sse_cvtss2si): Likewise.
8762 (sse_cvtss2si_2): Likewise.
8763 (sse_cvtss2siq): Likewise.
8764 (sse_cvtss2siq_2): Likewise.
8765 (sse_cvttss2si): Likewise.
8766 (sse_cvtss2siq_2): Likewise.
8767 (float<sseintvecmodelower><mode>2): Likewise.
8768 (sse2_cvtsd2si_2): Likewise.
8769 (sse2_cvtsd2siq_2): Likewise.
8770 (*<plusminus_insn><mode>3): Likewise.
8771 (*<sse2_avx2>_<plusminus_insn><mode>3): Likewise.
8772 (*<sse4_1_avx2>_mul<mode>3): Likewise.
8773 (ashr<mode>3): Likewise.
8774 (<shift_insn><mode>3): Likewise.
8775 (avx2_<code><mode>3): Likewise.
8776 (*avx2_<code><mode>3): Likewise.
8777 (*andnot<mode>3): Likewise.
8778 (*<code><mode>3) <any_logic>: Likewise.
8779 (abs<mode>2): Likewise.
8780 (avx2_permvar<mode>): Likewise.
8781 (avx2_perm<mode>_1): Likewise.
8782 (*avx_vpermilp<mode>): Likewise.
8783 (avx_vpermilvar<mode>3): Likewise.
8784 (avx2_ashrv<mode>): Likewise.
8785 (avx2_<shift_insn>v<mode>): Likewise.
8786 * doc/invoke.texi: Document -mavx512f, -mavx512pf, -mavx512er,
8787 -mavx512cd.
8788 * doc/rtl.texi: Document XImode.
8789
dd5e8423
JL
87902013-08-21 Jeff Law <law@redhat.com>
8791
b9ebee5d
JL
8792 * tree-flow.h (register_jump_thread): Pass vector of edges
8793 instead of each important edge.
8794 * tree-ssa-threadedge.c (thread_across_edge): Build the jump
8795 thread path into a vector and pass that to register_jump_thread.
8796 * tree-ssa-threadupdate.c (register_jump_thread): Conver the
8797 passed in edge vector to the current 3-edge form.
8798
dd5e8423
JL
8799 Revert:
8800 2013-08-20 Alexey Makhalov <makhaloff@gmail.com>
8801
9bb6628e 8802 * dce.c (fini_dce): Call df_analyze again just in case
dd5e8423
JL
8803 delete_unmarked_insns removed anything.
8804
445dc8df
JR
88052013-08-21 Joern Rennecke <joern.rennecke@embecosm.com>
8806
8807 * reload.h (struct reg_equivs): Rename to ..
8808 (struct reg_equivs_s): .. this.
8809
e0df53dd
ML
88102013-08-20 Martin Liska <marxin.liska@gmail.com>
8811
8812 * ipa.c (ipa_profile_read_summary): Fix buffer overflow.
8813
e7d1d3eb
RO
88142013-08-21 Rainer Orth <ro@CeBiTec.Uni-Bielefeld.DE>
8815
5339fc0c 8816 * config/sol2-10.h (TARGET_LIBC_HAS_FUNCTION): Don't nest comment.
e7d1d3eb 8817
5562e26e
JL
88182013-08-21 Jeff Law <law@redhat.com>
8819
8820 * tree-vrp.c (simplify_stmt_for_jump_threading): Try to
8821 simplify assignments too. If the RHS collapses to a singleton
8822 range, then return the value for the range.
8823
c7ecdec6
KY
88242013-08-21 Kirill Yukhin <kirill.yukhin@intel.com>
8825
7a5911d3
UB
8826 * config/i386/sse.md (V16): Rename to...
8827 (VMOVE): this.
8828 (mov<mode>): Update iterator name.
8829 (*mov<mode>_internal): Ditto.
8830 (push<mode>1): Ditto.
8831 (movmisalign<mode>): Ditto.
c7ecdec6 8832
bfa3b50a
JH
88332013-08-20 Jan Hubicka <jh@suse.cz>
8834
8835 PR bootstrap/58186
8836 * cgraph.c (cgraph_add_edge_to_call_site_hash): Overwrite hash
8837 entry for direct edges.
8838 (cgraph_turn_edge_to_speculative): Fix setting of can_throw_external.
8839
6cd4d135
DM
88402013-08-20 David Malcolm <dmalcolm@redhat.com>
8841
8842 Revert my last two changes, r201865 and r201864:
8843
8844 Revert r201865:
8845 2013-08-20 David Malcolm <dmalcolm@redhat.com>
8846
8847 Make opt_pass and gcc::pass_manager be GC-managed, so that pass
8848 instances can own GC refs.
8849
8850 * Makefile.in (GTFILES): Add pass_manager.h and tree-pass.h.
8851 * context.c (gcc::context::gt_ggc_mx): Traverse passes_.
8852 (gcc::context::gt_pch_nx): Likewise.
8853 (gcc::context::gt_pch_nx): Likewise.
8854 * ggc.h (gt_ggc_mx <T>): New.
8855 (gt_pch_nx_with_op <T>): New.
8856 (gt_pch_nx <T>): New.
8857 * passes.c (opt_pass::gt_ggc_mx): New.
8858 (opt_pass::gt_pch_nx): New.
8859 (opt_pass::gt_pch_nx_with_op): New.
8860 (pass_manager::gt_ggc_mx): New.
8861 (pass_manager::gt_pch_nx): New.
8862 (pass_manager::gt_pch_nx_with_op): New.
8863 (pass_manager::operator new): Use
8864 ggc_internal_cleared_alloc_stat rather than xcalloc.
8865 * pass_manager.h (class pass_manager): Add GTY((user)) marking.
8866 (pass_manager::gt_ggc_mx): New.
8867 (pass_manager::gt_pch_nx): New.
8868 (pass_manager::gt_pch_nx_with_op): New.
8869 * tree-pass.h (class opt_pass): Add GTY((user)) marking.
8870 (opt_pass::operator new): New.
8871 (opt_pass::gt_ggc_mx): New.
8872 (opt_pass::gt_pch_nx): New.
8873 (opt_pass::gt_pch_nx_with_op): New.
8874
8875 Revert r201864:
8876 2013-08-20 David Malcolm <dmalcolm@redhat.com>
8877
8878 * Makefile.in (GTFILES): Add context.h.
8879 * context.c (gcc::context::operator new): New.
8880 (gcc::context::gt_ggc_mx): New.
8881 (gcc::context::gt_pch_nx): New.
8882 (gcc::context::gt_pch_nx): New.
8883 * context.h (gcc::context): Add GTY((user)) marking.
8884 (gcc::context::operator new): New.
8885 (gcc::context::gt_ggc_mx): New.
8886 (gcc::context::gt_pch_nx): New.
8887 (gcc::context::gt_pch_nx): New.
8888 (g): Add GTY marking.
8889 (gt_ggc_mx (gcc::context *)): New.
8890 (gt_pch_nx (gcc::context *)): New.
8891 (gt_pch_nx (gcc::context *ctxt, gt_pointer_operator op,
8892 void *cookie)): New.
8893 * gengtype.c (open_base_files) <ifiles>: Add context.h.
8894
a046136a
AM
88952013-08-20 Alexey Makhalov <makhaloff@gmail.com>
8896
8897 * dce.c (fini_dce): Call df_analyze again just in case
8898 delete_unmarked_insns removed anything.
8899
67598720
TJ
89002013-08-20 Teresa Johnson <tejohnson@google.com>
8901
8902 PR rtl-optimizations/57451
8903 * final.c (reemit_insn_block_notes): Prevent lexical blocks
8904 from crossing split section boundaries.
8905
f8693faf
MGD
89062013-08-20 Matthew Gretton-Dann <matthew.gretton-dann@linaro.org>
8907
8908 * config/arm/linux-elf.h (MULTILIB_DEFAULTS): Remove definition.
8909 * config/arm/t-linux-eabi (MULTILIB_OPTIONS): Document association
8910 with MULTLIB_DEFAULTS.
8911
7d0b9a9c
NC
89122013-08-20 Nick Clifton <nickc@redhat.com>
8913
8914 * target.def (narrow_volatile_bitfield): Note that the default
8915 value is false, not !TARGET_STRICT_ALIGN.
8916 * doc/tm.texi: Regenerate.
8917
8edb8dc8
PC
89182013-08-20 Pavel Chupin <pavel.v.chupin@intel.com>
8919
8920 Fix LIB_SPEC for systems without libpthread.
8921
8922 * config/gnu-user.h: Introduce GNU_USER_TARGET_NO_PTHREADS_LIB_SPEC.
8923 * config/arm/linux-eabi.h: Use GNU_USER_TARGET_NO_PTHREADS_LIB_SPEC
8924 for Android.
8925 * config/i386/linux-common.h: Likewise.
8926 * config/mips/linux-common.h: Likewise.
8927
12211b99 89282013-08-20 Zhouyi Zhou <yizhouzhou@ict.ac.cn>
d71ff3fa
ZZ
8929
8930 * tree-ssa-ccp.c (get_default_value): Remove redundant condition
8931 checks.
8932
e42c64cb
DM
89332013-08-20 David Malcolm <dmalcolm@redhat.com>
8934
8935 Make opt_pass and gcc::pass_manager be GC-managed, so that pass
8936 instances can own GC refs.
8937
8938 * Makefile.in (GTFILES): Add pass_manager.h and tree-pass.h.
8939 * context.c (gcc::context::gt_ggc_mx): Traverse passes_.
8940 (gcc::context::gt_pch_nx): Likewise.
8941 (gcc::context::gt_pch_nx): Likewise.
8942 * ggc.h (gt_ggc_mx <T>): New.
8943 (gt_pch_nx_with_op <T>): New.
8944 (gt_pch_nx <T>): New.
8945 * passes.c (opt_pass::gt_ggc_mx): New.
8946 (opt_pass::gt_pch_nx): New.
8947 (opt_pass::gt_pch_nx_with_op): New.
8948 (pass_manager::gt_ggc_mx): New.
8949 (pass_manager::gt_pch_nx): New.
8950 (pass_manager::gt_pch_nx_with_op): New.
8951 (pass_manager::operator new): Use
8952 ggc_internal_cleared_alloc_stat rather than xcalloc.
8953 * pass_manager.h (class pass_manager): Add GTY((user)) marking.
8954 (pass_manager::gt_ggc_mx): New.
8955 (pass_manager::gt_pch_nx): New.
8956 (pass_manager::gt_pch_nx_with_op): New.
8957 * tree-pass.h (class opt_pass): Add GTY((user)) marking.
8958 (opt_pass::operator new): New.
8959 (opt_pass::gt_ggc_mx): New.
8960 (opt_pass::gt_pch_nx): New.
8961 (opt_pass::gt_pch_nx_with_op): New.
8962
910c02a0
DM
89632013-08-20 David Malcolm <dmalcolm@redhat.com>
8964
8965 * Makefile.in (GTFILES): Add context.h.
8966 * context.c (gcc::context::operator new): New.
8967 (gcc::context::gt_ggc_mx): New.
8968 (gcc::context::gt_pch_nx): New.
8969 (gcc::context::gt_pch_nx): New.
8970 * context.h (gcc::context): Add GTY((user)) marking.
8971 (gcc::context::operator new): New.
8972 (gcc::context::gt_ggc_mx): New.
8973 (gcc::context::gt_pch_nx): New.
8974 (gcc::context::gt_pch_nx): New.
8975 (g): Add GTY marking.
8976 (gt_ggc_mx (gcc::context *)): New.
8977 (gt_pch_nx (gcc::context *)): New.
8978 (gt_pch_nx (gcc::context *ctxt, gt_pointer_operator op,
8979 void *cookie)): New.
8980 * gengtype.c (open_base_files) <ifiles>: Add context.h.
8981
af4b4236
AM
89822013-08-20 Alan Modra <amodra@gmail.com>
8983
8984 PR target/57865
8985 * config/rs6000/rs6000.c (rs6000_emit_prologue): Correct ool_adjust.
8986 (rs6000_emit_epilogue): Likewise.
8987
12211b99 89882013-08-19 Dehao Chen <dehao@google.com>
0d633222
DC
8989
8990 * value-prof.c (gimple_ic): Fix the bug of adding EH edge.
8991
e2323f5b
PB
89922013-08-19 Peter Bergner <bergner@vnet.ibm.com>
8993 Jakub Jelinek <jakub@redhat.com>
8994
8995 * builtins.def (BUILT_IN_FABSD32): New DFP ABS builtin.
8996 (BUILT_IN_FABSD64): Likewise.
8997 (BUILT_IN_FABSD128): Likewise.
8998 * builtins.c (expand_builtin): Add support for new DFP ABS builtins.
8999 (fold_builtin_1): Likewise.
5339fc0c
UB
9000 * config/rs6000/dfp.md (*negtd2_fpr): Handle non-overlapping
9001 destination and source operands.
e2323f5b
PB
9002 (*abstd2_fpr): Likewise.
9003 (*nabstd2_fpr): Likewise.
9004
cfbf3ee8
RS
90052013-08-19 Richard Sandiford <rdsandiford@googlemail.com>
9006
9007 * config/mips/mips.c (mips_adjust_insn_length): Add checks for
9008 JUMP_P and INSN_P.
9009
d3136aeb
AH
90102013-08-19 Aldy Hernandez <aldyh@redhat.com>
9011
9012 * doc/invoke.texi (-fcilkplus): Clarify that implementation is
9013 incomplete.
9014
d33d9e47
AI
90152013-08-19 Alexander Ivchenko <alexander.ivchenko@intel.com>
9016
9017 * target.def (TARGET_LIBC_HAS_FUNCTION): New target hook.
9018 * builtins.c (default_libc_has_function): New.
9019 (gnu_libc_has_function): Ditto.
9020 (no_c99_libc_has_function): Ditto.
9021 (expand_builtin_cexpi): Using new target hook TARGET_LIBC_HAS_FUNCTION
9022 instead of TARGET_HAS_SINCOS and TARGET_C99_FUNCTIONS.
9023 (fold_builtin_sincos): Likewise.
9024 (fold_builtin_cexp): Likewise.
9025 * builtins.def (DEF_C94_BUILTIN): Likewise.
9026 (DEF_C99_BUILTIN): Likewise.
9027 (DEF_C99_C90RES_BUILTIN): Likewise.
9028 (DEF_C99_COMPL_BUILTIN): New define. Change all complex c99 builtin
9029 definitions to using this define.
9030 * config/darwin-protos.h (darwin_libc_has_function): New.
5339fc0c 9031 * config/darwin.c (darwin_libc_has_function): Ditto.
d33d9e47
AI
9032 * config/alpha/linux.h: Remove TARGET_C99_FUNCTIONS and
9033 TARGET_HAS_SINCOS. Redefine TARGET_LIBC_HAS_FUNCTION.
9034 * config/darwin.h: Ditto.
9035 * config/elfos.h: Ditto.
9036 * config/freebsd.h: Ditto.
9037 * config/i386/cygming.h: Ditto.
9038 * config/i386/djgpp.h: Ditto.
9039 * config/i386/i386-interix.h: Ditto.
9040 * config/microblaze/microblaze.h: Ditto.
9041 * config/mmix/mmix.h: Ditto.
9042 * config/gnu-user.h: Ditto.
9043 * config/ia64/hpux.h: Ditto.
9044 * config/pa/pa-hpux.h: Ditto.
9045 * config/pdp11/pdp11.h: Ditto.
9046 * config/picochip/picochip.h: Ditto.
9047 * config/linux.h: Ditto.
9048 * config/netbsd.h: Ditto.
9049 * config/openbsd.h: Ditto.
9050 * config/rs6000/aix43.h: Ditto.
9051 * config/rs6000/aix51.h: Ditto.
9052 * config/rs6000/aix52.h: Ditto.
9053 * config/rs6000/aix53.h: Ditto.
9054 * config/rs6000/aix61.h: Ditto.
9055 * config/rs6000/darwin.h: Ditto.
9056 * config/rs6000/linux.h: Ditto.
9057 * config/rs6000/linux64.h: Ditto.
9058 * config/s390/tpf.h: Ditto.
9059 * config/sol2-10.h: Ditto.
9060 * config/sol2.h: Ditto.
9061 * config/vms/vms.h: Ditto.
9062 * config/vxworks.h: Ditto.
9063 * config/linux-android.c (linux_android_libc_has_function):
9064 New linux-specific implementation of TARGET_LIBC_HAS_FUNCTION.
9065 * config/linux-protos.h (linux_android_libc_has_function):
9066 New declaration.
9067 * config/i386/i386.c (ix86_libc_has_function): New.
9068 * config/i386/i386-protos.h
9069 (ix86_libc_has_function): New declaration.
9070 * config/i386/i386.md
9071 ("isinfxf2"): Change condition for TARGET_LIBC_HAS_FUNCTION.
9072 ("isinf<mode>2): Likewise.
9073 * convert.c (convert_to_integer): Using new target hook
9074 TARGET_LIBC_HAS_FUNCTION istead of TARGET_HAS_SINCOS and
9075 TARGET_C99_FUNCTIONS.
9076 * fortran/f95-lang.c (gfc_init_builtin_functions): Ditto.
9077 * tree-ssa-math-opts.c (execute_cse_sincos): Ditto.
9078 * coretypes.h (function_class): New enum for different
9079 classes of functions.
9080 * defaults.h: Remove TARGET_C99_FUNCTIONS and TARGET_HAS_SINCOS.
9081 * doc/tm.texi.in (TARGET_C99_FUNCTIONS): Remove documentation.
9082 (TARGET_HAS_SINCOS): Likewise.
9083 (TARGET_LIBC_HAS_FUNCTION): New.
9084 * doc/tm.texi: Regenerated.
9085 * targhooks.h (default_libc_has_function): New declaration.
9086 (no_c99_libc_has_function): Ditto.
9087 (gnu_libc_has_function): Ditto.
9088 * system.h: Add the poisoning of TARGET_C99_FUNCTIONS
9089 and TARGET_HAS_SINCOS.
9090
eefe9a99
JH
90912013-08-18 Jan Hubicka <jh@suse.cz>
9092
9093 * Makeifle-in (ipa-devirt.o): New.
9094 (GTFILES): Add ipa-utils.h and ipa-devirt.c
9095 * cgraphunit.c (decide_is_symbol_needed): Do not care about virtuals.
9096 (analyze_functions): Look into possible targets of polymorphic call.
9097 * dumpfile.c (dump_files): Add type-inheritance dump.
9098 * dumpfile.h (TDI_inheritance): New.
9099 * ipa-devirt.c: New file.
9100 * ipa-utils.h (odr_type_d): Forward declare.
9101 (odr_type): New type.
9102 (build_type_inheritance_graph): Declare.
9103 (possible_polymorphic_call_targets): Declare and introduce inline
9104 variant when only edge is pased.
9105 (dump_possible_polymorphic_call_targets): Likewise.
9106 * timevar.def (TV_IPA_INHERITANCE, TV_IPA_VIRTUAL_CALL): New.
9107 * tree.c (type_in_anonymous_namespace_p): Break out from ...
9108 (types_same_for_odr): ... here.
9109 * tree.h (type_in_anonymous_namespace_p): Declare.
9110
29b89442
JJ
91112013-08-18 Jakub Jelinek <jakub@redhat.com>
9112
9113 PR tree-optimization/58006
9114 * tree-parloops.c (take_address_of): Don't ICE if get_name
9115 returns NULL.
9116 (eliminate_local_variables_stmt): Remove clobber stmts.
9117
4f219961
EB
91182013-08-18 Eric Botcazou <ebotcazou@adacore.com>
9119
9120 * cgraphunit.c (handle_alias_pairs): Reset the alias flag after the
9121 error message is issued for an alias to undefined symbol.
9122
1d5755ef
JH
91232013-08-18 Jan Hubicka <jh@suse.cz>
9124
9125 * cgraph.c (cgraph_create_indirect_edge): Discover
9126 polymorphic calls and record basic info into indirect_info.
9127 * gimple-fold.c (gimple_fold_call): When doing BINFO based
9128 devirtualization, ignore objc function calls.
9129 * ipa-cp.c (initialize_node_lattices): Be ready for polymorphic
9130 call with no parm index info.
9131 * ipa-prop.c (ipa_analyze_call_uses): Likewise.
9132 * tree.c (virtual_method_call_p): New function.
9133 * tree.h (virtual_method_call_p): Declare.
9134
5945bebf
JH
91352013-08-16 Jan Hubicka <jh@suse.cz>
9136
9137 PR middle-end/58179
9138 * tree.c (obj_type_ref_class): Do not ICE on non-method calls.
9139
47fa5d23
DE
91402013-08-16 David Edelsohn <dje.gcc@gmail.com>
9141
9142 * config/rs6000/rs6000.md (rs6000_get_timebase_ppc32): Add length
9143 attribute.
9144
c74f54a0
DM
91452013-08-16 David Malcolm <dmalcolm@redhat.com>
9146
9147 * gengtype.c (type_for_name): Add special-case support for
9148 locating types within the "gcc::" namespace.
9149 (open_base_files): Emit a "using namespace gcc" directive.
9150
f758f299
MM
91512013-08-16 Michael Meissner <meissner@linux.vnet.ibm.com>
9152
9153 PR target/58160
9154 * config/rs6000/predicates.md (fusion_gpr_mem_load): Allow the
9155 memory rtx to contain ZERO_EXTEND and SIGN_EXTEND.
9156
9157 * config/rs6000/rs6000-protos.h (fusion_gpr_load_p): Pass operands
9158 array instead of each individual operand as a separate argument.
9159 (emit_fusion_gpr_load): Likewise.
9160 (expand_fusion_gpr_load): Add new function declaration.
9161
9162 * config/rs6000/rs6000.c (fusion_gpr_load_p): Change the calling
9163 signature to have the operands passed as an array, instead of as
9164 separate arguments. Allow ZERO_EXTEND to be in the memory
9165 address, and also SIGN_EXTEND if -mpower8-fusion-sign. Do not
9166 depend on the register live/dead flags when peepholes are run.
9167 (expand_fusion_gpr_load): New function to be called from the
9168 peephole2 pass, to change the register that addis sets to be the
9169 target register.
9170 (emit_fusion_gpr_load): Change the calling signature to have the
9171 operands passed as an array, instead of as separate arguments.
9172 Allow ZERO_EXTEND to be in the memory address, and also
9173 SIGN_EXTEND if -mpower8-fusion-sign.
9174
9175 * config/rs6000/rs6000.md (UNSPEC_FUSION_GPR): Delete unused
9176 unspec enumeration.
9177 (power8 fusion peephole/peephole2): Rework the fusion peepholes to
9178 adjust the register addis loads up in the peephole2 pass. Do not
5339fc0c 9179 depend on the register live/dead state when the peephole pass is done.
f758f299 9180
158f4e4f
DM
91812013-08-16 David Malcolm <dmalcolm@redhat.com>
9182
9183 * gengtype.c (create_user_defined_type): Ensure that the kind
9184 is set to TYPE_USER_STRUCT, fixing a bug seen when an incomplete
9185 declaration is seen before the GTY((user)) marking.
9186
22f8cea5
BE
91872013-08-16 Bernd Edlinger <bernd.edlinger@hotmail.de>
9188
9189 PR target/58105
9190 * config/i386/i386.c (make_resolver_func): Set DECL_UNINLINABLE.
9191
c49bdb2e
JH
91922013-08-16 Jan Hubicka <jh@suse.cz>
9193
9194 * gimple-fold.c (gimple_extract_devirt_binfo_from_cst): Add new
9195 arugment expected_type.
9196 (gimple_fold_call): Use it.
9197 * gimple.h (gimple_extract_devirt_binfo_from_cst): Update prototype.
9198 * ipa-cp.c (ipa_get_indirect_edge_target_1): Update.
22f8cea5 9199 * ipa-prop.c (ipa_analyze_virtual_call_uses): Use obj_type_ref_class.
c49bdb2e
JH
9200 (try_make_edge_direct_virtual_call): Likewise.
9201 * tree.c (obj_type_ref_class): New.
9202 * tree.h (obj_type_ref_class): Use it.
9203
4042dca9
GDR
92042013-08-16 Gabriel Dos Reis <gdr@integrable-solutions.net>
9205
9206 * sched-vis.c (rtl_slim_pp_initialized): Remove.
9207 (rtl_slim_pp): Likewise.
9208 (init_rtl_slim_pretty_print): Likewise.
9209 (dump_value_slim): Don't call it. Use local pretty printer.
9210 (dump_insn_slim): Likewise.
9211 (dump_rtl_slim): Likewise.
9212 (str_pattern_slim): Likewise.
9213 * tree-mudflap.c (mf_varname_tree): Use local pretty printer.
9214 Simplify.
9215
7a460594
JJ
92162013-08-16 Jakub Jelinek <jakub@redhat.com>
9217
639dc669
JJ
9218 PR tree-optimization/58164
9219 * gimple.c (walk_stmt_load_store_addr_ops): For visit_addr
9220 walk gimple_goto_dest of GIMPLE_GOTO.
9221
7a460594
JJ
9222 PR tree-optimization/58165
9223 * tree-call-cdce.c (shrink_wrap_one_built_in_call): If
9224 bi_call must be the last stmt in a bb, don't split_block, instead
9225 use fallthru edge from it and give up if there is none.
9226 Release conds vector when returning early.
9227
e3c02cc7
XDL
92282013-08-14 Xinliang David Li <davidxl@google.com>
9229
9230 * config/i386/i386.c (ix86_option_override_internal):
9231 Remove unused variable and field.
9232
6469da2c
BS
92332013-08-14 Bill Schmidt <wschmidt@linux.vnet.ibm.com>
9234
9235 PR target/57949
22f8cea5 9236 * doc/invoke.texi: Add documentation of mcompat-align-parm option.
6469da2c
BS
9237 * config/rs6000/rs6000.opt: Add mcompat-align-parm option.
9238 * config/rs6000/rs6000.c (rs6000_function_arg_boundary): For AIX
9239 and Linux, correct BLKmode alignment when 128-bit alignment is
9240 required and compatibility flag is not set.
22f8cea5
BE
9241 (rs6000_gimplify_va_arg): For AIX and Linux, honor specified alignment
9242 for zero-size arguments when compatibility flag is not set.
6469da2c 9243
4ca890e2
JJ
92442013-08-14 Jakub Jelinek <jakub@redhat.com>
9245
9246 PR tree-optimization/58145
9247 * tree-sra.c (build_ref_for_offset): If prev_base has
9248 TREE_THIS_VOLATILE or TREE_SIDE_EFFECTS, propagate it to MEM_REF.
9249
daec1759 92502013-08-14 Xinliang David Li <davidxl@google.com>
4ca890e2 9251
daec1759 9252 * config/i386/i386.c (ix86_option_override_internal):
e3c02cc7 9253 Fix uninitialized variable error.
daec1759 9254
3ad20bd4
XDL
92552013-08-14 Xinliang David Li <davidxl@google.com>
9256
9257 * config/i386/i386.opt: Define two new options.
9258 * config/i386/x86-tune.def: Add arch selector field in macros.
9259 * config/i386/i386.h: Adjust macro definition.
9260 * config/i386/i386.c (ix86_option_override_internal):
9261 Refactor the code.
9262 (parse_mtune_ctrl_str): New function.
9263 (set_ix86_tune_features): New function.
9264 (ix86_function_specific_restore): Call the new helper function.
9265
e839e2a9
AB
92662013-08-14 Andrey Belevantsev <abel@ispras.ru>
9267
9268 PR rtl-optimization/57662
9269 * sel-sched.c (code_motion_process_successors): When the current insn
9270 is removed after the recursive traversal, break from the loop.
9271 Add comments and debug printouts.
9272
27ed665f
JJ
92732013-08-14 Jakub Jelinek <jakub@redhat.com>
9274 Alexandre Oliva <aoliva@redhat.com>
9275
9276 PR target/58067
9277 * config/i386/i386.c (ix86_delegitimize_address): For CM_MEDIUM_PIC
9278 and CM_LARGE_PIC ix86_cmodel fall thru into the -m32 code, handle
9279 there also UNSPEC_PLTOFF.
9280
61c1a609
MP
92812013-08-14 Marek Polacek <polacek@redhat.com>
9282
9283 * ipa-inline-analysis.c (add_clause): Avoid shifting integer
9284 NUM_CONDITIONS bit positions.
9285
0e901c67
CC
92862013-08-13 Cary Coutant <ccoutant@google.com>
9287
9288 * dwarf2out.c (CHECKSUM_BLOCK): New macro.
9289 (attr_checksum): Hash vector contents instead of pointer.
9290 (attr_checksum_ordered): Likewise.
9291
061eff6d
UB
92922013-08-13 Uros Bizjak <ubizjak@gmail.com>
9293
9294 * config/i386/sse.md (*sse2_maskmovdqu): Emit addr32 prefix
9295 when Pmode != word_mode. Add length_address attribute.
9296 (sse3_monitor_<mode>): Merge from sse3_monitor and
9297 sse3_monitor64_<mode> insn patterns. Emit addr32 prefix when
9298 Pmode != word_mode. Update insn length attribute.
9299 * config/i386/i386.c (ix86_option_override_internal): Update
9300 ix86_gen_monitor selection for merged sse3_monitor insn.
9301
8ec9249b
JB
93022013-08-13 Julian Brown <julian@codesourcery.com>
9303
9304 * config/rs6000/rs6000.c (rs6000_legitimize_reload_address): Don't
9305 perform invalid legitimization on greater-than-word-size modes for
9306 TARGET_E500_DOUBLE.
9307
761a8eb7
VM
93082013-08-13 Vladimir Makarov <vmakarov@redhat.com>
9309
9310 * ira.c (setup_class_translate_array): Use aclass instead of cl
9311 for classes not fully covered by allocno classes.
9312
99c2bd54
JJ
93132013-08-13 Jakub Jelinek <jakub@redhat.com>
9314
4029a5e0
JJ
9315 PR tree-optimization/57661
9316 * tree-inline.h (struct copy_body_data): Add blocks_to_copy field.
9317 * tree-inline.c (tree_function_versioning): Initialize it.
9318 (remap_gimple_stmt): Return GIMPLE_NOP for MEM_REF lhs clobber stmts
9319 if id->blocks_to_copy and MEM_REF's SSA_NAME is defined in a block
9320 that is not being copied.
9321
99c2bd54
JJ
9322 PR sanitizer/56417
9323 * asan.c (instrument_strlen_call): Fix typo in comment.
9324 Use char * type even for the lhs of POINTER_PLUS_EXPR.
9325
b5f8f063
SE
93262013-08-13 Steve Ellcey <sellcey@mips.com>
9327
9328 * config/mips/mips.md (prefetch): Use lw instead of ld on
9329 loongson in 32bit mode.
9330
4de80584
NC
93312013-08-13 Nick Clifton <nickc@redhat.com>
9332
9333 * config.gcc: (avr-linux): Allow for tmake_file not being empty.
9334
09ce3660
JH
93352013-08-13 Jan Hubicka <jh@suse.cz>
9336
9337 * cgraph.c (cgraph_turn_edge_to_speculative): Return newly
9338 introduced edge; fix typo in sanity check.
9339 (cgraph_resolve_speculation): Export; improve diagnostic.
9340 (cgraph_redirect_edge_call_stmt_to_callee): Better diagnostic; cancel
9341 speculation at type mismatch.
9342 * cgraph.h (cgraph_turn_edge_to_speculative): Update.
9343 (cgraph_resolve_speculation): Declare.
9344 (symtab_can_be_discarded): New function.
9345 * value-prof.c (gimple_ic_transform): Remove actual transform code.
9346 * ipa-inline-transform.c (speculation_removed): New global var.
9347 (clone_inlined_nodes): See if speculation can be removed.
9348 (inline_call): If speculations was removed, we growths may not match.
9349 * ipa-inline.c (can_inline_edge_p): Add DISREGARD_LIMITS parameter.
9350 (speculation_useful_p): New function.
9351 (resolve_noninline_speculation): New function.
9352 (inline_small_functions): Resolve useless speculations.
9353 * ipa-inline.h (speculation_useful_p): Declare
9354 * ipa.c (can_replace_by_local_alias): Simplify.
9355 (ipa_profile): Produce speculative calls in non-lto, too;
9356 add simple cost model; produce local aliases.
9357
f971dc24
DM
93582013-08-13 David Malcolm <dmalcolm@redhat.com>
9359
9360 * config/i386/t-i386 (i386.o): Rename stray PIPELINE_H to
061eff6d 9361 PASS_MANAGER_H.
f971dc24 9362
2e507b9a
PC
93632013-08-12 Paolo Carlini <paolo.carlini@oracle.com>
9364
9365 * config/i386/i386.c (ix86_function_versions): Use error + inform.
9366
c97fd4b8
UB
93672013-08-12 Uros Bizjak <ubizjak@gmail.com>
9368
9369 * config/i386/i386.md (floatunssi<mode>2 expand): Use MODEF mode
9370 iterator instead of X87MODEF.
9371
93722013-08-12 Perez Read <netfirewall@gmail.com>
9373
9374 PR target/58132
9375 * config/i386/i386.md (*movabs<mode>_1): Add <ptrsize> PTR before
9376 operand 0 for intel asm alternative.
9377 (*movabs<mode>_2): Ditto for operand 1.
9378
f4f4204c
JG
93792013-08-12 James Greenhalgh <james.greenhalgh@arm.com>
9380
9381 * config/aarch64/arm_none.h
9382 (vdup<bhsd>_lane_<su><8,16,32,64>): Fix macro call.
9383
133b1a8e
NC
93842013-08-12 Nick Clifton <nickc@redhat.com>
9385
9386 * config.gcc (m32r-linux): Allow for tmake_file not being empty.
9387
15dd8b3a
YR
93882013-08-12 Yuri Rumyantsev <ysrumyan@gmail.com>
9389
9390 * config/i386/i386.md (floatunssi<mode>2 expand): Add new
9391 expand for QI/HImode operand to produce more effictive code for
9392 unsigned char(short) --> float(double) conversion.
9393
94109a6a
AM
93942013-08-12 Alexander Monakov <amonakov@ispras.ru>
9395
9396 * doc/invoke.texi: Mention that -ftls-model does not force the final
9397 model.
9398
4e115102
MP
93992013-08-12 Marek Polacek <polacek@redhat.com>
9400 Marc Glisse <marc.glisse@inria.fr>
9401
9402 PR tree-optimization/57980
9403 * tree-tailcall.c (process_assignment): Call build_minus_one_cst
9404 when creating -1 constant.
9405
748d71f3
JH
94062013-08-10 Jan Hubicka <jh@suse.cz>
9407
1f97e193 9408 Workaround binutils PR14342.
748d71f3
JH
9409 * tree-profile.c (init_ic_make_global_vars): Add LTO path.
9410 (gimple_init_edge_profiler): Likewise.
9411 (gimple_gen_ic_func_profiler): Likewise.
9412
5979aa54
JH
94132013-08-09 Jan Hubicka <jh@suse.cz>
9414
9415 * cgraph.c (cgraph_create_edge_1): Clear speculative flag.
9416
ad83025e
XDL
94172013-08-09 Xinliang David Li <davidxl@google.com>
9418
9419 * config/i386/stringop.def: New file.
9420 * config/i386/stringop.opt: New file.
9421 * config/i386/i386-opts.h: Include stringopt.def.
9422 * config/i386/i386.opt: Include stringopt.opt.
9423 * config/i386/i386.c (ix86_option_override_internal):
c97fd4b8 9424 Override default size based stringop inline strategies with options.
ad83025e
XDL
9425 * config/i386/i386.c (ix86_parse_stringop_strategy_string):
9426 New function.
9427
e59df5fd
JH
94282013-08-09 Jan Hubicka <jh@suse.cz>
9429
c97fd4b8 9430 * ipa-ref.c (ipa_clear_stmts_in_references): Clear lto_stmt_uid, too.
e59df5fd 9431
634ab819
JH
94322013-08-09 Jan Hubicka <jh@suse.cz>
9433
9434 * cgraph.c (cgraph_resolve_speculation): Cut frequency to
9435 CGRAPH_FREQ_MAX.
9436 (dump_cgraph_node): Dump profile-id.
9437 * cgraph.h (cgraph_indirect_call_info): Add common_target_id
9438 and common_target_probability.
9439 * lto-cgraph.c (lto_output_edge): Stream common targets.
9440 (lto_output_node): Stream profile ids.
9441 (input_node): Stream profile ids.
9442 (input_edge): Stream common targets.
9443 * lto-streamer-in.c (fixup_call_stmt_edges_1): Fix formatting.
9444 * ipa.c: Include value-prof.h
9445 (ipa_profile_generate_summary): Turn indirect call statement histograms
9446 into common targets.
9447 (ipa_profile): Turn common targets into speculative edges.
9448
2fa3d31b
JH
94492013-08-09 Jan Hubicka <jh@suse.cz>
9450
9451 * cgraph.h (cgraph_node): Add profile_id.
9452 * value-prof.c (cgraph_node_map): Turn into pointer_map.
c97fd4b8 9453 (init_node_map): Rewrite to handle hashes increas of incremental IDs.
2fa3d31b
JH
9454 (del_node_map): Update.
9455 (find_func_by_funcdef_no): Replace by ...
9456 (find_func_by_profile_id): ... this one.
9457 (gimple_ic_transform): Do not remove useful histograms when
9458 speculation is not done; dump info when indirect call removal
9459 can happen at LTO.
9460 * value-prof.h (find_func_by_profile_id, gimple_ic): Declare.
9461 * gcov-io.h (__gcov_indirect_call_profiler): Replace by ...
9462 (__gcov_indirect_call_profiler_v2): .. this one.
9463 * profile.h (init_node_map): Update.
9464 * coverage.c (coverage_compute_profile_id): New function.
9465 * coverage.h (coverage_compute_profile_id): Declare.
9466 * tree-profile.c (init_ic_make_global_vars): Make
9467 __gcov_indirect_call_callee and __gcov_indirect_call_counters global.
9468 (gimple_init_edge_profiler): Update prototype of
9469 __gcov_indirect_call_profiler.
9470 (gimple_gen_ic_func_profiler): Simplify.
9471 (tree_profiling): Use init_node_map
9472
042ae7d2
JH
94732013-08-09 Jan Hubicka <jh@suse.cz>
9474
c97fd4b8
UB
9475 * cgraphbuild.c (cgraph_rebuild_references): Rebuild only
9476 non-speculative refs.
042ae7d2
JH
9477 * cgraph.c (cgraph_update_edge_in_call_site_hash): New function.
9478 (cgraph_add_edge_to_call_site_hash): Deal with speculative calls.
9479 (cgraph_set_call_stmt): Likewise.
9480 (cgraph_create_edge_1): Fix release checking compilatoin;
9481 clear lto_stmt_uid.
9482 (cgraph_free_edge): Free indirect info.
9483 (cgraph_turn_edge_to_speculative): New function.
9484 (cgraph_speculative_call_info): New function.
9485 (cgraph_make_edge_direct): Return direct edge; handle speculation.
c97fd4b8 9486 (cgraph_redirect_edge_call_stmt_to_callee): Expand speculative edges.
042ae7d2
JH
9487 (dump_cgraph_node): Dump speculation.
9488 (verify_edge_count_and_frequency): Accept speculative edges.
9489 (verify_edge_corresponds_to_fndecl): Handle partitioned cgraph.
9490 (verify_cgraph_node): Handle speculation.
9491 * cgraph.h (cgraph_edge): Add SPECULATIVE flag.
9492 (cgraph_set_call_stmt): Update prototype.
9493 (cgraph_make_edge_direct): Update prototype.
9494 (cgraph_speculative_call_info): Declare.
9495 * ipa-cp.c (ipcp_discover_new_direct_edges): Be ready for edge
9496 to change; update call of ipa_find_references.
9497 * ipa-ref.c (ipa_record_reference): Fix return value; clear
9498 lto_stmt_uid and speculative flags.
9499 (ipa_dump_references): Dump speculation.
9500 (ipa_clone_references): Clone speculative flag.
9501 (ipa_clone_referring): Likewise.
9502 (ipa_clone_ref): New function.
9503 (ipa_find_reference): Look into lto_stmt_uids
9504 (ipa_clear_stmts_in_references): Do not clear speculative calls.
9505 * ipa-ref.h (ipa_ref): Add lto_stmt_uid and speculative flags.
9506 (ipa_find_reference): Update declaration.
9507 (ipa_clone_ref): Declare.
9508 * lto-cgraph.c (lto_output_edge): Make lto_stmt_uids start from 0;
9509 stream speculative flag.
9510 (lto_output_ref): Stream statements uids and speculation.
9511 (input_ref): Likewise.
9512 (input_edge): Stream speuclation.
9513 * cgraphclones.c (cgraph_clone_edge): Clone speculation.
9514 (cgraph_set_call_stmt_including_clones): Handle speculation.
9515 * ipa-inline.c (heap_edge_removal_hook): New function.
9516 (inline_small_functions): Register it.
9517 * lto-streamer-in.c (fixup_call_stmt_edges_1): Bounds checking;
9518 also initialize refs.
9519 * ipa-prop.c (ipa_make_edge_direct_to_target): Be ready for
9520 edge to change.
9521 (try_make_edge_direct_simple_call): Likewise.
9522 (try_make_edge_direct_simple_call): Likewise.
9523 (update_indirect_edges_after_inlining): Likewise.
9524 (remove_described_reference): Look proper lto_stmt_uid.
9525 (propagate_controlled_uses): Likewise.
9526 (propagate_controlled_uses): Liekwise.
9527 * tree-inline.c (copy_bb): Copy speculative edges.
9528 (redirect_all_calls): New function.
c97fd4b8
UB
9529 (copy_cfg_body): Do redirection after loop info is updated.
9530 (delete_unreachable_blocks_update_callgraph): Updadte speculation.
042ae7d2 9531
aa1e10cc
JH
95322013-08-09 Jan Hubicka <jh@suse.cz>
9533
9534 * lto-streamer-out.c (output_function): Renumber PHIs.
9535 * lto-streamer-in.c (input_function): Likewise.
9536
66adb8eb
JG
95372013-08-09 James Greenhalgh <james.greenhalgh@arm.com>
9538
9539 * config/aarch64/aarch64-simd-builtins.def (get_lane_signed): Remove.
9540 (get_lane_unsigned): Likewise.
9541 (dup_lane_scalar): Likewise.
9542 (get_lane): enable for VALL.
9543 * config/aarch64/aarch64-simd.md
9544 (aarch64_dup_lane_scalar<mode>): Remove.
9545 (aarch64_get_lane_signed<mode>): Likewise.
9546 (aarch64_get_lane_unsigned<mode>): Likewise.
9547 (aarch64_get_lane_extend<GPI:mode><VDQQH:mode>): New.
9548 (aarch64_get_lane_zero_extendsi<mode>): Likewise.
9549 (aarch64_get_lane<mode>): Enable for all vector modes.
9550 (aarch64_get_lanedi): Remove misleading constraints.
9551 * config/aarch64/arm_neon.h
9552 (__aarch64_vget_lane_any): Define.
9553 (__aarch64_vget<q>_lane_<fpsu><8,16,32,64>): Likewise.
9554 (vget<q>_lane_<fpsu><8,16,32,64>): Use __aarch64_vget_lane macros.
9555 (vdup<bhsd>_lane_<su><8,16,32,64>): Likewise.
9556 * config/aarch64/iterators.md (VDQQH): New.
9557 (VDQQHS): Likewise.
9558 (vwcore): Likewise.
9559
1f65ae7a
EB
95602013-08-09 Eric Botcazou <ebotcazou@adacore.com>
9561
9562 * configure.ac: Add GAS check for LEON instructions on SPARC.
9563 * configure: Regenerate.
9564 * config.in: Likewise.
9565 * config.gcc (with_cpu): Remove sparc-leon*-* and deal with LEON in the
9566 sparc*-*-* block.
9567 * config/sparc/sparc.opt (LEON, LEON3): New masks.
9568 * config/sparc/sparc.h (ASM_CPU32_DEFAULT_SPEC): Set to AS_LEON_FLAG
9569 for LEON or LEON3.
9570 (ASM_CPU_SPEC): Pass AS_LEON_FLAG if -mcpu=leon or -mcpu=leon3.
9571 (AS_LEON_FLAG): New macro.
9572 * config/sparc/sparc.c (sparc_option_override): Set MASK_LEON for leon
9573 and MASK_LEON3 for leon3 and unset them if HAVE_AS_LEON is not defined.
9574 Deal with LEON and LEON3 for the memory model.
23a6cb78 9575 * config/sparc/sync.md (atomic_compare_and_swap<mode>): Enable if LEON3
1f65ae7a
EB
9576 (atomic_compare_and_swap<mode>_1): Likewise.
9577 (*atomic_compare_and_swap<mode>_1): Likewise.
9578
ff522f7f
ZC
95792013-08-09 Zhenqiang Chen <zhenqiang.chen@linaro.org>
9580
9581 * config/arm/neon.md (vcond): Fix floating-point vector
9582 comparisons against 0.
9583
cb1cca12
VM
95842013-08-08 Vladimir Makarov <vmakarov@redhat.com>
9585
9586 * lra-constraints.c (emit_spill_move): Remove assert.
9587 (process_alt_operands): Add more debugging
9588 output. Increase reject for spilling into memory. Decrease
9589 reject for reloading scratch.
9590 (split_reg): Use HARD_REGNO_CALLER_SAVE_MODE.
9591
a3719e31
SE
95922013-08-08 Steve Ellcey <sellcey@mips.com>
9593
9594 * config/mips/mti-linux.h (SYSROOT_SUFFIX_SPEC): Add nan2008.
9595 * config/mips/t-mti-elf (MULTILIB_OPTIONS): Make mips16 and
9596 micromips incompatible. Add nan2008.
9597 (MULTILIB_DIRNAMES): Add nan2008.
9598 (MULTILIB_EXCEPTIONS): Remove mips16/micromips entry.
9599 * config/mips/t-mti-linux (MULTILIB_OPTIONS): Make mips16
9600 and micromips incompatible. Add nan2008.
9601 (MULTILIB_DIRNAMES): Add nan2008.
9602 (MULTILIB_EXCEPTIONS): Remove mips16/micromips entry.
9603
f40423e2
RS
96042013-08-08 Richard Sandiford <rdsandiford@googlemail.com>
9605
9606 PR rtl-optimization/58079
9607 * combine.c (combine_simplify_rtx): Avoid using SUBST if
9608 simplify_comparison has widened a comparison with an integer.
9609
869b9125
KT
96102013-08-08 Kyrylo Tkachov <kyrylo.tkachov@arm.com>
9611
9612 * config/arm/neon.md (movmisalign<mode>): Disable when we
9613 don't allow unaligned accesses.
9614 (*movmisalign<mode>_neon_store): Likewise.
9615 (*movmisalign<mode>_neon_load): Likewise.
9616 (*movmisalign<mode>_neon_store): Likewise.
9617 (*movmisalign<mode>_neon_load): Likewise.
9618
71cafea9
JH
96192013-08-08 Jan Hubicka <jh@suse.cz>
9620
9621 * cgraphbuild.c (build_cgraph_edges): Do not walk into debugs.
9622 (make_pass_rebuild_cgraph_edges): Also clear references.
9623 * cgraph.c (verify_cgraph_node): Add basic ipa-ref verifier.
9624 * ipa-inline-transform.c (inline_transform): Remove all references
9625 after inlining.
c97fd4b8
UB
9626 * cgraphunit.c (expand_function): Remove all references after
9627 expansion.
71cafea9
JH
9628 * ipa-ref.c (ipa_ref_has_aliases_p): Fix formatting.
9629 (ipa_find_reference): Rewrite to iterator.
9630 (remove_stmt_references): Likewise.
9631 (ipa_clear_stmts_in_references): New function.
9632 * ipa-ref.h (ipa_clear_stmts_in_references): Declare.
03b0ee0a
UB
9633 * cgraphclones.c (cgraph_materialize_all_clones): Remove or
9634 clear references.
71cafea9
JH
9635 * ipa-split.c (split_function): Remove references in split function.
9636
27d2e612
RE
96372013-08-08 Richard Earnshaw <rearnsha@arm.com>
9638
9639 PR target/57431
1f65ae7a 9640 * config/arm/arm/neon.md (neon_vld1_dupdi): New expand pattern.
27d2e612
RE
9641 (neon_vld1_dup<mode> VD iterator): Iterate over VD not VDX.
9642
5efc661c
RE
96432013-08-08 Richard Earnshaw <rearnsha@arm.com>
9644
9645 PR target/56979
1f65ae7a 9646 * config/arm/arm.c (aapcs_vfp_allocate): Decompose the argument if the
5efc661c
RE
9647 suggested mode for the assignment isn't compatible with the
9648 registers required.
9649
f276d31d
BE
96502013-08-08 Bernd Edlinger <bernd.edlinger@hotmail.de>
9651
9652 PR target/58065
9653 * config/arm/arm.h (MALLOC_ABI_ALIGNMENT): Define.
9654
4b8bc035
XDL
96552013-08-07 Xinliang David Li <davidxl@google.com>
9656
9657 * config/i386/i386.opt: New option -mtune-ctrl=.
9658 * config/i386/x86-tune.def: New file.
9659 * config/i386/i386.h: include x86-tune.def.
9660 * config/i386/i386.c (ix86_option_override_internal):
9661 Parsing -mtune-ctrl= option and set tune features.
9662
12211b99 96632013-08-07 Oleg Endo <olegendo@gcc.gnu.org>
a5b38740
OE
9664
9665 PR other/12081
03b0ee0a 9666 * config/rs6000/rs6000.c (gen_2arg_fn_t): Remove typedef.
a5b38740
OE
9667 (rs6000_emit_swdiv, rs6000_emit_swrsqrt): Don't cast result of GEN_FCN
9668 to gen_2arg_fn_t.
9669
5cf6635b
EB
96702013-08-07 Eric Botcazou <ebotcazou@adacore.com>
9671
9672 * rtl.h (update_alignments): Declare.
9673 * final.c (grow_label_align): New function extracted from...
9674 (shorten_branches): ...here. Call it.
9675 (update_alignments): New function.
9676 * reorg.c (sibling_labels): New variable.
9677 (get_label_before): Add SIBLING parameter. If it is non-zero, push
9678 the new label along with it onto the sibling_labels vector.
9679 (fill_simple_delay_slots): Adjust call to get_label_before.
9680 (fill_slots_from_thread): Likewise.
9681 (relax_delay_slots): Likewise.
9682 (make_return_insns): Likewise.
9683 (dbr_schedule): Invoke update_alignment on the sibling_labels vector.
9684
3c8ca1ab
EB
96852013-08-07 Eric Botcazou <ebotcazou@adacore.com>
9686
9687 * diagnostic.c (diagnostic_classify_diagnostic): Accept zero index and
9688 document its semantics.
9689 (diagnostic_report_diagnostic): Adjust accordingly.
9690
5ee5b32c
DM
96912013-08-07 David Malcolm <dmalcolm@redhat.com>
9692
03b0ee0a 9693 * config/sparc/sparc.c (insert_pass_work_around_errata): Move into...
5ee5b32c
DM
9694 (sparc_option_override): ...and port to new C++ pass API.
9695 * config/sparc/t-sparc (sparc.o): Add dep on CONTEXT_H
9696
abd566fa
PB
96972013-08-07 Peter Bergner <bergner@vnet.ibm.com>
9698
9699 * config/rs6000/rs6000.c (htm_expand_builtin) <case 0>: Remove.
9700
2077db1b
CT
97012013-08-06 Caroline Tice <cmtice@google.com>
9702
9703 * gcc.c (VTABLE_VERIFICATION_SPEC): New definition.
9704 (LINK_COMMAND_SPEC): Add VTABLE_VERIFICATION_SPEC.
9705 * tree-pass.h: Add pass_vtable_verify.
9706 * varasm.c (assemble_variable): Add code to properly set the comdat
9707 section and name for the .vtable_map_vars section.
9708 (assemble_vtyv_preinit_initializer): New function.
9709 (default_sectin_type_flags): Make sure .vtable_map_vars section has
9710 LINK_ONCE flag.
9711 * output.h: Add function decl for assemble_vtv_preinit_initializer.
9712 * vtable-verify.c: New file.
9713 * vtable-verify.h: New file.
9714 * flag-types.h (enum vtv_priority): Defintions for flag_vtable_verify
9715 initialiation levels.
9716 * timevar.def (TV_VTABLE_VERIFICATION): New definition.
9717 * passes.def: Insert pass_vtable_verify.
9718 * aclocal.m4: Reorder includes.
9719 * doc/invoke.texi: Document the -fvtable-verify=, -fvtv-debug, and
9720 -fvtv-counts options.
9721 * config/gnu-user.h (GNU_USER_TARGET_STARTFILE_SPEC): Add vtv_start*.o,
9722 as appropriate, if -fvtable-verify=... is used.
9723 (GNU_USER_TARGET_ENDFILE_SPEC): Add vtv_end*.o as appropriate, if
9724 -fvtable-verify=... is used.
9725 * Makefile.in (OBJS): Add vtable-verify.o to list.
9726 (vtable-verify.o): Add new build rule.
9727 (GTFILES): Add vtable-verify.c to list.
9728 * common.opt (fvtable-verify=): New flag.
9729 (vtv_priority): Values for fvtable-verify= flag.
9730 (fvtv-counts): New flag.
9731 (fvtv-debug): New flag.
9732 * tree.h (save_vtable_map_decl): New extern function decl.
9733
03085d1c
DM
97342013-08-07 David Malcolm <dmalcolm@redhat.com>
9735
9736 * config/rl78/rl78.c (rl78_devirt_pass): Convert from a struct to...
9737 (pass_rl78_devirt): ...new subclass of rtl_opt_pass along with...
9738 (pass_data_rl78_devirt): ...new pass_data instance and...
9739 (make_pass_rl78_devirt): ...new function.
9740 (rl78_asm_file_start): Port pass registration to new C++ API.
9741
05555c4a
DM
97422013-08-07 David Malcolm <dmalcolm@redhat.com>
9743
9744 * coretypes.h (rtl_opt_pass): Add.
9745 (gcc::context): Add.
9746 * config/epiphany/epiphany.c (pass_mode_switch_use): New.
9747 (epiphany_init): Port to new C++ pass API.
9748 (epiphany_optimize_mode_switching): Likewise.
9749 * pass_manager.h (pass_manager::get_pass_split_all_insns): New.
9750 (pass_manager::get_pass_mode_switching): New.
9751 (pass_manager::get_pass_peephole2): New.
9752 * mode-switching.c (pass_mode_switching): Add clone method.
9753 * recog.c (pass_peephole2): Add clone method.
9754 (pass_split_all_insns): Add clone method.
9755
c5a12e26
DM
97562013-08-06 David Malcolm <dmalcolm@redhat.com>
9757
03b0ee0a
UB
9758 * config/mips/mips.c (insert_pass_mips_machine_reorg2): Move into...
9759 (mips_option_override): ...here, porting to new C++ API for passes.
c5a12e26 9760
a2e2a668
JH
97612013-08-06 Jan Hubicka <jh@suse.cz>
9762
9763 * cgraph.c (cgraph_get_body): New function based on lto.c
9764 implementation.
9765 * cgraph.h (cgraph_get_body): Declare.
03b0ee0a
UB
9766 * cgraphclones.c (cgraph_create_virtual_clone): Commonize WPA and
9767 LTO paths.
a2e2a668
JH
9768 * cgraphunit.c (expand_function): Get body prior expanding.
9769 * ipa.c (function_and_variable_visibility): Use gimple_has_body_p test.
03b0ee0a
UB
9770 * lto-cgraph.c (lto_output_node): Do not stream bodies we don't
9771 really need.
a2e2a668
JH
9772 * passes.c (do_per_function_toporder): Get body.
9773 * tree-inline.c (expand_call_inline): Get body prior inlining it.
9774 * tree-ssa-structalias.c (ipa_pta_execute): Get body; skip clones.
9775
15682f24
MJ
97762013-08-06 Martin Jambor <mjambor@suse.cz>
9777
9778 PR fortran/57987
9779 * cgraphunit.c (cgraph_finalize_function): Assert that nested function
9780 is not re-finalized. Rename second parameter to no_collect.
9781
78f6dd68
MJ
97822013-08-06 Martin Jambor <mjambor@suse.cz>
9783
9784 PR middle-end/58041
9785 * gimple-ssa-strength-reduction.c (replace_ref): Make sure built
9786 MEM_REF has proper alignment information.
9787
12211b99 97882013-08-05 Oleg Endo <olegendo@gcc.gnu.org>
eb923e54
OE
9789
9790 PR other/12081
9791 * recog.h (rtx (*insn_gen_fn) (rtx, ...)): Replace typedef with new
9792 class insn_gen_fn.
9793 * expr.c (move_by_pieces_1, store_by_pieces_2): Replace argument
9794 rtx (*) (rtx, ...) with insn_gen_fn.
9795 * genoutput.c (output_insn_data): Cast gen_? function pointers to
9796 insn_gen_fn::stored_funcptr. Add initializer braces.
9797
8ac69a6c
DM
97982013-08-05 David Malcolm <dmalcolm@redhat.com>
9799
9800 Rewrite how instances of passes are cloned to remove assumptions
9801 about their sizes (thus allowing pass subclasses to have
9802 additional data fields, albeit non-GC-managed ones at this point).
9803
9804 * passes.c (make_pass_instance): Now that passes have clone
9805 methods, rewrite this function to eliminate XNEW and memcpy
9806 calls that used hardcoded sizes. Since this function no longer
9807 creates pass instances, rename it to...
03b0ee0a
UB
9808 (add_pass_instance): ...this. Document the old way that passes were
9809 numbered and flagged, and rework this function to continue using it.
8ac69a6c
DM
9810 (next_pass_1): Add an initial_pass argument for use by
9811 add_pass_instance.
9812 (position_pass): When adding multiple instances of a pass, use
9813 the pass's clone method, rather than relying on the XNEW/memcpy
9814 within the former make_pass_instance (now add_pass_instance).
03b0ee0a
UB
9815 (pass_manager::pass_manager): When invoking next_pass_1, also supply
9816 the initial instance of the current pass within the pass manager.
8ac69a6c 9817
27a4cd48
DM
98182013-08-05 David Malcolm <dmalcolm@redhat.com>
9819
9820 This is the automated part of the conversion of passes from C
9821 structs to C++ classes.
9822
9823 Patch autogenerated by refactor_passes.py from
9824 https://github.com/davidmalcolm/gcc-refactoring-scripts
9825 revision 03fe39476a4c4ea450b49e087cfa817b5f92021e
9826
9827 * asan.c (pass_asan): Convert from a global struct to a subclass of
9828 gimple_opt_pass along with...
9829 (pass_data_asan): ...new pass_data instance and...
9830 (make_pass_asan): ...new function.
9831 (pass_asan_O0): Convert from a global struct to a subclass of
9832 gimple_opt_pass along with...
9833 (pass_data_asan_O0): ...new pass_data instance and...
9834 (make_pass_asan_O0): ...new function.
9835 * auto-inc-dec.c (pass_inc_dec): Convert from a global struct to a
9836 subclass of rtl_opt_pass along with...
9837 (pass_data_inc_dec): ...new pass_data instance and...
9838 (make_pass_inc_dec): ...new function.
9839 * bb-reorder.c (pass_reorder_blocks): Convert from a global struct to
9840 a subclass of rtl_opt_pass along with...
9841 (pass_data_reorder_blocks): ...new pass_data instance and...
9842 (make_pass_reorder_blocks): ...new function.
9843 (pass_duplicate_computed_gotos): Convert from a global struct to a
9844 subclass of rtl_opt_pass along with...
9845 (pass_data_duplicate_computed_gotos): ...new pass_data instance and...
9846 (make_pass_duplicate_computed_gotos): ...new function.
9847 (pass_partition_blocks): Convert from a global struct to a subclass of
9848 rtl_opt_pass along with...
9849 (pass_data_partition_blocks): ...new pass_data instance and...
9850 (make_pass_partition_blocks): ...new function.
9851 * bt-load.c (pass_branch_target_load_optimize1): Convert from a global
9852 struct to a subclass of rtl_opt_pass along with...
9853 (pass_data_branch_target_load_optimize1): ...new pass_data instance
9854 and...
9855 (make_pass_branch_target_load_optimize1): ...new function.
9856 (pass_branch_target_load_optimize2): Convert from a global struct to a
9857 subclass of rtl_opt_pass along with...
9858 (pass_data_branch_target_load_optimize2): ...new pass_data instance
9859 and...
9860 (make_pass_branch_target_load_optimize2): ...new function.
9861 * cfgcleanup.c (pass_jump): Convert from a global struct to a subclass
9862 of rtl_opt_pass along with...
9863 (pass_data_jump): ...new pass_data instance and...
9864 (make_pass_jump): ...new function.
9865 (pass_jump2): Convert from a global struct to a subclass of
9866 rtl_opt_pass along with...
9867 (pass_data_jump2): ...new pass_data instance and...
9868 (make_pass_jump2): ...new function.
9869 * cfgexpand.c (pass_expand): Convert from a global struct to a
9870 subclass of rtl_opt_pass along with...
9871 (pass_data_expand): ...new pass_data instance and...
9872 (make_pass_expand): ...new function.
9873 * cfgrtl.c (pass_free_cfg): Convert from a global struct to a subclass
9874 of rtl_opt_pass along with...
9875 (pass_data_free_cfg): ...new pass_data instance and...
9876 (make_pass_free_cfg): ...new function.
9877 (pass_into_cfg_layout_mode): Convert from a global struct to a
9878 subclass of rtl_opt_pass along with...
9879 (pass_data_into_cfg_layout_mode): ...new pass_data instance and...
9880 (make_pass_into_cfg_layout_mode): ...new function.
9881 (pass_outof_cfg_layout_mode): Convert from a global struct to a
9882 subclass of rtl_opt_pass along with...
9883 (pass_data_outof_cfg_layout_mode): ...new pass_data instance and...
9884 (make_pass_outof_cfg_layout_mode): ...new function.
9885 * cgraphbuild.c (pass_build_cgraph_edges): Convert from a global
9886 struct to a subclass of gimple_opt_pass along with...
9887 (pass_data_build_cgraph_edges): ...new pass_data instance and...
9888 (make_pass_build_cgraph_edges): ...new function.
9889 (pass_rebuild_cgraph_edges): Convert from a global struct to a
9890 subclass of gimple_opt_pass along with...
9891 (pass_data_rebuild_cgraph_edges): ...new pass_data instance and...
9892 (make_pass_rebuild_cgraph_edges): ...new function.
9893 (pass_remove_cgraph_callee_edges): Convert from a global struct to a
9894 subclass of gimple_opt_pass along with...
9895 (pass_data_remove_cgraph_callee_edges): ...new pass_data instance
9896 and...
9897 (make_pass_remove_cgraph_callee_edges): ...new function.
9898 * combine-stack-adj.c (pass_stack_adjustments): Convert from a global
9899 struct to a subclass of rtl_opt_pass along with...
9900 (pass_data_stack_adjustments): ...new pass_data instance and...
9901 (make_pass_stack_adjustments): ...new function.
9902 * combine.c (pass_combine): Convert from a global struct to a subclass
9903 of rtl_opt_pass along with...
9904 (pass_data_combine): ...new pass_data instance and...
9905 (make_pass_combine): ...new function.
9906 * compare-elim.c (pass_compare_elim_after_reload): Convert from a
9907 global struct to a subclass of rtl_opt_pass along with...
9908 (pass_data_compare_elim_after_reload): ...new pass_data instance
9909 and...
9910 (make_pass_compare_elim_after_reload): ...new function.
9911 * cprop.c (pass_rtl_cprop): Convert from a global struct to a subclass
9912 of rtl_opt_pass along with...
9913 (pass_data_rtl_cprop): ...new pass_data instance and...
9914 (make_pass_rtl_cprop): ...new function.
9915 * cse.c (pass_cse): Convert from a global struct to a subclass of
9916 rtl_opt_pass along with...
9917 (pass_data_cse): ...new pass_data instance and...
9918 (make_pass_cse): ...new function.
9919 (pass_cse2): Convert from a global struct to a subclass of
9920 rtl_opt_pass along with...
9921 (pass_data_cse2): ...new pass_data instance and...
9922 (make_pass_cse2): ...new function.
9923 (pass_cse_after_global_opts): Convert from a global struct to a
9924 subclass of rtl_opt_pass along with...
9925 (pass_data_cse_after_global_opts): ...new pass_data instance and...
9926 (make_pass_cse_after_global_opts): ...new function.
9927 * dce.c (pass_ud_rtl_dce): Convert from a global struct to a subclass
9928 of rtl_opt_pass along with...
9929 (pass_data_ud_rtl_dce): ...new pass_data instance and...
9930 (make_pass_ud_rtl_dce): ...new function.
9931 (pass_fast_rtl_dce): Convert from a global struct to a subclass of
9932 rtl_opt_pass along with...
9933 (pass_data_fast_rtl_dce): ...new pass_data instance and...
9934 (make_pass_fast_rtl_dce): ...new function.
9935 * df-core.c (pass_df_initialize_opt): Convert from a global struct to
9936 a subclass of rtl_opt_pass along with...
9937 (pass_data_df_initialize_opt): ...new pass_data instance and...
9938 (make_pass_df_initialize_opt): ...new function.
9939 (pass_df_initialize_no_opt): Convert from a global struct to a
9940 subclass of rtl_opt_pass along with...
9941 (pass_data_df_initialize_no_opt): ...new pass_data instance and...
9942 (make_pass_df_initialize_no_opt): ...new function.
9943 (pass_df_finish): Convert from a global struct to a subclass of
9944 rtl_opt_pass along with...
9945 (pass_data_df_finish): ...new pass_data instance and...
9946 (make_pass_df_finish): ...new function.
9947 * dse.c (pass_rtl_dse1): Convert from a global struct to a subclass of
9948 rtl_opt_pass along with...
9949 (pass_data_rtl_dse1): ...new pass_data instance and...
9950 (make_pass_rtl_dse1): ...new function.
9951 (pass_rtl_dse2): Convert from a global struct to a subclass of
9952 rtl_opt_pass along with...
9953 (pass_data_rtl_dse2): ...new pass_data instance and...
9954 (make_pass_rtl_dse2): ...new function.
9955 * dwarf2cfi.c (pass_dwarf2_frame): Convert from a global struct to a
9956 subclass of rtl_opt_pass along with...
9957 (pass_data_dwarf2_frame): ...new pass_data instance and...
9958 (make_pass_dwarf2_frame): ...new function.
9959 * except.c (pass_set_nothrow_function_flags): Convert from a global
9960 struct to a subclass of rtl_opt_pass along with...
9961 (pass_data_set_nothrow_function_flags): ...new pass_data instance
9962 and...
9963 (make_pass_set_nothrow_function_flags): ...new function.
9964 (pass_convert_to_eh_region_ranges): Convert from a global struct to a
9965 subclass of rtl_opt_pass along with...
9966 (pass_data_convert_to_eh_region_ranges): ...new pass_data instance
9967 and...
9968 (make_pass_convert_to_eh_region_ranges): ...new function.
9969 * final.c (pass_compute_alignments): Convert from a global struct to a
9970 subclass of rtl_opt_pass along with...
9971 (pass_data_compute_alignments): ...new pass_data instance and...
9972 (make_pass_compute_alignments): ...new function.
9973 (pass_final): Convert from a global struct to a subclass of
9974 rtl_opt_pass along with...
9975 (pass_data_final): ...new pass_data instance and...
9976 (make_pass_final): ...new function.
9977 (pass_shorten_branches): Convert from a global struct to a subclass of
9978 rtl_opt_pass along with...
9979 (pass_data_shorten_branches): ...new pass_data instance and...
9980 (make_pass_shorten_branches): ...new function.
9981 (pass_clean_state): Convert from a global struct to a subclass of
9982 rtl_opt_pass along with...
9983 (pass_data_clean_state): ...new pass_data instance and...
9984 (make_pass_clean_state): ...new function.
9985 * function.c (pass_instantiate_virtual_regs): Convert from a global
9986 struct to a subclass of rtl_opt_pass along with...
9987 (pass_data_instantiate_virtual_regs): ...new pass_data instance and...
9988 (make_pass_instantiate_virtual_regs): ...new function.
9989 (pass_leaf_regs): Convert from a global struct to a subclass of
9990 rtl_opt_pass along with...
9991 (pass_data_leaf_regs): ...new pass_data instance and...
9992 (make_pass_leaf_regs): ...new function.
9993 (pass_thread_prologue_and_epilogue): Convert from a global struct to a
9994 subclass of rtl_opt_pass along with...
9995 (pass_data_thread_prologue_and_epilogue): ...new pass_data instance
9996 and...
9997 (make_pass_thread_prologue_and_epilogue): ...new function.
9998 (pass_match_asm_constraints): Convert from a global struct to a
9999 subclass of rtl_opt_pass along with...
10000 (pass_data_match_asm_constraints): ...new pass_data instance and...
10001 (make_pass_match_asm_constraints): ...new function.
10002 * fwprop.c (pass_rtl_fwprop): Convert from a global struct to a
10003 subclass of rtl_opt_pass along with...
10004 (pass_data_rtl_fwprop): ...new pass_data instance and...
10005 (make_pass_rtl_fwprop): ...new function.
10006 (pass_rtl_fwprop_addr): Convert from a global struct to a subclass of
10007 rtl_opt_pass along with...
10008 (pass_data_rtl_fwprop_addr): ...new pass_data instance and...
10009 (make_pass_rtl_fwprop_addr): ...new function.
10010 * gcse.c (pass_rtl_pre): Convert from a global struct to a subclass of
10011 rtl_opt_pass along with...
10012 (pass_data_rtl_pre): ...new pass_data instance and...
10013 (make_pass_rtl_pre): ...new function.
10014 (pass_rtl_hoist): Convert from a global struct to a subclass of
10015 rtl_opt_pass along with...
10016 (pass_data_rtl_hoist): ...new pass_data instance and...
10017 (make_pass_rtl_hoist): ...new function.
10018 * gimple-low.c (pass_lower_cf): Convert from a global struct to a
10019 subclass of gimple_opt_pass along with...
10020 (pass_data_lower_cf): ...new pass_data instance and...
10021 (make_pass_lower_cf): ...new function.
10022 * gimple-ssa-strength-reduction.c (pass_strength_reduction): Convert
10023 from a global struct to a subclass of gimple_opt_pass along with...
10024 (pass_data_strength_reduction): ...new pass_data instance and...
10025 (make_pass_strength_reduction): ...new function.
10026 * ifcvt.c (pass_rtl_ifcvt): Convert from a global struct to a subclass
10027 of rtl_opt_pass along with...
10028 (pass_data_rtl_ifcvt): ...new pass_data instance and...
10029 (make_pass_rtl_ifcvt): ...new function.
10030 (pass_if_after_combine): Convert from a global struct to a subclass of
10031 rtl_opt_pass along with...
10032 (pass_data_if_after_combine): ...new pass_data instance and...
10033 (make_pass_if_after_combine): ...new function.
10034 (pass_if_after_reload): Convert from a global struct to a subclass of
10035 rtl_opt_pass along with...
10036 (pass_data_if_after_reload): ...new pass_data instance and...
10037 (make_pass_if_after_reload): ...new function.
10038 * init-regs.c (pass_initialize_regs): Convert from a global struct to
10039 a subclass of rtl_opt_pass along with...
10040 (pass_data_initialize_regs): ...new pass_data instance and...
10041 (make_pass_initialize_regs): ...new function.
10042 * ipa-cp.c (pass_ipa_cp): Convert from a global struct to a subclass
10043 of ipa_opt_pass_d along with...
10044 (pass_data_ipa_cp): ...new pass_data instance and...
10045 (make_pass_ipa_cp): ...new function.
10046 * ipa-inline-analysis.c (pass_inline_parameters): Convert from a
10047 global struct to a subclass of gimple_opt_pass along with...
10048 (pass_data_inline_parameters): ...new pass_data instance and...
10049 (make_pass_inline_parameters): ...new function.
10050 * ipa-inline.c (pass_early_inline): Convert from a global struct to a
10051 subclass of gimple_opt_pass along with...
10052 (pass_data_early_inline): ...new pass_data instance and...
10053 (make_pass_early_inline): ...new function.
10054 (pass_ipa_inline): Convert from a global struct to a subclass of
10055 ipa_opt_pass_d along with...
10056 (pass_data_ipa_inline): ...new pass_data instance and...
10057 (make_pass_ipa_inline): ...new function.
10058 * ipa-pure-const.c (pass_local_pure_const): Convert from a global
10059 struct to a subclass of gimple_opt_pass along with...
10060 (pass_data_local_pure_const): ...new pass_data instance and...
10061 (make_pass_local_pure_const): ...new function.
10062 (pass_ipa_pure_const): Convert from a global struct to a subclass of
10063 ipa_opt_pass_d along with...
10064 (pass_data_ipa_pure_const): ...new pass_data instance and...
10065 (make_pass_ipa_pure_const): ...new function.
10066 * ipa-reference.c (pass_ipa_reference): Convert from a global struct
10067 to a subclass of ipa_opt_pass_d along with...
10068 (pass_data_ipa_reference): ...new pass_data instance and...
10069 (make_pass_ipa_reference): ...new function.
10070 * ipa-split.c (pass_split_functions): Convert from a global struct to
10071 a subclass of gimple_opt_pass along with...
10072 (pass_data_split_functions): ...new pass_data instance and...
10073 (make_pass_split_functions): ...new function.
10074 (pass_feedback_split_functions): Convert from a global struct to a
10075 subclass of gimple_opt_pass along with...
10076 (pass_data_feedback_split_functions): ...new pass_data instance and...
10077 (make_pass_feedback_split_functions): ...new function.
10078 * ipa.c (pass_ipa_function_and_variable_visibility): Convert from a
10079 global struct to a subclass of simple_ipa_opt_pass along with...
10080 (pass_data_ipa_function_and_variable_visibility): ...new pass_data
10081 instance and...
10082 (make_pass_ipa_function_and_variable_visibility): ...new function.
10083 (pass_ipa_free_inline_summary): Convert from a global struct to a
10084 subclass of simple_ipa_opt_pass along with...
10085 (pass_data_ipa_free_inline_summary): ...new pass_data instance and...
10086 (make_pass_ipa_free_inline_summary): ...new function.
10087 (pass_ipa_whole_program_visibility): Convert from a global struct to a
10088 subclass of ipa_opt_pass_d along with...
10089 (pass_data_ipa_whole_program_visibility): ...new pass_data instance
10090 and...
10091 (make_pass_ipa_whole_program_visibility): ...new function.
10092 (pass_ipa_profile): Convert from a global struct to a subclass of
10093 ipa_opt_pass_d along with...
10094 (pass_data_ipa_profile): ...new pass_data instance and...
10095 (make_pass_ipa_profile): ...new function.
10096 (pass_ipa_cdtor_merge): Convert from a global struct to a subclass of
10097 ipa_opt_pass_d along with...
10098 (pass_data_ipa_cdtor_merge): ...new pass_data instance and...
10099 (make_pass_ipa_cdtor_merge): ...new function.
10100 * ira.c (pass_ira): Convert from a global struct to a subclass of
10101 rtl_opt_pass along with...
10102 (pass_data_ira): ...new pass_data instance and...
10103 (make_pass_ira): ...new function.
10104 (pass_reload): Convert from a global struct to a subclass of
10105 rtl_opt_pass along with...
10106 (pass_data_reload): ...new pass_data instance and...
10107 (make_pass_reload): ...new function.
10108 * jump.c (pass_cleanup_barriers): Convert from a global struct to a
10109 subclass of rtl_opt_pass along with...
10110 (pass_data_cleanup_barriers): ...new pass_data instance and...
10111 (make_pass_cleanup_barriers): ...new function.
10112 * loop-init.c (pass_loop2): Convert from a global struct to a subclass
10113 of rtl_opt_pass along with...
10114 (pass_data_loop2): ...new pass_data instance and...
10115 (make_pass_loop2): ...new function.
10116 (pass_rtl_loop_init): Convert from a global struct to a subclass of
10117 rtl_opt_pass along with...
10118 (pass_data_rtl_loop_init): ...new pass_data instance and...
10119 (make_pass_rtl_loop_init): ...new function.
10120 (pass_rtl_loop_done): Convert from a global struct to a subclass of
10121 rtl_opt_pass along with...
10122 (pass_data_rtl_loop_done): ...new pass_data instance and...
10123 (make_pass_rtl_loop_done): ...new function.
10124 (pass_rtl_move_loop_invariants): Convert from a global struct to a
10125 subclass of rtl_opt_pass along with...
10126 (pass_data_rtl_move_loop_invariants): ...new pass_data instance and...
10127 (make_pass_rtl_move_loop_invariants): ...new function.
10128 (pass_rtl_unswitch): Convert from a global struct to a subclass of
10129 rtl_opt_pass along with...
10130 (pass_data_rtl_unswitch): ...new pass_data instance and...
10131 (make_pass_rtl_unswitch): ...new function.
10132 (pass_rtl_unroll_and_peel_loops): Convert from a global struct to a
10133 subclass of rtl_opt_pass along with...
10134 (pass_data_rtl_unroll_and_peel_loops): ...new pass_data instance
10135 and...
10136 (make_pass_rtl_unroll_and_peel_loops): ...new function.
10137 (pass_rtl_doloop): Convert from a global struct to a subclass of
10138 rtl_opt_pass along with...
10139 (pass_data_rtl_doloop): ...new pass_data instance and...
10140 (make_pass_rtl_doloop): ...new function.
10141 * lower-subreg.c (pass_lower_subreg): Convert from a global struct to
10142 a subclass of rtl_opt_pass along with...
10143 (pass_data_lower_subreg): ...new pass_data instance and...
10144 (make_pass_lower_subreg): ...new function.
10145 (pass_lower_subreg2): Convert from a global struct to a subclass of
10146 rtl_opt_pass along with...
10147 (pass_data_lower_subreg2): ...new pass_data instance and...
10148 (make_pass_lower_subreg2): ...new function.
10149 * lto-streamer-out.c (pass_ipa_lto_gimple_out): Convert from a global
10150 struct to a subclass of ipa_opt_pass_d along with...
10151 (pass_data_ipa_lto_gimple_out): ...new pass_data instance and...
10152 (make_pass_ipa_lto_gimple_out): ...new function.
10153 (pass_ipa_lto_finish_out): Convert from a global struct to a subclass
10154 of ipa_opt_pass_d along with...
10155 (pass_data_ipa_lto_finish_out): ...new pass_data instance and...
10156 (make_pass_ipa_lto_finish_out): ...new function.
10157 * mode-switching.c (pass_mode_switching): Convert from a global struct
10158 to a subclass of rtl_opt_pass along with...
10159 (pass_data_mode_switching): ...new pass_data instance and...
10160 (make_pass_mode_switching): ...new function.
10161 * modulo-sched.c (pass_sms): Convert from a global struct to a
10162 subclass of rtl_opt_pass along with...
10163 (pass_data_sms): ...new pass_data instance and...
10164 (make_pass_sms): ...new function.
10165 * omp-low.c (pass_expand_omp): Convert from a global struct to a
10166 subclass of gimple_opt_pass along with...
10167 (pass_data_expand_omp): ...new pass_data instance and...
10168 (make_pass_expand_omp): ...new function.
10169 (pass_lower_omp): Convert from a global struct to a subclass of
10170 gimple_opt_pass along with...
10171 (pass_data_lower_omp): ...new pass_data instance and...
10172 (make_pass_lower_omp): ...new function.
10173 (pass_diagnose_omp_blocks): Convert from a global struct to a subclass
10174 of gimple_opt_pass along with...
10175 (pass_data_diagnose_omp_blocks): ...new pass_data instance and...
10176 (make_pass_diagnose_omp_blocks): ...new function.
10177 * passes.c (pass_early_local_passes): Convert from a global struct to
10178 a subclass of simple_ipa_opt_pass along with...
10179 (pass_data_early_local_passes): ...new pass_data instance and...
10180 (make_pass_early_local_passes): ...new function.
10181 (pass_all_early_optimizations): Convert from a global struct to a
10182 subclass of gimple_opt_pass along with...
10183 (pass_data_all_early_optimizations): ...new pass_data instance and...
10184 (make_pass_all_early_optimizations): ...new function.
10185 (pass_all_optimizations): Convert from a global struct to a subclass
10186 of gimple_opt_pass along with...
10187 (pass_data_all_optimizations): ...new pass_data instance and...
10188 (make_pass_all_optimizations): ...new function.
10189 (pass_all_optimizations_g): Convert from a global struct to a subclass
10190 of gimple_opt_pass along with...
10191 (pass_data_all_optimizations_g): ...new pass_data instance and...
10192 (make_pass_all_optimizations_g): ...new function.
10193 (pass_rest_of_compilation): Convert from a global struct to a subclass
10194 of rtl_opt_pass along with...
10195 (pass_data_rest_of_compilation): ...new pass_data instance and...
10196 (make_pass_rest_of_compilation): ...new function.
10197 (pass_postreload): Convert from a global struct to a subclass of
10198 rtl_opt_pass along with...
10199 (pass_data_postreload): ...new pass_data instance and...
10200 (make_pass_postreload): ...new function.
10201 * postreload-gcse.c (pass_gcse2): Convert from a global struct to a
10202 subclass of rtl_opt_pass along with...
10203 (pass_data_gcse2): ...new pass_data instance and...
10204 (make_pass_gcse2): ...new function.
10205 * postreload.c (pass_postreload_cse): Convert from a global struct to
10206 a subclass of rtl_opt_pass along with...
10207 (pass_data_postreload_cse): ...new pass_data instance and...
10208 (make_pass_postreload_cse): ...new function.
10209 * predict.c (pass_profile): Convert from a global struct to a subclass
10210 of gimple_opt_pass along with...
10211 (pass_data_profile): ...new pass_data instance and...
10212 (make_pass_profile): ...new function.
10213 (pass_strip_predict_hints): Convert from a global struct to a subclass
10214 of gimple_opt_pass along with...
10215 (pass_data_strip_predict_hints): ...new pass_data instance and...
10216 (make_pass_strip_predict_hints): ...new function.
10217 * recog.c (pass_peephole2): Convert from a global struct to a subclass
10218 of rtl_opt_pass along with...
10219 (pass_data_peephole2): ...new pass_data instance and...
10220 (make_pass_peephole2): ...new function.
10221 (pass_split_all_insns): Convert from a global struct to a subclass of
10222 rtl_opt_pass along with...
10223 (pass_data_split_all_insns): ...new pass_data instance and...
10224 (make_pass_split_all_insns): ...new function.
10225 (pass_split_after_reload): Convert from a global struct to a subclass
10226 of rtl_opt_pass along with...
10227 (pass_data_split_after_reload): ...new pass_data instance and...
10228 (make_pass_split_after_reload): ...new function.
10229 (pass_split_before_regstack): Convert from a global struct to a
10230 subclass of rtl_opt_pass along with...
10231 (pass_data_split_before_regstack): ...new pass_data instance and...
10232 (make_pass_split_before_regstack): ...new function.
10233 (pass_split_before_sched2): Convert from a global struct to a subclass
10234 of rtl_opt_pass along with...
10235 (pass_data_split_before_sched2): ...new pass_data instance and...
10236 (make_pass_split_before_sched2): ...new function.
10237 (pass_split_for_shorten_branches): Convert from a global struct to a
10238 subclass of rtl_opt_pass along with...
10239 (pass_data_split_for_shorten_branches): ...new pass_data instance
10240 and...
10241 (make_pass_split_for_shorten_branches): ...new function.
10242 * ree.c (pass_ree): Convert from a global struct to a subclass of
10243 rtl_opt_pass along with...
10244 (pass_data_ree): ...new pass_data instance and...
10245 (make_pass_ree): ...new function.
10246 * reg-stack.c (pass_stack_regs): Convert from a global struct to a
10247 subclass of rtl_opt_pass along with...
10248 (pass_data_stack_regs): ...new pass_data instance and...
10249 (make_pass_stack_regs): ...new function.
10250 (pass_stack_regs_run): Convert from a global struct to a subclass of
10251 rtl_opt_pass along with...
10252 (pass_data_stack_regs_run): ...new pass_data instance and...
10253 (make_pass_stack_regs_run): ...new function.
10254 * regcprop.c (pass_cprop_hardreg): Convert from a global struct to a
10255 subclass of rtl_opt_pass along with...
10256 (pass_data_cprop_hardreg): ...new pass_data instance and...
10257 (make_pass_cprop_hardreg): ...new function.
10258 * reginfo.c (pass_reginfo_init): Convert from a global struct to a
10259 subclass of rtl_opt_pass along with...
10260 (pass_data_reginfo_init): ...new pass_data instance and...
10261 (make_pass_reginfo_init): ...new function.
10262 * regmove.c (pass_regmove): Convert from a global struct to a subclass
10263 of rtl_opt_pass along with...
10264 (pass_data_regmove): ...new pass_data instance and...
10265 (make_pass_regmove): ...new function.
10266 * regrename.c (pass_regrename): Convert from a global struct to a
10267 subclass of rtl_opt_pass along with...
10268 (pass_data_regrename): ...new pass_data instance and...
10269 (make_pass_regrename): ...new function.
10270 * reorg.c (pass_delay_slots): Convert from a global struct to a
10271 subclass of rtl_opt_pass along with...
10272 (pass_data_delay_slots): ...new pass_data instance and...
10273 (make_pass_delay_slots): ...new function.
10274 (pass_machine_reorg): Convert from a global struct to a subclass of
10275 rtl_opt_pass along with...
10276 (pass_data_machine_reorg): ...new pass_data instance and...
10277 (make_pass_machine_reorg): ...new function.
10278 * sched-rgn.c (pass_sched): Convert from a global struct to a subclass
10279 of rtl_opt_pass along with...
10280 (pass_data_sched): ...new pass_data instance and...
10281 (make_pass_sched): ...new function.
10282 (pass_sched2): Convert from a global struct to a subclass of
10283 rtl_opt_pass along with...
10284 (pass_data_sched2): ...new pass_data instance and...
10285 (make_pass_sched2): ...new function.
10286 * stack-ptr-mod.c (pass_stack_ptr_mod): Convert from a global struct
10287 to a subclass of rtl_opt_pass along with...
10288 (pass_data_stack_ptr_mod): ...new pass_data instance and...
10289 (make_pass_stack_ptr_mod): ...new function.
10290 * store-motion.c (pass_rtl_store_motion): Convert from a global struct
10291 to a subclass of rtl_opt_pass along with...
10292 (pass_data_rtl_store_motion): ...new pass_data instance and...
10293 (make_pass_rtl_store_motion): ...new function.
10294 * tracer.c (pass_tracer): Convert from a global struct to a subclass
10295 of gimple_opt_pass along with...
10296 (pass_data_tracer): ...new pass_data instance and...
10297 (make_pass_tracer): ...new function.
10298 * trans-mem.c (pass_diagnose_tm_blocks): Convert from a global struct
10299 to a subclass of gimple_opt_pass along with...
10300 (pass_data_diagnose_tm_blocks): ...new pass_data instance and...
10301 (make_pass_diagnose_tm_blocks): ...new function.
10302 (pass_lower_tm): Convert from a global struct to a subclass of
10303 gimple_opt_pass along with...
10304 (pass_data_lower_tm): ...new pass_data instance and...
10305 (make_pass_lower_tm): ...new function.
10306 (pass_tm_init): Convert from a global struct to a subclass of
10307 gimple_opt_pass along with...
10308 (pass_data_tm_init): ...new pass_data instance and...
10309 (make_pass_tm_init): ...new function.
10310 (pass_tm_mark): Convert from a global struct to a subclass of
10311 gimple_opt_pass along with...
10312 (pass_data_tm_mark): ...new pass_data instance and...
10313 (make_pass_tm_mark): ...new function.
10314 (pass_tm_edges): Convert from a global struct to a subclass of
10315 gimple_opt_pass along with...
10316 (pass_data_tm_edges): ...new pass_data instance and...
10317 (make_pass_tm_edges): ...new function.
10318 (pass_tm_memopt): Convert from a global struct to a subclass of
10319 gimple_opt_pass along with...
10320 (pass_data_tm_memopt): ...new pass_data instance and...
10321 (make_pass_tm_memopt): ...new function.
10322 (pass_ipa_tm): Convert from a global struct to a subclass of
10323 simple_ipa_opt_pass along with...
10324 (pass_data_ipa_tm): ...new pass_data instance and...
10325 (make_pass_ipa_tm): ...new function.
10326 * tree-call-cdce.c (pass_call_cdce): Convert from a global struct to a
10327 subclass of gimple_opt_pass along with...
10328 (pass_data_call_cdce): ...new pass_data instance and...
10329 (make_pass_call_cdce): ...new function.
10330 * tree-cfg.c (pass_build_cfg): Convert from a global struct to a
10331 subclass of gimple_opt_pass along with...
10332 (pass_data_build_cfg): ...new pass_data instance and...
10333 (make_pass_build_cfg): ...new function.
10334 (pass_split_crit_edges): Convert from a global struct to a subclass of
10335 gimple_opt_pass along with...
10336 (pass_data_split_crit_edges): ...new pass_data instance and...
10337 (make_pass_split_crit_edges): ...new function.
10338 (pass_warn_function_return): Convert from a global struct to a
10339 subclass of gimple_opt_pass along with...
10340 (pass_data_warn_function_return): ...new pass_data instance and...
10341 (make_pass_warn_function_return): ...new function.
10342 (pass_warn_function_noreturn): Convert from a global struct to a
10343 subclass of gimple_opt_pass along with...
10344 (pass_data_warn_function_noreturn): ...new pass_data instance and...
10345 (make_pass_warn_function_noreturn): ...new function.
10346 (pass_warn_unused_result): Convert from a global struct to a subclass
10347 of gimple_opt_pass along with...
10348 (pass_data_warn_unused_result): ...new pass_data instance and...
10349 (make_pass_warn_unused_result): ...new function.
10350 * tree-cfgcleanup.c (pass_merge_phi): Convert from a global struct to
10351 a subclass of gimple_opt_pass along with...
10352 (pass_data_merge_phi): ...new pass_data instance and...
10353 (make_pass_merge_phi): ...new function.
10354 * tree-complex.c (pass_lower_complex): Convert from a global struct to
10355 a subclass of gimple_opt_pass along with...
10356 (pass_data_lower_complex): ...new pass_data instance and...
10357 (make_pass_lower_complex): ...new function.
10358 (pass_lower_complex_O0): Convert from a global struct to a subclass of
10359 gimple_opt_pass along with...
10360 (pass_data_lower_complex_O0): ...new pass_data instance and...
10361 (make_pass_lower_complex_O0): ...new function.
10362 * tree-eh.c (pass_lower_eh): Convert from a global struct to a
10363 subclass of gimple_opt_pass along with...
10364 (pass_data_lower_eh): ...new pass_data instance and...
10365 (make_pass_lower_eh): ...new function.
10366 (pass_refactor_eh): Convert from a global struct to a subclass of
10367 gimple_opt_pass along with...
10368 (pass_data_refactor_eh): ...new pass_data instance and...
10369 (make_pass_refactor_eh): ...new function.
10370 (pass_lower_resx): Convert from a global struct to a subclass of
10371 gimple_opt_pass along with...
10372 (pass_data_lower_resx): ...new pass_data instance and...
10373 (make_pass_lower_resx): ...new function.
10374 (pass_lower_eh_dispatch): Convert from a global struct to a subclass
10375 of gimple_opt_pass along with...
10376 (pass_data_lower_eh_dispatch): ...new pass_data instance and...
10377 (make_pass_lower_eh_dispatch): ...new function.
10378 (pass_cleanup_eh): Convert from a global struct to a subclass of
10379 gimple_opt_pass along with...
10380 (pass_data_cleanup_eh): ...new pass_data instance and...
10381 (make_pass_cleanup_eh): ...new function.
10382 * tree-emutls.c (pass_ipa_lower_emutls): Convert from a global struct
10383 to a subclass of simple_ipa_opt_pass along with...
10384 (pass_data_ipa_lower_emutls): ...new pass_data instance and...
10385 (make_pass_ipa_lower_emutls): ...new function.
10386 * tree-if-conv.c (pass_if_conversion): Convert from a global struct to
10387 a subclass of gimple_opt_pass along with...
10388 (pass_data_if_conversion): ...new pass_data instance and...
10389 (make_pass_if_conversion): ...new function.
10390 * tree-into-ssa.c (pass_build_ssa): Convert from a global struct to a
10391 subclass of gimple_opt_pass along with...
10392 (pass_data_build_ssa): ...new pass_data instance and...
10393 (make_pass_build_ssa): ...new function.
10394 * tree-loop-distribution.c (pass_loop_distribution): Convert from a
10395 global struct to a subclass of gimple_opt_pass along with...
10396 (pass_data_loop_distribution): ...new pass_data instance and...
10397 (make_pass_loop_distribution): ...new function.
10398 * tree-mudflap.c (pass_mudflap_1): Convert from a global struct to a
10399 subclass of gimple_opt_pass along with...
10400 (pass_data_mudflap_1): ...new pass_data instance and...
10401 (make_pass_mudflap_1): ...new function.
10402 (pass_mudflap_2): Convert from a global struct to a subclass of
10403 gimple_opt_pass along with...
10404 (pass_data_mudflap_2): ...new pass_data instance and...
10405 (make_pass_mudflap_2): ...new function.
10406 * tree-nomudflap.c (pass_mudflap_1): Convert from a global struct to a
10407 subclass of gimple_opt_pass along with...
10408 (pass_data_mudflap_1): ...new pass_data instance and...
10409 (make_pass_mudflap_1): ...new function.
10410 (pass_mudflap_2): Convert from a global struct to a subclass of
10411 gimple_opt_pass along with...
10412 (pass_data_mudflap_2): ...new pass_data instance and...
10413 (make_pass_mudflap_2): ...new function.
10414 * tree-nrv.c (pass_nrv): Convert from a global struct to a subclass of
10415 gimple_opt_pass along with...
10416 (pass_data_nrv): ...new pass_data instance and...
10417 (make_pass_nrv): ...new function.
10418 (pass_return_slot): Convert from a global struct to a subclass of
10419 gimple_opt_pass along with...
10420 (pass_data_return_slot): ...new pass_data instance and...
10421 (make_pass_return_slot): ...new function.
10422 * tree-object-size.c (pass_object_sizes): Convert from a global struct
10423 to a subclass of gimple_opt_pass along with...
10424 (pass_data_object_sizes): ...new pass_data instance and...
10425 (make_pass_object_sizes): ...new function.
10426 * tree-optimize.c (pass_cleanup_cfg_post_optimizing): Convert from a
10427 global struct to a subclass of gimple_opt_pass along with...
10428 (pass_data_cleanup_cfg_post_optimizing): ...new pass_data instance
10429 and...
10430 (make_pass_cleanup_cfg_post_optimizing): ...new function.
10431 (pass_fixup_cfg): Convert from a global struct to a subclass of
10432 gimple_opt_pass along with...
10433 (pass_data_fixup_cfg): ...new pass_data instance and...
10434 (make_pass_fixup_cfg): ...new function.
10435 * tree-pass.h (pass_mudflap_1): Replace declaration with that of...
10436 (make_pass_mudflap_1): ...new function.
10437 (pass_mudflap_2): Replace declaration with that of...
10438 (make_pass_mudflap_2): ...new function.
10439 (pass_asan): Replace declaration with that of...
10440 (make_pass_asan): ...new function.
10441 (pass_asan_O0): Replace declaration with that of...
10442 (make_pass_asan_O0): ...new function.
10443 (pass_tsan): Replace declaration with that of...
10444 (make_pass_tsan): ...new function.
10445 (pass_tsan_O0): Replace declaration with that of...
10446 (make_pass_tsan_O0): ...new function.
10447 (pass_lower_cf): Replace declaration with that of...
10448 (make_pass_lower_cf): ...new function.
10449 (pass_refactor_eh): Replace declaration with that of...
10450 (make_pass_refactor_eh): ...new function.
10451 (pass_lower_eh): Replace declaration with that of...
10452 (make_pass_lower_eh): ...new function.
10453 (pass_lower_eh_dispatch): Replace declaration with that of...
10454 (make_pass_lower_eh_dispatch): ...new function.
10455 (pass_lower_resx): Replace declaration with that of...
10456 (make_pass_lower_resx): ...new function.
10457 (pass_build_cfg): Replace declaration with that of...
10458 (make_pass_build_cfg): ...new function.
10459 (pass_early_tree_profile): Replace declaration with that of...
10460 (make_pass_early_tree_profile): ...new function.
10461 (pass_cleanup_eh): Replace declaration with that of...
10462 (make_pass_cleanup_eh): ...new function.
10463 (pass_sra): Replace declaration with that of...
10464 (make_pass_sra): ...new function.
10465 (pass_sra_early): Replace declaration with that of...
10466 (make_pass_sra_early): ...new function.
10467 (pass_early_ipa_sra): Replace declaration with that of...
10468 (make_pass_early_ipa_sra): ...new function.
10469 (pass_tail_recursion): Replace declaration with that of...
10470 (make_pass_tail_recursion): ...new function.
10471 (pass_tail_calls): Replace declaration with that of...
10472 (make_pass_tail_calls): ...new function.
10473 (pass_tree_loop): Replace declaration with that of...
10474 (make_pass_tree_loop): ...new function.
10475 (pass_tree_loop_init): Replace declaration with that of...
10476 (make_pass_tree_loop_init): ...new function.
10477 (pass_lim): Replace declaration with that of...
10478 (make_pass_lim): ...new function.
10479 (pass_tree_unswitch): Replace declaration with that of...
10480 (make_pass_tree_unswitch): ...new function.
10481 (pass_predcom): Replace declaration with that of...
10482 (make_pass_predcom): ...new function.
10483 (pass_iv_canon): Replace declaration with that of...
10484 (make_pass_iv_canon): ...new function.
10485 (pass_scev_cprop): Replace declaration with that of...
10486 (make_pass_scev_cprop): ...new function.
10487 (pass_empty_loop): Replace declaration with that of...
10488 (make_pass_empty_loop): ...new function.
10489 (pass_record_bounds): Replace declaration with that of...
10490 (make_pass_record_bounds): ...new function.
10491 (pass_graphite): Replace declaration with that of...
10492 (make_pass_graphite): ...new function.
10493 (pass_graphite_transforms): Replace declaration with that of...
10494 (make_pass_graphite_transforms): ...new function.
10495 (pass_if_conversion): Replace declaration with that of...
10496 (make_pass_if_conversion): ...new function.
10497 (pass_loop_distribution): Replace declaration with that of...
10498 (make_pass_loop_distribution): ...new function.
10499 (pass_vectorize): Replace declaration with that of...
10500 (make_pass_vectorize): ...new function.
10501 (pass_slp_vectorize): Replace declaration with that of...
10502 (make_pass_slp_vectorize): ...new function.
10503 (pass_complete_unroll): Replace declaration with that of...
10504 (make_pass_complete_unroll): ...new function.
10505 (pass_complete_unrolli): Replace declaration with that of...
10506 (make_pass_complete_unrolli): ...new function.
10507 (pass_parallelize_loops): Replace declaration with that of...
10508 (make_pass_parallelize_loops): ...new function.
10509 (pass_loop_prefetch): Replace declaration with that of...
10510 (make_pass_loop_prefetch): ...new function.
10511 (pass_iv_optimize): Replace declaration with that of...
10512 (make_pass_iv_optimize): ...new function.
10513 (pass_tree_loop_done): Replace declaration with that of...
10514 (make_pass_tree_loop_done): ...new function.
10515 (pass_ch): Replace declaration with that of...
10516 (make_pass_ch): ...new function.
10517 (pass_ccp): Replace declaration with that of...
10518 (make_pass_ccp): ...new function.
10519 (pass_phi_only_cprop): Replace declaration with that of...
10520 (make_pass_phi_only_cprop): ...new function.
10521 (pass_build_ssa): Replace declaration with that of...
10522 (make_pass_build_ssa): ...new function.
10523 (pass_build_alias): Replace declaration with that of...
10524 (make_pass_build_alias): ...new function.
10525 (pass_build_ealias): Replace declaration with that of...
10526 (make_pass_build_ealias): ...new function.
10527 (pass_dominator): Replace declaration with that of...
10528 (make_pass_dominator): ...new function.
10529 (pass_dce): Replace declaration with that of...
10530 (make_pass_dce): ...new function.
10531 (pass_dce_loop): Replace declaration with that of...
10532 (make_pass_dce_loop): ...new function.
10533 (pass_cd_dce): Replace declaration with that of...
10534 (make_pass_cd_dce): ...new function.
10535 (pass_call_cdce): Replace declaration with that of...
10536 (make_pass_call_cdce): ...new function.
10537 (pass_merge_phi): Replace declaration with that of...
10538 (make_pass_merge_phi): ...new function.
10539 (pass_split_crit_edges): Replace declaration with that of...
10540 (make_pass_split_crit_edges): ...new function.
10541 (pass_pre): Replace declaration with that of...
10542 (make_pass_pre): ...new function.
10543 (pass_profile): Replace declaration with that of...
10544 (make_pass_profile): ...new function.
10545 (pass_strip_predict_hints): Replace declaration with that of...
10546 (make_pass_strip_predict_hints): ...new function.
10547 (pass_lower_complex_O0): Replace declaration with that of...
10548 (make_pass_lower_complex_O0): ...new function.
10549 (pass_lower_complex): Replace declaration with that of...
10550 (make_pass_lower_complex): ...new function.
10551 (pass_lower_vector): Replace declaration with that of...
10552 (make_pass_lower_vector): ...new function.
10553 (pass_lower_vector_ssa): Replace declaration with that of...
10554 (make_pass_lower_vector_ssa): ...new function.
10555 (pass_lower_omp): Replace declaration with that of...
10556 (make_pass_lower_omp): ...new function.
10557 (pass_diagnose_omp_blocks): Replace declaration with that of...
10558 (make_pass_diagnose_omp_blocks): ...new function.
10559 (pass_expand_omp): Replace declaration with that of...
10560 (make_pass_expand_omp): ...new function.
10561 (pass_expand_omp_ssa): Replace declaration with that of...
10562 (make_pass_expand_omp_ssa): ...new function.
10563 (pass_object_sizes): Replace declaration with that of...
10564 (make_pass_object_sizes): ...new function.
10565 (pass_strlen): Replace declaration with that of...
10566 (make_pass_strlen): ...new function.
10567 (pass_fold_builtins): Replace declaration with that of...
10568 (make_pass_fold_builtins): ...new function.
10569 (pass_stdarg): Replace declaration with that of...
10570 (make_pass_stdarg): ...new function.
10571 (pass_early_warn_uninitialized): Replace declaration with that of...
10572 (make_pass_early_warn_uninitialized): ...new function.
10573 (pass_late_warn_uninitialized): Replace declaration with that of...
10574 (make_pass_late_warn_uninitialized): ...new function.
10575 (pass_cse_reciprocals): Replace declaration with that of...
10576 (make_pass_cse_reciprocals): ...new function.
10577 (pass_cse_sincos): Replace declaration with that of...
10578 (make_pass_cse_sincos): ...new function.
10579 (pass_optimize_bswap): Replace declaration with that of...
10580 (make_pass_optimize_bswap): ...new function.
10581 (pass_optimize_widening_mul): Replace declaration with that of...
10582 (make_pass_optimize_widening_mul): ...new function.
10583 (pass_warn_function_return): Replace declaration with that of...
10584 (make_pass_warn_function_return): ...new function.
10585 (pass_warn_function_noreturn): Replace declaration with that of...
10586 (make_pass_warn_function_noreturn): ...new function.
10587 (pass_cselim): Replace declaration with that of...
10588 (make_pass_cselim): ...new function.
10589 (pass_phiopt): Replace declaration with that of...
10590 (make_pass_phiopt): ...new function.
10591 (pass_forwprop): Replace declaration with that of...
10592 (make_pass_forwprop): ...new function.
10593 (pass_phiprop): Replace declaration with that of...
10594 (make_pass_phiprop): ...new function.
10595 (pass_tree_ifcombine): Replace declaration with that of...
10596 (make_pass_tree_ifcombine): ...new function.
10597 (pass_dse): Replace declaration with that of...
10598 (make_pass_dse): ...new function.
10599 (pass_nrv): Replace declaration with that of...
10600 (make_pass_nrv): ...new function.
10601 (pass_rename_ssa_copies): Replace declaration with that of...
10602 (make_pass_rename_ssa_copies): ...new function.
10603 (pass_sink_code): Replace declaration with that of...
10604 (make_pass_sink_code): ...new function.
10605 (pass_fre): Replace declaration with that of...
10606 (make_pass_fre): ...new function.
10607 (pass_check_data_deps): Replace declaration with that of...
10608 (make_pass_check_data_deps): ...new function.
10609 (pass_copy_prop): Replace declaration with that of...
10610 (make_pass_copy_prop): ...new function.
10611 (pass_vrp): Replace declaration with that of...
10612 (make_pass_vrp): ...new function.
10613 (pass_uncprop): Replace declaration with that of...
10614 (make_pass_uncprop): ...new function.
10615 (pass_return_slot): Replace declaration with that of...
10616 (make_pass_return_slot): ...new function.
10617 (pass_reassoc): Replace declaration with that of...
10618 (make_pass_reassoc): ...new function.
10619 (pass_rebuild_cgraph_edges): Replace declaration with that of...
10620 (make_pass_rebuild_cgraph_edges): ...new function.
10621 (pass_remove_cgraph_callee_edges): Replace declaration with that of...
10622 (make_pass_remove_cgraph_callee_edges): ...new function.
10623 (pass_build_cgraph_edges): Replace declaration with that of...
10624 (make_pass_build_cgraph_edges): ...new function.
10625 (pass_local_pure_const): Replace declaration with that of...
10626 (make_pass_local_pure_const): ...new function.
10627 (pass_tracer): Replace declaration with that of...
10628 (make_pass_tracer): ...new function.
10629 (pass_warn_unused_result): Replace declaration with that of...
10630 (make_pass_warn_unused_result): ...new function.
10631 (pass_diagnose_tm_blocks): Replace declaration with that of...
10632 (make_pass_diagnose_tm_blocks): ...new function.
10633 (pass_lower_tm): Replace declaration with that of...
10634 (make_pass_lower_tm): ...new function.
10635 (pass_tm_init): Replace declaration with that of...
10636 (make_pass_tm_init): ...new function.
10637 (pass_tm_mark): Replace declaration with that of...
10638 (make_pass_tm_mark): ...new function.
10639 (pass_tm_memopt): Replace declaration with that of...
10640 (make_pass_tm_memopt): ...new function.
10641 (pass_tm_edges): Replace declaration with that of...
10642 (make_pass_tm_edges): ...new function.
10643 (pass_split_functions): Replace declaration with that of...
10644 (make_pass_split_functions): ...new function.
10645 (pass_feedback_split_functions): Replace declaration with that of...
10646 (make_pass_feedback_split_functions): ...new function.
10647 (pass_strength_reduction): Replace declaration with that of...
10648 (make_pass_strength_reduction): ...new function.
10649 (pass_ipa_lower_emutls): Replace declaration with that of...
10650 (make_pass_ipa_lower_emutls): ...new function.
10651 (pass_ipa_function_and_variable_visibility): Replace declaration with
10652 that of...
10653 (make_pass_ipa_function_and_variable_visibility): ...new function.
10654 (pass_ipa_tree_profile): Replace declaration with that of...
10655 (make_pass_ipa_tree_profile): ...new function.
10656 (pass_early_local_passes): Replace declaration with that of...
10657 (make_pass_early_local_passes): ...new function.
10658 (pass_ipa_whole_program_visibility): Replace declaration with that
10659 of...
10660 (make_pass_ipa_whole_program_visibility): ...new function.
10661 (pass_ipa_lto_gimple_out): Replace declaration with that of...
10662 (make_pass_ipa_lto_gimple_out): ...new function.
10663 (pass_ipa_increase_alignment): Replace declaration with that of...
10664 (make_pass_ipa_increase_alignment): ...new function.
10665 (pass_ipa_inline): Replace declaration with that of...
10666 (make_pass_ipa_inline): ...new function.
10667 (pass_ipa_free_lang_data): Replace declaration with that of...
10668 (make_pass_ipa_free_lang_data): ...new function.
10669 (pass_ipa_free_inline_summary): Replace declaration with that of...
10670 (make_pass_ipa_free_inline_summary): ...new function.
10671 (pass_ipa_cp): Replace declaration with that of...
10672 (make_pass_ipa_cp): ...new function.
10673 (pass_ipa_reference): Replace declaration with that of...
10674 (make_pass_ipa_reference): ...new function.
10675 (pass_ipa_pure_const): Replace declaration with that of...
10676 (make_pass_ipa_pure_const): ...new function.
10677 (pass_ipa_pta): Replace declaration with that of...
10678 (make_pass_ipa_pta): ...new function.
10679 (pass_ipa_lto_finish_out): Replace declaration with that of...
10680 (make_pass_ipa_lto_finish_out): ...new function.
10681 (pass_ipa_tm): Replace declaration with that of...
10682 (make_pass_ipa_tm): ...new function.
10683 (pass_ipa_profile): Replace declaration with that of...
10684 (make_pass_ipa_profile): ...new function.
10685 (pass_ipa_cdtor_merge): Replace declaration with that of...
10686 (make_pass_ipa_cdtor_merge): ...new function.
10687 (pass_cleanup_cfg_post_optimizing): Replace declaration with that
10688 of...
10689 (make_pass_cleanup_cfg_post_optimizing): ...new function.
10690 (pass_init_datastructures): Replace declaration with that of...
10691 (make_pass_init_datastructures): ...new function.
10692 (pass_fixup_cfg): Replace declaration with that of...
10693 (make_pass_fixup_cfg): ...new function.
10694 (pass_expand): Replace declaration with that of...
10695 (make_pass_expand): ...new function.
10696 (pass_instantiate_virtual_regs): Replace declaration with that of...
10697 (make_pass_instantiate_virtual_regs): ...new function.
10698 (pass_rtl_fwprop): Replace declaration with that of...
10699 (make_pass_rtl_fwprop): ...new function.
10700 (pass_rtl_fwprop_addr): Replace declaration with that of...
10701 (make_pass_rtl_fwprop_addr): ...new function.
10702 (pass_jump): Replace declaration with that of...
10703 (make_pass_jump): ...new function.
10704 (pass_jump2): Replace declaration with that of...
10705 (make_pass_jump2): ...new function.
10706 (pass_lower_subreg): Replace declaration with that of...
10707 (make_pass_lower_subreg): ...new function.
10708 (pass_cse): Replace declaration with that of...
10709 (make_pass_cse): ...new function.
10710 (pass_fast_rtl_dce): Replace declaration with that of...
10711 (make_pass_fast_rtl_dce): ...new function.
10712 (pass_ud_rtl_dce): Replace declaration with that of...
10713 (make_pass_ud_rtl_dce): ...new function.
10714 (pass_rtl_dce): Replace declaration with that of...
10715 (make_pass_rtl_dce): ...new function.
10716 (pass_rtl_dse1): Replace declaration with that of...
10717 (make_pass_rtl_dse1): ...new function.
10718 (pass_rtl_dse2): Replace declaration with that of...
10719 (make_pass_rtl_dse2): ...new function.
10720 (pass_rtl_dse3): Replace declaration with that of...
10721 (make_pass_rtl_dse3): ...new function.
10722 (pass_rtl_cprop): Replace declaration with that of...
10723 (make_pass_rtl_cprop): ...new function.
10724 (pass_rtl_pre): Replace declaration with that of...
10725 (make_pass_rtl_pre): ...new function.
10726 (pass_rtl_hoist): Replace declaration with that of...
10727 (make_pass_rtl_hoist): ...new function.
10728 (pass_rtl_store_motion): Replace declaration with that of...
10729 (make_pass_rtl_store_motion): ...new function.
10730 (pass_cse_after_global_opts): Replace declaration with that of...
10731 (make_pass_cse_after_global_opts): ...new function.
10732 (pass_rtl_ifcvt): Replace declaration with that of...
10733 (make_pass_rtl_ifcvt): ...new function.
10734 (pass_into_cfg_layout_mode): Replace declaration with that of...
10735 (make_pass_into_cfg_layout_mode): ...new function.
10736 (pass_outof_cfg_layout_mode): Replace declaration with that of...
10737 (make_pass_outof_cfg_layout_mode): ...new function.
10738 (pass_loop2): Replace declaration with that of...
10739 (make_pass_loop2): ...new function.
10740 (pass_rtl_loop_init): Replace declaration with that of...
10741 (make_pass_rtl_loop_init): ...new function.
10742 (pass_rtl_move_loop_invariants): Replace declaration with that of...
10743 (make_pass_rtl_move_loop_invariants): ...new function.
10744 (pass_rtl_unswitch): Replace declaration with that of...
10745 (make_pass_rtl_unswitch): ...new function.
10746 (pass_rtl_unroll_and_peel_loops): Replace declaration with that of...
10747 (make_pass_rtl_unroll_and_peel_loops): ...new function.
10748 (pass_rtl_doloop): Replace declaration with that of...
10749 (make_pass_rtl_doloop): ...new function.
10750 (pass_rtl_loop_done): Replace declaration with that of...
10751 (make_pass_rtl_loop_done): ...new function.
10752 (pass_web): Replace declaration with that of...
10753 (make_pass_web): ...new function.
10754 (pass_cse2): Replace declaration with that of...
10755 (make_pass_cse2): ...new function.
10756 (pass_df_initialize_opt): Replace declaration with that of...
10757 (make_pass_df_initialize_opt): ...new function.
10758 (pass_df_initialize_no_opt): Replace declaration with that of...
10759 (make_pass_df_initialize_no_opt): ...new function.
10760 (pass_reginfo_init): Replace declaration with that of...
10761 (make_pass_reginfo_init): ...new function.
10762 (pass_inc_dec): Replace declaration with that of...
10763 (make_pass_inc_dec): ...new function.
10764 (pass_stack_ptr_mod): Replace declaration with that of...
10765 (make_pass_stack_ptr_mod): ...new function.
10766 (pass_initialize_regs): Replace declaration with that of...
10767 (make_pass_initialize_regs): ...new function.
10768 (pass_combine): Replace declaration with that of...
10769 (make_pass_combine): ...new function.
10770 (pass_if_after_combine): Replace declaration with that of...
10771 (make_pass_if_after_combine): ...new function.
10772 (pass_ree): Replace declaration with that of...
10773 (make_pass_ree): ...new function.
10774 (pass_partition_blocks): Replace declaration with that of...
10775 (make_pass_partition_blocks): ...new function.
10776 (pass_match_asm_constraints): Replace declaration with that of...
10777 (make_pass_match_asm_constraints): ...new function.
10778 (pass_regmove): Replace declaration with that of...
10779 (make_pass_regmove): ...new function.
10780 (pass_split_all_insns): Replace declaration with that of...
10781 (make_pass_split_all_insns): ...new function.
10782 (pass_fast_rtl_byte_dce): Replace declaration with that of...
10783 (make_pass_fast_rtl_byte_dce): ...new function.
10784 (pass_lower_subreg2): Replace declaration with that of...
10785 (make_pass_lower_subreg2): ...new function.
10786 (pass_mode_switching): Replace declaration with that of...
10787 (make_pass_mode_switching): ...new function.
10788 (pass_sms): Replace declaration with that of...
10789 (make_pass_sms): ...new function.
10790 (pass_sched): Replace declaration with that of...
10791 (make_pass_sched): ...new function.
10792 (pass_ira): Replace declaration with that of...
10793 (make_pass_ira): ...new function.
10794 (pass_reload): Replace declaration with that of...
10795 (make_pass_reload): ...new function.
10796 (pass_clean_state): Replace declaration with that of...
10797 (make_pass_clean_state): ...new function.
10798 (pass_branch_prob): Replace declaration with that of...
10799 (make_pass_branch_prob): ...new function.
10800 (pass_value_profile_transformations): Replace declaration with that
10801 of...
10802 (make_pass_value_profile_transformations): ...new function.
10803 (pass_postreload_cse): Replace declaration with that of...
10804 (make_pass_postreload_cse): ...new function.
10805 (pass_gcse2): Replace declaration with that of...
10806 (make_pass_gcse2): ...new function.
10807 (pass_split_after_reload): Replace declaration with that of...
10808 (make_pass_split_after_reload): ...new function.
10809 (pass_branch_target_load_optimize1): Replace declaration with that
10810 of...
10811 (make_pass_branch_target_load_optimize1): ...new function.
10812 (pass_thread_prologue_and_epilogue): Replace declaration with that
10813 of...
10814 (make_pass_thread_prologue_and_epilogue): ...new function.
10815 (pass_stack_adjustments): Replace declaration with that of...
10816 (make_pass_stack_adjustments): ...new function.
10817 (pass_peephole2): Replace declaration with that of...
10818 (make_pass_peephole2): ...new function.
10819 (pass_if_after_reload): Replace declaration with that of...
10820 (make_pass_if_after_reload): ...new function.
10821 (pass_regrename): Replace declaration with that of...
10822 (make_pass_regrename): ...new function.
10823 (pass_cprop_hardreg): Replace declaration with that of...
10824 (make_pass_cprop_hardreg): ...new function.
10825 (pass_reorder_blocks): Replace declaration with that of...
10826 (make_pass_reorder_blocks): ...new function.
10827 (pass_branch_target_load_optimize2): Replace declaration with that
10828 of...
10829 (make_pass_branch_target_load_optimize2): ...new function.
10830 (pass_leaf_regs): Replace declaration with that of...
10831 (make_pass_leaf_regs): ...new function.
10832 (pass_split_before_sched2): Replace declaration with that of...
10833 (make_pass_split_before_sched2): ...new function.
10834 (pass_compare_elim_after_reload): Replace declaration with that of...
10835 (make_pass_compare_elim_after_reload): ...new function.
10836 (pass_sched2): Replace declaration with that of...
10837 (make_pass_sched2): ...new function.
10838 (pass_stack_regs): Replace declaration with that of...
10839 (make_pass_stack_regs): ...new function.
10840 (pass_stack_regs_run): Replace declaration with that of...
10841 (make_pass_stack_regs_run): ...new function.
10842 (pass_df_finish): Replace declaration with that of...
10843 (make_pass_df_finish): ...new function.
10844 (pass_compute_alignments): Replace declaration with that of...
10845 (make_pass_compute_alignments): ...new function.
10846 (pass_duplicate_computed_gotos): Replace declaration with that of...
10847 (make_pass_duplicate_computed_gotos): ...new function.
10848 (pass_variable_tracking): Replace declaration with that of...
10849 (make_pass_variable_tracking): ...new function.
10850 (pass_free_cfg): Replace declaration with that of...
10851 (make_pass_free_cfg): ...new function.
10852 (pass_machine_reorg): Replace declaration with that of...
10853 (make_pass_machine_reorg): ...new function.
10854 (pass_cleanup_barriers): Replace declaration with that of...
10855 (make_pass_cleanup_barriers): ...new function.
10856 (pass_delay_slots): Replace declaration with that of...
10857 (make_pass_delay_slots): ...new function.
10858 (pass_split_for_shorten_branches): Replace declaration with that of...
10859 (make_pass_split_for_shorten_branches): ...new function.
10860 (pass_split_before_regstack): Replace declaration with that of...
10861 (make_pass_split_before_regstack): ...new function.
10862 (pass_convert_to_eh_region_ranges): Replace declaration with that
10863 of...
10864 (make_pass_convert_to_eh_region_ranges): ...new function.
10865 (pass_shorten_branches): Replace declaration with that of...
10866 (make_pass_shorten_branches): ...new function.
10867 (pass_set_nothrow_function_flags): Replace declaration with that of...
10868 (make_pass_set_nothrow_function_flags): ...new function.
10869 (pass_dwarf2_frame): Replace declaration with that of...
10870 (make_pass_dwarf2_frame): ...new function.
10871 (pass_final): Replace declaration with that of...
10872 (make_pass_final): ...new function.
10873 (pass_rtl_seqabstr): Replace declaration with that of...
10874 (make_pass_rtl_seqabstr): ...new function.
10875 (pass_release_ssa_names): Replace declaration with that of...
10876 (make_pass_release_ssa_names): ...new function.
10877 (pass_early_inline): Replace declaration with that of...
10878 (make_pass_early_inline): ...new function.
10879 (pass_inline_parameters): Replace declaration with that of...
10880 (make_pass_inline_parameters): ...new function.
10881 (pass_update_address_taken): Replace declaration with that of...
10882 (make_pass_update_address_taken): ...new function.
10883 (pass_convert_switch): Replace declaration with that of...
10884 (make_pass_convert_switch): ...new function.
10885 * tree-profile.c (pass_ipa_tree_profile): Convert from a global struct
10886 to a subclass of simple_ipa_opt_pass along with...
10887 (pass_data_ipa_tree_profile): ...new pass_data instance and...
10888 (make_pass_ipa_tree_profile): ...new function.
10889 * tree-sra.c (pass_sra_early): Convert from a global struct to a
10890 subclass of gimple_opt_pass along with...
10891 (pass_data_sra_early): ...new pass_data instance and...
10892 (make_pass_sra_early): ...new function.
10893 (pass_sra): Convert from a global struct to a subclass of
10894 gimple_opt_pass along with...
10895 (pass_data_sra): ...new pass_data instance and...
10896 (make_pass_sra): ...new function.
10897 (pass_early_ipa_sra): Convert from a global struct to a subclass of
10898 gimple_opt_pass along with...
10899 (pass_data_early_ipa_sra): ...new pass_data instance and...
10900 (make_pass_early_ipa_sra): ...new function.
10901 * tree-ssa-ccp.c (pass_ccp): Convert from a global struct to a
10902 subclass of gimple_opt_pass along with...
10903 (pass_data_ccp): ...new pass_data instance and...
10904 (make_pass_ccp): ...new function.
10905 (pass_fold_builtins): Convert from a global struct to a subclass of
10906 gimple_opt_pass along with...
10907 (pass_data_fold_builtins): ...new pass_data instance and...
10908 (make_pass_fold_builtins): ...new function.
10909 * tree-ssa-copy.c (pass_copy_prop): Convert from a global struct to a
10910 subclass of gimple_opt_pass along with...
10911 (pass_data_copy_prop): ...new pass_data instance and...
10912 (make_pass_copy_prop): ...new function.
10913 * tree-ssa-copyrename.c (pass_rename_ssa_copies): Convert from a
10914 global struct to a subclass of gimple_opt_pass along with...
10915 (pass_data_rename_ssa_copies): ...new pass_data instance and...
10916 (make_pass_rename_ssa_copies): ...new function.
10917 * tree-ssa-dce.c (pass_dce): Convert from a global struct to a
10918 subclass of gimple_opt_pass along with...
10919 (pass_data_dce): ...new pass_data instance and...
10920 (make_pass_dce): ...new function.
10921 (pass_dce_loop): Convert from a global struct to a subclass of
10922 gimple_opt_pass along with...
10923 (pass_data_dce_loop): ...new pass_data instance and...
10924 (make_pass_dce_loop): ...new function.
10925 (pass_cd_dce): Convert from a global struct to a subclass of
10926 gimple_opt_pass along with...
10927 (pass_data_cd_dce): ...new pass_data instance and...
10928 (make_pass_cd_dce): ...new function.
10929 * tree-ssa-dom.c (pass_dominator): Convert from a global struct to a
10930 subclass of gimple_opt_pass along with...
10931 (pass_data_dominator): ...new pass_data instance and...
10932 (make_pass_dominator): ...new function.
10933 (pass_phi_only_cprop): Convert from a global struct to a subclass of
10934 gimple_opt_pass along with...
10935 (pass_data_phi_only_cprop): ...new pass_data instance and...
10936 (make_pass_phi_only_cprop): ...new function.
10937 * tree-ssa-dse.c (pass_dse): Convert from a global struct to a
10938 subclass of gimple_opt_pass along with...
10939 (pass_data_dse): ...new pass_data instance and...
10940 (make_pass_dse): ...new function.
10941 * tree-ssa-forwprop.c (pass_forwprop): Convert from a global struct to
10942 a subclass of gimple_opt_pass along with...
10943 (pass_data_forwprop): ...new pass_data instance and...
10944 (make_pass_forwprop): ...new function.
10945 * tree-ssa-ifcombine.c (pass_tree_ifcombine): Convert from a global
10946 struct to a subclass of gimple_opt_pass along with...
10947 (pass_data_tree_ifcombine): ...new pass_data instance and...
10948 (make_pass_tree_ifcombine): ...new function.
10949 * tree-ssa-loop-ch.c (pass_ch): Convert from a global struct to a
10950 subclass of gimple_opt_pass along with...
10951 (pass_data_ch): ...new pass_data instance and...
10952 (make_pass_ch): ...new function.
10953 * tree-ssa-loop.c (pass_tree_loop): Convert from a global struct to a
10954 subclass of gimple_opt_pass along with...
10955 (pass_data_tree_loop): ...new pass_data instance and...
10956 (make_pass_tree_loop): ...new function.
10957 (pass_tree_loop_init): Convert from a global struct to a subclass of
10958 gimple_opt_pass along with...
10959 (pass_data_tree_loop_init): ...new pass_data instance and...
10960 (make_pass_tree_loop_init): ...new function.
10961 (pass_lim): Convert from a global struct to a subclass of
10962 gimple_opt_pass along with...
10963 (pass_data_lim): ...new pass_data instance and...
10964 (make_pass_lim): ...new function.
10965 (pass_tree_unswitch): Convert from a global struct to a subclass of
10966 gimple_opt_pass along with...
10967 (pass_data_tree_unswitch): ...new pass_data instance and...
10968 (make_pass_tree_unswitch): ...new function.
10969 (pass_predcom): Convert from a global struct to a subclass of
10970 gimple_opt_pass along with...
10971 (pass_data_predcom): ...new pass_data instance and...
10972 (make_pass_predcom): ...new function.
10973 (pass_vectorize): Convert from a global struct to a subclass of
10974 gimple_opt_pass along with...
10975 (pass_data_vectorize): ...new pass_data instance and...
10976 (make_pass_vectorize): ...new function.
10977 (pass_graphite): Convert from a global struct to a subclass of
10978 gimple_opt_pass along with...
10979 (pass_data_graphite): ...new pass_data instance and...
10980 (make_pass_graphite): ...new function.
10981 (pass_graphite_transforms): Convert from a global struct to a subclass
10982 of gimple_opt_pass along with...
10983 (pass_data_graphite_transforms): ...new pass_data instance and...
10984 (make_pass_graphite_transforms): ...new function.
10985 (pass_check_data_deps): Convert from a global struct to a subclass of
10986 gimple_opt_pass along with...
10987 (pass_data_check_data_deps): ...new pass_data instance and...
10988 (make_pass_check_data_deps): ...new function.
10989 (pass_iv_canon): Convert from a global struct to a subclass of
10990 gimple_opt_pass along with...
10991 (pass_data_iv_canon): ...new pass_data instance and...
10992 (make_pass_iv_canon): ...new function.
10993 (pass_scev_cprop): Convert from a global struct to a subclass of
10994 gimple_opt_pass along with...
10995 (pass_data_scev_cprop): ...new pass_data instance and...
10996 (make_pass_scev_cprop): ...new function.
10997 (pass_record_bounds): Convert from a global struct to a subclass of
10998 gimple_opt_pass along with...
10999 (pass_data_record_bounds): ...new pass_data instance and...
11000 (make_pass_record_bounds): ...new function.
11001 (pass_complete_unroll): Convert from a global struct to a subclass of
11002 gimple_opt_pass along with...
11003 (pass_data_complete_unroll): ...new pass_data instance and...
11004 (make_pass_complete_unroll): ...new function.
11005 (pass_complete_unrolli): Convert from a global struct to a subclass of
11006 gimple_opt_pass along with...
11007 (pass_data_complete_unrolli): ...new pass_data instance and...
11008 (make_pass_complete_unrolli): ...new function.
11009 (pass_parallelize_loops): Convert from a global struct to a subclass
11010 of gimple_opt_pass along with...
11011 (pass_data_parallelize_loops): ...new pass_data instance and...
11012 (make_pass_parallelize_loops): ...new function.
11013 (pass_loop_prefetch): Convert from a global struct to a subclass of
11014 gimple_opt_pass along with...
11015 (pass_data_loop_prefetch): ...new pass_data instance and...
11016 (make_pass_loop_prefetch): ...new function.
11017 (pass_iv_optimize): Convert from a global struct to a subclass of
11018 gimple_opt_pass along with...
11019 (pass_data_iv_optimize): ...new pass_data instance and...
11020 (make_pass_iv_optimize): ...new function.
11021 (pass_tree_loop_done): Convert from a global struct to a subclass of
11022 gimple_opt_pass along with...
11023 (pass_data_tree_loop_done): ...new pass_data instance and...
11024 (make_pass_tree_loop_done): ...new function.
11025 * tree-ssa-math-opts.c (pass_cse_reciprocals): Convert from a global
11026 struct to a subclass of gimple_opt_pass along with...
11027 (pass_data_cse_reciprocals): ...new pass_data instance and...
11028 (make_pass_cse_reciprocals): ...new function.
11029 (pass_cse_sincos): Convert from a global struct to a subclass of
11030 gimple_opt_pass along with...
11031 (pass_data_cse_sincos): ...new pass_data instance and...
11032 (make_pass_cse_sincos): ...new function.
11033 (pass_optimize_bswap): Convert from a global struct to a subclass of
11034 gimple_opt_pass along with...
11035 (pass_data_optimize_bswap): ...new pass_data instance and...
11036 (make_pass_optimize_bswap): ...new function.
11037 (pass_optimize_widening_mul): Convert from a global struct to a
11038 subclass of gimple_opt_pass along with...
11039 (pass_data_optimize_widening_mul): ...new pass_data instance and...
11040 (make_pass_optimize_widening_mul): ...new function.
11041 * tree-ssa-phiopt.c (pass_phiopt): Convert from a global struct to a
11042 subclass of gimple_opt_pass along with...
11043 (pass_data_phiopt): ...new pass_data instance and...
11044 (make_pass_phiopt): ...new function.
11045 (pass_cselim): Convert from a global struct to a subclass of
11046 gimple_opt_pass along with...
11047 (pass_data_cselim): ...new pass_data instance and...
11048 (make_pass_cselim): ...new function.
11049 * tree-ssa-phiprop.c (pass_phiprop): Convert from a global struct to a
11050 subclass of gimple_opt_pass along with...
11051 (pass_data_phiprop): ...new pass_data instance and...
11052 (make_pass_phiprop): ...new function.
11053 * tree-ssa-pre.c (pass_pre): Convert from a global struct to a
11054 subclass of gimple_opt_pass along with...
11055 (pass_data_pre): ...new pass_data instance and...
11056 (make_pass_pre): ...new function.
11057 (pass_fre): Convert from a global struct to a subclass of
11058 gimple_opt_pass along with...
11059 (pass_data_fre): ...new pass_data instance and...
11060 (make_pass_fre): ...new function.
11061 * tree-ssa-reassoc.c (pass_reassoc): Convert from a global struct to a
11062 subclass of gimple_opt_pass along with...
11063 (pass_data_reassoc): ...new pass_data instance and...
11064 (make_pass_reassoc): ...new function.
11065 * tree-ssa-sink.c (pass_sink_code): Convert from a global struct to a
11066 subclass of gimple_opt_pass along with...
11067 (pass_data_sink_code): ...new pass_data instance and...
11068 (make_pass_sink_code): ...new function.
11069 * tree-ssa-strlen.c (pass_strlen): Convert from a global struct to a
11070 subclass of gimple_opt_pass along with...
11071 (pass_data_strlen): ...new pass_data instance and...
11072 (make_pass_strlen): ...new function.
11073 * tree-ssa-structalias.c (pass_build_alias): Convert from a global
11074 struct to a subclass of gimple_opt_pass along with...
11075 (pass_data_build_alias): ...new pass_data instance and...
11076 (make_pass_build_alias): ...new function.
11077 (pass_build_ealias): Convert from a global struct to a subclass of
11078 gimple_opt_pass along with...
11079 (pass_data_build_ealias): ...new pass_data instance and...
11080 (make_pass_build_ealias): ...new function.
11081 (pass_ipa_pta): Convert from a global struct to a subclass of
11082 simple_ipa_opt_pass along with...
11083 (pass_data_ipa_pta): ...new pass_data instance and...
11084 (make_pass_ipa_pta): ...new function.
11085 * tree-ssa-uncprop.c (pass_uncprop): Convert from a global struct to a
11086 subclass of gimple_opt_pass along with...
11087 (pass_data_uncprop): ...new pass_data instance and...
11088 (make_pass_uncprop): ...new function.
11089 * tree-ssa-uninit.c (pass_late_warn_uninitialized): Convert from a
11090 global struct to a subclass of gimple_opt_pass along with...
11091 (pass_data_late_warn_uninitialized): ...new pass_data instance and...
11092 (make_pass_late_warn_uninitialized): ...new function.
11093 * tree-ssa.c (pass_init_datastructures): Convert from a global struct
11094 to a subclass of gimple_opt_pass along with...
11095 (pass_data_init_datastructures): ...new pass_data instance and...
11096 (make_pass_init_datastructures): ...new function.
11097 (pass_early_warn_uninitialized): Convert from a global struct to a
11098 subclass of gimple_opt_pass along with...
11099 (pass_data_early_warn_uninitialized): ...new pass_data instance and...
11100 (make_pass_early_warn_uninitialized): ...new function.
11101 (pass_update_address_taken): Convert from a global struct to a
11102 subclass of gimple_opt_pass along with...
11103 (pass_data_update_address_taken): ...new pass_data instance and...
11104 (make_pass_update_address_taken): ...new function.
11105 * tree-ssanames.c (pass_release_ssa_names): Convert from a global
11106 struct to a subclass of gimple_opt_pass along with...
11107 (pass_data_release_ssa_names): ...new pass_data instance and...
11108 (make_pass_release_ssa_names): ...new function.
11109 * tree-stdarg.c (pass_stdarg): Convert from a global struct to a
11110 subclass of gimple_opt_pass along with...
11111 (pass_data_stdarg): ...new pass_data instance and...
11112 (make_pass_stdarg): ...new function.
11113 * tree-switch-conversion.c (pass_convert_switch): Convert from a
11114 global struct to a subclass of gimple_opt_pass along with...
11115 (pass_data_convert_switch): ...new pass_data instance and...
11116 (make_pass_convert_switch): ...new function.
11117 * tree-tailcall.c (pass_tail_recursion): Convert from a global struct
11118 to a subclass of gimple_opt_pass along with...
11119 (pass_data_tail_recursion): ...new pass_data instance and...
11120 (make_pass_tail_recursion): ...new function.
11121 (pass_tail_calls): Convert from a global struct to a subclass of
11122 gimple_opt_pass along with...
11123 (pass_data_tail_calls): ...new pass_data instance and...
11124 (make_pass_tail_calls): ...new function.
11125 * tree-vect-generic.c (pass_lower_vector): Convert from a global
11126 struct to a subclass of gimple_opt_pass along with...
11127 (pass_data_lower_vector): ...new pass_data instance and...
11128 (make_pass_lower_vector): ...new function.
11129 (pass_lower_vector_ssa): Convert from a global struct to a subclass of
11130 gimple_opt_pass along with...
11131 (pass_data_lower_vector_ssa): ...new pass_data instance and...
11132 (make_pass_lower_vector_ssa): ...new function.
11133 * tree-vectorizer.c (pass_slp_vectorize): Convert from a global struct
11134 to a subclass of gimple_opt_pass along with...
11135 (pass_data_slp_vectorize): ...new pass_data instance and...
11136 (make_pass_slp_vectorize): ...new function.
11137 (pass_ipa_increase_alignment): Convert from a global struct to a
11138 subclass of simple_ipa_opt_pass along with...
11139 (pass_data_ipa_increase_alignment): ...new pass_data instance and...
11140 (make_pass_ipa_increase_alignment): ...new function.
11141 * tree-vrp.c (pass_vrp): Convert from a global struct to a subclass of
11142 gimple_opt_pass along with...
11143 (pass_data_vrp): ...new pass_data instance and...
11144 (make_pass_vrp): ...new function.
11145 * tree.c (pass_ipa_free_lang_data): Convert from a global struct to a
11146 subclass of simple_ipa_opt_pass along with...
11147 (pass_data_ipa_free_lang_data): ...new pass_data instance and...
11148 (make_pass_ipa_free_lang_data): ...new function.
11149 * tsan.c (pass_tsan): Convert from a global struct to a subclass of
11150 gimple_opt_pass along with...
11151 (pass_data_tsan): ...new pass_data instance and...
11152 (make_pass_tsan): ...new function.
11153 (pass_tsan_O0): Convert from a global struct to a subclass of
11154 gimple_opt_pass along with...
11155 (pass_data_tsan_O0): ...new pass_data instance and...
11156 (make_pass_tsan_O0): ...new function.
11157 * var-tracking.c (pass_variable_tracking): Convert from a global
11158 struct to a subclass of rtl_opt_pass along with...
11159 (pass_data_variable_tracking): ...new pass_data instance and...
11160 (make_pass_variable_tracking): ...new function.
11161 * web.c (pass_web): Convert from a global struct to a subclass of
11162 rtl_opt_pass along with...
11163 (pass_data_web): ...new pass_data instance and...
11164 (make_pass_web): ...new function.
11165 * config/epiphany/epiphany.h (pass_mode_switch_use): Replace
11166 declaration with that of...
11167 (make_pass_mode_switch_use): ...new function.
11168 (pass_resolve_sw_modes): Replace declaration with that of...
11169 (make_pass_resolve_sw_modes): ...new function.
11170 * config/epiphany/mode-switch-use.c (pass_mode_switch_use): Convert
11171 from a global struct to a subclass of rtl_opt_pass along with...
11172 (pass_data_mode_switch_use): ...new pass_data instance and...
11173 (make_pass_mode_switch_use): ...new function.
11174 * config/epiphany/resolve-sw-modes.c (pass_resolve_sw_modes): Convert
11175 from a global struct to a subclass of rtl_opt_pass along with...
11176 (pass_data_resolve_sw_modes): ...new pass_data instance and...
11177 (make_pass_resolve_sw_modes): ...new function.
11178 * config/i386/i386.c (pass_insert_vzeroupper): Convert from a global
11179 struct to a subclass of rtl_opt_pass along with...
11180 (pass_data_insert_vzeroupper): ...new pass_data instance and...
11181 (make_pass_insert_vzeroupper): ...new function.
11182 * config/sparc/sparc.c (pass_work_around_errata): Convert from a
11183 global struct to a subclass of rtl_opt_pass along with...
11184 (pass_data_work_around_errata): ...new pass_data instance and...
11185 (make_pass_work_around_errata): ...new function.
11186 * config/mips/mips.c (pass_mips_machine_reorg2): Convert from a global
11187 struct to a subclass of rtl_opt_pass along with...
11188 (pass_data_mips_machine_reorg2): ...new pass_data instance and...
11189 (make_pass_mips_machine_reorg2): ...new function.
11190
6a389ed5
DM
111912013-08-05 David Malcolm <dmalcolm@redhat.com>
11192
11193 * passes.c (pass_manager::operator new): New.
11194
f7695dbf
DM
111952013-08-05 David Malcolm <dmalcolm@redhat.com>
11196
11197 Handwritten part of conversion of passes to C++ classes.
11198
11199 * Makefile.in (PASS_MANAGER_H): Add dep on pass-instances.def.
11200 (toplev.o): Add dep on PASS_MANAGER_H.
11201 * cgraphunit.c (cgraph_process_new_functions): Rework invocation
11202 of early local pases to reflect this moving from a global to a
11203 member of gcc::pass_manager.
11204 (cgraph_add_new_function): Likewise.
11205 * lto-cgraph.c (lto_output_node): Update for conversion of
11206 struct ipa_opt_pass_d to a C++ subclass of opt_pass.
11207 * passes.c (opt_pass::clone): New.
11208 (opt_pass::gate): New.
11209 (opt_pass::execute): New.
11210 (opt_pass::opt_pass): New.
11211 (pass_manager::execute_early_local_passes): New.
11212 (pass_manager::execute_pass_mode_switching): new.
11213 (finish_optimization_passes): Convert to...
11214 (pass_manager::finish_optimization_passes): ...this.
11215 (finish_optimization_passes): Update for conversion of passes to
11216 C++ classes.
11217 (register_dump_files_1): Use has_gate since we cannot portably
11218 check a vtable entry against NULL.
11219 (dump_one_pass): Likewise.
11220 (ipa_write_summaries_2): Likewise.
11221 (ipa_write_optimization_summaries_1): Likewise.
11222 (ipa_read_summaries_1): Likewise.
11223 (ipa_read_optimization_summaries_1): Likewise.
11224 (execute_ipa_stmt_fixups): Likewise.
11225 (pass_manager::pass_manager): Rewrite pass-creation, invoking
11226 pass-creation functions rather than wiring up globals, and
11227 storing the results in fields of pass_manager generated using
11228 pass-instances.def.
11229 (pass_manager::dump_profile_report): Update for conversion of
11230 passes to C++ classes.
11231 (pass_manager::execute_ipa_summary_passes): Likewise.
11232 (execute_one_ipa_transform_pass): Likewise.
11233 (execute_one_pass): Use has_gate and has_execute since we cannot
11234 portably check a vtable entry against NULL.
11235 * pass_manager.h (pass_manager::finish_optimization_passes): New.
11236 (pass_manager): Use pass-instances.def to add fields for the
11237 various pass instances.
11238 * toplev.c (finalize): Update for move of
11239 finish_optimization_passes to a method of gcc::pass_manager.
11240 * toplev.h (finish_optimization_passes): Move to method of class
11241 pass_manager.
11242 * tree-pass.h (struct pass_data): New.
03b0ee0a 11243 (opt_pass): Convert to C++ class, make it a subclass of pass_data.
f7695dbf
DM
11244 (opt_pass::gate): Convert to virtual function.
11245 (opt_pass::~opt_pass): New.
11246 (opt_pass::clone): New.
11247 (opt_pass::execute): Convert to virtual function.
11248 (opt_pass::opt_pass): New.
11249 (opt_pass::ctxt_): new.
11250 (gimple_opt_pass): Convert to subclass of opt_pass.
11251 (gimple_opt_pass::gimple_opt_pass): New.
11252 (rtl_opt_pass): Convert to subclass of opt_pass.
11253 (rtl_opt_pass::rtl_opt_pass): New.
11254 (ipa_opt_pass_d): Convert to subclass of opt_pass.
11255 (ipa_opt_pass_d::ipa_opt_pass_d): New.
11256 (simple_ipa_opt_pass): Convert to subclass of opt_pass.
11257 (simple_ipa_opt_pass::simple_ipa_opt_pass): New.
11258 * config/i386/i386.c (rest_of_handle_insert_vzeroupper): Rework
11259 invocation of pass_mode_switching to reflect this moving from a
11260 global to a member of gcc::pass_manager.
11261 (ix86_option_override): Rework how pass_insert_vzeroupper is
11262 added to the pass_manager to reflect autogenerated changes.
03b0ee0a 11263 * config/i386/t-i386 (i386.o) Add deps on CONTEXT_H and PASS_MANAGER_H.
f7695dbf 11264
d0b2f831
RE
112652013-08-05 Richard Earnshaw <rearnsha@arm.com>
11266
11267 PR rtl-optimization/57708
11268 * recog.c (peep2_find_free_register): Validate all regs in a
11269 multi-reg mode.
11270
51a5c0c2
JH
112712013-08-05 Jan Hubicka <jh@suse.cz>
11272
11273 PR lto/57602
03b0ee0a
UB
11274 * cgraph.c (verify_cgraph_node): Accept local flags from other
11275 partitions.
51a5c0c2
JH
11276 * ipa.c (symtab_remove_unreachable_nodes): Do not clear local flag.
11277 (function_and_variable_visibility): Likewise.
11278 * trans-mem.c (ipa_tm_create_version): TM versions are not local.
11279
65f0a120
GDR
112802013-08-05 Gabriel Dos Reis <gdr@integrable-solutions.net>
11281
11282 * graph.c (init_graph_slim_pretty_print): Remove.
11283 (print_graph_cfg): Do not call it. Use local pretty printer.
11284 (start_graph_dump): Likewise.
11285
b3f80694
GDR
112862013-08-05 Gabriel Dos Reis <gdr@integrable-solutions.net>
11287
11288 * gimple-pretty-print.c (buffer): Remove.
11289 (initialized): Likewise.
11290 (maybe_init_pretty_print): Likewise.
11291 (print_gimple_stmt): Do not call it. Use non-static local
11292 pretty_printer variable.
11293 (print_gimple_expr): Likewise.
11294 (print_gimple_seq): Likewise.
11295 (gimple_dump_bb): Likewise.
11296
11a877b3
GDR
112972013-08-05 Gabriel Dos Reis <gdr@integrable-solutions.net>
11298
11299 * asan.c (asan_pp): Remove.
11300 (asan_pp_initialized): Likewise.
11301 (asan_pp_initialize): Likewise.
11302 (asan_pp_string): Take a pretty_printer parameter. Adjust callers.
11303 (asan_emit_stack_protection): Tidy. Use local pretty printer.
11304 (asan_add_global): Likewise.
11305
b066401f
GDR
113062013-08-04 Gabriel Dos Reis <gdr@integrable-solutions.net>
11307
11308 * pretty-print.h (pp_base): Remove. Adjust dependent macros.
11309 * diagnostic.h (diagnostic_flush_buffer): Adjust.
11310 * pretty-print.c (pp_formatted_text_data): Likewise.
11311 (pp_indent): Rename from pp_base_indent.
11312 (pp_format): Rename from pp_base_format.
11313 (pp_output_formatted_text): Rename from pp_base_output_formatted_text.
11314 (pp_format_verbatim): Rename from pp_base_format_verbatim.
11315 (pp_flush): Rename from pp_base_flush.
11316 (pp_set_line_maximum_length): Rename from
11317 pp_base_set_line_maximum_length.
11318 (pp_clear_output_area): Rename from pp_base_clear_output_area.
11319 (pp_set_prefix): Rename from pp_base_set_prefix.
11320 (pp_destroy_prefix): Rename from pp_base_destroy_prefix.
11321 (pp_emit_prefix): Rename from pp_base_emit_prefix.
11322 (pp_append_text): Rename from pp_base_append_text.
11323 (pp_formatted_text): Rename from pp_base_formatted_text.
11324 (pp_last_position_in_text): Rename from pp_base_last_position_in_text.
11325 (pp_remaining_character_count_for_line): Rename from
11326 pp_base_remaining_character_count_for_line.
11327 (pp_newline): Rename from pp_base_newline.
11328 (pp_character): Rename from pp_base_character.
11329 (pp_string): Rename from pp_base_string.
11330 (pp_maybe_space): Rename from pp_base_maybe_space.
11331 * asan.c (asan_pp_string): Adjust.
11332 (asan_emit_stack_protection): Likewise.
11333 (asan_add_global): Likewise.
11334 * sched-vis.c (str_pattern_slim): Adjust pretty printer function call.
11335 * tree-mudflap.c (mf_varname_tree): Likewise.
11336 * tree-pretty-print.c (pp_tree_identifier): Rename from
11337 pp_base_tree_identifier.
11338 * tree-pretty-print.h (pp_tree_identifier): Remove macro definition.
11339 Declare as function.
11340
137a1a27
GDR
113412013-08-03 Gabriel Dos Reis <gdr@integrable-solutions.net>
11342
11343 * pretty-print.h (pp_bar_bar): New.
11344 (pp_ampersand_ampersand): Likewise.
11345 (pp_less_equal): Likewise.
11346 (pp_greater_equal): Likewise.
11347 * gimple-pretty-print.c (dump_ternary_rhs): Use specialized pretty
11348 printer functions instead of pp_string or operators and punctuators.
11349 (dump_gimple_call): Likewise.
11350 (dump_gimple_omp_for): Likewise.
11351 (dump_gimple_transaction): Likewise.
11352 (dump_gimple_phi): Likewise.
11353 (pp_gimple_stmt_1): Likewise.
11354 * sched-vis.c (print_insn): Likewise.
11355 * tree-mudflap.c (mf_varname_tree): Likewise.
11356 * tree-pretty-print.c (dump_block_node): Likewise.
11357 (dump_generic_node): Likewise.
11358
815effe1
JH
113592013-08-02 Jan Hubicka <jh@suse.cz>
11360
11361 * lto-cgraph.c (compute_ltrans_boundary): Add abstract origins into
11362 boundaries.
11363 * lto-streamer-out.c (tree_is_indexable): Results decls and
11364 parm decls are not indexable.
11365 (DFS_write_tree_body): Do not follow args and results.
11366 (hash_tree): Likewise.
11367 (output_functions): Rearrange so struct function is needed
11368 only when real body is output; be able to also ouptut abstract
11369 functions; output DECL_ARGUMENTS and DECL_RESULT.
11370 (lto_output): When not in WPA, ale store abstract functions.
11371 (write_symbol): Do not care about RESULT_DECL.
11372 (output_symbol_p): Handle correctly sbtract decls.
11373 * lto-streamer-in.c (input_function): Rearrange so struct
11374 function can be NULL at entry; allow streaming of
11375 functions w/o body; store DECL_ARGUMENTS and DECL_RESULT.
11376 * ipa.c (symtab_remove_unreachable_nodes): Silence confused
11377 sanity check during LTO.
11378 * tree-streamer-out.c (write_ts_decl_non_common_tree_pointers): Skip
11379 RESULT_DECl and DECL_ARGUMENTS.
11380 * tree-streamer-in.c (lto_input_ts_decl_non_common_tree_pointers):
11381 Likewise.
11382
07838b13
GDR
113832013-08-03 Gabriel Dos Reis <gdr@integrable-solutions.net>
11384
11385 * pretty-print.h (pp_underscore): New.
11386 (pp_comma): Tidy.
11387 * gimple-pretty-print.c (dump_unary_rhs): Use specialized pretty
11388 printer functions instead of pp_character.
11389 (dump_binary_rhs): Likewise.
11390 (dump_ternary_rhs): Likewise.
11391 (dump_gimple_call_args): Likewise.
11392 (pp_points_to_solution): Likewise.
11393 (dump_gimple_call): Likewise.
11394 (dump_gimple_switch): Likewise.
11395 (dump_gimple_cond): Likewise.
11396 (dump_gimple_bind): Likewise.
11397 (dump_gimple_try): Likewise.
11398 (dump_gimple_omp_for): Likewise.
11399 (dump_gimple_omp_continue): Likewise.
11400 (dump_gimple_omp_single): Likewise.
11401 (dump_gimple_omp_sections): Likewise.
11402 (dump_gimple_omp_block): Likewise.
11403 (dump_gimple_omp_critical): Likewise.
11404 (dump_gimple_transaction): Likewise.
11405 (dump_gimple_asm): Likewise.
11406 (dump_gimple_phi): Likewise.
11407 (dump_gimple_omp_parallel): Likewise.
11408 (dump_gimple_omp_task): Likewise.
11409 (dump_gimple_omp_atomic_load): Likewise.
11410 (dump_gimple_omp_atomic_store): Likewise.
11411 (dump_gimple_mem_ops): Likewise.
11412 (pp_gimple_stmt_1): Likewise.
11413 (pp_cfg_jump): Likewise.
11414 (dump_implicit_edges): Likewise.
11415 (gimple_dump_bb_for_graph): Likewise.
11416 * graph.c (draw_cfg_node): Likewise.
11417 * langhooks.c (lhd_print_error_function): Likewise.
11418 * sched-vis.c (print_exp): Likewise.
11419 (print_value): Likewise.
11420 (print_pattern): Likewise.
11421 (print_insn): Likewise.
11422 (rtl_dump_bb_for_graph): Likewise.
11423 * tree-pretty-print.c (dump_function_declaration): Likewise.
11424 (dump_array_domain): Likewise.
11425 (dump_omp_clause): Likewise.
11426 (dump_location): Likewise.
11427 (dump_generic_node): Likewise.
11428 (print_struct_decl): Likewise.
11429 * diagnostic.c (diagnostic_show_locus): Use pp_space.
11430
bb0d2039
BS
114312013-08-03 Bill Schmidt <wschmidt@vnet.linux.ibm.com>
11432
11433 * gimple-ssa-strength-reduction.c (replace_mult_candidate): Update
11434 candidate table when replacing a candidate statement.
11435 (replace_rhs_if_not_dup): Likewise.
11436 (replace_one_candidate): Likewise.
11437
bc0ec027 114382013-08-02 Jan Hubicka <jh@suse.cz>
cdaa808a 11439 Martin Liska <marxin.liska@gmail.com>
bc0ec027
JH
11440
11441 * cgraphunit.c (add_new_function): Fix logic when adding from
11442 late IPA pass.
11443 (assemble_thunk): Rename to ...
03b0ee0a
UB
11444 (expand_thunk); .. this one; export; get it working with
11445 general functions; make produced gimple valid.
bc0ec027
JH
11446 * cgraph.h (expand_thunk): Declare.
11447
0e8853ee
JH
114482013-08-02 Jan Hubicka <jh@suse.cz>
11449
03b0ee0a
UB
11450 * ipa-cp.c (gather_context_independent_values): Use
11451 ipa_get_param_move_cost.
11452 (get_replacement_map): Remove PARAM; move parameter folding
11453 into tree-inline.c
0e8853ee
JH
11454 (create_specialized_node): Update.
11455 * ipa-prop.c (ipa_populate_param_decls): Do not look for origins;
11456 assert that we have gimple body; update move_cost.
11457 (count_formal_params): Assert that we have gimple body.
11458 (ipa_dump_param): New function.
11459 (ipa_alloc_node_params): Break out from ...
11460 (ipa_initialize_node_params): ... here.
11461 (ipa_get_vector_of_formal_parms): ICE when used in WPA.
11462 (ipa_write_node_info): Stream move costs.
11463 (ipa_read_node_info): Read move costs.
11464 (ipa_update_after_lto_read): Do not recompute node params.
11465 * ipa-prop.h (ipa_param_descriptor): Add move_cost.
11466 (ipa_get_param): Check we are not in WPA.
11467 (ipa_get_param_move_cost): New.
11468 * tree-inline.c (tree_function_versioning): Fold replacement as needed.
11469 * ipa-inline-analysis.c (inline_node_duplication_hook): Expect only
11470 parm numbers to be present.
11471
ecee672b
VM
114722013-08-02 Vladimir Makarov <vmakarov@redhat.com>
11473
11474 PR rtl-optimization/58048
11475 * lra-constraints.c (process_alt_operands): Don't check asm
11476 operand on register.
11477
13ec0527
EB
114782013-08-02 Eric Botcazou <ebotcazou@adacore.com>
11479
11480 * config/sparc/sparc.c (sparc_emit_membar_for_model) <SMM_TSO>: Add
11481 the implied StoreLoad barrier for atomic operations if before.
11482
af15184a
JH
114832013-08-02 Jan Hubicka <jh@suse.cz>
11484 Martin Liska <marxin.liska@gmail.com>
11485
03b0ee0a
UB
11486 * cgraph.c (cgraph_function_body_availability): Do not check
11487 cgraph flags.
af15184a
JH
11488 * cgraph.h (symtab_for_node_and_aliases, symtab_nonoverwritable_alias,
11489 symtab_node_availability): Declare.
11490 * ipa.c (can_replace_by_local_alias): New.
11491 (function_and_variable_visibility): Use it.
03b0ee0a
UB
11492 * symtab.c (symtab_for_node_and_aliases,
11493 symtab_nonoverwritable_alias_1, symtab_nonoverwritable_alias): New.
af15184a 11494
01e54ef8
VM
114952013-08-02 Vladimir Makarov <vmakarov@redhat.com>
11496
11497 PR rtl-optimization/57963
03b0ee0a 11498 * lra-constraints.c (reverse_equiv_p, contains_reloaded_insn_p): New.
01e54ef8
VM
11499 (lra_constraints): Use them.
11500
e2fd7ca7
SN
115012013-08-02 Sofiane Naci <sofiane.naci@arm.com>
11502
03b0ee0a
UB
11503 * config/arm/types.md (define_attr "type"): Add "load_acq"
11504 and "store_rel".
e2fd7ca7
SN
11505 * config/arm/cortex-a53.md (cortex_a53_load1): Update for attribute
11506 changes.
11507 (cortex_a53_store1): Likewise.
11508
0dcc26c3
JH
115092013-08-01 Jan Hubicka <jh@suse.cz>
11510
03b0ee0a
UB
11511 * ipa.c (symtab_remove_unreachable_nodes): Nodes in other
11512 partitions are not needed.
0dcc26c3 11513
0bd72901
UB
115142013-08-01 Uros Bizjak <ubizjak@gmail.com>
11515
11516 * config/i386/i386.h (MAYBE_NON_Q_CLASS_P): New.
11517 * config/i386/i386.c (ix86_secondary_reload): Use INTEGER_CLASS_P and
11518 MAYBE_NON_Q_CLASS_P where appropriate.
11519
12123452
JH
115202013-08-01 Jan Hubicka <jh@suse.cz>
11521
11522 * cgraph.h (release_function_body): Declare.
11523 * tree.c (free_lang_data_in_decl): Free, parameters and return values
11524 of unused delcarations.
11525
ce852f9c
KT
115262013-08-01 Kyrylo Tkachov <kyrylo.tkachov@arm.com>
11527
03b0ee0a
UB
11528 * config/arm/arm.md (minmax_arithsi_non_canon): Emit canonical
11529 RTL form when subtracting a constant.
ce852f9c 11530
c743b246
KT
115312013-08-01 Kyrylo Tkachov <kyrylo.tkachov@arm.com>
11532
11533 * config/arm/arm.md (peepholes for eq (reg1) (reg2/imm)):
11534 Generate canonical plus rtx with negated immediate instead of minus
11535 where appropriate.
11536 * config/arm/arm.c (thumb2_reorg): Handle ADCS <Rd>, <Rn> case.
11537
c0c123ef
JH
115382013-08-01 Jan Hubicka <jh@suse.cz>
11539
11540 * cgraph.c (cgraph_release_function_body): Use used_as_abstract_origin.
11541 (cgraph_release_function_body): Likewise.
11542 (cgraph_can_remove_if_no_direct_calls_p): Likewise.
11543 * cgraph.h (cgrpah_node): Rename abstract_and_needed
11544 to used_as_abstract_origin.
11545 * tree-inline-transfrom.c (can_remove_node_now_p_1): Do not remove
11546 symbols used as abstract origins.
11547 * cgraphunit.c (analyze_functions): Update.
0bd72901
UB
11548 * ipa.c (symtab_remove_unreachable_nodes): Recompute
11549 used_as_abstract_origin.
c0c123ef 11550 * tree-inline.c (tree_function_versioning): Update
0bd72901
UB
11551 used_as_abstract_origin; be ready for DECL_RESULT and
11552 DECL_ARGUMENTS to be NULL.
c0c123ef 11553
0bd72901
UB
11554 * lto-symtab.c (lto_symtab_merge_symbols): Merge duplicated nodes
11555 for abstract functions.
11556 * cgraph.h (symtab_real_symbol_p): Abstract declarations are not
11557 real symbols.
c0c123ef 11558
1f029433
JH
115592013-08-01 Jan Hubicka <jh@suse.cz>
11560
11561 * profile.c (compute_value_histograms): Fix thinko.
11562
c451f4d6
SN
115632013-08-01 Sofiane Naci <sofiane.naci@arm.com>
11564
11565 * config.gcc (aarch64*-*-*): Add aarch-common.o to extra_objs. Add
11566 aarch-common-protos.h to extra_headers.
11567 (aarch64*-*-*): Add arm/aarch-common-protos.h to tm_p_file.
11568 * config/aarch64/aarch64.md: Include "../arm/cortex-a53.md".
11569 * config/aarch64/t-aarch64 (aarch-common.o): Define.
11570
34b512f5
SN
115712013-08-01 Sofiane Naci <sofiane.naci@arm.com>
11572
11573 * config/aarch64/aarch64.md (define_attr "type"): Delete.
11574 Include "../arm/types.md". Define "type" attribute for all patterns.
11575 * config/aarch64/aarch64-simd.md (move_lo_quad_<mode>): Update for
11576 attribute changes.
11577
d86e633a
MM
115782013-07-31 Michael Meissner <meissner@linux.vnet.ibm.com>
11579
11580 * config/rs6000/predicates.md (fusion_gpr_addis): New predicates
11581 to support power8 load fusion.
11582 (fusion_gpr_mem_load): Likewise.
11583
11584 * config/rs6000/rs6000-modes.def (PTImode): Update a comment.
11585
11586 * config/rs6000/rs6000-protos.h (fusion_gpr_load_p): New
11587 declarations for power8 load fusion.
11588 (emit_fusion_gpr_load): Likewise.
11589
11590 * config/rs6000/rs6000.c (rs6000_option_override_internal): If
11591 tuning for power8, turn on fusion mode by default. Turn on sign
11592 extending fusion mode if normal fusion mode is on, and we are at
11593 -O2 or -O3.
11594 (fusion_gpr_load_p): New function, return true if we can fuse an
11595 addis instruction with a dependent load to a GPR.
11596 (emit_fusion_gpr_load): Emit the instructions for power8 load
11597 fusion to GPRs.
11598
0bd72901 11599 * config/rs6000/vsx.md (VSX_M2): New iterator for fusion peepholes.
d86e633a
MM
11600 (VSX load fusion peepholes): New peepholes to fuse together an
11601 addi instruction with a VSX load instruction.
11602
11603 * config/rs6000/rs6000.md (GPR load fusion peepholes): New
11604 peepholes to fuse an addis instruction with a load to a GPR base
11605 register. If we are supporting sign extending fusions, convert
11606 sign extending loads to zero extending loads and add an explicit
11607 sign extension.
11608
c3f35647
SN
116092013-07-31 Sofiane Naci <sofiane.naci@arm.com>
11610
11611 * config.gcc (arm*-*-*): Add aarch-common.o to extra_objs. Add
11612 aarch-common-protos.h to extra_headers.
11613 (arm*-*-*): Add arm/aarch-common-protos.h to tm_p_file.
11614 * config/arm/arm.c (arm_early_load_addr_dep): Move from here to ...
11615 (arm_early_store_addr_dep): Likewise.
0bd72901
UB
11616 (arm_no_early_alu_shift_dep): Likewise.
11617 (arm_no_early_alu_shift_value_dep): Likewise.
11618 (arm_no_early_mul_dep): Likewise.
11619 (arm_no_early_store_addr_dep): Likewise.
11620 (arm_mac_accumulator_is_mul_result): Likewise.
11621 (arm_mac_accumulator_is_result): Likewise.
c3f35647 11622 * config/arm/aarch-common.c: ... here. New file.
0bd72901
UB
11623 * config/arm/arm-protos.h (arm_early_load_addr_dep): Move from
11624 here to ...
c3f35647 11625 (arm_early_store_addr_dep): Likewise.
0bd72901
UB
11626 (arm_no_early_alu_shift_dep): Likewise.
11627 (arm_no_early_alu_shift_value_dep): Likewise.
11628 (arm_no_early_mul_dep): Likewise.
11629 (arm_no_early_store_addr_dep): Likewise.
11630 (arm_mac_accumulator_is_mul_result): Likewise.
11631 (arm_mac_accumulator_is_result): Likewise.
c3f35647
SN
11632 * config/arm/aarch-common-protos.h: ... here. New file.
11633 * config/arm/t-arm (aarch-common.o): Define.
11634
116352013-07-31 Sofiane Naci <sofiane.naci@arm.com>
11636
11637 * config/arm/arm.md: Include new file "types.md".
11638 (define_attr "type"): Move from here to ...
11639 (define_attr "mul32"): Likewise.
11640 (define_attr "mul64"): Likewise.
11641 * config/arm/types.md: ... here. New file.
11642
228c1313
SH
116432013-07-31 Sebastian Huber <sebastian.huber@embedded-brains.de>
11644
11645 * config.gcc (*-*-rtems*): Use __cxa_atexit by default.
11646 * config/rs6000/rtems.h (TARGET_LIBGCC_SDATA_SECTION): Define.
11647
d6e9046d 116482013-07-31 Jan-Benedict Glaw <jbglaw@lug-owl.de>
8ac499bb
JBG
11649
11650 * gen-pass-instances.awk: Fix offset of substr().
11651
a167b052
DM
116522013-07-31 David Malcolm <dmalcolm@redhat.com>
11653
11654 * Makefile.in (pass-instances.def): New.
11655 (passes.o): Replace dependency on passes.def with one on
11656 pass-instances.def
11657
11658 * gen-pass-instances.awk: New.
11659
11660 * passes.c (pass_manager::pass_manager): Use pass-instances.def
11661 rather than passes.def, updating local definition of NEXT_PASS
11662 macro to add an extra NUM parameter (currently unused).
11663
315f8c0e
DM
116642013-07-30 David Malcolm <dmalcolm@redhat.com>
11665
11666 * Makefile.in (PASS_MANAGER_H): New.
11667 (lto-cgraph.o): Depend on CONTEXT_H and PASS_MANAGER_H.
11668 (passes.o): Likewise.
11669 (statistics.o): Likewise.
11670 (cgraphunit.o): Likewise.
11671 (context.o): Depend on PASS_MANAGER_H.
11672
11673 * pass_manager.h: New.
11674
11675 * cgraphunit.c (cgraph_add_new_function): Update for moves
11676 of globals to fields of pass_manager.
11677 (analyze_function): Likewise.
11678 (expand_function): Likewise.
11679 (ipa_passes): Likewise.
11680 (compile): Likewise.
11681
11682 * context.c (context::context): New.
11683 * context.h (context::context): New.
11684 (context::get_passes): New.
11685 (context::passes_): New.
11686
11687 * lto-cgraph.c (input_node): Update for moves of globals to
11688 fields of pass_manager.
11689
11690 * passes.c (all_passes): Remove, in favor of a field of the
11691 same name within the new class pass_manager.
11692 (all_small_ipa_passes): Likewise.
11693 (all_lowering_passes): Likewise.
11694 (all_regular_ipa_passes): Likewise.
11695 (all_late_ipa_passes): Likewise.
11696 (all_lto_gen_passes): Likewise.
11697 (passes_by_id): Likewise.
11698 (passes_by_id_size): Likewise.
11699 (gcc_pass_lists): Remove, in favor of "pass_lists" field within
11700 the new class pass_manager.
11701 (set_pass_for_id): Convert to...
11702 (pass_manager::set_pass_for_id): ...method.
11703 (get_pass_for_id): Convert to...
11704 (pass_manager::get_pass_for_id): ...method.
11705 (register_one_dump_file): Move body of implementation into...
11706 (pass_manager::register_one_dump_file): ...here.
11707 (register_dump_files_1): Convert to...
11708 (pass_manager::register_dump_files_1): ...method.
11709 (register_dump_files): Convert to...
11710 (pass_manager::register_dump_files): ...method.
11711 (create_pass_tab): Update for moves of globals to fields of
11712 pass_manager.
11713 (dump_passes): Move body of implementation into...
11714 (pass_manager::dump_passes): ...here.
11715 (register_pass): Move body of implementation into...
11716 (pass_manager::register_pass): ...here.
11717 (init_optimization_passes): Convert into...
11718 (pass_manager::pass_manager): ...constructor for new
11719 pass_manager class, and initialize the pass_lists array.
11720 (check_profile_consistency): Update for moves of globals to
11721 fields of pass_manager.
11722 (dump_profile_report): Move body of implementation into...
11723 (pass_manager::dump_profile_report): ...here.
11724 (ipa_write_summaries_1): Update for moves of pass lists from
11725 being globals to fields of pass_manager.
11726 (ipa_write_optimization_summaries): Likewise.
11727 (ipa_read_summaries): Likewise.
11728 (ipa_read_optimization_summaries): Likewise.
11729 (execute_all_ipa_stmt_fixups): Likewise.
11730
11731 * statistics.c (statistics_fini): Update for moves of globals to
11732 fields of pass_manager.
11733
11734 * toplev.c (general_init): Replace call to
11735 init_optimization_passes with construction of the pass_manager
11736 instance.
11737
11738 * tree-pass.h (all_passes): Remove, in favor of a field of the
11739 same name within the new class pass_manager.
11740 (all_small_ipa_passes): Likewise.
11741 (all_lowering_passes): Likewise.
11742 (all_regular_ipa_passes): Likewise.
11743 (all_lto_gen_passes): Likewise.
11744 (all_late_ipa_passes): Likewise.
11745 (passes_by_id): Likewise.
11746 (passes_by_id_size): Likewise.
11747 (gcc_pass_lists): Remove, in favor of "pass_lists" field within
11748 the new class pass_manager.
11749 (get_pass_for_id): Remove.
11750
02cab1c1
RE
117512013-07-30 Richard Earnshaw <rearnsha@arm.com>
11752
11753 * config.gcc (arm): Require 64-bit host-wide-int for all ARM target
11754 configs.
11755
b2d3f886
RE
117562013-07-30 Richard Earnshaw <rearnsha@arm.com>
11757
11758 * arm.md (mulhi3): New expand pattern.
11759
605e86fa
JH
117602013-07-30 Jan Hubicka <jh@suse.cz>
11761 Martin Liska <marxin.liska@gmail.com>
11762
11763 * profile.c (compute_value_histograms): Do not ICE when
11764 there is mismatch only on some counters.
11765
14a87636
ZC
117662013-07-30 Zhenqiang Chen <zhenqiang.chen@linaro.org>
11767
11768 PR rtl-optimization/57637
11769 * function.c (move_insn_for_shrink_wrap): Also check the
11770 GEN set of the LIVE problem for the liveness analysis
11771 if it exists, otherwise give up.
11772
0100cd3f
BS
117732013-07-29 Bill Schmidt <wschmidt@vnet.linux.ibm.com>
11774
11775 PR tree-optimization/57993
11776 * gimple-ssa-strength-reduction.c (replace_mult_candidate): Record
11777 replaced statement in the candidate table.
11778 (phi_add_costs): Return infinite cost when the hidden basis does
11779 not dominate all phis on which the candidate is dependent.
11780 (replace_one_candidate): Record replaced statement in the
11781 candidate table.
11782
f0d811f7
JR
117832013-07-29 Joern Rennecke <joern.rennecke@embecosm.com>
11784
11785 * config/epiphany/epiphany.md (*isub_i+2): New peephole.
11786 (ashlv2si3): New expander.
11787 (*ashlv2si3_i): New define_insn_and_split.
11788 * predicates.md (float_operation): Allow patterns with three
11789 basic sub-patterns.
11790
bba33211
JR
11791 PR rtl-optimization/58021
11792 * mode-switching.c (create_pre_exit): Always split off preceding
11793 insns if we are not at the basic block head.
11794
ff3f3951
MR
117952013-07-29 Maciej W. Rozycki <macro@codesourcery.com>
11796
11797 * config/mips/linux.h (GLIBC_DYNAMIC_LINKER): Handle `-mnan=2008'.
11798 (UCLIBC_DYNAMIC_LINKER): New macro.
11799 * config/mips/linux64.h (GLIBC_DYNAMIC_LINKER32): Handle
11800 `-mnan=2008'.
11801 (GLIBC_DYNAMIC_LINKER64, GLIBC_DYNAMIC_LINKERN32): Likewise.
11802 (UCLIBC_DYNAMIC_LINKER32): Undefine macro first. Handle
11803 `-mnan=2008'.
11804 (UCLIBC_DYNAMIC_LINKER64): Redefine macro.
11805 (UCLIBC_DYNAMIC_LINKERN32): Likewise.
11806 * config/mips/mips-modes.def: Remove RESET_FLOAT_FORMAT calls
11807 for SF and DF modes. Use ieee_quad_format for TF mode.
11808 * config/mips/mips-opts.h (mips_ieee_754_setting): New enum.
11809 * config/mips/mips.c (mips_file_start): Output a `.nan' directive.
11810 (mips_option_override): Handle `-mnan=legacy'.
11811 * config/mips/mips.h (TARGET_CPU_CPP_BUILTINS): Handle
11812 `-mabs=2008' and `-mnan=2008'.
11813 (OPTION_DEFAULT_SPECS): Add "nan" default.
11814 (ASM_SPEC): Handle `-mnan='.
11815 [!HAVE_AS_NAN] (HAVE_AS_NAN): New macro.
11816 * config/mips/mips.md (abs<mode>2): Handle `-mabs=2008', update
11817 comment accordingly.
11818 (neg<mode>2): Likewise.
11819 * config/mips/mips.opt (mabs, mnan): New options.
11820 * doc/install.texi (Configuration): Document `--with-nan=' option.
11821 * doc/invoke.texi (Option Summary): List MIPS `-mabs=' and
11822 `-mnan=' options.
11823 (MIPS Options): Document them.
11824 * config.gcc <mips*-*-*>: Handle `--with-nan='.
11825 * configure.ac <mips*-*-*>: Check for GAS `-mnan=2008' support.
11826 * configure: Regenerate.
11827 * config.in: Regenerate.
11828
576e2f37
UB
118292013-07-29 Uros Bizjak <ubizjak@gmail.com>
11830
11831 * config/i386/i386.md (float post-reload splitters): Do not check
11832 for subregs of SSE registers.
11833
118342013-07-29 Uros Bizjak <ubizjak@gmail.com>
11835 H.J. Lu <hongjiu.lu@intel.com>
11836
11837 PR target/57954
11838 PR target/57988
11839 * config/i386/i386.md (post-reload splitter
11840 to avoid partial SSE reg dependency stalls): New pattern.
11841
a71f0749
DV
118422013-07-29 Dominik Vogt <vogt@linux.vnet.ibm.com>
11843
11844 * config/s390/s390.md ("movcc"): Swap load and store instructions.
11845
4c97f1cc
JR
118462013-07-27 Joern Rennecke <joern.rennecke@embecosm.com>
11847
11848 * config/epiphany/epiphany.c (epiphany_compute_frame_size):
11849 Also reserve space for saving UNKNOWN_REGNUM for leaf functions.
11850
b0c733d8
CC
118512013-07-26 Cary Coutant <ccoutant@google.com>
11852
11853 * dwarf2out.c (die_checksum_ordered): Don't include template
11854 instantiations in signature.
11855 (is_template_parameter): New function.
11856 (is_template_instantiation): New function.
11857 (generate_skeleton_bottom_up): Don't include template instantiations
11858 in type unit DIE.
11859 (generate_skeleton): Likewise.
11860 (break_out_comdat_types): Move recursive call to break out nested
11861 types earlier.
11862 (prune_unused_types_mark_generic_parms_dies): Call
11863 is_template_parameter.
11864
040d8a1c
IB
118652013-07-26 Ian Bolton <ian.bolton@arm.com>
11866
11867 * config/aarch64/aarch64.md (neg<mode>2): Offer alternative that
11868 uses vector registers.
89fdc743 11869 * config/aarch64/iterators.md: Add attributes rtn and vas.
040d8a1c 11870
85bd4ac6 118712013-07-26 Kyrylo Tkachov <kyrylo.tkachov@arm.com>
576e2f37 11872 Richard Earnshaw <richard.earnshaw@arm.com>
85bd4ac6
KT
11873
11874 * combine.c (simplify_comparison): Re-canonicalize operands
576e2f37 11875 where appropriate.
85bd4ac6
KT
11876 * config/arm/arm.md (movcond_addsi): New splitter.
11877
b6051207
SA
118782013-07-25 Sterling Augustine <saugustine@google.com>
11879
11880 * dwarf2out.c (size_of_pubnames): Move code to...
11881 (include_pubname_in_output): ...here. New.
11882 (want_pubnames): Rearrange.
11883 (output_pubnames): Call include_pubname_in_output. Move assertion.
11884
41a7c215
CM
118852013-07-25 Cameron McInally <cameron.mcinally@nyu.edu>
11886
11887 * doc/extend.texi: Fix return types for __builtin_ia32_cmp*s builtins.
11888
118892013-07-25 Cameron McInally <cameron.mcinally@nyu.edu>
11890
11891 PR target/38836
11892 * doc/extend.texi: Remove obsolete builtins. Fix
11893 typo for __builtin_ia32_loadss and __builtin_ia32_cmpnltss.
11894
49bde175
JH
118952013-07-25 Jan Hubicka <jh@suse.cz>
11896
11897 * cgraph.c (release_function_body): Break out from ...
11898 (cgraph_release_function_body): ... this one; also release DECL_RESULT
11899 and DECL_ARGUMENTS.
11900 * ipa-cp.c (get_replacement_map): Add parm_num argument; do not set
11901 old_tree in the map.
11902 (create_specialized_node): Update.
11903 * lto-cgraph.c (output_node_opt_summary): Do not translate old_tree
11904 into index.
41a7c215
CM
11905 * cgraphclones.c (cgraph_create_virtual_clone): Do not copy
11906 DECL_ARGUMENTS, DECL_INITIAL and DECL_RESULT.
49bde175
JH
11907 * ipa-prop.c (ipa_populate_param_decls): Look for origin of clones.
11908 * tree-inline.c (initialize_cfun): Initialize DECL_ARGUMENTS and
11909 DECL_RESULT.
11910
4870352d
KT
119112013-07-25 Kyrylo Tkachov <kyrylo.tkachov@arm.com>
11912
11913 * config/arm/arm.md (arm_addsi3, addsi3_carryin_<optab>,
11914 addsi3_carryin_alt2_<optab>): Correct output template.
11915
1d6aee1c
KT
119162013-07-25 Kyrylo Tkachov <kyrylo.tkachov@arm.com>
11917
11918 * config/arm/arm-fixed.md (ssmulsa3, usmulusa3):
11919 Adjust for arm_restrict_it.
11920 Remove trailing whitespace.
11921
c7cafd75
MK
119222013-07-25  Mark Kettenis  <kettenis@openbsd.org>
11923
017d38f5
MK
11924 * config/pa/pa.c (pa_trampoline_init): Emit __enable_execute_stack
11925 libcall if HAVE_ENABLE_EXECUTE_STACK is defined.
11926
c7cafd75
MK
11927 * config.gcc (hppa-*-openbsd*): Don't set tmake_file.
11928
00b7527b
VM
119292013-07-25 Vladimir Makarov <vmakarov@redhat.com>
11930
11931 PR rtl-optimization/57960
11932 * lra-constraints.c (process_alt_operands): Use the right mode
11933 when checking strict_low.
11934
4f63dfc6
JH
119352013-07-25 Jan Hubicka <jh@suse.cz>
11936
11937 * lto-symtab.c (lto_cgraph_replace_node): Release function body.
11938 * cgraph.c (cgraph_remove_node): Do not release function body
11939 when in cgraph streaming.
11940 * ipa.c (process_references, symtab_remove_unreachable_nodes): Objects
41a7c215
CM
11941 in other partitions are not considered reachable; fix handling of
11942 clones.
4f63dfc6 11943
e6dcfa18
RR
119442013-07-25 Ramana Radhakrishnan <ramana.radhakrishnan@arm.com>
11945
11946 * config/arm/arm.md (*sibcall_insn): Remove unnecessary space.
11947
0b93d3b6
RR
119482013-07-25 Ramana Radhakrishnan <ramana.radhakrishnan@arm.com>
11949
11950 PR target/19599
11951 PR target/57731
bb80c2eb 11952 PR target/57837
e6dcfa18 11953 * config/arm/arm.md ("*sibcall_insn): Replace use of
0b93d3b6 11954 Ss with US. Adjust output for v5 and v4t.
41a7c215 11955 (*sibcall_value_insn): Likewise and loosen predicate on operand0.
e6dcfa18 11956
0b93d3b6
RR
11957 * config/arm/constraints.md ("Ss"): Rename to US.
11958
690688b3
TG
119592013-07-25 Terry Guo <terry.guo@arm.com>
11960
11961 * config/arm/arm.c (thumb1_size_rtx_costs): Assign proper cost for
11962 shift_add/shift_sub0/shift_sub1 RTXs.
11963
a4ad093b 119642013-07-24 Bill Schmidt <wschmidt@linux.ibm.com>
12211b99 11965 Anton Blanchard <anton@au1.ibm.com>
a4ad093b
BS
11966
11967 * config/rs6000/altivec.md (altivec_vpkpx): Handle little endian.
11968 (altivec_vpks<VI_char>ss): Likewise.
11969 (altivec_vpks<VI_char>us): Likewise.
11970 (altivec_vpku<VI_char>us): Likewise.
11971 (altivec_vpku<VI_char>um): Likewise.
11972
8c5005ce
DM
119732013-07-24 David Malcolm <dmalcolm@redhat.com>
11974
11975 Introduce context class.
11976
11977 * Makefile.in (CONTEXT_H): New.
11978 (OBJS): Add context.o.
11979 (toplev.o): Add CONTEXT_H to dependencies.
11980 (context.o): New.
11981
41a7c215 11982 * toplev.c (general_init): Create the singleton gcc::context instance.
8c5005ce
DM
11983
11984 * context.c: New.
11985
11986 * context.h: New.
11987
ce4a9422
JR
119882013-07-24 Joern Rennecke <joern.rennecke@embecosm.com>
11989
11990 PR rtl-optimization/57968
11991 * mode-switching.c (create_pre_exit): Allow instructions that
11992 don't set a return register to need a non-exit mode.
11993
493f4c9e 119942013-07-24 Bill Schmidt <wschmidt@vnet.linux.ibm.com>
12211b99 11995 Anton Blanchard <anton@au1.ibm.com>
493f4c9e 11996
b5860fd3
WS
11997 * config/rs6000/vector.md (vec_realign_load_<mode>): Reorder input
11998 operands to vperm for little endian.
11999 * config/rs6000/rs6000.c (rs6000_expand_builtin): Use lvsr instead
12000 of lvsl to create the control mask for a vperm for little endian.
493f4c9e 12001
73792b92 120022013-07-23 Bill Schmidt <wschmidt@linux.vnet.ibm.com>
12211b99 12003 Anton Blanchard <anton@au1.ibm.com>
73792b92
BS
12004
12005 * config/rs6000/rs6000.c (altivec_expand_vec_perm_const): Reverse
12006 two operands for little-endian.
12007
14c21302
SE
120082013-07-23 Steve Ellcey <sellcey@mips.com>
12009
12010 * config/mips/mips.c (mips_case_values_threshold): New.
12011 (TARGET_CASE_VALUES_THRESHOLD): Define.
12012
a6056198 120132013-07-23 Bill Schmidt <wschmidt@linux.vnet.ibm.com>
12211b99 12014 Anton Blanchard <anton@au1.ibm.com>
a6056198
BS
12015
12016 * config/rs6000/rs6000.c (altivec_expand_vec_perm_const): Correct
12017 selection of field for vector splat in little endian mode.
12018
dd7a40e1
MM
120192013-07-23 Michael Meissner <meissner@linux.vnet.ibm.com>
12020
12021 * config/rs6000/vector.md (xor<mode>3): Move 128-bit boolean
12022 expanders to rs6000.md.
12023 (ior<mode>3): Likewise.
12024 (and<mode>3): Likewise.
12025 (one_cmpl<mode>2): Likewise.
12026 (nor<mode>3): Likewise.
12027 (andc<mode>3): Likewise.
12028 (eqv<mode>3): Likewise.
12029 (nand<mode>3): Likewise.
12030 (orc<mode>3): Likewise.
12031
12032 * config/rs6000/rs6000-protos.h (rs6000_split_logical): New
12033 declaration.
12034
12035 * config/rs6000/rs6000.c (rs6000_split_logical_inner): Add support
12036 to split multi-word logical operations.
12037 (rs6000_split_logical_di): Likewise.
12038 (rs6000_split_logical): Likewise.
12039
12040 * config/rs6000/vsx.md (VSX_L2): Delete, no longer used.
12041 (vsx_and<mode>3_32bit): Move 128-bit logical insns to rs6000.md,
12042 and allow TImode operations in 32-bit.
12043 (vsx_and<mode>3_64bit): Likewise.
12044 (vsx_ior<mode>3_32bit): Likewise.
12045 (vsx_ior<mode>3_64bit): Likewise.
12046 (vsx_xor<mode>3_32bit): Likewise.
12047 (vsx_xor<mode>3_64bit): Likewise.
12048 (vsx_one_cmpl<mode>2_32bit): Likewise.
12049 (vsx_one_cmpl<mode>2_64bit): Likewise.
12050 (vsx_nor<mode>3_32bit): Likewise.
12051 (vsx_nor<mode>3_64bit): Likewise.
12052 (vsx_andc<mode>3_32bit): Likewise.
12053 (vsx_andc<mode>3_64bit): Likewise.
12054 (vsx_eqv<mode>3_32bit): Likewise.
12055 (vsx_eqv<mode>3_64bit): Likewise.
12056 (vsx_nand<mode>3_32bit): Likewise.
12057 (vsx_nand<mode>3_64bit): Likewise.
12058 (vsx_orc<mode>3_32bit): Likewise.
12059 (vsx_orc<mode>3_64bit): Likewise.
12060
12061 * config/rs6000/rs6000.h (VLOGICAL_REGNO_P): Always allow vector
12062 logical types in GPRs.
12063
12064 * config/rs6000/altivec.md (altivec_and<mode>3): Move 128-bit
12065 logical insns to rs6000.md, and allow TImode operations in
12066 32-bit.
12067 (altivec_ior<mode>3): Likewise.
12068 (altivec_xor<mode>3): Likewise.
12069 (altivec_one_cmpl<mode>2): Likewise.
12070 (altivec_nor<mode>3): Likewise.
12071 (altivec_andc<mode>3): Likewise.
12072
12073 * config/rs6000/rs6000.md (BOOL_128): New mode iterators and mode
12074 attributes for moving the 128-bit logical operations into
12075 rs6000.md.
12076 (BOOL_REGS_OUTPUT): Likewise.
12077 (BOOL_REGS_OP1): Likewise.
12078 (BOOL_REGS_OP2): Likewise.
12079 (BOOL_REGS_UNARY): Likewise.
12080 (BOOL_REGS_AND_CR0): Likewise.
12081 (one_cmpl<mode>2): Add support for DI logical operations on
12082 32-bit, splitting the operations to 32-bit.
12083 (anddi3): Likewise.
12084 (iordi3): Likewise.
12085 (xordi3): Likewise.
12086 (and<mode>3, 128-bit types): Rewrite 2013-06-06 logical operator
12087 changes to combine the 32/64-bit code, allow logical operations on
12088 TI mode in 32-bit, and to use similar match_operator patterns like
12089 scalar mode uses. Combine the Altivec and VSX code for logical
12090 operations, and move it here.
12091 (ior<mode>3, 128-bit types): Likewise.
12092 (xor<mode>3, 128-bit types): Likewise.
12093 (one_cmpl<mode>3, 128-bit types): Likewise.
12094 (nor<mode>3, 128-bit types): Likewise.
12095 (andc<mode>3, 128-bit types): Likewise.
12096 (eqv<mode>3, 128-bit types): Likewise.
12097 (nand<mode>3, 128-bit types): Likewise.
12098 (orc<mode>3, 128-bit types): Likewise.
12099 (and<mode>3_internal): Likewise.
12100 (bool<mode>3_internal): Likewise.
12101 (boolc<mode>3_internal1): Likewise.
12102 (boolc<mode>3_internal2): Likewise.
12103 (boolcc<mode>3_internal1): Likewise.
12104 (boolcc<mode>3_internal2): Likewise.
12105 (eqv<mode>3_internal1): Likewise.
12106 (eqv<mode>3_internal2): Likewise.
12107 (one_cmpl1<mode>3_internal): Likewise.
12108
12211b99 121092013-07-23 David Holsgrove <david.holsgrove@xilinx.com>
9b69d4b4
DH
12110
12111 * config/microblaze/microblaze.c (microblaze_expand_prologue):
12112 Rename flag_stack_usage to flag_stack_usage_info.
12113
12211b99 121142013-07-23 David Holsgrove <david.holsgrove@xilinx.com>
450b0ebb 12115
a5f257fd
ME
12116 * config/microblaze/sync.md: New file.
12117 * config/microblaze/microblaze.md: Include sync.md
12118 * config/microblaze/microblaze.c: Add print_operand 'y'.
12119 * config/microblaze/constraints.md: Add memory_contraint
12120 'Q' which is a single register.
450b0ebb 12121
49fba14f
EB
121222013-07-23 Eric Botcazou <ebotcazou@adacore.com>
12123
12124 * doc/invoke.texi (SPARC Options): Document new leon3 processor value.
12125
c7b3b99f
PCC
121262013-07-22 Po-Chun Chang <pchang9@cs.wisc.edu>
12127
12128 * reload.c (find_reloads): Exit loop once we find this operand
12129 cannot be reloaded somehow for this alternative.
12130
12131 * reload.c (find_reloads): Exit loop once we find a hard register.
12132
12133 * rtlanal.c (computed_jump_p): Exit loop once we find label
12134 reference is used.
12135
12136 * i386.c (ix86_pad_returns): Exit loop after setting replace.
12137
12138 * cfgloopmanip.c (remove_path): Exit loop after setting
12139 irred_invalidated.
12140
12141 * gensupport.c (subst_dup): Avoid loop if code is not
12142 MATCH_DUP nor MATCH_OP_DUP.
12143
c8fbf1fa
NBJ
121442013-07-23 Nicklas Bo Jensen <nbjensen@gmail.com>
12145
12146 * doc/md.texi (Machine-Specific Peephole Optimizers): Fix a typo.
12147
9259db42
YZ
121482013-07-23 Yufeng Zhang <yufeng.zhang@arm.com>
12149
12150 * config/aarch64/aarch64.c (aarch64_hard_regno_mode_ok): Also return
12151 true for SP_REGNUM if mode == ptr_mode.
12152 * config/aarch64/aarch64.h (ADDITIONAL_REGISTER_NAMES): Add "wsp"
12153 with value R0_REGNUM + 31.
12154
00edcfbe
YZ
121552013-07-23 Yufeng Zhang <yufeng.zhang@arm.com>
12156
12157 * config/aarch64/aarch64.c (aarch64_pad_arg_upward): In big-endian,
12158 pad pointer-typed argument downward.
12159
43be9a95
YZ
121602013-07-23 Yufeng Zhang <yufeng.zhang@arm.com>
12161
12162 * config/aarch64/aarch64.h (TARGET_CPU_CPP_BUILTINS): Define _ILP32
12163 and __ILP32__ when the ILP32 model is in use.
12164
28514dda
YZ
121652013-07-23 Yufeng Zhang <yufeng.zhang@arm.com>
12166
12167 * config/aarch64/aarch64.c (POINTER_BYTES): New define.
12168 (aarch64_load_symref_appropriately): In the case of
12169 SYMBOL_SMALL_ABSOLUTE, use the mode of 'dest' instead of Pmode
12170 to generate new rtx; likewise to the case of SYMBOL_SMALL_GOT.
12171 (aarch64_expand_mov_immediate): In the case of SYMBOL_FORCE_TO_MEM,
12172 change to pass 'ptr_mode' to force_const_mem and zero-extend 'mem'
12173 if 'mode' doesn't equal to 'ptr_mode'.
12174 (aarch64_output_mi_thunk): Add an assertion on the alignment of
12175 'vcall_offset'; change to call aarch64_emit_move differently depending
12176 on whether 'Pmode' equals to 'ptr_mode' or not; use 'POINTER_BYTES'
12177 to calculate the upper bound of 'vcall_offset'.
12178 (aarch64_cannot_force_const_mem): Change to also return true if
12179 mode != ptr_mode.
12180 (aarch64_legitimize_reload_address): In the case of large
12181 displacements, add new local variable 'xmode' and an assertion
12182 based on it; change to use 'xmode' to generate the new rtx and
12183 reload.
12184 (aarch64_asm_trampoline_template): Change to generate the template
12185 differently depending on TARGET_ILP32 or not; change to use
12186 'POINTER_BYTES' in the argument passed to assemble_aligned_integer.
12187 (aarch64_trampoline_size): Removed.
12188 (aarch64_trampoline_init): Add new local constant 'tramp_code_sz'
12189 and replace immediate literals with it. Change to use 'ptr_mode'
12190 instead of 'DImode' and call convert_memory_address if the mode
12191 of 'fnaddr' doesn't equal to 'ptr_mode'.
12192 (aarch64_elf_asm_constructor): Change to use assemble_aligned_integer
12193 to output symbol.
12194 (aarch64_elf_asm_destructor): Likewise.
12195 * config/aarch64/aarch64.h (TRAMPOLINE_SIZE): Change to be dependent
12196 on TARGET_ILP32 instead of aarch64_trampoline_size.
12197 * config/aarch64/aarch64.md (movsi_aarch64): Add new alternatives
12198 of 'mov' between WSP and W registers as well as 'adr' and 'adrp'.
12199 (loadwb_pair<GPI:mode>_<PTR:mode>): Rename to ...
12200 (loadwb_pair<GPI:mode>_<P:mode>): ... this. Replace PTR with P.
12201 (storewb_pair<GPI:mode>_<PTR:mode>): Likewise; rename to ...
12202 (storewb_pair<GPI:mode>_<P:mode>): ... this.
12203 (add_losym): Change to 'define_expand' and call gen_add_losym_<mode>
12204 depending on the value of 'mode'.
12205 (add_losym_<mode>): New.
12206 (ldr_got_small_<mode>): New, based on ldr_got_small.
12207 (ldr_got_small): Remove.
12208 (ldr_got_small_sidi): New.
12209 * config/aarch64/iterators.md (P): New.
12210 (PTR): Change to 'ptr_mode' in the condition.
12211
17a819cb
YZ
122122013-07-23 Yufeng Zhang <yufeng.zhang@arm.com>
12213
12214 * config.gcc (aarch64*-*-*): Support --with-abi.
12215 (aarch64*-*-elf): Support --with-multilib-list.
12216 (aarch64*-*-linux*): Likewise.
12217 (supported_defaults): Add abi to aarch64*-*-*.
12218 * configure.ac: Mention AArch64 for --with-multilib-list.
12219 * configure: Re-generated.
12220 * config/aarch64/biarchilp32.h: New file.
12221 * config/aarch64/biarchlp64.h: New file.
12222 * config/aarch64/aarch64-elf.h (ENDIAN_SPEC): New define.
12223 (ABI_SPEC): Ditto.
12224 (MULTILIB_DEFAULTS): Ditto.
12225 (DRIVER_SELF_SPECS): Ditto.
12226 (ASM_SPEC): Update to also substitute -mabi.
12227 * config/aarch64/aarch64-elf-raw.h (LINK_SPEC): Add linker script
12228 file whose name depends on -mabi= and -mbig-endian.
12229 * config/aarch64/aarch64.h (LONG_TYPE_SIZE): Change to depend on
12230 TARGET_ILP32.
12231 (POINTER_SIZE): New define.
12232 (POINTERS_EXTEND_UNSIGNED): Ditto.
12233 (enum aarch64_abi_type): New enumeration tag.
12234 (AARCH64_ABI_LP64, AARCH64_ABI_ILP32): New enumerators.
12235 (AARCH64_ABI_DEFAULT): Define to AARCH64_ABI_LP64 if undefined.
12236 (TARGET_ILP32): New define.
12237 * config/aarch64/aarch64.opt (mabi): New.
12238 (aarch64_abi): New.
12239 (ilp32, lp64): New values for -mabi.
12240 * config/aarch64/t-aarch64 (comma): New define.
12241 (MULTILIB_OPTIONS): Ditto.
12242 (MULTILIB_DIRNAMES): Ditto.
12243 * config/aarch64/t-aarch64-linux (MULTIARCH_DIRNAME): New define.
12244 * doc/invoke.texi: Document -mabi for AArch64.
12245
9f9cbdce
GJL
122462013-07-23 Georg-Johann Lay <avr@gjlay.de>
12247
12248 * config/avr/avr.md: Explain asm print modifier 'r' for REG.
12249
7061977b 122502013-07-22 Bill Schmidt <wschmidt@vnet.linux.ibm.com>
12211b99 12251 Anton Blanchard <anton@au1.ibm.com>
7061977b
BS
12252
12253 * config/rs6000/rs6000.c (rs6000_expand_vector_init): Fix
12254 endianness when selecting field to splat.
12255
75a41878
EC
122562013-07-22 Eric Christopher <echristo@gmail.com>
12257
12258 * dwarf2out.c (die_odr_checksum): New function to use
12259 CHECKSUM_ macros and ULEB128 for DIE tag.
12260 (generate_type_signature): Use.
41a7c215 12261
38ae58ca
EB
122622013-07-22 Eric Botcazou <ebotcazou@adacore.com>
12263
12264 * config.gcc (sparc*-*-*): Accept leon3 processor.
12265 (sparc-leon*-*): Merge with sparc*-*-* and add leon3 support.
12266 * doc/invoke.texi (SPARC Options): Adjust -mfix-ut699 entry.
12267 * config/sparc/sparc-opts.h (enum processor_type): Add PROCESSOR_LEON3.
12268 * config/sparc/sparc.opt (enum processor_type): Add leon3.
12269 (mfix-ut699): Adjust comment.
12270 * config/sparc/sparc.h (TARGET_CPU_leon3): New define.
12271 (CPP_CPU32_DEFAULT_SPEC): Add leon3 support.
12272 (CPP_CPU_SPEC): Likewise.
12273 (ASM_CPU_SPEC): Likewise.
12274 * config/sparc/sparc.c (leon3_cost): New constant.
12275 (sparc_option_override): Add leon3 support.
12276 (mem_ref): New function.
12277 (sparc_gate_work_around_errata): Return true if -mfix-ut699 is enabled.
12278 (sparc_do_work_around_errata): Look into the instruction in the delay
12279 slot and adjust accordingly. Add fix for the data cache nullify issues
12280 of the UT699. Change insertion position for the NOP.
12281 * config/sparc/leon.md (leon_fpalu, leon_fpmds, write_buf): Delete.
12282 (leon3_load): New reservation.
12283 (leon_store): Bump latency to 2.
12284 (grfpu): New automaton.
12285 (grfpu_alu): New unit.
12286 (grfpu_ds): Likewise.
12287 (leon_fp_alu): Adjust.
12288 (leon_fp_mult): Delete.
12289 (leon_fp_div): Split into leon_fp_divs and leon_fp_divd.
12290 (leon_fp_sqrt): Split into leon_fp_sqrts and leon_fp_sqrtd.
12291 * config/sparc/sparc.md (cpu): Add leon3.
12292 * config/sparc/sync.md (atomic_exchangesi): Disable if -mfix-ut699.
12293 (swapsi): Likewise.
12294 (atomic_test_and_set): Likewise.
12295 (ldstub): Likewise.
12296
04dfc6df
JU
122972013-07-22 Jürgen Urban <JuergenUrban@gmx.de>
12298
12299 * config.gcc (mips*-*-*): Add --with-fpu support. Make single the
12300 default for R5900 targets.
12301 * config/mips/mips.h (OPTION_DEFAULT_SPECS): Handle --with-fpu.
12302 (ISA_HAS_LDC1_SDC1): Set to false for TARGET_MIPS5900.
12303 * config/mips/mips.c (mips_option_override): Report an error for
12304 -march=r5900 -mhard-float -mdouble-float. Use spu_single_format
12305 for -march=r5900 -mhard-float.
12306
c7b3b99f 123072013-07-22 Po-Chun Chang <pchang9@cs.wisc.edu>
ae382ebd
PCC
12308
12309 * df-problems.c (can_move_insns_across): Exit loop once we
12310 find a non-fixed, non-global register.
12311
12312 * ipa-pure-const.c (propagate_nothrow): Exit loop after
12313 setting can_throw.
12314
12315 * omega.c (omega_eliminate_red): Break after setting red_found.
12316 (omega_problem_has_red_equations): Similarly after setting found.
12317 (omega_query_variable): Similarly after setting coupled.
12318
9b6e6981
MP
123192013-07-22 Marek Polacek <polacek@redhat.com>
12320
12321 * gimplify.c: Don't include gimple.h twice.
12322
5e5f7673
KT
123232013-07-22 Kyrylo Tkachov <kyrylo.tkachov@arm.com>
12324
12325 * config/arm/constraints.md (Pd): Allow TARGET_THUMB
12326 instead of TARGET_THUMB1.
12327 (Pz): New constraint.
12328 * config/arm/arm.md (arm_addsi3): Add alternatives for 16-bit
12329 encodings.
12330 (compare_negsi_si): Likewise.
12331 (compare_addsi2_op0): Likewise.
12332 (compare_addsi2_op1): Likewise.
12333 (addsi3_carryin_<optab>): Likewise.
12334 (addsi3_carryin_alt2_<optab>): Likewise.
12335 (addsi3_carryin_shift_<optab>): Disable cond_exec variant
12336 for arm_restrict_it.
12337 (subsi3_carryin): Likewise.
12338 (arm_subsi3_insn): Add alternatives for 16-bit encoding.
12339 (minmax_arithsi): Disable for arm_restrict_it.
12340 (minmax_arithsi_non_canon): Adjust for arm_restrict_it.
12341 (satsi_<SAT:code>): Disable cond_exec variant for arm_restrict_it.
12342 (satsi_<SAT:code>_shift): Likewise.
12343 (arm_shiftsi3): Add alternative for 16-bit encoding.
12344 (arm32_movhf): Disable for arm_restrict_it.
12345 (arm_cmpdi_unsigned): Add alternatives for 16-bit encoding.
12346 (arm_movtas_ze): Disable cond_exec variant for arm_restrict_it.
12347
859abddd
SN
123482013-07-22 Sofiane Naci <sofiane.naci@arm.com>
12349
12350 * config/arm/arm.md (attribute "insn"): Delete.
12351 (attribute "type"): Add "mov_imm", "mov_reg", "mov_shift",
12352 "mov_shift_reg", "mvn_imm", "mvn_reg", "mvn_shift" and "mvn_shift_reg".
12353 (not_shiftsi): Update for attribute change.
12354 (not_shiftsi_compare0): Likewise.
12355 (not_shiftsi_compare0_scratch): Likewise.
12356 (arm_one_cmplsi2): Likewise.
12357 (thumb1_one_cmplsi2): Likewise.
12358 (notsi_compare0): Likewise.
12359 (notsi_compare0_scratch): Likewise.
12360 (thumb1_movdi_insn): Likewise.
12361 (arm_movsi_insn): Likewise.
12362 (movhi_insn_arch4): Likewise.
12363 (movhi_bytes): Likewise.
12364 (arm_movqi_insn): Likewise.
12365 (thumb1_movqi_insn): Likewise.
12366 (arm32_movhf): Likewise.
12367 (thumb1_movhf): Likewise.
12368 (arm_movsf_soft_insn): Likewise.
12369 (thumb1_movsf_insn): Likewise.
12370 (thumb_movdf_insn): Likewise.
12371 (movsicc_insn): Likewise.
12372 (movsfcc_soft_insn): Likewise.
12373 (and_scc): Likewise.
12374 (cond_move): Likewise.
12375 (if_move_not): Likewise.
12376 (if_not_move): Likewise.
12377 (if_shift_move): Likewise.
12378 (if_move_shift): Likewise.
12379 (if_shift_shift): Likewise.
12380 (if_not_arith): Likewise.
12381 (if_arith_not): Likewise.
12382 (cond_move_not): Likewise.
12383 * config/arm/neon.md (neon_mov<mode>): Update for attribute change.
12384 (neon_mov<mode>): Likewise.
12385 * config/arm/vfp.md (arm_movsi_vfp): Update for attribute change.
12386 (thumb2_movsi_vfp): Likewise.
12387 (movsf_vfp): Likewise.
12388 (thumb2_movsf_vfp): Likewise.
371e77e3
UB
12389 * config/arm/arm.c (xscale_sched_adjust_cost): Update for attribute
12390 change.
859abddd
SN
12391 (cortexa7_older_only): Likewise.
12392 (cortexa7_younger): Likewise.
12393 * config/arm/arm1020e.md (1020alu_op): Update for attribute change.
12394 (1020alu_shift_op): Likewise.
12395 (1020alu_shift_reg_op): Likewise.
12396 * config/arm/arm1026ejs.md (alu_op): Update for attribute change.
12397 (alu_shift_op): Likewise.
12398 (alu_shift_reg_op): Likewise.
12399 * config/arm/arm1136jfs.md (11_alu_op): Update for attribute change.
12400 (11_alu_shift_op): Likewise.
12401 (11_alu_shift_reg_op): Likewise.
12402 * config/arm/arm926ejs.md (9_alu_op): Update for attribute change.
12403 (9_alu_shift_reg_op): Likewise.
371e77e3
UB
12404 * config/arm/cortex-a15.md (cortex_a15_alu): Update for attribute
12405 change.
859abddd
SN
12406 (cortex_a15_alu_shift): Likewise.
12407 (cortex_a15_alu_shift_reg): Likewise.
12408 * config/arm/cortex-a5.md (cortex_a5_alu): Update for attribute change.
12409 (cortex_a5_alu_shift): Likewise.
371e77e3
UB
12410 * config/arm/cortex-a53.md (cortex_a53_alu): Update for attribute
12411 change.
859abddd 12412 (cortex_a53_alu_shift): Likewise.
371e77e3
UB
12413 * config/arm/cortex-a7.md (cortex_a7_alu_imm): Update for attribute
12414 change.
859abddd
SN
12415 (cortex_a7_alu_reg): Likewise.
12416 (cortex_a7_alu_shift): Likewise.
12417 * config/arm/cortex-a8.md (cortex_a8_alu): Update for attribute change.
12418 (cortex_a8_alu_shift): Likewise.
12419 (cortex_a8_alu_shift_reg): Likewise.
12420 (cortex_a8_mov): Likewise.
12421 * config/arm/cortex-a9.md (cortex_a9_dp): Update for attribute change.
12422 (cortex_a9_dp_shift): Likewise.
12423 * config/arm/cortex-m4.md (cortex_m4_alu): Update for attribute change.
12424 * config/arm/cortex-r4.md (cortex_r4_alu): Update for attribute change.
12425 (cortex_r4_mov): Likewise.
12426 (cortex_r4_alu_shift): Likewise.
12427 (cortex_r4_alu_shift_reg): Likewise.
12428 * config/arm/fa526.md (526_alu_op): Update for attribute change.
12429 (526_alu_shift_op): Likewise.
12430 * config/arm/fa606te.md (606te_alu_op): Update for attribute change.
12431 * config/arm/fa626te.md (626te_alu_op): Update for attribute change.
12432 (626te_alu_shift_op): Likewise.
12433 * config/arm/fa726te.md (726te_shift_op): Update for attribute change.
12434 (726te_alu_op): Likewise.
12435 (726te_alu_shift_op): Likewise.
12436 (726te_alu_shift_reg_op): Likewise.
12437 * config/arm/fmp626.md (mp626_alu_op): Update for attribute change.
12438 (mp626_alu_shift_op): Likewise.
12439 * config/arm/marvell-pj4.md (pj4_alu_e1): Update for attribute change.
12440 (pj4_alu_e1_conds): Likewise.
12441 (pj4_alu): Likewise.
12442 (pj4_alu_conds): Likewise.
12443 (pj4_shift): Likewise.
12444 (pj4_shift_conds): Likewise.
12445 (pj4_alu_shift): Likewise.
12446 (pj4_alu_shift_conds): Likewise.
12447
95b97fac
KT
124482013-07-22 Kyrylo Tkachov <kyrylo.tkachov@arm.com>
12449
12450 * config/arm/predicates.md (shiftable_operator_strict_it):
12451 New predicate.
12452 * config/arm/thumb2.md (thumb_andsi_not_shiftsi_si):
12453 Disable cond_exec version for arm_restrict_it.
12454 (thumb2_smaxsi3): Convert to generate cond_exec.
12455 (thumb2_sminsi3): Likewise.
12456 (thumb32_umaxsi3): Likewise.
12457 (thumb2_uminsi3): Likewise.
12458 (thumb2_abssi2): Adjust constraints for arm_restrict_it.
12459 (thumb2_neg_abssi2): Likewise.
12460 (thumb2_mov_scc): Add alternative for 16-bit encoding.
12461 (thumb2_movsicc_insn): Adjust alternatives.
12462 (thumb2_mov_negscc): Disable for arm_restrict_it.
12463 (thumb2_mov_negscc_strict_it): New pattern.
12464 (thumb2_mov_notscc_strict_it): New pattern.
12465 (thumb2_mov_notscc): Disable for arm_restrict_it.
12466 (thumb2_ior_scc): Likewise.
12467 (thumb2_ior_scc_strict_it): New pattern.
12468 (thumb2_cond_move): Adjust for arm_restrict_it.
12469 (thumb2_cond_arith): Disable for arm_restrict_it.
12470 (thumb2_cond_arith_strict_it): New pattern.
12471 (thumb2_cond_sub): Adjust for arm_restrict_it.
12472 (thumb2_movcond): Likewise.
12473 (thumb2_extendqisi_v6): Disable cond_exec variant for arm_restrict_it.
12474 (thumb2_zero_extendhisi2_v6): Likewise.
12475 (thumb2_zero_extendqisi2_v6): Likewise.
12476 (orsi_notsi_si): Likewise.
12477 (orsi_not_shiftsi_si): Likewise.
12478
d5a6ef82
GJL
124792013-07-22 Georg-Johann Lay <avr@gjlay.de>
12480
12481 * config/avr/avr.c (avr_out_xload): No SBIS around LPM so that
12482 instruction sequence is 1 byte shorter.
12483
d9e71154
UB
124842013-07-22 Uros Bizjak <ubizjak@gmail.com>
12485
12486 * config/i386/i386.md (nonlocal_goto_receiver): Delete insn if
12487 it is not needed after split.
12488
8c6381ec
IS
124892013-07-20 Iain Sandoe <iain@codesourcery.com>
12490
12491 PR target/51784
12492 * config/i386/i386.c (output_set_got) [TARGET_MACHO]: Adjust to emit a
12493 second label for nonlocal goto receivers. Don't output pic base labels
12494 unless we're producing PIC; mark that action unreachable().
12495 (ix86_save_reg): If the function contains a nonlocal label, save the
12496 PIC base reg.
12497 * config/darwin-protos.h (machopic_should_output_picbase_label): New.
f43245d1 12498 * config/darwin.c (emitted_pic_label_num): New GTY.
8c6381ec
IS
12499 (update_pic_label_number_if_needed): New.
12500 (machopic_output_function_base_name): Adjust for nonlocal receiver
12501 case.
12502 (machopic_should_output_picbase_label): New.
12503 * config/i386/i386.md (enum unspecv): UNSPECV_NLGR: New.
12504 (nonlocal_goto_receiver): New insn and split.
12505
d05d0709
JG
125062013-07-20 James Greenhalgh <james.greenhalgh@arm.com>
12507
12508 * config/aarch64/aarch64-builtins.c
12509 (aarch64_fold_builtin): Fold abs in all modes.
12510 * config/aarch64/aarch64-simd-builtins.def
12511 (abs): Enable for all modes.
12512 * config/aarch64/arm_neon.h
12513 (vabs<q>_s<8,16,32,64): Rewrite using builtins.
12514 (vabs_f64): Add missing intrinsic.
12515
ba590e63
IB
125162013-07-19 Ian Bolton <ian.bolton@arm.com>
12517
12518 * config/aarch64/arm_neon.h (vabs_s64): New function
12519
02371798
GJL
125202013-07-19 Georg-Johann Lay <avr@gjlay.de>
12521
12522 PR target/57516
12523 * config/avr/avr-fixed.md (round<mode>3_const): Turn expander to insn.
12524 * config/avr/avr.md (adjust_len): Add `round'.
12525 * config/avr/avr-protos.h (avr_out_round): New prototype.
12526 (avr_out_plus): Add `out_label' argument.
12527 * config/avr/avr.c (avr_out_plus_1): Add `out_label' argument.
12528 (avr_out_plus): Pass down `out_label' to avr_out_plus_1.
12529 Handle the case where `insn' is just a pattern.
12530 (avr_out_bitop): Handle the case where `insn' is just a pattern.
12531 (avr_out_round): New function.
12532 (avr_adjust_insn_length): Handle ADJUST_LEN_ROUND.
12533
12211b99 125342013-07-18 David Holsgrove <david.holsgrove@xilinx.com>
b950ac25
DH
12535
12536 * config/microblaze/microblaze.c (microblaze_expand_prologue):
12537 Add check for flag_stack_usage to handle -fstack-usage support
12538
12211b99 125392013-07-18 Pat Haugen <pthaugen@us.ibm.com>
ed1fd999
PH
12540
12541 * config/rs6000/rs6000.c (rs6000_option_override_internal): Adjust flag
12542 interaction for new Power8 flags and VSX.
12543
df9dda2d
ST
125442013-07-18 Sriraman Tallam <tmsriram@google.com>
12545
12546 PR middle-end/57698
12547 * tree-inline.c (expand_call_inline): Emit errors during
12548 early_inlining only if optimization is not turned on.
12549
0f9cdef4
DM
125502013-07-18 David Malcolm <dmalcolm@redhat.com>
12551
12552 * passes.def: New.
12553
12554 * passes.c (init_optimization_passes): Move the construction of
12555 the pass hierarchy into a new passes.def file.
12556
12557 * Makefile.in (passes.o): Add dependency on passes.def.
12558
2efa4087
DM
125592013-07-18 David Malcolm <dmalcolm@redhat.com>
12560
12561 * passes.c (init_optimization_passes): Introduce macros for
12562 constructing the tree of passes (INSERT_PASSES_AFTER,
12563 PUSH_INSERT_PASSES_WITHIN, POP_INSERT_PASSES,
12564 TERMINATE_PASS_LIST).
12565
bc404e1b
VM
125662013-07-18 Vladimir Makarov <vmakarov@redhat.com>
12567 Wei Mi <wmi@google.com>
12568
12569 PR rtl-optimization/57878
12570 * lra-assigns.c (assign_by_spills): Move non_reload_pseudos to the
12571 top.
12572 (reload_pseudo_compare_func): Check nregs first for reload
12573 pseudos.
12574
d675843f
DM
125752013-07-18 David Malcolm <dmalcolm@redhat.com>
12576
12577 * tree-pass.h (pass_ipa_lto_wpa_fixup): Remove redundant decl.
12578
c7b3b99f 125792013-07-18 Po-Chun Chang <pchang9@cs.wisc.edu>
bcb97fbd
PCC
12580
12581 * read-rtl.c (validate_const_int): Once an invalid character is
12582 seen, quit the loop.
12583
12584 * gengtype.c (write_roots): Similarly once we find the "deletable"
12585 or "if_marked" option.
12586
52fceb44
SN
125872013-07-18 Sofiane Naci <sofiane.naci@arm.com>
12588
12589 * config/arm/arm.md (attribute "insn"): Delete values "mrs", "msr",
12590 "xtab" and "sat". Move value "clz" from here to ...
12591 (attriubte "type"): ... here.
12592 (satsi_<SAT:code>): Delete "insn" attribute.
12593 (satsi_<SAT:code>_shift): Likewise.
12594 (arm_zero_extendqisi2addsi): Likewise.
12595 (arm_extendqisi2addsi): Likewise.
12596 (clzsi2): Update for attribute changes.
12597 (rbitsi2): Likewise.
371e77e3
UB
12598 * config/arm/arm-fixed.md (arm_ssatsihi_shift): Delete "insn"
12599 attribute.
52fceb44
SN
12600 (arm_usatsihi): Likewise.
12601 * config/arm/cortex-a8.md (cortex_a8_alu): Update for attribute change.
12602
006bd006
SN
126032013-07-18 Sofiane Naci <sofiane.naci@arm.com>
12604
12605 * config/arm/arm.md (attribute "type"): Rename "simple_alu_imm" to
371e77e3
UB
12606 "arlo_imm". Rename "alu_reg" to "arlo_reg". Rename "simple_alu_shift"
12607 to "extend". Split "alu_shift" into "shift" and "arlo_shift". Split
006bd006
SN
12608 "alu_shift_reg" into "shift_reg" and "arlo_shift_reg". List types
12609 in alphabetical order.
12610 (attribute "core_cycles"): Update for attribute changes.
12611 (arm_addsi3): Likewise.
12612 (addsi3_compare0): Likewise.
12613 (addsi3_compare0_scratch): Likewise.
12614 (addsi3_compare_op1): Likewise.
12615 (addsi3_compare_op2): Likewise.
12616 (compare_addsi2_op0): Likewise.
12617 (compare_addsi2_op1): Likewise.
12618 (addsi3_carryin_shift_<optab>): Likewise.
12619 (subsi3_carryin_shift): Likewise.
12620 (rsbsi3_carryin_shift): Likewise.
12621 (arm_subsi3_insn): Likewise.
12622 (subsi3_compare0): Likewise.
12623 (subsi3_compare): Likewise.
12624 (arm_andsi3_insn): Likewise.
12625 (thumb1_andsi3_insn): Likewise.
12626 (andsi3_compare0): Likewise.
12627 (andsi3_compare0_scratch): Likewise.
12628 (zeroextractsi_compare0_scratch
12629 (andsi_not_shiftsi_si): Likewise.
12630 (iorsi3_insn): Likewise.
12631 (iorsi3_compare0): Likewise.
12632 (iorsi3_compare0_scratch): Likewise.
12633 (arm_xorsi3): Likewise.
12634 (thumb1_xorsi3_insn): Likewise.
12635 (xorsi3_compare0): Likewise.
12636 (xorsi3_compare0_scratch): Likewise.
12637 (satsi_<SAT:code>_shift): Likewise.
12638 (rrx): Likewise.
12639 (arm_shiftsi3): Likewise.
12640 (shiftsi3_compare0): Likewise.
12641 (not_shiftsi): Likewise.
12642 (not_shiftsi_compare0): Likewise.
12643 (not_shiftsi_compare0_scratch): Likewise.
12644 (arm_one_cmplsi2): Likewise.
12645 (thumb_one_complsi2): Likewise.
12646 (notsi_compare0): Likewise.
12647 (notsi_compare0_scratch): Likewise.
12648 (thumb1_zero_extendhisi2): Likewise.
12649 (arm_zero_extendhisi2): Likewise.
12650 (arm_zero_extendhisi2_v6): Likewise.
12651 (arm_zero_extendhisi2addsi): Likewise.
12652 (thumb1_zero_extendqisi2): Likewise.
12653 (thumb1_zero_extendqisi2_v6): Likewise.
12654 (arm_zero_extendqisi2): Likewise.
12655 (arm_zero_extendqisi2_v6): Likewise.
12656 (arm_zero_extendqisi2addsi): Likewise.
12657 (thumb1_extendhisi2): Likewise.
12658 (arm_extendhisi2): Likewise.
12659 (arm_extendhisi2_v6): Likewise.
12660 (arm_extendqisi): Likewise.
12661 (arm_extendqisi_v6): Likewise.
12662 (arm_extendqisi2addsi): Likewise.
12663 (thumb1_extendqisi2): Likewise.
12664 (thumb1_movdi_insn): Likewise.
12665 (arm_movsi_insn): Likewise.
12666 (movsi_compare0): Likewise.
12667 (movhi_insn_arch4): Likewise.
12668 (movhi_bytes): Likewise.
12669 (arm_movqi_insn): Likewise.
12670 (thumb1_movqi_insn): Likewise.
12671 (arm32_movhf): Likewise.
12672 (thumb1_movhf): Likewise.
12673 (arm_movsf_soft_insn): Likewise.
12674 (thumb1_movsf_insn): Likewise.
12675 (movdf_soft_insn): Likewise.
12676 (thumb_movdf_insn): Likewise.
12677 (arm_cmpsi_insn): Likewise.
12678 (cmpsi_shiftsi): Likewise.
12679 (cmpsi_shiftsi_swp): Likewise.
12680 (arm_cmpsi_negshiftsi_si): Likewise.
12681 (movsicc_insn): Likewise.
12682 (movsfcc_soft_insn): Likewise.
12683 (arith_shiftsi): Likewise.
12684 (arith_shiftsi_compare0
12685 (arith_shiftsi_compare0_scratch
12686 (sub_shiftsi): Likewise.
12687 (sub_shiftsi_compare0
12688 (sub_shiftsi_compare0_scratch
12689 (and_scc): Likewise.
12690 (cond_move): Likewise.
12691 (if_plus_move): Likewise.
12692 (if_move_plus): Likewise.
12693 (if_move_not): Likewise.
12694 (if_not_move): Likewise.
12695 (if_shift_move): Likewise.
12696 (if_move_shift): Likewise.
12697 (if_shift_shift): Likewise.
12698 (if_not_arith): Likewise.
12699 (if_arith_not): Likewise.
12700 (cond_move_not): Likewise.
12701 (thumb1_ashlsi3): Set type attribute.
12702 (thumb1_ashrsi3): Likewise.
12703 (thumb1_lshrsi3): Likewise.
12704 (thumb1_rotrsi3): Likewise.
12705 (shiftsi3_compare0_scratch): Likewise.
12706 * config/arm/neon.md (neon_mov<mode>): Update for attribute changes.
12707 (neon_mov<mode>): Likewise.
371e77e3
UB
12708 * config/arm/thumb2.md (thumb_andsi_not_shiftsi_si): Update for
12709 attribute changes.
006bd006
SN
12710 (thumb2_movsi_insn): Likewise.
12711 (thumb2_cmpsi_neg_shiftsi): Likewise.
12712 (thumb2_extendqisi_v6): Likewise.
12713 (thumb2_zero_extendhisi2_v6): Likewise.
12714 (thumb2_zero_extendqisi2_v6): Likewise.
12715 (thumb2_shiftsi3_short): Likewise.
12716 (thumb2_addsi3_compare0_scratch): Likewise.
12717 (orsi_not_shiftsi_si): Likewise.
12718 * config/arm/vfp.md (arm_movsi_vfp): Update for attribute changes.
12719 * config/arm/arm-fixed.md (arm_ssatsihi_shift): Update for attribute
12720 changes.
12721 * config/arm/arm1020e.md (1020alu_op): Update for attribute changes.
12722 (1020alu_shift_op): Likewise.
12723 (1020alu_shift_reg_op): Likewise.
12724 * config/arm/arm1026ejs.md (alu_op): Update for attribute changes.
12725 (alu_shift_op): Likewise.
12726 (alu_shift_reg_op): Likewise.
12727 * config/arm/arm1136jfs.md (11_alu_op): Update for attribute changes.
12728 (11_alu_shift_op): Likewise.
12729 (11_alu_shift_reg_op): Likewise.
12730 * config/arm/arm926ejs.md (9_alu_op): Update for attribute changes.
12731 (9_alu_shift_reg_op): Likewise.
371e77e3
UB
12732 * config/arm/cortex-a15.md (cortex_a15_alu): Update for attribute
12733 changes.
006bd006
SN
12734 (cortex_a15_alu_shift): Likewise.
12735 (cortex_a15_alu_shift_reg): Likewise.
371e77e3
UB
12736 * config/arm/cortex-a5.md (cortex_a5_alu): Update for attribute
12737 changes.
006bd006
SN
12738 (cortex_a5_alu_shift): Likewise.
12739 * config/arm/cortex-a53.md (cortex_a53_alu) : Update for attribute
12740 changes.
12741 (cortex_a53_alu_shift): Likewise.
12742 * config/arm/cortex-a7.md (cortex_a7_alu_imm): Update for attribute
12743 changes.
12744 (cortex_a7_alu_reg): Likewise.
12745 (cortex_a7_alu_shift): Likewise.
371e77e3
UB
12746 * config/arm/cortex-a8.md (cortex_a8_alu): Update for attribute
12747 changes.
006bd006
SN
12748 (cortex_a8_alu_shift): Likewise.
12749 (cortex_a8_alu_shift_reg): Likewise.
12750 (cortex_a8_mov): Likewise.
12751 * config/arm/cortex-a9.md (cortex_a9_dp): Update for attribute changes.
12752 (cortex_a9_dp_shift): Likewise.
371e77e3
UB
12753 * config/arm/cortex-m4.md (cortex_m4_alu): Update for attribute
12754 changes.
12755 * config/arm/cortex-r4.md (cortex_r4_alu): Update for attribute
12756 changes.
006bd006
SN
12757 (cortex_r4_mov): Likewise.
12758 (cortex_r4_alu_shift): Likewise.
12759 (cortex_r4_alu_shift_reg): Likewise.
12760 * config/arm/fa526.md (526_alu_op): Update for attribute changes.
12761 (526_alu_shift_op): Likewise.
12762 * config/arm/fa606te.md (606te_alu_op): Update for attribute changes.
12763 * config/arm/fa626te.md (626te_alu_op): Update for attribute changes.
12764 (626te_alu_shift_op): Likewise.
12765 * config/arm/fa726te.md (726te_shift_op): Update for attribute changes.
12766 (726te_alu_op): Likewise.
12767 (726te_alu_shift_op): Likewise.
12768 (726te_alu_shift_reg_op): Likewise.
12769 * config/arm/fmp626.md (mp626_alu_op): Update for attribute changes.
12770 (mp626_alu_shift_op): Likewise.
12771 * config/arm/marvell-pj4.md (pj4_alu_e1): Update for attribute changes.
12772 (pj4_alu_e1_conds): Likewise.
12773 (pj4_alu): Likewise.
12774 (pj4_alu_conds): Likewise.
12775 (pj4_shift): Likewise.
12776 (pj4_shift_conds): Likewise.
12777 (pj4_alu_shift): Likewise.
12778 (pj4_alu_shift_conds): Likewise.
371e77e3
UB
12779 * config/arm/arm.c (xscale_sched_adjust_cost): Update for attribute
12780 changes.
006bd006
SN
12781 (cortexa7_older_only): Likewise.
12782 (cortexa7_younger): Likewise.
12783
651df1b2
DM
127842013-07-18 David Malcolm <dmalcolm@redhat.com>
12785
12786 * ipa-pure-const.c (generate_summary): Rename to...
12787 (pure_const_generate_summary): ... this.
12788
ea717bd7
IS
127892013-07-17 Iain Sandoe <iain@codesourcery.com>
12790
12791 * config/rs6000/darwin.h (REGISTER_NAMES): Add HTM registers.
12792
a1259a13
YR
127932013-07-17 Yvan Roux <yvan.roux@linaro.org>
12794
12795 PR target/57909
12796 * config/arm/arm.c (gen_movmem_ldrd_strd): Fix unaligned load/store
12797 usage in HI mode.
12798
167f68ed
AK
127992013-07-17 Andreas Krebbel <Andreas.Krebbel@de.ibm.com>
12800
12801 * config/s390/s390.c: (s390_expand_builtin): Allow -mhtm to be
12802 enabled without -march=zEC12.
12803 * config/s390/s390.h (TARGET_HTM): Do not require EC12 machine
12804 flags to be set.
12805
f900a982
MR
128062013-07-16 Maciej W. Rozycki <macro@codesourcery.com>
12807
12808 * config/mips/mips.h (ISA_HAS_FP4): Correct formatting.
12809 (ISA_HAS_FP_MADD4_MSUB4): Also enable for ISA_MIPS32R2.
12810 (ISA_HAS_NMADD4_NMSUB4): Remove the MODE argument; rewrite in
12811 terms of ISA_HAS_FP4, and also enable for ISA_MIPS32R2.
12812 (ISA_HAS_NMADD3_NMSUB3): Remove the MODE argument.
12813 * config/mips/mips.c (mips_rtx_costs) <PLUS>: Check for
12814 ISA_HAS_FP_MADD4_MSUB4 || ISA_HAS_FP_MADD3_MSUB3 rather than
12815 ISA_HAS_FP4.
12816 <MINUS, NEG>: Update according to changes to ISA_HAS_NMADD4_NMSUB4
12817 and ISA_HAS_NMADD3_NMSUB3.
12818 * config/mips/mips.md (nmadd4<mode>, nmadd3<mode>): Likewise.
12819 (nmadd4<mode>_fastmath, nmadd3<mode>_fastmath): Likewise.
12820 (nmsub4<mode>, nmsub3<mode>): Likewise.
12821 (nmsub4<mode>_fastmath, nmsub3<mode>_fastmath): Likewise.
12822
d4ed27eb
MR
128232013-07-16 Maciej W. Rozycki <macro@codesourcery.com>
12824
12825 * config/mips/mips.h (ISA_HAS_NMADD4_NMSUB4): Remove
12826 TARGET_MIPS5400 checking.
12827
7a5add18
PB
128282013-07-16 Jakub Jelinek <jakub@redhat.com>
12829 Peter Bergner <bergner@vnet.ibm.com>
12830
12831 * config/rs6000/rs6000.h (FIRST_PSEUDO_REGISTERS): Mention HTM
12832 registers in the comment.
12833 (DWARF_FRAME_REGISTERS): Subtract also the 3 HTM registers.
12834 (DWARF_REG_TO_UNWIND_COLUMN): Use DWARF_FRAME_REGISTERS
12835 rather than FIRST_PSEUDO_REGISTERS.
12836
128372013-07-16 Peter Bergner <bergner@vnet.ibm.com>
c694dc01
PB
12838
12839 * config/rs6000/rs6000.c (rs6000_option_override_internal): Do not
12840 enable extra ISA flags with TARGET_HTM.
12841
d2ab0929
MR
128422013-07-16 Maciej W. Rozycki <macro@codesourcery.com>
12843
12844 * config/mips/mips.h (ISA_HAS_MULS, ISA_HAS_MSAC, ISA_HAS_MACC):
12845 Fix comment typos.
12846
839c74bc
CH
128472013-07-15 Cong Hou <congh@google.com>
12848
12849 * tree-vect-data-refs.c (dr_group_sort_cmp): Do not use hash function
12850 in compare function for sorting.
12851
7a5add18 128522013-07-15 Peter Bergner <bergner@vnet.ibm.com>
0258b6e4
PB
12853
12854 * config.gcc (powerpc*-*-*): Install htmintrin.h and htmxlintrin.h.
12855 * config/rs6000/t-rs6000 (MD_INCLUDES): Add htm.md.
12856 * config/rs6000/rs6000.opt: Add -mhtm option.
12857 * config/rs6000/rs6000-cpus.def (POWERPC_MASKS): Add OPTION_MASK_HTM.
12858 (ISA_2_7_MASKS_SERVER): Add OPTION_MASK_HTM.
12859 * config/rs6000/rs6000-c.c (rs6000_target_modify_macros): Define
12860 __HTM__ if the HTM instructions are available.
12861 * config/rs6000/predicates.md (u3bit_cint_operand, u10bit_cint_operand,
12862 htm_spr_reg_operand): New define_predicates.
12863 * config/rs6000/rs6000.md (define_attr "type"): Add htm.
12864 (TFHAR_REGNO, TFIAR_REGNO, TEXASR_REGNO): New define_constants.
12865 Include htm.md.
12866 * config/rs6000/rs6000-builtin.def (BU_HTM_0, BU_HTM_1, BU_HTM_2,
12867 BU_HTM_3, BU_HTM_SPR0, BU_HTM_SPR1): Add support macros for defining
12868 HTM builtin functions.
12869 * config/rs6000/rs6000.c (RS6000_BUILTIN_H): New macro.
12870 (rs6000_reg_names, alt_reg_names): Add HTM SPR register names.
12871 (rs6000_init_hard_regno_mode_ok): Add support for HTM instructions.
12872 (rs6000_builtin_mask_calculate): Likewise.
12873 (rs6000_option_override_internal): Likewise.
12874 (bdesc_htm): Add new HTM builtin support.
12875 (htm_spr_num): New function.
12876 (htm_spr_regno): Likewise.
12877 (rs6000_htm_spr_icode): Likewise.
12878 (htm_expand_builtin): Likewise.
12879 (htm_init_builtins): Likewise.
12880 (rs6000_expand_builtin): Add support for HTM builtin functions.
12881 (rs6000_init_builtins): Likewise.
371e77e3
UB
12882 (rs6000_invalid_builtin, rs6000_opt_mask): Add support for -mhtm
12883 option.
0258b6e4
PB
12884 * config/rs6000/rs6000.h (ASM_CPU_SPEC): Add support for -mhtm.
12885 (TARGET_HTM, MASK_HTM): Define macros.
12886 (FIRST_PSEUDO_REGISTER): Adjust for new HTM SPR registers.
12887 (FIXED_REGISTERS): Likewise.
12888 (CALL_USED_REGISTERS): Likewise.
12889 (CALL_REALLY_USED_REGISTERS): Likewise.
12890 (REG_ALLOC_ORDER): Likewise.
12891 (enum reg_class): Likewise.
12892 (REG_CLASS_NAMES): Likewise.
12893 (REG_CLASS_CONTENTS): Likewise.
12894 (REGISTER_NAMES): Likewise.
12895 (ADDITIONAL_REGISTER_NAMES): Likewise.
12896 (RS6000_BTC_SPR, RS6000_BTC_VOID, RS6000_BTC_32BIT, RS6000_BTC_64BIT,
12897 RS6000_BTC_MISC_MASK, RS6000_BTM_HTM): New macros.
12898 (RS6000_BTM_COMMON): Add RS6000_BTM_HTM.
12899 * config/rs6000/htm.md: New file.
12900 * config/rs6000/htmintrin.h: New file.
12901 * config/rs6000/htmxlintrin.h: New file.
12902
87dd8ab0
MS
129032013-07-15 Marcus Shawcroft <marcus.shawcroft@arm.com>
12904
12905 * config/aarch64/aarch64-protos.h (aarch64_symbol_type):
12906 Define SYMBOL_TINY_GOT, update comment.
12907 * config/aarch64/aarch64.c
12908 (aarch64_load_symref_appropriately): Handle SYMBOL_TINY_GOT.
12909 (aarch64_expand_mov_immediate): Likewise.
12910 (aarch64_print_operand): Likewise.
12911 (aarch64_classify_symbol): Likewise.
12912 * config/aarch64/aarch64.md (UNSPEC_GOTTINYPIC): Define.
12913 (ldr_got_tiny): Define.
12914
52d676b6
TG
129152013-07-13 Tobias Grosser <tobias@grosser.es>
12916
12917 PR tree-optimization/54094
12918 * graphite-clast-to-gimple.c (translate_clast_for_loop): Derive the
12919 scheduling dimension for the parallelism check from the polyhedral
12920 information in the AST.
12921 * graphite-dependences.c (carries_deps): Do not assume the schedule is
12922 in 2D + 1 form.
12923
286e8fc1
JM
129242013-07-13 Jason Merrill <jason@redhat.com>
12925
12926 * print-tree.c (debug_vec_tree): Use debug_raw.
12927 (debug_raw (vec<tree, va_gc> &)): New.
12928 (debug_raw (vec<tree, va_gc> *)): New.
12929 * tree.h: Declare them.
12930
e43257e8
BC
129312013-07-13 Bin Cheng <bin.cheng@arm.com>
12932
12933 * ifcvt.c (ifcvt_after_combine): New static variable.
12934 (cheap_bb_rtx_cost_p): Set scale to REG_BR_PROB_BASE when optimizing
12935 for size.
12936 (if_convert): New parameter after_combine. Set ifcvt_after_combine.
12937 (rest_of_handle_if_conversion, rest_of_handle_if_after_combine,
12938 rest_of_handle_if_after_reload): Pass new argument for if_convert.
12939
da582d46
MR
129402013-07-12 Maciej W. Rozycki <macro@codesourcery.com>
12941
12942 * config/mips/mips.c (mips_expand_call): Remove empty statement.
12943
18b0ea8f
MM
129442013-07-12 Michael Matz <matz@suse.de>
12945
12946 PR middle-end/55771
12947 * convert.c (convert_to_real): Reject non-float inner types.
12948
e4f0f84d
TB
129492013-07-12 Tejas Belagod <tejas.belagod@arm.com>
12950
12951 * config/aarch64/aarch64-protos.h
12952 (aarch64_simd_immediate_valid_for_move): Remove.
12953 * config/aarch64/aarch64.c (simd_immediate_info): New member.
12954 (aarch64_simd_valid_immediate): Recognize idioms for shifting ones
12955 cases.
12956 (aarch64_output_simd_mov_immediate): Print the correct shift specifier.
12957
76500022
SE
129582013-07-11 Steve Ellcey <sellcey@mips.com>
12959
12960 * config/mips/mips.c (mips_conditional_register_usage): Do not
12961 use t[0-7] registers in MIPS16 mode when optimizing for size.
12962
78039734
ST
129632013-07-11 Sriraman Tallam <tmsriram@google.com>
12964
12965 * config/i386/i386.c (dispatch_function_versions): Fix array
12966 indexing of function_version_info to match actual_versions.
12967
7b24b675
TJ
129682013-07-11 Teresa Johnson <tejohnson@google.com>
12969
12970 * vec.h (struct va_gc): Move release out-of-line.
12971 (va_gc::release): Call ggc_free on released vec.
12972
85d4cbb8
UW
129732013-07-11 Ulrich Weigand <Ulrich.Weigand@de.ibm.com>
12974
12975 * config/rs6000/rs6000.md (""*tls_gd_low<TLSmode:tls_abi_suffix>"):
12976 Require GOT register as additional operand in UNSPEC.
12977 ("*tls_ld_low<TLSmode:tls_abi_suffix>"): Likewise.
12978 ("*tls_got_dtprel_low<TLSmode:tls_abi_suffix>"): Likewise.
12979 ("*tls_got_tprel_low<TLSmode:tls_abi_suffix>"): Likewise.
12980 ("*tls_gd<TLSmode:tls_abi_suffix>"): Update splitter.
12981 ("*tls_ld<TLSmode:tls_abi_suffix>"): Likewise.
12982 ("tls_got_dtprel_<TLSmode:tls_abi_suffix>"): Likewise.
12983 ("tls_got_tprel_<TLSmode:tls_abi_suffix>"): Likewise.
12984
52b3f9e2
GJL
129852013-07-11 Georg-Johann Lay <avr@gjlay.de>
12986
12987 PR target/57631
12988 * config/avr/avr.c (avr_set_current_function): Sanity-check signal
12989 name seen by assembler/linker rather if available.
12990
3f3e266e
AS
129912013-07-11 Andreas Schwab <schwab@suse.de>
12992
12993 * config/aarch64/aarch64-linux.h (CPP_SPEC): Define.
12994
21231ca6
VM
129952013-07-10 Vladimir Makarov <vmakarov@redhat.com>
12996
371e77e3 12997 * lra-constraints.c (curr_insn_transform): Switch off optional reloads.
21231ca6 12998
d331d374
JM
129992013-07-10 Joseph Myers <joseph@codesourcery.com>
13000
673c2f63
JM
13001 * doc/tm.texi.in: Move hook documentation to ....
13002 * target.def: ... here.
13003
d331d374
JM
13004 * doc/tm.texi.in (TARGET_CANONICALIZE_COMPARISON): Remove stray
13005 text on @hook line.
13006 * doc/tm.texi: Regenerate.
13007
dfeadaa0
PC
130082013-07-10 Paolo Carlini <paolo.carlini@oracle.com>
13009
13010 PR c++/57869
13011 * doc/invoke.texi: Document -Wconditionally-supported.
13012
bb6f2bac
GJL
130132013-07-10 Georg-Johann Lay <avr@gjlay.de>
13014
13015 PR target/57844
13016 * config/avr/avr.c (avr_prologue_setup_frame): Trunk -size to mode
13017 of my_fp.
13018
5219b232
GJL
130192013-07-10 Georg-Johann Lay <avr@gjlay.de>
13020
13021 PR target/57506
13022 * config/avr/avr-mcus.def (atmega16hva, atmega16hva2, atmega16hvb)
13023 (atmega16m1, atmega16u4, atmega32a, atmega32c1, atmega32hvb)
13024 (atmega32m1, atmega32u4, atmega32u6, atmega64c1, atmega64m1):
13025 Remove duplicate devices.
13026 * config/avr/gen-avr-mmcu-texi.c (print_mcus): Fail on duplicate MCUs.
13027 * config/avr/t-multilib: Regenerate.
13028 * config/avr/avr-tables.opt: Regenerate.
13029 * doc/avr-mmcu.texi: Regenerate.
13030
283045c0
GJL
130312013-07-10 Georg-Johann Lay <avr@gjlay.de>
13032
13033 PR target/56987
13034 * config/avr/avr.opt (Waddr-space-convert): Fix typo.
13035
ceb2d59f 130362013-07-10 Graham Stott <graham.stott@btinternet.com>
6a2fa4b2 13037
ceb2d59f
GS
13038 * config/mips/mips.c (mips_rtx_costs): Very slightly increase
13039 the cost of MULT when optimizing for size.
13040
055e0a99
JBG
130412013-07-10 Jan-Benedict Glaw <jbglaw@lug-owl.de>
13042
13043 * config/cr16/cr16-protos.h: Don't include target.h.
13044
ef43e868
JM
130452013-07-09 Joseph Myers <joseph@codesourcery.com>
13046
13047 * config/rs6000/rs6000.c (rs6000_init_hard_regno_mode_ok): Only
13048 adjust register size for TDmode and TFmode for VSX registers.
13049
4a283fdf
KT
130502013-07-08 Kai Tietz <ktietz@redhat.com>
13051
13052 PR target/56892
13053 * config/i386/i386.c (TARGET_FUNCTION_ATTRIBUTE_INLINABLE_P): Define as
13054 hook_bool_const_tree_true.
13055
2cf4c39e
AK
130562013-07-08 Andreas Krebbel <Andreas.Krebbel@de.ibm.com>
13057
13058 * config/s390/s390.c: Replace F*_REGNUM with FPR*_REGNUM.
13059 * config/s390/s390.h: Remove F*_REGNUM macro definitions.
13060 * config/s390/s390.md: Define FPR*_REGNUM constants.
13061 Fix FPR2_REGNUM constant (18 -> 17).
13062 ("*trunc<BFP:mode><DFP_ALL:mode>2")
13063 ("*trunc<DFP_ALL:mode><BFP:mode>2")
13064 ("trunc<BFP:mode><DFP_ALL:mode>2")
13065 ("trunc<DFP_ALL:mode><BFP:mode>2")
13066 ("*extend<BFP:mode><DFP_ALL:mode>2")
13067 ("*extend<DFP_ALL:mode><BFP:mode>2")
13068 ("extend<BFP:mode><DFP_ALL:mode>2")
13069 ("extend<DFP_ALL:mode><BFP:mode>2"): Replace FPR2_REGNUM with
13070 FPR4_REGNUM.
13071
e6ac0270 130722013-07-08 Graham Stott <graham.stott@btinternet.com>
e6ac0270 13073
6a2fa4b2 13074 * Makefile.in: (c-family-warn): Define to $(STRICT_WARN)
64a92610 13075
b89b22fc
AK
130762013-07-08 Andreas Krebbel <Andreas.Krebbel@de.ibm.com>
13077
13078 * config/s390/s390.c: Rename cfun_set_fpr_bit to cfun_set_fpr_save
13079 and cfun_fpr_bit_p to cfun_fpr_save_p.
13080 (s390_frame_area, s390_register_info, s390_frame_info)
13081 (s390_emit_prologue, s390_emit_epilogue)
13082 (s390_conditional_register_usage): Use the *_REGNUM macros for FPR
13083 register numbers.
13084 * config/s390/s390.h: Define *_REGNUM macros for floating point
13085 register numbers.
13086
07f398aa
EB
130872013-07-08 Eric Botcazou <ebotcazou@adacore.com>
13088
13089 * Makefile.in (tree-ssa-reassoc.o): Add dependency on $(PARAMS_H).
13090
75fefa91
PCC
130912013-07-08 Po-Chun Chang <pchang9@cs.wisc.edu>
13092
13093 PR rtl-optimization/57786
13094 * combine.c (distribute_notes) <case REG_DEAD>: Change all_used to bool
13095 and break out of the loop when it is set to false.
13096
046f1eee
JJ
130972013-07-08 Jakub Jelinek <jakub@redhat.com>
13098
8140c065
JJ
13099 PR target/57819
13100 * simplify-rtx.c (simplify_unary_operation_1) <case ZERO_EXTEND>:
13101 Simplify (zero_extend:SI (subreg:QI (and:SI (reg:SI)
13102 (const_int 63)) 0)).
13103 * combine.c (make_extraction): Create ZERO_EXTEND or SIGN_EXTEND
13104 using simplify_gen_unary instead of gen_rtx_*_EXTEND.
13105 * config/i386/i386.md (*jcc_bt<mode>_1): New define_insn_and_split.
13106
046f1eee
JJ
13107 PR rtl-optimization/57829
13108 * simplify-rtx.c (simplify_binary_operation_1) <case IOR>: Ensure that
13109 mask bits outside of mode are just sign-extension from mode to HWI.
13110
930b700b
MZ
131112013-07-08 Michael Zolotukhin <michael.v.zolotukhin@gmail.com>
13112
13113 * config/i386/i386-opts.h (enum stringop_alg): Add vector_loop.
13114 * config/i386/i386.c (expand_set_or_movmem_via_loop): Use
13115 adjust_address instead of change_address to keep info about alignment.
13116 (emit_strmov): Remove.
13117 (emit_memmov): New function.
13118 (expand_movmem_epilogue): Refactor to properly handle bigger sizes.
13119 (expand_movmem_epilogue): Likewise and return updated rtx for
13120 destination.
13121 (expand_constant_movmem_prologue): Likewise and return updated rtx for
13122 destination and source.
13123 (decide_alignment): Refactor, handle vector_loop.
13124 (ix86_expand_movmem): Likewise.
13125 (ix86_expand_setmem): Likewise.
13126 * config/i386/i386.opt (Enum): Add vector_loop to option stringop_alg.
13127
7b9d1bd8
UB
131282013-07-07 Uros Bizjak <ubizjak@gmail.com>
13129
13130 * config/i386/driver-i386.c (host_detect_local_cpu): Do not check
13131 signature_TM2_ebx, it interferes with signature_INTEL_ebx.
13132
eabb5f48
UB
131332013-07-06 Uros Bizjak <ubizjak@gmail.com>
13134
13135 * config/i386/sse.md (sse_movlhps): Change alternative 3
13136 of operand 2 to "m".
13137
131382013-07-06 Uros Bizjak <ubizjak@gmail.com>
13139
13140 PR target/57807
13141 * config/i386/sse.md (iptr): New mode attribute.
13142 (sse2_movq128): Add pointer size overrides for Intel asm dialect.
13143 (<sse>_vm<plusminus_insn><mode>3): Ditto.
13144 (<sse>_vmmul<mode>3): Ditto.
13145 (<sse>_vmdiv<mode>3): Ditto.
13146 (sse_vmrcpv4sf2): Ditto.
13147 (<sse>_vmsqrt<mode>2): Ditto.
13148 (sse_vmrsqrtv4sf2): Ditto.
13149 (<sse>_vm<code><mode>3): Ditto.
13150 (avx_vmcmp<mode>3): Ditto.
13151 (<sse>_vmmaskcmp<mode>3): Ditto.
13152 (<sse>_comi): Ditto.
13153 (<sse>_ucomi): Ditto.
13154 (*xop_vmfrcz_<mode>): Ditto.
13155 (*fmai_fmadd_<mode>): Ditto.
13156 (*fmai_fmsub_<mode>): Ditto.
13157 (*fmai_fnmadd_<mode>): Ditto.
13158 (*fmai_fnmsub_<mode>): Ditto.
13159 (*fma4i_vmfmadd_<mode>): Ditto.
13160 (*fma4i_vmfmsub_<mode>): Ditto.
13161 (*fma4i_vmfnmadd_<mode>): Ditto.
13162 (*fma4i_vmfnmsub_<mode>): Ditto.
13163 (*xop_vmfrcz_<mode>): Ditto.
13164 (sse_cvtps2pi): Ditto.
13165 (sse_cvttps2pi): Ditto.
13166 (sse_cvtss2si): Ditto.
13167 (sse_cvtss2si_2): Ditto.
13168 (sse_cvtss2siq_2): Ditto.
13169 (sse_cvttss2si): Ditto.
13170 (sse_cvttss2siq): Ditto.
13171 (sse_cvtsd2si): Ditto.
13172 (sse_cvtsd2si_2): Ditto.
13173 (sse_cvtsd2siq_2): Ditto.
13174 (sse_cvttsd2si): Ditto.
13175 (sse_cvttsd2siq): Ditto.
13176 (sse_cvtsd2ss): Ditto.
13177 (sse_cvtss2sd): Ditto.
13178 (avx2_pbroadcast<mode>): Ditto.
13179 (avx2_pbroadcast<mode>_1): Ditto.
13180 (*avx_vperm_broadcast_v4sf): Ditto.
13181
13182 (sse_movhlps): Ditto for movlp[sd]/movhp[sd] alternatives.
13183 (sse_movlhps): Ditto.
13184 (sse_storehps): Ditto.
13185 (sse_loadhps): Ditto.
13186 (sse_storelps): Ditto.
13187 (sse_loadlps): Ditto.
13188 (*vec_concatv4sf): Ditto.
13189 (*vec_interleave_highv2df): Ditto.
13190 (*vec_interleave_lowv2df): Ditto.
13191 (*vec_extractv2df_1_sse): Ditto.
13192 (*vec_extractv2df_0_sse): Ditto.
13193 (sse2_storelpd): Ditto.
13194 (sse2_loadlpd): Ditto.
13195 (sse2_movsd): Ditto.
13196 (*vec_concatv4si): Ditto.
13197 (vec_concatv2di): Ditto.
13198
13199 * config/i386/mmx.md (mmx_punpcklbw): Add pointer size overrides
13200 for Intel asm dialect.
13201 (mmx_punpcklwd): Ditto.
13202 (mmx_punpckldq): Ditto.
13203
13204 * config/i386/i386.c (ix86_print_operand) ['H']: Output 'qword ptr'
13205 for intel assembler dialect.
13206
1f6eac90
JJ
132072013-07-06 Jakub Jelinek <jakub@redhat.com>
13208
13209 PR target/29776
13210 * fold-const.c (tree_call_nonnegative_warnv_p): Return true
13211 for BUILT_IN_C{LZ,LRSB}*.
13212 * tree.h (CASE_INT_FN): Add FN##IMAX case.
13213 * tree-vrp.c (extract_range_basic): Handle
13214 BUILT_IN_{FFS,PARITY,POPCOUNT,C{LZ,TZ,LRSB}}*. For
13215 BUILT_IN_CONSTANT_P if argument isn't (D) of PARM_DECL,
13216 fall thru to code calling set_value*.
13217 * builtins.c (expand_builtin): Remove *IMAX cases.
13218 (fold_builtin_bitop): For BUILT_IN_CLRSB* return NULL_TREE
13219 if width is bigger than 2*HWI.
13220
2b778c9d
VM
132212013-07-05 Vladimir Makarov <vmakarov@redhat.com>
13222
13223 PR rtl-optimization/55342
13224 * lra-int.h (lra_subreg_reload_pseudos): New.
13225 * lra.c: Add undoing optional reloads to the block diagram.
13226 (lra_subreg_reload_pseudos): New.
13227 (lra_optional_reload_pseudos): Change comments.
13228 (lra): Init and clear lra_subreg_reload_pseudos. Clear
13229 lra_optional_reload_pseudos after undo transformations.
13230 * lra-assigns.c (pseudo_prefix_title): New.
13231 (lra_setup_reg_renumber): Use it.
13232 (spill_for): Ditto. Check subreg reload pseudos too.
13233 (assign_by_spills): Consider subreg reload pseudos too.
13234 * lra-constraints.c (simplify_operand_subreg): Use
13235 lra_subreg_reload_pseudos instead of lra_optional_reload_pseudos.
13236 (curr_insn_transform): Recognize and do optional reloads.
13237 (undo_optional_reloads): New.
13238 (lra_undo_inheritance): Call undo_optional_reloads.
13239
207156e4
TQ
132402013-07-05 Thomas Quinot <quinot@adacore.com>
13241
13242 * tree-complex.c (expand_complex_operations_1): Fix typo.
13243
8990e73a
TB
132442013-07-04 Tejas Belagod <tejas.belagod@arm.com>
13245
13246 * config/aarch64/aarch64-protos.h (cpu_vector_cost): New.
13247 (tune_params): New member 'const vec_costs'.
13248 * config/aarch64/aarch64.c (generic_vector_cost): New.
13249 (generic_tunings): New member 'generic_vector_cost'.
13250 (aarch64_builtin_vectorization_cost): New.
13251 (aarch64_add_stmt_cost): New.
13252 (TARGET_VECTORIZE_ADD_STMT_COST): New.
13253 (TARGET_VECTORIZE_BUILTIN_VECTORIZATION_COST): New.
13254
f7bc421d
JJ
132552013-07-03 Jakub Jelinek <jakub@redhat.com>
13256
13257 PR target/57777
13258 * config/i386/predicates.md (vsib_address_operand): Disallow
13259 SYMBOL_REF or LABEL_REF in parts.disp if TARGET_64BIT && flag_pic.
13260
e90d1568
HPN
132612013-07-03 Hans-Peter Nilsson <hp@bitrange.com>
13262
13263 PR middle-end/55030
13264 * stmt.c (expand_nl_goto_receiver): Remove almost-copy of
13265 expand_builtin_setjmp_receiver.
13266 (expand_label): Adjust, call expand_builtin_setjmp_receiver
13267 with NULL for the label parameter.
13268 * builtins.c (expand_builtin_setjmp_receiver): Don't clobber
13269 the frame-pointer. Adjust comments.
13270 [HAVE_builtin_setjmp_receiver]: Emit builtin_setjmp_receiver
13271 only if LABEL is non-NULL.
13272
cbd7413b
YZ
132732013-07-03 Yufeng Zhang <yufeng.zhang@arm.com>
13274
13275 * config/aarch64/aarch64.h (enum arm_abi_type): Remove.
13276 (ARM_ABI_AAPCS64): Ditto.
13277 (arm_abi): Ditto.
13278 (ARM_DEFAULT_ABI): Ditto.
13279
dec11868
JG
132802013-07-03 James Greenhalgh <james.greenhalgh@arm.com>
13281
13282 * config/aarch64/aarch64-builtins.c
13283 (aarch64_simd_expand_builtin): Handle AARCH64_SIMD_STORE1.
13284 * config/aarch64/aarch64-simd-builtins.def (ld1): New.
13285 (st1): Likewise.
13286 * config/aarch64/aarch64-simd.md
13287 (aarch64_ld1<VALL:mode>): New.
13288 (aarch64_st1<VALL:mode>): Likewise.
13289 * config/aarch64/arm_neon.h
13290 (vld1<q>_<fpsu><8, 16, 32, 64>): Convert to RTL builtins.
13291
786fb9b6
ST
132922013-07-02 Sriraman Tallam <tmsriram@google.com>
13293
0ba6aff3
UB
13294 * config/i386/i386.c (gate_insert_vzeroupper): Check if
13295 target ISA is AVX.
786fb9b6
ST
13296 (ix86_option_override_internal):Turn on all -mavx target flags by
13297 default as they are dependent on AVX anyway.
13298
4b043553
CC
132992013-07-02 Cary Coutant <ccoutant@google.com>
13300
13301 * dwarf2out.c (loc_checksum): Call hash_loc_operands for a
13302 deterministic hash.
13303 (loc_checksum_ordered): Likewise.
13304 (hash_loc_operands): Remove inline keyword.
13305
c4e87a13
JJ
133062013-07-02 Jakub Jelinek <jakub@redhat.com>
13307
13308 PR tree-optimization/57741
13309 * tree-vect-loop.c (vect_is_simple_iv_evolution): Disallow
13310 non-INTEGRAL_TYPE_P non-SCALAR_FLOAT_TYPE_P SSA_NAME step_exprs,
13311 or SCALAR_FLOAT_TYPE_P SSA_NAMEs if !flag_associative_math.
13312 Allow REAL_CST step_exprs if flag_associative_math.
13313 (get_initial_def_for_induction): Handle SCALAR_FLOAT_TYPE_P step_expr.
13314
26366d28
IB
133152013-07-02 Ian Bolton <ian.bolton@arm.com>
13316
0ba6aff3 13317 * config/aarch64/aarch64-simd.md (absdi2): Support abs for DI mode.
26366d28 13318
2879bb2b
IB
133192013-07-02 Ian Bolton <ian.bolton@arm.com>
13320
13321 * config/aarch64/aarch64.md (*extr_insv_reg<mode>): New pattern.
13322
afc5e8a6
KT
133232013-07-02 Kyrylo Tkachov <kyrylo.tkachov@arm.com>
13324
13325 * config/arm/arm.md (arm_andsi3_insn): Add alternatives for 16-bit
13326 encoding.
13327 (iorsi3_insn): Likewise.
13328 (arm_xorsi3): Likewise.
13329
31f8442b
SN
133302013-07-01 Sofiane Naci <sofiane.naci@arm.com>
13331
13332 * arm.md (attribute "wtype"): Delete. Move attribute values from here
13333 to ...
13334 (attribute "type"): ... here, and prefix with "wmmx_".
13335 (attribute "core_cycles"): Update for attribute changes.
13336 * iwmmxt.md (tbcstv8qi): Update for attribute changes.
13337 (tbcstv4hi): Likewise.
13338 (tbcstv2si): Likewise.
13339 (iwmmxt_iordi3): Likewise.
13340 (iwmmxt_xordi3): Likewise.
13341 (iwmmxt_anddi3): Likewise.
13342 (iwmmxt_nanddi3): Likewise.
13343 (iwmmxt_arm_movdi): Likewise.
13344 (iwmmxt_movsi_insn): Likewise.
13345 (mov<mode>_internal): Likewise.
13346 (and<mode>3_iwmmxt): Likewise.
13347 (ior<mode>3_iwmmxt): Likewise.
13348 (xor<mode>3_iwmmxt): Likewise.
13349 (add<mode>3_iwmmxt): Likewise.
13350 (ssaddv8qi3): Likewise.
13351 (ssaddv4hi3): Likewise.
13352 (ssaddv2si3): Likewise.
13353 (usaddv8qi3): Likewise.
13354 (usaddv4hi3): Likewise.
13355 (usaddv2si3): Likewise.
13356 (sub<mode>3_iwmmxt): Likewise.
13357 (sssubv8qi3): Likewise.
13358 (sssubv4hi3): Likewise.
13359 (sssubv2si3): Likewise.
13360 (ussubv8qi3): Likewise.
13361 (ussubv4hi3): Likewise.
13362 (ussubv2si3): Likewise.
13363 (mulv4hi3_iwmmxt): Likewise.
13364 (smulv4hi3_highpart): Likewise.
13365 (umulv4hi3_highpart): Likewise.
13366 (iwmmxt_wmacs): Likewise.
13367 (iwmmxt_wmacsz): Likewise.
13368 (iwmmxt_wmacu): Likewise.
13369 (iwmmxt_wmacuz): Likewise.
13370 (iwmmxt_clrdi): Likewise.
13371 (iwmmxt_clrv8qi): Likewise.
13372 (iwmmxt_clr4hi): Likewise.
13373 (iwmmxt_clr2si): Likewise.
13374 (iwmmxt_uavgrndv8qi3): Likewise.
13375 (iwmmxt_uavgrndv4hi3): Likewise.
13376 (iwmmxt_uavgv8qi3): Likewise.
13377 (iwmmxt_uavgv4hi3): Likewise.
13378 (iwmmxt_tinsrb): Likewise.
13379 (iwmmxt_tinsrh): Likewise.
13380 (iwmmxt_tinsrw): Likewise.
13381 (iwmmxt_textrmub): Likewise.
13382 (iwmmxt_textrmsb): Likewise.
13383 (iwmmxt_textrmuh): Likewise.
13384 (iwmmxt_textrmsh): Likewise.
13385 (iwmmxt_textrmw): Likewise.
13386 (iwmxxt_wshufh): Likewise.
13387 (eqv8qi3): Likewise.
13388 (eqv4hi3): Likewise.
13389 (eqv2si3): Likewise.
13390 (gtuv8qi3): Likewise.
13391 (gtuv4hi3): Likewise.
13392 (gtuv2si3): Likewise.
13393 (gtv8qi3): Likewise.
13394 (gtv4hi3): Likewise.
13395 (gtv2si3): Likewise.
13396 (smax<mode>3_iwmmxt): Likewise.
13397 (umax<mode>3_iwmmxt): Likewise.
13398 (smin<mode>3_iwmmxt): Likewise.
13399 (umin<mode>3_iwmmxt): Likewise.
13400 (iwmmxt_wpackhss): Likewise.
13401 (iwmmxt_wpackwss): Likewise.
13402 (iwmmxt_wpackdss): Likewise.
13403 (iwmmxt_wpackhus): Likewise.
13404 (iwmmxt_wpackwus): Likewise.
13405 (iwmmxt_wpackdus): Likewise.
13406 (iwmmxt_wunpckihb): Likewise.
13407 (iwmmxt_wunpckihh): Likewise.
13408 (iwmmxt_wunpckihw): Likewise.
13409 (iwmmxt_wunpckilb): Likewise.
13410 (iwmmxt_wunpckilh): Likewise.
13411 (iwmmxt_wunpckilw): Likewise.
13412 (iwmmxt_wunpckehub): Likewise.
13413 (iwmmxt_wunpckehuh): Likewise.
13414 (iwmmxt_wunpckehuw): Likewise.
13415 (iwmmxt_wunpckehsb): Likewise.
13416 (iwmmxt_wunpckehsh): Likewise.
13417 (iwmmxt_wunpckehsw): Likewise.
13418 (iwmmxt_wunpckelub): Likewise.
13419 (iwmmxt_wunpckeluh): Likewise.
13420 (iwmmxt_wunpckeluw): Likewise.
13421 (iwmmxt_wunpckelsb): Likewise.
13422 (iwmmxt_wunpckelsh): Likewise.
13423 (iwmmxt_wunpckelsw): Likewise.
13424 (ror<mode>3): Likewise.
13425 (ashr<mode>3_iwmmxt): Likewise.
13426 (lshr<mode>3_iwmmxt): Likewise.
13427 (ashl<mode>3_iwmmxt): Likewise.
13428 (ror<mode>3_di): Likewise.
13429 (ashr<mode>3_di): Likewise.
13430 (lshr<mode>3_di): Likewise.
13431 (ashl<mode>3_di): Likewise.
13432 (iwmmxt_wmadds): Likewise.
13433 (iwmmxt_wmaddu): Likewise.
13434 (iwmmxt_tmia): Likewise.
13435 (iwmmxt_tmiaph): Likewise.
13436 (iwmmxt_tmiabb): Likewise.
13437 (iwmmxt_tmiatb): Likewise.
13438 (iwmmxt_tmiabt): Likewise.
13439 (iwmmxt_tmiatt): Likewise.
13440 (iwmmxt_tmovmskb): Likewise.
13441 (iwmmxt_tmovmskh): Likewise.
13442 (iwmmxt_tmovmskw): Likewise.
13443 (iwmmxt_waccb): Likewise.
13444 (iwmmxt_wacch): Likewise.
13445 (iwmmxt_waccw): Likewise.
13446 (iwmmxt_waligni): Likewise.
13447 (iwmmxt_walignr): Likewise.
13448 (iwmmxt_walignr0): Likewise.
13449 (iwmmxt_walignr1): Likewise.
13450 (iwmmxt_walignr2): Likewise.
13451 (iwmmxt_walignr3): Likewise.
13452 (iwmmxt_wsadb): Likewise.
13453 (iwmmxt_wsadh): Likewise.
13454 (iwmmxt_wsadbz): Likewise.
13455 (iwmmxt_wsadhz): Likewise.
13456 * iwmmxt2.md (iwmmxt_wabs<mode>3): Update for attribute changes.
13457 (iwmmxt_wabsdiffb): Likewise.
13458 (iwmmxt_wabsdiffh): Likewise.
13459 (iwmmxt_wabsdiffw): Likewise.
13460 (iwmmxt_waddsubhx): Likewise
13461 (iwmmxt_wsubaddhx): Likewise.
13462 (addc<mode>3): Likewise.
13463 (iwmmxt_avg4): Likewise.
13464 (iwmmxt_avg4r): Likewise.
13465 (iwmmxt_wmaddsx): Likewise.
13466 (iwmmxt_wmaddux): Likewise.
13467 (iwmmxt_wmaddsn): Likewise.
13468 (iwmmxt_wmaddun): Likewise.
13469 (iwmmxt_wmulwsm): Likewise.
13470 (iwmmxt_wmulwum): Likewise.
13471 (iwmmxt_wmulsmr): Likewise.
13472 (iwmmxt_wmulumr): Likewise.
13473 (iwmmxt_wmulwsmr): Likewise.
13474 (iwmmxt_wmulwumr): Likewise.
13475 (iwmmxt_wmulwl): Likewise.
13476 (iwmmxt_wqmulm): Likewise.
13477 (iwmmxt_wqmulwm): Likewise.
13478 (iwmmxt_wqmulmr): Likewise.
13479 (iwmmxt_wqmulwmr): Likewise.
13480 (iwmmxt_waddbhusm): Likewise.
13481 (iwmmxt_waddbhusl): Likewise.
13482 (iwmmxt_wqmiabb): Likewise.
13483 (iwmmxt_wqmiabt): Likewise.
13484 (iwmmxt_wqmiatb): Likewise.
13485 (iwmmxt_wqmiatt): Likewise.
13486 (iwmmxt_wqmiabbn): Likewise.
13487 (iwmmxt_wqmiabtn): Likewise.
13488 (iwmmxt_wqmiatbn): Likewise.
13489 (iwmmxt_wqmiattn): Likewise.
13490 (iwmmxt_wmiabb): Likewise.
13491 (iwmmxt_wmiabt): Likewise.
13492 (iwmmxt_wmiatb): Likewise.
13493 (iwmmxt_wmiatt): Likewise.
13494 (iwmmxt_wmiabbn): Likewise.
13495 (iwmmxt_wmiabtn): Likewise.
13496 (iwmmxt_wmiatbn): Likewise.
13497 (iwmmxt_wmiattn): Likewise.
13498 (iwmmxt_wmiawbb): Likewise.
13499 (iwmmxt_wmiawbt): Likewise.
13500 (iwmmxt_wmiawtb): Likewise.
13501 (iwmmxt_wmiawtt): Likewise.
13502 (iwmmxt_wmiawbbn): Likewise.
13503 (iwmmxt_wmiawbtn): Likewise.
13504 (iwmmxt_wmiawtbn): Likewise.
13505 (iwmmxt_wmiawttn): Likewise.
13506 (iwmmxt_wmerge): Likewise.
13507 (iwmmxt_tandc<mode>3): Likewise.
13508 (iwmmxt_torc<mode>3): Likewise.
13509 (iwmmxt_torvsc<mode>3): Likewise.
13510 (iwmmxt_textrc<mode>3): Likewise.
13511 * marvell-f-iwmmxt.md (wmmxt_shift): Update for attribute changes.
13512 (wmmxt_pack): Likewise.
13513 (wmmxt_mult_c1): Likewise.
13514 (wmmxt_mult_c2): Likewise.
13515 (wmmxt_alu_c1): Likewise.
13516 (wmmxt_alu_c2): Likewise.
13517 (wmmxt_alu_c3): Likewise.
13518 (wmmxt_transfer_c1): Likewise.
13519 (wmmxt_transfer_c2): Likewise.
13520 (wmmxt_transfer_c3): Likewise.
13521 (marvell_f_iwmmxt_wstr): Likewise.
13522 (marvell_f_iwmmxt_wldr): Likewise.
13523
bdb7bf8a
YZ
135242013-06-29 Yufeng Zhang <yufeng.zhang@arm.com>
13525
0ba6aff3 13526 * config/aarch64/aarch64.c: Remove junk from the beginning of the file.
bdb7bf8a 13527
e6631d38
VM
135282013-06-28 Vladimir Makarov <vmakarov@redhat.com>
13529
13530 Revert:
13531 2013-06-28 Vladimir Makarov <vmakarov@redhat.com>
13532 * lra-constraints.c (need_for_split_p): Check call used hard regs
13533 living through calls.
13534
13535 * lra-constraints.c (inherit_in_ebb): Reset live_hard_regs for
13536 call used regs for call insn.
13537
53c1275b
JJ
135382013-06-28 Jakub Jelinek <jakub@redhat.com>
13539
13540 PR target/57736
0ba6aff3
UB
13541 * config/i386/i386.c (ix86_expand_builtin): If target == NULL and
13542 mode is VOIDmode, don't create a VOIDmode pseudo to copy result into.
53c1275b 13543
1141ed3f
BI
135442013-06-28 Balaji V. Iyer <balaji.v.iyer@intel.com>
13545
13546 * builtins.def: Fixed the function type of CILKPLUS_BUILTIN.
e6631d38 13547
e32e4c4a
VM
135482013-06-28 Vladimir Makarov <vmakarov@redhat.com>
13549
13550 * lra-constraints.c (need_for_split_p): Check call used hard regs
13551 living through calls.
13552
f161bfd3
MM
135532013-06-28 Michael Meissner <meissner@linux.vnet.ibm.com>
13554
13555 PR target/57744
13556 * config/rs6000/rs6000.h (MODES_TIEABLE_P): Do not allow PTImode
13557 to tie with any other modes. Eliminate Altivec vector mode tests,
13558 since these are a subset of ALTIVEC or VSX vector modes. Simplify
13559 code, to return 0 if testing MODE2 for a condition, if we've
13560 already tested MODE1 for the same condition.
13561
7eda14e1
MS
135622013-06-28 Marcus Shawcroft <marcus.shawcroft@arm.com>
13563
13564 * config/aarch64/aarch64.c (aarch64_cannot_force_const_mem): Adjust
13565 layout.
13566
c822f852
MS
135672013-06-28 Marcus Shawcroft <marcus.shawcroft@arm.com>
13568
13569 * config/aarch64/aarch64-protos.h (aarch64_symbol_type):
13570 Update comment w.r.t SYMBOL_TINY_ABSOLUTE.
13571
da4f13a4
MS
135722013-06-28 Marcus Shawcroft <marcus.shawcroft@arm.com>
13573
0ba6aff3
UB
13574 * config/aarch64/aarch64-protos.h (aarch64_classify_symbol_expression):
13575 Define.
da4f13a4
MS
13576 (aarch64_symbolic_constant_p): Remove.
13577 * config/aarch64/aarch64.c (aarch64_classify_symbol_expression): Remove
13578 static. Fix line length and white space.
13579 (aarch64_symbolic_constant_p): Remove.
13580 * config/aarch64/predicates.md (aarch64_valid_symref):
13581 Use aarch64_classify_symbol_expression.
13582
956a95a5
KT
135832013-06-28 Kyrylo Tkachov <kyrylo.tkachov@arm.com>
13584
13585 * config/arm/constraints.md (Ts): New constraint.
13586 * config/arm/arm.md (arm_movqi_insn): Add alternatives for
13587 16-bit encodings.
13588 (compare_scc): Use "Ts" constraint for operand 0.
13589 (ior_scc_scc): Likewise.
13590 (and_scc_scc): Likewise.
13591 (and_scc_scc_nodom): Likewise.
13592 (ior_scc_scc_cmp): Likewise for operand 7.
13593 (and_scc_scc_cmp): Likewise.
13594 * config/arm/thumb2.md (thumb2_movsi_insn):
13595 Add alternatives for 16-bit encodings.
13596 (thumb2_movhi_insn): Likewise.
13597 (thumb2_movsicc_insn): Likewise.
13598 (thumb2_and_scc): Take 'and' outside cond_exec. Use "Ts" constraint.
13599 (thumb2_negscc): Use "Ts" constraint.
13600 Move mvn instruction outside cond_exec block.
13601 * config/arm/vfp.md (thumb2_movsi_vfp): Add alternatives
13602 for 16-bit encodings.
13603
31dfce10
KT
136042013-06-28 Kyrylo Tkachov <kyrylo.tkachov@arm.com>
13605
13606 * config/arm/arm.md (arm_mulsi3_v6): Add alternative for 16-bit
13607 encoding.
13608 (mulsi3addsi_v6): Disable predicable variant for arm_restrict_it.
13609 (mulsi3subsi): Likewise.
13610 (mulsidi3adddi): Likewise.
13611 (mulsidi3_v6): Likewise.
13612 (umulsidi3_v6): Likewise.
13613 (umulsidi3adddi_v6): Likewise.
13614 (smulsi3_highpart_v6): Likewise.
13615 (umulsi3_highpart_v6): Likewise.
13616 (mulhisi3tb): Likewise.
13617 (mulhisi3bt): Likewise.
13618 (mulhisi3tt): Likewise.
13619 (maddhisi4): Likewise.
13620 (maddhisi4tb): Likewise.
13621 (maddhisi4tt): Likewise.
13622 (maddhidi4): Likewise.
13623 (maddhidi4tb): Likewise.
13624 (maddhidi4tt): Likewise.
13625 (zeroextractsi_compare0_scratch): Likewise.
13626 (insv_zero): Likewise.
13627 (insv_t2): Likewise.
13628 (anddi_notzesidi_di): Likewise.
13629 (anddi_notsesidi_di): Likewise.
13630 (andsi_notsi_si): Likewise.
13631 (iordi_zesidi_di): Likewise.
13632 (xordi_zesidi_di): Likewise.
13633 (andsi_iorsi3_notsi): Likewise.
13634 (smax_0): Likewise.
13635 (smax_m1): Likewise.
13636 (smin_0): Likewise.
13637 (not_shiftsi): Likewise.
13638 (unaligned_loadsi): Likewise.
13639 (unaligned_loadhis): Likewise.
13640 (unaligned_loadhiu): Likewise.
13641 (unaligned_storesi): Likewise.
13642 (unaligned_storehi): Likewise.
13643 (extv_reg): Likewise.
13644 (extzv_t2): Likewise.
13645 (divsi3): Likewise.
13646 (udivsi3): Likewise.
13647 (arm_zero_extendhisi2addsi): Likewise.
13648 (arm_zero_extendqisi2addsi): Likewise.
13649 (compareqi_eq0): Likewise.
13650 (arm_extendhisi2_v6): Likewise.
13651 (arm_extendqisi2addsi): Likewise.
13652 (arm_movt): Likewise.
13653 (thumb2_ldrd): Likewise.
13654 (thumb2_ldrd_base): Likewise.
13655 (thumb2_ldrd_base_neg): Likewise.
13656 (thumb2_strd): Likewise.
13657 (thumb2_strd_base): Likewise.
13658 (thumb2_strd_base_neg): Likewise.
13659 (arm_negsi2): Add alternative for 16-bit encoding.
13660 (arm_one_cmplsi2): Likewise.
13661
1572e697
KT
136622013-06-28 Kyrylo Tkachov <kyrylo.tkachov@arm.com>
13663
13664 * config/arm/predicates.md (arm_cond_move_operator): New predicate.
13665 * config/arm/arm.md (movsfcc): Use arm_cond_move_operator predicate.
13666 (movdfcc): Likewise.
13667 * config/arm/vfp.md (*thumb2_movsf_vfp):
13668 Disable predication for arm_restrict_it.
13669 (*thumb2_movsfcc_vfp): Disable for arm_restrict_it.
13670 (*thumb2_movdfcc_vfp): Likewise.
13671 (*abssf2_vfp, *absdf2_vfp, *negsf2_vfp, *negdf2_vfp,*addsf3_vfp,
13672 *adddf3_vfp, *subsf3_vfp, *subdf3_vfpc, *divsf3_vfp,*divdf3_vfp,
13673 *mulsf3_vfp, *muldf3_vfp, *mulsf3negsf_vfp, *muldf3negdf_vfp,
13674 *mulsf3addsf_vfp, *muldf3adddf_vfp, *mulsf3subsf_vfp,
13675 *muldf3subdf_vfp, *mulsf3negsfaddsf_vfp, *fmuldf3negdfadddf_vfp,
13676 *mulsf3negsfsubsf_vfp, *muldf3negdfsubdf_vfp, *fma<SDF:mode>4,
13677 *fmsub<SDF:mode>4, *fnmsub<SDF:mode>4, *fnmadd<SDF:mode>4,
13678 *extendsfdf2_vfp, *truncdfsf2_vfp, *extendhfsf2, *truncsfhf2,
13679 *truncsisf2_vfp, *truncsidf2_vfp, fixuns_truncsfsi2, fixuns_truncdfsi2,
13680 *floatsisf2_vfp, *floatsidf2_vfp, floatunssisf2, floatunssidf2,
13681 *sqrtsf2_vfp, *sqrtdf2_vfp, *cmpsf_vfp, *cmpsf_trap_vfp, *cmpdf_vfp,
13682 *cmpdf_trap_vfp, <vrint_pattern><SDF:mode>2):
13683 Disable predication for arm_restrict_it.
13684
ebf8f0ea
KY
136852013-06-28 Kirill Yukhin <kirill.yukhin@intel.com>
13686
13687 * config/i386/bmiintrin.h (_bextr_u32): New.
13688 (_bextr_u64): Ditto.
13689
a290fcda
RS
136902013-06-27 Richard Sandiford <rdsandiford@googlemail.com>
13691
13692 * config.gcc (mips*-mti-elf*, mips*-sde-elf*, mips64r5900-*-elf*)
13693 (mips64r5900el-*-elf*): Include mips/n32-elf.h.
13694 * config/mips/sde.h (LOCAL_LABEL_PREFIX, NO_DOLLAR_IN_LABEL)
13695 (LONG_DOUBLE_TYPE_SIZE, LIBGCC2_LONG_DOUBLE_TYPE_SIZE): Move to...
13696 * config/mips/n32-elf.h: ...this new file.
13697
ea169996
MG
136982013-06-27 Marc Glisse <marc.glisse@inria.fr>
13699
13700 PR target/57224
13701 * config/i386/i386.c (enum ix86_builtins, bdesc_args): Remove
13702 IX86_BUILTIN_CMPNGTSS and IX86_BUILTIN_CMPNGESS.
13703
12211b99 137042013-06-27 Catherine Moore <clm@codesourcery.com>
8e932114
CM
13705
13706 * config/mips/mips-tables.opt: Regenerate.
13707 * config/mips/mips-cpus.def: Add m14ke and m14kec.
0ba6aff3 13708 * config/mips/mips.h (BASE_DRIVER_SELF_SPECS): m14ke* implies -mdspr2.
8e932114 13709 * doc/invoke.texi: Add -m14kc.
ebf8f0ea 13710
2444244e
JJ
137112013-06-27 Jakub Jelinek <jakub@redhat.com>
13712
cd06d2a2
JJ
13713 PR target/57623
13714 * config/i386/i386.md (bmi_bextr_<mode>): Swap predicates and
13715 constraints of operand 1 and 2.
13716
2444244e
JJ
13717 PR target/57623
13718 * config/i386/i386.md (bmi2_bzhi_<mode>3): Swap AND arguments
13719 to match RTL canonicalization. Swap predicates and
13720 constraints of operand 1 and 2.
13721
2f259720
VM
137222013-06-27 Vladimir Makarov <vmakarov@redhat.com>
13723
0ba6aff3
UB
13724 * lra-constraints.c (inherit_in_ebb): Process static hard regs too.
13725 Process OP_INOUT regs for splitting too.
2f259720 13726
fef4d2b3
JJ
137272013-06-27 Jakub Jelinek <jakub@redhat.com>
13728
13729 * tree-vect-stmts.c (vectorizable_store): Move ptr_incr var
13730 decl before the loop, initialize to NULL.
13731 (vectorizable_load): Initialize ptr_incr to NULL.
13732
79ee9826
MJ
137332013-06-27 Martin Jambor <mjambor@suse.cz>
13734
13735 PR lto/57208
13736 * ipa-ref.h (ipa_maybe_record_reference): Declare.
13737 * ipa-ref.c (ipa_maybe_record_reference): New function.
13738 * cgraphclones.c (cgraph_create_virtual_clone): Use it.
13739 * ipa-cp.c (create_specialized_node): Record potential references from
13740 aggvals.
13741 * Makefile.in (ipa-ref.o): Add IPA_REF_H to dependencies.
13742
e18b4a81
YZ
137432013-06-27 Yufeng Zhang <yufeng.zhang@arm.com>
13744
13745 * config/aarch64/aarch64.c (aarch64_force_temporary): Add an extra
13746 parameter 'mode' of type 'enum machine_mode mode'; change to pass
13747 'mode' to force_reg.
13748 (aarch64_add_offset): Update calls to aarch64_force_temporary.
13749 (aarch64_expand_mov_immediate): Likewise.
13750
9c023bf0
YZ
137512013-06-27 Yufeng Zhang <yufeng.zhang@arm.com>
13752
13753 * config/aarch64/aarch64.c (aarch64_add_offset): Change to pass
13754 'mode' to aarch64_plus_immediate and gen_rtx_PLUS.
13755
5a3fe9b6
AK
137562013-06-27 Andreas Krebbel <Andreas.Krebbel@de.ibm.com>
13757
13758 * config/s390/s390.c: Rename UNSPEC_CCU_TO_INT to
13759 UNSPEC_STRCMPCC_TO_INT and UNSPEC_CCZ_TO_INT to UNSPEC_CC_TO_INT.
13760 (struct machine_function): Add tbegin_p.
13761 (s390_canonicalize_comparison): Fold CC mode compares to
13762 conditional jump if possible.
13763 (s390_emit_jump): Return the emitted jump.
13764 (s390_branch_condition_mask, s390_branch_condition_mnemonic):
13765 Handle CCRAWmode compares.
13766 (s390_option_override): Default to -mhtm if available.
13767 (s390_reg_clobbered_rtx): Handle floating point regs as well.
13768 (s390_regs_ever_clobbered): Use s390_regs_ever_clobbered also for
13769 FPRs instead of df_regs_ever_live_p.
13770 (s390_optimize_nonescaping_tx): New function.
13771 (s390_init_frame_layout): Extend clobbered_regs array to cover
13772 FPRs as well.
13773 (s390_emit_prologue): Call s390_optimize_nonescaping_tx.
13774 (s390_expand_tbegin): New function.
13775 (enum s390_builtin): New enum definition.
13776 (code_for_builtin): New array definition.
13777 (s390_init_builtins): New function.
13778 (s390_expand_builtin): New function.
13779 (TARGET_INIT_BUILTINS): Define.
13780 (TARGET_EXPAND_BUILTIN): Define.
0ba6aff3 13781 * common/config/s390/s390-common.c (processor_flags_table): Add PF_TX.
5a3fe9b6
AK
13782 * config/s390/predicates.md (s390_comparison): Handle CCRAWmode.
13783 (s390_alc_comparison): Likewise.
13784 * config/s390/s390-modes.def: Add CCRAWmode.
13785 * config/s390/s390.h (processor_flags): Add PF_TX.
13786 (TARGET_CPU_HTM): Define macro.
13787 (TARGET_HTM): Define macro.
13788 (TARGET_CPU_CPP_BUILTINS): Define __HTM__ for htm.
13789 * config/s390/s390.md: Rename UNSPEC_CCU_TO_INT to
13790 UNSPEC_STRCMPCC_TO_INT and UNSPEC_CCZ_TO_INT to UNSPEC_CC_TO_INT.
13791 (UNSPECV_TBEGIN, UNSPECV_TBEGINC, UNSPECV_TEND, UNSPECV_TABORT)
0ba6aff3 13792 (UNSPECV_ETND, UNSPECV_NTSTG, UNSPECV_PPA): New unspecv enum values.
5a3fe9b6
AK
13793 (TBEGIN_MASK, TBEGINC_MASK): New constants.
13794 ("*cc_to_int"): Move up.
13795 ("*mov<mode>cc", "*cjump_64", "*cjump_31"): Accept integer
13796 constants other than 0.
13797 ("*ccraw_to_int"): New insn and splitter definition.
13798 ("tbegin", "tbegin_nofloat", "tbegin_retry")
13799 ("tbegin_retry_nofloat", "tbeginc", "tend", "tabort")
13800 ("tx_assist"): New expander.
13801 ("tbegin_1", "tbegin_nofloat_1", "*tbeginc_1", "*tend_1")
13802 ("*tabort_1", "etnd", "ntstg", "*ppa"): New insn definition.
13803 * config/s390/s390.opt: Add -mhtm option.
13804 * config/s390/s390-protos.h (s390_emit_jump): Add return type.
13805 * config/s390/htmxlintrin.h: New file.
13806 * config/s390/htmintrin.h: New file.
13807 * config/s390/s390intrin.h: New file.
13808 * doc/extend.texi: Document htm builtins.
13809 * config.gcc: Add the new header files to extra_headers.
13810
9e216629
TS
138112013-06-26 Thomas Schwinge <thomas@codesourcery.com>
13812
13813 * config/i386/gnu.h [TARGET_LIBC_PROVIDES_SSP]
0ba6aff3 13814 (TARGET_CAN_SPLIT_STACK, TARGET_THREAD_SPLIT_STACK_OFFSET): Undefine.
9e216629 13815
fd8c65e7 138162013-06-26 Michael Meissner <meissner@linux.vnet.ibm.com>
12211b99
UB
13817 Pat Haugen <pthaugen@us.ibm.com>
13818 Peter Bergner <bergner@vnet.ibm.com>
fd8c65e7
MM
13819
13820 * config/rs6000/power8.md: New.
13821 * config/rs6000/rs6000-cpus.def (RS6000_CPU table): Adjust processor
13822 setting for power8 entry.
13823 * config/rs6000/t-rs6000 (MD_INCLUDES): Add power8.md.
13824 * config/rs6000/rs6000.c (is_microcoded_insn, is_cracked_insn): Adjust
13825 test for Power4/Power5 only.
13826 (insn_must_be_first_in_group, insn_must_be_last_in_group): Add Power8
13827 support.
13828 (force_new_group): Adjust comment.
13829 * config/rs6000/rs6000.md: Include power8.md.
13830
b24a2ce5
GY
138312013-06-26 Greta Yorsh <Greta.Yorsh@arm.com>
13832
13833 * config/arm/arm.h (MAX_CONDITIONAL_EXECUTE): Define macro.
13834 * config/arm/arm-protos.h (arm_max_conditional_execute): New
13835 declaration.
13836 (tune_params): Update comment.
13837 * config/arm/arm.c (arm_cortex_a15_tune): Set max_cond_insns to 2.
13838 (arm_max_conditional_execute): New function.
13839 (thumb2_final_prescan_insn): Use max_insn_skipped and
13840 MAX_INSN_PER_IT_BLOCK to compute maximum instructions in a block.
13841
f6fdeb62
JJ
138422013-06-25 Jakub Jelinek <jakub@redhat.com>
13843
13844 PR tree-optimization/57705
13845 * tree-vect-loop.c (vect_is_simple_iv_evolution): Allow
13846 SSA_NAME step, provided that it is not defined inside the loop.
0ba6aff3 13847 (vect_analyze_scalar_cycles_1): Disallow SSA_NAME step in nested loop.
f6fdeb62
JJ
13848 (get_initial_def_for_induction): Handle SSA_NAME IV step.
13849
c13bc3d9
MJ
138502013-06-25 Martin Jambor <mjambor@suse.cz>
13851
13852 PR middle-end/57670
13853 * cgraph.h (cgraph_indirect_call_info): New flag member_ptr.
13854 * ipa-prop.c (ipa_print_node_jump_functions): Mark member pointer
13855 calls in the dump.
13856 (ipa_note_param_call): Initialize member_ptr flag.
13857 (ipa_analyze_indirect_call_uses): Set member_ptr flag.
13858 (ipa_make_edge_direct_to_target): Bail out if member_ptr is set.
13859 (ipa_write_indirect_edge_info): Stream member_ptr flag.
13860 (ipa_read_indirect_edge_info): Likewise.
13861
8a845901
RB
138622013-06-25 Richard Biener <rguenther@suse.de>
13863
13864 PR middle-end/56977
13865 * passes.c (init_optimization_passes): Move pass_fold_builtins
13866 and pass_dce earlier with -Og.
13867
33be0bec
EB
138682013-06-25 Eric Botcazou <ebotcazou@adacore.com>
13869
13870 * expr.c (expand_expr_real_1) <ARRAY_REF>: Fix formatting glitches.
13871 <BIT_FIELD_REF>: Remove trailing TAB.
13872 * varasm.c (output_constructor_bitfield): Fix formatting glitch and
13873 remove blank line.
13874
5fe8e757
MJ
138752013-06-24 Martin Jambor <mjambor@suse.cz>
13876
13877 PR tree-optimization/57358
13878 * ipa-prop.c (ipa_func_spec_opts_forbid_analysis_p): New function.
13879 (ipa_compute_jump_functions_for_edge): Bail out if it returns true.
13880 (ipa_analyze_params_uses): Generate pessimistic info when true.
13881
44a60244
MJ
138822013-06-24 Martin Jambor <mjambor@suse.cz>
13883
13884 PR tree-optimization/57539
13885 * cgraphclones.c (cgraph_clone_node): Add parameter new_inlined_to, set
13886 global.inlined_to of the new node to it. All callers changed.
13887 * ipa-inline-transform.c (clone_inlined_nodes): New variable
13888 inlining_into, pass it to cgraph_clone_node.
13889 * ipa-prop.c (ipa_propagate_indirect_call_infos): Do not call
13890 ipa_free_edge_args_substructures.
13891 (ipa_edge_duplication_hook): Only add edges from inlined nodes to
13892 rdesc linked list. Do not assert rdesc edges have inlined caller.
13893 Assert we have found an rdesc in the rdesc list.
13894
7c5848b8
RB
138952013-06-24 Richard Biener <rguenther@suse.de>
13896
0ba6aff3 13897 * pointer-set.h (struct pointer_set_t): Move here from pointer-set.c.
7c5848b8
RB
13898 (pointer_set_lookup): Declare.
13899 (class pointer_map): New template class implementing a
13900 generic pointer to T map.
13901 (pointer_map<T>::pointer_map, pointer_map<T>::~pointer_map,
13902 pointer_map<T>::contains, pointer_map<T>::insert,
13903 pointer_map<T>::traverse): New functions.
13904 * pointer-set.c (struct pointer_set_t): Moved to pointer-set.h.
13905 (pointer_set_lookup): New function.
13906 (pointer_set_contains): Use pointer_set_lookup.
13907 (pointer_set_insert): Likewise.
13908 (insert_aux): Remove.
13909 (struct pointer_map_t): Embed a pointer_set_t.
13910 (pointer_map_create): Adjust.
13911 (pointer_map_destroy): Likewise.
13912 (pointer_map_contains): Likewise.
13913 (pointer_map_insert): Likewise.
13914 (pointer_map_traverse): Likewise.
13915 * tree-streamer.h (struct streamer_tree_cache_d): Use a
13916 pointer_map<unsigned> instead of a pointer_map_t.
13917 * tree-streamer.c (streamer_tree_cache_insert_1): Adjust.
13918 (streamer_tree_cache_lookup): Likewise.
13919 (streamer_tree_cache_create): Likewise.
13920 (streamer_tree_cache_delete): Likewise.
13921 * lto-streamer.h (struct lto_tree_ref_encoder): Use a
13922 pointer_map<unsigned> instead of a pointer_map_t.
13923 (lto_init_tree_ref_encoder): Adjust.
13924 (lto_destroy_tree_ref_encoder): Likewise.
13925 * lto-section-out.c (lto_output_decl_index): Likewise.
13926 (lto_record_function_out_decl_state): Likewise.
13927 * dominance.c (iterate_fix_dominators): Use pointer_map<int>.
13928
3dbc97a9
RB
139292013-06-24 Richard Biener <rguenther@suse.de>
13930
13931 PR tree-optimization/57488
13932 * tree-ssa-pre.c (insert): Clear NEW sets before each iteration.
13933
8345965a
AM
139342013-06-24 Alan Modra <amodra@gmail.com>
13935
13936 * config/rs6000/rs6000.c (vspltis_constant): Correct for little-endian.
13937 (gen_easy_altivec_constant): Likewise.
13938 * config/rs6000/predicates.md (easy_vector_constant_add_self,
13939 easy_vector_constant_msb): Likewise.
13940
b7df379f
JJ
139412013-06-23 Jakub Jelinek <jakub@redhat.com>
13942
13943 PR target/57688
13944 * common/config/i386/i386-common.c (ix86_handle_option): For OPT_mlzcnt
13945 add missing return true.
13946
12211b99 139472013-06-23 Oleg Endo <olegendo@gcc.gnu.org>
0bcf9a09
OE
13948
13949 PR target/52483
13950 * config/sh/predicates.md (general_extend_operand): Invoke
13951 general_movsrc_operand for memory operands.
13952 (general_movsrc_operand): Allow reg+reg addressing, do not use
13953 general_operand for memory operands.
13954
97db2bf7
ST
139552013-06-23 Sriraman Tallam <tmsriram@google.com>
13956
13957 * config/i386/i386.c (ix86_pragma_target_parse): Restore target
13958 when current target options does not apply.
13959 * config/i386/i386-protos.h (ix86_reset_previous_fndecl): New function.
13960 * config/i386/i386.c (ix86_reset_previous_fndecl): Ditto.
0ba6aff3
UB
13961 * config/i386/bmiintrin.h: Pass appropriate target
13962 attributes to header.
97db2bf7
ST
13963 * config/i386/mmintrin.h: Ditto.
13964 * config/i386/nmmintrin.h: Ditto.
13965 * config/i386/avx2intrin.h: Ditto.
13966 * config/i386/fxsrintrin.h: Ditto.
13967 * config/i386/tbmintrin.h: Ditto.
13968 * config/i386/xsaveintrin.h: Ditto.
13969 * config/i386/f16cintrin.h: Ditto.
13970 * config/i386/xtestintrin.h: Ditto.
13971 * config/i386/xsaveoptintrin.h: Ditto.
13972 * config/i386/bmi2intrin.h: Ditto.
13973 * config/i386/lzcntintrin.h: Ditto.
13974 * config/i386/smmintrin.h: Ditto.
13975 * config/i386/wmmintrin.h: Ditto.
13976 * config/i386/x86intrin.h: Remove all header include guards.
13977 * config/i386/prfchwintrin.h: Ditto.
13978 * config/i386/pmmintrin.h: Ditto.
13979 * config/i386/tmmintrin.h: Ditto.
13980 * config/i386/xmmintrin.h: Ditto.
13981 * config/i386/popcntintrin.h: Ditto.
13982 * config/i386/rdseedintrin.h: Ditto.
13983 * config/i386/ammintrin.h: Ditto.
13984 * config/i386/emmintrin.h: Ditto.
13985 * config/i386/immintrin.h: Remove all header include guards.
13986 * config/i386/fma4intrin.h: Ditto.
13987 * config/i386/lwpintrin.h: Ditto.
13988 * config/i386/xopintrin.h: Ditto.
13989 * config/i386/ia32intrin.h: Ditto.
13990 * config/i386/avxintrin.h: Ditto.
13991 * config/i386/rtmintrin.h: Ditto.
13992 * config/i386/fmaintrin.h: Ditto.
13993 * config/i386/mm3dnow.h: Ditto.
13994
495e6879
ST
139952013-06-22 Sriraman Tallam <tmsriram@google.com>
13996
13997 * common/config/i386/i386-common.c: Handle LZCNT.
13998
11e8f687
AK
139992013-06-22 Andi Kleen <ak@linux.intel.com>
14000
14001 * doc/extend.texi: Use __atomic_store_n instead of
14002 __atomic_store in HLE example.
14003
12211b99 140042013-06-22 Oleg Endo <olegendo@gcc.gnu.org>
664dd721
OE
14005
14006 * config/sh/sh.c: Remove <cstdlib> workaround.
14007
927f908b
AK
140082013-06-21 Andi Kleen <ak@linux.intel.com>
14009
0ba6aff3 14010 * doc/extend.texi: Dont use __atomic_clear in HLE example. Fix typo.
927f908b 14011
cca410b6
AK
140122013-06-21 Andi Kleen <ak@linux.intel.com>
14013
14014 * doc/extend.texi: Document that __atomic_clear and
14015 __atomic_test_and_set should only be used with bool.
14016
d84db1ed
JH
140172013-06-20 Jan Hubicka <jh@suse.cz>
14018
14019 * gimple-fold.c (gimple_extract_devirt_binfo_from_cst): Use
14020 types_same_for_odr.
14021 * tree.c (decls_same_for_odr): New function.
14022 (same_for_edr): New function.
14023 (types_same_for_odr): New function.
14024 (get_binfo_at_offset): Use it.
14025 * tree.h (types_same_for_odr): Declare.
14026
12211b99 140272013-06-20 Oleg Endo <olegendo@gcc.gnu.org>
469a31a9
OE
14028 Jason Merrill <jason@redhat.com>
14029
14030 * system.h: Include <cstdlib> as well as <stdlib.h>.
14031
49b4ff63
UB
140322013-06-20 Uros Bizjak <ubizjak@gmail.com>
14033
14034 PR target/57655
14035 * config/i386/i386.c (construct_container): Report error if
14036 long double is used with disabled x87 float returns.
14037
7a89ad00
JH
140382013-06-20 Jan Hubicka <jh@suse.cz>
14039
14040 * lto-cgraph.c (input_symtab): Do not set cgraph state.
14041
12211b99 140422013-06-20 Joern Rennecke <joern.rennecke@embecosm.com>
bd280792
JR
14043
14044 PR rtl-optimization/57425
14045 PR rtl-optimization/57569
14046 * alias.c (write_dependence_p): Remove parameters mem_mode and
14047 canon_mem_addr. Add parameters x_mode, x_addr and x_canonicalized.
14048 Changed all callers.
14049 (canon_anti_dependence): Get comments and semantics in sync.
14050 Add parameter mem_canonicalized. Changed all callers.
14051 * rtl.h (canon_anti_dependence): Update prototype.
14052
d16e9a99
RB
140532013-06-20 Richard Biener <rguenther@suse.de>
14054
14055 * data-streamer-in.c (streamer_read_uhwi): Optimize single
14056 byte case, inline streamer_read_uchar and defer section
14057 overrun check.
14058
c3a9b91b
RB
140592013-06-20 Richard Biener <rguenther@suse.de>
14060
14061 PR tree-optimization/57584
14062 * tree-ssa-loop-niter.c (expand_simple_operations): Avoid including
14063 SSA names into the expanded expression that take part in
14064 abnormal coalescing.
14065
371e77e3 140662013-06-19 Sharad Singhai <singhai@google.com>
fc8a650e
SS
14067
14068 * gcov.c (print_usage): Handle new option.
14069 (process_args): Ditto.
14070 (get_gcov_intermediate_filename): New function.
14071 (output_intermediate_file): New function.
14072 (output_gcov_file): New function
14073 (generate_results): Handle new option.
14074 (release_function): Relase demangled name.
14075 (read_graph_file): Handle demangled name.
14076 (output_lines): Ditto.
14077 * doc/gcov.texi: Document gcov intermediate format.
14078
11f5b71a
VM
140792013-06-19 Vladimir Makarov <vmakarov@redhat.com>
14080
14081 PR bootstrap/57604
14082 * lra.c (emit_add3_insn, emit_add2_insn): New functions.
14083 (lra_emit_add): Use the functions. Add comment about Y as an
14084 address segment.
14085
e1f9340b
DE
140862013-06-19 David Edelsohn <dje.gcc@gmail.com>
14087
14088 PR driver/57652
14089 * collect2.c (collect_atexit): New.
14090 (collect_exit): Delete.
14091 (main): Register collect_atexit with atexit.
14092 (collect_wait): Change collect_exit to exit.
14093 (do_wait): Same.
14094 * collect2.h (collect_exit): Delete.
14095 * tlink.c (do_tlink): Rename exit to ret. Change collect_exit to exit.
14096
e3f9e0ac
WM
140972013-06-19 Wei Mi <wmi@google.com>
14098
14099 PR rtl-optimization/57518
14100 * ira.c (set_paradoxical_subreg): Set pdx_subregs[regno]
14101 if regno is used in paradoxical subreg.
14102 (update_equiv_regs): Check pdx_subregs[regno] before
14103 set a reg to be equivalent with a mem.
14104
a09f1a76
MK
141052013-06-19 Matthias Klose <doko@ubuntu.com>
14106
14107 PR driver/57651
14108 * file-find.h (find_a_file): Add a mode parameter.
14109 * file-find.c (find_a_file): Likewise.
14110 * gcc-ar.c (main): Call find_a_file with R_OK for the plugin,
14111 with X_OK for the executables.
14112 * collect2.c (main): Call find_a_file with X_OK.
14113
7d18b0ad
SE
141142013-06-19 Steve Ellcey <sellcey@mips.com>
14115
14116 PR target/56942
14117 * config/mips/mips.md (casesi_internal_mips16_<mode>):
14118 Use NEXT_INSN instead of next_real_insn.
14119
6a6dac52
JH
141202013-06-19 Jan Hubicka <jh@suse.cz>
14121
14122 * cgraph.h (const_value_known_p): Replace by ...
14123 (ctor_for_folding): .. this one.
14124 * cgraphunit.c (process_function_and_variable_attributes): Use it.
14125 * lto-cgraph.c (compute_ltrans_boundary): Use ctor_for_folding.
14126 * expr.c (expand_expr_real_1): Likewise.
14127 (string_constant): Likewise.
14128 * tree-ssa-loop-ivcanon.c (constant_after_peeling): Likewise.
14129 * ipa.c (process_references): Likewise.
14130 (symtab_remove_unreachable_nodes): Likewise.
14131 * ipa-inline-analysis.c (param_change_prob): Likewise.
14132 * gimple-fold.c (canonicalize_constructor_val): Likewise.
14133 (get_base_constructor): Likwise.
14134 * varpool.c (varpool_remove_node): Likewise.
14135 (varpool_remove_initializer): LIkewise.
14136 (dump_varpool_node): LIkwise.
14137 (const_value_known_p): Rewrite to ...
14138 (ctor_for_folding): ... this one.
14139
216c12ab
JJ
141402013-06-19 Jakub Jelinek <jakub@redhat.com>
14141
14142 PR driver/57651
14143 * gcc-ar.c (main): If not CROSS_DIRECTORY_STRUCTURE, look for
14144 PERSONALITY in $PATH derived prefixes.
14145
d53e2f99
JL
141462013-06-19 Jeff Law <law@redhat.com>
14147
ecdbf306
JL
14148 * tree-ssa-forwprop.c (simplify_bitwise_binary_boolean): Fix typo
14149 in comment.
14150
d53e2f99
JL
14151 * tree-ssa-forwprop.c (simplify_bitwise_binary_boolean): New function.
14152 (simplify_bitwise_binary): Use it to simpify certain binary ops on
14153 booleans.
14154
89e7fc04
SN
141552013-06-19 Sofiane Naci <sofiane.naci@arm.com>
14156
14157 * config/arm/vfp.md: Move VFP instruction classification documentation
14158 to ...
14159 * config/arm/arm.md: ... here. Update instruction classification
14160 documentation.
14161
5e7f6aaa
RE
141622013-06-19 Richard Earnshaw <rearnsha@arm.com>
14163
14164 arm.md (split for eq(reg, 0)): Add variants for ARMv5 and Thumb2.
14165 (peepholes for eq(reg, not-0)): Ensure condition register is dead after
14166 pattern. Use more efficient sequences on ARMv5 and Thumb2.
14167
5fdc1e5d
SB
141682013-06-19 Steven Bosscher <steven@gcc.gnu.org>
14169
14170 PR target/57609
14171 * config/s390/s390.c (s390_chunkify_start): Replace next_real_insn
14172 with NEXT_INSN. Use tablejump_p to check for jump table data
14173 insns.
14174
e952afba
PC
141752013-06-19 Paolo Carlini <paolo.carlini@oracle.com>
14176
14177 PR c++/56544
14178 * doc/cpp.texi [Standard Predefined Macros, __cplusplus]: Document
14179 that now in C++ the value is correct per the C++ standards.
14180
3fd0303a
RB
141812013-06-19 Richard Biener <rguenther@suse.de>
14182
14183 * expr.c (expand_expr_real_1): Use SCOPE_FILE_SCOPE_P to check
14184 for global context.
14185
b04e4b08
AK
141862013-06-19 Andreas Krebbel <Andreas.Krebbel@de.ibm.com>
14187
14188 Revert:
14189 2013-06-18 Andreas Krebbel <Andreas.Krebbel@de.ibm.com>
14190
14191 PR target/57609
14192 * config/s390/s390.c (s390_chunkify_start): Replace next_real_insn
14193 with next_active_insn.
14194
bef8491a
ST
141952013-06-18 Sriraman Tallam <tmsriram@google.com>
14196
14197 * ipa-inline.c (inline_always_inline_functions): Pretend always_inline
14198 functions are inlined during failures to flag an error.
14199 * tree-inline.c (expand_call_inline): Allow the error to be flagged
14200 in early inline pass.
22fc7d3d 14201
c21b257d
L
142022013-06-18 H.J. Lu <hongjiu.lu@intel.com>
14203
14204 * config/i386/i386.c (initial_ix86_tune_features): Fix a typo
14205 in comments.
14206
33255ae3
JB
142072013-06-18 Julian Brown <julian@codesourcery.com>
14208
14209 * config/arm/arm.c (neon_vector_mem_operand): Add strict argument.
14210 Permit virtual register pre-reload if !strict.
14211 (coproc_secondary_reload_class): Adjust for neon_vector_mem_operand
14212 change.
14213 * config/arm/arm-protos.h (neon_vector_mem_operand): Adjust
14214 prototype.
14215 * config/arm/neon.md (movmisalign<mode>): Use
14216 neon_perm_struct_or_reg_operand instead of
14217 neon_struct_or_register_operand.
14218 (*movmisalign<mode>_neon_load, *movmisalign<mode>_neon_store): Use
14219 neon_permissive_struct_operand instead of neon_struct_operand.
14220 * config/arm/constraints.md (Un, Um, Us): Adjust calls to
14221 neon_vector_mem_operand.
14222 * config/arm/predicates.md (neon_struct_operand): Adjust call to
14223 neon_vector_mem_operand.
14224 (neon_permissive_struct_operand): New.
14225 (neon_struct_or_register_operand): Rename to...
14226 (neon_perm_struct_or_reg_operand): This. Adjust call to
14227 neon_vector_mem_operand.
14228
d579fcda
RB
142292013-06-18 Richard Biener <rguenther@suse.de>
14230
14231 * Makefile.in (LTO_STREAMER_H): Add pointer-set.h dependency.
14232 * lto-streamer.h: Include pointer-set.h.
14233 (struct lto_decl_slot): Remove.
14234 (struct lto_tree_ref_encoder): Make tree_hash_table a pointer-map.
14235 Remove next_index entry.
14236 (lto_hash_decl_slot_node, lto_eq_decl_slot_node,
14237 lto_hash_type_slot_node, lto_eq_type_slot_node): Remove.
14238 (lto_init_tree_ref_encoder): Adjust.
14239 (lto_destroy_tree_ref_encoder): Likewise.
14240 * lto-section-out.c (lto_hash_decl_slot_node, lto_eq_decl_slot_node,
14241 lto_hash_type_slot_node, lto_eq_type_slot_node): Remove.
14242 (lto_output_decl_index): Adjust.
14243 (lto_new_out_decl_state): Likewise.
14244 (lto_record_function_out_decl_state): Likewise.
14245 * lto-streamer-out.c (copy_function): Likewise.
14246
31ee20ba
RB
142472013-06-18 Richard Biener <rguenther@suse.de>
14248
14249 * Makefile.in (cgraphunit.o): Add $(CFGLOOP_H) dependency.
14250 * cgraphunit.c: Include cfgloop.h.
14251 (init_lowered_empty_function): Initialize the loop tree.
14252 (assemble_thunk): Insert new BBs into loops.
14253
ac0511f2
RB
142542013-06-18 Richard Biener <rguenther@suse.de>
14255
14256 * tree-streamer.h (streamer_tree_cache_create): Adjust prototype.
14257 * tree-streamer.c (streamer_tree_cache_create): Make maintaining
14258 the map from cache entry to cache index optional.
14259 (streamer_tree_cache_replace_tree): Adjust accordingly.
14260 (streamer_tree_cache_append): Likewise.
14261 (streamer_tree_cache_delete): Likewise.
14262 * lto-streamer-in.c (lto_data_in_create): Do not maintain the
14263 streamer cache map from cache entry to cache index.
14264 * lto-streamer-out.c (create_output_block): Adjust.
14265
09485a08
SN
142662013-06-18 Sofiane Naci <sofiane.naci@arm.com>
14267
22fc7d3d
UB
14268 * config/arm/arm.md (attribute "insn"): Move multiplication and
14269 division attributes to...
09485a08
SN
14270 (attribute "type"): ... here. Remove mult.
14271 (attribute "mul32"): New attribute.
14272 (attribute "mul64"): Add umaal.
14273 (*arm_mulsi3): Update attributes.
14274 (*arm_mulsi3_v6): Likewise.
14275 (*thumb_mulsi3): Likewise.
14276 (*thumb_mulsi3_v6): Likewise.
14277 (*mulsi3_compare0): Likewise.
14278 (*mulsi3_compare0_v6): Likewise.
14279 (*mulsi_compare0_scratch): Likewise.
14280 (*mulsi_compare0_scratch_v6): Likewise.
14281 (*mulsi3addsi): Likewise.
14282 (*mulsi3addsi_v6): Likewise.
14283 (*mulsi3addsi_compare0): Likewise.
14284 (*mulsi3addsi_compare0_v6): Likewise.
14285 (*mulsi3addsi_compare0_scratch): Likewise.
14286 (*mulsi3addsi_compare0_scratch_v6): Likewise.
14287 (*mulsi3subsi): Likewise.
14288 (*mulsidi3adddi): Likewise.
14289 (*mulsi3addsi_v6): Likewise.
14290 (*mulsidi3adddi_v6): Likewise.
14291 (*mulsidi3_nov6): Likewise.
14292 (*mulsidi3_v6): Likewise.
14293 (*umulsidi3_nov6): Likewise.
14294 (*umulsidi3_v6): Likewise.
14295 (*umulsidi3adddi): Likewise.
14296 (*umulsidi3adddi_v6): Likewise.
14297 (*smulsi3_highpart_nov6): Likewise.
14298 (*smulsi3_highpart_v6): Likewise.
14299 (*umulsi3_highpart_nov6): Likewise.
14300 (*umulsi3_highpart_v6): Likewise.
14301 (mulhisi3): Likewise.
14302 (*mulhisi3tb): Likewise.
14303 (*mulhisi3bt): Likewise.
14304 (*mulhisi3tt): Likewise.
14305 (maddhisi4): Likewise.
14306 (*maddhisi4tb): Likewise.
14307 (*maddhisi4tt): Likewise.
14308 (maddhidi4): Likewise.
14309 (*maddhidi4tb): Likewise.
14310 (*maddhidi4tt): Likewise.
14311 (divsi3): Likewise.
14312 (udivsi3): Likewise.
14313 * config/arm/thumb2.md (thumb2_mulsi_short): Update attributes.
14314 (thumb2_mulsi_short_compare0): Likewise.
14315 (thumb2_mulsi_short_compare0_scratch): Likewise.
14316 * config/arm/arm1020e.md (1020mult1): Update attribute change.
14317 (1020mult2): Likewise.
14318 (1020mult3): Likewise.
14319 (1020mult4): Likewise.
14320 (1020mult5): Likewise.
14321 (1020mult6): Likewise.
22fc7d3d
UB
14322 * config/arm/cortex-a15.md (cortex_a15_mult32): Update attribute
14323 change.
09485a08
SN
14324 (cortex_a15_mult64): Likewise.
14325 (cortex_a15_sdiv): Likewise.
14326 (cortex_a15_udiv): Likewise.
14327 * config/arm/arm1026ejs.md (mult1): Update attribute change.
14328 (mult2): Likewise.
14329 (mult3): Likewise.
14330 (mult4): Likewise.
14331 (mult5): Likewise.
14332 (mult6): Likewise.
14333 * config/arm/marvell-pj4.md (pj4_ir_mul): Update attribute change.
14334 (pj4_ir_div): Likewise.
14335 * config/arm/arm1136jfs.md (11_mult1): Update attribute change.
14336 (11_mult2): Likewise.
14337 (11_mult3): Likewise.
14338 (11_mult4): Likewise.
14339 (11_mult5): Likewise.
14340 (11_mult6): Likewise.
14341 (11_mult7): Likewise.
14342 * config/arm/cortex-a8.md (cortex_a8_mul): Update attribute change.
14343 (cortex_a8_mla): Likewise.
14344 (cortex_a8_mull): Likewise.
14345 (cortex_a8_smulwy): Likewise.
14346 (cortex_a8_smlald): Likewise.
14347 * config/arm/cortex-m4.md (cortex_m4_alu): Update attribute change.
14348 * config/arm/cortex-r4.md (cortex_r4_mul_4): Update attribute change.
14349 (cortex_r4_mul_3): Likewise.
14350 (cortex_r4_mla_4): Likewise.
14351 (cortex_r4_mla_3): Likewise.
14352 (cortex_r4_smlald): Likewise.
14353 (cortex_r4_mull): Likewise.
14354 (cortex_r4_sdiv): Likewise.
14355 (cortex_r4_udiv): Likewise.
14356 * config/arm/cortex-a7.md (cortex_a7_mul): Update attribute change.
14357 (cortex_a7_idiv): Likewise.
14358 * config/arm/arm926ejs.md (9_mult1): Update attribute change.
14359 (9_mult2): Likewise.
14360 (9_mult3): Likewise.
14361 (9_mult4): Likewise.
14362 (9_mult5): Likewise.
14363 (9_mult6): Likewise.
14364 * config/arm/cortex-a53.md (cortex_a53_mul): Update attribute change.
14365 (cortex_a53_sdiv): Likewise.
14366 (cortex_a53_udiv): Likewise.
14367 * config/arm/fa726te.md (726te_mult_op): Update attribute change.
14368 * config/arm/fmp626.md (mp626_mult1): Update attribute change.
14369 (mp626_mult2): Likewise.
14370 (mp626_mult3): Likewise.
14371 (mp626_mult4): Likewise.
14372 * config/arm/fa526.md (526_mult1): Update attribute change.
14373 (526_mult2): Likewise.
14374 * config/arm/arm-generic.md (mult): Update attribute change.
14375 (mult_ldsched_strongarm): Likewise.
14376 (mult_ldsched): Likewise.
14377 (multi_cycle): Likewise.
14378 * config/arm/cortex-a5.md (cortex_a5_mul): Update attribute change.
14379 * config/arm/fa606te.md (606te_mult1): Update attribute change.
14380 (606te_mult2): Likewise.
14381 (606te_mult3): Likewise.
14382 (606te_mult4): Likewise.
14383 * config/arm/cortex-a9.md (cortex_a9_mult16): Update attribute change.
14384 (cortex_a9_mac16): Likewise.
14385 (cortex_a9_multiply): Likewise.
14386 (cortex_a9_mac): Likewise.
14387 (cortex_a9_multiply_long): Likewise.
14388 * config/arm/fa626te.md (626te_mult1): Update attribute change.
14389 (626te_mult2): Likewise.
14390 (626te_mult3): Likewise.
14391 (626te_mult4): Likewise.
14392
8951f345
RB
143932013-06-18 Richard Biener <rguenther@suse.de>
14394
14395 PR lto/57334
14396 * lto-symtab.c (lto_symtab_merge_decls): Process nodes properly.
14397
d8101d05
AK
143982013-06-18 Andreas Krebbel <Andreas.Krebbel@de.ibm.com>
14399
14400 PR target/57609
14401 * config/s390/s390.c (s390_chunkify_start): Replace next_real_insn
14402 with next_active_insn.
14403
69eff9da
AM
144042013-06-18 Alan Modra <amodra@gmail.com>
14405
14406 * config/rs6000/rs6000.h (enum data_align): New.
14407 (LOCAL_ALIGNMENT, DATA_ALIGNMENT): Use rs6000_data_alignment.
14408 (DATA_ABI_ALIGNMENT): Define.
14409 (CONSTANT_ALIGNMENT): Correct comment.
14410 * config/rs6000/rs6000-protos.h (rs6000_data_alignment): Declare.
14411 * config/rs6000/rs6000.c (rs6000_data_alignment): New function.
14412
8ed00d76
DM
144132013-06-17 David Malcolm <dmalcolm@redhat.com>
14414
14415 * ggc-page.c (ggc_pch_write_object) <d>: Remove erroneous
14416 ATTRIBUTE_UNUSED marking.
14417
3d33d06b
SN
144182013-06-17 Sofiane Naci <sofiane.naci@arm.com>
14419
14420 * config/aarch64/aarch64-simd.md (aarch64_dup_lane<mode>): Add r<-w
14421 alternative and update.
14422 (aarch64_dup_lanedi): Delete.
14423 * config/aarch64/arm_neon.h (vdup<bhsd>_lane_*): Update.
14424 * config/aarch64/aarch64-simd-builtins.def: Update.
14425
ee03e71d
RB
144262013-06-17 Richard Biener <rguenther@suse.de>
14427
14428 * lto-streamer.h (enum LTO_tags): Add LTO_tree_scc.
14429 (lto_input_scc): Declare.
14430 (lto_input_tree_1): Likewise.
14431 (struct lto_stats_d): Add num_tree_bodies_output and
14432 num_pickle_refs_output.
14433 * lto-streamer-in.c (lto_read_body): Use streamer_tree_cache_get_tree.
14434 (lto_read_tree_1): Split out from ...
14435 (lto_read_tree): ... this.
14436 (lto_input_scc): New function.
14437 (lto_input_tree_1): Split out from ...
14438 (lto_input_tree): ... this. Handle LTO_tree_scc.
14439 (lto_data_in_create): Create the streamer cache without hashes.
14440 * lto-streamer-out.c (create_output_block): Create the streamer
14441 cache with hashes when not doing WPA.
14442 (lto_write_tree_1): Split out from ...
14443 (lto_write_tree): ... this.
14444 (get_symbol_initial_value): New function.
14445 (lto_output_tree_1): Split out from ...
14446 (lto_output_tree): ... this. Write trees as series of SCCs
14447 using a DFS walk via DFS_write_tree.
14448 (struct sccs, struct scc_entry): New types.
14449 (next_dfs_num, sccstack, sccstate, sccstate_obstack): New globals.
14450 (DFS_write_tree_body): New function.
14451 (DFS_write_tree): Likewise.
14452 (hash_tree): Likewise.
14453 (scc_entry_compare): Likewise.
14454 (hash_scc): Likewise.
14455 (tree_is_indexable): DEBUG_EXPR_DECLs are local entities.
14456 * tree-streamer-in.c (lto_input_ts_list_tree_pointers): Stream
14457 TREE_CHAIN as regular reference.
14458 (streamer_read_integer_cst): Remove.
14459 (streamer_get_pickled_tree): Adjust.
14460 * tree-streamer-out.c (streamer_write_chain): Disable streaming
14461 of DECL_EXTERNALs in BLOCK_VARS for now.
14462 (write_ts_list_tree_pointers): Stream TREE_CHAIN as regular
14463 reference.
14464 * tree-streamer.c (streamer_tree_cache_add_to_node_array):
14465 Add hash value argument and record that if hashes are recorded
14466 in the cache.
14467 (streamer_tree_cache_insert_1): Adjust.
14468 (streamer_tree_cache_insert): Likewise.
14469 (streamer_tree_cache_insert_at): Rename to ...
14470 (streamer_tree_cache_replace_tree): ... this and adjust.
14471 (streamer_tree_cache_append): Adjust.
14472 (record_common_node): Likewise.
14473 (streamer_tree_cache_create): Add argument whether to
14474 record hash values together with trees.
14475 (streamer_tree_cache_delete): Adjust.
14476 * tree-streamer.h (struct streamer_tree_cache_d): Add
14477 vector of hashes.
14478 (streamer_read_integer_cst): Remove.
14479 (streamer_tree_cache_insert): Adjust.
14480 (streamer_tree_cache_append): Likewise.
14481 (streamer_tree_cache_insert_at): Rename to ...
14482 (streamer_tree_cache_replace_tree): ... this and adjust.
14483 (streamer_tree_cache_create): Add argument whether to record hashes.
14484 (streamer_tree_cache_get): Rename to ...
14485 (streamer_tree_cache_get_tree): ... this.
14486 (streamer_tree_cache_get_hash): New function.
14487 * tree.c (cache_integer_cst): New function.
14488 * tree.h (cache_integer_cst): Declare.
14489 (ANON_AGGRNAME_FORMAT, ANON_AGGRNAME_P): Move here from cp/cp-tree.h.
14490 * lto-symtab.c (lto_varpool_replace_node): Only release
14491 DECL_INITIAL of non-prevailing decls.
14492 * varpool.c (varpool_remove_initializer): Do not release
14493 DECL_INITIAL when we are still in CGRAPH_LTO_STREAMING.
14494
0de86a92
JU
144952013-06-16 Jürgen Urban <JuergenUrban@gmx.de>
14496
14497 * config/mips/mips.h (ISA_HAS_MUL3): Include TARGET_MIPS5900.
14498 (ISA_HAS_MULT, ISA_HAS_DMULT, ISA_HAS_DIV, ISA_HAS_DDIV): New macros.
14499 * config/mips/mips.md (mul<mode>3, mul<mode>3_internal)
14500 (mul<mode>3_r4000): Require ISA_HAS_<D>MULT.
14501 (mul<mode>3_mul3): Handle TARGET_MIPS5900.
14502 (mulsidi3_64bit_dmul): Remove redundant TARGET_64BIT test.
14503 (<su>muldi3_highpart, <su>muldi3_highpart_internal, <u>mulditi3)
14504 (<u>mulditi3_internal, <u>mulditi3_r4000): Require ISA_HAS_DMULT
14505 instead of TARGET_64BIT.
14506 (divmod<mode>4, udivmod<mode>4, <u>divmod<GPR:mode>4_hilo_<HILO:mode>):
14507 Require ISA_HAS_<D>DIV.
14508
74240413
RS
145092013-06-16 Richard Sandiford <rdsandiford@googlemail.com>
14510
14511 * config.gcc (mips*-mti-linux*, mips64*-*-linux*, mipsisa64*-*-linux*)
14512 (mips*-*-linux*): Move default with_llsc setting to where other
14513 defaults are set.
14514 (mips*-*-vxworks*): Move with_arch default from with_cpu block to
14515 with_arch block.
14516 (mips64r5900-*-*, mips64r5900el-*-*, mipsr5900-*-*, mipsr5900el-*-*):
14517 Likewise. Remove default with_tune setting. Move default float
14518 setting to its own block. Handle with_llsc in the same block as above.
14519
12211b99 145202013-06-16 Joern Rennecke <joern.rennecke@embecosm.com>
393f9fed
JR
14521
14522 PR rtl-optimization/57425
14523 PR rtl-optimization/57569
299a5f6a 14524 * alias.c (write_dependence_p): Add new parameters mem_mode,
393f9fed
JR
14525 canon_mem_addr and mem_canonicalized. Change type of writep to bool.
14526 Changed all callers.
14527 (canon_anti_dependence): New function.
14528 * cse.c (check_dependence): Use canon_anti_dependence.
14529 * cselib.c (cselib_invalidate_mem): Likewise.
14530 * rtl.h (canon_anti_dependence): Declare.
14531
b259d352
JU
145322013-06-16 Jürgen Urban <JuergenUrban@gmx.de>
14533
14534 * config/mips/mips.h (ISA_HAS_LL_SC): Exclude TARGET_MIPS5900.
14535 * config/mips/mips.c (mips_start_ll_sc_sync_block): Output
14536 ".set mips3" for 64-bit targets.
14537
4de09b85
DC
145382013-06-15 Dehao Chen <dehao@google.com>
14539
14540 * tree-flow.h (gimple_check_call_matching_types): Add new argument.
14541 * gimple-low.c (gimple_check_call_matching_types): Likewise.
14542 (gimple_check_call_args): Likewise.
14543 * value-prof.c (check_ic_target): Likewise.
14544 * ipa-inline.c (early_inliner): Likewise.
14545 * ipa-prop.c (update_indirect_edges_after_inlining): Likewise.
14546 * cgraph.c (cgraph_create_edge_1): Likewise.
14547 (cgraph_make_edge_direct): Likewise.
14548
d1aa4a76
MM
145492013-06-14 Michael Meissner <meissner@linux.vnet.ibm.com>
14550
14551 PR target/57615
14552 * config/rs6000/rs6000.md (mov<mode>_ppc64): Call
14553 rs6000_output_move_128bit to handle emitting quad memory
14554 operations. Set attribute length to 8 bytes.
14555
12211b99 145562013-06-14 Vidya Praveen <vidyapraveen@arm.com>
24244e4d
VP
14557
14558 * config/aarch64/aarch64-simd.md (aarch64_<su>mlal_lo<mode>):
14559 New pattern.
14560 (aarch64_<su>mlal_hi<mode>, aarch64_<su>mlsl_lo<mode>): Likewise.
14561 (aarch64_<su>mlsl_hi<mode>, aarch64_<su>mlal<mode>): Likewise.
14562 (aarch64_<su>mlsl<mode>): Likewise.
14563
f5273574
MS
145642013-06-14 Mike Stump <mikestump@comcast.net>
14565
14566 * Makefile.in (TARGET_H): Add insn-codes.h.
14567
ef7e7ab2
AM
145682013-06-14 Alan Modra <amodra@gmail.com>
14569
14570 PR middle-end/57134
14571 PR middle-end/57586
14572 * expr.c (expand_expr_real_1 <normal_inner_ref>): Pass
14573 EXPAND_MEMORY and EXPAND_WRITE to recursive call. Don't use
14574 bitfield expansion when EXPAND_MEMORY.
14575 (expand_expr_real_1 <VIEW_CONVERT_EXPR>): Pass modifier likewise.
14576
135faab6
MM
145772013-06-13 Michael Meissner <meissner@linux.vnet.ibm.com>
14578
14579 * config/rs6000/rs6000.c (rs6000_option_override_internal): Move
14580 test for clearing quad memory on 32-bit later.
14581
948a1fd9
MG
145822013-06-13 Marc Glisse <marc.glisse@inria.fr>
14583
14584 * fold-const.c (negate_expr_p): Handle VECTOR_CST.
14585 (fold_negate_expr): Likewise.
14586 (fold_real_zero_addition_p): Handle vectors.
14587 (fold_binary_loc) <PLUS_EXPR, MINUS_EXPR>: Likewise.
14588
0597fb9c
AM
145892013-06-14 Alan Modra <amodra@gmail.com>
14590
14591 * varasm.c (force_const_mem): Revert 2013-06-07 change.
14592
e5b962d0
JH
145932013-06-13 Jan Hubicka <jh@suse.cz>
14594
14595 * ipa.c (cgraph_externally_visible_p, varpool_externally_visible_p):
14596 Local comdats are not externally visible.
14597 * symtab.c (dump_symtab_base): Dump externally visible.
14598 (verify_symtab_base): Verify back links in the symtab hash.
14599
a8355e51
BC
146002013-06-13 Bin Cheng <bin.cheng@arm.com>
14601
14602 * fold-const.c (operand_equal_p): Consider NOP_EXPR and
14603 CONVERT_EXPR as equal nodes.
14604
8f7e6e33
BC
146052013-06-13 Bin Cheng <bin.cheng@arm.com>
14606
22fc7d3d 14607 * rtlanal.c (noop_move_p): Check the code to be executed for COND_EXEC.
8f7e6e33 14608
5d418483
MG
146092013-06-13 Marc Glisse <marc.glisse@inria.fr>
14610
14611 * tree-ssa-forwprop.c (simplify_bitwise_binary, associate_plusminus):
14612 Generalize to complex and vector.
14613 * tree.c (build_all_ones_cst): New function.
14614 * tree.h (build_all_ones_cst): Declare it.
14615
50751417
AM
146162013-06-13 Alan Modra <amodra@gmail.com>
14617
14618 * config/rs6000/rs6000.h (LONG_DOUBLE_LARGE_FIRST): Define.
14619 * config/rs6000/rs6000.md (signbittf2): New insn.
14620 (extenddftf2_internal): Use LONG_DOUBLE_LARGE_FIRST.
14621 (abstf2_internal, cmptf_internal2): Likewise.
14622 * config/rs6000/spe.md (spe_abstf2_cmp, spe_abstf2_tst): Likewise.
14623
4b02c962 146242013-06-12 Michael Meissner <meissner@linux.vnet.ibm.com>
12211b99
UB
14625 Pat Haugen <pthaugen@us.ibm.com>
14626 Peter Bergner <bergner@vnet.ibm.com>
4b02c962
MM
14627
14628 * config/rs6000/rs6000.c (emit_load_locked): Add support for
14629 power8 byte, half-word, and quad-word atomic instructions.
14630 (emit_store_conditional): Likewise.
14631 (rs6000_expand_atomic_compare_and_swap): Likewise.
14632 (rs6000_expand_atomic_op): Likewise.
14633
14634 * config/rs6000/sync.md (larx): Add new modes for power8.
14635 (stcx): Likewise.
14636 (AINT): New mode iterator to include TImode as well as normal
14637 integer modes on power8.
14638 (fetchop_pred): Use int_reg_operand instead of gpc_reg_operand so
14639 that VSX registers are not considered. Use AINT mode iterator
14640 instead of INT1 to allow inclusion of quad word atomic operations
14641 on power8.
14642 (load_locked<mode>): Likewise.
14643 (store_conditional<mode>): Likewise.
14644 (atomic_compare_and_swap<mode>): Likewise.
14645 (atomic_exchange<mode>): Likewise.
14646 (atomic_nand<mode>): Likewise.
14647 (atomic_fetch_<fetchop_name><mode>): Likewise.
14648 (atomic_nand_fetch<mode>): Likewise.
14649 (mem_thread_fence): Use gen_loadsync_<mode> instead of enumerating
14650 each type.
14651 (ATOMIC): On power8, add QImode, HImode modes.
14652 (load_locked<QHI:mode>_si): Varients of load_locked for QI/HI
14653 modes that promote to SImode.
14654 (load_lockedti): Convert TImode arguments to PTImode, so that we
14655 get a guaranteed even/odd register pair.
14656 (load_lockedpti): Likewise.
14657 (store_conditionalti): Likewise.
14658 (store_conditionalpti): Likewise.
14659
14660 * config/rs6000/rs6000.md (QHI): New mode iterator for power8
14661 atomic load/store instructions.
14662 (HSI): Likewise.
14663
710b6cc1
RS
146642013-06-12 Richard Sandiford <rdsandiford@googlemail.com>
14665
14666 * config/mips/mips.md (extended_mips16): Include GOT and constant-pool
14667 loads.
14668 (insn_count): New attribute, with most cases extracted from...
14669 (length): ...here. Redefine most cases in terms of insn_count.
14670 (single_insn): Delete.
14671 (can_delay): Use insn_count to check for single instructions.
14672 (*mul<mode>3_r4300, mul<mode>3_r4000, *mul_acc_si, *mul_acc_si_r3900)
14673 (*msac_using_macc, *mul_sub_si, <u>mulsidi3_32bit_r4000)
14674 (<u>mulsidi3_64bit_r4000, <su>muldi3_highpart_internal)
14675 (<su>mulsi3_highpart_split, <su>muldi3_highpart_internal)
14676 (<u>mulditi3_r4000, *div<mode>3, *recip<mode>3, divmod<mode>4)
14677 (udivmod<mode>4, sqrt<mode>2, *rsqrt<mode>a, *rsqrt<mode>b)
14678 (fix_truncdfsi2_macro, fix_truncsfsi2_macro, *lea_high64)
14679 (*lea64, cprestore_<mode>, clear_hazard_<mode>, <unnamed insn>)
14680 (casesi_internal_mips16_<mode>, *tls_get_tp_<mode>_split)
14681 (tls_get_tp_mips16, *tls_get_tp_mips16_call_<mode>): Use "insn_count"
14682 rather than "length".
14683 (tls_get_tp_<mode>): Likewise. Remove redundant "no_delay" attribute.
14684 * config/mips/mips-ps-3d.md (mips_c_cond_4s, mips_cabs_cond_4s):
14685 Use "insn_count" rather than "length".
14686 * config/mips/mips-dsp.md
14687 (mips_l<SHORT:size><u>x_ext<GPR:mode>_<P:mode>)
14688 (mips_l<GPR:size>x_<P:mode>, *mips_lw<u>x_<P:mode>_ext): Remove
14689 length attributes.
14690
1951f101
MG
146912013-06-12 Marc Glisse <marc.glisse@inria.fr>
14692
14693 PR tree-optimization/57361
14694 * tree-ssa-dse.c (dse_possible_dead_store_p): Handle self-assignment.
14695
8b033a8a
SN
146962013-06-12 Sofiane Naci <sofiane.naci@arm.com>
14697
22fc7d3d
UB
14698 * config/aarch64/aarch64-simd.md (aarch64_combine<mode>): Convert
14699 to split.
8b033a8a
SN
14700 (aarch64_simd_combine<mode>): New instruction expansion.
14701 * config/aarch64/aarch64-protos.h (aarch64_split_simd_combine): New
14702 function prototype.
14703 * config/aarch64/aarch64.c (aarch64_split_combine): New function.
14704 * config/aarch64/iterators.md (Vdbl): Add entry for DF.
14705
ca0f62a8
JH
147062013-06-12 Jan Hubicka <jh@suse.cz>
14707
14708 * cgraph.c (verify_edge_corresponds_to_fndecl): Be lax about
14709 decl has when in streaming stage.
14710 * lto-symtab.c (lto_symtab_merge_symbols): Likewise.
14711 * cgraph.h (cgraph_state): Add CGRAPH_LTO_STREAMING.
14712
12211b99 147132013-06-12 Roland Stigge <stigge@antcom.de>
cddd69eb
RS
14714
14715 PR target/57578
14716 * config/rs6000/t-linux (MULTIARCH_DIRNAME): Fix SPE version detection.
14717
bfdeda2c
JJ
147182013-06-12 Jakub Jelinek <jakub@redhat.com>
14719
14720 PR tree-optimization/57537
14721 * tree-vect-patterns.c (vect_recog_widen_mult_pattern): If
14722 vect_handle_widen_op_by_const, convert oprnd1 to half_type1.
14723
a4fa02d1
RB
147242013-06-12 Richard Biener <rguenther@suse.de>
14725
14726 * data-streamer.h (streamer_write_char_stream): CSE
14727 obs->current_pointer.
14728 * data-streamer-out.c (streamer_write_uhwi_stream): Inline
14729 streamer_write_char_stream manually and optimize the resulting loop.
14730 (streamer_write_hwi_stream): Likewise.
14731
bbf9ad07
JH
147322013-06-12 Jan Hubicka <jh@suse.cz>
14733
14734 * lto-symtab.c (lto_symtab_merge_symbols): Populate symtab hashtable.
14735 * cgraph.h (varpool_create_empty_node): Declare.
14736 * lto-cgraph.c (input_node, input_varpool_node): Forcingly create
14737 duplicated nodes.
14738 * symtab.c (symtab_unregister_node): Be lax about missin entries
14739 in node hash.
14740 (symtab_get_node): Update comment.
14741 * varpool.c (varpool_create_empty_node): Break out from ...
14742 (varpool_node_for_decl): ... here.
14743 * lto-streamer.h (lto_file_decl_data): Add RESOLUTION_MAP.
14744
a53f90ad
EB
147452013-06-12 Eric Botcazou <ebotcazou@adacore.com>
14746
14747 * expr.c (expand_expr_real_1) <TARGET_MEM_REF>: Use straight-line flow.
14748 <MEM_REF>: Use 'type' instead of TREE_TYPE (exp) and tidy up the first
14749 part. Use straight-line flow at the end.
14750 <COMPONENT_REF>: Remove superfluous else.
14751 <VIEW_CONVERT_EXPR>: Use 'type' instead of TREE_TYPE (exp).
14752
946f9306
JJ
147532013-06-12 Jakub Jelinek <jakub@redhat.com>
14754
14755 PR target/56564
14756 * varasm.c (decl_binds_to_current_def_p): Call binds_local_p
14757 target hook even for !TREE_PUBLIC decls. If no resolution info
14758 is available, return false for common and external decls.
14759
a7fca6f0
KP
147602013-06-12 Kaushik Phatak <kaushik.phatak@kpitcummins.com>
14761
14762 * config/rl78/constraints.md (U): New constraint.
14763 * config/rl78/rl78.md (*mulqi3_rl78,*mulhi3_rl78,*mulhi3_g13): Add
14764 valloc attribute.
14765
c7d8f446
MM
147662013-06-11 Michael Meissner <meissner@linux.vnet.ibm.com>
14767
14768 PR target/57589
14769 * config/rs6000/driver-rs6000.c (elf_platform): Make buffer static
14770 to allow returning address to AT_PLATFORM name.
14771
08346abd
JH
147722013-06-11 Jan Hubicka <jh@suse.cz>
14773
14774 * cgraph.c (cgraph_create_function_alias): Set weakref flag.
14775 * cgraph.h (symtab_node_base): Add weakref flag.
14776 * cgraphunit.c (cgraph_reset_node): Clear weakref flag.
14777 (handle_alias_pairs): Set weakref flag, do not set DECL_EXTERNAL.
14778 (output_weakrefs): Use weakref flag.
14779 * fold-const.c (simple_operand_p): Handle WEAK.
14780 * gimple-fold.c (can_refer_decl_in_current_unit_p): Drop weakref.
14781 * ipa.c (varpool_externally_visible_p): Drop weakref.
14782 (function_and_variable_visibility): Update comment; fix weakref
14783 sanity checks; do not clear DECL_WEAK on them.
14784 * lto-cgraph.c (lto_output_node): update.
14785 (lto_output_varpool_node): Update.
14786 (input_overwrite_node): Update.
14787 (input_node): Update.
14788 (input_varpool_node): Update.
14789 * lto-symtab.c (lto_symtab_symbol_p): Do not special case weakrefs.
14790 (lto_symtab_merge_symbols): Add sanity check.
14791 (lto_symtab_prevailing_decl): Do not special case weakrefs.
14792 * passes.c (rest_of_decl_compilation): Set static flag, too.
14793 * symtab.c (dump_symtab_base): Dump weakref.
14794 (verify_symtab_base): Sanity check weakrefs.
14795 (symtab_make_decl_local): Remove duplicated code.
14796 (symtab_alias_ultimate_target): Simplify.
14797 * varpool.c (varpool_create_variable_alias): Set weakref flag.
c7d8f446 14798
2e6a2f88
TV
147992013-06-11 Tom de Vries <tom@codesourcery.com>
14800
14801 * genautomata.c (gen_regexp_sequence): Handle els_num == -1. Handle
14802 sequence_vect == NULL.
14803
d9af3652
DD
148042013-06-11 DJ Delorie <dj@redhat.com>
14805
14806 * config/rl78/rl78.c (TARGET_UNWIND_WORD_MODE): Define.
14807 (rl78_unwind_word_mode): New.
14808
ffa66012
DM
148092013-06-11 David Malcolm <dmalcolm@redhat.com>
14810
14811 * final.c (debug_prefix_maps): Make static.
14812
783f2d35
DM
148132013-06-11 David Malcolm <dmalcolm@redhat.com>
14814
14815 * function.c (initial_trampoline): Remove stray copy.
14816
7af26def
SN
148172013-06-11 Sofiane Naci <sofiane.naci@arm.com>
14818
14819 * config/aarch64/aarch64-simd.md (move_lo_quad_<mode>): Update.
14820
97756c0e
MJ
148212013-06-11 Martin Jambor <mjambor@suse.cz>
14822
14823 * ipa-cp.c (ipa_get_indirect_edge_target_1): Check that param_index is
14824 within bounds at the beginning of the function.
14825
2c7b8bf6
AM
148262013-06-11 Alan Modra <amodra@gmail.com>
14827
14828 * varasm.c (get_section): Don't die on !DECL_P decl. Tidy error
14829 reporting.
14830 (get_named_section): Don't NULL !DECL_P decl.
14831
77cea46e
IZ
148322013-06-11 Igor Zamyatin <igor.zamyatin@intel.com>
14833
14834 * doc/invoke.texi (core-avx2): Document.
14835 (slm): Likewise.
14836 (atom): Updated with MOVBE.
14837
888f0920
RB
148382013-06-11 Richard Biener <rguenther@suse.de>
14839
22fc7d3d 14840 * collect2.c (main): Do not redirect ld stdout/stderr when debugging.
888f0920 14841
6b438d58
AB
148422013-06-11 Anton Blanchard <anton@samba.org>
14843
14844 * config/rs6000/rs6000.c (rs6000_adjust_atomic_subword): Calculate
14845 correct shift value in little-endian mode.
14846
3bc66938
JJ
148472013-06-11 Jakub Jelinek <jakub@redhat.com>
14848
14849 PR target/56564
14850 * varasm.c (get_variable_align): Move #endif to the right place.
14851
a5dc7f88
CC
148522013-06-10 Cary Coutant <ccoutant@google.com>
14853
14854 * dwarf2out.c (hash_external_ref): Use die_symbol or signature
14855 for hash so that hash table traversal order is deterministic.
14856
19be72ab 148572013-06-10 Michael Meissner <meissner@linux.vnet.ibm.com>
12211b99
UB
14858 Pat Haugen <pthaugen@us.ibm.com>
14859 Peter Bergner <bergner@vnet.ibm.com>
19be72ab
MM
14860
14861 * config/rs6000/vector.md (GPR move splitter): Do not split moves
14862 of vectors in GPRS if they are direct moves or quad word load or
14863 store moves.
14864
14865 * config/rs6000/rs6000-protos.h (rs6000_output_move_128bit): Add
14866 declaration.
14867 (direct_move_p): Likewise.
14868 (quad_load_store_p): Likewise.
14869
14870 * config/rs6000/rs6000.c (enum rs6000_reg_type): Simplify register
14871 classes into bins based on the physical register type.
14872 (reg_class_to_reg_type): Likewise.
14873 (IS_STD_REG_TYPE): Likewise.
14874 (IS_FP_VECT_REG_TYPE): Likewise.
14875 (reload_fpr_gpr): Arrays to determine what insn to use if we can
14876 use direct move instructions.
14877 (reload_gpr_vsx): Likewise.
14878 (reload_vsx_gpr): Likewise.
14879 (rs6000_init_hard_regno_mode_ok): Precalculate the register type
14880 information that is a simplification of register classes. Also
14881 precalculate direct move reload helpers.
14882 (direct_move_p): New function to return true if the operation can
14883 be done as a direct move instruciton.
14884 (quad_load_store_p): New function to return true if the operation
14885 is a quad memory operation.
14886 (rs6000_legitimize_address): If quad memory, only allow register
14887 indirect for TImode addresses.
14888 (rs6000_legitimate_address_p): Likewise.
14889 (enum reload_reg_type): Delete, replace with rs6000_reg_type.
14890 (rs6000_reload_register_type): Likewise.
14891 (register_to_reg_type): Return register type.
14892 (rs6000_secondary_reload_simple_move): New helper function for
14893 secondary reload and secondary memory needed to identify anything
14894 that is a simple move, and does not need reloading.
14895 (rs6000_secondary_reload_direct_move): New helper function for
14896 secondary reload to identify cases that can be done with several
14897 instructions via the direct move instructions.
14898 (rs6000_secondary_reload_move): New helper function for secondary
14899 reload to identify moves between register types that can be done.
14900 (rs6000_secondary_reload): Add support for quad memory operations
14901 and for direct move.
14902 (rs6000_secondary_memory_needed): Likewise.
14903 (rs6000_debug_secondary_memory_needed): Change argument names.
14904 (rs6000_output_move_128bit): New function to return the move to
14905 use for 128-bit moves, including knowing about the various
14906 limitations of quad memory operations.
14907
14908 * config/rs6000/vsx.md (vsx_mov<mode>): Add support for quad
14909 memory operations. call rs6000_output_move_128bit for the actual
14910 instruciton(s) to generate.
14911 (vsx_movti_64bit): Likewise.
14912
14913 * config/rs6000/rs6000.md (UNSPEC_P8V_FMRGOW): New unspec values.
14914 (UNSPEC_P8V_MTVSRWZ): Likewise.
14915 (UNSPEC_P8V_RELOAD_FROM_GPR): Likewise.
14916 (UNSPEC_P8V_MTVSRD): Likewise.
14917 (UNSPEC_P8V_XXPERMDI): Likewise.
14918 (UNSPEC_P8V_RELOAD_FROM_VSX): Likewise.
14919 (UNSPEC_FUSION_GPR): Likewise.
14920 (FMOVE128_GPR): New iterator for direct move.
22fc7d3d 14921 (f32_lv): New mode attribute for load/store of SFmode/SDmode values.
19be72ab
MM
14922 (f32_sv): Likewise.
14923 (f32_dm): Likewise.
14924 (zero_extend<mode>di2_internal1): Add support for power8 32-bit
14925 loads and direct move instructions.
14926 (zero_extendsidi2_lfiwzx): Likewise.
14927 (extendsidi2_lfiwax): Likewise.
14928 (extendsidi2_nocell): Likewise.
14929 (floatsi<mode>2_lfiwax): Likewise.
14930 (lfiwax): Likewise.
14931 (floatunssi<mode>2_lfiwzx): Likewise.
14932 (lfiwzx): Likewise.
14933 (fix_trunc<mode>_stfiwx): Likewise.
14934 (fixuns_trunc<mode>_stfiwx): Likewise.
14935 (mov<mode>_hardfloat, 32-bit floating point): Likewise.
14936 (mov<move>_hardfloat64, 64-bit floating point): Likewise.
14937 (parity<mode>2_cmpb): Set length/type attr.
14938 (unnamed shift right patterns, mov<mode>_internal2): Change type attr
14939 for 'mr.' to fast_compare.
14940 (bpermd_<mode>): Change type attr to popcnt.
14941 (p8_fmrgow_<mode>): New insns for power8 direct move support.
14942 (p8_mtvsrwz_1): Likewise.
14943 (p8_mtvsrwz_2): Likewise.
14944 (reload_fpr_from_gpr<mode>): Likewise.
14945 (p8_mtvsrd_1): Likewise.
14946 (p8_mtvsrd_2): Likewise.
14947 (p8_xxpermdi_<mode>): Likewise.
14948 (reload_vsx_from_gpr<mode>): Likewise.
14949 (reload_vsx_from_gprsf): Likewise.
14950 (p8_mfvsrd_3_<mode>): LIkewise.
14951 (reload_gpr_from_vsx<mode>): Likewise.
14952 (reload_gpr_from_vsxsf): Likewise.
14953 (p8_mfvsrd_4_disf): Likewise.
14954 (multi-word GPR splits): Do not split direct moves or quad memory
14955 operations.
14956
16876bdc
DM
149572013-06-10 David Malcolm <dmalcolm@redhat.com>
14958
14959 * tree-into-ssa.c (interesting_blocks): Make static.
14960
df8a1d28
JJ
149612013-06-10 Jakub Jelinek <jakub@redhat.com>
14962
14963 PR target/56564
14964 * varasm.c (align_variable): Don't use DATA_ALIGNMENT or
14965 CONSTANT_ALIGNMENT if !decl_binds_to_current_def_p (decl).
14966 Use DATA_ABI_ALIGNMENT for that case instead if defined.
14967 (get_variable_align): New function.
14968 (get_variable_section, emit_bss, emit_common,
14969 assemble_variable_contents, place_block_symbol): Use
14970 get_variable_align instead of DECL_ALIGN.
14971 (assemble_noswitch_variable): Add align argument, use it
14972 instead of DECL_ALIGN.
14973 (assemble_variable): Adjust caller. Use get_variable_align
14974 instead of DECL_ALIGN.
14975 * config/i386/i386.h (DATA_ALIGNMENT): Adjust x86_data_alignment
14976 caller.
14977 (DATA_ABI_ALIGNMENT): Define.
14978 * config/i386/i386-protos.h (x86_data_alignment): Adjust prototype.
14979 * config/i386/i386.c (x86_data_alignment): Add opt argument. If
14980 opt is false, only return the psABI mandated alignment increase.
14981 * config/c6x/c6x.h (DATA_ALIGNMENT): Renamed to...
14982 (DATA_ABI_ALIGNMENT): ... this.
14983 * config/mmix/mmix.h (DATA_ALIGNMENT): Renamed to...
14984 (DATA_ABI_ALIGNMENT): ... this.
14985 * config/mmix/mmix.c (mmix_data_alignment): Adjust function comment.
14986 * config/s390/s390.h (DATA_ALIGNMENT): Renamed to...
14987 (DATA_ABI_ALIGNMENT): ... this.
14988 * doc/tm.texi.in (DATA_ABI_ALIGNMENT): Document.
14989 * doc/tm.texi: Regenerated.
14990
57b29ca6
UB
149912013-06-10 Uros Bizjak <ubizjak@gmail.com>
14992
14993 * config/alpha/alpha.c (alpha_emit_xfloating_compare): Also use
14994 cmp_code to construct REG_EQUAL note.
14995
3e56ed50
JJ
149962013-06-09 Jakub Jelinek <jakub@redhat.com>
14997
14998 PR target/57568
14999 * config/i386/i386.md (TARGET_READ_MODIFY_WRITE peepholes): Ensure
15000 that operands[2] doesn't overlap with operands[0].
15001
74fe2a1d
DE
150022013-06-09 David Edelsohn <dje.gcc@gmail.com>
15003 Jan Hubicka <jh@suse.cz>
15004
15005 * config/rs6000/rs6000.c (print_operand, 'z'): Remove historical
15006 hack to mark symbols as used.
15007
ed52a84e
VM
150082013-06-08 Vladimir Makarov <vmakarov@redhat.com>
15009
15010 PR rtl-optimization/57559
15011 * lra-constraints.c (process_alt_operands): Don't discourage
15012 memory with known offset for offsetable memory constraint.
15013 * lra.c (lra_emit_add): Exchange y and z for 2-op add insn.
15014
299316ed
EB
150152013-06-08 Eric Botcazou <ebotcazou@adacore.com>
15016
15017 * varasm.c (struct oc_local_state): Reorder fields.
15018 (output_constructor_bitfield): Replace OUTER parameter with BIT_OFFSET
15019 and adjust accordingly.
15020 (output_constructor): Reorder initialization code and adjust call to
15021 output_constructor_bitfield.
15022
88ac60d3
JH
150232013-06-07 Jan Hubicka <jh@suse.cz>
15024
15025 * symtab.c (symtab_resolve_alias): Do not remove alias attribute.
15026
4cdce1a8
DM
150272013-06-07 David Malcolm <dmalcolm@redhat.com>
15028
15029 * tree-object-size.c (unknown): Make const.
15030
b5e0425c
AK
150312013-06-07 Andreas Krebbel <Andreas.Krebbel@de.ibm.com>
15032
15033 * config/s390/s390.md (cpu_facility): Add cpu_zarch.
15034 ("*movmem_short", "*clrmem_short", "*cmpmem_short): Use cpu_zarch
15035 for last alternative in the cpu_facility attribute.
15036
79678d04
KT
150372013-06-07 Kyrylo Tkachov <kyrylo.tkachov@arm.com>
15038
15039 PR target/56315
15040 * config/arm/arm.md (*xordi3_insn): Change to insn_and_split.
15041 (xordi3): Change operand 2 constraint to arm_xordi_operand.
15042 * config/arm/arm.c (const_ok_for_dimode_op): Handle XOR.
15043 * config/arm/constraints.md (Dg): New constraint.
15044 * config/arm/neon.md (xordi3_neon): Remove.
15045 (neon_veor<mode>): Generate xordi3 instead of xordi3_neon.
15046 * config/arm/predicates.md (arm_xordi_operand): New predicate.
15047
b31ddbdb
KT
150482013-06-07 Kyrylo Tkachov <kyrylo.tkachov@arm.com>
15049
15050 * config/arm/arm.md (anddi3_insn): Remove duplicate alternatives.
15051 Clean up alternatives.
15052
8f90b7d4
AM
150532013-06-07 Alan Modra <amodra@gmail.com>
15054
15055 * config/rs6000/rs6000.c (setup_incoming_varargs): Round up
15056 va_list_gpr_size.
15057
961ce119
AM
150582013-06-07 Alan Modra <amodra@gmail.com>
15059
15060 * varasm.c (force_const_mem): Assert mode is not VOID or BLK.
15061
33e49835
KT
150622013-06-07 Kyrylo Tkachov <kyrylo.tkachov@arm.com>
15063
15064 * config/arm/constraints.md (Df): New constraint.
15065 * config/arm/arm.md (iordi3_insn): Use Df constraint instead of De.
15066 Correct length attribute for last two alternatives.
15067
acd17ae6
AM
150682013-06-07 Alan Modra <amodra@gmail.com>
15069
15070 * config/rs6000/rs6000.c (rs6000_option_override_internal): Don't
15071 override user -mfp-in-toc.
15072 (offsettable_ok_by_alignment): Consider just the current access
15073 rather than the whole object, unless BLKmode. Handle
15074 CONSTANT_POOL_ADDRESS_P constants that lack a decl too.
15075 (use_toc_relative_ref): Allow CONSTANT_POOL_ADDRESS_P constants
15076 for -mcmodel=medium.
15077 * config/rs6000/linux64.h (SUBSUBTARGET_OVERRIDE_OPTIONS): Don't
15078 override user -mfp-in-toc or -msum-in-toc. Default to
15079 -mno-fp-in-toc for -mcmodel=medium.
15080
73310b0e
DD
150812013-06-06 DJ Delorie <dj@redhat.com>
15082
15083 * config/rl78/rl78.c (rl78_valid_pointer_mode): New, implements
15084 TARGET_VALID_POINTER_MODE.
15085
0bd62dca 150862013-06-06 Michael Meissner <meissner@linux.vnet.ibm.com>
12211b99
UB
15087 Pat Haugen <pthaugen@us.ibm.com>
15088 Peter Bergner <bergner@vnet.ibm.com>
0bd62dca
MM
15089
15090 * doc/extend.texi (PowerPC AltiVec/VSX Built-in Functions):
15091 Document new power8 builtins.
15092
15093 * config/rs6000/vector.md (and<mode>3): Add a clobber/scratch of a
15094 condition code register, to allow 128-bit logical operations to be
15095 done in the VSX or GPR registers.
15096 (nor<mode>3): Use the canonical form for nor.
15097 (eqv<mode>3): Add expanders for power8 xxleqv, xxlnand, xxlorc,
15098 vclz*, and vpopcnt* vector instructions.
15099 (nand<mode>3): Likewise.
15100 (orc<mode>3): Likewise.
15101 (clz<mode>2): LIkewise.
15102 (popcount<mode>2): Likewise.
15103
15104 * config/rs6000/predicates.md (int_reg_operand): Rework tests so
15105 that only the GPRs are recognized.
15106
15107 * config/rs6000/rs6000-c.c (altivec_overloaded_builtins): Add
15108 support for new power8 builtins.
15109
15110 * config/rs6000/rs6000-builtin.def (xscvspdpn): Add new power8
15111 builtin functions.
15112 (xscvdpspn): Likewise.
15113 (vclz): Likewise.
15114 (vclzb): Likewise.
15115 (vclzh): Likewise.
15116 (vclzw): Likewise.
15117 (vclzd): Likewise.
15118 (vpopcnt): Likewise.
15119 (vpopcntb): Likewise.
15120 (vpopcnth): Likewise.
15121 (vpopcntw): Likewise.
15122 (vpopcntd): Likewise.
15123 (vgbbd): Likewise.
15124 (vmrgew): Likewise.
15125 (vmrgow): Likewise.
15126 (eqv): Likewise.
15127 (eqv_v16qi3): Likewise.
15128 (eqv_v8hi3): Likewise.
15129 (eqv_v4si3): Likewise.
15130 (eqv_v2di3): Likewise.
15131 (eqv_v4sf3): Likewise.
15132 (eqv_v2df3): Likewise.
15133 (nand): Likewise.
15134 (nand_v16qi3): Likewise.
15135 (nand_v8hi3): Likewise.
15136 (nand_v4si3): Likewise.
15137 (nand_v2di3): Likewise.
15138 (nand_v4sf3): Likewise.
15139 (nand_v2df3): Likewise.
15140 (orc): Likewise.
15141 (orc_v16qi3): Likewise.
15142 (orc_v8hi3): Likewise.
15143 (orc_v4si3): Likewise.
15144 (orc_v2di3): Likewise.
15145 (orc_v4sf3): Likewise.
15146 (orc_v2df3): Likewise.
15147
15148 * config/rs6000/rs6000.c (rs6000_option_override_internal): Only
15149 allow power8 quad mode in 64-bit.
15150 (rs6000_builtin_vectorized_function): Add support to vectorize
15151 ISA 2.07 count leading zeros, population count builtins.
15152 (rs6000_expand_vector_init): On ISA 2.07 use xscvdpspn to form
15153 V4SF vectors instead of xscvdpsp to avoid IEEE related traps.
15154 (builtin_function_type): Add vgbbd builtin function which takes an
15155 unsigned argument.
15156 (altivec_expand_vec_perm_const): Add support for new power8 merge
15157 instructions.
15158
15159 * config/rs6000/vsx.md (VSX_L2): New iterator for 128-bit types,
15160 that does not include TImdoe for use with 32-bit.
15161 (UNSPEC_VSX_CVSPDPN): Support for power8 xscvdpspn and xscvspdpn
15162 instructions.
15163 (UNSPEC_VSX_CVDPSPN): Likewise.
15164 (vsx_xscvdpspn): Likewise.
15165 (vsx_xscvspdpn): Likewise.
15166 (vsx_xscvdpspn_scalar): Likewise.
15167 (vsx_xscvspdpn_directmove): Likewise.
15168 (vsx_and<mode>3): Split logical operations into 32-bit and
15169 64-bit. Add support to do logical operations on TImode as well as
15170 VSX vector types. Allow logical operations to be done in either
15171 VSX registers or in general purpose registers in 64-bit mode. Add
15172 splitters if GPRs were used. For AND, add clobber of CCmode to
15173 allow use of ANDI on GPRs. Rewrite nor to use the canonical RTL
15174 encoding.
15175 (vsx_and<mode>3_32bit): Likewise.
15176 (vsx_and<mode>3_64bit): Likewise.
15177 (vsx_ior<mode>3): Likewise.
15178 (vsx_ior<mode>3_32bit): Likewise.
15179 (vsx_ior<mode>3_64bit): Likewise.
15180 (vsx_xor<mode>3): Likewise.
15181 (vsx_xor<mode>3_32bit): Likewise.
15182 (vsx_xor<mode>3_64bit): Likewise.
15183 (vsx_one_cmpl<mode>2): Likewise.
15184 (vsx_one_cmpl<mode>2_32bit): Likewise.
15185 (vsx_one_cmpl<mode>2_64bit): Likewise.
15186 (vsx_nor<mode>3): Likewise.
15187 (vsx_nor<mode>3_32bit): Likewise.
15188 (vsx_nor<mode>3_64bit): Likewise.
15189 (vsx_andc<mode>3): Likewise.
15190 (vsx_andc<mode>3_32bit): Likewise.
15191 (vsx_andc<mode>3_64bit): Likewise.
15192 (vsx_eqv<mode>3_32bit): Add support for power8 xxleqv, xxlnand,
15193 and xxlorc instructions.
15194 (vsx_eqv<mode>3_64bit): Likewise.
15195 (vsx_nand<mode>3_32bit): Likewise.
15196 (vsx_nand<mode>3_64bit): Likewise.
15197 (vsx_orc<mode>3_32bit): Likewise.
15198 (vsx_orc<mode>3_64bit): Likewise.
15199
15200 * config/rs6000/rs6000.h (VLOGICAL_REGNO_P): Update comment.
15201
15202 * config/rs6000/altivec.md (UNSPEC_VGBBD): Add power8 vgbbd
15203 instruction.
15204 (p8_vmrgew): Add power8 vmrgew and vmrgow instructions.
15205 (p8_vmrgow): Likewise.
15206 (altivec_and<mode>3): Add clobber of CCmode to allow AND using
15207 GPRs to be split under VSX.
15208 (p8v_clz<mode>2): Add power8 count leading zero support.
15209 (p8v_popcount<mode>2): Add power8 population count support.
15210 (p8v_vgbbd): Add power8 gather bits by bytes by doubleword
15211 support.
15212
15213 * config/rs6000/rs6000.md (eqv<mode>3): Add support for powerp eqv
15214 instruction.
15215
15216 * config/rs6000/altivec.h (vec_eqv): Add defines to export power8
15217 builtin functions.
15218 (vec_nand): Likewise.
15219 (vec_vclz): Likewise.
15220 (vec_vclzb): Likewise.
15221 (vec_vclzd): Likewise.
15222 (vec_vclzh): Likewise.
15223 (vec_vclzw): Likewise.
15224 (vec_vgbbd): Likewise.
15225 (vec_vmrgew): Likewise.
15226 (vec_vmrgow): Likewise.
15227 (vec_vpopcnt): Likewise.
15228 (vec_vpopcntb): Likewise.
15229 (vec_vpopcntd): Likewise.
15230 (vec_vpopcnth): Likewise.
15231 (vec_vpopcntw): Likewise.
15232
37684c46
VM
152332013-06-06 Vladimir Makarov <vmakarov@redhat.com>
15234
15235 PR rtl-optimization/57468
15236 * config/i386/i386.c (inline_secondary_memory_needed): Ignore
15237 spilled pseudos.
15238
d89ae27c
VM
152392013-06-06 Vladimir Makarov <vmakarov@redhat.com>
15240
15241 PR rtl-optimization/57459
15242 * lra-constraints.c (update_ebb_live_info): Fix typo for operand
15243 type when setting live regs.
15244
3597e113
VM
152452013-06-06 Vladimir Makarov <vmakarov@redhat.com>
15246
15247 * config/s390/s390.opt (mlra): New option.
15248 * config/s390/s390.c (s390_decompose_address): Check displacement
15249 for all registers for LRA.
15250 (s390_secondary_reload): Don't used secondary reloads for LRA.
15251 (s390_lra_p): New function.
15252 (TARGET_LRA_P): Define.
15253 * config/s390/s390.md (*movmem_short, *clrmem_short): Change value
15254 of attribute cpu_facility to zarch for the last alternative.
15255 (*cmpmem_short): Ditto.
15256
01037aeb
EB
152572013-06-06 Eric Botcazou <ebotcazou@adacore.com>
15258
15259 * config/arm/arm.c (arm_r3_live_at_start_p): New predicate.
15260 (arm_compute_static_chain_stack_bytes): Use it. Tidy up.
15261 (arm_expand_prologue): Likewise.
15262
3371a64f
TJ
152632013-06-06 Teresa Johnson <tejohnson@google.com>
15264
15265 PR c++/53743
15266 * ifcvt.c (find_if_case_1): Replace BB_COPY_PARTITION with assert
15267 as this is now done by redirect_edge_and_branch_force.
15268 * function.c (thread_prologue_and_epilogue_insns): Insert new bb after
15269 barriers, and fix interaction with splitting.
15270 * emit-rtl.c (try_split): Copy REG_CROSSING_JUMP notes.
15271 * cfgcleanup.c (try_forward_edges): Fix early return value to properly
15272 reflect changes made in the routine.
15273 * bb-reorder.c (emit_barrier_after_bb): Move to cfgrtl.c.
15274 (fix_up_fall_thru_edges): Remove incorrect check for bb layout order
15275 since this is called in cfglayout mode, and replace partition fixup
15276 with assert as that is now done by force_nonfallthru_and_redirect.
15277 (add_reg_crossing_jump_notes): Handle the fact that some jumps may
15278 already be marked with region crossing note.
15279 (insert_section_boundary_note): Make non-static, gate on flag
15280 has_bb_partition, rewrite to also check for multiple partitions.
15281 (rest_of_handle_reorder_blocks): Remove call to
15282 insert_section_boundary_note, now done later during free_cfg.
15283 (duplicate_computed_gotos): Don't duplicate partition crossing edge.
15284 * bb-reorder.h (insert_section_boundary_note): Declare.
15285 * Makefile.in (cfgrtl.o): Depend on bb-reorder.h
15286 * cfgrtl.c (rest_of_pass_free_cfg): If partitions exist
15287 invoke insert_section_boundary_note.
15288 (try_redirect_by_replacing_jump): Remove unnecessary
15289 check for region crossing note.
15290 (fixup_partition_crossing): New function.
15291 (rtl_redirect_edge_and_branch): Fixup partition boundaries.
15292 (emit_barrier_after_bb): Move here from bb-reorder.c, handle insertion
15293 in non-cfglayout mode.
15294 (force_nonfallthru_and_redirect): Fixup partition boundaries,
15295 remove old code that tried to do this. Emit barrier correctly
15296 when we are in cfglayout mode.
15297 (last_bb_in_partition): New function.
15298 (rtl_split_edge): Correctly fixup partition boundaries.
15299 (commit_one_edge_insertion): Remove old code that tried to
15300 fixup region crossing edge since this is now handled in
15301 split_block, and set up insertion point correctly since
15302 block may now end in a jump.
15303 (verify_hot_cold_block_grouping): Guard against checking when not in
15304 linearized RTL mode.
15305 (rtl_verify_edges): Add checks for incorrect/missing REG_CROSSING_JUMP
15306 notes.
15307 (rtl_verify_flow_info_1): Move verify_hot_cold_block_grouping to
15308 rtl_verify_flow_info, so not called in cfglayout mode.
15309 (rtl_verify_flow_info): Move verify_hot_cold_block_grouping here.
15310 (fixup_reorder_chain): Remove old code that attempted to fixup region
15311 crossing note as this is now handled in force_nonfallthru_and_redirect.
15312 (duplicate_insn_chain): Don't duplicate switch section notes.
15313 (rtl_can_remove_branch_p): Remove unnecessary check for region crossing
15314 note.
15315 * basic-block.h (emit_barrier_after_bb): Declare.
15316
66071e10
KT
153172013-06-06 Kyrylo Tkachov <kyrylo.tkachov@arm.com>
15318
15319 * config/arm/arm-fixed.md (add<mode>3,usadd<mode>3,ssadd<mode>3,
15320 sub<mode>3, ussub<mode>3, sssub<mode>3, arm_ssatsihi_shift,
15321 arm_usatsihi): Adjust alternatives for arm_restrict_it.
15322
0e26bf3d
KT
153232013-06-06 Kyrylo Tkachov <kyrylo.tkachov@arm.com>
15324
15325 * config/arm/arm-ldmstm.ml: Set "predicable_short_it" to "no"
15326 where appropriate.
15327 * config/arm/ldmstm.md: Regenerate.
15328
12b4e7ef
KT
153292013-06-06 Kyrylo Tkachov <kyrylo.tkachov@arm.com>
15330
15331 * config/arm/sync.md (atomic_loaddi_1):
15332 Disable predication for arm_restrict_it.
15333 (arm_load_exclusive<mode>): Likewise.
15334 (arm_load_exclusivesi): Likewise.
15335 (arm_load_exclusivedi): Likewise.
15336 (arm_load_acquire_exclusive<mode>): Likewise.
15337 (arm_load_acquire_exclusivesi): Likewise.
15338 (arm_load_acquire_exclusivedi): Likewise.
15339 (arm_store_exclusive<mode>): Likewise.
15340 (arm_store_exclusive<mode>): Likewise.
15341 (arm_store_release_exclusivedi): Likewise.
15342 (arm_store_release_exclusive<mode>): Likewise.
15343
15d16c8a
RB
153442013-06-06 Richard Biener <rguenther@suse.de>
15345
15346 * lto-streamer.h (enum LTO_tags): Move LTO_tree_pickle_reference
15347 after LTO_null.
15348 (lto_tag_is_tree_code_p): Adjust.
15349 (lto_tag_is_gimple_code_p): Likewise.
15350 (lto_gimple_code_to_tag): Likewise.
15351 (lto_tag_to_gimple_code): Likewise.
15352 (lto_tree_code_to_tag): Likewise.
15353 (lto_tag_to_tree_code): Likewise.
15354 * data-streamer.h (streamer_write_hwi_in_range): Use
15355 uhwi streaming to stream the normalized range.
15356 (streamer_read_hwi_in_range): Likewise.
15357
17a7fc37
KT
153582013-06-05 Kyrylo Tkachov <kyrylo.tkachov@arm.com>
15359
15360 * config/arm/arm.md (enabled_for_depr_it): New attribute.
15361 (predicable_short_it): Likewise.
15362 (predicated): Likewise.
15363 (enabled): Handle above.
15364 (define_cond_exec): Set predicated attribute to yes.
15365
b57ca59b
MS
153662013-06-05 Mike Stump <mikestump@comcast.net>
15367
15368 * gdbinit.in (__FUNCTION__): Add.
15369
c1e183a9
UB
153702013-06-05 Uros Bizjak <ubizjak@gmail.com>
15371
15372 * config/alpha/alpha.c (alpha_emit_conditional_move): Swap all
15373 GE, GT, GEU and GTU compares, modulo DImode compares with zero.
15374
0378bddb
JH
153752013-06-05 Jan Hubicka <jh@suse.cz>
15376
15377 * varasm.c (mark_decl_referenced): Revert the removal until targets
15378 are fixed.
15379
8f8a2057
DE
153802013-06-05 David Edelsohn <dje.gcc@gmail.com>
15381
15382 * config/rs6000/rs6000.c (print_operand, 'z'): Use DECL_PRESERVE_P
15383 instead of mark_decl_referenced.
15384
edb983b2
JH
153852013-06-05 Jan Hubicka <jh@suse.cz>
15386
15387 * cgraph.c (cgraph_remove_node): Clear forced_by_abi.
15388 (cgraph_node_cannot_be_local_p_1): Honnor symbol.forced_by_abi
15389 and symtab_used_from_object_file_p.
15390 (cgraph_make_node_local_1): Clear forced_by_abi.
15391 (cgraph_can_remove_if_no_direct_calls_and): Use forced_by_abi
15392 * cgraph.h (symtab_node_base): Add forced_by_abi.
15393 (decide_is_variable_needed): Remove.
15394 (varpool_can_remove_if_no_refs): Honnor symbol.forced_by_abi.
15395 * cgraphunit.c (cgraph_decide_is_function_needed): Rename to ..
15396 (decide_is_symbol_needed): ... this one; handle symbols in general;
15397 always analyze virtuals; honnor forced_by_abi.
15398 (cgraph_finalize_function): Update.
15399 (varpool_finalize_decl): Update.
15400 (symbol_defined_and_needed): Remove.
15401 (analyze_functions): Update.
15402 * lto-cgraph.c (lto_output_node, lto_output_varpool_node,
15403 output_refs, input_overwrite_node): Handle forced_by_abi.
15404 * ipa.c (cgraph_address_taken_from_non_vtable_p): Rename to ...
15405 (address_taken_from_non_vtable_p): ... this one.
15406 (comdat_can_be_unshared_p_1): New function.
15407 (cgraph_comdat_can_be_unshared_p): Rename to ...
15408 (comdat_can_be_unshared_p): ... this one; handle symbols in general.
15409 (varpool_externally_visible_p): Use comdat_can_be_unshared_p.
15410 (function_and_variable_visibility): Clear forced_by_abi as needed.
15411 * trans-mem.c (ipa_tm_mark_forced_by_abi_node): New functoin.
15412 (ipa_tm_create_version_alias, ipa_tm_create_version): Update.
edb983b2
JH
15413 * symtab.c (dump_symtab_base): Dump forced_by_abi.
15414 * varpool.c (decide_is_variable_needed): Remove.
15415
9912dbe5
KT
154162013-06-05 Kyrylo Tkachov <kyrylo.tkachov@arm.com>
15417
15418 * config/arm/arm.c (MAX_INSN_PER_IT_BLOCK): New macro.
15419 (arm_option_override): Override arm_restrict_it where appropriate.
15420 (thumb2_final_prescan_insn): Use MAX_INSN_PER_IT_BLOCK.
15421 * config/arm/arm.opt (mrestrict-it): New command-line option.
15422 * doc/invoke.texi: Document -mrestrict-it.
15423
34c136b6
DM
154242013-06-05 David Malcolm <dmalcolm@redhat.com>
15425
15426 * tsan.c (tsan_atomic_table): Make const.
15427
a367df53
RB
154282013-06-05 Richard Biener <rguenther@suse.de>
15429
15430 * tree-streamer.c (streamer_tree_cache_insert_1): Update the
15431 index associated with the tree we are supposed to replace.
15432 * tree-streamer-out.c (pack_ts_base_value_fields): Output
15433 TREE_ASM_WRITTEN as zero for everything but SSA names.
15434
70d51a19
DM
154352013-06-05 David Malcolm <dmalcolm@redhat.com>
15436
15437 * tree-ssa-structalias.c (call_stmt_vars): Make static.
15438
b9bd2075
JH
154392013-06-04 Jan Hubicka <jh@suse.cz>
15440
15441 * lto-cgraph.c (get_alias_symbol): Remove weakref sanity check.
15442 (input_node, input_varpool_node): Handle correctly external same
15443 body aliases.
15444 * ipa.c (symtab_remove_unreachable_nodes): Do not remove external
15445 nodes at ltrans stage.
15446
aaae719d
JH
154472013-06-04 Jan Hubicka <jh@suse.cz>
15448
15449 * ipa-inline.c (update_caller_keys): Fix availability test.
15450 (update_callee_keys): Likewise.
15451 * symtab.c (symtab_alias_ultimate_target): Make availaiblity logic
15452 to follow ELF standard.
15453
107eea2c
JU
154542013-06-04 Jürgen Urban <JuergenUrban@gmx.de>
15455
15456 * config.gcc (mipsr5900-*-elf*, mipsr5900el-*-elf*, mips64r5900-*-elf*)
15457 (mips64r5900el-*-elf*): New configurations.
15458 * config/mips/mips-cpus.def (r5900): New processor.
15459 * config/mips/mips-tables.opt: Regenerate.
15460 * config/mips/mips.c (mips_rtx_cost_data): Add an R5900 entry.
15461 (mips_issue_rate): Handle PROCESSOR_R5900.
15462 (mips_reorg_process_insns): Force reorder mode for the R5900.
15463 * config/mips/mips.h (TARGET_MIPS5900): Define.
15464 (ISA_HAS_CONDMOVE, ISA_HAS_PREFETCH, ISA_HAS_HILO_INTERLOCKS): Include
15465 TARGET_MIPS5900.
15466 (ISA_HAS_LOAD_DELAY, ISA_HAS_XFER_DELAY, ISA_HAS_FCMP_DELAY): Exclude
15467 TARGET_MIPS5900.
15468 * config/mips/mips.md (processor): Add r5900.
15469 (MOVECC): Disallow CCmode conditions for TARGET_MIPS5900.
15470
b7342d25
IB
154712013-06-04 Ian Bolton <ian.bolton@arm.com>
15472
15473 * config/aarch64/aarch64.md (*mov<mode>_aarch64): Call
15474 into function to generate MOVI instruction.
c1e183a9 15475 * config/aarch64/aarch64.c (aarch64_simd_container_mode): New function.
b7342d25
IB
15476 (aarch64_preferred_simd_mode): Turn into wrapper.
15477 (aarch64_output_scalar_simd_mov_immediate): New function.
15478 * config/aarch64/aarch64-protos.h: Add prototype for above.
15479
81c2dfb9
IB
154802013-06-04 Ian Bolton <ian.bolton@arm.com>
15481
15482 * config/aarch64/aarch64.c (simd_immediate_info): Remove
15483 element_char member.
15484 (sizetochar): Return signed char.
15485 (aarch64_simd_valid_immediate): Remove elchar and other
15486 unnecessary variables.
15487 (aarch64_output_simd_mov_immediate): Take rtx instead of &rtx.
15488 Calculate element_char as required.
15489 * config/aarch64/aarch64-protos.h: Update and move prototype
15490 for aarch64_output_simd_mov_immediate.
15491 * config/aarch64/aarch64-simd.md (*aarch64_simd_mov<mode>):
15492 Update arguments.
15493
48063b9d
IB
154942013-06-04 Ian Bolton <ian.bolton@arm.com>
15495
15496 * config/aarch64/aarch64.c (simd_immediate_info): Struct to hold
15497 information completed by aarch64_simd_valid_immediate.
15498 (aarch64_legitimate_constant_p): Update arguments.
15499 (aarch64_simd_valid_immediate): Work with struct rather than many
15500 pointers.
15501 (aarch64_simd_scalar_immediate_valid_for_move): Update arguments.
15502 (aarch64_simd_make_constant): Update arguments.
15503 (aarch64_output_simd_mov_immediate): Work with struct rather than
15504 many pointers. Output immediate directly rather than as operand.
15505 * config/aarch64/aarch64-protos.h (aarch64_simd_valid_immediate):
15506 Update prototype.
15507 * config/aarch64/constraints.md (Dn): Update arguments.
15508
3ea63f60
IB
155092013-06-04 Ian Bolton <ian.bolton@arm.com>
15510
15511 * config/aarch64/aarch64.c (aarch64_simd_valid_immediate): No
15512 longer static.
15513 (aarch64_simd_immediate_valid_for_move): Remove.
15514 (aarch64_simd_scalar_immediate_valid_for_move): Update call.
15515 (aarch64_simd_make_constant): Update call.
15516 (aarch64_output_simd_mov_immediate): Update call.
15517 * config/aarch64/aarch64-protos.h (aarch64_simd_valid_immediate):
15518 Add prototype.
15519 * config/aarch64/constraints.md (Dn): Update call.
15520
d8edd899
IB
155212013-06-04 Ian Bolton <ian.bolton@arm.com>
15522
15523 * config/aarch64/aarch64.c (aarch64_simd_valid_immediate): Change
15524 return type to bool for prototype.
15525 (aarch64_legitimate_constant_p): Check for true instead of not -1.
15526 (aarch64_simd_valid_immediate): Fix up each return to return a bool.
15527 (aarch64_simd_immediate_valid_for_move): Update retval for bool.
15528
44b20bb8
CM
155292013-06-04 Catherine Moore <clm@codesourcery.com>
15530
15531 * config/mips/mips.opt (meva): New.
c1e183a9 15532 * config/mips/mips.h (TARGET_CPU_CPP_BUILTINS): Define __mips_eva.
44b20bb8
CM
15533 (ASM_SPEC): Handle -meva.
15534 * doc/invoke.texi (meva): Document.
15535
52befbd8
AM
155362013-06-04 Alan Modra <amodra@gmail.com>
15537
15538 * config/rs6000/rs6000.c (output_toc): Correct little-endian float
15539 constant output.
15540
aadaf24e
KT
155412013-06-04 Kyrylo Tkachov <kyrylo.tkachov@arm.com>
15542
15543 * rtl.def: Add extra fourth optional field to define_cond_exec.
15544 * gensupport.c (process_one_cond_exec): Process attributes from
15545 define_cond_exec.
15546 * doc/md.texi: Document fourth field in define_cond_exec.
15547
0cd9e9ee
EB
155482013-06-04 Eric Botcazou <ebotcazou@adacore.com>
15549
15550 * expmed.c (extract_bit_field_1): In the larger-than-a-word case, factor
15551 out the processing order as in store_bit_field_1.
15552
cb0f665d
JH
155532013-06-04 Jan Hubicka <jh@suse.cz>
15554
15555 PR middle-end/57500
15556 * cgraphunit.c (cgraph_process_same_body_aliases): Create
15557 non-VAR_DECL node if it does not exist yet.
15558
53984b9b
RS
155592013-06-03 Richard Sandiford <rdsandiford@googlemail.com>
15560
15561 * config.gcc (mipsisa64sr71k-*-elf*, mipsisa64sb1-*-elf*)
15562 (mipsisa64sb1el-*-elf*, mips64-*-elf*, mips64el-*-elf*)
15563 (mips64orion-*-elf*, mips64orionel-*-elf*): Remove
15564 target_cpu_default setting.
15565
5d318fd4
TJ
155662013-06-03 Teresa Johnson <tejohnson@google.com>
15567
15568 * dumpfile.c (opt_info_switch_p): Change -fopt-info
15569 default to -fopt-info=optimized instead of all.
15570 * doc/invoke.texi: Ditto.
15571 * tree-vectorizer.c (vectorize_loops): Emit loop vectorization
15572 success under MSG_OPTIMIZED_LOCATIONS, and use dump_printf_loc.
15573 (execute_vect_slp): Emit BB vectorization success under
15574 MSG_OPTIMIZED_LOCATIONS.
15575 * tree-vect-slp.c (vect_slp_transform_bb): Change
15576 MSG_OPTIMIZED_LOCATIONS to MSG_NOTE.
15577 * tree-vect-loop.c (vect_transform_loop): Ditto.
15578
ec3af349
JM
155792013-06-03 Jason Merrill <jason@redhat.com>
15580
15581 PR c++/57415
15582 * config/rs6000/rs6000-c.c (altivec_resolve_overloaded_builtin):
15583 Use TARGET_EXPR for C++.
15584
1baf9159
JJ
155852013-06-03 Jakub Jelinek <jakub@redhat.com>
15586
15587 PR rtl-optimization/57268
0cd9e9ee 15588 * sched-deps.c (sched_analyze_2): Don't flush_pending_lists
1baf9159
JJ
15589 if DEBUG_INSN_P (insn).
15590
15591 Reapply
15592 2013-05-31 Dinar Temirbulatov <dinar@kugelworks.com>
15593
15594 PR rtl-optimization/57268
15595 * sched-deps.c (sched_analyze_2): Flush dependence lists if
15596 the sum of the read and write lists exceeds MAX_PENDING_LIST_LENGTH.
15597
0f1d3965
YR
155982013-06-03 Yuri Rumyantsev <yuri.s.rumyantsev@intel.com>
15599
15600 * config/i386/i386.c (ix86_lea_outperforms): Fix formatting.
15601 (ix86_avoid_lea_for_addr): Likewise.
15602 (exact_dependency_1): Likewise.
15603 (ix86_adjust_cost): Likewise.
15604 (swap_top_of_ready_list): Fix formatting and !reload_completed check
15605 removed.
15606 (do_reorder_for_imul): Fix typo, formatting and
15607 !reload_completed check removed.
15608 (ix86_sched_reorder): Fix typo and formatting.
15609 (fold_builtin_cpu): Move M_INTEL_SLM at the end of processor types
15610 list.
15611
3bc30db4
SN
156122013-06-03 Sofiane Naci <sofiane.naci@arm.com>
15613
15614 * config/aarch64/aarch64.md (*movdi_aarch64): Define "simd" attribute.
15615
19cc6fac
EB
156162013-06-03 Eric Botcazou <ebotcazou@adacore.com>
15617
15618 * varasm.c (output_constant) <CONSTRUCTOR>: Minor formatting tweak.
15619 <STRING_CST>: Likewise.
15620 <VECTOR_CST>: Likewise.
15621
c68a6e08
JW
156222013-06-01 Janus Weil <janus@gcc.gnu.org>
15623 Mikael Morin <mikael@gcc.gnu.org>
15624
15625 * configure.ac: Add AC_HEADER_TIOCGWINSZ macro.
15626 * config.in: Regenerated.
15627 * configure: Regenerated.
15628
38e55e5c
JH
156292013-06-01 Jan Hubicka <jh@suse.cz>
15630
15631 PR middle-end/57366
15632 * cgraphunit.c (compile): When weakref is not supported,
15633 set up transparent aliases before final output pass.
15634 * varasm.c (assemble_alias): Do not try to do it here.
15635
eb51d2ff
JH
156362013-06-01 Jan Hubicka <jh@suse.cz>
15637
15638 PR middle-end/57467
15639 * passes.c (for_per_function): Skip unanalyzed functions.
15640
40a7fe1e
JH
156412013-06-01 Jan Hubicka <jh@suse.cz>
15642
15643 * lto-symtab.c (lto_symtab_merge_cgraph_nodes_1): Rename to ...
15644 (lto_symtab_merge_symbols_1): ... this one.
15645 (lto_symtab_merge_cgraph_nodes): Rename to ...
15646 (lto_symtab_merge_symbols): ... this one; simplify.
15647 * cgraph.c (same_body_aliases_done): Rename to ...
15648 (cpp_implicit_aliases_done): ... this one.
15649 (cgraph_create_function_alias): Update.
15650 (cgraph_same_body_alias): Update.
c1e183a9 15651 (dump_cgraph_node): Remove alias dumping; simplify thunk dumping.
40a7fe1e 15652 (verify_edge_corresponds_to_fndecl): Simplify.
c1e183a9 15653 * cgraph.h (symtab_node_base): Add cpp_implicit_alias, alias_target.
40a7fe1e
JH
15654 (cgraph_node): Remove same_body_alias.
15655 (varpool_node): Remove alias_of and extra_name_alias.
15656 (same_body_aliases_done): Rename to ..
15657 (cpp_implicit_aliases_done): ... this one.
15658 (symtab_alias_ultimate_target): Add default parameter.
15659 (symtab_resolve_alias): New function.
15660 (fixup_same_cpp_alias_visibility): Declare.
15661 (cgraph_function_node): Add default parameter.
15662 (cgraph_node_asm_name): Likewise.
15663 (cgraph_function_or_thunk_node): Add default parameter; do
15664 not ICE when it is NULL.
15665 (varpool_variable_node): Likewise.
15666 * tree-emutls.c (create_emultls_var): Update.
15667 (ipa_lower_emutls): Update.
15668 * cgraphunit.c (cgraph_decide_is_function_needed): Update.
15669 (cgraph_reset_node): Reset alias info.
15670 (cgraph_finalize_function): Update.
15671 (fixup_same_cpp_alias_visibility): Move to symtab.c.
15672 (analyze_function): Simplify.
15673 (cgraph_process_same_body_aliases): Simplify.
15674 (analyze_functions): Fixup same body aliases.
15675 (handle_alias_pairs): Simplify.
15676 (assemble_thunk): Update.
15677 (assemble_thunks_and_aliases): Update.
15678 (output_weakrefs): Rewrite.
15679 * lto-cgraph.c (lto_output_node): Rewrite alias handling.
15680 (lto_output_varpool_node): Likewise.
15681 (compute_ltrans_boundary): Remve assert.
15682 (get_alias_symbol): New functoin.
15683 (input_node): Rewrite alias handling.
15684 (input_varpool_node): Likewise.
15685 * ipa-pure-const.c (propagate_pure_const): Fix formating.
15686 * ipa.c (process_references): Handle weakrefs correctly.
15687 (symtab_remove_unreachable_nodes): Likewise.
15688 * trans-mem.c (get_cg_data): Update.
15689 (ipa_tm_create_version_alias): Update.
15690 (ipa_tm_execute): Update.
15691 * symtab.c (dump_symtab_base): Dump aliases.
15692 (verify_symtab_base): Verify aliases.
15693 (symtab_node_availability): New function.
15694 (symtab_alias_ultimate_target): Simplify.
15695 (fixup_same_cpp_alias_visibility): Move here from cgraphunit.c;
15696 handle all the fixup cases.
15697 (symtab_resolve_alias): New function.
15698 * passes.c (ipa_write_summaries): Handle weakrefs.
15699 * varpool.c (varpool_analyze_node): Simplify.
15700 (assemble_aliases): Update.
15701 (varpool_create_variable_alias): Simplify.
15702 (varpool_extra_name_alias): Simplify.
15703 * lto-streamer.h (lto_symtab_merge_cgraph_nodes): Rename to...
15704 (lto_symtab_merge_symbols): ... this one.
15705
21759881
DT
157062013-06-01 Dinar Temirbulatov <dinar@kugelworks.com>
15707
15708 Revert
15709 PR rtl-optimization/57268
15710 * sched-deps.c (sched_analyze_2): Flush dependence lists if
15711 the sum of the read and write lists exceeds MAX_PENDING_LIST_LENGTH.
15712
daf4e940
TB
157132013-06-01 Tobias Burnus <burnus@net-b.de>
15714
15715 Partially reverted:
15716 2013-05-31 Tobias Burnus <burnus@net-b.de>
15717
15718 PR middle-end/57073
15719 * tree-ssa-math-opts.c (execute_cse_sincos): Move check
15720 further up.
15721
c359d8d0
DT
157222013-05-31 Dinar Temirbulatov <dinar@kugelworks.com>
15723
15724 PR rtl-optimization/57268
15725 * sched-deps.c (sched_analyze_2): Flush dependence lists if
15726 the sum of the read and write lists exceeds MAX_PENDING_LIST_LENGTH.
15727
4172215d
EB
157282013-05-31 Eric Botcazou <ebotcazou@adacore.com>
15729
15730 * config/rs6000/predicates.md (rs6000_cbranch_operator): Accept some
15731 unordered comparison operators when -fno-trapping-math is in effect
15732 on the e500.
15733 * config/rs6000/rs6000.c (rs6000_generate_compare): Remove dead code
15734 and implement unordered comparison operators properly on the e500.
15735
a8c50132
EB
157362013-05-31 Eric Botcazou <ebotcazou@adacore.com>
15737
15738 * simplify-rtx.c (simplify_byte_swapping_operation): Use proper macro
15739 for constant scalar integers.
15740 (simplify_relational_operation_1): Likewise.
15741
44626634
SB
157422013-05-31 Segher Boessenkool <segher@kernel.crashing.org>
15743
15744 * config/rs6000/rs6000-opts.h (enum processor_type): Reorder.
15745 * config/rs6000/rs6000.md (cpu): Reorder. Split long line.
15746 Fix comment.
15747
55805e54
YR
157482013-05-31 Yuri Rumyantsev <yuri.s.rumyantsev@intel.com>
15749 Igor Zamyatin <igor.zamyatin@intel.com>
15750
15751 Silvermont (SLM) architecture performance tuning.
15752 * config/i386/i386.h (enum ix86_tune_indices): Add
15753 X86_TUNE_SPLIT_MEM_OPND_FOR_FP_CONVERTS.
15754 (TARGET_SPLIT_MEM_OPND_FOR_FP_CONVERTS): New define.
15755
15756 * config/i386/i386.c (initial_ix86_tune_features)
15757 <X86_TUNE_SPLIT_MEM_OPND_FOR_FP_CONVERTS>: Initialize.
15758 (ix86_lea_outperforms): Handle Silvermont tuning.
15759 (ix86_avoid_lea_for_add): Add new argument to ix86_lea_outperforms
15760 call.
15761 (ix86_use_lea_for_mov): Likewise.
15762 (ix86_avoid_lea_for_addr): Likewise.
15763 (ix86_lea_for_add_ok): Likewise.
15764 (exact_dependency_1): New function.
15765 (exact_store_load_dependency): Likewise.
15766 (ix86_adjust_cost): Handle Silvermont tuning.
15767 (do_reoder_for_imul): Likewise.
15768 (swap_top_of_ready_list): New function.
15769 (ix86_sched_reorder): Changed to handle Silvermont tuning.
15770
15771 * config/i386/i386.md (peepholes that split memory operand in fp
15772 converts): New.
15773
e19c9de2
MS
157742013-05-31 Marcus Shawcroft <marcus.shawcroft@arm.com>
15775
15776 * config/aarch64/aarch64.c (aarch64_load_symref_appropriately):
15777 Remove un-necessary braces.
15778
38e6c9a6
MS
157792013-05-31 Marcus Shawcroft <marcus.shawcroft@arm.com>
15780
15781 * config/aarch64/aarch64.c (aarch64_classify_symbol):
15782 Use SYMBOL_TINY_ABSOLUTE for AARCH64_CMODEL_TINY_PIC.
15783
e3530904
TB
157842013-05-31 Tobias Burnus <burnus@net-b.de>
15785
15786 PR middle-end/57073
c1e183a9 15787 * tree-ssa-math-opts.c (execute_cse_sincos): Move check further up.
e3530904 15788
b6af05a9
KT
157892013-05-31 Kyrylo Tkachov <kyrylo.tkachov@arm.com>
15790
15791 PR target/56315
15792 * config/arm/arm.c (const_ok_for_dimode_op): Handle IOR.
15793 * config/arm/arm.md (*iordi3_insn): Change to insn_and_split.
15794 * config/arm/neon.md (iordi3_neon): Remove.
15795 (neon_vorr<mode>): Generate iordi3 instead of iordi3_neon.
15796 * config/arm/predicates.md (imm_for_neon_logic_operand):
15797 Move to earlier in the file.
15798 (neon_logic_op2): Likewise.
15799 (arm_iordi_operand_neon): New predicate.
15800
f800c166
RB
158012013-05-31 Richard Biener <rguenther@suse.de>
15802
15803 PR tree-optimization/57478
15804 PR tree-optimization/57453
15805 * tree-vect-slp.c (vect_bb_slp_scalar_cost): Uses in PHI nodes
15806 are life as well.
15807
e571fa59
KP
158082013-05-31 Kaushik Phatak <kaushik.phatak@kpitcummins.com>
15809
15810 * config/rl78/rl78.md (mulqi3,mulhi3): New define_expands.
15811 (*mulqi3_rl78,*mulhi3_rl78,*mulhi3_g13): New define_insns.
15812
0fa6e0ef
TB
158132013-05-30 Tobias Burnus <burnus@net-b.de>
15814 Thomas Koenig <tkoenig@gcc.gnu.org>
15815
15816 PR middle-end/57073
15817 * tree-ssa-math-opts.c (execute_cse_sincos): Optimize
15818 powi (-1.0, k) to (k & 1) ? -1.0 : 1.0.
15819
4688ddf5
SB
158202013-05-30 Steven Bosscher <steven@gcc.gnu.org>
15821
0fa6e0ef 15822 * rtlanal.c (tablejump_p): Expect table and label to be adjacent.
4688ddf5 15823
3b9ceb4b
VM
158242013-05-30 Vladimir Makarov <vmakarov@redhat.com>
15825
15826 * target.def (register_usage_leveling_p): New hook.
15827 * targhooks.c (default_register_usage_leveling_p): New.
15828 * targhooks.h (default_register_usage_leveling_p): New prototype.
15829 * lra-assigns.c (register_usage_leveling_p): Use the hook.
15830 * doc/tm.texi.in (TARGET_REGISTER_USAGE_LEVELING_P): New hook.
15831 * doc/tm.texi: Update.
15832 * config/i386/i386.c (TARGET_REGISTER_USAGE_LEVELING_P): Define.
15833
65f9e789
IB
158342013-05-30 Ian Bolton <ian.bolton@arm.com>
15835
15836 * config/aarch64/aarch64.md (insv<mode>): New define_expand.
15837 (*insv_reg<mode>): New define_insn.
15838
12211b99 158392013-05-30 Joern Rennecke <joern.rennecke@embecosm.com>
ca035367
JR
15840
15841 PR rtl-optimization/57439
15842 * postreload.c (move2add_valid_value_p): Check that we have
15843 a zero subreg_regno_offset when accessing the register in
15844 the requested mode.
15845
0b871ccf
YR
158462013-05-30 Yuri Rumyantsev <yuri.s.rumyantsev@intel.com>
15847 Igor Zamyatin <igor.zamyatin@intel.com>
15848
15849 Silvermont (SLM) architecture pipeline model, tuning and
15850 insn selection.
15851 * config.gcc: Add slm config options and target.
15852
15853 * config/i386/slm.md: New.
15854
15855 * config/i386/driver-i386.c (host_detect_local_cpu): Check movbe.
15856
f43245d1 15857 * config/i386/i386-c.c (ix86_target_macros_internal): New case
0b871ccf
YR
15858 PROCESSOR_SLM.
15859 (ix86_target_macros_internal): Likewise.
15860
f43245d1 15861 * config/i386/i386.c (slm_cost): New cost.
0b871ccf
YR
15862 (m_SLM): New macro flag.
15863 (initial_ix86_tune_features): Set m_SLM.
15864 (x86_accumulate_outgoing_args): Likewise.
15865 (x86_arch_always_fancy_math_387): Likewise.
15866 (processor_target_table): Add slm cost.
15867 (cpu_names): Add slm cpu name.
15868 (x86_option_override_internal): Set SLM ISA.
15869 (ix86_issue_rate): New case PROCESSOR_SLM.
15870 (ia32_multipass_dfa_lookahead): Likewise.
15871 (fold_builtin_cpu): Add slm.
15872
15873 * config/i386/i386.h (TARGET_SLM): New target macro.
15874 (target_cpu_default): Add TARGET_CPU_DEFAULT_slm.
15875 (processor_type): Add PROCESSOR_SLM.
15876
15877 * config/i386/i386.md (cpu): Add new value "slm".
15878 (slm.md): Include slm.md.
15879
24d5b097
XG
158802013-05-30 Bernd Schmidt <bernds@codesourcery.com>
15881 Zhenqiang Chen <zhenqiang.chen@linaro.org>
15882
15883 * config/arm/arm-protos.h: Add and update function protos.
15884 * config/arm/arm.c (use_simple_return_p): New added.
15885 (thumb2_expand_return): Check simple_return flag.
15886 * config/arm/arm.md: Add simple_return and conditional simple_return.
15887 * config/arm/iterators.md: Add iterator for return and simple_return.
15888
c1cccc15
ZC
158892013-05-30 Zhenqiang Chen <zhenqiang.chen@linaro.org>
15890
15891 * config/arm/arm.c (arm_add_cfa_adjust_cfa_note): New added.
15892 (arm_emit_multi_reg_pop): Add REG_CFA_ADJUST_CFA notes.
15893 (arm_emit_vfp_multi_reg_pop): Likewise.
15894 (thumb2_emit_ldrd_pop): Likewise.
15895 (arm_expand_epilogue): Add misc REG_CFA notes.
15896 (arm_unwind_emit): Skip REG_CFA_ADJUST_CFA and REG_CFA_RESTORE.
15897
3a4f280b
LC
158982013-05-29 Lawrence Crowl <crowl@google.com>
15899
15900 * config/arm/t-arm: Update for below.
15901
15902 * config/arm/arm.c (arm_libcall_uses_aapcs_base::libcall_htab):
15903 Change type to hash_table. Update dependent calls and types.
15904
15905 * config/i386/t-cygming: Update for below.
15906
15907 * config/i386/t-interix: Update for below.
15908
15909 * config/i386/winnt.c (i386_pe_section_type_flags::htab):
15910 Change type to hash_table. Update dependent calls and types.
15911 (i386_find_on_wrapper_list::wrappers): Likewise.
15912
15913 * config/ia64/t-ia64: Update for below.
15914
15915 * config/ia64/ia64.c (bundle_state_table):
15916 Change type to hash_table. Update dependent calls and types.
15917
15918 * config/mips/mips.c (mips_reorg_process_insns::htab):
15919 Change type to hash_table. Update dependent calls and types.
15920
15921 * config/sol2.c (solaris_comdat_htab):
15922 Change type to hash_table. Update dependent calls and types.
15923
15924 * config/t-sol2: Update for above.
15925
1388a0e3
TJ
159262013-05-29 Teresa Johnson <tejohnson@google.com>
15927
15928 * passes.c (dump_passes): Use FOR_EACH_FUNCTION since
15929 functions are not yet marked as defined.
15930
a5965b52 159312013-05-29 Michael Meissner <meissner@linux.vnet.ibm.com>
12211b99
UB
15932 Pat Haugen <pthaugen@us.ibm.com>
15933 Peter Bergner <bergner@vnet.ibm.com>
a5965b52
MM
15934
15935 * config/rs6000/vector.md (VEC_I): Add support for new power8 V2DI
15936 instructions.
15937 (VEC_A): Likewise.
15938 (VEC_C): Likewise.
15939 (vrotl<mode>3): Likewise.
15940 (vashl<mode>3): Likewise.
15941 (vlshr<mode>3): Likewise.
15942 (vashr<mode>3): Likewise.
15943
15944 * config/rs6000/rs6000-c.c (altivec_overloaded_builtins): Add
15945 support for power8 V2DI builtins.
15946
15947 * config/rs6000/rs6000-builtin.def (abs_v2di): Add support for
15948 power8 V2DI builtins.
15949 (vupkhsw): Likewise.
15950 (vupklsw): Likewise.
15951 (vaddudm): Likewise.
15952 (vminsd): Likewise.
15953 (vmaxsd): Likewise.
15954 (vminud): Likewise.
15955 (vmaxud): Likewise.
15956 (vpkudum): Likewise.
15957 (vpksdss): Likewise.
15958 (vpkudus): Likewise.
15959 (vpksdus): Likewise.
15960 (vrld): Likewise.
15961 (vsld): Likewise.
15962 (vsrd): Likewise.
15963 (vsrad): Likewise.
15964 (vsubudm): Likewise.
15965 (vcmpequd): Likewise.
15966 (vcmpgtsd): Likewise.
15967 (vcmpgtud): Likewise.
15968 (vcmpequd_p): Likewise.
15969 (vcmpgtsd_p): Likewise.
15970 (vcmpgtud_p): Likewise.
15971 (vupkhsw): Likewise.
15972 (vupklsw): Likewise.
15973 (vaddudm): Likewise.
15974 (vmaxsd): Likewise.
15975 (vmaxud): Likewise.
15976 (vminsd): Likewise.
15977 (vminud): Likewise.
15978 (vpksdss): Likewise.
15979 (vpksdus): Likewise.
15980 (vpkudum): Likewise.
15981 (vpkudus): Likewise.
15982 (vrld): Likewise.
15983 (vsld): Likewise.
15984 (vsrad): Likewise.
15985 (vsrd): Likewise.
15986 (vsubudm): Likewise.
15987
15988 * config/rs6000/rs6000.c (rs6000_init_hard_regno_mode_ok): Add
15989 support for power8 V2DI instructions.
15990
15991 * config/rs6000/altivec.md (UNSPEC_VPKUHUM): Add support for
15992 power8 V2DI instructions. Combine pack and unpack insns to use an
15993 iterator for each mode. Check whether a particular mode supports
15994 Altivec instructions instead of just checking TARGET_ALTIVEC.
15995 (UNSPEC_VPKUWUM): Likewise.
15996 (UNSPEC_VPKSHSS): Likewise.
15997 (UNSPEC_VPKSWSS): Likewise.
15998 (UNSPEC_VPKUHUS): Likewise.
15999 (UNSPEC_VPKSHUS): Likewise.
16000 (UNSPEC_VPKUWUS): Likewise.
16001 (UNSPEC_VPKSWUS): Likewise.
16002 (UNSPEC_VPACK_SIGN_SIGN_SAT): Likewise.
16003 (UNSPEC_VPACK_SIGN_UNS_SAT): Likewise.
16004 (UNSPEC_VPACK_UNS_UNS_SAT): Likewise.
16005 (UNSPEC_VPACK_UNS_UNS_MOD): Likewise.
16006 (UNSPEC_VUPKHSB): Likewise.
16007 (UNSPEC_VUNPACK_HI_SIGN): Likewise.
16008 (UNSPEC_VUNPACK_LO_SIGN): Likewise.
16009 (UNSPEC_VUPKHSH): Likewise.
16010 (UNSPEC_VUPKLSB): Likewise.
16011 (UNSPEC_VUPKLSH): Likewise.
16012 (VI2): Likewise.
16013 (VI_char): Likewise.
16014 (VI_scalar): Likewise.
16015 (VI_unit): Likewise.
16016 (VP): Likewise.
16017 (VP_small): Likewise.
16018 (VP_small_lc): Likewise.
16019 (VU_char): Likewise.
16020 (add<mode>3): Likewise.
16021 (altivec_vaddcuw): Likewise.
16022 (altivec_vaddu<VI_char>s): Likewise.
16023 (altivec_vadds<VI_char>s): Likewise.
16024 (sub<mode>3): Likewise.
16025 (altivec_vsubcuw): Likewise.
16026 (altivec_vsubu<VI_char>s): Likewise.
16027 (altivec_vsubs<VI_char>s): Likewise.
16028 (altivec_vavgs<VI_char>): Likewise.
16029 (altivec_vcmpbfp): Likewise.
16030 (altivec_eq<mode>): Likewise.
16031 (altivec_gt<mode>): Likewise.
16032 (altivec_gtu<mode>): Likewise.
16033 (umax<mode>3): Likewise.
16034 (smax<mode>3): Likewise.
16035 (umin<mode>3): Likewise.
16036 (smin<mode>3): Likewise.
16037 (altivec_vpkuhum): Likewise.
16038 (altivec_vpkuwum): Likewise.
16039 (altivec_vpkshss): Likewise.
16040 (altivec_vpkswss): Likewise.
16041 (altivec_vpkuhus): Likewise.
16042 (altivec_vpkshus): Likewise.
16043 (altivec_vpkuwus): Likewise.
16044 (altivec_vpkswus): Likewise.
16045 (altivec_vpks<VI_char>ss): Likewise.
16046 (altivec_vpks<VI_char>us): Likewise.
16047 (altivec_vpku<VI_char>us): Likewise.
16048 (altivec_vpku<VI_char>um): Likewise.
16049 (altivec_vrl<VI_char>): Likewise.
16050 (altivec_vsl<VI_char>): Likewise.
16051 (altivec_vsr<VI_char>): Likewise.
16052 (altivec_vsra<VI_char>): Likewise.
16053 (altivec_vsldoi_<mode>): Likewise.
16054 (altivec_vupkhsb): Likewise.
16055 (altivec_vupkhs<VU_char>): Likewise.
16056 (altivec_vupkls<VU_char>): Likewise.
16057 (altivec_vupkhsh): Likewise.
16058 (altivec_vupklsb): Likewise.
16059 (altivec_vupklsh): Likewise.
16060 (altivec_vcmpequ<VI_char>_p): Likewise.
16061 (altivec_vcmpgts<VI_char>_p): Likewise.
16062 (altivec_vcmpgtu<VI_char>_p): Likewise.
16063 (abs<mode>2): Likewise.
16064 (vec_unpacks_hi_v16qi): Likewise.
16065 (vec_unpacks_hi_v8hi): Likewise.
16066 (vec_unpacks_lo_v16qi): Likewise.
16067 (vec_unpacks_hi_<VP_small_lc>): Likewise.
16068 (vec_unpacks_lo_v8hi): Likewise.
16069 (vec_unpacks_lo_<VP_small_lc>): Likewise.
16070 (vec_pack_trunc_v8h): Likewise.
16071 (vec_pack_trunc_v4si): Likewise.
16072 (vec_pack_trunc_<mode>): Likewise.
16073
16074 * config/rs6000/altivec.h (vec_vaddudm): Add defines for power8
16075 V2DI builtins.
16076 (vec_vmaxsd): Likewise.
16077 (vec_vmaxud): Likewise.
16078 (vec_vminsd): Likewise.
16079 (vec_vminud): Likewise.
16080 (vec_vpksdss): Likewise.
16081 (vec_vpksdus): Likewise.
16082 (vec_vpkudum): Likewise.
16083 (vec_vpkudus): Likewise.
16084 (vec_vrld): Likewise.
16085 (vec_vsld): Likewise.
16086 (vec_vsrad): Likewise.
16087 (vec_vsrd): Likewise.
16088 (vec_vsubudm): Likewise.
16089 (vec_vupkhsw): Likewise.
16090 (vec_vupklsw): Likewise.
16091
e70670cf
JH
160922013-05-29 Jan Hubicka <jh@suse.cz>
16093
16094 * cgraph.h (symtab_node_base): Add definition, alias and analyzed
16095 flags; reorder rest of fields in more consistent way.
16096 (varpool_node): Remove analyzed, finalized and alias.
16097 (cgraph_ndoe): Likewise.
16098 (symtab_alias_ultimate_target): New function.
16099 (cgraph_function_node): Move offline.
16100 (cgraph_reset_node): Declare.
16101 (cgraph_comdat_can_be_unshared_p): Remove.
16102 (varpool_remove_initializer): Declare.
16103 (varpool_first_defined_variable, varpool_next_defined_variable
16104 cgraph_first_defined_function, cgraph_next_defined_function): Update.
16105 (cgraph_function_with_gimple_body_p): Update.
16106 (varpool_all_refs_explicit_p): Update.
16107 (symtab_alias_target): New function.
16108 (cgraph_alias_aliased_node, varpool_alias_aliased_node): Rename to ...
16109 (cgraph_alias_target, varpool_alias_target): .. this one; simplify.
c1e183a9
UB
16110 (cgraph_function_or_thunk_node): Simplify using
16111 symtab_alias_ultimate_target.
e70670cf
JH
16112 (varpool_variable_node): Likewise.
16113 * cgraph.c (cgraph_create_function_alias): Update.
16114 (cgraph_add_thunk): Update.
16115 (cgraph_remove_node): Update.
16116 (dump_cgraph_node): Do not dump removed flags.
16117 (cgraph_function_body_availability): Update.
16118 (cgraph_propagate_frequency): Update.
16119 (verify_cgraph_node): Check sanity of local flag.
16120 (cgraph_function_node): Move here from cgraph.h; revamp for
16121 cgraph_function_or_thunk_node.
16122 * lto-symtab.c (lto_varpool_replace_node): Update.
16123 (lto_symtab_resolve_can_prevail_p): Update.
16124 (lto_symtab_merge_cgraph_nodes): Update.
16125 * ipa-cp.c (determine_versionability, initialize_node_lattices,
16126 propagate_constants_accross_call, devirtualization_time_bonus,
16127 ipcp_propagate_stage): Update.
16128 * tree-emutls.c (create_emultls_var, ipa_lower_emutls): Update.
c1e183a9
UB
16129 * ipa-inline-transform.c (clone_inlined_nodes,
16130 preserve_function_body_p): Update.
e70670cf
JH
16131 * ipa-reference.c (propagate): Update.
16132 (write_node_summary_p): Update.
16133 * toplev.c (wrapup_global_declaration_2): Update.
16134 * cgraphunit.c (cgraph_analyze_function): Rename to ...
16135 (analyze_function) ... this one.
16136 (cgraph_process_new_functions): Update.
16137 (cgraph_reset_node): Export.
16138 (cgraph_finalize_function): Update.
16139 (cgraph_add_new_function): Update.
16140 (process_function_and_variable_attributes): Update.
16141 (varpool_finalize_decl): Update.
16142 (symbol_finalized): Remove.
16143 (symbol_finalized_and_needed): Rename to ...
16144 (symbol_defined_and_needed): ... update.
16145 (cgraph_analyze_functions): Update.
16146 (handle_alias_pairs): Update.
16147 (mark_functions_to_output): Update.
16148 (assemble_thunk): Update.
16149 (output_in_order): Update.
16150 (output_weakrefs): Update.
16151 (finalize_compilation_unit): Update.
16152 * lto-cgraph.c (reachable_from_other_partition_p, lto_output_node,
16153 lto_output_varpool_node, compute_ltrans_boundary, input_overwrite_node,
16154 input_node, input_varpool_node): Update.
16155 * dbxout.c (dbxout_expand_expr): Update.
16156 * cgraphclones.c (cgraph_clone_node): Update.
16157 (cgraph_copy_node_for_versioning): Update.
16158 (cgraph_materialize_clone): Update.
16159 (cgraph_materialize_all_clones): Update.
16160 * ipa-pure-const.c (analyze_function, pure_const_write_summary,
16161 propagate_pure_const, propagate_nothrow): Update.
16162 * lto-streamer-out.c (lto_output, write_symbol): Update.
16163 * ipa-utils.c (ipa_reverse_postorder): Update.
16164 * ipa-inline.c (can_inline_edge_p): Update.
16165 (update_caller_keys, ipa_inline): Update.
16166 * dwarf2out.c (reference_to_unused,
16167 premark_types_used_by_global_vars_helper): Update.
16168 * tree-eh.c (tree_could_trap_p): Update.
16169 * ipa-split.c (consider_split, execute_split_functions): Update.
16170 * ipa.c (cgraph_non_local_node_p_1, cgraph_local_node_p,
c1e183a9
UB
16171 has_addr_references_p): Update; move ahead in file for better
16172 readability.
e70670cf
JH
16173 (process_references): Simplify.
16174 (symtab_remove_unreachable_nodes): Update; cleanup way function/var
16175 bodies are removed.
16176 (cgraph_comdat_can_be_unshared_p): Make static.
16177 (cgraph_externally_visible_p): Update.
16178 (varpool_externally_visible_p): Update.
16179 (function_and_variable_visibility): Update.
16180 * trans-mem.c (get_cg_data, ipa_tm_mayenterirr_function,
16181 ipa_tm_mark_force_output_node): Update.
16182 * ipa-inline-analysis.c (dump_inline_summary, initialize_inline_failed,
16183 estimate_edge_devirt_benefit, inline_generate_summary,
16184 inline_write_summary): Update.
16185 * gimple-fold.c (can_refer_decl_in_current_unit_p): Update.
16186 * ipa-prop.c (ipa_compute_jump_functions): Update.
c1e183a9
UB
16187 (ipa_print_node_params, ipa_prop_read_section,
16188 ipa_update_after_lto_read, read_replacements_section): Update.
e70670cf
JH
16189 * varasm.c (mark_decl_referenced): Update.
16190 (assemble_alias, dump_tm_clone_pairs): Update.
16191 * tree-inline.c (copy_bb): Update.
16192 (estimate_num_insns, optimize_inline_calls, tree_function_versioning):
16193 Update.
16194 * symtab.c (dump_symtab_base): Print new flags.
16195 (verify_symtab_base): Verify new flags.
16196 (symtab_alias_ultimate_target): New function.
16197 * tree-ssa-structalias.c (get_constraint_for_ssa_var,
16198 create_variable_info_for, associate_varinfo_to_alias, ipa_pta_execute):
16199 Update.
c1e183a9
UB
16200 * passes.c (ipa_write_summaries, ipa_write_optimization_summaries):
16201 Update.
e70670cf
JH
16202 * i386.c (ix86_get_function_versions_dispatcher,
16203 ix86_generate_version_dispatcher_body): Update.
16204 (fold_builtin_cpu): Use varpool_add_new_variable.
16205 * varpool.c (varpool_remove_initializer): Break out from ...
16206 (varpool_remove_node): ... this one.
16207 (dump_varpool_node, varpool_node_for_asm,
16208 cgraph_variable_initializer_availability, varpool_analyze_node,
16209 varpool_assemble_decl, varpool_remove_unreferenced_decls,
c1e183a9
UB
16210 varpool_finalize_named_section_flags, varpool_create_variable_alias):
16211 Update.
e70670cf 16212
182802ad
JH
162132013-05-29 Jan Hubicka <jh@suse.cz>
16214
16215 * passes.c (init_optimization_passes): Move OMP expansion into lowering.
16216
8190b609
ER
162172013-05-29 Easwaran Raman <eraman@google.com>
16218
16219 PR tree-optimization/57442
16220 * tree-ssa-reassoc.c (appears_later_in_bb): Return correct value
16221 when control exits the main loop.
16222
69f5aa9b
SKS
162232013-05-29 Sandeep Kumar Singh <Sandeep.Singh2@kpitcummins.com>
16224
16225 * rx/rx.h (TARGET_CPU_CPP_BUILTINS): Add macros for RX100, RX200,
16226 and RX600.
c1e183a9 16227 * rx/rx.opt: Add macro for rx100 with string rx100 and value RX100.
69f5aa9b
SKS
16228 * rx/rx-opts.h (rx_cpu_types): Add new cpu type rx100.
16229 * rx/t-rx: Add rx100 under multi library matches option for nofpu
16230 option.
16231
4b847da9
BS
162322013-05-29 Bill Schmidt <wschmidt@linux.vnet.ibm.com>
16233
16234 PR tree-optimization/57441
16235 * gimple-ssa-strength-reduction.c (analyze_candidates_and_replace):
16236 Don't limit size of incr_vec to number of candidates.
16237
4fc43c39
SE
162382013-05-29 Steve Ellcey <sellcey@imgtec.com>
16239
16240 * config/mips/mti-linux.h (SYSROOT_SUFFIX_SPEC): Add micromips
16241 and mips16 directories.
c1e183a9 16242 * config/mips/t-mti-linux (MULTILIB_OPTIONS): Add micromips and mips16.
4fc43c39
SE
16243 (MULTILIB_DIRNAMES): Ditto.
16244 (MULTILIB_EXCEPTIONS): Add new exceptions.
16245 * config/mips/t-mti-elf (MULTILIB_OPTIONS): Add micromips.
16246 (MULTILIB_DIRNAMES): Ditto.
16247 (MULTILIB_EXCEPTIONS): Add new exceptions.
16248
12211b99 162492012-05-29 Chris Schlumberger-Socha <chris.schlumberger-socha@arm.com>
a5350ddc
CSS
16250 Marcus Shawcroft <marcus.shawcroft@arm.com>
16251
16252 * config/aarch64/aarch64-protos.h (aarch64_symbol_type): Define
16253 SYMBOL_TINY_ABSOLUTE.
16254 * config/aarch64/aarch64.c (aarch64_load_symref_appropriately): Handle
16255 SYMBOL_TINY_ABSOLUTE.
16256 (aarch64_expand_mov_immediate): Likewise.
16257 (aarch64_classify_symbol): Likewise.
16258 (aarch64_mov_operand_p): Remove ATTRIBUTE_UNUSED.
16259 Permit SYMBOL_TINY_ABSOLUTE.
16260 * config/aarch64/predicates.md (aarch64_mov_operand): Permit CONST.
16261
12211b99 162622013-05-29 Chris Schlumberger-Socha <chris.schlumberger-socha@arm.com>
17f4d4bf
CSS
16263 Marcus Shawcroft <marcus.shawcroft@arm.com>
16264
16265 * config/aarch64/aarch64.c (aarch64_classify_symbol): Remove comment.
16266 Refactor if/switch. Replace gcc_assert with if.
16267
c0186656
GG
162682013-05-29 Ganesh Gopalasubramanian <Ganesh.Gopalasubramanian@amd.com>
16269
16270 * config/i386/i386.c (initial_ix86_tune_features): Enable
16271 FP Reassociation for AMD bdver1 and bdver2.
16272
d20188f3
MJ
162732013-05-29 Martin Jambor <mjambor@suse.cz>
16274
16275 * tree-cfg.c (verify_expr): Verify that BIT_FIELD_REF, REALPART_EXPR
16276 and IMAGPART_EXPR do not occur within other handled_components.
16277
292cba13
RB
162782013-05-29 Richard Biener <rguenther@suse.de>
16279
16280 * tree-vect-slp.c (vect_bb_slp_scalar_cost): Guard vinfo
16281 access on whether the use is in the BB we currently try to
16282 vectorize.
16283 (vect_bb_vectorization_profitable_p): Pass the BB we currently
16284 vectorize to vect_bb_slp_scalar_cost.
16285
6eddf228
RB
162862013-05-29 Richard Biener <rguenther@suse.de>
16287
16288 * tree-vect-slp.c (vect_bb_slp_scalar_cost): New function
16289 computing scalar cost offsetted by stmts that are kept live
16290 by scalar uses.
16291 (vect_bb_vectorization_profitable_p): Use vect_bb_slp_scalar_cost
16292 for computation of scalar cost.
16293
7df36117
SE
162942013-05-28 Steve Ellcey <sellcey@mips.com>
16295
16296 * config/mips/mips-cpus.def (mips32r2): Change processor type.
16297
36536d79
BI
162982013-05-28 Balaji V. Iyer <balaji.v.iyer@intel.com>
16299
16300 * doc/extend.texi (C Extensions): Added documentation about Cilk Plus
16301 array notation built-in reduction functions.
16302 * doc/passes.texi (Passes): Added documentation about changes done
16303 for Cilk Plus.
16304 * doc/invoke.texi (C Dialect Options): Added documentation about
16305 the -fcilkplus flag.
16306 * Makefile.in (C_COMMON_OBJS): Added c-family/array-notation-common.o.
16307 (BUILTINS_DEF): Depend on cilkplus.def.
16308 * builtins.def: Include cilkplus.def. Define DEF_CILKPLUS_BUILTIN.
16309 * builtin-types.def: Define BT_FN_INT_PTR_PTR_PTR.
16310 * cilkplus.def: New file.
16311
12211b99 163122013-05-28 Joern Rennecke <joern.rennecke@embecosm.com>
b49eefa5
JR
16313
16314 PR rtl-optimization/57439
16315 * postreload.c (move2add_use_add2_insn): Use gen_lowpart_common.
16316
9daf6dbb
ER
163172013-05-28 Easwaran Raman <eraman@google.com>
16318
16319 PR tree-optimization/57337
16320 * tree-ssa-reassoc.c (appears_later_in_bb): New function.
16321 (find_insert_point): Correctly identify the insertion point
16322 when two statements with the same UID is compared.
16323
fbd7e877
RB
163242013-05-28 Richard Biener <rguenther@suse.de>
16325
16326 PR tree-optimization/56787
16327 * tree-vect-data-refs.c (vect_analyze_data_refs): Drop clobbers
16328 from the list of data references.
16329 * tree-vect-loop.c (vect_determine_vectorization_factor): Skip
16330 clobbers.
16331 (vect_analyze_loop_operations): Likewise.
16332 (vect_transform_loop): Remove clobbers.
16333
bbba1117
MJ
163342013-05-28 Martin Jambor <mjambor@suse.cz>
16335
16336 * tree-cfg.c (verify_expr): Verify that BIT_FIELD_REFs, IMAGPART_EXPRs
16337 and REALPART_EXPRs have scalar type.
16338
bd388c2a
RB
163392013-05-28 Richard Biener <rguenther@suse.de>
16340
16341 PR tree-optimization/57411
16342 * tree-ssa-copy.c (may_propagate_copy): Cannot propagate
16343 virtual operands.
16344 * tree-ssa-dom.c (eliminate_const_or_copy): Special-case
16345 virtual operand propagation.
16346
2f56a311
EB
163472013-05-28 Eric Botcazou <ebotcazou@adacore.com>
16348
16349 * config/sparc/sparc.c (sparc_expand_vec_perm_bmask): Use %g0 as
16350 destination register for bmasksi_vis.
16351 (vector_init_bshuffle): Likewise.
16352 * config/sparc/sparc.md (vec_perm_constv8qi): Likewise.
16353
5c3eacbb
EB
163542013-05-28 Eric Botcazou <ebotcazou@adacore.com>
16355
16356 * doc/invoke.texi (SPARC Options): Document -mfix-ut699.
16357 * builtins.c (expand_builtin_mathfn) <BUILT_IN_SQRT>: Try to widen the
16358 mode if the instruction isn't available in the original mode.
16359 * config/sparc/sparc.opt (mfix-ut699): New option.
16360 * config/sparc/sparc.md (muldf3_extend): Disable if -mfix-ut699.
16361 (divdf3): Turn into expander.
16362 (divdf3_nofix): New insn.
16363 (divdf3_fix): Likewise.
16364 (divsf3): Disable if -mfix-ut699.
16365 (sqrtdf2): Turn into expander.
16366 (sqrtdf2_nofix): New insn.
16367 (sqrtdf2_fix): Likewise.
16368 (sqrtsf2): Disable if -mfix-ut699.
16369
a1756c0a
RB
163702013-05-27 Richard Biener <rguenther@suse.de>
16371
16372 PR middle-end/57412
16373 * omp-low.c (expand_omp_atomic_pipeline): Use the correct latch
16374 block for the new loop.
16375
5a892248
RB
163762013-05-27 Richard Biener <rguenther@suse.de>
16377
16378 PR tree-optimization/57343
16379 * tree-ssa-loop-niter.c (number_of_iterations_ne_max): Do not
16380 use multiple_of_p if not TYPE_OVERFLOW_UNDEFINED.
16381 (number_of_iterations_cond): Do not build the folded tree.
16382
d1de852b
RB
163832013-05-27 Richard Biener <rguenther@suse.de>
16384
16385 Revert
16386 PR middle-end/57381
16387 * fold-const.c (operand_equal_p): Compare FIELD_DECLs with
16388 OEP_CONSTANT_ADDRESS_OF retained.
16389
16390 PR tree-optimization/57417
16391 * tree-ssa-sccvn.c (vn_reference_fold_indirect): Fix test
16392 for unchanged base.
16393 (set_ssa_val_to): Compare addresses using
16394 get_addr_base_and_unit_offset.
16395
12211b99 163962013-05-27 Joern Rennecke <joern.rennecke@embecosm.com>
7894bc6b
JR
16397
16398 PR rtl-optimization/56833
16399 * postreload.c (move2add_record_mode): New function.
16400 (move2add_record_sym_value, move2add_valid_value_p): Likewise.
16401 (move2add_use_add2_insn): Use move2add_record_sym_value.
16402 (move2add_use_add3_insn): Likewise.
16403 (reload_cse_move2add): Use move2add_valid_value_p and
16404 move2add_record_mode. Invalidate call-clobbered and REG_INC
16405 affected regs by setting reg_mode to VOIDmode.
16406 (move2add_note_store): Don't pretend the inside of a SUBREG is
16407 the actual destination. Invalidate single/leading registers by
16408 setting reg_mode to VOIDmode.
16409 Use move2add_record_sym_value, move2add_valid_value_p and
16410 move2add_record_mode.
16411
b03be25f
RB
164122013-05-27 Richard Biener <rguenther@suse.de>
16413
16414 PR tree-optimization/57396
16415 * tree-affine.c (double_int_constant_multiple_p): Properly
16416 return false for val == 0 and div != 0.
16417
44e88db2
RS
164182013-05-25 Richard Sandiford <rdsandiford@googlemail.com>
16419
16420 * config/mips/mips.h: Use #elif in preprocessor conditions.
16421
3b859704
RS
164222013-05-25 Richard Sandiford <rdsandiford@googlemail.com>
16423
16424 PR target/53916
16425 * config/mips/constraints.md (kl): New constraint.
16426 * config/mips/mips.md (divmod<mode>4, udivmod<mode>4): Delete.
16427 (divmod<mode>4_internal): Rename to divmod<mode>4. Use "kl" as the
16428 constraint for operand 0. Split after CSE for MIPS16. Emit a move
16429 from LO for MIPS16.
16430 (udivmod<mode>4_internal): Likewise udivmod<mode>4.
16431
c979d5f5
RS
164322013-05-25 Richard Sandiford <rdsandiford@googlemail.com>
16433
16434 PR target/55777
16435 * config/mips/mips.c (mips_can_inline_p): New function.
16436 (TARGET_CAN_INLINE_P): Define.
16437
8e90de43
SB
164382013-05-25 Steven Bosscher <steven@gcc.gnu.org>
16439
16440 * sched-int.h (ds_t, dw_t): Make unsigned int.
16441 Fix documentation that describes how all the ds_t bits are used.
16442 Reserve the last bit for delayed-branch scheduling.
16443 (BITS_PER_DEP_STATUS): Move to ds_t typedef.
16444 (BITS_PER_DEP_WEAK): Fix definition and documentation.
16445 (gen_dep_weak_1): Remove prototype.
16446 * sched-deps.c (get_dep_weak_1): Make static.
16447 * target.def (speculate_insn, needs_block_p, gen_spec_check,
16448 get_insn_spec_ds, get_insn_checked_ds): Adjust hook prototypes.
16449 * doc/tm.texi: Regenerate.
16450 * config/ia64/ia64.c (ia64_needs_block_p): Update prototype.
16451
cb5cb194
SB
164522013-05-24 Steven Bosscher <steven@gcc.gnu.org>
16453
16454 PR debug/56950
16455 * haifa-sched.c (sched_extend_bb): Ignore DEBUG_INSNs.
16456
5bd93ff6
NS
164572013-05-24 Nathan Sidwell <nathan@codesourcery.com>
16458 Sandra Loosemore <sandra@codesourcery.com>
16459
16460 * config.gcc (powerpc-*): Allow native for with-cpu.
16461
2343af65
JL
164622013-05-24 Jeff Law <law@redhat.com>
16463
16464 PR tree-optimization/57124
16465 * tree-vrp.c (simplify_cond_using_ranges): Only simplify a
16466 conversion feeding a condition if the range has an overflow
16467 if -fstrict-overflow. Add warnings for when we do make the
16468 transformation.
16469
3ad695b9
DC
164702013-05-24 Dehao Chen <dehao@google.com>
16471
04960246 16472 * tree-cfg.c (locus_discrim_map): Fix the typo.
3ad695b9
DC
16473 (locus_discrim_hasher): Likewise.
16474 (locus_discrim_hasher::hash): Likewise.
16475 (locus_discrim_hasher::equal): Likewise.
16476
82338059
MJ
164772013-05-24 Martin Jambor <mjambor@suse.cz>
16478
16479 PR tree-optimization/57294
16480 * cgraph.h (ipa_record_stmt_references): Declare.
16481 * cgraphbuild.c (ipa_record_stmt_references): New function.
16482 (build_cgraph_edges): Use ipa_record_stmt_references.
16483 (rebuild_cgraph_edges): Likewise.
16484 (cgraph_rebuild_references): Likewise.
16485 * ipa-prop.c (ipa_modify_call_arguments): Discard references
16486 associated with the old statement and build references from the
16487 newly built statements.
16488 * ipa-ref.c (ipa_remove_stmt_references): New function.
16489 * ipa-ref.h (ipa_remove_stmt_references): Declare.
16490
1ccd4874
VM
164912013-05-24 Vladimir Makarov <vmakarov@redhat.com>
16492
55805e54 16493 * lra-constraints.c (emit_spill_move): Use smaller mode for
1ccd4874 16494 mem-mem moves.
55805e54 16495 (check_and_process_move): Consider mem-reg moves for secondary
1ccd4874 16496 too.
55805e54 16497 (curr_insn_transform): Don't lose insns emitted before for
1ccd4874
VM
16498 secondary memory moves.
16499 (inherit_in_ebb): Mark defined reg. Add usage only if it is not a
16500 reg set up in the current insn.
16501
25e25c73
DC
165022013-05-24 Dehao Chen <dehao@google.com>
16503
04960246 16504 * tree-cfg.c (locus_descrim_hasher::hash): Change discriminator
25e25c73
DC
16505 hash function.
16506 (locus_descrim_hasher::equal): Likewise.
04960246 16507 (build_gimple_cfg): New discriminator assignment algorithm.
25e25c73
DC
16508 (make_edges): Likewise.
16509 (next_discriminator_for_locus): Likewise.
16510 (same_line_p): Likewise.
16511 (assign_discriminators): Likewise.
16512 (make_cond_expr_edges): Likewise.
16513 (make_gimple_switch_edges): Likewise.
16514 (make_goto_expr_edges): Likewise.
16515 (make_gimple_asm_edges): Likewise.
16516
50d38551
IB
165172013-05-24 Ian Bolton <ian.bolton@arm.com>
16518
16519 * config/aarch64/aarch64.c (aarch64_print_operand): Change the
16520 X format specifier to only display bottom 16 bits.
16521 * config/aarch64/aarch64.md (insv_imm<mode>): Allow any size of
16522 immediate to match for operand 2, since it will be masked.
16523
aea0101d
RB
165242013-05-24 Richard Biener <rguenther@suse.de>
16525
16526 PR tree-optimization/57287
16527 * tree-ssa-uninit.c (compute_uninit_opnds_pos): Disregard
16528 all SSA names that occur in abnormal PHIs.
16529
634e03d3
AI
165302013-05-24 Alexander Ivchenko <alexander.ivchenko@intel.com>
16531
16532 PR tree-ssa/57385
16533 * tree-ssa-sccvn.c (fully_constant_vn_reference_p): Check
16534 that index is not negative.
16535
b17c024f
EB
165362013-05-24 Eric Botcazou <ebotcazou@adacore.com>
16537
16538 PR rtl-optimization/55177
16539 * simplify-rtx.c (simplify_unary_operation_1) <NOT>: Deal with BSWAP.
16540 (simplify_byte_swapping_operation): New.
16541 (simplify_binary_operation_1): Call it for AND, IOR and XOR.
16542 (simplify_relational_operation_1): Deal with BSWAP.
16543
46aeac1b
RH
165442013-05-23 Richard Henderson <rth@redhat.com>
16545
16546 PR target/56742
16547 * config/i386/i386.c (ix86_seh_fixup_eh_fallthru): New.
16548 (ix86_reorg): Call it.
16549
70cc1536
UB
165502013-05-23 Uros Bizjak <ubizjak@gmail.com>
16551
16552 PR target/57379
16553 * config/alpha/alpha.md (unspec): Add UNSPEC_XFLT_COMPARE.
16554 * config/alpha/alpha.c (alpha_emit_xfloating_compare): Construct
16555 REG_EQUAL note as UNSPEC_XFLT_COMPARE unspec.
16556
b3851501
CB
165572013-05-23 Christian Bruel <christian.bruel@st.com>
16558
16559 PR debug/57351
16560 * config/arm/arm.c (arm_dwarf_register_span): Do not use dbx number.
16561
12211b99 165622013-05-23 Chris Schlumberger-Socha <chris.schlumberger-socha@arm.com>
4e194912
CSS
16563 Marcus Shawcroft <marcus.shawcroft@arm.com>
16564
16565 * config/aarch64/aarch64.md (*movdi_aarch64): Replace Usa with S.
16566 * config/aarch64/constraints.md (Usa): Remove.
16567 * doc/md.texi (AArch64 Usa): Remove.
16568
12211b99 165692013-05-23 Chris Schlumberger-Socha <chris.schlumberger-socha@arm.com>
83f8c414
CSS
16570 Marcus Shawcroft <marcus.shawcroft@arm.com>
16571
16572 * config/aarch64/aarch64-protos.h (aarch64_mov_operand_p): Define.
16573 * config/aarch64/aarch64.c (aarch64_mov_operand_p): Define.
16574 * config/aarch64/predicates.md (aarch64_const_address): Remove.
16575 (aarch64_mov_operand): Use aarch64_mov_operand_p.
16576
12211b99 165772013-05-23 Vidya Praveen <vidyapraveen@arm.com>
b5574232
VP
16578
16579 * config/aarch64/aarch64-simd.md (clzv4si2): Support for CLZ
16580 instruction (AdvSIMD).
16581 * config/aarch64/aarch64-builtins.c
16582 (aarch64_builtin_vectorized_function): Handler for BUILT_IN_CLZ.
16583 * config/aarch64/aarch-simd-builtins.def: Entry for CLZ.
16584
76c36cb1 165852013-05-23 Martin Jambor <mjambor@suse.cz>
7d2fb524
MJ
16586
16587 PR middle-end/57347
16588 * tree.h (contains_bitfld_component_ref_p): Declare.
16589 * tree-sra.c (contains_bitfld_comp_ref_p): Move...
70cc1536
UB
16590 * tree.c (contains_bitfld_component_ref_p): ...here. Adjust its
16591 caller.
7d2fb524
MJ
16592 * ipa-prop.c (determine_known_aggregate_parts): Check that LHS does
16593 not access a bit-field. Assert all final offsets are byte-aligned.
16594
ce521ae6
RB
165952013-05-23 Richard Biener <rguenther@suse.de>
16596
16597 PR tree-optimization/57380
16598 * tree-ssa-phiprop.c (propagate_with_phi): Do not require at
16599 least one invariant or re-used load.
16600 * passes.c (init_optimization_passes): Move pass_phiprop before
16601 pass_forwprop.
16602
75c7257f
JG
166032013-05-23 James Greenhalgh <james.greenhalgh@arm.com>
16604
16605 * config/aarch64/aarch64-simd.md
16606 (aarch64_cm<optab>di): Add clobber of CC_REGNUM to unsplit pattern.
16607
af6d44b5
RB
166082013-05-23 Richard Biener <rguenther@suse.de>
16609
16610 PR middle-end/57381
16611 * fold-const.c (operand_equal_p): Compare FIELD_DECLs with
16612 OEP_CONSTANT_ADDRESS_OF retained.
16613
bd3647bf
JJ
166142013-05-23 Jakub Jelinek <jakub@redhat.com>
16615
16616 PR middle-end/57344
70cc1536
UB
16617 * expmed.c (store_split_bit_field): If op0 is a REG or SUBREG of a REG,
16618 don't lower unit. Handle unit not being always BITS_PER_WORD.
bd3647bf 16619
a55757ea
RB
166202013-05-23 Richard Biener <rguenther@suse.de>
16621
16622 PR rtl-optimization/57341
16623 * ira.c (validate_equiv_mem_from_store): Use anti_dependence
16624 instead of true_dependence.
16625
cfb00b41
DM
166262013-05-22 David Malcolm <dmalcolm@redhat.com>
16627
dd1c676f
DM
16628 * bb-reorder.c (branch_threshold): Make const.
16629 (exec_threshold): Ditto.
cfb00b41 16630
f62511da 166312013-05-22 Michael Meissner <meissner@linux.vnet.ibm.com>
12211b99
UB
16632 Pat Haugen <pthaugen@us.ibm.com>
16633 Peter Bergner <bergner@vnet.ibm.com>
f62511da
MM
16634
16635 * doc/extend.texi (PowerPC AltiVec/VSX Built-in Functions): Add
16636 documentation for the power8 crypto builtins.
16637
16638 * config/rs6000/t-rs6000 (MD_INCLUDES): Add crypto.md.
16639
16640 * config/rs6000/rs6000-builtin.def (BU_P8V_AV_1): Add support
16641 macros for defining power8 builtin functions.
16642 (BU_P8V_AV_2): Likewise.
16643 (BU_P8V_AV_P): Likewise.
16644 (BU_P8V_VSX_1): Likewise.
16645 (BU_P8V_OVERLOAD_1): Likewise.
16646 (BU_P8V_OVERLOAD_2): Likewise.
16647 (BU_CRYPTO_1): Likewise.
16648 (BU_CRYPTO_2): Likewise.
16649 (BU_CRYPTO_3): Likewise.
16650 (BU_CRYPTO_OVERLOAD_1): Likewise.
16651 (BU_CRYPTO_OVERLOAD_2): Likewise.
16652 (XSCVSPDP): Fix typo, point to the correct instruction.
16653 (VCIPHER): Add power8 crypto builtins.
16654 (VCIPHERLAST): Likewise.
16655 (VNCIPHER): Likewise.
16656 (VNCIPHERLAST): Likewise.
16657 (VPMSUMB): Likewise.
16658 (VPMSUMH): Likewise.
16659 (VPMSUMW): Likewise.
16660 (VPERMXOR_V2DI): Likewise.
16661 (VPERMXOR_V4SI: Likewise.
16662 (VPERMXOR_V8HI: Likewise.
16663 (VPERMXOR_V16QI: Likewise.
16664 (VSHASIGMAW): Likewise.
16665 (VSHASIGMAD): Likewise.
16666 (VPMSUM): Likewise.
16667 (VPERMXOR): Likewise.
16668 (VSHASIGMA): Likewise.
16669
16670 * config/rs6000/rs6000-c.c (rs6000_target_modify_macros): Define
16671 __CRYPTO__ if the crypto instructions are available.
16672 (altivec_overloaded_builtins): Add support for overloaded power8
16673 builtins.
16674
16675 * config/rs6000/rs6000.c (rs6000_expand_ternop_builtin): Add
16676 support for power8 crypto builtins.
16677 (builtin_function_type): Likewise.
16678 (altivec_init_builtins): Add support for builtins that take vector
16679 long long (V2DI) arguments.
16680
16681 * config/rs6000/crypto.md: New file, define power8 crypto
16682 instructions.
16683
166842013-05-22 Michael Meissner <meissner@linux.vnet.ibm.com>
12211b99
UB
16685 Pat Haugen <pthaugen@us.ibm.com>
16686 Peter Bergner <bergner@vnet.ibm.com>
f62511da
MM
16687
16688 * doc/invoke.texi (Option Summary): Add power8 options.
16689 (RS/6000 and PowerPC Options): Likewise.
16690
16691 * doc/md.texi (PowerPC and IBM RS6000 constraints): Update to use
16692 constraints.md instead of rs6000.h. Reorder w* constraints. Add
16693 wm, wn, wr documentation.
16694
f43245d1 16695 * config/rs6000/constraints.md (wm): New constraint for VSX
f62511da
MM
16696 registers if direct move instructions are enabled.
16697 (wn): New constraint for no registers.
16698 (wq): New constraint for quad word even GPR registers.
16699 (wr): New constraint if 64-bit instructions are enabled.
16700 (wv): New constraint if power8 vector instructions are enabled.
16701 (wQ): New constraint for quad word memory locations.
16702
f43245d1 16703 * config/rs6000/predicates.md (const_0_to_15_operand): New
f62511da
MM
16704 constraint for 0..15 for crypto instructions.
16705 (gpc_reg_operand): If VSX allow registers in VSX registers as well
16706 as GPR and floating point registers.
16707 (int_reg_operand): New predicate to match only GPR registers.
16708 (base_reg_operand): New predicate to match base registers.
16709 (quad_int_reg_operand): New predicate to match even GPR registers
16710 for quad memory operations.
16711 (vsx_reg_or_cint_operand): New predicate to allow vector logical
16712 operations in both GPR and VSX registers.
16713 (quad_memory_operand): New predicate for quad memory operations.
16714 (reg_or_indexed_operand): New predicate for direct move support.
16715
f43245d1 16716 * config/rs6000/rs6000-cpus.def (ISA_2_5_MASKS_EMBEDDED):
f62511da
MM
16717 Inherit from ISA_2_4_MASKS, not ISA_2_2_MASKS.
16718 (ISA_2_7_MASKS_SERVER): New mask for ISA 2.07 (i.e. power8).
16719 (POWERPC_MASKS): Add power8 options.
16720 (power8 cpu): Use ISA_2_7_MASKS_SERVER instead of specifying the
16721 various options.
16722
f43245d1 16723 * config/rs6000/rs6000-c.c (rs6000_target_modify_macros):
f62511da
MM
16724 Define _ARCH_PWR8 and __POWER8_VECTOR__ for power8.
16725
f43245d1 16726 * config/rs6000/rs6000.opt (-mvsx-timode): Add documentation.
f62511da
MM
16727 (-mpower8-fusion): New power8 options.
16728 (-mpower8-fusion-sign): Likewise.
16729 (-mpower8-vector): Likewise.
16730 (-mcrypto): Likewise.
16731 (-mdirect-move): Likewise.
16732 (-mquad-memory): Likewise.
16733
f43245d1 16734 * config/rs6000/rs6000.c (power8_cost): Initial definition for power8.
f62511da
MM
16735 (rs6000_hard_regno_mode_ok): Make PTImode only match even GPR
16736 registers.
70cc1536 16737 (rs6000_debug_reg_print): Print the base register class if -mdebug=reg.
f62511da
MM
16738 (rs6000_debug_vector_unit): Add p8_vector.
16739 (rs6000_debug_reg_global): If -mdebug=reg, print power8 constraint
16740 definitions. Also print fusion state.
16741 (rs6000_init_hard_regno_mode_ok): Set up power8 constraints.
16742 (rs6000_builtin_mask_calculate): Add power8 builtin support.
16743 (rs6000_option_override_internal): Add support for power8.
16744 (rs6000_common_init_builtins): Add debugging for skipped builtins
16745 if -mdebug=builtin.
16746 (rs6000_adjust_cost): Add power8 support.
16747 (rs6000_issue_rate): Likewise.
16748 (insn_must_be_first_in_group): Likewise.
16749 (insn_must_be_last_in_group): Likewise.
16750 (force_new_group): Likewise.
16751 (rs6000_register_move_cost): Likewise.
16752 (rs6000_opt_masks): Likewise.
16753
16754 * config/rs6000/rs6000.h (ASM_CPU_POWER8_SPEC): If we don't have a
16755 power8 capable assembler, default to power7 options.
16756 (TARGET_DIRECT_MOVE): Likewise.
16757 (TARGET_CRYPTO): Likewise.
16758 (TARGET_P8_VECTOR): Likewise.
16759 (VECTOR_UNIT_P8_VECTOR_P): Define power8 vector support.
16760 (VECTOR_UNIT_VSX_OR_P8_VECTOR_P): Likewise.
16761 (VECTOR_MEM_P8_VECTOR_P): Likewise.
16762 (VECTOR_MEM_VSX_OR_P8_VECTOR_P): Likewise.
16763 (VECTOR_MEM_ALTIVEC_OR_VSX_P): Likewise.
16764 (TARGET_XSCVDPSPN): Likewise.
16765 (TARGET_XSCVSPDPN): Likewsie.
16766 (TARGET_SYNC_HI_QI): Likewise.
16767 (TARGET_SYNC_TI): Likewise.
16768 (MASK_CRYPTO): Likewise.
16769 (MASK_DIRECT_MOVE): Likewise.
16770 (MASK_P8_FUSION): Likewise.
16771 (MASK_P8_VECTOR): Likewise.
70cc1536
UB
16772 (REG_ALLOC_ORDER): Move fr13 to be lower in priority so that the TFmode
16773 temporary used by some of the direct move instructions to get two FP
16774 temporary registers does not force creation of a stack frame.
f62511da
MM
16775 (VLOGICAL_REGNO_P): Allow vector logical operations in GPRs.
16776 (MODES_TIEABLE_P): Move the VSX tests above the Altivec tests so
16777 that any VSX registers are tieable, even if they are also an
16778 Altivec vector mode.
16779 (r6000_reg_class_enum): Add wm, wr, wv constraints.
16780 (RS6000_BTM_P8_VECTOR): Power8 builtin support.
16781 (RS6000_BTM_CRYPTO): Likewise.
16782 (RS6000_BTM_COMMON): Likewise.
16783
16784 * config/rs6000/rs6000.md (cpu attribute): Add power8.
16785 * config/rs6000/rs6000-opts.h (PROCESSOR_POWER8): Likewise.
16786 (enum rs6000_vector): Add power8 vector support.
16787
73a1a707
RR
167882013-05-22 Ramana Radhakrishnan <ramana.radhakrishnan@arm.com>
16789
16790 PR target/19599
16791 PR target/57340
16792 * config/arm/arm.c (any_sibcall_uses_r3): Rename to ..
16793 (any_sibcall_could_use_r3): this and handle indirect calls.
16794 (arm_get_frame_offsets): Rename use of any_sibcall_uses_r3.
16795
d28073d4
BS
167962013-05-22 Bill Schmidt <wschmidt@linux.vnet.ibm.com>
16797
16798 * config/rs6000/rs6000.h (MALLOC_ABI_ALIGNMENT): New #define.
16799
bee0b10c
RB
168002013-05-22 Richard Biener <rguenther@suse.de>
16801
16802 PR middle-end/57349
16803 * profile.c (branch_prob): Do not split blocks that are
16804 abnormally receiving from ECF_RETURNS_TWICE functions.
16805
98409b51
RS
168062013-05-22 Richard Sandiford <rsandifo@linux.vnet.ibm.com>
16807
16808 * recog.c (offsettable_address_addr_space_p): Fix calculation of
16809 address mode. Move pointer mode initialization to the same place.
16810
c0602ab8
MZ
168112013-05-22 Michael Zolotukhin <michael.v.zolotukhin@intel.com>
16812
16813 * read-rtl.c (copy_rtx_for_iterators): Continue applying iterators
16814 while it has any effect.
16815
4f7a634e
ER
168162013-05-21 Easwaran Raman <eraman@google.com>
16817
16818 PR tree-optimization/57322
9daf6dbb
ER
16819 * tree-ssa-reassoc.c (build_and_add_sum): If a BB is empty, set the
16820 UID of the statement added to the BB to be 1.
4f7a634e 16821
a4ce1258
JJ
168222013-05-21 Jakub Jelinek <jakub@redhat.com>
16823
16824 PR tree-optimization/57331
70cc1536
UB
16825 * tree-vrp.c (simplify_cond_using_ranges): Don't optimize comparison
16826 of conversion from pointer type to integral type with integer.
a4ce1258 16827
1b14621a
MJ
168282013-05-21 Martin Jambor <mjambor@suse.cz>
16829
16830 PR lto/57289
16831 * ipa-prop.c (ipa_read_node_info): Process param_used and
16832 controlled_uses in the same order as when writing.
16833
e60661f0
MG
168342013-05-21 Magnus Granberg <baldrick@free.fr>
16835
16836 PR plugins/56754
ee49aa34 16837 * Makefile.in (PLUGIN_HEADERS): Add $(TARGET_H).
e60661f0 16838
4a61be9a
RB
168392013-05-21 Richard Biener <rguenther@suse.de>
16840
16841 PR tree-optimization/57318
16842 * tree-ssa-loop-ivcanon.c (tree_estimate_loop_size): Do not
16843 estimate stmts with side-effects as likely eliminated.
16844
c52da5f7
RB
168452013-05-21 Richard Biener <rguenther@suse.de>
16846
16847 PR tree-optimization/57330
16848 * cgraph.c (cgraph_redirect_edge_call_stmt_to_callee): Properly
16849 preserve the call stmts fntype.
16850
7ec67e2a
RB
168512013-05-21 Richard Biener <rguenther@suse.de>
16852
16853 PR tree-optimization/57303
16854 * tree-ssa-sink.c (statement_sink_location): Improve killing
16855 stmt detection and properly handle self-assignments.
16856
b112d513
CB
168572013-05-21 Christian Bruel <christian.bruel@st.com>
16858
55805e54
YR
16859 * dwarf2out.c (multiple_reg_loc_descriptor): Use dbx_reg_number for
16860 spanning registers. LEAF_REG_REMAP is supported only for contiguous
b112d513
CB
16861 registers. Set register size out of the PARALLEL loop.
16862
14c2ec26
OE
168632013-05-20 Oleg Endo <olegendo@gcc.gnu.org>
16864
16865 PR target/56547
16866 * config/sh/sh.md (fmasf4): Remove empty constraints strings.
16867 (*fmasf4, *fmasf4_media): New insns.
16868
da734fa1
RS
168692013-05-19 Richard Sandiford <rdsandiford@googlemail.com>
16870
16871 * config/mips/mips.h (BASE_INSN_LENGTH, NOP_INSN_LENGTH): New macros.
16872 * config/mips/mips.c (mips_symbol_insns, mips_address_insns)
16873 (mips_const_insns, mips_split_const_insns, mips_load_store_insns)
16874 (mips_idiv_insns): Update the comments to say that the returned
16875 instruction counts are in units of BASE_INSN_LENGTH.
16876 (mips_adjust_insn_length): Multiply the mips_load_label_num_insns
16877 by BASE_INSN_LENGTH rather than 4. Add the jump separately,
16878 using 2 rather than 4 as the length of indirect MIPS16 and
16879 microMIPS jumps. Use NOP_INSN_LENGTH rather than 4 as the
16880 length of a NOP. Don't divide MIPS16 lengths by 2.
16881 (mips16_split_long_branches): Assume a branch is long if the
16882 length is greater than 4 rather than 8.
16883 * config/mips/mips.md (length): Give MIPS16 lengths directly,
16884 rather than multiplying them by 2. Multiply instruction counts
16885 by BASE_INSN_LENGTH rather than 4.
16886 (*jump_mips16, tls_get_tp_mips16_<mode>)
16887 (*tls_get_tp_mips16_call_<mode>): Divide lengths by 2.
16888
13719e8b
RS
168892013-05-19 Richard Sandiford <rdsandiford@googlemail.com>
16890
16891 * config/mips/mips.md (extended_mips16): Remove branch case.
16892 (length): Remove duplicated extended_mips16 test.
16893
c3850d14
RS
168942013-05-19 Richard Sandiford <rdsandiford@googlemail.com>
16895
16896 * config/mips/t-sde: Don't build 64-bit microMIPS multilibs.
16897
8da2e059
RS
168982013-05-18 Richard Sandiford <rdsandiford@googlemail.com>
16899
16900 * recog.h (Recog_data): Rename to...
16901 (recog_data_d): ...this.
16902 (recog_data): Update accordingly.
16903 * recog.c (recog_data): Likewise.
16904 * reload.c (save_recog_data): Likewise.
16905 * config/picochip/picochip.c (picochip_saved_recog_data): Likewise.
16906 (picochip_save_recog_data, picochip_restore_recog_data): Likewise.
16907
2df013f3
JB
169082013-05-17 Julian Brown <julian@codesourcery.com>
16909
16910 * gcse.c (compute_ld_motion_mems): If a non-simple MEM is
16911 found in a REG_EQUAL note, invalidate it.
16912
371e77e3 169132013-05-17 Easwaran Raman <eraman@google.com>
933f507d
ER
16914
16915 * tree-ssa-reassoc.c (find_insert_point): New function.
16916 (insert_stmt_after): Likewise.
16917 (get_def_stmt): Likewise.
16918 (ensure_ops_are_available): Likewise.
16919 (not_dominated_by): Likewise.
16920 (rewrite_expr_tree): Do not move statements beyond what is
16921 necessary. Remove call to swap_ops_for_binary_stmt...
16922 (reassociate_bb): ... and move it here.
16923 (build_and_add_sum): Assign UIDs for new statements.
16924 (linearize_expr): Likewise.
16925 (do_reassoc): Renumber gimple statement UIDs.
16926
e01c7cca
JH
169272013-05-17 Jan Hubicka <jh@suse.cz>
16928
16929 * lto-symtab.c (lto_symtab_merge_cgraph_nodes): Resolve cross module
16930 weakrefs.
16931 * cgraph.c (dump_cgraph_node): Do not ice on unresolved alias.
e2420687
UB
16932 * cgraphunit.c (handle_alias_pairs): Store target of unresolved
16933 weakrefs.
e01c7cca
JH
16934 (output_weakrefs): Update.
16935
c3272a92
PCC
169362013-05-17 Po-Chun Chang <pchang9@cs.wisc.edu>
16937 Martin Jambor <mjambor@suse.cz>
16938
16939 PR middle-end/57276
16940 * ipa-cp.c (cgraph_edge_brings_all_agg_vals_for_node): Break when a
16941 value that corresponds to the given aggval is found in values vector.
16942
11c2aa39
UB
169432013-05-17 Uros Bizjak <ubizjak@gmail.com>
16944
16945 * config/i386/driver-i386.c (host_detect_local_cpu): Pass mmx, 3dnow,
16946 sse, sse2, sse3, ssse3 and sse4a flags to options.
16947
abecc8c6
DM
169482013-05-17 David Malcolm <dmalcolm@redhat.com>
16949
16950 * gengtype-state.c: (s_expr_writer): New class, to handle
16951 prettifying of output layout of s-expressions.
16952 (state_writer): New class, to write out gtype.state.
16953 (state_written_type_count): Move this variable into member data of
16954 state_writer.
16955 (s_expr_writer::s_expr_writer): New code: constructor for new class
16956 (state_writer::state_writer(): ditto
16957 (s_expr_writer::write_new_line): New function
16958 (s_expr_writer::write_any_indent): ditto
16959 (s_expr_writer::begin_s_expr): ditto
16960 (s_expr_writer::end_s_expr): ditto
16961 (write_state_fileloc): convert to method of state_writer...
16962 (state_writer:: write_state_fileloc): ...and use methods of
16963 s_expr_writer to write indentation into the gtype.state output file
16964 to visually represent the hierarchical structure of the list
16965 structures
16966 (write_state_fields): ditto, renaming to...
16967 (state_writer::write_state_fields)
16968 (write_state_a_string): ditto, renaming to...
16969 (state_writer::write_state_a_string)
16970 (write_state_string_option): ditto, renaming to...
16971 (state_writer::write_state_string_option)
16972 (write_state_type_option): ditto, renaming to...
16973 (state_writer::write_state_type_option)
16974 (write_state_nested_option): ditto, renaming to...
16975 (state_writer::write_state_nested_option)
16976 (write_state_option): ditto, renaming to...
16977 (state_writer::write_state_option)
16978 (write_state_options): ditto, renaming to...
16979 (state_writer::write_state_options)
16980 (write_state_lang_bitmap): ditto, renaming to...
16981 (state_writer::write_state_lang_bitmap)
16982 (write_state_version): ditto, renaming to...
16983 (state_writer::write_state_version)
16984 (write_state_scalar_type): ditto, renaming to...
16985 (state_writer::write_state_scalar_type)
16986 (write_state_string_type): ditto, renaming to...
16987 (state_writer::write_state_string_type)
16988 (write_state_undefined_type): ditto, renaming to...
16989 (state_writer::write_state_undefined_type)
16990 (write_state_struct_union_type): ditto, renaming to...
16991 (state_writer::write_state_struct_union_type)
16992 (write_state_struct_type): ditto, renaming to...
16993 (state_writer::write_state_struct_type)
16994 (write_state_user_struct_type): ditto, renaming to...
16995 (state_writer::write_state_user_struct_type)
16996 (write_state_lang_struct_type): ditto, renaming to...
16997 (state_writer::write_state_lang_struct_type)
16998 (write_state_param_struct_type): ditto, renaming to...
16999 (state_writer::write_state_param_struct_type)
17000 (write_state_pointer_type): ditto, renaming to...
17001 (state_writer::write_state_pointer_type)
17002 (write_state_array_type): ditto, renaming to...
17003 (state_writer::write_state_array_type)
17004 (write_state_gc_used): ditto, renaming to...
17005 (state_writer::write_state_gc_used)
17006 (write_state_common_type_content): ditto, renaming to...
17007 (state_writer::write_state_common_type_content)
17008 (write_state_type): ditto, renaming to...
17009 (state_writer::write_state_type)
17010 (write_state_pair_list): ditto, renaming to...
17011 (state_writer::write_state_pair_list)
17012 (write_state_pair): ditto, renaming to...
17013 (state_writer::write_state_pair)
17014 (write_state_typedefs): ditto, renaming to...
17015 (state_writer::write_state_typedefs)
17016 (write_state_structures): ditto, renaming to...
17017 (state_writer::write_state_structures)
17018 (write_state_param_structs): ditto, renaming to...
17019 (state_writer::write_state_param_structs)
17020 (write_state_variables): ditto, renaming to...
17021 (state_writer::write_state_variables)
17022 (write_state_srcdir): ditto, renaming to...
17023 (state_writer::write_state_srcdir)
17024 (write_state_files_list): ditto, renaming to...
17025 (state_writer::write_state_files_list)
17026 (write_state_languages): ditto, renaming to...
17027 (state_writer::write_state_languages)
17028 (write_state): create a state_writer instance and use it when
17029 writing out the state file
17030
d6545f29
MS
170312013-05-17 Mike Stump <mikestump@comcast.net>
17032
816a3f73 17033 PR rtl-optimization/57304
d6545f29
MS
17034 * web.c (union_match_dups): Ensure that DF_REF_LOC exists before
17035 accessing DF_REF_REAL_LOC.
17036
38c821cf
JJ
170372013-05-17 Jakub Jelinek <jakub@redhat.com>
17038
17039 PR rtl-optimization/57281
17040 PR rtl-optimization/57300
17041 * config/i386/i386.md (extendsidi2_1 dead reg splitter): Remove.
17042 (extendsidi2_1 peephole2s): Add instead 2 new peephole2s, that undo
17043 what the other splitter did if the registers are dead.
17044
2a293391
RB
170452013-05-17 Richard Biener <rguenther@suse.de>
17046
17047 * tree-ssa-alias.c (stmt_kills_ref_p_1): Properly compare
17048 MEM_REF offsets.
17049
17042d2b
JJ
170502013-05-17 Jakub Jelinek <jakub@redhat.com>
17051
17052 * gcc.c (SANITIZER_SPEC): Reject -fsanitize=address -fsanitize=thread
17053 linking.
17054
5b115c1f
MP
170552013-05-17 Marek Polacek <polacek@redhat.com>
17056
11c2aa39
UB
17057 * tree-ssa-strlen.c (handle_char_store): Don't invalidate cached
17058 length when doing non-zero store of storing '\0' to '\0'.
5b115c1f 17059
52d84413
JJ
170602013-05-17 Jakub Jelinek <jakub@redhat.com>
17061
68119618
JJ
17062 * tree-vect-patterns.c (vect_recog_rotate_pattern): For
17063 vect_external_def oprnd1 with loop_vinfo, try to emit
17064 optional cast, negation and and stmts on the loop preheader
17065 edge instead of into the pattern def seq.
17066
52d84413
JJ
17067 PR tree-optimization/57051
17068 * fold-const.c (const_binop) <case VEC_LSHIFT_EXPR,
17069 case VEC_RSHIFT_EXPR>: Fix BYTES_BIG_ENDIAN handling.
17070
17b962bd
NC
170712013-05-16 Nick Clifton <nickc@redhat.com>
17072
17073 * config/rl78/rl78.c (rl78_attribute_table): Add naked.
17074 (rl78_is_naked_func): New function.
56aefbf7
UB
17075 (rl78_expand_prologue): Skip prologue generation for naked functions.
17076 (rl78_expand_epilogue): Skip epilogue generation for naked functions.
17b962bd
NC
17077 * doc/extend.texi (naked): Add RL78 to the list of processors
17078 that supports this attribute.
17079
b1a0f84e
JL
170802013-05-16 Jeff Law <law@redhat.com>
17081
17082 * Makefile.in (tree-switch-conversion.o): Depend on $(OPTABS_H).
17083
af0e415b
UB
170842013-05-16 Uros Bizjak <ubizjak@gmail.com>
17085
17086 * config/i386/driver-i386.c (host_detect_local_cpu): Determine
17087 cache parameters using detect_caches_amd also for CYRIX,
17088 NSC and TM2 signatures.
17089
19db293a
UB
170902013-05-16 Uros Bizjak <ubizjak@gmail.com>
17091 Dzianis Kahanovich <mahatma@eu.by>
17092
17093 PR target/45359
17094 PR target/46396
17095 * config/i386/driver-i386.c (host_detect_local_cpu): Detect
17096 VIA/Centaur processors and determine their cache parameters
17097 using detect_caches_amd.
17098
251a41b9
TJ
170992013-05-16 Teresa Johnson <tejohnson@google.com>
17100
17101 * cfgrtl.c (verify_hot_cold_block_grouping): Return err.
17102 (rtl_verify_edges): New function.
17103 (rtl_verify_bb_insns): Ditto.
17104 (rtl_verify_bb_pointers): Ditto.
17105 (rtl_verify_bb_insn_chain): Ditto.
17106 (rtl_verify_fallthru): Ditto.
17107 (rtl_verify_bb_layout): Ditto.
17108 (rtl_verify_flow_info_1): Outline checks into new functions.
17109 (rtl_verify_flow_info): Ditto.
17110
f14540b6
SE
171112013-05-16 Steve Ellcey <sellcey@imgtec.com>
17112
17113 * cfghooks.c (copy_bbs): Add update_dominance argument.
17114 * cfghooks.h (copy_bbs): Update prototype.
17115 * tree-cfg.c (gimple_duplicate_sese_region):
17116 Add update_dominance argument.
17117 * tree-flow.h (gimple_duplicate_sese_region): Update prototype.
17118 * tree-ssa-loop-ch.c (copy_loop_headers): Update
17119 gimple_duplicate_sese_region call.
17120 * tree-vect-loop-manip.c (slpeel_tree_duplicate_loop_to_edge_cfg):
17121 Update copy_bbs call.
17122 * cfgloopmanip.c (duplicate_loop_to_header_edge): Ditto.
17123 * trans-mem.c (ipa_uninstrument_transaction): Ditto.
17124
7e9a3abb
JJ
171252013-05-16 Jakub Jelinek <jakub@redhat.com>
17126
17127 * tree-vectorizer.h (NUM_PATTERNS): Increment.
17128 * tree-vect-patterns.c (vect_vect_recog_func_ptrs): Add
17129 vect_recog_rotate_pattern.
17130 (vect_recog_rotate_pattern): New function.
17131
427b248d
JM
171322013-05-16 Jason Merrill <jason@redhat.com>
17133
17134 * Makefile.in (LLINKER): New variable.
17135 (mostlyclean): Remove link mutex.
17136 * configure.ac: Handle --enable-link-mutex.
17137 * lock-and-run.sh: New script.
17138
b871e3d2
RR
171392013-05-16 Ramana Radhakrishnan <ramana.radhakrishnan@arm.com>
17140
17141 PR target/19599
17142 * config/arm/arm.c (arm_function_ok_for_sibcall): Add check
17143 for NULL decl.
17144
ce858126
RO
171452013-05-16 Rainer Orth <ro@CeBiTec.Uni-Bielefeld.DE>
17146
17147 * reorg.c (link_cc0_insns): Wrap in #ifdef HAVE_cc0.
17148
798d3d04
GY
171492013-05-16 Greta Yorsh <Greta.Yorsh@arm.com>
17150
17151 * config/arm/arm-protos.h (gen_movmem_ldrd_strd): New declaration.
17152 * config/arm/arm.c (next_consecutive_mem): New function.
17153 (gen_movmem_ldrd_strd): Likewise.
17154 * config/arm/arm.md (movmemqi): Update condition and code.
17155 (unaligned_loaddi, unaligned_storedi): New patterns.
17156
0baddc45
RO
171572013-05-16 Rainer Orth <ro@CeBiTec.Uni-Bielefeld.DE>
17158
17159 * config.gcc: Obsolete *-*-solaris2.9*.
17160 * doc/install.texi (Specific, *-*-solaris2*): Document it.
17161
aa9d5bcf
RB
171622013-05-16 Richard Biener <rguenther@suse.de>
17163
17164 * passes.c (init_optimization_passes): Move pass_parallelize_loops
17165 earlier, after GRAPHITE transforms and IV canonicalization.
17166
5a0f4dd3
JJ
171672013-05-16 Jakub Jelinek <jakub@redhat.com>
17168
17169 * omp-low.c (extract_omp_for_data): For collapsed loops,
17170 if at least one of the loops is known at compile time to
17171 iterate zero times, set count to 0.
17172 (expand_omp_regimplify_p): New function.
17173 (expand_omp_for_generic): For collapsed loops, if at least
17174 one of the loops isn't known to iterate at least once,
17175 add runtime check with setting count to 0.
17176 (expand_omp_for_static_nochunk, expand_omp_for_static_chunk):
17177 For unsigned types if it isn't known at compile time that
17178 the loop will iterate at least once, add runtime check to bypass
17179 the whole loop if initial condition isn't true.
17180
e3753785
NS
171812013-05-16 Nathan Sidwell <nathan@codesourcery.com>
17182
17183 * varasm.c (default_use_anchors_for_symbol_p): Use decl_replaceable_p.
17184
43bb4dd1
MG
171852013-05-16 Marc Glisse <marc.glisse@inria.fr>
17186
17187 PR middle-end/57286
17188 * fold-const.c (fold_ternary_loc) <VEC_COND_EXPR>: Disable some
17189 transformations to avoid an infinite loop.
17190
3571dde6
MP
171912013-05-16 Marek Polacek <polacek@redhat.com>
17192
17193 * tree-scalar-evolution.c (scev_const_prop): Add more dumps.
17194
b8b3f0ca
LE
171952013-05-15 Leif Ekblad <leif@rdos.net>
17196
17197 * config/i386/i386.c (ix86_decompose_address): Use
17198 DEFAULT_TLS_SEG_REG to access TLS segment register.
17199 * config/i386/i386.h (DEFAULT_TLS_SEG_REG): New define.
17200 * config/i386/rdos.h (DEFAULT_TLS_SEG_REG): Ditto.
17201 (TARGET_TLS_DIRECT_SEG_REFS_DEFAULT): Ditto.
17202
e299a383
RS
172032013-05-15 Richard Sandiford <rdsandiford@googlemail.com>
17204
17205 PR target/57260
17206 * config/mips/mips.c (mips_function_ok_for_sibcall): Don't allow
17207 sibling calls to functions that would normally be lazily bound,
17208 unless $gp is call-clobbered.
17209
e7208ea3
UB
172102013-05-15 Uros Bizjak <ubizjak@gmail.com>
17211
19db293a 17212 * config/i386/i386.c (ix86_option_override_internal): Update
e7208ea3 17213 processor_alias_table for missing PTA_PRFCHW and PTA_FXSR flags. Add
5d280e2b
UB
17214 PTA_POPCNT to corei7 entry. Do not enable SSE prefetch on
17215 non-SSE 3dNow! targets. Enable TARGET_PRFCHW for TARGET_3DNOW targets.
e7208ea3
UB
17216 * config/i386/i386.md (prefetch): Enable for TARGET_PRFCHW instead
17217 of TARGET_3DNOW.
17218 (*prefetch_3dnow): Enable for TARGET_PRFCHW only.
17219
e7413f3d
AS
172202013-05-15 Andreas Schwab <schwab@suse.de>
17221
17222 * config/m68k/m68k.md (*rotlhi3_lowpart, *rotlqi3_lowpart): Name
17223 for rotlhi3+1 and rotlqi3+1, resp. Fix reference to non-existing
17224 third operand.
17225
ad4db775
TJ
172262013-05-15 Teresa Johnson <tejohnson@google.com>
17227
17228 * loop-unroll.c (report_unroll_peel): Check decision before
17229 emitting unroll/peel message.
17230
af205f67
TJ
172312013-05-15 Teresa Johnson <tejohnson@google.com>
17232
17233 * function.h (has_bb_partition): New rtl_data flag.
17234 (bb_reorder_complete): Ditto.
17235 * cfgcleanup.c (try_crossjump_to_edge): Check for has_bb_partition
17236 instead of flag_reorder_blocks_and_partition.
17237 * cfgrtl.c (verify_hot_cold_block_grouping): Moved from bb-reorder.c,
17238 with some enhancements.
17239 (rtl_verify_flow_info_1): Call verify_hot_cold_block_grouping.
17240 * bb-reorder.c (connect_traces): Check for has_bb_partition
17241 instead of flag_reorder_blocks_and_partition.
17242 (verify_hot_cold_block_grouping): Moved to cfgrtl.c.
17243 (reorder_basic_blocks): Set bb_reorder_complete flag, remove call to
17244 verify_hot_cold_block_grouping.
17245 (partition_hot_cold_basic_blocks): Set has_bb_partition.
17246
9adcfa3c
RR
172472013-05-15 Ramana Radhakrishnan <ramana.radhakrishnan@arm.com>
17248
17249 PR target/19599
17250 * config/arm/predicates.md (call_insn_operand): New predicate.
17251 * config/arm/constraints.md ("Cs", "Ss"): New constraints.
17252 * config/arm/arm.md (*call_insn, *call_value_insn): Match only
17253 if insn is not a tail call.
17254 (*sibcall_insn, *sibcall_value_insn): Adjust for tailcalling through
17255 registers.
17256 * config/arm/arm.h (enum reg_class): New caller save register class.
17257 (REG_CLASS_NAMES): Likewise.
17258 (REG_CLASS_CONTENTS): Likewise.
17259 * config/arm/arm.c (arm_function_ok_for_sibcall): Allow tailcalling
17260 without decls.
17261
ccb3ad87
RB
172622013-05-15 Richard Biener <rguenther@suse.de>
17263
17264 * tree-vect-loop.c (vect_transform_loop): Use MSG_NOTE instead
17265 of MSG_OPTIMIZED_LOCATIONS.
17266 * tree-vect-slp.c (vect_make_slp_decision): Likewise.
17267 (vect_slp_transform_bb): Indicate location in MSG_OPTIMIZED_LOCATIONS
17268 message.
17269 * tree-vectorizer.c (vectorize_loops): Use MSG_NOTE instead
17270 of MSG_OPTIMIZED_LOCATIONS.
17271 (execute_vect_slp): Likewise.
17272 * tree-vect-loop-manip.c (vect_do_peeling_for_loop_bound): Likewise.
17273 (vect_create_cond_for_alias_checks): Likewise.
17274 * tree-vect-patterns.c (vect_recog_dot_prod_pattern): Likewise.
17275 (vect_recog_widen_mult_pattern): Likewise.
17276 (vect_recog_widen_sum_pattern): Likewise.
17277 (vect_recog_over_widening_pattern): Likewise.
17278 (vect_recog_widen_shift_pattern): Likewise.
17279 (vect_recog_vector_vector_shift_pattern): Likewise.
17280 (vect_recog_divmod_pattern): Likewise.
17281 (vect_recog_mixed_size_cond_pattern): Likewise.
17282 (vect_recog_bool_pattern): Likewise.
17283 (vect_pattern_recog_1): Likewise.
17284
48b1474e
MJ
172852013-05-15 Martin Jambor <mjambor@suse.cz>
17286
17287 * ipa-prop.c (ipa_make_edge_direct_to_target): Redirect calls to
17288 non-functions to builtin_unreachable.
17289 * ipa-inline-transform.c (inline_call): Do not assert estimates were
17290 correct when new direct edges were discovered.
17291
9de04252
MJ
172922013-05-15 Martin Jambor <mjambor@suse.cz>
17293
17294 * ipa-prop.c (ipa_print_node_jump_functions): Print symbol order in
17295 header, print symbol order instead of node uid, print more information
17296 about indirect edge targets.
17297 (ipa_make_edge_direct_to_target): Print symbol order instead of node
17298 uids.
17299 (ipa_make_edge_direct_to_target): Likewise.
17300 (remove_described_reference): Likewise.
17301 (propagate_controlled_uses): Likewise.
17302 (ipa_print_node_params): Also print symbol order.
17303 (ipcp_transform_function): Print symbol order instead of node uids.
17304 * cgraph.c (cgraph_redirect_edge_call_stmt_to_callee): Likewise.
17305 (cgraph_get_create_real_symbol_node): Likewise.
17306 * ipa-cp.c (print_lattice): Likewise.
17307 (print_all_lattices): Likewise.
17308 (determine_versionability): Likewise.
17309 (initialize_node_lattices): Likewise.
17310 (estimate_local_effects): Likewise.
17311 (update_profiling_info): Likewise.
17312 (create_specialized_node): Likewise.
17313 (perhaps_add_new_callers): Likewise.
17314 (decide_about_value): Likewise.
17315 (decide_whether_version_node): Likewise.
17316 (identify_dead_nodes): Likewise.
17317 * ipa-inline-analysis.c (dump_inline_edge_summary): Likewise.
17318 (dump_inline_summary): Likewise.
17319 (estimate_node_size_and_time): Likewise.
17320 (inline_analyze_function): Likewise.
17321 * ipa-inline.c (report_inline_failed_reason): Likewise.
17322 (want_early_inline_function_p): Likewise.
17323 (edge_badness): Likewise.
17324 (update_edge_key): Likewise.
17325 (inline_small_functions): Likewise. Add dumping of order to two other
17326 dumps.
17327 * ipa-pure-const.c (pure_const_read_summary): Print symbol order
17328 instead of node uids.
17329 (propagate_pure_const): Likewise.
17330 (propagate_pure_const): Likewise.
17331 * ipa-utils.c (dump_cgraph_node_set): Likewise.
17332 * lto-cgraph.c (input_node): Explicitly specify we dump uid.
17333 * lto-symtab.c (lto_cgraph_replace_node): Print symbol order instead
17334 of node uids.
17335 * tree-pretty-print.c (dump_function_header): Likewise.
17336 * tree-sra.c (convert_callers_for_node): Dump in traditional format.
17337 Print symbol order instead of node uids.
17338
1dd03b91
AK
173392013-05-15 Andreas Krebbel <Andreas.Krebbel@de.ibm.com>
17340
17341 * config/s390/s390.c (s390_register_move_cost): Don't impose the
17342 FPR<->GPR move cost penalty if ldgr/lgdr can be used.
17343
fdf6a7b9
RB
173442013-05-15 Richard Biener <rguenther@suse.de>
17345
17346 PR tree-optimization/57275
17347 * tree-vect-data-refs.c (vect_analyze_data_ref_dependence): Fix
17348 return value for fail to do runtime alias checks for gather loads.
17349
2d6e4603
JH
173502013-05-15 Jan Hubicka <jh@suse.cz>
17351
17352 PR lto/57038
17353 PR lto/47375
e7208ea3
UB
17354 * lto-symtab.c (lto_symtab_symbol_p): Add external symbol;
17355 weakrefs are not external.
17356 (lto_symtab_merge_decls): Fix thinko when dealing with
17357 non-lto_symtab decls.
2d6e4603
JH
17358 (lto_symtab_merge_cgraph_nodes): Use lto_symtab_symbol_p.
17359 (lto_symtab_prevailing_decl): Get int sync with lto_symtab_symbol_p.
17360 * varpool.c (dump_varpool_node): Dump more flags.
17361
83f44b39
GG
173622013-05-15 Ganesh Gopalasubramanian <Ganesh.Gopalasubramanian@amd.com>
17363
17364 * config/i386/i386.c (processor_alias_table): Add instruction
17365 FSGSBASE for AMD bdver3 architecture.
17366
2e55d062
JJ
173672013-05-14 Jakub Jelinek <jakub@redhat.com>
17368
17369 * tree.c (warn_deprecated_use): Print file:line using locus color.
17370 * diagnostic.c (diagnostic_report_current_module): Print file:line
17371 and file:line:column using locus color.
17372
0bfdb81e
MS
173732013-05-14 Mike Stump <mikestump@comcast.net>
17374
17375 * gdbinit.in: Add __null.
17376
a508ef22
MS
173772013-05-14 Mike Stump <mikestump@comcast.net>
17378
17379 * recog.h: Rename struct recog_data to Recog_data.
17380 * recog.c: Likewise.
17381 * reload.c (can_reload_into): Likewise.
17382 * config/picochip/picochip.c: Likewise.
17383
e7180acb
MS
173842013-05-14 Mike Stump <mikestump@comcast.net>
17385
17386 * web.c (union_match_dups): Also check DF_REF_REAL_LOC.
17387
61216c26
SB
173882013-05-14 Steven Bosscher <steven@gcc.gnu.org>
17389
fd6beed4
SB
17390 * resource.h (struct resources): Remove unch_memory member.
17391 (CLEAR_RESOURCE): Don't clear unch_memory.
17392 * resource.c (mark_referenced_resources): Don't set it.
17393 (mark_set_resources): Likewise.
17394 (mark_target_live_regs): Don't clear it.
17395 (init_resource_info): Likewise.
17396 * reorg.c (resource_conflicts_p): Don't compare it.
17397 (redundant_insn): Don't set it.
17398
61216c26
SB
17399 * rtl.h (next_label, skip_consecutive_labels, link_cc0_insns):
17400 Remove prototypes.
17401 * emit-rtl.c (next_label): Remove unused function.
17402 (skip_consecutive_labels, link_cc0_insns): Move to ...
17403 * reorg.c (skip_consecutive_labels, link_cc0_insns): ... here, the
17404 only place where these functions are used, and make them static.
17405
418d1b87
MG
174062013-05-14 Marc Glisse <marc.glisse@inria.fr>
17407
17408 * fold-const.c (fold_negate_expr): Handle vectors.
17409 (fold_truth_not_expr): Make it static.
17410 (fold_invert_truthvalue): New static function.
17411 (invert_truthvalue_loc): Handle vectors. Do not call
17412 fold_truth_not_expr directly.
17413 (fold_unary_loc) <BIT_NOT_EXPR>: Handle comparisons.
17414 <TRUTH_NOT_EXPR>: Do not cast to boolean.
17415 (fold_comparison): Handle vector constants.
17416 (fold_binary_loc) <TRUTH_XOR_EXPR>: Remove redundant code.
17417 (fold_ternary_loc) <VEC_COND_EXPR>: Adapt more COND_EXPR optimizations.
17418 * tree.h (fold_truth_not_expr): Remove declaration.
17419
fc21784d
JG
174202013-05-14 James Greenhalgh <james.greenhalgh@arm.com>
17421
17422 * config/aarch64/aarch64-simd.md
17423 (aarch64_vcond_internal<mode>): Rename to...
17424 (aarch64_vcond_internal<mode><mode>): ...This, for integer modes.
17425 (aarch64_vcond_internal<VDQF_COND:mode><VDQF:mode>): ...This for
17426 float modes. Clarify all iterator modes.
17427 (vcond<mode><mode>): Use new name for vcond expanders.
17428 (vcond<v_cmp_result><mode>): Likewise.
17429 (vcondu<mode><mode>: Likewise.
17430 * config/aarch64/iterators.md (VDQF_COND): New.
17431
d4c52634
MG
174322013-05-14 Marc Glisse <marc.glisse@inria.fr>
17433
17434 PR bootstrap/57266
17435 * fold-const.c (fold_binary_loc) <shift>: Use an unsigned
17436 variable for the shift amount. Check that we shift by non-negative
17437 amounts.
17438
2b261262
CLT
174392013-05-14 Chung-Lin Tang <cltang@codesourcery.com>
17440
17441 PR target/42017
17442 * config/arm/arm.h (EPILOGUE_USES): Only return true
17443 for LR_REGNUM after epilogue_completed.
17444
12211b99 174452013-05-14 Joern Rennecke <joern.rennecke@embecosm.com>
6e022d7b
JR
17446
17447 * config/avr/avr.c (avr_encode_section_info): Bail out if the type
e7208ea3 17448 is error_mark_node.
6e022d7b 17449
56cf7859
RO
174502013-05-14 Rainer Orth <ro@CeBiTec.Uni-Bielefeld.DE>
17451
17452 PR target/57261
17453 * configure.ac (gcc_cv_ld_as_needed): Disable before Solaris 11
17454 and Solaris 11+/x86 with gld.
17455 * configure: Regenerate.
17456
75776c6d
JJ
174572013-05-14 Jakub Jelinek <jakub@redhat.com>
17458
17459 * expmed.c (expand_shift_1): Canonicalize rotates by
17460 constant bitsize / 2 to bitsize - 1.
7f998021 17461 * simplify-rtx.c (simplify_binary_operation_1) <case ROTATE,
75776c6d
JJ
17462 case ROTATERT>: Likewise.
17463
17464 Revert:
17465 2013-05-10 Jakub Jelinek <jakub@redhat.com>
17466
17467 * config/i386/i386.md (rotateinv): New code attr.
17468 (*<rotate_insn><mode>3_1, *<rotate_insn>si3_1_zext,
17469 *<rotate_insn>qi3_1_slp): Emit rorl %eax instead of
17470 roll $31, %eax, etc.
17471
df35498a
RB
174722013-05-14 Richard Biener <rguenther@suse.de>
17473
17474 PR middle-end/57235
17475 * tree-eh.c (sink_clobbers): Give up for successors with
17476 multiple predecessors and no virtual uses.
17477
cc6e7ece
EB
174782013-05-14 Eric Botcazou <ebotcazou@adacore.com>
17479
17480 * config/sparc/sp64-elf.h (CPP_SUBTARGET_SPEC): Delete.
17481 * config/sparc/openbsd64.h (CPP_SUBTARGET_SPEC): Likewise.
17482
f778dd4d
JJ
174832013-05-14 Jakub Jelinek <jakub@redhat.com>
17484
17485 PR middle-end/57251
17486 * expr.c (expand_expr_real_2) <case WIDEN_MULT_EXPR>: Handle
17487 the case when both op0 and op1 have VOIDmode.
17488
bad4df9b
KP
174892013-05-14 Kaushik Phatak <kaushik.phatak@kpitcummins.com>
17490
17491 * config/rl78/rl78.md(mulsi3_g13): Add additional 'nop' required
17492 in multiply-accumulate mode.
17493
56f3e9ac
GW
174942013-05-13 Guozhi Wei <carrot@google.com>
17495
17496 * dwarf2asm.c (dw2_output_indirect_constant_1): Mark new decl STATIC.
17497
a3d7ab92
KT
174982013-05-13 Kai Tietz <ktietz@redhat.com>
17499
17500 PR target/56975
17501 * config/i386/cygming.h (TARGET_PECOFF): Define as true.
e7208ea3 17502 * config/i386/i386.h (TARGET_PECOFF): Define by default as false.
a3d7ab92 17503 (PIC_OFFSET_TABLE_REGNUM): Use TARGET_PECOFF.
e7208ea3 17504 * config/i386/i386.c (ix86_option_override_internal): Likewise.
a3d7ab92
KT
17505 (ix86_expand_prologue): Likewise.
17506 (ix86_expand_split_stack_prologue): Likewise.
17507 (legitimate_pic_address_disp_p): Likewise.
17508 (legitimize_pic_address): Likewise.
17509 (legitimize_tls_address): Likewise.
17510 (legitimize_pe_coff_symbol): Likewise.
17511 (output_pic_addr_const): Likewise.
17512 (construct_plt_address): Likewise.
17513 (ix86_expand_call): Likewise.
17514 (x86_output_mi_thunk): Likewise.
17515 (x86_function_profiler): Likewise.
17516
c59b7e28
SN
175172013-05-13 Sofiane Naci <sofiane.naci@arm.com>
17518
17519 * config/aarch64/aarch64-simd.md (aarch64_simd_mov<mode>): Group
17520 similar switch cases.
17521 (aarch64_simd_mov): Rename to aarch64_split_simd_mov. Update.
17522 (aarch64_simd_mov_to_<mode>low): Delete.
17523 (aarch64_simd_mov_to_<mode>high): Delete.
17524 (move_lo_quad_<mode>): Add w<-r alternative.
17525 (aarch64_simd_move_hi_quad_<mode>): Likewise.
17526 (aarch64_simd_mov_from_*): Update type attribute.
17527 * config/aarch64/aarch64.c (aarch64_split_simd_move): Refacror switch
17528 statement.
17529
5f28524a
JH
175302013-05-13 Jan Hubicka <jh@suse.cz>
17531
17532 * mode-switching.c (optimize_mode_switching): Set correct RTL profile.
17533 * config/i386/i386.c (ix86_compute_frame_layout,
e7208ea3
UB
17534 ix86_expand_epilogue, emit_i387_cw_initialization,
17535 ix86_expand_vector_move_misalign, ix86_fp_comparison_strategy,
17536 ix86_local_alignment): Fix use of size/speed predicates.
5f28524a 17537
ae6fa899
JJ
175382013-05-13 Jakub Jelinek <jakub@redhat.com>
17539
17540 PR tree-optimization/45216
17541 PR tree-optimization/57157
17542 * tree-ssa-forwprop.c (simplify_rotate): Only recognize
17543 the (-Y) & (B - 1) variant if OP is |.
17544 * expmed.c (expand_shift_1): For rotations by const0_rtx just
17545 return shifted. Use (-op1) & (prec - 1) as other_amount
17546 instead of prec - op1.
17547
4502fe8d
MJ
175482013-05-13 Martin Jambor <mjambor@suse.cz>
17549
17550 PR middle-end/42371
17551 * ipa-prop.h (IPA_UNDESCRIBED_USE): New macro.
17552 (ipa_constant_data): New type.
17553 (ipa_jump_func): Use ipa_constant_data to hold information about
17554 constant jump functions.
17555 (ipa_get_jf_constant): Adjust to jump function type changes.
17556 (ipa_get_jf_constant_rdesc): New function.
17557 (ipa_param_descriptor): New field controlled_uses.
17558 (ipa_get_controlled_uses): New function.
17559 (ipa_set_controlled_uses): Likewise.
17560 * ipa-ref.h (ipa_find_reference): Declare.
17561 * ipa-prop.c (ipa_cst_ref_desc): New type.
17562 (ipa_print_node_jump_functions_for_edge): Adjust for jump function type
17563 changes.
17564 (ipa_set_jf_constant): Likewise. Also create reference descriptions.
17565 New parameter cs. Adjust all callers.
17566 (ipa_analyze_params_uses): Detect uncontrolled and controlled uses.
17567 (remove_described_reference): New function.
17568 (jfunc_rdesc_usable): Likewise.
17569 (try_make_edge_direct_simple_call): Decrement controlled use count,
17570 attempt to remove reference if it hits zero.
17571 (combine_controlled_uses_counters): New function.
17572 (propagate_controlled_uses): Likewise.
17573 (ipa_propagate_indirect_call_infos): Call propagate_controlled_uses.
17574 (ipa_edge_duplication_hook): Duplicate reference descriptions.
17575 (ipa_print_node_params): Print described use counter.
17576 (ipa_write_jump_function): Adjust to jump function type changes.
17577 (ipa_read_jump_function): New parameter CS, pass it to
17578 ipa_set_jf_constant. Adjust caller.
17579 (ipa_write_node_info): Stream controlled use count
17580 (ipa_read_node_info): Likewise.
17581 * cgraph.c (cgraph_mark_address_taken_node): Bail out instead of
17582 asserting.
17583 * ipa-cp.c (ipcp_discover_new_direct_edges): Decrement controlled use
17584 count. Remove cloning-added reference if it reaches zero.
17585 * ipa-ref.c (ipa_find_reference): New function.
17586
0864bfc2
GG
175872013-05-13 Ganesh Gopalasubramanian <Ganesh.Gopalasubramanian@amd.com>
17588
e7208ea3 17589 * config/i386/i386.c (processor_target_table): Modified default
0864bfc2
GG
17590 alignment values for AMD BD and BT architectures.
17591
640bfeb2
MG
175922013-05-13 Marc Glisse <marc.glisse@inria.fr>
17593
17594 * tree-vect-generic.c (uniform_vector_p): Move ...
17595 * tree.c (uniform_vector_p): ... here.
17596 * tree.h (uniform_vector_p): Declare it.
17597 * fold-const.c (fold_binary_loc) <shift>: Turn the second argument
17598 into a scalar.
17599
3a60f32b
JJ
176002013-05-13 Jakub Jelinek <jakub@redhat.com>
17601
198fe1bf
JJ
17602 PR tree-optimization/57230
17603 * tree-ssa-strlen.c (handle_char_store): Record length for
17604 array store from STRING_CST.
17605
3a60f32b
JJ
17606 PR tree-optimization/57230
17607 * tree-ssa-strlen.c (handle_char_store): Add missing integer_zerop
17608 check.
17609
566be57c
JR
176102013-05-12 Joern Rennecke <joern.rennecke@embecosm.com>
17611
17612 * config/epiphany/epiphany.c (epiphany_init): Check size of
17613 NUM_MODES_FOR_MODE_SWITCHING.
17614 (epiphany_expand_prologue):
17615 Remove CONFIG_REGNUM initial value handling code.
17616 (epiphany_optimize_mode_switching): Handle EPIPHANY_MSW_ENTITY_CONFIG.
17617 (epiphany_mode_needed, epiphany_mode_entry_exit): Likewise.
2710a27a 17618 (emit_set_fp_mode, epiphany_mode_after): Likewise.
566be57c
JR
17619 (epiphany_mode_needed) <Handle EPIPHANY_MSW_ENTITY_AND>:
17620 Don't return 1 for FP_MODE_NONE.
17621 * config/epiphany/epiphany.h (NUM_MODES_FOR_MODE_SWITCHING):
17622 Add value for EPIPHANY_MSW_ENTITY_CONFIG.
17623 (EPIPHANY_MSW_ENTITY_CONFIG, EPIPHANY_MSW_ENTITY_NUM): Define.
17624 * config/epiphany/epiphany.md (save_config): New pattern.
17625
0f2c2331
UB
176262013-05-12 Uros Bizjak <ubizjak@gmail.com>
17627
17628 * config/i386/i386.md (*zero_extendsidi2): Add *x->?r alternative.
17629
5b3f93c7
UB
176302013-05-10 Uros Bizjak <ubizjak@gmail.com>
17631
17632 * config/i386/i386.md (memory): Handle sseishft1.
17633 * config/i386/sse.md (*vec_extractv4si): Remove memory attribute.
17634 (*vec_extractv2di_1): Ditto.
17635
1f873f0e
VM
176362013-05-10 Vladimir Makarov <vmakarov@redhat.com>
17637
17638 * lra-assigns.c (find_hard_regno_for): Add 1 to the cost of call
17639 saved registers.
17640
341427fa 176412013-05-10 Sebastian Huber <sebastian.huber@embedded-brains.de>
d7de6455
SH
17642
17643 * config/arm/t-rtems-eabi: Remove mthumb/march=armv7 multilib.
17644 Add mthumb/march=armv7-a multilib.
17645 Add mthumb/march=armv7-r multilib.
17646 Add mthumb/march=armv7-a/mfpu=neon/mfloat-abi=hard multilib.
17647
9e69bdde
RC
176482013-05-10 Ralf Corsépius <ralf.corsepius@rtems.org>
17649
17650 * config/v850/t-rtems: Add more multilibs.
17651
9ff09a22
RB
176522013-05-10 Richard Biener <rguenther@suse.de>
17653
17654 PR tree-optimization/57214
17655 * tree-ssa-loop-ivcanon.c (propagate_constants_for_unrolling): Do
17656 not propagate from SSA names that occur in abnormal PHI nodes.
17657
a5e0cd1d
MG
176582013-05-10 Marc Glisse <marc.glisse@inria.fr>
17659
17660 * stor-layout.c (element_precision): New function.
17661 * machmode.h (element_precision): Declare it.
17662 * tree.c (build_minus_one_cst): New function.
17663 (element_precision): Likewise.
17664 * tree.h (build_minus_one_cst): Declare new function.
17665 (element_precision): Likewise.
17666 * fold-const.c (operand_equal_p): Use element_precision.
17667 (fold_binary_loc): Handle vector types.
17668 * convert.c (convert_to_integer): Use element_precision.
17669 * gimple.c (iterative_hash_canonical_type): Handle complex and vectors
17670 separately.
17671
cb2558bc
RS
176722013-05-10 Richard Sandiford <rdsandiford@googlemail.com>
17673
17674 * config/mips/mips-protos.h (m16_uimm3_b, m16_simm4_1, m16_nsimm4_1)
17675 (m16_simm5_1, m16_nsimm5_1, m16_uimm5_4, m16_nuimm5_4, m16_simm8_1)
17676 (m16_nsimm8_1, m16_uimm8_1, m16_nuimm8_1, m16_uimm8_m1_1, m16_uimm8_4)
17677 (m16_nuimm8_4, m16_simm8_8, m16_nsimm8_8): Delete.
17678 * config/mips/mips.c (m16_check_op, m16_uimm3_b, m16_simm4_1)
17679 (m16_nsimm4_1, m16_simm5_1, m16_nsimm5_1, m16_uimm5_4, m16_nuimm5_4)
17680 (m16_simm8_1, m16_nsimm8_1, m16_uimm8_1, m16_nuimm8_1, m16_uimm8_m1_1)
17681 (m16_uimm8_4, m16_nuimm8_4, m16_simm8_8, m16_nsimm8_8): Delete.
17682 * config/mips/constraints.md (Udb8, Usb5, Usb8, Usd8, Uub8, Uuw5)
17683 (Uuw8): New constraints.
17684 (Usb4): Move into alphabetical order.
17685 * config/mips/predicates.md (db8_operand, sb5_operand, sb8_operand)
17686 (sd8_operand, ub8_operand, uw8_operand): New predicates.
17687 * config/mips/mips.md (*xor<mode>3, *xor<mode>3_mips16): Name
17688 previously unnamed patterns.
17689 (*add<mode>3_mips16, *xor<mode>3_mips16, *<optab>si3_mips16)
17690 (*ashldi3_mips16, *ashrdi3_mips16, *lshrdi3_mips16)
17691 (*slt<u>_<GPR:mode><GPR2:mode>_mips16)
17692 (*sle<u>_<GPR:mode><GPR2:mode>_mips16): Use constraints instead
17693 of set_attr_alternative/if_then_else. Use extended_mips16 instead
17694 of specific lengths.
17695
cb3b8d33
JJ
176962013-05-10 Jakub Jelinek <jakub@redhat.com>
17697
6f93c008
JJ
17698 * config/i386/i386.md (rotateinv): New code attr.
17699 (*<rotate_insn><mode>3_1, *<rotate_insn>si3_1_zext,
17700 *<rotate_insn>qi3_1_slp): Emit rorl %eax instead of
17701 roll $31, %eax, etc.
17702
cb3b8d33
JJ
17703 PR tree-optimization/45216
17704 PR tree-optimization/57157
17705 * tree-ssa-forwprop.c (simplify_rotate): New function.
17706 (ssa_forward_propagate_and_combine): Call it.
17707
afb119be
RB
177082013-05-10 Richard Biener <rguenther@suse.de>
17709
17710 * tree-vect-data-refs.c (vect_enhance_data_refs_alignment): Do not
17711 disable peeling when we version for aliasing.
17712 (vector_alignment_reachable_p): Honor explicit user alignment.
17713 (vect_supportable_dr_alignment): Likewise.
17714 * tree-vect-loop-manip.c (vect_can_advance_ivs_p): Use
17715 STMT_VINFO_LOOP_PHI_EVOLUTION_PART instead of recomputing it.
17716 * tree-vect-loop.c (vect_transform_loop): First apply versioning,
17717 then peeling to arrange for the cost-model check to come first.
17718
01ae4861
AM
177192013-05-10 Alan Modra <amodra@gmail.com>
17720
17721 * configure.ac (HAVE_AS_TLS): Swap powerpc64 and powerpc cases.
17722 (HAVE_LD_LARGE_TOC): Don't mention AIX in help text.
17723 * configure: Regenerate.
17724
ebc9a431
AM
177252013-05-10 Alan Modra <amodra@gmail.com>
17726
17727 PR target/55033
17728 * varasm.c (default_elf_select_section): Move !DECL_P check..
17729 (get_named_section): ..to here before calling get_section_name.
17730 Adjust assertion.
17731 (default_section_type_flags): Add DECL_P check.
17732 * config/i386/winnt.c (i386_pe_section_type_flags): Likewise.
17733 * config/rs6000/rs6000.c (rs6000_xcoff_section_type_flags): Likewise.
17734
d4bca93c
JR
177352013-05-09 Joern Rennecke <joern.rennecke@embecosm.com>
17736
17737 * config/epiphany/epiphany.c (epiphany_expand_prologue):
17738 When using gen_stack_adjust_str with a register offset, add a
17739 REG_FRAME_RELATED_EXPR note.
17740
60ca9a65
UB
177412013-05-09 Uros Bizjak <ubizjak@gmail.com>
17742
17743 * config/i386/sse.md (*vec_extractv4si_0_zext): New pattern.
17744 (*vec_extractv4si_zext_mem): Ditto.
17745 (*vec_extractv2di): Add 0->x and x->x alternatives.
17746 * config/i386/mmx.md (*vec_extractv2si_zext_mem): New pattern.
17747 * config/i386/i386.md (*zero_extendsidi2): Add *Yj->?r alternative.
17748
a3409c02
JM
177492013-05-09 Jason Merrill <jason@redhat.com>
17750
0138d6b2
JM
17751 N3639 C++1y VLA support
17752 * gimplify.c (gimplify_vla_decl): Don't touch an existing
17753 DECL_VALUE_EXPR.
17754
a3409c02
JM
17755 * tree.c (build_constructor_va): New.
17756 * tree.h: Declare it.
17757
66e6b990
MJ
177582013-05-09 Martin Jambor <mjambor@suse.cz>
17759
17760 PR lto/57084
17761 * gimple-fold.c (canonicalize_constructor_val): Call
17762 cgraph_get_create_real_symbol_node instead of cgraph_get_create_node.
17763
64cfa6c0
JH
177642013-05-09 Jan Hubicka <jh@suse.cz>
17765 Richard Biener <rguenther@suse.de>
17766
17767 PR lto/54095
17768 * symtab.c (symtab_make_decl_local): Do not add private names.
17769
c3167b00
JH
177702013-05-09 Jan Hubicka <jh@suse.cz>
17771
17772 PR lto/54095
17773 * symtab.c (insert_to_assembler_name_hash): Handle clones.
17774 (unlink_from_assembler_name_hash): Likewise.
17775 (symtab_prevail_in_asm_name_hash, symtab_register_node,
17776 symtab_unregister_node, symtab_initialize_asm_name_hash,
17777 change_decl_assembler_name): Update.
17778
12dc6974
SN
177792013-05-09 Sofiane Naci <sofiane.naci@arm.com>
17780
17781 * config/aarch64/aarch64.md: New movtf split.
17782 (*movtf_aarch64): Update.
17783 (aarch64_movdi_tilow): Handle TF modes and rename to
17784 aarch64_movdi_<mode>low.
17785 (aarch64_movdi_tihigh): Handle TF modes and rename to
17786 aarch64_movdi_<mode>high
17787 (aarch64_movtihigh_di): Handle TF modes and rename to
17788 aarch64_mov<mode>high_di
17789 (aarch64_movtilow_di): Handle TF modes and rename to
17790 aarch64_mov<mode>low_di
17791 (aarch64_movtilow_tilow): Remove spurious whitespace.
17792 * config/aarch64/aarch64.c (aarch64_split_128bit_move): Handle TFmode
17793 splits.
17794 (aarch64_print_operand): Update.
17795
227eb343
AM
177962013-05-09 Alan Modra <amodra@gmail.com>
17797
17798 * configure.ac (HAVE_AS_TLS): Enable tests for powerpcle and
17799 powerpc64le.
17800 * configure: Regenerate.
17801
0b013847
UB
178022013-05-08 Uros Bizjak <ubizjak@gmail.com>
17803
17804 * config/i386/mmx.md (*vec_extract* splitters): Simplify post-reload
17805 splitter preparation statements.
17806 * config/i386/sse.md (*vec_extract* splitters): Ditto.
17807 (*avx_vperm_broadcast_<mode>): Use adjust_address instead of
17808 adjust_address_nv.
17809
1dc3d6e9
BS
178102013-05-08 Bill Schmidt <wschmidt@linux.vnet.ibm.com>
17811
17812 * gimple-ssa-strength-reduction.c (count_candidates): Change
17813 return value to int.
17814 (analyze_candidates_and_replace): Change type of length to int.
17815
e61e7d28
UB
178162013-05-08 Uros Bizjak <ubizjak@gmail.com>
17817
17818 * config/i386/sse.md (PEXTR_MODE, PEXTR_MODEx): Remove.
17819 (*vec_extract<mode>): Use VI12_128 mode iterator.
17820 (*vec_extract<mode>_mem): Ditto.
17821 (*vec_extract*_mem splitters): Merge splitters using VI_128 mode
17822 attribute.
17823
4fbfcf44
DN
178242013-05-08 Diego Novillo <dnovillo@google.com>
17825
17826 PR bootstrap/54659
17827
17828 Revert:
4fbfcf44
DN
17829 2012-08-17 Diego Novillo <dnovillo@google.com>
17830
e61e7d28
UB
17831 PR bootstrap/54281
17832 * configure.ac: Add libintl.h to AC_CHECK_HEADERS list.
17833 * config.in: Regenerate.
17834 * configure: Regenerate.
17835 * intl.h: Always include libintl.h if HAVE_LIBINTL_H is set.
4fbfcf44 17836
702d8703
JH
178372013-05-08 Jan Hubicka <jh@suse.cz>
17838
17839 PR lto/54095
17840 * cgraph.c (cgraph_make_node_local_1): Se unique_name.
17841 * cgraph.h (symtab_node_base): Add unique_name.
17842 * lto-cgraph.c (lto_output_node, lto_output_varpool_node,
17843 input_overwrite_node, input_varpool_node): Stream unique_name.
17844 * cgraphclones.c (cgraph_create_virtual_clone,
17845 cgraph_function_versioning): Set unique_name.
17846 * ipa.c (function_and_variable_visibility): Set unique_name.
17847
8b28cf47
BS
178482013-05-08 Bill Schmidt <wschmidt@linux.vnet.ibm.com>
17849
17850 * gimple-ssa-strength-reduction.c (find_phi_def): Revert former "fix."
17851 (alloc_cand_and_find_basis): Restrict conditional candidate
17852 processing to CAND_MULTs.
17853
e86074fd
JH
178542013-05-08 Jan Hubicka <jh@suse.cz>
17855
17856 PR lto/54095
17857 lto-symtab.c (lto_symtab_symbol_p): New function.
17858 (lto_symtab_resolve_can_prevail_p, lto_symtab_resolve_symbols,
17859 lto_symtab_resolve_symbols, lto_symtab_merge_decls_2,
17860 lto_symtab_merge_decls_1, lto_symtab_merge_cgraph_nodes_1):
17861 Skip static symbols.
17862
44398cbe
PC
178632013-05-08 Paolo Carlini <paolo.carlini@oracle.com>
17864
17865 PR tree-optimization/57200
17866 * tree-ssa-loop-niter.c (do_warn_aggressive_loop_optimizations):
17867 Only call inform if the preceding warning_at returns true.
17868
f6bc1c4a
HS
178692013-05-07 Han Shen <shenhan@google.com>
17870
17871 * cfgexpand.c (record_or_union_type_has_array_p): New function.
17872 (expand_used_vars): Add logic handling '-fstack-protector-strong'.
17873 * common.opt (fstack-protector-strong): New option.
17874 * doc/cpp.texi (__SSP_STRONG__): New builtin "__SSP_STRONG__".
17875 * doc/invoke.texi (Optimization Options): Document
17876 "-fstack-protector-strong".
17877 * gcc.c (LINK_SSP_SPEC): Add 'fstack-protector-strong'.
17878
4ffecb1f
SB
178792013-05-06 Steven Bosscher <steven@gcc.gnu.org>
17880
17881 * config/mips/mips.c (mips_machine_reorg2): Return 0.
17882
5a107a0f
VM
178832013-05-07 Vladimir Makarov <vmakarov@redhat.com>
17884
17885 * ira.c (update_equiv_regs): Add insn having equiv memory even if
17886 it is not lhs of the insn.
17887 (setup_reg_equiv): Remove insn having equiv memory which it is not
17888 lhs of the insn.
17889 * lra-constraints.c (process_address): Try to improve generation
17890 code for address base + disp.
17891 (lra_constraints): Make correct the code for checking insn setting
17892 up backward equivalence. Remove insn only if it is in the init
17893 insn list.
17894 * lra-eliminations.c (update_reg_eliminate): Change return value.
17895 (lra_eliminate): Use the result.
17896
3f5783ea
UB
178972013-05-07 Uros Bizjak <ubizjak@gmail.com>
17898
17899 * config/i386/sse.md (ssescalarnummask): New mode attribute.
17900 (PEXTR_MODE, PEXTR_MODEx): New mode iterators.
17901 (*vec_extract<mode>): Merge from *sse4_1_pextrb_memory and
17902 *sse4_1_pextrw_memory using PEXTR_MODE mode iterator. Handle
17903 register target operands.
17904 (*vec_extractv8hi_sse2): New pattern.
17905 (*vec_extractv16qi_zext): Rename from *sse4_1_pextrb_<mode>.
17906 (*vec_extractv8hi_zext): Rename from *sse2_pextrw_<mode>.
17907 (*vec_extract<mode>_mem): New insn and split pattern.
17908
8a5800b8
CL
179092013-05-07 Christophe Lyon <christophe.lyon@linaro.org>
17910
17911 * config/arm/arm.c (arm_asan_shadow_offset): New function.
17912 (TARGET_ASAN_SHADOW_OFFSET): Define.
17913 * config/arm/linux-eabi.h (ASAN_CC1_SPEC): Define.
17914 (LINUX_OR_ANDROID_CC): Add ASAN_CC1_SPEC.
17915
7bf55a70
BS
179162013-05-07 Bill Schmidt <wschmidt@linux.vnet.ibm.com>
17917
17918 * gimple-ssa-strength-reduction.c (MAX_INCR_VEC_LEN): New constant.
17919 (incr_vec_index): Return -1 if increment not found.
17920 (create_add_on_incoming_edge): Assert if increment not found.
17921 (record_increment): Limit number of increments recorded.
17922 (all_phi_incrs_profitable): Return false if an increment not found.
17923 (replace_profitable_candidates): Don't process increments that were
17924 not recorded.
17925 (analyze_candidates_and_replace): Limit size of incr_vec.
17926
3f8825c0
RB
179272013-05-07 Richard Biener <rguenther@suse.de>
17928
17929 * calls.c (special_function_p): setjmp-like functions are leaf.
17930 * builtins.def (BUILT_IN_SETJMP): setjmp is leaf.
17931 * tree-inline.c (update_ssa_across_abnormal_edges): Remove assert.
17932
fd4842cd
SN
179332013-05-07 Sofiane Naci <sofiane.naci@arm.com>
17934
17935 * config/aarch64/aarch64-simd.md (*aarch64_simd_mov<mode>): call splitter.
17936 (aarch64_simd_mov<mode>): New expander.
17937 (aarch64_simd_mov_to_<mode>low): New instruction pattern.
17938 (aarch64_simd_mov_to_<mode>high): Likewise.
17939 (aarch64_simd_mov_from_<mode>low): Likewise.
17940 (aarch64_simd_mov_from_<mode>high): Likewise.
17941 (aarch64_dup_lane<mode>): Update.
17942 (aarch64_dup_lanedi): New instruction pattern.
17943 * config/aarch64/aarch64-protos.h (aarch64_split_simd_move): New prototype.
17944 * config/aarch64/aarch64.c (aarch64_split_simd_move): New function.
17945
a7a7d10e
BS
179462013-05-07 Bill Schmidt <wschmidt@linux.vnet.ibm.com>
17947
17948 * gimple-ssa-strength-reduction.c (lazy_create_slsr_reg): Remove.
17949 (replace_mult_candidate): Remove unnecessary argument; remove
17950 unnecessary parameter from call to introduce_cast_before_cand.
17951 (replace_unconditional_candidate): Remove unnecessary parameter
17952 from call to replace_mult_candidate.
17953 (replace_conditional_candidate): Likewise.
17954 (insert_initializers): Use make_temp_ssa_name.
17955 (introduce_cast_before_cand): Remove unnecessary argument; use
17956 make_temp_ssa_name.
17957 (replace_one_candidate): Remove unnecessary argument; remove
17958 unnecessary parameter from calls to introduce_cast_before_cand.
17959 (replace_profitable_candidates): Remove unnecessary parameters
17960 from calls to replace_one_candidate.
17961
29105868
BS
179622013-05-07 Bill Schmidt <wschmidt@linux.vnet.ibm.com>
17963
17964 * gimple-ssa-strength-reduction.c (find_phi_def): Don't record a
17965 phi def as possibly hiding a basis for a CAND_ADD whose operands
17966 have been commuted in the analysis.
17967 (alloc_cand_and_find_basis): Add parms to call to find_phi_def.
17968
4095f9fa
N
179692013-05-07 Naveen H.S <Naveen.Hurugalawadi@caviumnetworks.com>
17970
17971 * config/aarch64/aarch64.md
17972 (cmp_swp_<optab><ALLX:mode>_shft_<GPI:mode>): Restrict the
17973 shift value between 0-4.
17974
39e843e8
RB
179752013-05-07 Richard Biener <rguenther@suse.de>
17976
17977 * double-int.h (rshift): New overload.
17978 * double-int.c (rshift): New function.
17979 * tree-ssa-sccvn.c (copy_reference_ops_from_ref): Optimize.
17980 (create_reference_ops_from_ref): Remove.
17981 (vn_reference_insert): Use shared ops for constructing the
17982 reference and copy it.
17983
0a1a83cb
RB
179842013-05-07 Richard Biener <rguenther@suse.de>
17985
17986 PR middle-end/57190
17987 * tree-eh.c (sink_clobbers): Properly propagate
17988 SSA_NAME_OCCURS_IN_ABNORMAL_PHI.
17989
43303d6f
JJ
179902013-05-07 Jakub Jelinek <jakub@redhat.com>
17991
ba7e83f8
JJ
17992 PR tree-optimization/57149
17993 * tree-ssa-uninit.c (uninit_undefined_value_p): New inline.
17994 (can_skip_redundant_opnd, compute_uninit_opnds_pos,
17995 collect_phi_def_edges, execute_late_warn_uninitialized): Use
17996 uninit_undefined_value_p instead of ssa_undefined_value_p.
17997
43303d6f
JJ
17998 PR debug/57184
17999 * expr.c (expand_expr_addr_expr_1): Handle COMPOUND_LITERAL_EXPR
18000 for modifier == EXPAND_INITIALIZER.
18001
14523c25
AB
180022013-05-07 Anton Blanchard <anton@samba.org>
18003
18004 * configure.ac (HAVE_LD_LARGE_TOC): Use correct linker emulation
18005 for powerpc64 little endian.
18006 * configure: Regenerate.
18007
cb7c8be9
GS
180082013-05-06 Graham Stott <grahams@btinternet.com>
18009
18010 * expmed.c (init_expmed_rtl): Remove unused fields reg_fld, plus_fld,
18011 mult_fld, sdiv_fld1, udiv_fld1, sdiv_32_fld1, smod_32_fld1,
18012 wide_mult_fld1, wide_lshr_fld1, shift_fld1, shift_mult_fld1,
18013 shift_add_fld1, shift_sub0_fld1, shift_sub1_fld1.
18014
180152013-05-06 Graham Stott <grahams@btinternet.com>
a8fd4225
GS
18016
18017 * gensupport.c (add_predicate_code): Also exclude SCRATCH from rtx
18018 codes which allow non-lvalues.
18019
9a0ee7b0
MG
180202013-05-06 Marc Glisse <marc.glisse@inria.fr>
18021
18022 * tree.c (integer_all_onesp) <COMPLEX_CST>: Test that both
18023 components are all 1s.
18024 (integer_minus_onep): New function.
18025 * tree.h (integer_minus_onep): Declare it.
18026 * fold-const.c (fold_binary_loc) <MULT_EXPR>: Test
18027 integer_minus_onep instead of integer_all_onesp.
18028
f2c17ea9
OE
180292013-05-06 Oleg Endo <olegendo@gcc.gnu.org>
18030
18031 PR target/52933
18032 * config/sh/sh.md (*cmp_div0s_0, *cmp_div0s_1, *movsicc_div0s): Add
18033 variations of these patterns.
18034
f75e6a51
UB
180352013-05-06 Uros Bizjak <ubizjak@gmail.com>
18036
18037 * config/i386/i386.md (isa): Add x64_sse4 member.
18038 (enabled): Handle x64_sse4.
18039 (*movdi_internal): Add *x->?r alternative to emit pextrq $0,%xmm,%reg
18040 instruction for 64bit SSE4_1 targets. Update insn attributes.
18041 (*movsi_internal): Add *x->?r alternative to emit pextrd $0,%xmm,%reg
18042 instruction for SSE4_1 targets. Update insn attributes.
18043 * config/i386/sse.md (*vec_extract<ssevecmodelower>_0): Merge
18044 with *sse4_1_pextrd and *sse4_1_pextrq having const_0 selector.
18045 (*vec_extractv2di_1): Merge with *sse4_1_pextrq having
18046 const_1 selector.
18047 (*vec_extractv4si): Rename from *sse4_1_pextrd.
18048 (*vec_extractv4si_zext): Rename from *sse4_1_pextrd_zext.
18049 (*vec_extract<ssevecmodelower>_0 splitters): Merge splitters together.
18050
a986d468
OE
180512013-05-06 Oleg Endo <olegendo@gcc.gnu.org>
18052
18053 PR target/57108
18054 * config/sh/sh.md (tstsi_t_zero_extract_eq): Use QIHISIDI mode iterator.
18055
382522cb
MK
180562013-05-06 Maxim Kuznetsov <maks.kuznetsov@gmail.com>
18057
18058 * final.c (do_assembler_dialects): Don't handle curly braces and
18059 vertical bar escaped by % as dialect delimiters.
18060 (output_asm_insn): Print curly braces and vertical bar if escaped
18061 by % and ASSEMBLER_DIALECT defined.
18062 * doc/tm.texi.in (ASSEMBLER_DIALECT): Document new standard escapes.
18063 * doc/tm.texi: Regenerated.
18064
fb0d5c60
SB
180652013-05-06 Steven Bosscher <steven@gcc.gnu.org>
18066
fb0d5c60
SB
18067 * config/mips/mips.c: Include tree-pass.h.
18068 (mips_reorg): Split in pre- and post-dbr_schedule parts.
18069 (mips_machine_reorg2): Move mips_reorg post-dbr_schedule parts here.
18070 (pass_mips_machine_reorg2): New machine specific pass.
18071 (insert_pass_mips_machine_reorg2): New pass plugin definition.
18072 (mips_option_override): Register the new pass.
18073 * rtl.h (cleanup_barriers): Remove prototype.
18074 (dbr_schedule): Likewise.
18075 * jump.c (cleanup_barriers): Make static.
18076 * reorg.c (dbr_schedule): Likewise.
18077
aa06a978
RB
180782013-05-06 Richard Biener <rguenther@suse.de>
18079
18080 PR tree-optimization/57185
18081 * tree-parloops.c (add_field_for_reduction): Handle anonymous
18082 SSA names properly.
18083
0b953bec
UB
180842013-05-06 Uros Bizjak <ubizjak@gmail.com>
18085
18086 PR target/57106
18087 * config/i386/i386.c (add_parameter_dependencies): Add dependence
18088 between "first_arg" and "insn", not "last" and "insn".
18089
28708525
WS
180902013-05-06 Bill Schmidt <wschmidt@linux.vnet.ibm.com>
18091
18092 * gimple-ssa-strength-reduction.c (slsr_process_phi): Re-enable.
18093 (find_candidates_in_block): Re-enable slsr_process_phi.
18094 (create_phi_basis): Fix double counting of candidate adjustment.
18095
0107dca2
RB
180962013-05-06 Richard Biener <rguenther@suse.de>
18097
18098 PR middle-end/57147
18099 * tree-cfg.c (gimple_purge_dead_abnormal_call_edges): If
18100 the edge is also fallthru, preserve it and just clear the
18101 abnormal flag.
18102 * tree-cfgcleanup.c (remove_fallthru_edge): If the edge is
18103 also complex, preserve that and just clear the fallthru flag.
18104 * tree-inline.c (update_ssa_across_abnormal_edges): Also
18105 update virtual operands.
18106
470d4d13
AM
181072013-05-06 Alan Modra <amodra@gmail.com>
18108
18109 * config/rs6000/linux.h (DEFAULT_ASM_ENDIAN): Define.
18110 (LINK_OS_LINUX_EMUL): Use ENDIAN_SELECT.
18111 * config/rs6000/linux64.h (DEFAULT_ASM_ENDIAN): Define.
18112 * config/rs6000/sysv4le.h (DEFAULT_ASM_ENDIAN): Define.
18113 (LINK_TARGET_SPEC): Use ENDIAN_SELECT.
18114 * config/rs6000/sysv4.h (DEFAULT_ASM_ENDIAN): Define as -mbig.
18115
8f1dbf8d
AM
181162013-05-06 Alan Modra <amodra@gmail.com>
18117
18118 * config/rs6000/sysv4.h (ENDIAN_SELECT): Define, extracted from
18119 (ASM_SPEC): ..here. Emit DEFAULT_ASM_ENDIAN too.
18120 (DEFAULT_ASM_ENDIAN): Define.
18121 (CC1_SPEC, LINK_TARGET_SPEC): Use ENDIAN_SELECT.
18122 * config/rs6000/linux64.h (ASM_SPEC32): Remove endian options.
18123 Update -K PIC clause from sysv4.h.
18124 (ASM_SPEC_COMMON): Use ENDIAN_SELECT.
18125 (LINK_OS_LINUX_EMUL32, LINK_OS_LINUX_EMUL64): Likewise.
18126
54890767
AM
181272013-05-06 Alan Modra <amodra@gmail.com>
18128
18129 * config/rs6000/rs6000.md (bswapdi 2nd splitter): Don't swap words
18130 twice for little-endian.
18131 (ashrdi3_no_power, ashrdi3): Support little-endian.
18132
2353515d
OE
181332013-05-06 Oleg Endo <olegendo@gcc.gnu.org>
18134
18135 PR target/55303
18136 * config/sh/sh.c (sh_rtx_costs): Handle SMIN and SMAX cases.
18137 * config/sh/sh.md (*clips, uminsi3, *clipu, clipu_one): New insns and
18138 related expanders.
18139 * config/sh/iterators.md (SMIN_SMAX): New code iterator.
18140 * config/sh/predicates.md (arith_reg_or_0_or_1_operand,
18141 clips_min_const_int, clips_max_const_int, clipu_max_const_int):
18142 New predicates.
18143
33e67557
SB
181442013-05-05 Steven Bosscher <steven@gcc.gnu.org>
18145 John David Anglin <dave.anglin@nrc-cnrc.gc.ca>
18146
18147 * config.gcc (hppa*-*-*): Remove MASK_BIG_SWITCH from CPU default.
18148 * config/pa/pa.opt: Make mbig-switch a no-op.
18149 * config/pa/pa.h (TARGET_DEFAULT): Remove MASK_BIG_SWITCH.
18150 (CASE_VECTOR_MODE): Always return SImode.
18151 (ASM_OUTPUT_ADDR_VEC_ELT, ASM_OUTPUT_ADDR_DIFF_ELT): Remove code
18152 for the !TARGET_BIG_SWITCH case.
18153 * config/pa/pa-linux.h: Likewise.
18154 * config/pa/pa-openbsd.h: Likewise.
18155 * config/pa/pa-hpux.h: Define TARGET_DEFAULT to 0.
18156 * config/pa/pa.md (short_jump): Remove define_insn.
18157 (casesi): Remove code for the !TARGET_BIG_SWITCH case.
18158 (casesi0): Remove define_insn.
18159 (type): Remove btable_branch.
18160 (pa_combine_type): Likewise.
18161 (in_nullified_branch_delay): Likewise.
18162 (in_call_delay): Likewise.
18163 (define_delay): Likewise.
18164 (define_insn_reservation "Z3"): Likewise.
18165 (define_insn_reservation "Z4"): Likewise.
18166 * config/pa/pa.c (pa_reorg): Remove code for !TARGET_BIG_SWITCH.
18167 (pa_adjust_insn_length): Remove adjustment for btable branches.
18168 * doc/invoke.texi (HPPA Options): Delete documentation for mbig-switch
18169 and mno-big-switch
18170
3095685e
UB
181712013-05-05 Uros Bizjak <ubizjak@gmail.com>
18172
18173 * config/i386/sse.md (*vec_extract<ssevecmodelower>_0): Merge
18174 from sse2_stored and *sse2_storeq_rex64 using SWI48 mode iterator.
18175 Add m->r,x alternatives.
18176 (*vec_extract<ssevecmodelower>_0 splitters): Merge V2DI and V4SI
18177 splitters using SWI48x mode iterator.
18178 (*vec_extract_v2di_0_sse): Rename from *sse2_storeq. Disable for
18179 TARGET_64BIT. Add m->x alternative.
18180 (*vec_extractv4si_mem): Rename from *vec_ext_v4si_mem.
18181 Add o->x alternative. Enable for TARGET_SSE.
18182 (sse_storeq): Remove expander.
18183 (*vec_extractv2di_1): Enable for TARGET_SSE. Split alternatives
18184 with memory input operand.
18185 (*vec_extractv2di_1 splitter): New.
18186 (*vec_extractv4sf_mem): Rename from *vec_extract_v4sf_mem.
18187 * config/i386/i386.md (ssevecmodelower): New mode attribute.
18188
4b36ae28
SB
181892013-05-04 Segher Boessenkool <segher@kernel.crashing.org>
18190
18191 * config/rs6000/rs6000.c (INT_P): Reformat. Delete obsolete comment.
18192 (INT_LOWPART): Delete.
18193 (extract_MB): Adjust.
18194 (extract_ME): Adjust.
18195 (print_operand): Adjust.
18196
da226db2
SB
181972013-05-04 Segher Boessenkool <segher@kernel.crashing.org>
18198
18199 * config/rs6000/predicates.md (reg_or_add_cint_operand,
18200 reg_or_sub_cint_operand): Delete "HOST_BITS_PER_WIDE_INT == 32" case.
18201 (reg_or_logical_cint_operand, easy_fp_constant,
18202 logical_const_operand): Delete "CONST_DOUBLE" case.
18203 * config/rs6000/rs6000.c (num_insns_constant_wide): Delete
18204 "HOST_BITS_PER_WIDE_INT == 64" test.
3095685e 18205 (num_insns_constant): Ditto. Delete CONST_DOUBLE DImode/VOIDmode case.
da226db2
SB
18206 (build_mask64_2_operands): Delete "HOST_BITS_PER_WIDE_INT >= 64" test.
18207 (rs6000_emit_set_const): Delete CONST_DOUBLE case.
18208 (rs6000_emit_set_long_const): Delete "HOST_BITS_PER_WIDE_INT >= 64"
18209 test.
18210 (includes_rldic_lshift_p, includes_rldicr_lshift_p): Delete
18211 CONST_DOUBLE DImode/VOIDmode case.
18212 (INT_P, INT_LOWPART): Delete CONST_DOUBLE case.
18213 (print_operand): Delete "HOST_BITS_PER_WIDE_INT == 32" case. Delete
18214 CONST_DOUBLE VOIDmode case.
18215 (output_toc): Delete "HOST_BITS_PER_WIDE_INT == 32" case.
18216 (rs6000_rtx_costs): Delete CONST_DOUBLE DImode/VOIDmode case.
18217 * config/rs6000/rs6000.md (iordi3, xordi3, splitter for these):
18218 Delete CONST_DOUBLE case.
18219 (splitters for mov FMOVE64 const_double): Delete
18220 "HOST_BITS_PER_WIDE_INT == 32" case. Delete
18221 "HOST_BITS_PER_WIDE_INT >= 64" test.
18222 (splitter for mov DI const_int): Delete "HOST_BITS_PER_WIDE_INT == 32"
18223 case.
18224 (mov DI const_double): Delete.
18225
40de22d6
JJ
182262013-05-04 Jakub Jelinek <jakub@redhat.com>
18227
18228 * combine.c (combine_simplify_rtx) <case SUBREG>: If nonzero_bits
18229 on op shows all bits zero in mode of a lowpart subreg, return zero.
18230
5ec6aff2
MM
182312013-05-03 Michael Meissner <meissner@linux.vnet.ibm.com>
18232
18233 PR target/57150
18234 * config/rs6000/rs6000.h (HARD_REGNO_CALLER_SAVE_MODE): Use DFmode
18235 to save TFmode registers and DImode to save TImode registers for
18236 caller save operations.
18237 (HARD_REGNO_CALL_PART_CLOBBERED): TFmode and TDmode do not need to
18238 mark being partially clobbered since they only use the first
18239 double word.
18240
18241 * config/rs6000/rs6000.c (rs6000_init_hard_regno_mode_ok): TFmode
18242 and TDmode only use the upper 64-bits of each VSX register.
18243
2cefad90
BS
182442013-05-03 Bill Schmidt <wschmidt@linux.vnet.ibm.com>
18245
18246 * gimple-ssa-strength-reduction.c (slsr_process_phi): Disable.
18247 (find_candidates_in_block): Disable slsr_process_phi.
18248
d6d7eee1
GW
182492013-05-03 Guozhi Wei <carrot@google.com>
18250
18251 * coverage.c (coverage_obj_init): Move the construction of gcov
3095685e
UB
18252 constructor to ...
18253 (build_init_ctor): ... here.
d6d7eee1 18254
9b92d12b
BS
182552013-05-03 Bill Schmidt <wschmidt@linux.vnet.ibm.com>
18256
18257 * gimple-ssa-strength-reduction.c (cand_kind): Add CAND_PHI.
18258 (slsr_cand_d): Redefine def_phi.
18259 (stride_status, phi_adjust_status, count_phis_status): New enums.
18260 (find_phi_def): New.
18261 (find_basis_for_base_expr): New.
18262 (find_basis_for_candidate): Handle hidden bases.
18263 (alloc_cand_and_find_basis): Handle phi candidates.
18264 (slsr_process_phi): New.
18265 (create_mul_ssa_cand): Exclude phi base candidates; use integer_onep.
18266 (create_mul_imm_cand): Likewise.
18267 (create_add_ssa_cand): Exclude phi base candidates.
18268 (create_add_imm_cand): Likewise.
18269 (slsr_process_cast): Likewise.
18270 (slsr_process_copy): Likewise.
18271 (find_candidates_in_block): Handle phi candidates.
18272 (dump_candidate): Likewise.
18273 (unconditional_cands): Delete.
18274 (unconditional_cands_with_known_stride_p): Delete.
18275 (phi_dependent_cand_p): New.
18276 (cand_increment): Handle phi-dependent candidates.
18277 (replace_dependent): Delete.
18278 (replace_mult_candidate): New.
18279 (replace_unconditional_candidate): New.
18280 (incr_vec_index): Move to avoid forward reference.
18281 (create_add_on_incoming_edge): New.
18282 (create_phi_basis): New.
18283 (replace_dependents): Delete.
18284 (replace_conditional_candidate): New.
18285 (phi_add_costs): New.
18286 (replace_uncond_cands_and_profitable_phis): New.
18287 (record_increment): Handle phi adjustments.
18288 (record_phi_increments): New.
18289 (record_increments): Handle phi adjustments.
18290 (phi_incr_cost): New.
18291 (lowest_cost_path): Handle phis.
18292 (total_savings): Likewise.
18293 (analyze_increments): Likewise.
18294 (ncd_with_phi): New.
18295 (ncd_of_cand_and_phis): New.
18296 (nearest_common_dominator_for_cands): Handle phi increments.
18297 (all_phi_incrs_profitable): New.
18298 (replace_profitable_candidates): Handle phi-dependent candidates.
18299 (analyze_candidates_and_replace): Likewise.
18300
68f073d4
TJ
183012013-05-03 Teresa Johnson <tejohnson@google.com>
18302
18303 PR bootstrap/57154
18304 * sched-rgn.c (compute_dom_prob_ps): Ensure accumulated probabilities
18305 do not exceed REG_BR_PROB_BASE.
18306
a4ee7cb9
JL
183072013-05-03 Jeff Law <law@redhat.com>
18308
ade67f70 18309 PR tree-optimization/57144
a4ee7cb9
JL
18310 * tree-vrp.c (simplify_cond_using_ranges): Verify the constant
18311 operand of the condition will bit into the new type when eliminating
18312 a cast feeding a condition.
18313
47954c4e
JJ
183142013-05-03 Jakub Jelinek <jakub@redhat.com>
18315
18316 PR rtl-optimization/57130
3095685e
UB
18317 * combine.c (make_compound_operation) <case SUBREG>: Pass SET instead
18318 of COMPARE as in_code to the recursive call if needed.
47954c4e 18319
3c21604f
UB
183202013-05-03 Uros Bizjak <ubizjak@gmail.com>
18321
18322 * config/i386/i386.md (isa): Add x64_sse4_noavx and x64_avx members.
18323 (enabled): Handle new members.
18324 * config/i386/sse.md (*vec_concatv2si): Merge from
18325 *vec_concatv2si_sse2 and vec_concatv2si_sse.
18326 (vec_concatv2di): Merge with *vec_concatv2di_rex64.
18327
12211b99 183282013-05-03 Joern Rennecke <joern.rennecke@embecosm.com>
95c03b36
JR
18329
18330 PR tree-optimization/57027
18331 * tree-ssa-math-opts.c (convert_mult_to_fma): When checking
18332 for fnms opportunity, check we got the prerequisite kind
18333 of tree / gimple before using accessor functions.
18334
07bfc9ec
RB
183352013-05-03 Richard Biener <rguenther@suse.de>
18336
18337 * double-int.h (lshift): New overload without precision
18338 and arith argument.
18339 (operator *=, operator +=, operator -=): Move ...
18340 * double-int.c (operator *=, operator +=, operator -=): ... here
18341 and implement more efficiently.
18342 (mul_double_with_sign): Remove.
18343 (lshift_double): Adjust to take unsinged shift argument, push
18344 dispatching code to callers.
18345 (mul_double_wide_with_sign): Add early out for callers that
18346 are not interested in high parts or overflow.
18347 (lshift): New function.
18348 (lshift, rshift, alshift, arshift, llshift, lrshift): Add
18349 dispatch code here.
18350 (lrotate, rrotate): Use logical shifts.
18351 * expr.c (get_inner_reference): Use lshift.
18352 * fixed-value.c (do_fixed_divide): Likewise.
18353 * tree-dfa.c (get_ref_base_and_extent): Likewise.
18354 * tree-ssa-alias.c (indirect_ref_may_alias_decl_p): Likewise.
18355 (indirect_refs_may_alias_p): Likewise.
18356 (stmt_kills_ref_p_1): Likewise.
18357
7769bb64
VP
183582013-05-03 Vidya Praveen <vidyapraveen@arm.com>
18359
18360 * config/aarch64/aarch64-simd.md (simd_fabd): Correct the description.
18361
1d0c8e5c
VP
183622013-05-03 Vidya Praveen <vidyapraveen@arm.com>
18363
18364 * config/aarch64/aarch64-simd.md (*fabd_scalar<mode>3): Support
18365 scalar form of FABD instruction.
18366
f15feaf9
VM
183672013-05-02 Vladimir Makarov <vmakarov@redhat.com>
18368
18369 * lra-constraints.c (process_alt_operands): Add checking alt
18370 number to choose the best alternative.
18371
d90e76d4
RB
183722013-05-02 Richard Biener <rguenther@suse.de>
18373
18374 * tree-eh.c (cleanup_empty_eh_merge_phis): Remove rename_virts
18375 bitmap and its handling.
18376 (pass_cleanup_eh): Set todo_flags_finish to TODO_verify_ssa.
18377
9f8e7a96
RB
183782013-05-02 Richard Biener <rguenther@suse.de>
18379
18380 PR middle-end/57140
18381 * tree-inline.c (copy_loops): Properly handle removed loops.
18382 (copy_cfg_body): Mark destination loops for fixup if source
18383 loops needed fixup.
18384
f3a81b39
GY
183852013-05-02 Greta Yorsh <Greta.Yorsh@arm.com>
18386
18387 PR target/56732
18388 * config/arm/arm.c (arm_expand_epilogue): Check really_return before
18389 generating simple_return for naked functions.
18390
7b920a9a
MJ
183912013-05-02 Martin Jambor <mjambor@suse.cz>
18392
18393 PR middle-end/56988
18394 * ipa-prop.h (ipa_agg_replacement_value): New flag by_ref.
18395 * ipa-cp.c (ipa_get_indirect_edge_target_1): Also check that by_ref
18396 flags match.
18397 (find_aggregate_values_for_callers_subset): Fill in the by_ref flag of
18398 ipa_agg_replacement_value structures.
18399 (known_aggs_to_agg_replacement_list): Likewise.
18400 * ipa-prop.c (write_agg_replacement_chain): Stream by_ref flag.
18401 (read_agg_replacement_chain): Likewise.
18402 (ipcp_transform_function): Also check that by_ref flags match.
18403
2c41c19d
RB
184042013-05-02 Richard Biener <rguenther@suse.de>
18405
18406 * graphds.h (struct graph): Add obstack member.
18407 * graphds.c (new_graph): Initialize obstack and allocate
18408 vertices from it.
18409 (add_edge): Allocate edge from the obstack.
3c21604f 18410 (free_graph): Free the obstack instead of all edges and vertices.
2c41c19d 18411
8b47039c
TJ
184122013-05-02 Teresa Johnson <tejohnson@google.com>
18413
18414 * loop-unswitch.c (unswitch_loop): Use helper routines with rounding
18415 divides.
18416 * cfg.c (update_bb_profile_for_threading): Ditto.
18417 * tree-inline.c (copy_bb): Ditto.
18418 (copy_edges_for_bb): Ditto.
18419 (initialize_cfun): Ditto.
18420 (copy_cfg_body): Ditto.
18421 (expand_call_inline): Ditto.
18422 * ipa-inline-analysis.c (estimate_edge_size_and_time): Ditto.
18423 (estimate_node_size_and_time): Ditto.
18424 (inline_merge_summary): Ditto.
18425 * cgraphclones.c (cgraph_clone_edge): Ditto.
18426 (cgraph_clone_node): Ditto.
18427 * sched-rgn.c (compute_dom_prob_ps): Ditto.
18428 (compute_trg_info): Ditto.
18429
da65928c
IB
184302013-05-02 Ian Bolton <ian.bolton@arm.com>
18431
18432 * config/aarch64/aarch64.md (movsi_aarch64): Only allow to/from
18433 S reg when fp attribute set.
18434 (movdi_aarch64): Only allow to/from D reg when fp attribute set.
18435
add3c965
IB
184362013-05-02 Ian Bolton <ian.bolton@arm.com>
18437
18438 * config/aarch64/aarch64.md (*and_one_cmpl<mode>3_compare0):
18439 New pattern.
18440 (*and_one_cmplsi3_compare0_uxtw): Likewise.
18441 (*and_one_cmpl_<SHIFT:optab><mode>3_compare0): Likewise.
18442 (*and_one_cmpl_<SHIFT:optab>si3_compare0_uxtw): Likewise.
18443
fdd43ac4
RB
184442013-05-02 Richard Biener <rguenther@suse.de>
18445
18446 * tree-scalar-evolution.c (scev_info_hasher): Remove.
18447 (struct instantiate_cache_entry): New type.
18448 (struct instantiate_cache_entry_hasher): New hashtable descriptor.
18449 (struct instantiate_cache_type): New type.
18450 (set_instantiated_value, get_instantiated_value): Remove.
18451 (get_instantiated_value_entry): New function.
18452 (instantiate_scev_name): Use the new cache and adjust.
18453 (instantiate_scev_poly): Adjust.
18454 (instantiate_scev_binary): Likewise.
18455 (instantiate_array_ref): Likewise.
18456 (instantiate_scev_convert): Likewise.
18457 (instantiate_scev_not): Likewise.
18458 (instantiate_scev_3): Likewise.
18459 (instantiate_scev_2): Likewise.
18460 (instantiate_scev_r): Likewise.
18461 (instantiate_scev): Likewise.
18462 (resolve_mixers): Likewise.
18463
36ff9dfb
VM
184642013-05-01 Vladimir Makarov <vmakarov@redhat.com>
18465
18466 PR target/57091
18467 * lra-constraints.c (best_small_class_operands_num): Remove.
18468 (process_alt_operands): Remove small_class_operands_num. Take
18469 small classes operands into losers and only if the operand is not
18470 matched. Modify debugging output.
18471 (curr_insn_transform): Remove best_small_class_operands_num.
18472 Print insn name.
18473
36054fab
JG
184742013-05-01 James Greenhalgh <james.greenhalgh@arm.com>
18475
18476 * config/aarch64/aarch64-builtins.c
18477 (aarch64_gimple_fold_builtin.c): Fold more modes for reduc_splus_.
18478 * config/aarch64/aarch64-simd-builtins.def
18479 (reduc_splus_): Add new modes.
18480 (reduc_uplus_): New.
18481 * config/aarch64/aarch64-simd.md (aarch64_addvv4sf): Remove.
18482 (reduc_uplus_v4sf): Likewise.
18483 (reduc_splus_v4sf): Likewise.
18484 (aarch64_addv<mode>): Likewise.
18485 (reduc_uplus_<mode>): Likewise.
18486 (reduc_splus_<mode>): Likewise.
18487 (aarch64_addvv2di): Likewise.
18488 (reduc_uplus_v2di): Likewise.
18489 (reduc_splus_v2di): Likewise.
18490 (aarch64_addvv2si): Likewise.
18491 (reduc_uplus_v2si): Likewise.
18492 (reduc_splus_v2si): Likewise.
18493 (reduc_<sur>plus_<mode>): New.
18494 (reduc_<sur>plus_v2di): Likewise.
18495 (reduc_<sur>plus_v2si): Likewise.
18496 (reduc_<sur>plus_v4sf): Likewise.
18497 (aarch64_addpv4sf): Likewise.
18498 * config/aarch64/arm_neon.h
18499 (vaddv<q>_<s,u,f><8, 16, 32, 64): Rewrite using builtins.
18500 * config/aarch64/iterators.md (unspec): Remove UNSPEC_ADDV,
18501 add UNSPEC_SADDV, UNSPEC_UADDV.
18502 (SUADDV): New.
18503 (sur): Add UNSPEC_SADDV, UNSPEC_UADDV.
18504
6dce23a8
JG
185052013-05-01 James Greenhalgh <james.greenhalgh@arm.com>
18506
18507 * config/aarch64/arm_neon.h
18508 (v<max,min><nm><q><v>_<sfu><8, 16, 32, 64>): Rewrite using builtins.
18509
1598945b
JG
185102013-05-01 James Greenhalgh <james.greenhalgh@arm.com>
18511
18512 * config/aarch64/aarch64-builtins
18513 (aarch64_gimple_fold_builtin): Fold reduc_<su><maxmin>_ builtins.
18514
998eaf97
JG
185152013-05-01 James Greenhalgh <james.greenhalgh@arm.com>
18516
18517 * config/aarch64/aarch64-simd-builtins.def
18518 (reduc_smax_): New.
18519 (reduc_smin_): Likewise.
18520 (reduc_umax_): Likewise.
18521 (reduc_umin_): Likewise.
18522 (reduc_smax_nan_): Likewise.
18523 (reduc_smin_nan_): Likewise.
18524 (fmax): Remove.
18525 (fmin): Likewise.
18526 (smax): Update for V2SF, V4SF and V2DF modes.
18527 (smin): Likewise.
18528 (smax_nan): New.
18529 (smin_nan): Likewise.
18530 * config/aarch64/aarch64-simd.md (<maxmin><mode>3): Rename to...
18531 (<su><maxmin><mode>3): ...This, refactor.
18532 (s<maxmin><mode>3): New.
18533 (<maxmin_uns><mode>3): Likewise.
18534 (reduc_<maxmin_uns>_<mode>): Refactor.
18535 (reduc_<maxmin_uns>_v4sf): Likewise.
18536 (reduc_<maxmin_uns>_v2si): Likewise.
18537 (aarch64_<fmaxmin><mode>: Remove.
18538 * config/aarch64/arm_neon.h (vmax<q>_f<32,64>): Rewrite to use
18539 new builtin names.
18540 (vmin<q>_f<32,64>): Likewise.
18541 * config/iterators.md (unspec): Add UNSPEC_FMAXNMV, UNSPEC_FMINNMV.
18542 (FMAXMIN): New.
18543 (su): Add mappings for smax, smin, umax, umin.
18544 (maxmin): New.
18545 (FMAXMINV): Add UNSPEC_FMAXNMV, UNSPEC_FMINNMV.
18546 (FMAXMIN): Rename as...
18547 (FMAXMIN_UNS): ...This.
18548 (maxminv): Remove.
18549 (fmaxminv): Likewise.
18550 (fmaxmin): Likewise.
18551 (maxmin_uns): New.
18552 (maxmin_uns_op): Likewise.
18553
bd11644e
JG
185542013-05-01 James Greenhalgh <james.greenhalgh@arm.com>
18555
18556 * config/aarch64/arm_neon.h
18557 (vac<ge, gt><sd>_f<32, 64>): Rename to...
18558 (vca<ge, gt><sd>_f<32, 64>): ...this, reimpliment in C.
18559 (vca<ge, gt, lt, le><q>_f<32, 64>): Reimpliment in C.
18560
75dd5ace
JG
185612013-05-01 James Greenhalgh <james.greenhalgh@arm.com>
18562
18563 * config/aarch64/aarch64-simd.md (*aarch64_fac<optab><mode>): New.
18564 * config/aarch64/iterators.md (FAC_COMPARISONS): New.
18565
7c19979f
JG
185662013-05-01 James Greenhalgh <james.greenhalgh@arm.com>
18567
18568 * config/aarch64/aarch64-simd.md
18569 (vcond<mode>_internal): Handle special cases for constant masks.
18570 (vcond<mode><mode>): Allow nonmemory_operands for outcome vectors.
18571 (vcondu<mode><mode>): Likewise.
18572 (vcond<v_cmp_result><mode>): New.
18573
bb60efd9
JG
185742013-05-01 James Greenhalgh <james.greenhalgh@arm.com>
18575
18576 * config/aarch64/aarch64-builtins.c (BUILTIN_VALLDI): Define.
18577 (aarch64_fold_builtin): Add folding for cm<eq,ge,gt,tst>.
18578 * config/aarch64/aarch64-simd-builtins.def
18579 (cmeq): Update to BUILTIN_VALLDI.
18580 (cmgt): Likewise.
18581 (cmge): Likewise.
18582 (cmle): Likewise.
18583 (cmlt): Likewise.
18584 * config/aarch64/arm_neon.h
18585 (vc<eq, lt, le, gt, ge, tst><z><qsd>_<fpsu><8,16,32,64>): Remap
18586 to builtins or C as appropriate.
18587
889b9412
JG
185882013-05-01 James Greenhalgh <james.greenhalgh@arm.com>
18589
18590 * config/aarch64/aarch64-simd-builtins.def (cmhs): Rename to...
18591 (cmgeu): ...This.
18592 (cmhi): Rename to...
18593 (cmgtu): ...This.
18594 * config/aarch64/aarch64-simd.md
18595 (simd_mode): Add SF.
18596 (aarch64_vcond_internal): Use new names for unsigned comparison insns.
18597 (aarch64_cm<optab><mode>): Rewrite to not use UNSPECs.
18598 * config/aarch64/aarch64.md (*cstore<mode>_neg): Rename to...
18599 (cstore<mode>_neg): ...This.
18600 * config/aarch64/iterators.md
18601 (VALLF): new.
18602 (unspec): Remove UNSPEC_CM<EQ, LE, LT, GE, GT, HS, HI, TST>.
18603 (COMPARISONS): New.
18604 (UCOMPARISONS): Likewise.
18605 (optab): Add missing comparisons.
18606 (n_optab): New.
18607 (cmp_1): Likewise.
18608 (cmp_2): Likewise.
18609 (CMP): Likewise.
18610 (cmp): Remove.
18611 (VCMP_S): Likewise.
18612 (VCMP_U): Likewise.
18613 (V_cmp_result): Add DF, SF modes.
18614 (v_cmp_result): Likewise.
18615 (v): Likewise.
18616 (vmtype): Likewise.
18617 * config/aarch64/predicates.md (aarch64_reg_or_fp_zero): New.
18618
0a7dbb76
GY
186192013-05-01 Greta Yorsh <Greta.Yorsh@arm.com>
18620
18621 * config/arm/thumb2.md (thumb2_smaxsi3,thumb2_sminsi3): Convert
18622 define_insn to define_insn_and_split.
18623 (thumb32_umaxsi3,thumb2_uminsi3): Likewise.
18624 (thumb2_negdi2,thumb2_abssi2,thumb2_neg_abssi2): Likewise.
18625 (thumb2_mov_scc,thumb2_mov_negscc,thumb2_mov_notscc): Likewise.
18626 (thumb2_movsicc_insn,thumb2_and_scc,thumb2_ior_scc): Likewise.
18627 (thumb2_negscc): Likewise.
18628
fb614ca6
GY
186292013-04-30 Greta Yorsh <Greta.Yorsh@arm.com>
18630
18631 * config/arm/thumb2.md (thumb2_incscc, thumb2_decscc): Delete.
18632
9e64a0bf
GY
186332013-04-30 Greta Yorsh <Greta.Yorsh@arm.com>
18634
18635 * config/arm/thumb2.md: Remove trailing whitespaces.
18636
d6b28156
RS
186372013-04-30 Richard Sandiford <rsandifo@linux.vnet.ibm.com>
18638
18639 * explow.c (plus_constant): Pass "mode" to immed_double_int_const.
18640 Use gen_int_mode rather than GEN_INT.
18641
f91674c3
L
186422013-04-30 H.J. Lu <hongjiu.lu@intel.com>
18643
b0dec607 18644 * value-prof.c (stream_in_histogram_value): Remove the strayed
f91674c3
L
18645 debug_gimple_stmt.
18646
3551257c
RB
186472013-04-30 Richard Biener <rguenther@suse.de>
18648
18649 PR middle-end/57122
3c21604f 18650 * cfghooks.c (split_edge): Properly check for the loop latch edge.
3551257c 18651
4c1aff1c
RB
186522013-04-30 Richard Biener <rguenther@suse.de>
18653
18654 PR middle-end/57107
18655 * tree-eh.c (sink_clobbers): Preserve virtual SSA form.
18656
54b8379a
AB
186572013-04-30 Andrey Belevantsev <abel@ispras.ru>
18658
18659 PR rtl-optimization/56957
18660 PR rtl-optimization/57105
54b8379a
AB
18661 * sel-sched.c (move_op_orig_expr_found): Remove insn_emitted
18662 variable. Use just INSN_UID for determining whether an insn
18663 should be only disconnected from the insn stream.
18664 * sel-sched-ir.h (EXPR_WAS_CHANGED): Remove.
18665
abc27962
JJ
186662013-04-30 Jakub Jelinek <jakub@redhat.com>
18667
18668 PR tree-optimization/57104
18669 * tsan.c (instrument_expr): Don't instrument accesses to
18670 DECL_HARD_REGISTER VAR_DECLs.
18671
0fc822d0
RB
186722013-04-30 Richard Biener <rguenther@suse.de>
18673
18674 * function.h (loops_for_fn): New inline function.
18675 (set_loops_for_fn): Likewise.
18676 * cfgloop.h (place_new_loop): Add struct function parameter.
18677 (get_loop): Likewise.
18678 (get_loops): Likewise.
18679 (number_of_loops): Likewise.
18680 (fel_next): Adjust.
18681 (fel_init): Likewise.
18682 * cfg.c (get_loop_copy): Adjust.
18683 * cfgloop.c (flow_loops_dump): Likewise.
18684 (record_loop_exits): Likewise.
18685 (verify_loop_structure): Likewise.
18686 * cfgloopanal.c (mark_irreducible_loops): Likewise.
18687 (estimate_reg_pressure_cost): Likewise.
18688 (mark_loop_exit_edges): Likewise.
18689 * cfgloopmanip.c (place_new_loop): Likewise.
18690 (add_loop): Likewise.
18691 (duplicate_loop): Likewise.
18692 * graph.c (draw_cfg_nodes): Likewise.
18693 * graphite-clast-to-gimple.c (translate_clast_user): Likewise.
18694 * graphite-sese-to-poly.c (build_scop_scattering): Likewise.
18695 (extract_affine_chrec): Likewise.
18696 (build_scop_iteration_domain): Likewise.
18697 * graphite.c (graphite_initialize): Likewise.
18698 * ira-build.c (create_loop_tree_nodes): Likewise.
18699 (more_one_region_p): Likewise.
18700 (rebuild_regno_allocno_maps): Likewise.
18701 (mark_loops_for_removal): Likewise.
18702 (mark_all_loops_for_removal): Likewise.
18703 (remove_unnecessary_regions): Likewise.
18704 (ira_build): Likewise.
18705 * ira-emit.c (setup_entered_from_non_parent_p): Likewise.
18706 * loop-init.c (fix_loop_structure): Likewise.
18707 (gate_rtl_move_loop_invariants): Likewise.
18708 (gate_rtl_unswitch): Likewise.
18709 (gate_rtl_unroll_and_peel_loops): Likewise.
18710 (rtl_doloop): Likewise.
18711 * lto-streamer-in.c (input_cfg): Likewise.
18712 * lto-streamer-out.c (output_cfg): Likewise.
18713 * modulo-sched.c (sms_schedule): Likewise.
18714 * predict.c (tree_estimate_probability): Likewise.
18715 (tree_estimate_probability_driver): Likewise.
18716 (estimate_loops): Likewise.
18717 * tree-cfg.c (fixup_loop_arrays_after_move): Likewise.
18718 (move_sese_region_to_fn): Likewise.
18719 (debug_loop_num): Likewise.
18720 * tree-chrec.c (chrec_evaluate): Likewise.
18721 (hide_evolution_in_other_loops_than_loop): Likewise.
18722 (chrec_component_in_loop_num): Likewise.
18723 (reset_evolution_in_loop): Likewise.
18724 (evolution_function_is_invariant_rec_p): Likewise.
18725 * tree-if-conv.c (main_tree_if_conversion): Likewise.
18726 * tree-inline.c (copy_loops): Likewise.
18727 (copy_cfg_body): Likewise.
18728 (tree_function_versioning): Likewise.
18729 * tree-loop-distribution.c (rdg_flag_loop_exits): Likewise.
18730 * tree-scalar-evolution.c (chrec_contains_symbols_defined_in_loop):
18731 Likewise.
18732 (add_to_evolution_1): Likewise.
18733 (scev_const_prop): Likewise.
18734 * tree-scalar-evolution.h (get_chrec_loop): Likewise.
18735 * tree-ssa-loop-ch.c (copy_loop_headers): Likewise.
18736 * tree-ssa-loop-im.c (analyze_memory_references): Likewise.
18737 (tree_ssa_lim_initialize): Likewise.
18738 * tree-ssa-loop-manip.c (rewrite_into_loop_closed_ssa): Likewise.
18739 (verify_loop_closed_ssa): Likewise.
18740 * tree-ssa-loop.c (tree_ssa_loop_init): Likewise.
18741 (tree_ssa_loop_im): Likewise.
18742 (tree_ssa_loop_unswitch): Likewise.
18743 (tree_vectorize): Likewise.
18744 (check_data_deps): Likewise.
18745 (tree_ssa_loop_ivcanon): Likewise.
18746 (tree_ssa_loop_bounds): Likewise.
18747 (tree_complete_unroll): Likewise.
18748 (tree_complete_unroll_inner): Likewise.
18749 (tree_parallelize_loops): Likewise.
18750 (tree_ssa_loop_prefetch): Likewise.
18751 (tree_ssa_loop_ivopts): Likewise.
18752 * tree-ssa.c (execute_update_addresses_taken): Liekwise.
18753 * tree-vectorizer.c (vectorize_loops): Likewise.
18754
37953bd3
MF
187552013-04-29 Mike Frysinger <vapier@gentoo.org>
18756
18757 * config/arm/bpabi.h (EABI_LINK_SPEC): Define.
18758 (BPABI_LINK_SPEC): Use new EABI_LINK_SPEC.
18759 * config/arm/linux-eabi.h (LINK_SPEC): Replace BE8_LINK_SPEC
18760 with EABI_LINK_SPEC.
18761
f9ed28db
UB
187622013-04-29 Uros Bizjak <ubizjak@gmail.com>
18763
18764 PR target/44578
18765 * config/i386/i386.md (*zero_extendsidi2): Add "!" to m->?*y
18766 alternative.
18767
deca73f5
VM
187682013-04-29 Vladimir Makarov <vmakarov@redhat.com>
18769
18770 PR target/57097
37953bd3 18771 * lra-constraints.c (process_alt_operands): Discourage a bit more
deca73f5
VM
18772 using memory for pseudos. Print cost dump for alternatives.
18773 Modify cost values for conflicts with early clobbers.
18774 (curr_insn_transform): Spill pseudos reassigned to NO_REGS.
18775
fdca7d03
UB
187762013-04-29 Uros Bizjak <ubizjak@gmail.com>
18777
18778 PR target/57098
18779 * config/i386/i386.c (ix86_expand_vec_perm): Validize constant memory.
18780
0b064172
IB
187812013-04-29 Ian Bolton <ian.bolton@arm.com>
18782
18783 * config/aarch64/aarch64.md (movsi_aarch64): Support LDR/STR
18784 from/to S register.
18785 (movdi_aarch64): Support LDR/STR from/to D register.
18786
473cec55
IB
187872013-04-29 Ian Bolton <ian.bolton@arm.com>
18788
18789 * common/config/aarch64/aarch64-common.c: Enable REE pass at O2
18790 or higher by default.
18791
a6f30e66
RB
187922013-04-29 Richard Biener <rguenther@suse.de>
18793
18794 PR middle-end/57075
18795 * tree-inline.c (copy_edges_for_bb): Still split the bbs,
18796 even if not adding abnormal edges for calls that can make
18797 abnormal gotos.
18798
0c2b2040
RB
187992013-04-29 Richard Biener <rguenther@suse.de>
18800
18801 PR middle-end/57103
18802 * tree-cfg.c (move_stmt_op): Fix condition under which to update
18803 TREE_BLOCK.
18804 (move_stmt_r): Remove redundant checking.
18805
f41f80f9
TJ
188062013-04-29 Teresa Johnson <tejohnson@google.com>
18807
18808 PR bootstrap/57077
18809 * basic-block.h (apply_scale): New function.
18810 (apply_probability): Use apply_scale.
18811 * gimple-streamer-in.c (input_bb): Ditto.
18812 * lto-streamer-in.c (input_cfg): Ditto.
18813 * lto-cgraph.c (merge_profile_summaries): Ditto.
18814 * tree-optimize.c (execute_fixup_cfg): Ditto.
fdca7d03 18815 * tree-inline.c (copy_bb): Update comment to use apply_scale.
f41f80f9
TJ
18816 (copy_edges_for_bb): Ditto.
18817 (copy_cfg_body): Ditto.
18818
315bbd2e
TV
188192013-04-29 Tom de Vries <tom@codesourcery.com>
18820
18821 * tree-ssa-tail-merge.c (find_same_succ_bb): Skip loop latch bbs.
18822 (replace_block_by): Don't set LOOPS_NEED_FIXUP.
18823 (tail_merge_optimize): Handle current_loops == NULL.
18824
ebbd90d8
JL
188252013-04-26 Jeff Law <law@redhat.com>
18826
18827 * tree-vrp.c (range_fits_type_p): Move to earlier point in file.
18828 (simplify_cond_using_ranges): Generalize code to simplify
18829 COND_EXPRs where one argument is a constant and the other
18830 is an SSA_NAME created by an integral type conversion.
18831
8b9b57eb
KT
188322013-04-29 Kyrylo Tkachov <kyrylo.tkachov@arm.com>
18833
18834 * config/arm/arm.md (store_minmaxsi): Use only when
18835 optimize_insn_for_size_p.
18836
9498e5dc
CB
188372013-04-29 Christian Bruel <christian.bruel@st.com>
18838
18839 PR target/57108
18840 * sh.md (tstsi_t_zero_extract_eq): Set mode for operand 0.
18841
6093bc06
RB
188422013-04-29 Richard Biener <rguenther@suse.de>
18843
18844 PR middle-end/57089
fdca7d03
UB
18845 * omp-low.c (expand_omp_taskreg): If the parent function had a broken
18846 loop tree make sure to schedule a fixup for the child as well.
6093bc06
RB
18847 (expand_omp_for_generic): Properly add loops.
18848 (expand_omp_for_static_nochunk): Likewise.
18849 (expand_omp_for_static_chunk): Likewise.
18850 (expand_omp_for): For the degenerate case fixup loops.
18851 (expand_omp_sections): Fix default bb placement in loops.
18852 (expand_omp_atomic_pipeline): Properly add loops.
18853
84aacbfd
KT
188542013-04-29 Kyrylo Tkachov <kyrylo.tkachov@arm.com>
18855
18856 * predict.c: Fix typo in comment above #define PROB_VERY_UNLIKELY.
18857
8dee4479
TV
188582013-04-29 Tom de Vries <tom@codesourcery.com>
18859
18860 * tree-ssa-tail-merge.c: Update header comment.
18861
47934dc4
JG
188622013-04-29 James Greenhalgh <james.greenhalgh@arm.com>
18863
18864 * config/aarch64/arm_neon.h
18865 (vcvt<sd>_f<32,64>_s<32,64>): Rewrite in C.
18866 (vcvt<q>_f<32,64>_s<32,64>): Rewrite using builtins.
18867 (vcvt_<high_>_f<32,64>_f<32,64>): Likewise.
18868 (vcvt<qsd>_<su><32,64>_f<32,64>): Likewise.
18869 (vcvta<qsd>_<su><32,64>_f<32,64>): Likewise.
18870 (vcvtm<qsd>_<su><32,64>_f<32,64>): Likewise.
18871 (vcvtn<qsd>_<su><32,64>_f<32,64>): Likewise.
18872 (vcvtp<qsd>_<su><32,64>_f<32,64>): Likewise.
18873
384be29f
JG
188742013-04-29 James Greenhalgh <james.greenhalgh@arm.com>
18875
18876 * config/aarch64/aarch64-simd.md
18877 (<optab><VDQF:mode><fcvt_target>2): New, maps to fix, fixuns.
18878 (<fix_trunc_optab><VDQF:mode><fcvt_target>2): New, maps to
18879 fix_trunc, fixuns_trunc.
18880 (ftrunc<VDQF:mode>2): New.
18881 * config/aarch64/iterators.md (optab): Add fix, fixuns.
18882 (fix_trunc_optab): New.
18883
0386b123
JG
188842013-04-29 James Greenhalgh <james.greenhalgh@arm.com>
18885
18886 * config/aarch64/aarch64-builtins.c
18887 (aarch64_builtin_vectorized_function): Vectorize over ifloorf,
18888 iceilf, lround, iroundf.
18889
00fcb892
UB
188902013-04-29 Uros Bizjak <ubizjak@gmail.com>
18891
18892 PR target/54349
18893 * config/i386/i386.h (enum ix86_tune_indices)
18894 <X86_TUNE_INTER_UNIT_MOVES_TO_VEC, X86_TUNE_INTER_UNIT_MOVES_FROM_VEC>:
18895 New, split from X86_TUNE_INTER_UNIT_MOVES.
18896 <X86_TUNE_INTER_UNIT_MOVES>: Remove.
18897 (TARGET_INTER_UNIT_MOVES_TO_VEC): New define.
18898 (TARGET_INTER_UNIT_MOVES_FROM_VEC): Ditto.
18899 (TARGET_INTER_UNIT_MOVES): Remove.
18900 * config/i386/i386.c (initial_ix86_tune_features): Update.
18901 Disable X86_TUNE_INTER_UNIT_MOVES_FROM_VEC for m_ATHLON_K8 only.
18902 (ix86_expand_convert_uns_didf_sse): Use
18903 TARGET_INTER_UNIT_MOVES_TO_VEC instead of TARGET_INTER_UNIT_MOVES.
18904 (ix86_expand_vector_init_one_nonzero): Ditto.
18905 (ix86_expand_vector_init_interleave): Ditto.
18906 (inline_secondary_memory_needed): Return true for moves from SSE class
18907 registers for !TARGET_INTER_UNIT_MOVES_FROM_VEC targets and for moves
18908 to SSE class registers for !TARGET_INTER_UNIT_MOVES_TO_VEC targets.
18909 * config/i386/constraints.md (Yi, Ym): Depend on
18910 TARGET_INTER_UNIT_MOVES_TO_VEC.
18911 (Yj, Yn): New constraints.
18912 * config/i386/i386.md (*movdi_internal): Change constraints of
18913 operand 1 from Yi to Yj and from Ym to Yn.
18914 (*movsi_internal): Ditto.
18915 (*movdf_internal): Ditto.
18916 (*movsf_internal): Ditto.
18917 (*float<SWI48x:mode><X87MODEF:mode>2_1): Use
18918 TARGET_INTER_UNIT_MOVES_TO_VEC instead of TARGET_INTER_UNIT_MOVES.
18919 (*float<SWI48x:mode><X87MODEF:mode>2_1 splitters): Ditto.
18920 (floatdi<X87MODEF:mode>2_i387_with_xmm): Ditto.
18921 (floatdi<X87MODEF:mode>2_i387_with_xmm splitters): Ditto.
18922 * config/i386/sse.md (movdi_to_sse): Ditto.
18923 (sse2_stored): Change constraint of operand 1 from Yi to Yj.
18924 Use TARGET_INTER_UNIT_MOVES_FROM_VEC instead of
18925 TARGET_INTER_UNIT_MOVES.
18926 (sse_storeq_rex64): Change constraint of operand 1 from Yi to Yj.
18927 (sse_storeq_rex64 splitter): Use TARGET_INTER_UNIT_MOVES_FROM_VEC
18928 instead of TARGET_INTER_UNIT_MOVES.
18929 * config/i386/mmx.md (*mov<mode>_internal): Change constraint of
18930 operand 1 from Yi to Yj and from Ym to Yn.
18931
4c871069
JG
189322013-04-29 James Greenhalgh <james.greenhalgh@arm.com>
18933
18934 * config/aarch64/aarch64-simd-builtins.def (vec_unpacks_hi_): New.
18935 (float_truncate_hi_): Likewise.
18936 (float_extend_lo_): Likewise.
18937 (float_truncate_lo_): Likewise.
18938 * config/aarch64/aarch64-simd.md (vec_unpacks_lo_v4sf): New.
18939 (aarch64_float_extend_lo_v2df): Likewise.
18940 (vec_unpacks_hi_v4sf): Likewise.
18941 (aarch64_float_truncate_lo_v2sf): Likewise.
18942 (aarch64_float_truncate_hi_v4sf): Likewise.
18943 (vec_pack_trunc_v2df): Likewise.
18944 (vec_pack_trunc_df): Likewise.
18945
1709ff9b
JG
189462013-04-29 James Greenhalgh <james.greenhalgh@arm.com>
18947
18948 * config/aarch64/aarch64-builtins.c
18949 (aarch64_fold_builtin): Fold float conversions.
18950 * config/aarch64/aarch64-simd-builtins.def
18951 (floatv2si, floatv4si, floatv2di): New.
18952 (floatunsv2si, floatunsv4si, floatunsv2di): Likewise.
18953 * config/aarch64/aarch64-simd.md
18954 (<optab><fcvt_target><VDQF:mode>2): New, expands to float and floatuns.
18955 * config/aarch64/iterators.md (FLOATUORS): New.
18956 (optab): Add float, floatuns.
18957 (su_optab): Likewise.
18958
ce966824
JG
189592013-04-29 James Greenhalgh <james.greenhalgh@arm.com>
18960
18961 * config/aarch64/aarch64-builtins.c
18962 (aarch64_builtin_vectorized_function): Use new names for
18963 fcvt builtins.
18964 * config/aarch64/aarch64-simd-builtins.def (fcvtzs): Split as...
18965 (lbtruncv2sf, lbtruncv4sf, lbtruncv2df): ...This.
18966 (fcvtzu): Split as...
18967 (lbtruncuv2sf, lbtruncuv4sf, lbtruncuv2df): ...This.
18968 (fcvtas): Split as...
18969 (lroundv2sf, lroundv4sf, lroundv2df, lroundsf, lrounddf): ...This.
18970 (fcvtau): Split as...
18971 (lrounduv2sf, lrounduv4sf, lrounduv2df, lroundusf, lroundudf): ...This.
18972 (fcvtps): Split as...
18973 (lceilv2sf, lceilv4sf, lceilv2df): ...This.
18974 (fcvtpu): Split as...
18975 (lceiluv2sf, lceiluv4sf, lceiluv2df, lceilusf, lceiludf): ...This.
18976 (fcvtms): Split as...
18977 (lfloorv2sf, lfloorv4sf, lfloorv2df): ...This.
18978 (fcvtmu): Split as...
18979 (lflooruv2sf, lflooruv4sf, lflooruv2df, lfloorusf, lfloorudf): ...This.
18980 (lfrintnv2sf, lfrintnv4sf, lfrintnv2df, lfrintnsf, lfrintndf): New.
18981 (lfrintnuv2sf, lfrintnuv4sf, lfrintnuv2df): Likewise.
18982 (lfrintnusf, lfrintnudf): Likewise.
18983 * config/aarch64/aarch64-simd.md
18984 (l<fcvt_pattern><su_optab><fcvt_target><VDQF:mode>2): Convert to
18985 define_insn.
18986 (aarch64_fcvt<frint_suffix><su><mode>): Remove.
18987 * config/aarch64/iterators.md (FCVT): Include UNSPEC_FRINTN.
18988 (fcvt_pattern): Likewise.
18989
b9de24fe
JG
189902013-04-29 James Greenhalgh <james.greenhalgh@arm.com>
18991
18992 * config/aarch64/aarch64-simd.md
18993 (l<fcvt_pattern><su_optab><fcvt_target><VDQF:mode>2): Rename to...
18994 (l<fcvt_pattern><su_optab><VDQF:mode><fcvt_target>2): ... This.
18995
77a205be
JG
189962013-04-29 James Greenhalgh <james.greenhalgh@arm.com>
18997
18998 * config/aarch64/arm_neon.h (vrndq<a,m,n,p>_f<32, 64>): Rename to...
18999 (vrnd<a,m,n,p>q_f<32, 64>): ...This, implement using builtin.
19000 (vrnd<a,m,n,p>_f32): Implement using builtins.
19001 (vrnd<i,x><q>_f<32, 64>): New.
19002
0659ce6f
JG
190032013-04-29 James Greenhalgh <james.greenhalgh@arm.com>
19004
19005 * config/aarch64/aarch64-builtins.c
19006 (aarch64_builtin_vectorized_function): Fold to standard pattern names.
19007 * config/aarch64/aarch64-simd-builtins.def (frintn): New.
19008 (frintz): Rename to...
19009 (btrunc): ...this.
19010 (frintp): Rename to...
19011 (ceil): ...this.
19012 (frintm): Rename to...
19013 (floor): ...this.
19014 (frinti): Rename to...
19015 (nearbyint): ...this.
19016 (frintx): Rename to...
19017 (rint): ...this.
19018 (frinta): Rename to...
19019 (round): ...this.
19020 * config/aarch64/aarch64-simd.md
19021 (aarch64_frint<frint_suffix><mode>): Delete.
19022 (<frint_pattern><mode>2): Convert to insn.
19023 * config/aarch64/aarch64.md (unspec): Add UNSPEC_FRINTN.
19024 * config/aarch64/iterators.md (FRINT): Add UNSPEC_FRINTN.
19025 (frint_pattern): Likewise.
19026 (frint_suffix): Likewise.
19027
ea78906a
RB
190282013-04-29 Richard Biener <rguenther@suse.de>
19029
19030 PR tree-optimization/57081
19031 * loop-init.c: Include tree-flow.h.
19032 (loop_optimizer_finalize): Free number of iteration estimates.
19033 * Makefile.in (loop-init.o): Add $(TREE_FLOW_H) dependency.
19034
baee1763
JJ
190352013-04-29 Jakub Jelinek <jakub@redhat.com>
19036
94dc5332
JJ
19037 PR tree-optimization/57083
19038 * tree-vrp.c (extract_range_from_binary_expr_1): For LSHIFT_EXPR with
19039 non-singleton shift count range, zero extend low_bound for uns case.
19040
baee1763
JJ
19041 * config/i386/predicates.md (general_vector_operand): New predicate.
19042 * config/i386/i386.c (const_vector_equal_evenodd_p): New function.
19043 (ix86_expand_mul_widen_evenodd): Force op1 resp. op2 into register
19044 if they aren't nonimmediate operands. If their original values
19045 satisfy const_vector_equal_evenodd_p, don't shift them.
19046 * config/i386/sse.md (mul<mode>3): Use general_vector_operand
19047 predicates. For the SSE4.1 case force operands[{1,2}] into registers
19048 if not nonimmediate_operand.
19049 (vec_widen_smult_even_v4si): Use nonimmediate_operand predicates
19050 instead of register_operand.
19051 (vec_widen_<s>mult_odd_<mode>): Use general_vector_operand predicates.
19052
a9073727 190532013-04-28 Eric Botcazou <ebotcazou@adacore.com>
22ad5e0a
EB
19054
19055 * stor-layout.c (finalize_size_functions): Allocate a structure and
19056 reset cfun before dumping the functions.
19057
ba8011e6
JJ
190582013-04-27 Jakub Jelinek <jakub@redhat.com>
19059
d6fde69e
JJ
19060 * config/i386/i386.c (ix86_expand_call): Make cregs_size unsigned.
19061
ba8011e6
JJ
19062 PR target/56866
19063 * config/i386/i386.c (ix86_expand_mul_widen_evenodd): Don't
19064 use xop_pmacsdqh if uns_p.
19065 * config/i386/sse.md (xop_rotr<mode>3): Fix up computation of
19066 the immediate rotate count.
19067
2c62cbaa
VM
190682013-04-26 Vladimir Makarov <vmakarov@redhat.com>
19069
19070 * rtl.h (struct rtx_def): Add comment for field jump.
19071 (LRA_SUBREG_P): New macro.
19072 * recog.c (register_operand): Check LRA_SUBREG_P.
19073 * lra.c (lra): Add note at the end of RTL code. Align non-empty
19074 stack frame.
19075 * lra-spills.c (lra_spill): Align stack after spilling pseudos.
19076 (lra_final_code_change): Skip subreg change for operators.
19077 * lra-eliminations.c (eliminate_regs_in_insn): Make return earlier
19078 if there are no operand changes.
19079 * lra-constraints.c (curr_insn_set): New.
19080 (match_reload): Set LRA_SUBREG_P.
19081 (emit_spill_move): Ditto.
19082 (check_and_process_move): Use curr_insn_set. Process only single
19083 set insns. Don't initialize sec_mem_p and change_p.
19084 (simplify_operand_subreg): Use LRA_SUBREG_P.
19085 (reg_in_class_p): New function.
19086 (process_alt_operands): Use it. Use #if HAVE_ATTR_enabled instead
19087 of #ifdef. Add code to remove cycling.
19088 (process_address): Check EXTRA_CONSTRAINT_STR. Process even if
19089 non-null disp. Reload inner instead of disp when base and index
19090 are null. Try to put lo_sum into register.
19091 (EBB_PROBABILITY_CUTOFF): Redefine probability in percents.
19092 (check_and_process_move): Move code for move cost check to
19093 simple_move_p. Remove equiv_substitution.
19094 (simple_move_p): New function.
19095 (curr_insn_transform): Initialize sec_mem_p and change_p. Set up
19096 curr_insn_set. Call check_and_process_move only for single set
19097 insns. Use the new function. Move call of check_and_process_move
19098 after operand equiv substitution and address process.
19099
e7d764f3
JJ
191002013-04-26 Jakub Jelinek <jakub@redhat.com>
19101
19102 PR go/57045
19103 * tree-ssa-uninit.c (compute_uninit_opnds_pos): In functions
19104 with nonlocal goto receivers or returns twice calls, ignore
19105 unininitialized values from abnormal edges to nl goto receiver
19106 or returns twice call.
19107
41e10689
JJ
191082013-04-26 Jakub Jelinek <jakub@redhat.com>
19109
19110 PR tree-optimization/57051
19111 * fold-const.c (const_binop): Handle VEC_LSHIFT_EXPR
19112 and VEC_RSHIFT_EXPR if shift count is a multiple of element
19113 bitsize.
19114
d7ed20db
RB
191152013-04-26 Richard Biener <rguenther@suse.de>
19116
19117 * omp-low.c (finalize_task_copyfn): Do not drop PROP_loops.
19118 (expand_omp_taskreg): Likewise. Mark loops for fixup.
19119 * tree-cfg.c (move_block_to_fn): Remap loop fathers.
19120 (fixup_loop_arrays_after_move): New function.
19121 (move_sese_region_to_fn): Properly outline the loop tree parts
19122 of the SESE region.
19123
df93505e
UB
191242013-04-26 Uros Bizjak <ubizjak@gmail.com>
19125
19126 * config/i386/i386.md (type, unit): Fix long lines.
19127
dd366ec3
RB
191282013-04-26 Richard Biener <rguenther@suse.de>
19129
19130 * Makefile.in (lto-streamer-in.o): Add $(CFGLOOP_H) dependency.
19131 (lto-streamer-out.o): Likewise.
19132 * cfgloop.c (init_loops_structure): Export, add struct function
19133 argument and adjust.
19134 (flow_loops_find): Adjust.
19135 * cfgloop.h (enum loop_estimation): Add EST_LAST.
19136 (init_loops_structure): Declare.
19137 * lto-streamer-in.c: Include cfgloop.h.
19138 (input_cfg): Input the loop tree.
19139 * lto-streamer-out.c: Include cfgloop.h.
19140 (output_cfg): Output the loop tree.
19141 (output_struct_function_base): Do not drop PROP_loops.
19142
a9e0d843
RB
191432013-03-26 Richard Biener <rguenther@suse.de>
19144
19145 * tree-cfg.c (execute_build_cfg): Build the loop tree.
19146 (pass_build_cfg): Provide PROP_loops.
19147 (move_sese_region_to_fn): Remove loops that are outlined into fn
19148 for now.
19149 * tree-inline.c: Include cfgloop.h.
19150 (initialize_cfun): Do not drop PROP_loops.
19151 (copy_loops): New function.
19152 (copy_cfg_body): Copy loop structure.
19153 (tree_function_versioning): Initialize destination loop tree.
19154 * tree-ssa-loop.c (pass_tree_loop_init): Do not provide PROP_loops.
19155 (pass_parallelize_loops): Do IL verification.
19156 * loop-init.c (loop_optimizer_init): Fixup loops if required.
19157 * tree-optimize.c (execute_fixup_cfg): If we need to cleanup
19158 the CFG make sure we fixup loops as well.
19159 * tree-ssa-tail-merge.c: Include cfgloop.h.
19160 (replace_block_by): When merging loop latches mark loops for fixup.
19161 * lto-streamer-out.c (output_struct_function_base): Drop
19162 PROP_loops for now.
19163 * tree-ssa-phiopt.c: Include tree-scalar-evolution.h.
19164 (tree_ssa_cs_elim): Initialize the loop optimizer and SCEV.
19165 * ipa-split.c: Include cfgloop.h.
19166 (split_function): Add the new return block to the loop tree root.
19167 * tree-cfgcleanup.c (remove_forwarder_block_with_phi): Return
19168 whether we have removed the forwarder block.
19169 (merge_phi_nodes): If we removed a forwarder mark loops for fixup.
19170 * cfgloop.h (place_new_loop): Declare.
19171 * cfgloopmanip.c (place_new_loop): Export.
19172 * Makefile.in (asan.o): Add $(CFGLOOP_H) dependency.
19173 (tree-switch-conversion.o): Likewise.
19174 (tree-complex.o): Likewise.
19175 (tree-inline.o): Likewise.
19176 (tree-ssa-tailmerge.o): Likewise.
19177 (ipa-split.o): Likewise.
19178 (tree-ssa-phiopt.o): Add $(SCEV_H) dependency.
19179 (tree-ssa-copy.o): Likewise.
19180 * tree-switch-conversion.c: Include cfgloop.h
19181 (process_switch): If we emit a bit-test cascade, schedule loops
19182 for fixup.
19183 * tree-complex.c: Include cfgloop.h.
19184 (expand_complex_div_wide): Properly add new basic-blocks to loops.
19185 * asan.c: Include cfgloop.h.
19186 (create_cond_insert_point): Properly add new basic-blocks to
19187 loops, schedule loop fixup.
19188 * cfgloop.c (verify_loop_structure): Check that looks are not
19189 marked for fixup.
19190 * omp-low.c (expand_parallel_call): Properly add new basic-blocks
19191 to loops.
19192 (expand_omp_for_generic): Likewise.
19193 (expand_omp_sections): Likewise.
19194 (expand_omp_atomic_pipeline): Schedule loops for fixup.
19195 * tree-ssa-copy.c: Include tree-scalar-evolution.h.
19196 (fini_copy_prop): Disable DCE in substitute_and_fold if SCEV
19197 is initialized, not when loops are present.
19198 * tree-parloops.c (parallelize_loops): Remove checking here.
19199 * passes.c (init_optimization_passes): Schedule a copy-propagation
19200 pass before complete unrolling of inner loops.
19201
e78e8a0b
JJ
192022013-04-26 Jakub Jelinek <jakub@redhat.com>
19203
a2e836b2
JJ
19204 * Makefile.in (toplev.o): Depend on diagnostic-color.h.
19205 * diagnostic-color.c (should_colorize): Remove _WIN32 version.
19206 (colorize_init): Add argument to _WIN32 version.
19207 * toplev.c: Include diagnostic-color.h.
19208 (process_options): Default to -fdiagnostics-color=auto if
19209 GCC_COLORS env var is in the environment.
19210 * common.opt (fdiagnostics-color=): Add Var and Init.
df93505e
UB
19211 * doc/invoke.texi (-fdiagnostics-color=): Document that if GCC_COLORS
19212 env var is in the environment, the default is auto rather than never.
a2e836b2 19213
e78e8a0b
JJ
19214 * diagnostic.h (file_name_as_prefix): Add context argument.
19215 * diagnostic.c (file_name_as_prefix): Likewise. Colorize
19216 the string as locus.
19217 * langhooks.c (lhd_print_error_function): Adjust caller.
19218
013e5ef9
LC
192192013-04-25 Lawrence Crowl <crowl@google.com>
19220
19221 * var-tracking.c (shared_hash_def::htab):
19222 Change type to hash_table. Update dependent calls and types.
19223
4a8fb1a1
LC
192242013-04-25 Lawrence Crowl <crowl@google.com>
19225
19226 * Makefile.in: Update as needed below.
19227
19228 * alloc-pool.c (static hash_table <alloc_pool_hasher> alloc_pool_hash):
19229 Move declaration to after the type's method definitons.
19230
19231 * attribs.c (htab_t scoped_attributes::attribute_hash):
19232 Change type to hash_table. Update dependent calls and types.
19233
19234 * bitmap.c (htab_t bitmap_desc_hash):
19235 Change type to hash_table. Update dependent calls and types.
19236
19237 * cselib.c (htab_t cselib_hash_table):
19238 Change type to hash_table. Update dependent calls and types.
19239
19240 * data-streamer.h (struct string_slot): Move to lto-streamer.h.
19241 (hash_string_slot_node): Move implementation into lto-streamer.h
19242 struct string_slot_hasher.
19243 (eq_string_slot_node): Likewise.
19244
19245 * data-streamer-out.c: Update output_block::string_hash_table
19246 dependent calls and types.
19247
19248 * dwarf2cfi.c (htab_t trace_index):
19249 Change type to hash_table. Update dependent calls and types.
19250
19251 * dwarf2out.c (htab_t break_out_includes::cu_hash_table):
19252 Change type to hash_table. Update dependent calls and types.
19253 (htab_t copy_decls_for_unworthy_types::decl_table): Likewise.
19254 (htab_t optimize_external_refs::map): Likewise.
19255 (htab_t output_comp_unit::extern_map): Likewise.
19256 (htab_t output_comdat_type_unit::extern_map): Likewise.
19257 (htab_t output_macinfo::macinfo_htab): Likewise.
19258 (htab_t optimize_location_lists::htab): Likewise.
19259 (htab_t dwarf2out_finish::comdat_type_table): Likewise.
19260
19261 * except.c (htab_t ehspec_hash_type):
19262 Change type to hash_table. Update dependent calls and types.
19263 (assign_filter_values::ttypes): Likewise.
19264 (assign_filter_values::ehspec): Likewise.
19265 (sjlj_assign_call_site_values::ar_hash): Likewise.
19266 (convert_to_eh_region_ranges::ar_hash): Likewise.
19267
19268 * gcse.c (htab_t pre_ldst_table):
19269 Change type to hash_table. Update dependent calls and types.
19270
19271 * ggc-common.c (htab_t saving_htab):
19272 Change type to hash_table. Update dependent calls and types.
19273 (htab_t loc_hash): Likewise.
19274 (htab_t ptr_hash): Likewise.
19275 (call_count): Rename ggc_call_count.
19276 (call_alloc): Rename ggc_call_alloc.
19277 (loc_descriptor): Rename make_loc_descriptor.
19278 (add_statistics): Rename ggc_add_statistics.
19279
19280 * ggc-common.c (saving_htab):
19281 Change type to hash_table. Update dependent calls and types.
19282
19283 * gimple.h (struct gimplify_ctx): Move to gimplify-ctx.h.
19284 (push_gimplify_context): Likewise.
19285 (pop_gimplify_context): Likewise.
19286 (struct gimple_temp_hash_elt): Added.
19287 (struct gimplify_hasher): Likewise.
19288 (struct gimplify_ctx.temp_htab):
19289 Change type to hash_table. Update dependent calls and types.
19290
19291 * gimple-fold.c: Include gimplify-ctx.h.
19292
19293 * gimple-ssa-strength-reduction.c (htab_t base_cand_map):
19294 Change type to hash_table. Update dependent calls and types.
19295 (base_cand_dump_callback): Rename to ssa_base_cand_dump_callback to
19296 avoid potential global name collision.
19297
19298 * gimplify.c: Include gimplify-ctx.h.
19299 (struct gimple_temp_hash_elt): Move to gimplify-ctx.h.
19300 (htab_t gimplify_ctx::temp_htab):
19301 Update dependent calls and types for new type hash_table.
19302 (gimple_tree_hash): Move into gimplify_hasher in gimplify-ctx.h.
19303 (gimple_tree_eq): Move into gimplify_hasher in gimplify-ctx.h.
19304
19305 * gimplify-ctx.h: New.
19306 (struct gimple_temp_hash_elt): Move from gimplify.c.
19307 (class gimplify_hasher): New.
19308 (struct gimplify_ctx): Move from gimple.h.
19309 (htab_t gimplify_ctx::temp_htab):
19310 Change type to hash_table. Update dependent calls and types.
19311
19312 * graphite-clast-to-gimple.c: Include graphite-htab.h.
19313 (htab_t ivs_params::newivs_index):
19314 Change type to hash_table. Update dependent calls and types.
19315 (htab_t ivs_params::params_index): Likewise.
19316 (htab_t print_generated_program::params_index): Likewise.
19317 (htab_t gloog::newivs_index): Likewise.
19318 (htab_t gloog::params_index): Likewise.
19319
19320 * graphite.c: Include graphite-htab.h.
19321 4htab_t graphite_transform_loops::bb_pbb_mapping):
19322 Change type to hash_table. Update dependent calls and types.
19323
19324 * graphite-clast-to-gimple.h: (extern gloog) Move to graphite-htab.h.
19325 (bb_pbb_map_hash): Fold into bb_pbb_htab_type in graphite-htab.h.
19326 (eq_bb_pbb_map): Fold into bb_pbb_htab_type in graphite-htab.h.
19327
19328 * graphite-dependences.c: Include graphite-htab.h.
19329 (loop_is_parallel_p): Change hash table type of parameter.
19330
19331 * graphite-htab.h: New.
19332 (typedef hash_table <bb_pbb_hasher> bb_pbb_htab_type): New.
19333 (extern find_pbb_via_hash): Move from graphite-poly.h.
19334 (extern loop_is_parallel_p): Move from graphite-poly.h.
19335 (extern get_loop_body_pbbs): Move from graphite-poly.h.
19336
19337 * graphite-poly.h (extern find_pbb_via_hash): Move to graphite-htab.h.
19338 (extern loop_is_parallel_p): Move to graphite-htab.h.
19339 (extern get_loop_body_pbbs): Move to graphite-htab.h.
19340
19341 * haifa-sched.c (htab_t delay_htab):
19342 Change type to hash_table. Update dependent calls and types.
19343 (htab_t delay_htab_i2): Likewise.
19344
19345 * ira-color.c (htab_t allocno_hard_regs_htab):
19346 Change type to hash_table. Update dependent calls and types.
19347
19348 * ira-costs.c (htab_t cost_classes_htab):
19349 Change type to hash_table. Update dependent calls and types.
19350
19351 * loop-invariant.c (htab_t merge_identical_invariants::eq):
19352 Change type to hash_table. Update dependent calls and types.
19353
19354 * loop-iv.c (htab_t bivs):
19355 Change type to hash_table. Update dependent calls and types.
19356
19357 * loop-unroll.c (htab_t opt_info::insns_to_split):
19358 Change type to hash_table. Update dependent calls and types.
19359 (htab_t opt_info::insns_with_var_to_expand): Likewise.
19360
19361 * lto-streamer.h (struct string_slot): Move from data-streamer.h
19362 (struct string_slot_hasher): New.
19363 (htab_t output_block::string_hash_table):
19364 Change type to hash_table. Update dependent calls and types.
19365
19366 * lto-streamer-in.c (freeing_string_slot_hasher): New.
19367 (htab_t file_name_hash_table):
19368 Change type to hash_table. Update dependent calls and types.
19369
19370 * lto-streamer-out.c: Update output_block::string_hash_table dependent
19371 calls and types.
19372
19373 * lto-streamer.c (htab_t tree_htab):
19374 Change type to hash_table. Update dependent calls and types.
19375
19376 * omp-low.c: Include gimplify-ctx.h.
19377
19378 * passes.c (htab_t name_to_pass_map):
19379 Change type to hash_table. Update dependent calls and types.
19380 (pass_traverse): Rename to passes_pass_traverse.
19381
19382 * plugin.c (htab_t event_tab):
19383 Change type to hash_table. Update dependent calls and types.
19384
19385 * postreload-gcse.c (htab_t expr_table):
19386 Change type to hash_table. Update dependent calls and types.
19387 (dump_hash_table_entry): Rename dump_expr_hash_table_entry.
19388
19389 * sese.c (debug_rename_map_1): Make extern.
19390 (htab_t copy_bb_and_scalar_dependences::rename_map):
19391 Change type to hash_table. Update dependent calls and types.
19392
19393 * sese.h (extern debug_rename_map): Move to .c file.
19394
19395 * store-motion.c (htab_t store_motion_mems_table):
19396 Change type to hash_table. Update dependent calls and types.
19397
19398 * trans-mem.c (htab_t tm_new_mem_hash):
19399 Change type to hash_table. Update dependent calls and types.
19400
19401 * tree-browser.c (htab_t TB_up_ht):
19402 Change type to hash_table. Update dependent calls and types.
19403
19404 * tree-cfg.c (htab_t discriminator_per_locus):
19405 Change type to hash_table. Update dependent calls and types.
19406
19407 * tree-complex.c: Include tree-hasher.h
19408 (htab_t complex_variable_components):
19409 Change type to hash_table. Update dependent calls and types.
19410
19411 * tree-eh.c (htab_t finally_tree):
19412 Change type to hash_table. Update dependent calls and types.
19413
19414 * tree-flow.h (extern int_tree_map_hash): Moved into tree-hasher
19415 struct int_tree_hasher.
19416 (extern int_tree_map_eq): Likewise.
19417 (uid_decl_map_hash): Removed.
19418 (extern decl_tree_map_eq): Likewise.
19419
19420 * tree-hasher.h: New.
19421 (struct int_tree_hasher): New.
19422 (typedef int_tree_htab_type): New.
19423
19424 * tree-inline.c: Include gimplify-ctx.h.
19425
19426 * tree-mudflap.c: Include gimplify-ctx.h.
19427
19428 * tree-parloops.c: Include tree-hasher.h.
19429 (htab_t eliminate_local_variables_stmt::decl_address):
19430 Change type to hash_table. Update dependent calls and types.
19431 (htab_t separate_decls_in_region::decl_copies): Likewise.
19432
19433 * tree-scalar-evolution.c (htab_t resolve_mixers::cache):
19434 Change type to hash_table. Update dependent calls and types.
19435
19436 * tree-sra.c (candidates):
19437 Change type to hash_table. Update dependent calls and types.
19438
19439 * tree-ssa.c (int_tree_map_eq): Moved into struct int_tree_hasher
19440 in tree-flow.h.
19441 (int_tree_map_hash): Likewise.
19442
19443 * tree-ssa-dom.c (htab_t avail_exprs):
19444 Change type to hash_table. Update dependent calls and types.
19445
19446 * tree-ssa-live.c (var_map_base_init::tree_to_index):
19447 Change type to hash_table. Update dependent calls and types.
19448
19449 * tree-ssa-loop-ivopts.c (struct ivopts_data.inv_expr_tab):
19450 Change type to hash_table. Update dependent calls and types.
19451
19452 * tree-ssa-phiopt.c (seen_ssa_names):
19453 Change type to hash_table. Update dependent calls and types.
19454
19455 * tree-ssa-strlen.c (decl_to_stridxlist_htab):
19456 Change type to hash_table. Update dependent calls and types.
19457
19458 * tree-ssa-uncprop.c (equiv):
19459 Change type to hash_table. Update dependent calls and types.
19460
c5a44004
JJ
194612013-04-25 Jakub Jelinek <jakub@redhat.com>
19462
19463 PR rtl-optimization/57003
19464 * regcprop.c (copyprop_hardreg_forward_1): If ksvd.ignore_set_reg,
19465 call note_stores with kill_clobbered_value callback again after
19466 killing regs_invalidated_by_call.
19467
09962a4a
JG
194682013-04-25 James Greenhalgh <james.greenhalgh@arm.com>
19469
19470 * config/aarch64/aarch64-simd.md
19471 (aarch64_simd_bsl<mode>_internal): Rewrite RTL to not use UNSPEC_BSL.
19472 (aarch64_simd_bsl<mode>): Likewise.
19473 * config/aarch64/iterators.md (unspec): Remove UNSPEC_BSL.
19474
ea28bb0b
MP
194752013-04-25 Marek Polacek <polacek@redhat.com>
19476
19477 PR tree-optimization/57066
3c21604f 19478 * builtins.c (fold_builtin_logb): Return +Inf for -Inf.
ea28bb0b 19479
96659611
JG
194802013-04-25 James Greenhalgh <jame.greenhalgh@arm.com>
19481
19482 * config/aarch64/aarch64-simd.md (neg<mode>2): Use VDQ iterator.
19483
9697e620
JG
194842013-04-25 James Greenhalgh <james.greenhalgh@arm.com>
19485
19486 * config/aarch64/aarch64-builtins.c
19487 (aarch64_fold_builtin): New.
19488 * config/aarch64/aarch64-protos.h (aarch64_fold_builtin): New.
19489 * config/aarch64/aarch64.c (TARGET_FOLD_BUILTIN): Define.
19490 * config/aarch64/aarch64-simd-builtins.def (abs): New.
19491 * config/aarch64/arm_neon.h
19492 (vabs<q>_<f32, 64>): Implement using __builtin_aarch64_fabs.
19493
0ac198d3
JG
194942013-04-25 James Greenhalgh <james.greenhalgh@arm.com>
19495 Tejas Belagod <tejas.belagod@arm.com>
19496
19497 * config/aarch64/aarch64-builtins.c
19498 (aarch64_gimple_fold_builtin): New.
19499 * config/aarch64/aarch64-protos.h (aarch64_gimple_fold_builtin): New.
19500 * config/aarch64/aarch64-simd-builtins.def (addv): New.
19501 * config/aarch64/aarch64-simd.md (addpv4sf): New.
19502 (addvv4sf): Update.
19503 * config/aarch64/aarch64.c (TARGET_GIMPLE_FOLD_BUILTIN): Define.
19504
58cff58c
N
195052013-04-25 Naveen H.S <Naveen.Hurugalawadi@caviumnetworks.com>
19506
df93505e 19507 * config/aarch64/aarch64.md
58cff58c
N
19508 (*cmp_swp_<optab><ALLX:mode>_shft_<GPI:mode>): New pattern.
19509
7e0228bf
N
195102013-04-25 Naveen H.S <Naveen.Hurugalawadi@caviumnetworks.com>
19511
19512 * config/aarch64/aarch64.md (*ngc<mode>): New pattern.
19513 (*ngcsi_uxtw): New pattern.
19514
5819f96f 195152013-04-25 Kyrylo Tkachov <kyrylo.tkachov@arm.com>
3c21604f 19516 Julian Brown <julian@codesourcery.com>
5819f96f
KT
19517
19518 * config/arm/arm.c (neon_builtin_type_mode): Add T_V4HF.
19519 (TB_DREG): Add T_V4HF.
19520 (v4hf_UP): New macro.
19521 (neon_itype): Add NEON_FLOAT_WIDEN, NEON_FLOAT_NARROW.
df93505e 19522 (arm_init_neon_builtins): Handle NEON_FLOAT_WIDEN, NEON_FLOAT_NARROW.
5819f96f
KT
19523 Handle initialisation of V4HF. Adjust initialisation of reinterpret
19524 built-ins.
df93505e 19525 (arm_expand_neon_builtin): Handle NEON_FLOAT_WIDEN, NEON_FLOAT_NARROW.
5819f96f
KT
19526 (arm_vector_mode_supported_p): Handle V4HF.
19527 (arm_mangle_map): Handle V4HFmode.
19528 * config/arm/arm.h (VALID_NEON_DREG_MODE): Add V4HF.
19529 * config/arm/arm_neon_builtins.def: Add entries for
19530 vcvtv4hfv4sf, vcvtv4sfv4hf.
19531 * config/arm/neon.md (neon_vcvtv4sfv4hf): New pattern.
19532 (neon_vcvtv4hfv4sf): Likewise.
19533 * config/arm/neon-gen.ml: Handle half-precision floating point
19534 features.
19535 * config/arm/neon-testgen.ml: Handle Requires_FP_bit feature.
19536 * config/arm/arm_neon.h: Regenerate.
19537 * config/arm/neon.ml (type elts): Add F16.
19538 (type vectype): Add T_float16x4, T_floatHF.
19539 (type vecmode): Add V4HF.
19540 (type features): Add Requires_FP_bit feature.
19541 (elt_width): Handle F16.
19542 (elt_class): Likewise.
19543 (elt_of_class_width): Likewise.
19544 (mode_of_elt): Refactor.
19545 (type_for_elt): Handle F16, fix error messages.
19546 (vectype_size): Handle T_float16x4.
19547 (vcvt_sh): New function.
19548 (ops): Add entries for vcvt_f16_f32, vcvt_f32_f16.
19549 (string_of_vectype): Handle T_floatHF, T_float16, T_float16x4.
19550 (string_of_mode): Handle V4HF.
19551 * doc/arm-neon-intrinsics.texi: Regenerate.
19552
1ef395e4
JG
195532013-04-25 James Greenhalgh <james.greenhalgh@arm.com>
19554
19555 * config/aarch64/aarch64.c (aarch64_print_operand): Fix asm_fprintf
19556 format specifier in 'X' case.
19557
41c34e94
AM
195582013-04-25 Alan Modra <amodra@gmail.com>
19559
19560 PR target/57052
19561 * config/rs6000/rs6000.md (rotlsi3_internal7): Rename to
19562 rotlsi3_internal7le and condition on !BYTES_BIG_ENDIAN.
19563 (rotlsi3_internal8be): New BYTES_BIG_ENDIAN insn.
19564 Repeat for many other rotate/shift and mask patterns using subregs.
19565 Name lshiftrt insns.
19566 (ashrdisi3_noppc64): Rename to ashrdisi3_noppc64be and condition
19567 on WORDS_BIG_ENDIAN.
19568
b9a7eb5d
AM
195692013-04-25 Alan Modra <amodra@gmail.com>
19570
19571 * config.gcc: Support little-endian powerpc-linux targets.
19572 * config/rs6000/linux.h (LINK_OS_LINUX_EMUL): Define.
19573 (LINK_OS_LINUX_SPEC): Define.
19574 * config/rs6000/linuxspe.h (TARGET_DEFAULT):
19575 Preserve MASK_LITTLE_ENDIAN.
19576 * config/rs6000/default64.h (TARGET_DEFAULT): Likewise.
19577 * config/rs6000/linuxaltivec.h (TARGET_DEFAULT): Likewise.
19578 * config/rs6000/linux64.h (OPTION_LITTLE_ENDIAN): Don't zero.
19579 (LINK_OS_LINUX_EMUL32, LINK_OS_LINUX_EMUL64): Define.
19580 (LINK_OS_LINUX_SPEC32, LINK_OS_LINUX_SPEC64): Use above.
19581 * config/rs6000/rs6000.c (output_toc): Don't use .tc for TARGET_ELF.
19582 Correct fp word order for little-endian. Don't shift toc entries
19583 smaller than a word for little-endian.
19584 * config/rs6000/rs6000.md (bswaphi2, bswapsi2 split): Comment.
19585 (bswapdi2 splits): Correct low-part subreg for little-endian.
19586 Remove wrong BYTES_BIG_ENDIAN tests, and rename vars to remove
19587 low/high where such is correct only for be.
19588 * config/rs6000/sysv4.h (SUBTARGET_OVERRIDE_OPTIONS): Allow
19589 little-endian for -mcall-aixdesc.
19590
87f73374
AM
195912013-04-25 Alan Modra <amodra@gmail.com>
19592
19593 * config/rs6000/rs6000.c (rs6000_secondary_reload_inner): Use
19594 replace_equiv_address_nv.
19595
cabf91cd
AM
195962013-04-25 Alan Modra <amodra@gmail.com>
19597
19598 * config/rs6000/rs6000.c (rs6000_emit_set_long_const): Tidy.
19599
0ae24cc8
VM
196002013-04-24 Vladimir Makarov <vmakarov@redhat.com>
19601
19602 Revert:
19603 2013-04-24 Vladimir Makarov <vmakarov@redhat.com>
19604 * rtl.h (struct rtx_def): ...
cabf91cd 19605
77bce07c
VM
196062013-04-24 Vladimir Makarov <vmakarov@redhat.com>
19607
19608 PR rtl-optimizations/57046
19609 * lra-constraints (split_reg): Set up lra_risky_transformations_p
19610 for multi-reg splits.
19611
0db63e7f
L
196122013-04-24 H.J. Lu <hongjiu.lu@intel.com>
19613
19614 * config/i386/x86-64.h (ASM_SPEC): Support -mx32.
19615
3362b6b6
SA
196162013-04-24 Sterling Augustine <saugustine@google.com>
19617
19618 * dwarf2out.c (skeleton_debug_str_hash, add_skeleton_AT_string)
19619 (comp_dir_string, debug_str_dwo_section): New.
19620 (DEBUG_STR_DWO_SECTION): Rename to ...
19621 (DEBUG_DWO_STR_SECTION): ... this.
19622 (DEBUG_NORM_STR_SECTION): Delete.
19623 (DEBUG_STR_SECTION, DEBUG_STR_SECTION_FLAGS): Edit definitions.
19624 (DEBUG_STR_DWO_SECTION_FLAGS): New.
19625 (find_AT_string): Move most logic to ...
19626 (find_AT_string_in_table): ... here. New.
19627 (add_top_level_skeleton_die_attrs): Call comp_dir_string and
19628 add_skeleton_AT_string. Delete logic.
19629 (output_skeleton_debug_sections): Remove call to
19630 add_top_level_skeleton_die_attrs.
19631 (add_comp_dir_attribute): Move logic to comp_dir_string.
19632 (dwarf2out_init): Initialize debug_str_dwo_section.
19633 (output_indirect_string): Call find_string_form.
19634 (output_indirect_strings): Rewrite.
19635 (prune_unused_types): Empty skeleton_debug_str_hash.
19636 Call get_skeleton_type_unit and add_top_level_skeleton_die_attrs.
19637 (dwarf2out_finish): Call output_indirect_strings.
19638
e93e18e9
PC
196392013-04-24 Paolo Carlini <paolo.carlini@oracle.com>
19640
19641 * doc/cpp.texi: Remove __GXX_EXPERIMENTAL_CXX1Y__.
19642
f6ce35ac
VM
196432013-04-24 Vladimir Makarov <vmakarov@redhat.com>
19644
cabf91cd 19645 * rtl.h (struct rtx_def): Add comment for field jump.
f6ce35ac
VM
19646 (LRA_SUBREG_P): New macro.
19647 * recog.c (register_operand): Check LRA_SUBREG_P.
19648 * lra.c (lra): Add note at the end of RTL code. Align non-empty
19649 stack frame.
19650 * lra-spills.c (lra_spill): Align stack after spilling pseudos.
19651 (lra_final_code_change): Skip subreg change for operators.
19652 * lra-eliminations.c (eliminate_regs_in_insn): Make return earlier
19653 if there are no operand changes.
19654 * lra-constraints.c (curr_insn_set): New.
19655 (match_reload): Set LRA_SUBREG_P.
19656 (emit_spill_move): Ditto.
19657 (check_and_process_move): Use curr_insn_set. Process only single
19658 set insns. Don't initialize sec_mem_p and change_p.
19659 (simplify_operand_subreg): Use LRA_SUBREG_P.
19660 (reg_in_class_p): New function.
19661 (process_alt_operands): Use it. Use #if HAVE_ATTR_enabled instead
19662 of #ifdef. Add code to remove cycling.
19663 (process_address): Check EXTRA_CONSTRAINT_STR. Process even if
19664 non-null disp. Reload inner instead of disp when base and index
19665 are null. Try to put lo_sum into register.
19666 (EBB_PROBABILITY_CUTOFF): Redefine probability in percents.
cabf91cd 19667 (check_and_process_move): Move code for move cost check to
f6ce35ac 19668 simple_move_p. Remove equiv_substitution.
cabf91cd 19669 (simple_move_p): New function.
f6ce35ac
VM
19670 (curr_insn_transform): Initialize sec_mem_p and change_p. Set up
19671 curr_insn_set. Call check_and_process_move only for single set
19672 insns. Use the new function. Move call of check_and_process_move
19673 after operand equiv substitution and address process.
19674
38047d90
JG
196752013-04-24 James Greenhalgh <james.greenhalgh@arm.com>
19676
19677 * config/aarch64/arm_neon.h (vld1<q>_lane*): Fix constraints.
19678 (vld1<q>_dup_<sufp><8, 16, 32, 64>): Likewise.
19679 (vld1<q>_<sufp><8, 16, 32, 64>): Likewise.
19680
13f39b2e
PC
196812013-04-24 Paolo Carlini <paolo.carlini@oracle.com>
19682
19683 * doc/cpp.texi: Document __GXX_EXPERIMENTAL_CXX1Y__.
19684
97c116dc
MP
196852013-04-24 Marek Polacek <polacek@redhat.com>
19686
19687 * tree-scalar-evolution.h (analyze_scalar_evolution): Remove.
19688 * tree-scalar-evolution.c (get_exit_conditions_rec): Likewise.
19689 (select_loops_exit_conditions): Likewise.
19690 (number_of_iterations_for_all_loops): Likewise.
19691 (analyze_scalar_evolution_for_all_loop_phi_nodes): Likewise.
19692 (scev_analysis): Likewise.
19693
83082391 196942013-04-02 Catherine Moore <clm@codesourcery.com>
12211b99 19695 Chao-ying Fu <fu@mips.com>
83082391 19696
cabf91cd
AM
19697 * config/mips/micromips.md (jraddiusp): New pattern.
19698 * config/mips/mips.c (mips_expand_epilogue): Use the JRADDIUSP
19699 instruction if possible.
83082391 19700
19e34aa2
AM
197012013-04-24 Alan Modra <amodra@gmail.com>
19702
19703 * config/rs6000/driver-rs6000.c (elf_dcachebsize): Fix comment pasto.
19704
fdb6603c
JB
197052013-04-24 Julian Brown <julian@codesourcery.com>
19706 Chung-Lin Tang <cltang@codesourcery.com>
19707
19708 * dwarf2out.c (gen_enumeration_type_die): Fix HOST_BITS_PER_WIDE_INT
df93505e
UB
19709 dependency behavior in enumeration type DIE generation. Add TODO note
19710 to comments about future DW_FORM_sdata/udata re-work of related code.
fdb6603c 19711
bf190e8d
LC
197122013-04-23 Lawrence Crowl <crowl@google.com>
19713
19714 * Makefile.in: Update as needed below.
19715
19716 * hash-table.h (class hash_table):
19717 Correct many methods with parameter types compare_type to the correct
19718 value_type. (Correct code was unlikely to notice the change.)
19719 (hash_table::elements_with_deleted) New.
19720 (class hashtable::iterator): New.
19721 (hashtable::begin()): New.
19722 (hashtable::end()): New.
19723 (FOR_EACH_HASH_TABLE_ELEMENT): New.
19724
19725 * statistics.c (statistics_hashes):
19726 Change type to hash_table. Update dependent calls and types.
19727
19728 * tree-into-ssa.c (var_infos):
19729 Change type to hash_table. Update dependent calls and types.
19730
19731 * tree-ssa-coalesce.c (struct coalesce_list_d.list):
19732 Change type to hash_table. Update dependent calls and types.
19733
19734 * tree-ssa-loop-im.c (struct mem_ref.refs):
19735 Change type to hash_table. Update dependent calls and types.
19736
19737 * tree-ssa-reassoc.c (undistribute_ops_list::ctable):
19738 Change type to hash_table. Update dependent calls and types.
19739
19740 * tree-ssa-sccvn.c (vn_tables_s::nary):
19741 Change type to hash_table. Update dependent calls and types.
19742 (vn_tables_s::phis): Likewise.
19743 (vn_tables_s::references): Likewise.
19744
19745 * tree-ssa-sccvn.h (vn_nary_op_eq): Update parameter and return types.
19746 (vn_reference_eq): Update parameter and return types.
19747
19748 * tree-ssa-structalias.c (pointer_equiv_class_table):
19749 Change type to hash_table. Update dependent calls and types.
19750 (location_equiv_class_table): Likewise.
19751
19752 * tree-vect-data-refs.c: Consequential changes for making
19753 peeling a hash_table.
19754
19755 * tree-vect-loop.c (new_loop_vec_info): Dependent hash_table update.
19756 (destroy_loop_vec_info): Dependent hash_table update.
19757
19758 * tree-vectorizer.h (peeling_htab):
19759 Change type to hash_table. Update dependent calls and types.
19760
d70a81dd
SC
197612013-04-23 Shiva Chen <shiva0217@gmail.com>
19762
cabf91cd
AM
19763 * lra-assigns.c (find_hard_regno_for): Use lra_reg_val_equal_p
19764 to check the register content is equal or not.
19765 * lra-constraints.c (match_reload): Use lra_assign_reg_val
19766 to assign register content record.
19767 * lra-eliminations.c (update_reg_eliminate): Use
d70a81dd 19768 lra_update_reg_val_offset to update register content offset.
cabf91cd
AM
19769 * lra-int.h (struct lra_reg): Add offset member.
19770 (lra_reg_val_equal_p): New static inline function.
19771 (lra_update_reg_val_offset): New static inline function.
19772 (lra_assign_reg_val): New static inline function.
19773 * lra.c (lra_create_new_reg): Use lra_assign_reg_val
19774 to assign register content record.
19775 (initialize_lra_reg_info_element): Initial offset to zero.
d70a81dd 19776
b894a1f3
CM
197772013-04-23 Catherine Moore <clm@codesourcery.com>
19778
19779 * config/mips/mips.md (*movhi_internal, *movqi_internal): New
19780 operands. Record compression.
19781
ecd14de9
XDL
197822013-04-23 Xinliang David Li <davidxl@google.com>
19783
19784 * cfghhooks.c (dump_bb_for_graph): Support 'slim' graph dump.
19785
92e776e9
RB
197862013-04-23 Richard Biener <rguenther@suse.de>
19787
19788 PR middle-end/57036
19789 * tree-inline.c (copy_edges_for_bb): Add can_make_abnormal_goto
19790 parameter, only add abnormal goto edges from the copied body
19791 if the call could perform abnormal gotos.
19792 (copy_cfg_body): Adjust.
19793
a15ee567
SN
197942013-04-23 Sofiane Naci <sofiane.naci@arm.com>
19795
19796 * config/aarch64/aarch64.md (*mov<mode>_aarch64): Add simd attribute.
19797
08c52234
AS
197982013-04-23 Andreas Schwab <schwab@linux-m68k.org>
19799
19800 * coretypes.h (gimple_stmt_iterator): Add struct to make
19801 compatible with C.
19802
999c1171
RB
198032013-04-23 Richard Biener <rguenther@suse.de>
19804
19805 PR tree-optimization/57026
19806 * tree-vrp.c (simplify_conversion_using_ranges): Do not propagate
19807 from SSA names occuring in abnormal PHI nodes.
19808
53669259
AK
198092013-04-22 Andi Kleen <ak@linux.intel.com>
19810
19811 * lto/lto.c (print_lto_report_1): Fix LTO report names.
19812
1a0ad150
AK
198132013-04-22 Andi Kleen <ak@linux.intel.com>
19814
19815 * lto/lto.c (print_lto_report_1): Declare early.
19816 (read_cgraph_and_symbols): Call print_lto_report_1 early.
19817
057f8f20
AK
198182013-04-22 Andi Kleen <ak@linux.intel.com>
19819
19820 * common.opt (-flto-report-wpa): Add.
19821 * doc/invoke.texi (-flto-report-wpa): Add.
19822 * lto/lto.c (do_whole_program_analysis): Check for lto-report-wpa.
19823 (lto_main): dito.
19824
473b1e05
XDL
198252013-04-22 Xinliang David Li <davidxl@google.com>
19826
19827 * graph.c (draw_cfg_node_succ_edges): Add branch probility as label.
19828 * cfghhooks.c (dump_bb_for_graph): Dump profile count and frquency.
19829 * Makefile.in: New dependency
19830
cabf91cd 19831 David Daney <ddaney.cavm@gmail.com>
b1485a33 19832
cabf91cd
AM
19833 * configure.ac (gcc_cv_as_micromips_support): Use the
19834 --fatal-warnings option.
19835 * configure: Regenerate.
b1485a33 19836
829d0168
MP
198372013-04-22 Marek Polacek <polacek@redhat.com>
19838
19839 PR sanitizer/56990
19840 * tsan.c (instrument_expr): Don't instrument expression
19841 in case its size is zero.
19842
6d9b7208
UB
198432013-04-22 Uros Bizjak <ubizjak@gmail.com>
19844
19845 PR target/57032
19846 Revert:
19847 2013-03-17 Uros Bizjak <ubizjak@gmail.com>
19848
19849 * config/alpha/alpha.c (TARGET_LRA_P): New define.
19850
ea679d55
JG
198512013-04-22 James Greenhalgh <james.greenhalgh@arm.com>
19852
19853 * coretypes.h (gimple_stmt_iterator_d): Forward declare.
19854 (gimple_stmt_iterator): New typedef.
19855 * gimple.h (gimple_stmt_iterator): Rename to...
19856 (gimple_stmt_iterator_d): ... This.
19857 * doc/tm.texi.in (TARGET_FOLD_BUILTIN): Detail restriction that
19858 trees be valid for GIMPLE and GENERIC.
19859 (TARGET_GIMPLE_FOLD_BUILTIN): New.
19860 * gimple-fold.c (gimple_fold_call): Call target hook
19861 gimple_fold_builtin.
19862 * hooks.c (hook_bool_gsiptr_false): New.
19863 * hooks.h (hook_bool_gsiptr_false): New.
19864 * target.def (fold_stmt): New.
19865 * doc/tm.texi: Regenerate.
19866
88a581da
VM
198672013-04-22 Vladimir Makarov <vmakarov@redhat.com>
19868
19869 PR target/57018
19870 * lra-eliminations.c (mark_not_eliminable): Prevent elimination of
19871 a set sp if no stack realignment.
19872
92be22dc
NC
198732013-04-22 Nick Clifton <nickc@redhat.com>
19874
19875 * config.gcc (tilegx-linux): Extend extra_objs rather than
19876 overwriting it.
19877 (tilepro-linux): Likewise.
19878
0ddec79f
JG
198792013-04-22 James Greenhalgh <james.greenhalgh@arm.com>
19880
19881 * config/aarch64/aarch64-builtins.c
19882 (CF): Remove.
19883 (CF0, CF1, CF2, CF3, CF4, CF10): New.
19884 (VAR<1-12>): Add MAP parameter.
19885 (BUILTIN_*): Likewise.
19886 * config/aarch64/aarch64-simd-builtins.def: Set MAP parameter.
19887 * config/aarch64/aarch64-simd.md (aarch64_sshl_n<mode>): Remove.
19888 (aarch64_ushl_n<mode>): Likewise.
19889 (aarch64_sshr_n<mode>): Likewise.
19890 (aarch64_ushr_n<mode>): Likewise.
19891 (aarch64_<maxmin><mode>): Likewise.
19892 (aarch64_sqrt<mode>): Likewise.
19893 * config/aarch64/arm_neon.h (vshl<q>_n_*): Use new builtin names.
19894 (vshr<q>_n_*): Likewise.
19895
0050faf8
JG
198962013-04-22 James Greenhalgh <james.greenhalgh@arm.com>
19897
19898 * config/aarch64/aarch64-builtins.c
19899 (aarch64_simd_builtin_type_mode): Handle SF types.
19900 (sf_UP): Define.
19901 (BUILTIN_GPF): Define.
19902 (aarch64_init_simd_builtins): Handle SF types.
19903 * config/aarch64/aarch64-simd-builtins.def (frecpe): Add support.
19904 (frecps): Likewise.
19905 (frecpx): Likewise.
19906 * config/aarch64/aarch64-simd.md
19907 (simd_types): Update simd_frcp<esx> to simd_frecp<esx>.
19908 (aarch64_frecpe<mode>): New.
19909 (aarch64_frecps<mode>): Likewise.
19910 * config/aarch64/aarch64.md (unspec): Add UNSPEC_FRECP<ESX>.
19911 (v8type): Add frecp<esx>.
19912 (aarch64_frecp<FRECP:frecp_suffix><mode>): New.
19913 (aarch64_frecps<mode>): Likewise.
19914 * config/aarch64/iterators.md (FRECP): New.
19915 (frecp_suffix): Likewise.
19916 * config/aarch64/arm_neon.h
19917 (vrecp<esx><qsd>_<fd><32, 64>): Convert to using builtins.
19918
0fad3dbc 199192013-04-22 Christian Bruel <christian.bruel@st.com>
1818d01c
CB
19920
19921 PR target/56995
19922 * config/sh/sh.h (enum reg_class): Remove DF_HI_REGS.
19923 (REG_CLASS_NAMES): Idem.
19924 (REG_CLASS_CONTENTS): Idem.
19925 (REGCLASS_HAS_FP_REG): Idem.
19926 * config/sh/sh.c (sh_cannot_change_mode_class): Idem.
19927 (sh_conditional_register_usage): Idem.
19928
3e8a33f9
JL
199292013-04-21 Jeff Law <law@redhat.com>
19930
19931 * tree-ssa-forwprop.c (simplify_conversion_from_bitmask): New function.
19932 (ssa_forward_propagate_and_combine): Use it.
19933
f38e1b0a
VM
199342013-04-19 Vladimir Makarov <vmakarov@redhat.com>
19935
19936 * lra.c: Update the flow chart diagram.
19937
682303da
VM
199382013-04-19 Vladimir Makarov <vmakarov@redhat.com>
19939
19940 PR rtl-optimization/56847
19941 * lra-constraints.c (process_alt_operands): Discourage alternative
19942 with non-matche doffsettable memory constraint fro memory with
19943 known offset.
19944
f6b64c35
RB
199452013-04-19 Richard Biener <rguenther@suse.de>
19946
19947 PR tree-optimization/56982
19948 * builtins.def (BUILT_IN_LONGJMP): longjmp is not a leaf
19949 function.
19950 * gimplify.c (gimplify_call_expr): Notice special calls.
19951 (gimplify_modify_expr): Likewise.
19952 * tree-cfg.c (make_abnormal_goto_edges): Handle setjmp-like
19953 abnormal control flow receivers.
19954 (call_can_make_abnormal_goto): Handle cfun->calls_setjmp
19955 in the same way as cfun->has_nonlocal_labels.
19956 (gimple_purge_dead_abnormal_call_edges): Likewise.
19957 (stmt_starts_bb_p): Make setjmp-like abnormal control flow
19958 receivers start a basic-block.
19959
01d8bf07
RB
199602013-04-19 Richard Biener <rguenther@suse.de>
19961
19962 * tree-vectorizer.h (struct _slp_instance): Move load_permutation
19963 member ...
19964 (struct _slp_tree): ... here. Make it a vector of unsigned ints.
19965 (SLP_INSTANCE_LOAD_PERMUTATION): Remove.
19966 (SLP_TREE_LOAD_PERMUTATION): Add.
19967 (vect_transform_slp_perm_load): Adjust prototype.
19968 * tree-vect-slp.c (vect_free_slp_tree): Adjust.
19969 (vect_free_slp_instance): Likewise.
19970 (vect_create_new_slp_node): Likewise.
19971 (vect_supported_slp_permutation_p): Remove.
19972 (vect_slp_rearrange_stmts): Adjust.
19973 (vect_supported_load_permutation_p): Likewise. Inline
19974 vect_supported_slp_permutation_p here.
19975 (vect_analyze_slp_instance): Compute load permutations per
19976 slp node instead of per instance.
19977 (vect_get_slp_defs): Adjust.
19978 (vect_transform_slp_perm_load): Likewise.
19979 (vect_schedule_slp_instance): Remove redundant code.
19980 (vect_schedule_slp): Remove hack for PR56270, add it ...
19981 * tree-vect-stmts.c (vectorizable_load): ... here, do not
19982 CSE loads for SLP. Adjust.
19983
ede22fc3
GY
199842013-04-19 Greta Yorsh <Greta.Yorsh@arm.com>
19985
19986 * config/arm/arm.c (load_multiple_sequence, ldm_stm_operation_p): Fix
19987 spelling in two comments.
19988
67bc84fb
GY
199892013-04-19 Greta Yorsh <Greta.Yorsh@arm.com>
19990
19991 PR target/56797
19992 * config/arm/arm.c (load_multiple_sequence): Require SP
19993 as base register for loads if SP is in the register list.
19994
e248d83f
MJ
199952013-04-19 Martin Jambor <mjambor@suse.cz>
19996
19997 PR tree-optimization/56718
19998 * ipa-cp.c (ipa_value_from_known_type_jfunc): Moved...
19999 * ipa-prop.c (ipa_binfo_from_known_type_jfunc): ...here, renamed
20000 and made public. Adjusted all callers.
20001 (ipa_intraprocedural_devirtualization): New function.
20002 * ipa-prop.h (ipa_binfo_from_known_type_jfunc): Declare.
20003 (ipa_intraprocedural_devirtualization): Likewise.
20004 * Makefile.in (tree-ssa-pre.o): Add ipa-prop.h to dependencies.
20005
4891e8f8
RB
200062013-04-19 Richard Biener <rguenther@suse.de>
20007
20008 PR tree-optimization/57000
20009 * tree-ssa-reassoc.c (pass_reassoc): Add TODO_update_ssa_only_virtuals.
20010
dad89f7c
TG
200112013-04-19 Terry Guo <terry.guo@arm.com>
20012
20013 * config/arm/cortex-m4-fpu.md (cortex_m4_v): Delete cpu unit.
20014 Replace with ...
20015 (cortex_m4_v_a, cortex_m4_v_b): ... new cpu units.
20016 (cortex_m4_v, cortex_m4_exa_va, cortex_m4_exb_vb): New reservations.
20017 (cortex_m4_fmacs): Use new reservations.
20018 (cortex_m4_f_load, cortex_m4_f_store): Likewise.
20019
72ea0d47
VM
200202013-04-18 Vladimir Makarov <vmakarov@redhat.com>
20021
f1e6512c 20022 PR rtl-optimization/56999
72ea0d47
VM
20023 * lra-coalesce.c (coalescable_pseudo_p): Remove 2nd parameter and
20024 related code.
20025 (lra_coalesce): Remove split_origin_bitmap and related code.
20026 * lra.c (lra): Coalesce after undoing inheritance. Recreate live
20027 ranges if necessary.
20028
780a5b71
UB
200292013-04-18 Uros Bizjak <ubizjak@gmail.com>
20030
20031 * config/i386/i386.c (x86_64_ms_sysv_extra_clobbered_registers):
20032 New array.
20033 (ix86_expand_call): Remove clobbered_registers array and use
20034 x86_64_ms_sysv_extra_clobbered_registers instead.
20035 * config/i386/i386.h (x86_64_ms_sysv_extra_clobbered_registers):
20036 Declare here.
20037 * config/i386/predicates.md (call_rex64_ms_sysv_operation): New
20038 predicate.
20039 * config/i386/i386.md (*call_rex64_ms_sysv): Use
20040 call_rex64_ms_sysv_operation predicate. Remove explicit clobbers.
20041 (*call_value_rex64_ms_sysv): Ditto.
20042
6f5a366a
CC
200432013-04-18 Cary Coutant <ccoutant@google.com>
20044
20045 * dwarf2out.c (output_pubnames): Check die_perennial_p of
20046 parent instead of die_mark.
20047
475b8f37
DN
200482013-04-18 Diego Novillo <dnovillo@google.com>
20049
20050 * gimple.c (create_gimple_tmp): New.
20051 (get_expr_type): New.
20052 (build_assign): New.
20053 (build_type_cast): New.
20054 * gimple.h (enum ssa_mode): Define.
20055 (gimple_seq_set_location): New.
20056 * asan.c (build_check_stmt): Change some gimple_build_* calls
780a5b71 20057 to use build_assign and build_type_cast.
475b8f37 20058
08940f33
RB
200592013-04-18 Richard Biener <rguenther@suse.de>
20060
20061 * tree-vect-data-refs.c (vect_analyze_group_access): Properly
20062 handle negative step. Remove redundant checks.
20063 (vect_create_data_ref_ptr): Avoid ICEs with non-constant steps.
20064 * tree-vect-stmts.c (vectorizable_load): Instead of asserting
20065 for negative step and grouped loads fail to vectorize.
20066
0e0f87d4
SB
200672013-04-18 Steven Bosscher <steven@gcc.gnu.org>
20068
20069 * emit-rtl.c (reset_insn_used_flags): New function.
20070 (reset_all_used_flags): Use it.
20071 (verify_insn_sharing): New function.
20072 (verify_rtl_sharing): Fix verification for SEQUENCEs.
20073
4c445590
JJ
200742013-04-18 Jakub Jelinek <jakub@redhat.com>
20075
20076 PR tree-optimization/56984
20077 * tree-vrp.c (register_edge_assert_for_2): For (x >> M) < N
20078 and (x >> M) >= N don't register any assertion if N << M is the
20079 minimum value.
20080
6873ecab
SB
200812013-04-18 Steven Bosscher <steven@gcc.gnu.org>
20082
20083 * lower-subreg.c (resolve_simple_move): If called self-recursive,
20084 do not delete_insn insns that have not yet been emitted, only
20085 unlink them with remove_insn.
20086 * df-scan.c (df_insn_delete): Revert r197492.
20087
3ccb989e
SB
200882013-04-17 Steven Bosscher <steven@gcc.gnu.org>
20089
20090 * emit-rtl.c (link_insn_into_chain): Handle chaining of SEQUENCEs.
20091 * reorg.c (emit_delay_sequence): Simplify with emit-rtl API.
20092
200932013-04-17 Greta Yorsh <Greta.Yorsh@arm.com>
0480fff0
GY
20094
20095 * config/arm/arm.md (movsicc_insn): Convert define_insn into
20096 define_insn_and_split.
20097 (and_scc,ior_scc,negscc): Likewise.
20098 (cmpsi2_addneg, subsi3_compare): Convert to named patterns.
20099
3ccb989e 201002013-04-17 Greta Yorsh <Greta.Yorsh@arm.com>
367075fe
GY
20101
20102 * config/arm/arm.c (use_return_insn): Return 0 for targets that
20103 can benefit from using a sequence of LDRD instructions in epilogue
20104 instead of a single LDM instruction.
20105
6d10a203
MLI
201062013-04-17 Manuel López-Ibáñez <manu@gcc.gnu.org>
20107
20108 PR 45688
20109 * doc/extend.texi: Fix typo.
20110
6983e6b5
RB
201112013-04-17 Richard Biener <rguenther@suse.de>
20112
20113 * tree-vect-slp.c (vect_build_slp_tree_1): Split out from ...
20114 (vect_build_slp_tree): ... here.
20115 (vect_build_slp_tree_1): Compute which stmts of the SLP group
20116 match. Remove special-casing of mismatched complex loads.
20117 (vect_build_slp_tree): Based on the result from vect_build_slp_tree_1
20118 re-try the match with swapped commutative operands.
20119 (vect_supported_load_permutation_p): Remove special-casing of
20120 mismatched complex loads.
20121 (vect_analyze_slp_instance): Adjust.
20122
ef23e6a2
RB
201232013-04-17 Richard Biener <rguenther@suse.de>
20124
20125 PR rtl-optimization/56921
20126 * cfgloop.h (struct loop): Add simple_loop_desc member.
20127 (struct niter_desc): Mark with GTY(()).
20128 (simple_loop_desc): Do not use aux field but simple_loop_desc.
20129 * loop-iv.c (get_simple_loop_desc): Likewise.
20130 (free_simple_loop_desc): Likewise.
20131
20132 Revert
20133 2013-04-16 Richard Biener <rguenther@suse.de>
20134
20135 PR rtl-optimization/56921
20136 * loop-init.c (pass_rtl_move_loop_invariants): Add
20137 TODO_do_not_ggc_collect to todo_flags_finish.
20138 (pass_rtl_unswitch): Same.
20139 (pass_rtl_unroll_and_peel_loops): Same.
20140 (pass_rtl_doloop): Same.
20141
fc6f94f5
EB
201422013-04-17 Eric Botcazou <ebotcazou@adacore.com>
20143
20144 * tree-ssa-alias.c (nonoverlapping_component_refs_of_decl_p): New.
20145 (decl_refs_may_alias_p): Add REF1 and REF2 parameters.
20146 Use nonoverlapping_component_refs_of_decl_p to disambiguate component
20147 references.
20148 (refs_may_alias_p_1): Adjust call to decl_refs_may_alias_p.
20149 * tree-streamer.c (record_common_node): Adjust reference in comment.
20150
10a88311
TG
201512013-04-17 Terry Guo <terry.guo@arm.com>
20152
20153 * config/arm/cortex-m4.md: Add a new bypass.
20154
6d9b7208 201552013-04-16 Naveen H.S <Naveen.Hurugalawadi@caviumnetworks.com>
f3d096b0
N
20156
20157 * config/aarch64/aarch64.md (*adds_<optab><mode>_multp2):
20158 New pattern.
20159 (*subs_<optab><mode>_multp2): New pattern.
20160 (*adds_<optab><ALLX:mode>_<GPI:mode>): New pattern.
20161 (*subs_<optab><ALLX:mode>_<GPI:mode>): New pattern.
20162
6d9b7208 201632013-04-16 Naveen H.S <Naveen.Hurugalawadi@caviumnetworks.com>
860ad33f
N
20164
20165 * config/aarch64/aarch64.md (*adds_mul_imm_<mode>): New pattern.
20166 (*subs_mul_imm_<mode>): New pattern.
20167
18a6701e
DE
201682013-04-16 David Edelsohn <dje.gcc@gmail.com>
20169
20170 PR target/56948
20171 * config/rs6000/vsx.md (vsx_mov<mode>): Add j->r alternative.
20172 (vsx_movti_64bit): Change j->wa to O->wa. Add n->r alternative.
20173 (vsx_movti_32bit): Change j->wa to O->wa.
20174
07c37b2f
RB
201752013-04-16 Richard Biener <rguenther@suse.de>
20176
20177 PR rtl-optimization/56921
20178 * loop-init.c (pass_rtl_move_loop_invariants): Add
20179 TODO_do_not_ggc_collect to todo_flags_finish.
20180 (pass_rtl_unswitch): Same.
20181 (pass_rtl_unroll_and_peel_loops): Same.
20182 (pass_rtl_doloop): Same.
20183
0e0f87d4 201842013-04-16 Greta Yorsh <Greta.Yorsh@arm.com>
974ba22d
GY
20185
20186 * config/arm/arm.c (emit_multi_reg_push): New declaration
20187 for an existing function.
20188 (arm_emit_strd_push): New function.
20189 (arm_expand_prologue): Used here.
20190 (arm_emit_ldrd_pop): New function.
20191 (arm_expand_epilogue): Used here.
20192 (arm_get_frame_offsets): Update condition.
20193 (arm_emit_multi_reg_pop): Add a special case for load of a single
20194 register with writeback.
20195
5e8e2af4
UB
201962013-04-16 Uros Bizjak <ubizjak@gmail.com>
20197
20198 * doc/invoke.texi (i386 Option): Reword -mstack-protector-guard
20199 description.
20200
9fd9ccf7
RB
202012013-04-16 Richard Biener <rguenther@suse.de>
20202
20203 PR tree-optimization/56756
20204 * tree-ssa-loop-im.c (struct first_mem_ref_loc_1): New functor.
20205 (first_mem_ref_loc): New.
20206 (execute_sm): Place the load temporarily before a previous
20207 access instead of in the latch edge to ensure its SSA dependencies
20208 are defined at points dominating the load.
20209
96fba521
SB
202102013-04-16 Steven Bosscher <steven@gcc.gnu.org>
20211
4c8af858
SB
20212 * cfgrtl.c (cfg_layout_merge_blocks): Revert r184005, implement
20213 correct fix by moving header and footer insn to the footer of
20214 the merged basic block. Clear BB_END of the merged-away block.
20215
96fba521
SB
20216 PR middle-end/43631
20217 * emit-rtl.c (make_note_raw): New function.
20218 (link_insn_into_chain): New static inline function.
20219 (add_insn): Use it.
20220 (add_insn_before, add_insn_after): Factor insn chain linking code...
20221 (add_insn_before_nobb, add_insn_after_nobb): ...here, new functions
20222 using link_insn_into_chain.
20223 (note_outside_basic_block_p): New helper function for emit_note_after
20224 and emit_note_before.
20225 (emit_note_after): Use nobb variant of add_insn_after if the note
20226 should not be contained in a basic block.
20227 (emit_note_before): Use nobb variant of add_insn_before if the note
20228 should not be contained in a basic block.
20229 (emit_note_copy): Use make_note_raw.
20230 (emit_note): Likewise.
20231 * bb-reorder.c (insert_section_boundary_note): Remove hack to set
20232 BLOCK_FOR_INSN to NULL manually for NOTE_INSN_SWITCH_TEXT_SECTIONS.
20233 * jump.c (cleanup_barriers): Use reorder_insns_nobb to avoid making
20234 the moved barrier the tail of the basic block it follows.
20235 * var-tracking.c (pass_variable_tracking): Add TODO_verify_flow.
20236
7b8265ba
JJ
202372013-04-15 Jakub Jelinek <jakub@redhat.com>
20238
20239 PR tree-optimization/56962
20240 * gimple-ssa-strength-reduction.c (record_increment): Only set
20241 initializer if gimple_assign_rhs_code is {,POINTER_}PLUS_EXPR and
20242 either rhs1 or rhs2 is equal to c->base_expr.
20243
5185d248
RB
202442013-04-15 Richard Biener <rguenther@suse.de>
20245
20246 PR tree-optimization/56933
20247 * tree-vectorizer.h (struct _stmt_vec_info): Remove read_write_dep
20248 member.
20249 (GROUP_READ_WRITE_DEPENDENCE): Remove.
20250 (STMT_VINFO_GROUP_READ_WRITE_DEPENDENCE): Likewise.
20251 * tree-vect-data-refs.c (vect_analyze_group_access): Move
20252 dependence check ...
20253 vect_analyze_data_ref_dependence (vect_analyze_data_ref_dependence):
20254 ... here.
20255 * tree-vect-stmts.c (new_stmt_vec_info): Do not initialize
20256 GROUP_READ_WRITE_DEPENDENCE.
20257
a24243a0
AK
202582013-04-15 Andreas Krebbel <Andreas.Krebbel@de.ibm.com>
20259
20260 * emit-rtl.c (reset_all_used_flags): New function.
20261 (verify_rtl_sharing): Call reset_all_used_flags before and after
20262 performing the checks.
20263
1c50eada
KT
202642013-04-15 Kyrylo Tkachov <kyrylo.tkachov@arm.com>
20265
20266 * config/arm/arm.c (const_ok_for_dimode_op): Handle AND case.
20267 * config/arm/arm.md (*anddi3_insn): Change to insn_and_split.
20268 * config/arm/constraints.md (De): New constraint.
20269 * config/arm/neon.md (anddi3_neon): Delete.
20270 (neon_vand<mode>): Expand to standard anddi3 pattern.
20271 * config/arm/predicates.md (imm_for_neon_inv_logic_operand):
20272 Move earlier in the file.
20273 (neon_inv_logic_op2): Likewise.
20274 (arm_anddi_operand_neon): New predicate.
20275
e927b6ad
RO
202762013-04-15 Rainer Orth <ro@CeBiTec.Uni-Bielefeld.DE>
20277
20278 * configure.ac (gcc_cv_ld_as_needed): Set
20279 gcc_cv_ld_as_needed_option, gcc_cv_no_as_needed_option.
20280 Use -z ignore, -z record on *-*-solaris2*.
20281 (HAVE_LD_AS_NEEDED): Update comment.
20282 (LD_AS_NEEDED_OPTION, LD_NO_AS_NEEDED_OPTION): Define.
20283 * configure: Regenerate.
20284 * config.in: Regenerate.
20285 * gcc.c (init_gcc_specs) [USE_LD_AS_NEEDED]: Use
20286 LD_AS_NEEDED_OPTION, LD_NO_AS_NEEDED_OPTION.
20287 * config/sol2.h [HAVE_LD_AS_NEEDED] (USE_LD_AS_NEEDED): Define.
20288 * doc/tm.texi.in (USE_LD_AS_NEEDED): Allow for --as-needed
20289 equivalents. Fix markup.
20290 * doc/tm.texi: Regenerate.
20291
e0ea8797
AH
202922013-04-15 Andrew Hsieh <andrewhsieh.google.com>
20293
20294 * config/i386/i386.opt: New option mstack-protector-guard=.
20295 * config/i386/i386-opts.h: Add enum stack_protector_guard.
20296 * config/i386/i386.h: Define TARGET_SSP_GLOBAL_GUARD and
20297 TARGET_SSP_TLS_GUARD.
20298 * config/i386/i386.c (ix86_option_override_internal): Set
20299 ix86_stack_protector_guard.
20300 * config/i386/i386.md (stack_protect_set): Enable for
20301 TARGET_SSP_TLS_GUARD only.
20302 (stack_protect_set_<mode>): Ditto.
20303 (stack_protect_test): Ditto.
20304 (stack_protect_test_<mode>): Ditto.
20305 * doc/invoke.texi (i386 Option): Document.
20306
811b72f9
EB
203072013-04-15 Eric Botcazou <ebotcazou@adacore.com>
20308
20309 PR target/56890
20310 * config/sparc/sparc.c (enum sparc_mode_class): Add H_MODE value.
20311 (S_MODES): Set H_MODE bit.
20312 (SF_MODES): Set only S_MODE and SF_MODE bits.
20313 (DF_MODES): Set SF_MODES and only D_MODE and DF_MODE bits.
20314 (sparc_init_modes) <MODE_INT>: Set H_MODE bit for sub-word modes.
20315 <MODE_VECTOR_INT>: Do not set SF_MODE for sub-word modes.
20316 <MODE_FLOAT>: Likewise.
20317
5529fdd6
JY
203182013-04-15 Joey Ye <joey.ye@arm.com>
20319
20320 * config/arm/arm.c (thumb_far_jump_used_p): Fix typo in comments.
20321
517b1da2
JY
203222013-04-15 Joey Ye <joey.ye@arm.com>
20323
20324 * config/arm/arm.c (thumb1_final_prescan_insn): Assert lr save
20325 for real far jump.
20326 (thumb_far_jump_used_p): Count instruction size and set
20327 far_jump_used.
20328
01007ae0
EB
203292013-04-14 Eric Botcazou <ebotcazou@adacore.com>
20330
20331 * reorg.c (fill_simple_delay_slots): Reindent block of code.
20332 * resource.c (mark_target_live_regs): Reformat conditional block.
20333
c46f6580
SB
203342013-04-13 Steven Bosscher <steven@gcc.gnu.org>
20335
20336 * sched-deps.c (deps_analyze_insn): Do not check for EH_REGION insn
20337 notes, they are emitted only just before final.
20338 * sched-int.h: Include insn-attr.h before checking INSN_SCHEDULING.
20339
1f397f45
SB
203402013-04-13 Steven Bosscher <steven@gcc.gnu.org>
20341
20342 * emit-rtl.c (remove_insn): Do not call df_insn_delete here.
20343 * cfgrtl.c (delete_insn): Call it here instead.
20344 * lra-spills.c (lra_final_code_change): Use delete_insn.
20345 * haifa-sched.c (sched_remove_insn): Likewise.
20346 * sel-sched-ir.c (return_nop_to_pool): Clear INSN_DELETED_P for nops
20347 returning to the nop pool.
20348 (sel_remove_insn): Simplify the only_disconnect case via remove_insn,
20349 use delete_insn for definitive removal. Clear BLOCK_FOR_INSN.
20350
58a51369
SB
203512013-04-12 Steven Bosscher <steven@gcc.gnu.org>
20352
20353 * doc/tm.texi.in (LOOP_ALIGN): Remove loop note references.
20354 * doc/tm.texi: Regenerated.
20355
33159866
UB
203562013-04-12 Uros Bizjak <ubizjak@gmail.com>
20357
20358 * config/i386/i386.c (ix86_hard_regno_mode_ok): Use ANY_QI_REGNO_P in
20359 QImode checks.
20360
226e378f
SB
203612013-04-12 Steven Bosscher <steven@gcc.gnu.org>
20362
20363 * df-core.c (df_find_def): Compare register numbers.
20364 (df_find_use): Likewise.
20365
fafb9b18
VM
203662013-04-12 Vladimir Makarov <vmakarov@redhat.com>
20367
20368 PR target/56903
20369 * config/i386/i386.c (ix86_hard_regno_mode_ok): Add
20370 lra_in_progress for return.
20371
9a946fd6
GY
203722013-04-12 Greta Yorsh <Greta.Yorsh@arm.com>
20373
20374 * config/arm/arm.md (mov_scc,mov_negscc,mov_notscc): Convert
20375 define_insn into define_insn_and_split and emit movsicc patterns.
20376
41b83758
GY
203772013-04-12 Greta Yorsh <Greta.Yorsh@arm.com>
20378
20379 * config/arm/arm.c (gen_operands_ldrd_strd): Initialize "base".
20380
d3afd9aa
RB
203812013-04-12 Richard Biener <rguenther@suse.de>
20382
20383 * tree-pass.h (TODO_do_not_ggc_collect): New.
20384 * passes.c (execute_one_ipa_transform_pass): Honor
20385 TODO_do_not_ggc_collect.
20386 (execute_one_pass): Likewise.
20387
20388 Revert
20389 2013-04-10 Richard Biener <rguenther@suse.de>
20390
20391 * passes.c (init_optimization_passes): Remove reload pass.
20392 * ira.c (do_reload): Merge into ...
20393 (ira): ... this.
20394 (rest_of_handle_reload): Remove.
20395 (pass_reload): Likewise.
20396 * config/i386/i386.c (ix86_option_override): Refer to ira instead
20397 of reload for vzeroupper pass placement.
20398
06f9b387
JJ
203992013-04-12 Jakub Jelinek <jakub@redhat.com>
20400
20401 PR tree-optimization/56918
20402 PR tree-optimization/56920
20403 * fold-const.c (int_const_binop_1): Use op1.mul_with_sign (op2, ...)
20404 instead of op1 - op2. Pass 2 * TYPE_PRECISION (type) as second
20405 argument to rshift method. For 2 * HOST_BITS_PER_WIDE_INT precision
20406 use wide_mul_with_sign method.
20407
953094d2
RB
204082013-04-12 Richard Biener <rguenther@suse.de>
20409
20410 * gimple.c (is_gimple_constant): Vector CONSTRUCTORs should
20411 not be considered a gimple constant.
20412
bb506982
MG
204132013-04-12 Marc Glisse <marc.glisse@inria.fr>
20414
20415 * fold-const.c (const_binop): Handle vector shifts by a scalar.
20416 (fold_binary_loc): Call const_binop also for mixed vector-scalar
20417 operations.
20418
4b84d650
JJ
204192013-04-12 Manuel López-Ibáñez <manu@gcc.gnu.org>
20420 Jakub Jelinek <jakub@redhat.com>
20421
20422 * opts.c: Include diagnostic-color.h.
20423 (common_handle_option): Handle OPT_fdiagnostics_color_.
20424 * Makefile.in (OBJS-libcommon): Add diagnostic-color.o.
20425 (diagnostic.o, opts.o, pretty-print.o): Depend on diagnostic-color.h.
20426 (diagnostic-color.o): New.
20427 * common.opt (fdiagnostics-color, fdiagnostics-color=): New options.
20428 (diagnostic_color_rule): New enum.
20429 * dwarf2out.c (gen_producer_string): Don't print -fdiagnostics-color*.
20430 * langhooks.c (lhd_print_error_function): Add %r "locus" and %R around
20431 the location string.
20432 * diagnostic.def: Add 3rd argument to DEFINE_DIAGNOSTIC_KIND macros,
20433 either NULL, or color kind.
20434 * diagnostic-color.c: New file.
20435 * diagnostic-color.h: New file.
20436 * diagnostic-core.h (DEFINE_DIAGNOSTIC_KIND): Adjust macro for 3
20437 arguments.
20438 * doc/invoke.texi (-fdiagnostics-color): Document.
20439 * pretty-print.h (pp_show_color): Define.
20440 (struct pretty_print_info): Add show_color field.
20441 * diagnostic.c: Include diagnostic-color.h.
20442 (diagnostic_build_prefix): Adjust for 3 argument DEFINE_DIAGNOSTIC_KIND
20443 macros. Colorize error:, warning: etc. strings and also the location
20444 string.
20445 (diagnostic_show_locus): Colorize the caret line.
20446 * pretty-print.c: Include diagnostic-color.h.
20447 (pp_base_format): Handle %r and %R format specifiers. Colorize strings
20448 inside of %< %> quotes or quoted through q format modifier.
20449
067a1e71
AK
204502013-04-12 Andreas Krebbel <Andreas.Krebbel@de.ibm.com>
20451
33159866 20452 * ifcvt.c (end_ifcvt_sequence): Mark a and b for unsharing as well.
067a1e71 20453
33159866 204542013-04-11 Naveen H.S <Naveen.Hurugalawadi@caviumnetworks.com>
b056c910
N
20455
20456 * config/aarch64/aarch64.c (aarch64_select_cc_mode): Allow NEG
20457 code in CC_NZ mode.
20458 * config/aarch64/aarch64.md (*neg_<shift><mode>3_compare0): New
20459 pattern.
20460
7b55f98f
MP
204612013-04-11 Marek Polacek <polacek@redhat.com>
20462
20463 PR tree-optimization/48184
33159866 20464 * params.def (PARAM_ALIGN_THRESHOLD): Increase the minimum value to 1.
7b55f98f 20465
966b587e
EB
204662013-04-11 Eric Botcazou <ebotcazou@adacore.com>
20467
20468 * stor-layout.c (skip_simple_constant_arithmetic): Move to...
20469 * tree.c (skip_simple_constant_arithmetic): ...here and make public.
20470 (skip_simple_arithmetic): Tidy up.
20471 * tree.h (skip_simple_constant_arithmetic): Declare.
20472
33159866 204732013-04-11 Naveen H.S <Naveen.Hurugalawadi@caviumnetworks.com>
f8bf91ab
N
20474
20475 * config/aarch64/aarch64.h (REVERSIBLE_CC_MODE): Define.
20476
1d42c1ec
RB
204772013-04-11 Richard Biener <rguenther@suse.de>
20478
20479 * tree-vect-loop.c (get_initial_def_for_induction): Properly
20480 generate vector constants.
20481
4ba5ea11
RB
204822013-04-11 Richard Biener <rguenther@suse.de>
20483
20484 PR tree-optimization/56878
20485 * tree-flow.h (outermost_invariant_loop_for_expr): Declare.
20486 * tree-ssa-loop-ivopts.c (outermost_invariant_loop_for_expr):
20487 New function.
20488 * tree-vect-data-refs.c (vect_enhance_data_refs_alignment):
20489 Prefer to align the DR with the most invariant base address.
20490
f0defe58
SKS
204912013-04-11 Senthil Kumar Selvaraj <senthil_kumar.selvaraj@atmel.com>
20492
20493 * opts.c (common_handle_option): Fix formatting and add FALLTHRU
20494 comment.
20495
d07458be
JG
204962013-04-11 James Greenhalgh <james.greenhalgh@arm.com>
20497
20498 * config/aarch64/aarch64-simd.md (aarch64_vcond_internal): Fix
20499 floating-point vector comparisons against 0.
20500
146b8692
JJ
205012013-04-11 Jakub Jelinek <jakub@redhat.com>
20502
20503 PR tree-optimization/56899
20504 * fold-const.c (extract_muldiv_1): Apply distributive law
20505 only if TYPE_OVERFLOW_WRAPS (ctype).
20506
b8578ff7
BC
205072013-04-11 Bin Cheng <bin.cheng@arm.com>
20508
20509 PR target/56124
20510 * ira-costs.c (scan_one_insn): Check whether the source rtx of
20511 loading has side effect.
20512
0ea8a6f9
SB
205132013-04-10 Steven Bosscher <steven@gcc.gnu.org>
20514
20515 * config/sparc/sparc.c: Include tree-pass.h.
20516 (TARGET_MACHINE_DEPENDENT_REORG): Do not redefine.
20517 (sparc_reorg): Rename to sparc_do_work_around_errata. Move to
20518 head of file. Change return type. Split off gate function.
20519 (sparc_gate_work_around_errata): New function.
20520 (pass_work_around_errata): New pass definition.
20521 (insert_pass_work_around_errata) New pass insert definition to
20522 insert pass_work_around_errata just after delayed-branch scheduling.
20523 (sparc_option_override): Insert the pass.
20524 * config/sparc/t-sparc (sparc.o): Add TREE_PASS_H dependence.
20525
42e37616
DM
205262013-04-10 David S. Miller <davem@davemloft.net>
20527
89deeb3b
DM
20528 * config/sparc/sparc.h (ASM_CPU_SPEC): Pass -Av8 if -mcpu=supersparc
20529 or -mcpu=hypersparc.
20530
42e37616
DM
20531 * target.def (cstore_mode): New hook.
20532 * target.h: Include insn-codes.h
20533 * targhooks.c: Likewise.
20534 (default_cstore_mode): New function.
20535 * targhooks.h: Declare it.
20536 * doc/tm.texi.in: New hook slot for TARGET_CSTORE_MODE.
20537 * doc/tm.texi: Rebuild.
20538 * expmed.c (emit_cstore): Obtain cstore boolean result mode using
20539 target hook, rather than inspecting the insn_data.
20540 * config/sparc/sparc.c (sparc_cstore_mode): New function.
20541 (TARGET_CSTORE_MODE): Redefine.
20542 (emit_scc_insn): When TARGET_ARCH64, emit new 64-bit boolean
20543 result patterns.
20544 * config/sparc/predicates.md (cstore_result_operand): New special
20545 predicate.
20546 * config/sparc/sparc.md (cstoresi4, cstoredi4, cstore<F:mode>4):
20547 Use it for operand 0.
20548 (*seqsi_special): Rewrite using 'P' mode iterator on operand 0.
20549 (*snesi_special): Likewise.
20550 (*snesi_zero): Likewise.
20551 (*seqsi_zero): Likewise.
20552 (*sltu_insn): Likewise.
20553 (*sgeu_insn): Likewise.
20554 (*seqdi_special): Make operand 0 and comparison operation be of
20555 DImode.
20556 (*snedi_special): Likewise.
20557 (*snedi_special_vis3): Likewise.
20558 (*neg_snesi_zero): Rename to *neg_snesisi_zero.
20559 (*neg_snesi_sign_extend): Rename to *neg_snesidi_zero.
20560 (*snesi_zero_extend): Delete, covered by 'P' mode iterator.
20561 (*neg_seqsi_zero): Rename to *neg_seqsisi_zero.
20562 (*neg_seqsi_sign_extend): Rename to *neg_seqsidi_zero.
20563 (*seqsi_zero_extend): Delete, covered by 'P' mode iterator.
20564 (*sltu_extend_sp64): Likewise.
20565 (*neg_sltu_insn): Rename to *neg_sltusi_insn.
20566 (*neg_sltu_extend_sp64): Rename to *neg_sltudi_insn.
20567 (*sgeu_extend_sp64): Delete, covered by 'P' mode iterator.
20568 (*neg_sgeu_insn): Rename to *neg_sgeusi_insn.
20569 (*neg_sgeu_extend_sp64): Rename to *neg_sgeudi_insn.
20570
95ca411e
YZ
205712013-04-10 Yufeng Zhang <yufeng.zhang@arm.com>
20572
20573 * config/aarch64/aarch64.c (aarch64_print_extension): New function.
20574 (aarch64_start_file): Use the new function.
20575
6782438d 205762013-04-10 Senthil Kumar Selvaraj <senthil_kumar.selvaraj@atmel.com>
33159866 20577 Jason Merrill <jason@redhat.com>
6782438d
SKS
20578
20579 * common.opt: Add -gdwarf.
20580 * opts.c (common_handle_option): Handle it.
20581 * gcc.c (ASM_DEBUG_SPEC): Don't expect "-2" for DWARF.
20582
bb313b93
RB
205832013-04-10 Richard Biener <rguenther@suse.de>
20584
20585 * passes.c (execute_todo): Do not call ggc_collect conditional here.
20586 (execute_one_ipa_transform_pass): But unconditionally here.
20587 (execute_one_pass): And here.
20588 (init_optimization_passes): Remove reload pass.
20589 * tree-pass.h (TODO_ggc_collect): Remove.
20590 (pass_reload): Likewise.
20591 * ira.c (do_reload): Merge into ...
20592 (ira): ... this.
20593 (rest_of_handle_reload): Remove.
20594 (pass_reload): Likewise.
20595 * config/i386/i386.c (ix86_option_override): Refer to ira instead
20596 of reload for vzeroupper pass placement.
20597 * <everywhere>: Remove TODO_ggc_collect from todo_flags_start
20598 and todo_flags_finish of all passes.
20599
793d9a16
RB
206002013-04-10 Richard Biener <rguenther@suse.de>
20601
20602 * tree-vectorizer.h (struct _slp_oprnd_info): Remove
20603 first_const_oprnd field, rename first_def_type to first_op_type.
20604 * tree-vect-slp.c (vect_create_oprnd_info): Adjust.
20605 (vect_get_and_check_slp_defs): Always use the type of the
20606 operand. Allow mixed vect_external_def, vect_constant_def types.
20607 (vect_get_constant_vectors): Handle mixed vect_external_def,
20608 vect_constant_def types.
20609
12211b99 206102013-04-10 Joern Rennecke <joern.rennecke@embecosm.com>
ee8a9b7b
JR
20611
20612 PR tree-optimization/55524
20613 * tree-ssa-math-opts.c
20614 (convert_mult_to_fma): Don't use an fms construct
20615 when we don't have an fms operation, but fnma, and it looks
20616 likely that we'll be able to use the latter.
20617
12211b99 206182013-04-10 Zhouyi Zhou <yizhouzhou@ict.ac.cn>
6957a6f6
ZZ
20619
20620 * cif-code.def (OVERWRITABLE): Correct the comment for overwritable
20621 function.
20622 * ipa-inline.c (can_inline_edge_p): Let dump mechanism report the
20623 inline fail caused by overwritable functions.
20624
34ab4a5b
CJW
206252013-04-10 Chung-Ju Wu <jasonwucj@gmail.com>
20626
20627 * combine.c (simplify_compare_const): Use GET_MODE_MASK to filter out
20628 unnecessary bits in the constant power of two case.
20629
abf9bfbc
RB
206302013-04-10 Richard Biener <rguenther@suse.de>
20631
20632 * tree-vect-slp.c (vect_get_and_check_slp_defs): Remove
20633 broken code swapping operands.
20634 (vect_build_slp_tree): Do not compute load permutations here.
20635 (vect_analyze_slp_instance): Compute load permutations here,
20636 after building the SLP tree.
20637
f408477e
CB
206382013-04-09 Christian Bruel <christian.bruel@st.com>
20639
20640 * config/sh/sh.md (barrier_align): Use next/prev_active_insn instead
20641 of next/prev_real_insn.
20642
5ac42672
JH
206432013-04-09 Jan Hubicka <jh@suse.cz>
20644
abf9bfbc
RB
20645 * ipa.c (cgraph_externally_visible_p, varpool_externally_visible_p):
20646 Drop aliased parameter.
5ac42672
JH
20647 (function_and_variable_visibility): Do not handle alias pairs.
20648 * cgraph.c (varpool_externally_visible_p): Update prototype.
20649 * varpool.c (varpool_add_new_variable): Update.
20650
5017f1d2
KT
206512013-04-09 Kyrylo Tkachov <kyrylo.tkachov@arm.com>
20652
20653 * config/arm/arm.md (minmax_arithsi_non_canon): New pattern.
20654
48eecbee
SB
206552013-04-09 Steven Bosscher <steven@gcc.gnu.org>
20656
a949cf1c
SB
20657 * sched-vis.c (print_pattern): Print SEQUENCE of insns as insns.
20658
48eecbee
SB
20659 * config/sparc/sparc.md: Use define_c_enum for "unspec" and "unspecv".
20660
75ef8e3d
MP
206612013-04-09 Marek Polacek <polacek@redhat.com>
20662
20663 PR tree-optimization/48762
33159866 20664 * params.def (PARAM_MAX_CSE_INSNS): Increase the minimum value to 1.
75ef8e3d 20665
23847df4
RB
206662013-04-09 Richard Biener <rguenther@suse.de>
20667
20668 * tree-vect-slp.c (vect_get_and_check_slp_defs): Remove code
20669 dealing with cost.
20670 (vect_build_slp_tree): Likewise.
20671 (vect_analyze_slp_cost_1, vect_analyze_slp_cost): New functions
20672 calculating the cost of a SLP instance.
20673 (vect_analyze_slp_instance): Use it from here, after building
20674 the SLP tree.
20675
ea3a0fde
JJ
206762013-04-09 Jakub Jelinek <jakub@redhat.com>
20677
20678 PR middle-end/56883
20679 * omp-low.c (expand_omp_for_generic, expand_omp_for_static_nochunk,
20680 expand_omp_for_static_chunk): Use simple_p = true in
20681 force_gimple_operand_gsi calls when assigning to addressable decls.
20682
a32dfe9d
JL
206832013-04-09 Jeff Law <law@redhat.com>
20684
20685 * tree-vrp.c (simplify_cond_using_ranges): Simplify test of boolean
20686 when the boolean was created by converting a wider object which
20687 had a boolean range.
20688
d755c7ef
RB
206892013-04-09 Richard Biener <rguenther@suse.de>
20690
20691 * tree-vectorizer.h (slp_void_p): Remove.
20692 (slp_tree): Typedef before _slp_tree declaration.
20693 (struct _slp_tree): Use a vector of slp_tree as children.
20694 (vect_get_place_in_interleaving_chain): Remove.
20695 * tree-vect-data-refs.c (vect_get_place_in_interleaving_chain):
20696 Move ...
20697 * tree-vect-slp.c (vect_get_place_in_interleaving_chain): ... here
20698 and make static.
20699 (vect_free_slp_tree, vect_print_slp_tree, vect_mark_slp_stmts,
20700 vect_mark_slp_stmts_relevant, vect_slp_rearrange_stmts,
20701 vect_detect_hybrid_slp_stmts, vect_slp_analyze_node_operations,
20702 vect_schedule_slp_instance, vect_remove_slp_scalar_calls):
20703 Use slp_node instead of slp_void_p and adjust.
20704
3d741091
RB
207052013-04-09 Richard Biener <rguenther@suse.de>
20706
20707 * tree-ssa-loop-manip.c (rewrite_into_loop_closed_ssa): Avoid
20708 work that is not necessary.
20709
39307ba7
JJ
207102013-04-09 Jakub Jelinek <jakub@redhat.com>
20711
20712 PR tree-optimization/56854
20713 * tree-ssa-forwprop.c (forward_propagate_addr_expr_1): Don't
20714 forward into clobber stmts if it would change MEM_REF lhs into
20715 non-MEM_REF.
20716
343881fd
MK
207172013-04-09 Maxim Kuvyrkov <maxim@kugelworks.com>
20718
20719 * tree.c (type_hash_lookup, type_hash_add): Make static.
20720 * tree.h (type_hash_lookup, type_hash_add): Remove global declarations.
20721
3922658a
RB
207222013-04-09 Richard Biener <rguenther@suse.de>
20723
20724 * tree.h (unsave_expr_now): Remove.
20725 * tree-inline.c (mark_local_for_remap_r): Remove.
20726 (unsave_expr_1): Likewise.
20727 (unsave_r): Likewise.
20728 (unsave_expr_now): Likewise.
20729 * tree-ssa-copy.c (replace_exp_1): Use unshare_expr.
20730 (propagate_tree_value): Likewise.
20731
9fb6b620
SB
207322013-04-08 Steven Bosscher <steven@gcc.gnu.org>
20733
20734 * doc/rtl.texi (sequence): Rewrite documentation to match the
20735 current use of SEQUENCE rtl objects.
20736 * rtl.def (SEQUENCE): Likewise.
20737
20738 * doc/rtl.texi (NOTE_INSN_EH_REGION_BEG, NOTE_INSN_EH_REGION_END):
20739 Update documentation.
20740 (NOTE_INSN_LOOP_BEG, NOTE_INSN_LOOP_END, NOTE_INSN_LOOP_CONT,
20741 NOTE_INSN_LOOP_VTOP): Remove documentation for non-existing notes.
20742
20743 * reg-notes.def (REG_EH_CONTEXT): Remove unused note.
20744
8ddb5a29
TJ
207452013-04-08 Teresa Johnson <tejohnson@google.com>
20746
20747 * basic-block.h (GCOV_COMPUTE_SCALE): Define.
20748 * ipa-inline-analysis.c (param_change_prob): Use helper rounding divide
33159866 20749 methods.
8ddb5a29
TJ
20750 (estimate_edge_size_and_time): Add comment to suggest using rounding
20751 methods.
20752 (estimate_node_size_and_time): Ditto.
20753 (remap_edge_change_prob): Use helper rounding divide methods.
20754 * value-prof.c (gimple_divmod_fixed_value_transform): Ditto.
20755 (gimple_mod_pow2_value_transform): Ditto.
20756 (gimple_mod_subtract_transform): Ditto.
20757 (gimple_ic_transform): Ditto.
20758 (gimple_stringops_transform): Ditto.
20759 * stmt.c (conditional_probability): Ditto.
20760 (emit_case_dispatch_table): Ditto.
20761 * lto-cgraph.c (merge_profile_summaries): Ditto.
20762 * tree-optimize.c (execute_fixup_cfg): Ditto.
20763 * cfgcleanup.c (try_forward_edges): Ditto.
20764 * cfgloopmanip.c (scale_loop_profile): Ditto.
20765 (loopify): Ditto.
20766 (duplicate_loop_to_header_edge): Ditto.
20767 (lv_adjust_loop_entry_edge): Ditto.
20768 * tree-vect-loop.c (vect_transform_loop): Ditto.
20769 * profile.c (compute_branch_probabilities): Ditto.
20770 * cfgbuild.c (compute_outgoing_frequencies): Ditto.
20771 * lto-streamer-in.c (input_cfg): Ditto.
20772 * gimple-streamer-in.c (input_bb): Ditto.
20773 * ipa-cp.c (update_profiling_info): Ditto.
20774 (update_specialized_profile): Ditto.
20775 * tree-vect-loop-manip.c (slpeel_tree_peel_loop_to_edge): Ditto.
20776 * cfg.c (update_bb_profile_for_threading): Add comment to suggest using
33159866 20777 rounding methods.
8ddb5a29
TJ
20778 * sched-rgn.c (compute_dom_prob_ps): Ditto.
20779 (compute_trg_info): Ditto.
20780 * cfgrtl.c (force_nonfallthru_and_redirect): Ditto.
20781 (purge_dead_edges): Ditto.
20782 * loop-unswitch.c (unswitch_loop): Ditto.
20783 * cgraphclones.c (cgraph_clone_edge): Ditto.
20784 (cgraph_clone_node): Ditto.
20785 * tree-inline.c (copy_bb): Ditto.
20786 (copy_edges_for_bb): Ditto.
20787 (initialize_cfun): Ditto.
20788 (copy_cfg_body): Ditto.
20789 (expand_call_inline): Ditto.
20790
661e6bd7
KT
207912013-04-08 Kai Tietz <ktietz@redhat.com>
20792
20793 * config/i386/cygwin.h (EXTRA_OS_CPP_BUILTINS): Replaced
20794 TARGET_CYGWIN64 by TARGET_64BIT.
20795
105766f3
JR
207962013-04-08 Joern Rennecke <joern.rennecke@embecosm.com>
20797
20798 * config/epiphany/epiphany.md (GPR_1): New constant.
20799 (define_expand "mov<mode>cc): FAIL if gen_compare_reg returned 0.
20800 * config/epiphany/epiphany.c (gen_compare_reg):
20801 For flag_finite_math_only, avoid swapping operands when r0 and/or r1
20802 is already in place.
20803 Use GPR_0 / GPR_1 instead of 0/1 for r0/r1 register numbers.
20804 Don't require being called during rtl expansion; If y operlaps r0,
20805 return 0.
20806 (epiphany_compute_frame_size, epiphany_expand_prologue): Use GPR_1.
20807 (epiphany_expand_epilogue): Likewise.
20808
8afab237
JR
20809 * config/epiphany/epiphany.c (epiphany_select_cc_mode):
20810 Don't use CC_FPmode for ORDERED / UNORDERED.
33159866 20811 * config/epiphany/epiphany.md (cmpsf_ord): Make pattern unconditional.
8afab237 20812
fa7707d6
JR
20813 * config/epiphany/constraints.md (CnL): New constraint.
20814 * config/epiphany/epiphany.md (addsi3_i): Add r/r/CnL alternative.
20815 * config/epiphany/predicates.md (add_operand): Allow 1024.
20816
5138e18d
JR
20817 * config/epiphany/epiphany.md (logical_op): New code iterator.
20818 (op_mnc): New code attribute.
20819 (<op_mnc>_f, mov_f, cstoresi4): New patterns.
20820 (mov_f+1, mov_f+2): New peephole2 patterns.
20821
2ccc703d
JR
20822 * config/epiphany/epiphany.md (mov_f+2): New peephole2 pattern.
20823 (cstoresi4): Also allow re-use of zero result when doing a NE
20824 comparison to a non-zero operand.
aefb0819 20825 Use (clobber (scratch)) for first insn if the gpr output is not needed.
2ccc703d 20826
093ac0a5
JR
20827 * config/epiphany/epiphany.md (<insn_opname>v2si3):
20828 Use gen_addsi3_i / gen_subsi3_i.
20829
f223bb13
JJ
208302013-04-08 Jakub Jelinek <jakub@redhat.com>
20831
20832 PR c++/34949
20833 PR c++/50243
20834 * tree-eh.c (optimize_clobbers): Only remove clobbers if bb doesn't
20835 contain anything but clobbers, at most one __builtin_stack_restore,
20836 optionally debug stmts and final resx, and if it has at least one
20837 incoming EH edge. Don't check for SSA_NAME on LHS of a clobber.
20838 (sink_clobbers): Don't check for SSA_NAME on LHS of a clobber.
20839 Instead of moving clobbers with MEM_REF LHS with SSA_NAME address
20840 which isn't defaut definition, remove them.
20841 (unsplit_eh, cleanup_empty_eh): Use single_{pred,succ}_{p,edge}
20842 instead of EDGE_COUNT comparisons or EDGE_{PRED,SUCC}.
20843 * tree-ssa-ccp.c (execute_fold_all_builtins): Remove clobbers
20844 with MEM_REF LHS with SSA_NAME address.
20845
4481581f
JL
208462013-04-08 Jeff Law <law@redhat.com>
20847
20848 * gimple.c (canonicalize_cond_expr_cond): Rewrite x ^ y into x != y.
a32dfe9d 20849
451dabda
RB
208502013-04-08 Richard Biener <rguenther@suse.de>
20851
20852 * gimple-pretty-print.c (debug_gimple_stmt): Do not print
20853 extra newline.
20854 * tree-vect-loop.c (vect_determine_vectorization_factor): Dump
20855 determined vector type.
20856 (vect_analyze_data_refs): Likewise.
20857 (vect_get_new_vect_var): Adjust.
20858 (vect_create_destination_var): Preserve SSA name versions.
20859 * tree-vect-stmts.c (get_vectype_for_scalar_type_and_size): Do
20860 not dump anything here.
20861
3b088b47
JR
208622013-04-08 Joern Rennecke <joern.rennecke@embecosm.com>
20863
20864 * config/epiphany/epiphany.h (struct GTY (()) machine_function):
20865 Add member lr_slot_known.
20866 * config/epiphany/epiphany.md (reload_insi_ra): Compute lr_slot_offs
20867 if necessary.
20868 * config/epiphany/epiphany.c (epiphany_compute_frame_size):
20869 Remove code that sets lr_slot_offset according to what a previous
20870 version of epiphany_emit_save_restore used to do.
20871 (epiphany_emit_save_restore): When doing an lr save or restore,
20872 set/verify lr_slot_known and lr_slot_offset.
20873
d8484d41
XQ
208742013-04-08 Xinyu Qi <xyqi@marvell.com>
20875
33159866
UB
20876 PR target/54338
20877 * config/arm/arm.h (REG_CLASS_CONTENTS): Include IWMMXT_GR_REGS
d8484d41
XQ
20878 in ALL_REGS.
20879
481be1c4
RB
208802013-04-08 Richard Biener <rguenther@suse.de>
20881
20882 * alias.c (find_base_term): Fix thinko in previous change.
20883
401f3a81
JJ
208842013-04-08 Jakub Jelinek <jakub@redhat.com>
20885
20886 * tree-loop-distribution.c (const_with_all_bytes_same): New function.
20887 (generate_memset_builtin): Only handle integer_all_onesp as -1 val if
20888 TYPE_PRECISION is equal to mode bitsize. Use const_with_all_bytes_same
20889 if possible to compute val.
20890 (classify_partition): Verify CONSTRUCTOR doesn't have any elts.
20891 For QImode integers don't require anything about precision. Use
20892 const_with_all_bytes_same to find out if the constant doesn't have
20893 repeated bytes in it.
20894
03ed99a8
AK
208952013-04-08 Andreas Krebbel <Andreas.Krebbel@de.ibm.com>
20896
20897 * config/s390/s390.c (s390_expand_insv): Only accept insertions
20898 within mode size.
20899
781b2e62
MP
209002013-04-08 Marek Polacek <polacek@redhat.com>
20901
20902 PR rtl-optimization/48182
20903 * params.def (PARAM_MIN_CROSSJUMP_INSNS): Increase the minimum
20904 value to 1.
20905
27e430a2
JDA
209062013-04-06 John David Anglin <dave.anglin@nrc-cnrc.gc.ca>
20907
20908 PR target/55487
20909 * config/pa/pa.c (legitimize_pic_address): Before incrementing label
20910 nuses, make sure we have a label.
20911
4902aa64
BS
209122013-04-05 Bill Schmidt <wschmidt@linux.vnet.ibm.com>
20913
20914 PR target/56843
20915 * config/rs6000/rs6000.c (rs6000_emit_swdiv_high_precision): Remove.
20916 (rs6000_emit_swdiv_low_precision): Remove.
20917 (rs6000_emit_swdiv): Rewrite to handle between one and four
20918 iterations of Newton-Raphson generally; modify required number of
20919 iterations for some cases.
20920 * config/rs6000/rs6000.h (RS6000_RECIP_HIGH_PRECISION_P): Remove.
20921
7bca81dc
SB
209222013-04-05 Steven Bosscher <steven@gcc.gnu.org>
20923
20924 * bb-reorder.c (fix_crossing_unconditional_branches): Remove a
20925 set-but-unused variable.
20926
20927 * cgraph.c (cgraph_release_function_body): Clear cfun->cfg to make
20928 basic blocks of released function bodies garbage-collectable.
20929
20930 * ree.c (find_and_remove_re): Do not call df_finish_pass here.
20931 (struct rtl_opt_pass): Add TODO_df_finish.
20932
20933 * rtl.def (DEFINE_SUBST, DEFINE_SUBST_ATTR): Add documentation.
20934
4542a38a
GY
209352013-04-05 Greta Yorsh <Greta.Yorsh@arm.com>
20936
20937 * config/arm/constraints.md (q): New constraint.
20938 * config/arm/ldrdstrd.md: New file.
20939 * config/arm/arm.md (ldrdstrd.md) New include.
20940 (arm_movdi): Use "q" instead of "r" constraint
20941 for double-word memory access.
20942 (movdf_soft_insn): Likewise.
20943 * config/arm/vfp.md (movdi_vfp): Likewise.
20944 * config/arm/t-arm (MD_INCLUDES): Add ldrdstrd.md.
33159866 20945 * config/arm/arm-protos.h (gen_operands_ldrd_strd): New declaration.
4542a38a
GY
20946 * config/arm/arm.c (gen_operands_ldrd_strd): New function.
20947 (mem_ok_for_ldrd_strd): Likewise.
20948 (output_move_double): Update assertion.
20949
2385b218
GY
209502013-04-05 Greta Yorsh <Greta.Yorsh@arm.com>
20951
20952 * config/arm/arm.md: Comment on splitting Thumb1 patterns.
20953
75fe1cb5
GY
209542013-04-05 Greta Yorsh <Greta.Yorsh@arm.com>
20955
20956 * config/arm/arm.md (arm_smax_insn): Convert define_insn into
20957 define_insn_and_split.
20958 (arm_smin_insn,arm_umaxsi3,arm_uminsi3): Likewise.
20959
dd660e8e
GY
209602013-04-05 Greta Yorsh <Greta.Yorsh@arm.com>
20961
20962 * config/arm/arm.md (arm_ashldi3_1bit): Convert define_insn into
20963 define_insn_and_split.
33159866 20964 (arm_ashrdi3_1bit,arm_lshrdi3_1bit): Likewise.
dd660e8e
GY
20965 (shiftsi3_compare): New pattern.
20966 (rrx): New pattern.
20967 * config/arm/unspecs.md (UNSPEC_RRX): New.
20968
045e472c
GY
209692013-04-05 Greta Yorsh <Greta.Yorsh@arm.com>
20970
20971 * config/arm/arm.md (negdi_extendsidi): New pattern.
20972 (negdi_zero_extendsidi): Likewise.
20973
3f3bf1a8
GY
209742013-04-05 Greta Yorsh <Greta.Yorsh@arm.com>
20975
20976 * config/arm/arm.md (andsi_iorsi3_notsi): Convert define_insn into
20977 define_insn_and_split.
20978 (arm_negdi2,arm_abssi2,arm_neg_abssi2): Likewise.
20979 (arm_cmpdi_insn,arm_cmpdi_unsigned): Likewise.
20980
b0b49556
GY
209812013-04-05 Greta Yorsh <Greta.Yorsh@arm.com>
20982
20983 * config/arm/arm.md (arm_subdi3): Convert define_insn into
20984 define_insn_and_split.
20985 (subdi_di_zesidi,subdi_di_sesidi): Likewise.
20986 (subdi_zesidi_di,subdi_sesidi_di,subdi_zesidi_zesidi): Likewise.
20987
d633dd84
GY
209882013-04-05 Greta Yorsh <Greta.Yorsh@arm.com>
20989
20990 * config/arm/arm.md (subsi3_carryin): New pattern.
20991 (subsi3_carryin_const): Likewise.
20992 (subsi3_carryin_compare,subsi3_carryin_compare_const): Likewise.
20993 (subsi3_carryin_shift,rsbsi3_carryin_shift): Likewise.
20994
f4499066
GY
209952013-04-05 Greta Yorsh <Greta.Yorsh@arm.com>
20996
20997 * config/arm/arm.md (incscc,arm_incscc,decscc,arm_decscc): Delete.
20998
ceef6fd9
GY
209992013-04-05 Greta Yorsh <Greta.Yorsh@arm.com>
21000
21001 * config/arm/arm.md (addsi3_carryin_<optab>): Set attribute predicable.
d633dd84 21002 (addsi3_carryin_alt2_<optab>,addsi3_carryin_shift_<optab>): Likewise.
ceef6fd9 21003
ddbdd8a7
KT
210042013-04-05 Kyrylo Tkachov <kyrylo.tkachov@arm.com>
21005
21006 * config/arm/arm.c (arm_expand_builtin): Change fcode
21007 type to unsigned int.
21008
8456d78a
RR
210092013-04-05 Ramana Radhakrishnan <ramana.radhakrishnan@arm.com>
21010
21011 * doc/invoke.texi (ARM Options): Document cortex-a53 support.
21012
526c230f
ILT
210132013-04-04 Ian Lance Taylor <iant@google.com>
21014
33159866
UB
21015 * doc/standards.texi (Standards): The Go frontend supports the Go 1
21016 language standard.
526c230f 21017
3cfbe04d
SB
210182013-04-04 Steven Bosscher <steven@gcc.gnu.org>
21019
21020 PR middle-end/56729
21021 * df-scan.c (df_insn_delete): Disable failing assert.
21022
dfa3f8d0
KT
210232013-04-04 Kyrylo Tkachov <kyrylo.tkachov@arm.com>
21024
21025 * config/arm/arm-protos.h (arm_builtin_vectorized_function):
21026 New function prototype.
21027 * config/arm/arm.c (TARGET_VECTORIZE_BUILTINS): Define.
21028 (TARGET_VECTORIZE_BUILTIN_VECTORIZED_FUNCTION): Likewise.
21029 (arm_builtin_vectorized_function): New function.
21030
f7837758
KT
210312013-04-04 Kyrylo Tkachov <kyrylo.tkachov@arm.com>
21032
21033 * config/arm/arm_neon_builtins.def: New file.
21034 * config/arm/arm.c (neon_builtin_data): Move contents to
21035 arm_neon_builtins.def.
21036 (enum arm_builtins): Include neon builtin definitions.
21037 (ARM_BUILTIN_NEON_BASE): Move from enum to macro.
33159866 21038 * config/arm/t-arm (arm.o): Add dependency on arm_neon_builtins.def.
f7837758 21039
39c1b6db
MP
210402013-04-04 Marek Polacek <polacek@redhat.com>
21041
21042 PR tree-optimization/48186
21043 * predict.c (maybe_hot_frequency_p): Return false if
21044 HOT_BB_FREQUENCY_FRACTION is 0.
21045 (cgraph_maybe_hot_edge_p): Likewise.
21046
314f64eb
RB
210472013-04-04 Richard Biener <rguenther@suse.de>
21048
21049 PR tree-optimization/56826
21050 * tree-vect-slp.c (vect_build_slp_tree): Compute ncopies
21051 more accurately.
21052
51a905b2
RB
210532013-04-04 Richard Biener <rguenther@suse.de>
21054
21055 PR tree-optimization/56213
21056 * tree-vect-data-refs.c (vect_check_strided_load): Remove.
33159866 21057 (vect_analyze_data_refs): Allow all non-nested loads as strided loads.
51a905b2 21058
f1bf4f3a
RB
210592013-04-04 Richard Biener <rguenther@suse.de>
21060
21061 PR tree-optimization/56837
21062 * tree-loop-distribution.c (classify_partition): For non-zero
21063 values require that the value has the same precision as its
21064 mode to be useful as memset value.
21065
0bca7ded
NC
210662013-04-03 Nick Clifton <nickc@redhat.com>
21067
33159866 21068 * config/v850/v850e3v5.md (fmasf4): Use fmaf.s on E3V5 architectures.
0bca7ded
NC
21069 (fmssf4): Use fmsf.s on E3V5 architectures.
21070 (fnmasf4): Use fnmaf.s on E3V5 architectures.
21071 (fnmssf4): Use fnmsf.s on E3V5 architectures.
21072
b4019227
JL
210732013-04-03 Jeff Law <law@redhat.com>
21074
21075 * Makefile.in (lra-constraints.o): Depend on $(OPTABS_H).
21076 (lra-eliminations.o): Likewise.
21077
f57ddb5b
TJ
210782013-04-03 Teresa Johnson <tejohnson@google.com>
21079
21080 * gcov-io.c (compute_working_sets): Moved most of body of old
33159866 21081 compute_working_sets here from profile.c.
f57ddb5b 21082 * gcov-io.h (NUM_GCOV_WORKING_SETS): Moved here from profile.c.
33159866
UB
21083 (gcov_working_set_t): Moved typedef here from basic-block.h
21084 (compute_working_set): Declare.
f57ddb5b
TJ
21085 * profile.c (NUM_GCOV_WORKING_SETS): Moved to gcov-io.h.
21086 (get_working_sets): Renamed from compute_working_set,
33159866 21087 replace most of body with call to new compute_working_sets.
f57ddb5b 21088 (get_exec_counts): Replace call to compute_working_sets
33159866
UB
21089 to get_working_sets.
21090 * profile.h (get_working_sets): Renamed from compute_working_set.
f57ddb5b 21091 * lto-cgraph.c (input_symtab): Replace call to compute_working_sets
33159866 21092 to get_working_sets.
f57ddb5b
TJ
21093 * basic-block.h (gcov_working_set_t): Moved to gcov-io.h.
21094 * gcov-dump.c (dump_working_sets): New function.
21095
12211b99 210962013-04-03 Kenneth Zadeck <zadeck@naturalbridge.com>
5e0919f1
KZ
21097
21098 * hwint.c (sext_hwi, zext_hwi): New functions.
21099 * hwint.h (HOST_BITS_PER_HALF_WIDE_INT, HOST_HALF_WIDE_INT,
21100 HOST_HALF_WIDE_INT_PRINT, HOST_HALF_WIDE_INT_PRINT_C,
21101 HOST_HALF_WIDE_INT_PRINT_DEC, HOST_HALF_WIDE_INT_PRINT_DEC_C,
21102 HOST_HALF_WIDE_INT_PRINT_UNSIGNED, HOST_HALF_WIDE_INT_PRINT_HEX,
21103 HOST_HALF_WIDE_INT_PRINT_HEX_PURE): New symbols.
21104 (sext_hwi, zext_hwi): New functions.
0bca7ded 21105
be672e08
JL
211062013-04-03 Jeff Law <law@redhat.com>
21107
21108 PR tree-optimization/56799
33159866
UB
21109 * tree-ssa-dom.c (record_equivalences_from_incoming_edge): Bring
21110 back test for widening conversion erroneously dropped in prior change.
be672e08 21111
9d821fa5
KT
211122013-04-03 Kyrylo Tkachov <kyrylo.tkachov@arm.com>
21113
21114 PR target/56809
21115 * config/aarch64/aarch64.c (is_jump_table): Use next_active_insn
21116 instead of next_real_insn.
21117
71c581e7
MP
211182013-04-03 Marek Polacek <polacek@redhat.com>
21119
21120 PR sanitizer/55702
33159866 21121 * tsan.c (instrument_func_exit): Allow BUILT_IN_RETURN functions.
71c581e7 21122
4a32ef80
KT
211232013-04-03 Kyrylo Tkachov <kyrylo.tkachov@arm.com>
21124
21125 PR target/56809
21126 * config/arm/arm.c (is_jump_table): Use next_active_insn instead of
21127 next_real_insn.
21128 (thumb1_output_casesi): Likewise.
21129 (thumb2_output_casesi): Likewise.
21130
1b2253d4
RB
211312013-04-03 Richard Biener <rguenther@suse.de>
21132
21133 PR tree-optimization/56817
21134 * tree-ssa-loop-ivcanon.c (tree_unroll_loops_completely):
21135 Split out ...
21136 (tree_unroll_loops_completely_1): ... new function to manually
21137 walk the loop tree, properly defering outer loops of unrolled
21138 loops to later iterations.
21139
38000232
MG
211402013-04-03 Marc Glisse <marc.glisse@inria.fr>
21141
21142 * tree-vect-stmts.c (vectorizable_store): Accept BIT_FIELD_REF.
21143 (vectorizable_load): Likewise.
21144 * tree-vect-slp.c (vect_build_slp_tree): Likewise.
21145 * tree-vect-data-refs.c (vect_create_data_ref_ptr): Handle VECTOR_TYPE.
21146
3fa35298
MG
211472013-04-03 Marc Glisse <marc.glisse@inria.fr>
21148
21149 * tree-flow-inline.h (get_addr_base_and_unit_offset_1): Handle
21150 BIT_FIELD_REF.
21151
b3d45ff0
UW
211522013-04-03 Ulrich Weigand <Ulrich.Weigand@de.ibm.com>
21153
21154 * config/spu/spu.c (emit_nop_for_insn): Handle JUMP_TABLE_DATA.
21155
ec9202a8
BC
211562013-04-03 Bin Cheng <bin.cheng@arm.com>
21157
21158 * rtl.h (AUTO_INC_DEC): Fix typo of HAVE_POST_MODIFY_DISP.
21159
6805bd36
MG
211602013-04-03 Marc Glisse <marc.glisse@inria.fr>
21161
21162 PR tree-optimization/56790
33159866
UB
21163 * fold-const.c (fold_ternary_loc) <VEC_COND_EXPR>: Add constant
21164 folding.
6805bd36 21165
da694a77
MG
211662013-04-03 Marc Glisse <marc.glisse@inria.fr>
21167
21168 * simplify-rtx.c (simplify_binary_operation_1) <VEC_SELECT>:
21169 Handle VEC_MERGE.
21170 (simplify_ternary_operation) <VEC_MERGE>: Use unsigned HOST_WIDE_INT
21171 for masks. Test for side effects. Handle nested VEC_MERGE. Handle
21172 equal arguments.
21173
4e7d7b3d
JJ
211742013-04-03 Jakub Jelinek <jakub@redhat.com>
21175
21176 PR c/19449
21177 * tree.h (force_folding_builtin_constant_p): New decl.
21178 * builtins.c (force_folding_builtin_constant_p): New variable.
21179 (fold_builtin_constant_p): Fold immediately also if
21180 force_folding_builtin_constant_p.
21181
e6c9d234
RB
211822013-04-03 Richard Biener <rguenther@suse.de>
21183
21184 PR tree-optimization/56812
21185 * tree-vect-data-refs.c (vect_slp_analyze_data_ref_dependence):
21186 DRs of the same interleaving chain are independent.
21187
984d07dd
JM
211882013-04-02 Jason Merrill <jason@redhat.com>
21189
21190 * gdbinit.in (pbb): Use debug fn.
21191
622849c9
LC
211922013-04-02 Lawrence Crowl <crowl@google.com>
21193
21194 * sese.h (struct ivtype_map_elt_s): Remove unused.
21195 (extern debug_ivtype_map): Remove unused.
21196 (extern eq_ivtype_map_elts): Remove unused.
21197 * sese.c (debug_ivtype_map): Removed unused.
21198 (debug_ivtype_map_1): Removed unused.
21199 (debug_ivtype_elt): Remove unused.
21200 (eq_ivtype_map_elts): Remove unused.
21201
21202
82c0e1a0
KT
212032013-04-02 Kai Tietz <ktietz@redhat.com>
21204
21205 PR target/52790
21206 * config/i386/cygming.h (SUB_TARGET_RECORD_STUB): New sub-target macro.
21207 * config/i386/i386-protos.h (i386_pe_record_stub): Add new prototype.
33159866
UB
21208 * config/i386/i386.c (legitimize_pe_coff_extern_decl): New static
21209 function.
82c0e1a0
KT
21210 (legitimize_pe_coff_symbol): Likewise.
21211 (is_imported_p): New helper-function.
21212 (ix86_option_override_internal): Make MEDIUM_PIC the default code-model
21213 for Windows x64 targets.
21214 (ix86_expand_prologue): Optimize for pe-coff targets.
21215 (ix86_expand_split_stack_prologue): Adjust for pe-coff targets.
21216 (legitimate_pic_address_disp_p): Adjust for x64 pe-coff to support
21217 medium/large code-model.
21218 (legitimize_pic_address): Likewise.
21219 (legitimize_tls_address): Likewise.
21220 (ix86_expand_call): Likewise.
21221 (x86_output_mi_thunk): Likewise.
21222 (get_dllimport_decl): Add new beimport argument.
21223 (construct_plt_address): Don't assert for x64 pe-coff targets.
21224 * config/i386/i386.h (PIC_OFFSET_TABLE_REGNUM): Adjust for x64 pe-coff
21225 targets.
21226 (SYMBOL_FLAG_STUBVAR): New macro.
21227 (SYMBOL_REF_STUBVAR_P): Likewise.
21228 * config/i386/winnt.c (stub_list): New structure.
21229 (stub_head): New local variable.
21230 (i386_pe_record_stub): New function.
21231 (i386_pe_file_end): Emit refptr-stubs.
21232
5d751b0c
JJ
212332013-04-02 Jakub Jelinek <jakub@redhat.com>
21234
09bb4c99
JJ
21235 PR rtl-optimization/56745
21236 * ifcvt.c (cond_exec_find_if_block): Don't try to optimize
21237 if then_bb has no successors and else_bb is EXIT_BLOCK_PTR.
21238
a1d8947a
JJ
21239 PR c++/34949
21240 * tree-ssa-alias.c (stmt_kills_ref_p_1): If base != ref->base
21241 and both of them are MEM_REFs, just compare first argument for
21242 equality and attempt to deal even with differing offsets.
21243
5d751b0c
JJ
21244 PR c++/34949
21245 * tree-cfg.c (verify_gimple_assign_single): Allow lhs
21246 of gimple_clobber_p to be MEM_REF.
21247 * gimplify.c (gimplify_modify_expr): Gimplify *to_p of
21248 an assignment from TREE_CLOBBER_P. Allow it to be MEM_REF
21249 after gimplification.
21250 * asan.c (get_mem_ref_of_assignment): Don't instrument
21251 gimple_clobber_p stmts.
21252 * tree-ssa-dse.c (dse_optimize_stmt): Allow DSE of
21253 gimple_clobber_p stmt if they have MEM_REF lhs and
21254 are dead because of another gimple_clobber_p stmt.
21255 * tree-ssa-live.c (clear_unused_block_pointer): Treat
21256 gimple_clobber_p stmts like debug stmts.
21257 (remove_unused_locals): Remove clobbers with MEM_REF lhs
21258 that refer to unused VAR_DECLs or uninitialized values.
21259 * tree-sra.c (sra_ipa_reset_debug_stmts): Also remove
21260 gimple_clobber_p stmts if they refer to removed parameters.
21261 (get_repl_default_def_ssa_name, sra_ipa_modify_expr): Fix up
21262 formatting.
21263
e594716a
UB
212642013-04-02 Uros Bizjak <ubizjak@gmail.com>
21265
21266 * config/i386/i386.md (*testqi_ext_3): Merge with *testqi_ext_3_rex64
21267 using SWI48 mode attribute.
21268
7a80735b
WM
212692013-04-02 Wei Mi <wmi@google.com>
21270
21271 * config/i386/i386.c (ix86_rtx_costs): Set proper rtx cost for
21272 ashl<mode>3_mask, *<shift_insn><mode>3_mask and
21273 *<rotate_insn><mode>3_mask in i386.md.
21274
f423a9e4
AI
212752013-04-02 Alexander Ivchenko <alexander.ivchenko@intel.com>
21276
21277 * config.gcc (arm*-*-linux-*): Remove duplicate t-linux-android.
21278
90eb75f2
RB
212792013-04-02 Richard Biener <rguenther@suse.de>
21280
21281 PR tree-optimization/56778
21282 * tree-vect-data-refs.c (vect_analyze_data_ref_dependence):
21283 Runtime alias tests are not supported for gather loads.
21284 * tree-vect-loop-manip.c (vect_loop_versioning): Insert
21285 stmts referenced from SSA operands before updating SSA form.
21286
d8c69a92
IC
212872013-04-02 Ian Caulfield <ian.caulfield@arm.com>
21288 Ramana Radhakrishnan <ramana.radhakrishnan@arm.com>
21289
21290 * config/arm/arm-arches.def (armv8-a): Default to cortex-a53.
e594716a
UB
21291 * config/arm/t-arm (MD_INCLUDES): Depend on cortex-a53.md.
21292 * config/arm/cortex-a53.md: New file.
21293 * config/arm/bpabi.h (BE8_LINK_SPEC): Handle cortex-a53.
21294 * config/arm/arm.md (generic_sched, generic_vfp): Handle cortex-a53.
21295 * config/arm/arm.c (arm_issue_rate): Likewise.
21296 * config/arm/arm-tune.md: Regenerate
21297 * config/arm/arm-tables.opt: Regenerate.
21298 * config/arm/arm-cores.def: Add cortex-a53.
d8c69a92 21299
239eb04c
ZC
213002013-04-02 Zhenqiang Chen <zhenqiang.chen@arm.com>
21301
21302 * config/arm/uclinux-elf.h: Add %L to LINK_GCC_C_SEQUENCE_SPEC for
21303 non-static link.
21304
c902d3c8
SN
213052013-04-02 Sofiane Naci <sofiane.naci@arm.com>
21306
d8c69a92
IC
21307 * config/aarch64/aarch64.md (*mov<mode>_aarch64): Add variants for
21308 scalar load/store operations using B/H registers.
c902d3c8
SN
21309 (*zero_extend<SHORT:mode><GPI:mode>2_aarch64): Likewise.
21310
051d0e2f
SN
213112013-04-02 Sofiane Naci <sofiane.naci@arm.com>
21312
21313 * config/aarch64/aarch64.md (*mov<mode>_aarch64): Add alternatives for
21314 scalar move.
21315 * config/aarch64/aarch64.c
21316 (aarch64_simd_scalar_immediate_valid_for_move): New.
21317 * config/aarch64/aarch64-protos.h
21318 (aarch64_simd_scalar_immediate_valid_for_move): New.
21319 * config/aarch64/constraints.md (Dh, Dq): New.
21320 * config/aarch64/iterators.md (hq): New.
21321
0ee1e3d9
EB
213222013-04-02 Eric Botcazou <ebotcazou@adacore.com>
21323
21324 * reorg.c (get_branch_condition): Deal with conditional returns.
21325 (fill_simple_delay_slots): Remove dead code dealing with jumps.
21326
136fb3f7
WM
213272013-04-01 Wei Mi <wmi@google.com>
21328
21329 * config/i386/i386.md (*ashl<mode>3_mask): Rewrite as define_insn.
21330 Truncate operand 2 using %b asm operand modifier.
21331 (*<shift_insn><mode>3_mask): Ditto.
21332 (*<rotate_insn><mode>3_mask): Ditto.
21333
6388c738
SB
213342013-04-01 Steven Bosscher <steven@gcc.gnu.org>
21335
21336 PR middle-end/56798
21337 * cfgbuild.c (inside_basic_block_p): Restore check broken at r197234.
21338
f7a4d826
KK
213392013-03-31 Kaz Kojima <kkojima@gcc.gnu.org>
21340
21341 * config/sh/sh.md (casesi_worker_1): Use next_active_insn instead
21342 of next_real_insn.
21343 (casesi_worker_2, casesi_shift_media, casesi_load_media): Likewise.
21344
4b943a49
LC
213452013-03-30 Lawrence Crowl <crowl@google.com>
21346
21347 * dse.c (clear_alias_sets): Remove never set.
21348 (disqualified_clear_alias_sets): Remove never set.
21349 (clear_alias_mode_pool): Remove never set.
21350 (dse_step0): Remove condition that is never true.
21351 (canon_address): Remove condition that is never true.
21352 (dse_step7): Remove condition that is never true.
21353 (rest_of_handle_dse): Remove condition that is never true.
21354 (rest_of_handle_dse::did_global): Remove never read from above.
21355 (dse_step2_spill): Remove never called from above.
21356 (dse_step5_spill): Remove never called from above.
21357
39718607
SB
213582013-03-30 Steven Bosscher <steven@gcc.gnu.org>
21359
da5c6bde
SB
21360 * doc/md.texi (Standard Names) <casesi>: Update documentation for
21361 JUMP_TABLE_DATA changes.
21362 * doc/tm.texi.in (Dispatch Tables) <ASM_OUTPUT_CASE_LABEL>: Likewise.
21363 * doc/rtl.texi (Flags) <INSN_DELETED_P, SCHED_GROUP_P>: Likewise.
21364 (Insns) <jump_table_data>: New entry.
21365 * doc/tm.texi: Regenerate.
21366
39718607
SB
21367 * cfgrtl.c (fixup_reorder_chain): Do not emit barriers to BB_FOOTER.
21368
21369 * postreload-gcse.c (bb_has_well_behaved_predecessors): Correct test
21370 for table jump at the end of a basic block using tablejump_p.
21371 * targhooks.c (default_invalid_within_doloop): Likewise.
21372 * config/rs6000/rs6000.c (TARGET_INVALID_WITHIN_DOLOOP): Remove
21373 target hook implementation that is identical to the default hook.
21374 (rs6000_invalid_within_doloop): Remove.
21375
21376 * bb-reorder.c (fix_crossing_unconditional_branches): Remove set but
21377 unused variable from tablejump_p call.
21378
21379 * rtl.def (JUMP_TABLE_DATA): New RTX_INSN object.
21380 * rtl.h (RTX_PREV, RTX_NEXT): Adjust for new JUMP_TABLE_DATA.
21381 (INSN_DELETED_P): Likewise.
21382 (emit_jump_table_data): New prototype.
21383 * gengtype.c (adjust_field_rtx_def): Handle JUMP_TABLE_DATA fields
21384 after 4th as unused.
21385 * print-rtl.c (print_rtl): Handle JUMP_TABLE_DATA.
21386 * sched-vis.c (print_insn): Likewise.
21387 * emit-rtl.c (active_insn_p): Consider JUMP_TABLE_DATA an active
21388 insn for compatibility with back ends that use next_active_insn to
21389 identify jump table data.
21390 (set_insn_deleted): Remove no longer useful JUMP_TABLE_DATA_P check.
21391 (remove_insn): Likewise.
21392 (emit_insn): Do not accept JUMP_TABLE_DATA objects in insn chains
21393 to be emitted.
21394 (emit_debug_insn, emit_jump_insn, emit_call_insn, emit_label): Idem.
21395 (emit_jump_table_data): New function.
21396
21397 * cfgbuild.c (inside_basic_block_p): A JUMP_INSN is always inside a
21398 basic block, a JUMP_TABLE_DATA never is.
21399 (control_flow_insn_p): JUMP_TABLE_DATA is not a control flow insn.
21400 * cfgrtl.c (duplicate_insn_chain): Split handling of JUMP_TABLE_DATA
21401 off from code handling real insns.
21402 * final.c (get_attr_length_1): Simplify for JUMP_INSNs.
21403 * function.c (instantiate_virtual_regs): Remove JUMP_TABLE_DATA_P
21404 test, now redundant because JUMP_TABLE_DATA is not an INSN_P insn.
21405 * gcse.c (insert_insn_end_basic_block): Likewise, JUMP_TABLE_DATA_P
21406 is not a NONDEBUG_INSN_P.
21407 * ira-costs.c (scan_one_insn): Likewise.
21408 * jump.c (mark_all_labels): Likewise.
21409 (mark_jump_label_1): Likewise.
21410 * lra-eliminations.c (eliminate_regs_in_insn): Likewise.
21411 * lra.c (get_insn_freq): Expect all insns reaching here to be in
21412 a basic block.
21413 (check_rtl): Remove JUMP_TABLE_DATA_P test, not a NONDEBUG_INSN_P insn.
21414 * predict.c (expensive_function_p): Use FOR_BB_INSNS.
21415 * reload1.c (calculate_needs_all_insns): Call set_label_offsets for
21416 JUMP_TABLE_DATA_P insns.
21417 (calculate_elim_costs_all_insns): Likewise.
21418 (set_label_offsets): Recurse on the PATTERN of JUMP_TABLE_DATA insns.
21419 (elimination_costs_in_insn): Remove redundant JUMP_TABLE_DATA_P test.
21420 (delete_output_reload): Code style fixups.
21421 * reorg.c (dbr_schedule): Move JUMP_TABLE_DATA_P up to avoid setting
21422 insn flags on this non-insn.
21423 * sched-rgn.c (add_branch_dependences): Treat JUMP_TABLE_DATA insns
21424 as scheduling barriers, for pre-change compatibility.
21425 * stmt.c (emit_case_dispatch_table): Emit jump table data not as
21426 JUMP_INSN objects but instead as JUMP_TABLE_DATA objects.
21427
21428 * config/alpha/alpha.c (alpha_does_function_need_gp): Remove
21429 redundant JUMP_TABLE_DATA_P test.
21430 * config/arm/arm.c (thumb_far_jump_used_p): Likewise.
21431 * config/frv/frv.c (frv_function_contains_far_jump): Likewise.
21432 (frv_for_each_packet): Likewise.
21433 * config/i386/i386.c (min_insn_size): Likewise.
21434 (ix86_avoid_jump_mispredicts): Likewise.
21435 * config/m32r/m32r.c (m32r_is_insn): Likewise.
21436 * config/mep/mep.c (mep_reorg_erepeat): Likewise.
21437 * config/mips/mips.c (USEFUL_INSN_P): Likewise.
21438 (mips16_insn_length): Robustify.
21439 (mips_has_long_branch_p): Remove redundant JUMP_TABLE_DATA_P test.
21440 (mips16_split_long_branches): Likewise.
21441 * config/pa/pa.c (pa_combine_instructions): Likewise.
21442 * config/rs6000/rs6000.c (get_next_active_insn): Treat
21443 JUMP_TABLE_DATA objects as active insns, like in active_insn_p.
21444 * config/s390/s390.c (s390_chunkify_start): Treat JUMP_TABLE_DATA
21445 as contributing to pool range lengths.
21446 * config/sh/sh.c (find_barrier): Restore check for ADDR_DIFF_VEC.
21447 Remove redundant JUMP_TABLE_DATA_P test.
21448 (sh_loop_align): Likewise.
21449 (split_branches): Likewise.
21450 (sh_insn_length_adjustment): Likewise.
21451 * config/spu/spu.c (get_branch_target): Likewise.
21452
0208f7da
JH
214532013-03-29 Jan Hubicka <jh@suse.cz>
21454
21455 * lto-cgraph.c (output_profile_summary, input_profile_summary): Use
21456 gcov streaming; stream hot bb threshold to ltrans.
21457 * predict.c (get_hot_bb_threshold): Break out from ....
21458 (maybe_hot_count_p): ... here.
21459 (set_hot_bb_threshold): New function.
21460 * lto-section-in.c (lto_section_name): Add profile.
21461 * profile.h (get_hot_bb_threshold, set_hot_bb_threshold): Declare.
21462 * ipa.c: Include hash-table.h, tree-inline.h, profile.h, lto-streamer.h
21463 and data-streamer.h
21464 (histogram_entry): New structure.
21465 (histogram, histogram_pool): New global vars.
21466 (histogram_hash): New structure.
21467 (histogram_hash::hash): New method.
21468 (histogram_hash::equal): Likewise.
21469 (account_time_size): New function.
21470 (cmp_counts): New function.
21471 (dump_histogram): New function.
21472 (ipa_profile_generate_summary): New function.
21473 (ipa_profile_write_summary): New function.
21474 (ipa_profile_read_summary): New function.
21475 (ipa_profile): Decide on threshold.
e594716a
UB
21476 (pass_ipa_profile): Add ipa_profile_write_summary and
21477 ipa_profile_read_summary.
0208f7da
JH
21478 * Makefile.in (ipa.o): Update dependencies.
21479 * lto-streamer.h (LTO_section_ipa_profile): New section.
21480
5a6ccc94
GDR
214812013-03-29 Gabriel Dos Reis <gdr@integrable-solutions.net>
21482
21483 * tree.h (VAR_P): New.
21484
39385fa6
PC
214852013-03-29 Paolo Carlini <paolo.carlini@oracle.com>
21486
21487 PR lto/56777
21488 * doc/invoke.texi ([-fwhole-program]): Fix typo.
21489
34f0d87a
SB
214902013-03-29 Steven Bosscher <steven@gcc.gnu.org>
21491
21492 * cfgbuild.c (inside_basic_block_p): Use JUMP_TABLE_DATA_P in lieu
21493 of tests for JUMP_P and a ADDR_DIFF_VEC or ADDR_VEC pattern.
21494 (control_flow_insn_p): Likewise.
21495 * cfgrtl.c (duplicate_insn_chain): Likewise.
21496 * final.c (get_attr_length_1): Likewise.
21497 (shorten_branches): Likewise.
21498 (final_scan_insn): Likewise.
21499 * function.c (instantiate_virtual_regs): Likewise.
21500 * gcse.c (insert_insn_end_basic_block): Likewise.
21501 * ira-costs.c (scan_one_insn): Likewise.
21502 * lra-eliminations.c (eliminate_regs_in_insn): Likewise.
21503 * lra.c (check_rtl): Likewise.
21504 * reload1.c (elimination_costs_in_insn): Likewise.
21505 * reorg.c (follow_jumps): Likewise.
21506
21507 * config/arm/arm.c (is_jump_table): Use JUMP_TABLE_DATA_P in lieu
21508 of tests for JUMP_P and a ADDR_DIFF_VEC or ADDR_VEC pattern.
21509 (thumb_far_jump_used_p): Likewise.
21510 * config/bfin/bfin.c (workaround_rts_anomaly): Likewise.
21511 (workaround_speculation): Likewise.
21512 (add_sched_insns_for_speculation): Likewise.
21513 * config/c6x/c6x.c (reorg_emit_nops): Likewise.
21514 * config/frv/frv.c (frv_function_contains_far_jump): Likewise.
21515 (frv_for_each_packet): Likewise.
21516 * config/i386/i386.c (ix86_avoid_jump_mispredicts): Likewise.
21517 * config/ia64/ia64.c (emit_all_insn_group_barriers): Likewise.
21518 (final_emit_insn_group_barriers): Likewise.
21519 * config/m32r/m32r.c (m32r_is_insn): Likewise.
21520 * config/mips/mips.c (USEFUL_INSN_P): Likewise.
21521 (mips16_insn_length): Likewise.
21522 * config/pa/pa.c (pa_reorg): Likewise.
21523 (pa_combine_instructions): Likewise.
21524 * config/rs6000/rs6000.c (rs6000_invalid_within_doloop): Likewise.
21525 * config/sh/sh.c (fixup_addr_diff_vecs): Likewise.
21526 (sh_reorg): Likewise.
21527 (split_branches): Likewise.
21528 * config/spu/spu.c (get_branch_target): Likewise.
21529
21530 * config/s390/s390.c (s390_chunkify_start): Simplify logic using
21531 JUMP_TABLE_DATA_P.
21532
4ac761b0
KY
215332013-03-29 Kirill Yukhin <kirill.yukhin@intel.com>
21534
39385fa6 21535 * config/i386/avx2intrin.h (_mm256_broadcastsi128_si256):
4ac761b0
KY
21536 Fix declaration name.
21537
58a49189
LC
215382013-03-28 Lawrence Crowl <crowl@google.com>
21539
21540 * graphds.h (struct graph.indicies): Remove unused.
21541 * graphite-poly.h (struct graph.original_pddrs): Remove unused.
21542 (SCOP_ORIGINAL_PDDRS): Remove unused.
21543 * sese.h (extern insert_loop_close_phis): Removed unused.
21544 (extern insert_guard_phis): Removed unused.
21545 (extern ivtype_map_elt_info): Removed unused.
21546 (new_ivtype_map_elt): Removed unused.
21547 * sese.c (ivtype_map_elt_info): Removed unused.
21548
7b3b6ae4
LC
215492013-03-28 Lawrence Crowl <crowl@google.com>
21550
21551 * Makefile.in: Add several missing include dependences.
21552 (DUMPFILE_H): New.
21553 (test-dump.o): New. This object is not added to any executable,
21554 but is present for ad-hoc testing.
21555 * bitmap.c
21556 (debug (const bitmap_head_def &)): New.
21557 (debug (const bitmap_head_def *)): New.
21558 * bitmap.h
21559 (extern debug (const bitmap_head_def &)): New.
21560 (extern debug (const bitmap_head_def *)): New.
21561 * cfg.c
21562 (debug (edge_def &)): New.
21563 (debug (edge_def *)): New.
21564 * cfghooks.c
21565 (debug (basic_block_def &)): New.
21566 (debug (basic_block_def *)): New.
21567 * dumpfile.h
21568 (dump_node (const_tree, int, FILE *)): Correct source file.
21569 * dwarf2out.c
21570 (debug (die_struct &)): New.
21571 (debug (die_struct *)): New.
21572 * dwarf2out.h
21573 (extern debug (die_struct &)): New.
21574 (extern debug (die_struct *)): New.
21575 * gimple-pretty-print.c
21576 (debug (gimple_statement_d &)): New.
21577 (debug (gimple_statement_d *)): New.
21578 * gimple-pretty-print.h
21579 (extern debug (gimple_statement_d &)): New.
21580 (extern debug (gimple_statement_d *)): New.
21581 * ira-build.c
21582 (debug (ira_allocno_copy &)): New.
21583 (debug (ira_allocno_copy *)): New.
21584 (debug (ira_allocno &)): New.
21585 (debug (ira_allocno *)): New.
21586 * ira-int.h
21587 (extern debug (ira_allocno_copy &)): New.
21588 (extern debug (ira_allocno_copy *)): New.
21589 (extern debug (ira_allocno &)): New.
21590 (extern debug (ira_allocno *)): New.
21591 * ira-lives.c
21592 (debug (live_range &)): New.
21593 (debug (live_range *)): New.
21594 * lra-int.h
21595 (debug (lra_live_range &)): New.
21596 (debug (lra_live_range *)): New.
21597 * lra-lives.c
21598 (debug (lra_live_range &)): New.
21599 (debug (lra_live_range *)): New.
21600 * omega.c
21601 (debug (omega_pb_d &)): New.
21602 (debug (omega_pb_d *)): New.
21603 * omega.h
21604 (extern debug (omega_pb_d &)): New.
21605 (extern debug (omega_pb_d *)): New.
21606 * print-rtl.c
21607 (debug (const rtx_def &)): New.
21608 (debug (const rtx_def *)): New.
21609 * print-tree.c
21610 (debug_tree (tree): Move within file.
21611 (debug_raw (const tree_node &)): New.
21612 (debug_raw (const tree_node *)): New.
21613 (dump_tree_via_hooks (const tree_node *, int)): New.
21614 (debug (const tree_node &)): New.
21615 (debug (const tree_node *)): New.
21616 (debug_verbose (const tree_node &)): New.
21617 (debug_verbose (const tree_node *)): New.
21618 (debug_head (const tree_node &)): New.
21619 (debug_head (const tree_node *)): New.
21620 (debug_body (const tree_node &)): New.
21621 (debug_body (const tree_node *)): New.
21622 (debug_vec_tree (tree): Move and reimplement in terms of dump.
21623 (debug (vec<tree, va_gc> &)): New.
21624 (debug (vec<tree, va_gc> *)): New.
21625 * rtl.h
21626 (extern debug (const rtx_def &)): New.
21627 (extern debug (const rtx_def *)): New.
21628 * sbitmap.c
21629 (debug_raw (simple_bitmap_def &)): New.
21630 (debug_raw (simple_bitmap_def *)): New.
21631 (debug (simple_bitmap_def &)): New.
21632 (debug (simple_bitmap_def *)): New.
21633 * sbitmap.h
21634 (extern debug (simple_bitmap_def &)): New.
21635 (extern debug (simple_bitmap_def *)): New.
21636 (extern debug_raw (simple_bitmap_def &)): New.
21637 (extern debug_raw (simple_bitmap_def *)): New.
21638 * sel-sched-dump.c
21639 (debug (vinsn_def &)): New.
21640 (debug (vinsn_def *)): New.
21641 (debug_verbose (vinsn_def &)): New.
21642 (debug_verbose (vinsn_def *)): New.
21643 (debug (expr_def &)): New.
21644 (debug (expr_def *)): New.
21645 (debug_verbose (expr_def &)): New.
21646 (debug_verbose (expr_def *)): New.
21647 (debug (vec<rtx> &)): New.
21648 (debug (vec<rtx> *)): New.
21649 * sel-sched-dump.h
21650 (extern debug (vinsn_def &)): New.
21651 (extern debug (vinsn_def *)): New.
21652 (extern debug_verbose (vinsn_def &)): New.
21653 (extern debug_verbose (vinsn_def *)): New.
21654 (extern debug (expr_def &)): New.
21655 (extern debug (expr_def *)): New.
21656 (extern debug_verbose (expr_def &)): New.
21657 (extern debug_verbose (expr_def *)): New.
21658 (extern debug (vec<rtx> &)): New.
21659 (extern debug (vec<rtx> *)): New.
21660 * sel-sched-ir.h
21661 (_list_iter_cond_expr): Make inline instead of static.
21662 * sreal.c
21663 (debug (sreal &)): New.
21664 (debug (sreal *)): New.
21665 * sreal.h
21666 (extern debug (sreal &)): New.
21667 (extern debug (sreal *)): New.
21668 * tree.h
21669 (extern debug_raw (const tree_node &)): New.
21670 (extern debug_raw (const tree_node *)): New.
21671 (extern debug (const tree_node &)): New.
21672 (extern debug (const tree_node *)): New.
21673 (extern debug_verbose (const tree_node &)): New.
21674 (extern debug_verbose (const tree_node *)): New.
21675 (extern debug_head (const tree_node &)): New.
21676 (extern debug_head (const tree_node *)): New.
21677 (extern debug_body (const tree_node &)): New.
21678 (extern debug_body (const tree_node *)): New.
21679 (extern debug (vec<tree, va_gc> &)): New.
21680 (extern debug (vec<tree, va_gc> *)): New.
21681 * tree-cfg.c
21682 (debug (struct loop &)): New.
21683 (debug (struct loop *)): New.
21684 (debug_verbose (struct loop &)): New.
21685 (debug_verbose (struct loop *)): New.
21686 * tree-dump.c: Add header dependence.
21687 * tree-flow.h
21688 (extern debug (struct loop &)): New.
21689 (extern debug (struct loop *)): New.
21690 (extern debug_verbose (struct loop &)): New.
21691 (extern debug_verbose (struct loop *)): New.
21692 * tree-data-ref.c
21693 (debug (data_reference &)): New.
21694 (debug (data_reference *)): New.
21695 (debug (vec<data_reference_p> &)): New.
21696 (debug (vec<data_reference_p> *)): New.
21697 (debug (vec<ddr_p> &)): New.
21698 (debug (vec<ddr_p> *)): New.
21699 * tree-data-ref.h
21700 (extern debug (data_reference &)): New.
21701 (extern debug (data_reference *)): New.
21702 (extern debug (vec<data_reference_p> &)): New.
21703 (extern debug (vec<data_reference_p> *)): New.
21704 (extern debug (vec<ddr_p> &)): New.
21705 (extern debug (vec<ddr_p> *)): New.
21706 * tree-ssa-alias.c
21707 (debug (pt_solution &)): New.
21708 (debug (pt_solution *)): New.
21709 * tree-ssa-alias.h
21710 (extern debug (pt_solution &)): New.
21711 (extern debug (pt_solution *)): New.
21712 * tree-ssa-alias.c
21713 (debug (_var_map &)): New.
21714 (debug (_var_map *)): New.
21715 (debug (tree_live_info_d &)): New.
21716 (debug (tree_live_info_d *)): New.
21717 * tree-ssa-alias.h
21718 (extern debug (_var_map &)): New.
21719 (extern debug (_var_map *)): New.
21720 (extern debug (tree_live_info_d &)): New.
21721 (extern debug (tree_live_info_d *)): New.
21722
be77e1e5
JH
217232013-03-28 Jan Hubicka <jh@suse.cz>
21724
21725 * lto-cgraph.c (merge_profile_summaries): Fix overflows.
21726
777e6976
IB
217272013-03-28 Ian Bolton <ian.bolton@arm.com>
21728
21729 * config/aarch64/aarch64.md (aarch64_can_eliminate): Keep frame
21730 record only when desired or required.
21731
9cd347ae
UB
217322013-03-28 Uros Bizjak <ubizjak@gmail.com>
21733
21734 * config/i386/i386.md (*vec_extract2vdi_1): Merge with
21735 *vec_extractv2di_1_rex64. Use x64 isa attribute.
21736
33159866 217372013-03-28 Naveen H.S <Naveen.Hurugalawadi@caviumnetworks.com>
2a1d1991
N
21738
21739 * config/aarch64/aarch64.md (*and<mode>3_compare0): New pattern.
21740 (*andsi3_compare0_uxtw): New pattern.
21741 (*and_<SHIFT:optab><mode>3_compare0): New pattern.
21742 (*and_<SHIFT:optab>si3_compare0_uxtw): New pattern.
21743
89ab31c1
JH
217442013-03-28 Jan Hubicka <jh@suse.cz>
21745
21746 * data-streamer-in.c (streamer_read_gcov_count): New function.
21747 * gimple-streamer-out.c: Include value-prof.h.
21748 (output_gimple_stmt): Output histogram.
21749 (output_bb): Use streamer_write_gcov_count.
21750 * value-prof.c: Include data-streamer.h
21751 (dump_histogram_value): Add HIST_TYPE_MAX.
21752 (stream_out_histogram_value): New function.
21753 (stream_in_histogram_value): New function.
21754 * value-prof.h (enum hist_type): Add HIST_TYPE_MAX.
21755 (stream_out_histogram_value, stream_in_histogram_value): Declare.
21756 * data-streamer-out.c (streamer_write_gcov_count): New function.
21757 (streamer_write_gcov_count_stream): New function.
21758 * lto-cgraph.c (lto_output_edge): Update counter streaming.
21759 (lto_output_node): Likewise.
21760 (input_node, input_edge): Likewise.
21761 * lto-streamer-out.c (output_cfg): Update streaming.
21762 * lto-streamer-in.c (input_cfg): Likewise.
21763 * data-streamer.h (streamer_write_gcov_count,
21764 streamer_write_gcov_count_stream, streamer_read_gcov_count): Declare.
21765 * gimple-streamer-in.c: Include value-prof.h
21766 (input_gimple_stmt): Input histograms.
21767 (input_bb): Update profile streaming.
21768
e594716a 217692013-03-28 Kenneth Zadeck <zadeck@naturalbridge.com>
b4c73eed 21770
33159866 21771 * genmodes.c (emit_max_int): New function.
8697be17 21772 (emit_insn_modes_h): Added call to emit_max_function.
39385fa6
PC
21773 * doc/rtl.texi (MAX_BITSIZE_MODE_ANY_INT, MAX_BITSIZE_MODE_ANY_MODE):
21774 Added doc.
8697be17 21775 * machmode.def: Fixed comment.
89ab31c1 21776
e594716a 217772013-03-28 Kenneth Zadeck <zadeck@naturalbridge.com>
8697be17
KZ
21778
21779 * combine.c (try_combine): Removed useless assert.
b4c73eed
KZ
21780 * cselib.c (rtx_equal_for_cselib_1): Removed unnecessary parens.
21781
74946978
MP
217822013-03-28 Marek Polacek <polacek@redhat.com>
21783 Richard Biener <rguenther@suse.de>
21784
21785 PR tree-optimization/56695
21786 * tree-vect-stmts.c (vectorizable_condition): Unconditionally
21787 build signed result of a vector comparison.
21788 * tree-cfg.c (verify_gimple_comparison): Check that a result
21789 of a vector comparison has signed type.
21790
a64b9c26
RB
217912013-03-28 Richard Biener <rguenther@suse.de>
21792
21793 PR tree-optimization/37021
21794 * tree-vect-slp.c (vect_build_slp_tree): When not unrolling
21795 do not restrict gaps between groups.
21796 * tree-vect-stmts.c (vectorizable_load): Properly account for
21797 a gap between groups.
21798
a9dc2a2f
EB
217992013-03-28 Eric Botcazou <ebotcazou@adacore.com>
21800
21801 * toplev.c (process_options): Do not disable -fomit-frame-pointer on a
21802 general basis if unwind info is requested and ACCUMULATE_OUTGOING_ARGS
21803 is not enabled.
21804
53cb97f9
GP
218052013-03-27 Gerald Pfeifer <gerald@pfeifer.com>
21806
21807 * doc/invoke.texi (AVR Options): Tweak link for AVR-LibC user manual.
21808 * doc/extend.texi (Named Address Spaces): Ditto.
21809 (Variable Attributes): Ditto.
21810
b802ae5c
KT
218112013-03-27 Kai Tietz <ktietz@redhat.com>
21812
eddae10a
KT
21813 * config.build: Add support for cygwin x64 target.
21814 * config.gcc: Likewise.
21815 * config.host: Likewise.
21816 * configure.ac: Likewise
21817 * configure: Regenerated.
b802ae5c 21818
371e77e3 218192013-03-27 Kai Tietz <ktietz@redhat.com>
8269de54
KT
21820
21821 * config/i386/cygwin-stdint.h: Add support for cygwin x64 target.
21822 * config/i386/t-cygwin-w64: New file.
21823 * config/i386/cygwin-w64.h: New file.
21824 * config/i386/cygwin.h (EXTRA_OS_CPP_BUILTINS): Extend
21825 and add support for x64-cygwin target.
21826 (CPP_SPEC): Likewise.
21827 (CXX_WRAP_SPEC_LIST): Undefine before define.
21828 (LIBGCJ_SONAME): Use 15 as version.
21829
f49b33cb
RB
218302013-03-27 Richard Biener <rguenther@suse.de>
21831
21832 PR tree-optimization/56716
21833 * tree-ssa-structalias.c (perform_var_substitution): Adjust
21834 dumping for ref nodes.
21835
b37a6ce5
MJ
218362013-03-27 Martin Jambor <mjambor@suse.cz>
21837
21838 PR tree-optimization/55334
21839 * ipa-cp.c (initialize_node_lattices): Allow IPA-CP through and to
21840 restricted pointers to arrays.
21841
9469b9b2
GDR
218422013-03-27 Gabriel Dos Reis <gdr@integrable-solutions.net>
21843
21844 * Makefile.in (.SUFFIXES): Add .cc.
21845 (.c.o): Apply same recipe for implicit rule .cc.o.
21846
7d24f650
RB
218472013-03-27 Richard Biener <rguenther@suse.de>
21848
21849 PR tree-optimization/37021
21850 * tree-vect-data-refs.c (vect_check_strided_load): Allow
21851 REALPART/IMAGPART_EXPRs around the supported refs.
21852 * tree-ssa-structalias.c (find_func_aliases): Assume that
21853 floating-point values are not used to transfer pointers.
21854
2f251a05
AI
218552013-03-27 Alexander Ivchenko <alexander.ivchenko@intel.com>
21856
f43245d1
UB
21857 * target.def (TARGET_HAS_IFUNC_P): New target hook.
21858 * doc/tm.texi.in (TARGET_HAS_IFUNC_P): New.
21859 * doc/tm.texi: Regenerate.
21860 * targhooks.h (default_has_ifunc_p): New.
21861 * targhooks.c (default_has_ifunc_p): Ditto.
21862 * config/linux-protos.h: New file.
21863 * config/linux-android.h (TARGET_HAS_IFUNC_P): Using version of this
21864 hook for linux which disables support of indirect functions in android.
21865 * config/linux-android.c: New file.
21866 * config/t-linux-android.c: Ditto.
21867 * config.gcc: Added new object file linux-android.o.
21868 * config/i386/i386.c (ix86_get_function_versions_dispatcher):
2f251a05 21869 Using TARGET_HAS_IFUNC hook instead of HAVE_GNU_INDIRECT_FUNCTION.
f43245d1 21870 * varasm.c (do_assemble_alias): Likewise.
2f251a05
AI
21871 * configure.ac: Define HAVE_GNU_INDIRECT_FUNCTION as zero if the target
21872 doesn't support indirect functions.
21873 * configure: Regenerate.
21874
78b4e425
BC
218752013-03-27 Bin Cheng <bin.cheng@arm.com>
21876
21877 PR target/56102
21878 * config/arm/arm.c (thumb1_rtx_costs, thumb1_size_rtx_costs): Fix
21879 rtx costs for SET/ASHIFT/ASHIFTRT/LSHIFTRT/ROTATERT patterns with
21880 mult-word mode.
21881
a5ba7b92
AK
218822013-03-27 Andreas Krebbel <Andreas.Krebbel@de.ibm.com>
21883
21884 * config/s390/s390.h (TARGET_FLT_EVAL_METHOD): Define.
21885
be624986
TG
218862013-03-27 Terry Guo <terry.guo@arm.com>
21887
21888 * config/arm/arm-cores.def: Added core cortex-r7.
21889 * config/arm/arm-tune.md: Regenerated.
21890 * config/arm/arm-tables.opt: Regenerated.
21891 * doc/invoke.texi: Added entry for core cortex-r7.
21892
0a514f47
WL
218932013-03-27 Walter Lee <walt@tilera.com>
21894
21895 * config/tilegx/tilegx.c (tilegx_expand_prologue): Avoid
21896 double-decrement of next_scratch_regno.
21897
9332b0d2
WL
218982013-03-27 Walter Lee <walt@tilera.com>
21899
801d9b2a 21900 * config/tilegx/tilegx.md (insn_v1mulu): Fix predicates on
9332b0d2
WL
21901 input operands.
21902 (insn_v1mulus): Ditto.
21903 (insn_v2muls): Ditto.
21904
f54ea5dd
WL
219052013-03-27 Walter Lee <walt@tilera.com>
21906
e594716a 21907 * config/tilegx/tilegx.h (ASM_OUTPUT_ADDR_VEC_ELT): Delete extra tab.
f54ea5dd
WL
21908 (ASM_OUTPUT_ADDR_DIFF_ELT): Ditto.
21909
bf60f4ca
WL
219102013-03-27 Walter Lee <walt@tilera.com>
21911
e594716a 21912 * config/tilegx/tilegx.md (*sibcall_insn): Fix type atribute for jr.
bf60f4ca
WL
21913 (*sibcall_value): Ditto.
21914
e3b51eeb
WL
219152013-03-27 Walter Lee <walt@tilera.com>
21916
21917 * config/tilegx/tilegx.md (insn_mnz_<mode>): Replaced by ...
21918 (insn_mnz_v8qi): ... this ...
21919 (insn_mnz_v4hi): ... and this. Replace (const_int 0) with the
21920 vector equivalent.
21921 (insn_v<n>mnz): Replaced by ...
21922 (insn_v1mnz): ... this ...
21923 (insn_v2mnz): ... and this. Replace (const_int 0) with the vector
21924 equivalent.
21925 (insn_mz_<mode>): Replaced by ...
21926 (insn_mz_v8qi): ... this ...
21927 (insn_mz_v4hi): ... and this. Replace (const_int 0) with the
21928 vector equivalent.
21929 (insn_v<n>mz): Replaced by ...
21930 (insn_v1mz): ... this ...
21931 (insn_v2mz): ... and this. Replace (const_int 0) with the vector
21932 equivalent.
21933
4fc7b145
EB
219342013-03-26 Eric Botcazou <ebotcazou@adacore.com>
21935
21936 * doc/invoke.texi (SPARC options): Remove -mlittle-endian.
21937
6f33817e
RM
219382013-03-26 Roland McGrath <mcgrathr@google.com>
21939
21940 * config/arm/arm.c (arm_print_operand: case 'w'): Use fputs rather
21941 than fprintf with a non-constant, non-format string.
21942
93a4145b
UB
219432013-03-26 Uros Bizjak <ubizjak@gmail.com>
21944
21945 * config/i386/i386.md (*cmpqi_ext_1): Merge with *cmpqi_ext_1_rex64
21946 using nox64 isa attribute. Use nonimmediate_x86nomem_operand as
21947 operand 0 predicate.
21948 (*cmpqi_ext_3): Merge with *cmpqi_ext_3_rex64 using nox64 isa
21949 attribute. Use general_x64nomem_operand as operand 1 predicate.
21950 (*movqi_extv_1): Merge with *movqi_extv_1_rex64 using nox64 isa
21951 attribute. Use nonimmediate_x64nomem_operand as operand 0 predicate.
21952 (*movqi_extzv_2): Merge with *movqi_extzv_2_rex64 using nox64 isa
21953 attribute. Use nonimmediate_x64nomem_operand as operand 0 predicate.
21954 (mov<mode>_insv_1): Remove expander. Merge insn with
21955 movsi_insv_1 using SWI48 mode iterator and nox64 isa attribute.
21956 Use general_x64nomem_operand as operand 1 predicate.
21957 (addqi_ext_1): Merge with *addqi_ext_1_rex64 using nox64 isa attribute.
21958 (*testqi_ext_1): Merge with *testqi_ext_1_rex64 using nox64 isa
21959 attribute. Use nonimmediate_x64nomem_operand as operand 1 predicate.
21960 (*andqi_ext_1): Merge with *andqi_ext_1_rex64 using nox64 isa
21961 attribute. Use nonimmediate_x64nomem_operand as operand 2 predicate.
21962 (*<code>qi_ext_1): Merge with *<code>qi_ext_1_rex64 using nox64 isa
21963 attribute. Use nonimmediate_x64nomem_operand as operand 1 predicate.
21964 (*xorqi_cc_ext_1): Merge with *xorqi_cc_ext_1_rex64 using nox64
21965 isa attribute. Use general_x64nomem_operand as operand 2 predicate.
21966 * config/i386/predicates.md (nonimmediate_x64nomem_operand): New.
21967 (general_x64nomem_operand): Ditto.
21968
c6a9ed5a 219692013-03-26 Sebastian Huber <sebastian.huber@embedded-brains.de>
df599020
SH
21970
21971 * config/rtems.opt: Add -pthread option.
21972
31b0a960
RB
219732013-03-26 Richard Biener <rguenther@suse.de>
21974
93a4145b
UB
21975 * alias.c (find_base_term): Avoid redundant and not used recursion.
21976 (base_alias_check): Get the initial base term from the caller.
31b0a960
RB
21977 (true_dependence_1): Compute and pass base terms to base_alias_check.
21978 (write_dependence_p): Likewise.
21979 (may_alias_p): Likewise.
21980
79517551
SN
219812013-03-26 Sofiane Naci <sofiane.naci@arm.com>
21982
21983 * config/aarch64/aarch64.c (aarch64_classify_address): Support
21984 PC-relative load in SI modes and above only.
21985
a76213b9
XQ
219862013-03-26 Xinyu Qi <xyqi@marvell.com>
21987
21988 * config/arm/arm.h (FIRST_IWMMXT_GR_REGNUM): Add comment.
33159866
UB
21989 * config/arm/iwmmxt.md (WCGR0): Update.
21990 (WCGR1, WCGR2, WCGR3): Likewise.
a76213b9 21991
37ff9355
UB
219922013-03-26 Uros Bizjak <ubizjak@gmail.com>
21993
21994 * config/i386/i386.md (*movdfcc_1): Merge with *movdfcc_1_rex64.
21995 Use x64 and nox64 isa attributes.
21996
30b0317c
RB
219972013-03-26 Richard Biener <rguenther@suse.de>
21998
21999 * emit-rtl.c (set_mem_attributes_minus_bitpos): Remove
22000 alignment computations and rely on get_object_alignment_1
22001 for the !TYPE_P case.
22002 Commonize DECL/COMPONENT_REF handling in the ARRAY_REF path.
22003
cbcdb140
WL
220042013-03-26 Walter Lee <walt@tilera.com>
22005
22006 * config/tilegx/tilegx.h (PROFILE_BEFORE_PROLOGUE): Define.
22007 * config/tilegx/tilepro.h (PROFILE_BEFORE_PROLOGUE): Define.
22008
dd884525 220092013-03-25 Jeff Law <law@redhat.com>
be672e08 22010
33159866
UB
22011 * tree-ssa-dom.c (record_equivalences_from_incoming_edge): Add missing
22012 check for INTEGRAL_TYPE_P that was missing due to checking in
dd884525
JL
22013 wrong version of prior patch.
22014
0051d3ec
WL
220152013-03-25 Walter Lee <walt@tilera.com>
22016
22017 * config/tilegx/tilegx-builtins.h (enum tilegx_builtin): Add
22018 TILEGX_INSN_SHUFFLEBYTES1.
22019 * config/tilegx/tilegx.c (tilegx_builtin_info): Add entry for
22020 shufflebytes1.
22021 (tilegx_builtins): Ditto.
22022 * config/tilegx/tilegx.md (insn_shufflebytes1): New pattern.
22023
065a3605
WL
220242013-03-25 Walter Lee <walt@tilera.com>
22025
22026 * config/tilegx/tilegx.md (floatsisf2): New pattern.
0051d3ec
WL
22027 (floatunssisf2): New pattern.
22028 (floatsidf2): New pattern.
22029 (floatunssidf2): New pattern.
065a3605 22030
5b2a3c82
WL
220312013-03-25 Walter Lee <walt@tilera.com>
22032
22033 * config/tilegx/tilegx.c (expand_set_cint64_one_inst): Inline
22034 tests for constraint J, K, N, P.
22035
192ea533
WL
220362013-03-25 Walter Lee <walt@tilera.com>
22037
22038 * config/tilegx/tilegx.c (tilegx_asm_preferred_eh_data_format):
22039 Use indirect/pcrel encoding.
22040 * config/tilepro/tilepro.c (tilepro_asm_preferred_eh_data_format):
22041 Ditto.
22042
855e0d0b
SE
220432013-03-25 Steve Ellcey <sellcey@mips.com>
22044
22045 * config/mips/mmips-cpus.def (74kc, 74kf2_1, 74kf, 74kf, 74kf1_1,
22046 74kfx, 74kx, 74kf3_2): Add PTF_AVOID_IMADD.
22047 * config/mips/mips.c (mips_option_override): Set IMADD default.
22048 * config/mips/mips.h (PTF_AVOID_IMADD): New.
22049 (ISA_HAS_MADD_MSUB): Remove MIPS16 check.
22050 (GENERATE_MADD_MSUB): Remove TUNE_74K check, add MIPS16 check.
22051 * config/mips/mips.md (mimadd): New flag for integer madd/msub.
22052 * doc/invoke.texi (-mimadd/-mno-imadd): New.
22053
39e45653
JL
220542013-03-25 Jeff Law <law@redhat.com>
22055
22056 * tree-ssa-dom.c (record_equivalences_from_incoming_edge): Rework
22057 slightly to avoid creating and folding useless trees. Simplify
22058 slightly by restricting to INTEGER_CSTs and using int_fits_type_p.
22059
7482c470
UB
220602013-03-25 Uros Bizjak <ubizjak@gmail.com>
22061
22062 * config/i386/i386.md (*zero_extendsidi2): Merge with
22063 *zero_extendsidi2_rex64. Use x64 and nox64 isa attributes.
22064 * config/i386/predicates.md (x86_64_zext_operand): Rename from
22065 x86_64_zext_general_operand. Use nonimmediate_operand on 32bit
22066 targets. Clarify comment.
22067
4a53743e
MJ
220682013-03-25 Martin Jambor <mjambor@suse.cz>
22069
22070 * ipa-prop.c (ipa_write_jump_function): Stream simple and aritmetic
22071 pass-through jump functions differently.
22072 (ipa_read_jump_function): Likewise. Also use setter functions to set
22073 up jump functions.
22074
162712de
MJ
220752013-03-25 Martin Jambor <mjambor@suse.cz>
22076
22077 * ipa-cp.c (ipa_get_indirect_edge_target): Renamed to
22078 ipa_get_indirect_edge_target_1, added parameter agg_reps and ability to
22079 process it.
22080 (ipa_get_indirect_edge_target): New function.
22081 (devirtualization_time_bonus): New parameter known_aggs, pass it to
22082 ipa_get_indirect_edge_target. Update all callers.
22083 (ipcp_discover_new_direct_edges): New parameter aggvals. Pass it to
22084 ipa_get_indirect_edge_target_1 instead of calling
22085 ipa_get_indirect_edge_target.
22086 (create_specialized_node): Pass aggvlas to
22087 ipcp_discover_new_direct_edges.
22088
a5a4c20a
KT
220892013-03-25 Kyrylo Tkachov <kyrylo.tkachov@arm.com>
22090
22091 * config/arm/arm.md (f_sels, f_seld): New types.
22092 (*cmov<mode>): New pattern.
22093 * config/arm/predicates.md (arm_vsel_comparison_operator): New
22094 predicate.
22095
f992b9fc
KT
220962013-03-25 Kai Tietz <ktietz@redhat.com>
22097
22098 * config/i386/xm-mingw32.h (__USE_MINGW_ANSI_STDIO): Enable
22099 POSIX-printf for mingw-hosted builds.
22100
c16fd676
RB
221012013-03-25 Richard Biener <rguenther@suse.de>
22102
22103 PR middle-end/56694
22104 * tree-eh.c (lower_eh_must_not_throw): Strip BLOCKs from the
22105 must-not-throw stmt location.
22106
5ad29f12
KT
221072013-03-25 Kyrylo Tkachov <kyrylo.tkachov@arm.com>
22108
22109 * config/arm/arm.c (arm_emit_load_exclusive): Add acq parameter.
22110 Emit load-acquire versions when acq is true.
22111 (arm_emit_store_exclusive): Add rel parameter.
22112 Emit store-release versions when rel is true.
22113 (arm_split_compare_and_swap): Use acquire-release instructions
22114 instead.
22115 of barriers when appropriate.
22116 (arm_split_atomic_op): Likewise.
22117 * config/arm/arm.h (TARGET_HAVE_LDACQ): New macro.
22118 * config/arm/unspecs.md (VUNSPEC_LAX): New unspec.
22119 (VUNSPEC_SLX): Likewise.
22120 (VUNSPEC_LDA): Likewise.
22121 (VUNSPEC_STL): Likewise.
22122 * config/arm/sync.md (atomic_load<mode>): New pattern.
22123 (atomic_store<mode>): Likewise.
22124 (arm_load_acquire_exclusive<mode>): Likewise.
22125 (arm_load_acquire_exclusivesi): Likewise.
22126 (arm_load_acquire_exclusivedi): Likewise.
22127 (arm_store_release_exclusive<mode>): Likewise.
22128
03a7dddb
CM
221292013-03-25 Catherine Moore <clm@codesourcery.com>
22130
22131 * config/mips/constraints.md (u, Udb7 Uead, Uean, Uesp, Uib3,
22132 Uuw6, Usb4, ZS, ZT, ZU, ZV, ZW): New constraints.
22133 * config/mip/predicates.md (lwsp_swsp_operand,
22134 lw16_sw16_operand, lhu16_sh16_operand, lbu16_operand,
22135 sb16_operand, db4_operand, db7_operand, ib3_operand,
22136 sb4_operand, ub4_operand, uh4_operand, uw4_operand,
22137 uw5_operand, uw6_operand, addiur2_operand, addiusp_operand,
22138 andi16_operand): New predicates.
22139 * config/mips/mips.md (compression): New attribute.
22140 (enabled): New attribute.
22141 (length): Consider compression in computing length.
22142 (shift_compression): New code attribute.
22143 (*add<mode>3): New operands. Record compression.
22144 (sub<mode>3): Likewise.
22145 (one_cmpl<mode>2): Likewise.
22146 (*and<mode>3): Likewise.
22147 (*ior<mode>3): Likewise.
22148 (unnamed pattern for xor): Likewise.
22149 (*zero_extend<SHORT:mode><GPR:mode>2): Likewise.
22150 (*<optab><mode>3): Likewise.
22151 (*mov<mode>_internal: Likewise.
22152 * config/mips/mips-protos.h (mips_signed_immediate_p): New.
22153 (mips_unsigned_immediate_p): New.
22154 (umips_lwsp_swsp_address_p): New.
22155 (m16_based_address_p): New.
22156 * config/mips/mips-protos.h (mips_signed_immediate_p): New prototype.
22157 (mips_unsigned_immediate_p): New prototype.
22158 (lwsp_swsp_address_p): New prototype.
22159 (m16_based_address_p): New prototype.
22160 * config/mips/mips.c (mips_unsigned_immediate_p): New function.
22161 (mips_signed_immediate_p): New function.
22162 (m16_based_address_p): New function.
22163 (lwsp_swsp_address_p): New function.
22164 (mips_print_operand_punctuation): Recognize short delay slot insns
22165 for microMIPS.add<mode>3"
22166
f35c297f 221672013-03-25 Kyrylo Tkachov <kyrylo.tkachov@arm.com>
a5784ce6 22168
f35c297f
KT
22169 PR target/56720
22170 * config/arm/iterators.md (v_cmp_result): New mode attribute.
22171 * config/arm/neon.md (vcond<mode><mode>): Handle unordered cases.
22172
051b9446
RB
221732013-03-25 Richard Biener <rguenther@suse.de>
22174
22175 PR tree-optimization/56689
22176 * tree-vrp.c (execute_vrp): Mark loops for fixup if we removed
22177 any edge.
22178
374001cb
RB
221792013-03-25 Richard Biener <rguenther@suse.de>
22180
22181 * tree-ssa-loop-im.c (struct mem_ref): Use bitmap_head instead
22182 of bitmap.
22183 (memory_references): Likewise.
22184 (outermost_indep_loop, mem_ref_alloc, mark_ref_stored,
22185 gather_mem_refs_stmt, record_dep_loop, ref_indep_loop_p_1,
22186 ref_indep_loop_p_2, find_refs_for_sm): Adjust.
22187 (gather_mem_refs_in_loops): Fold into ...
22188 (analyze_memory_references): ... this. Move initialization
22189 to tree_ssa_lim_initialize.
22190 (fill_always_executed_in): Rename to ...
22191 (fill_always_executed_in_1): ... this.
22192 (fill_always_executed_in): Move contains_call computation to
22193 this new function from ...
22194 (tree_ssa_lim_initialize): ... here.
22195 (tree_ssa_lim): Call fill_always_executed_in.
22196
57895947
EB
221972013-03-25 Eric Botcazou <ebotcazou@adacore.com>
22198
22199 * postreload.c (reload_combine): Fix code detecting returns.
22200
1e1b18c1
EB
222012013-03-25 Eric Botcazou <ebotcazou@adacore.com>
22202
22203 * function.c (emit_use_return_register_into_block): On cc0 targets,
22204 do not emit the sequence between cc0 setter and user.
22205
9216baf1
KT
222062013-03-25 Kai Tietz <ktietz@redhat.com>
22207
1e1b18c1
EB
22208 * config/i386/predicates.md (local_symbolic_operand): Interpret
22209 dll-imported symbols as none-local.
9216baf1 22210
76421b44
RB
222112013-03-25 Richard Biener <rguenther@suse.de>
22212
22213 * tree-ssa-loop-im.c (struct depend): Remove.
22214 (struct lim_aux_data): Make depends a vec of gimples.
22215 (free_lim_aux_data): Adjust.
22216 (add_dependency): Likewise.
22217 (set_level): Likewise.
22218
d154bfa2
RB
222192013-03-25 Richard Biener <rguenther@suse.de>
22220
22221 PR middle-end/56434
22222 * calls.c (expand_call): Use MALLOC_ABI_ALIGNMENT to annotate
22223 the pointer returned by calls with ECF_MALLOC set.
22224
2cd9804e
UB
222252013-03-24 Uros Bizjak <ubizjak@gmail.com>
22226
acef5fe0
UB
22227 * config/i386/mmx.md (mov<mode>): Add ?!Ym,r and r,?!Ym alternatives.
22228
222292013-03-24 Uros Bizjak <ubizjak@gmail.com>
22230
22231 * config/i386/mmx.md (mov<mode>): Merge with movv2sf expander
2cd9804e
UB
22232 using MMXMODE mode iterator.
22233 (*move<mode>_internal): Merge with *movv2sf_internal and
22234 *movv2sf_internal_rex64 using MMXMODE mode iterator.
22235
225ccc68
SB
222362013-03-23 Steven Bosscher <steven@gcc.gnu.org>
22237
85c0f02d
SB
22238 * gcse.c (oprs_unchanged_p): Respect flag_gcse_lm.
22239 (record_last_mem_set_info): Likewise.
22240
225ccc68
SB
22241 * df-core.c (rest_of_handle_df_initialize): Use XCNEWVEC instead
22242 of XNEWVEC followed by memset.
22243 (df_worklist_dataflow): Use XNEWVEC instead of xmalloc with a cast.
22244
b64925dc
SB
222452013-03-23 Steven Bosscher <steven@gcc.gnu.org>
22246
22247 * config/avr/avr.c, config/bfin/bfin.c, config/c6x/c6x.c,
22248 config/epiphany/epiphany.c, config/frv/frv.c, config/ia64/ia64.c,
22249 config/iq2000/iq2000.c, config/mcore/mcore.c, config/mep/mep.c,
22250 config/mmix/mmix.c, config/pa/pa.c, config/rs6000/rs6000.c,
22251 config/s390/s390.c, config/sparc/sparc.c, config/spu/spu.c,
22252 config/stormy16/stormy16.c, config/v850/v850.c, config/xtensa/xtensa.c,
22253 dwarf2out.c, hw-doloop.c, resource.c, rtl.h : Where applicable, use
22254 the predicates NOTE_P, NONJUMP_INSN_P, JUMP_P, CALL_P, LABEL_P, and
22255 BARRIER_P instead of GET_CODE.
22256
b9aaf52e
EB
222572013-03-23 Eric Botcazou <ebotcazou@adacore.com>
22258
22259 * config/sparc/sparc.c (sparc_emit_probe_stack_range): Fix small
22260 inaccuracy in the probing code.
22261
22262 * config/sparc/sparc.md (ctrapsi4): Add predicate for operand #3.
22263 (ctrapdi4): Likewise.
22264
66de4d7c
EB
222652013-03-23 Eric Botcazou <ebotcazou@adacore.com>
22266
22267 * calls.c (expand_call): Add missing guard to code handling return
22268 of non-BLKmode structures in MSB.
22269 * function.c (expand_function_end): Likewise.
22270
4f25c8fb
EB
222712013-03-23 Eric Botcazou <ebotcazou@adacore.com>
22272
22273 * combine.c (try_combine): Adjust comment. Do not add the set of
22274 insn #0 if the destination indirectly is set or dies in insn #2.
22275 Tidy up code to distribute a new note.
22276
23b7850d
UB
222772013-03-22 Uros Bizjak <ubizjak@gmail.com>
22278
22279 * config/i386/i386.md (*movdi_internal): Set prefix_rex attribute
22280 also for alternatives 16 and 17.
22281
bed852cf
UB
222822013-03-22 Uros Bizjak <ubizjak@gmail.com>
22283
22284 * config/i386/sse.md (*mov<mode>_internal): Merge with
22285 *mov<mode>_internal_rex64. Use x64 and nox64 isa attributes.
22286 Emit insn template depending on type attribute. Use
22287 HAVE_AS_IX86_INTERUNIT_MOVQ to handle broken assemblers that require
22288 movd instead of movq mnemonic for interunit moves. Rewrite mode
22289 attribute calculation. Remove unit attribute calculation.
22290 Set prefix attribute to maybe_vex for sselog1 and ssemov types.
22291 Set prefix_data16 attribute for DImode ssemov types.
22292 Use Ym instead of y for SSE-MMX conversion alternatives.
22293 Reorder operand constraints.
22294
1e8a7937
SB
222952013-03-22 Steven Bosscher <steven@gcc.gnu.org>
22296
22297 * df.h (df_insn_delete): Adjust prototype.
22298 * emit-rtl.c (remove_insn): Pass a basic block to df_insn_delete
22299 and let it decide whether mark the basic block dirty.
22300 (set_insn_deleted): Only pass INSN_P insns to df_insn_delete.
22301 * df-scan.c (df_insn_info_delete): New helper function, split
22302 off from df_insn_delete.
22303 (df_scan_free_bb_info): Use it.
22304 (df_insn_rescan, df_insn_rescan_all, df_process_deferred_rescans):
22305 Likewise.
22306 (df_insn_delete): Likewise. Take insn rtx as argument. Verify
22307 that the insn is actually an insn and it has a non-NULL basic block.
22308 Do not mark basic block dirty if only deleting a DEBUG_INSN.
22309
e14d094c
RB
223102013-03-22 Richard Biener <rguenther@suse.de>
22311
22312 * tree-ssa-loop-im.c (struct mem_ref): Remove indep_ref and
22313 dep_ref members.
22314 (mem_ref_alloc): Do not allocate them.
22315 (refs_independent_p): Do not query or maintain a cache.
22316
e9cf7316
RB
223172013-03-22 Richard Biener <rguenther@suse.de>
22318
22319 * tree-ssa-loop-im.c (memory_references): Drop all_refs_in_loop.
22320 (gather_mem_refs_in_loops): Do not compute it.
22321 (analyze_memory_references): Do not allocate it.
22322 (tree_ssa_lim_finalize): Do not free it.
22323 (for_all_locs_in_loop): Do not query all_refs_in_loop.
22324
f046e81b
RB
223252013-03-22 Richard Biener <rguenther@suse.de>
22326
22327 * is-a.h (as_a): Use gcc_checking_assert.
bed852cf 22328
a05c0ddf
IB
223292013-03-22 Ian Bolton <ian.bolton@arm.com>
22330
22331 * config/aarch64/aarch64.c (aarch64_print_operand): New
22332 format specifier for printing a constant in hex.
22333 * config/aarch64/aarch64.md (insv_imm<mode>): Use the X
22334 format specifier for printing second operand.
22335
c00217fc
RB
223362013-03-22 Richard Biener <rguenther@suse.de>
22337
22338 * tree-ssa-loop-im.c (memory_references): Add refs_stored_in_loop
22339 bitmaps.
22340 (gather_mem_refs_in_loops): Perform store accumulation here.
22341 (create_vop_ref_mapping_loop): Remove.
22342 (create_vop_ref_mapping): Likewise.
22343 (analyze_memory_references): Initialize refs_stored_in_loop.
bed852cf 22344 (LOOP_DEP_BIT): New define to map to bits in (in)dep_loop bitmaps.
c00217fc
RB
22345 (record_indep_loop): Remove.
22346 (record_dep_loop): New function.
22347 (ref_indep_loop_p_1): Adjust to only walk over references
22348 in the loop, not its subloops.
22349 (ref_indep_loop_p): Rename to ...
22350 (ref_indep_loop_p_2): ... this and recurse over the loop tree,
22351 maintaining a more fine-grained cache.
22352 (ref_indep_loop_p): Wrap ref_indep_loop_p_2.
22353 (tree_ssa_lim_finalize): Free refs_stored_in_loop.
22354
15d19bf8
RB
223552013-03-22 Richard Biener <rguenther@suse.de>
22356
22357 * tree-ssa-loop-im.c (struct mem_ref_locs): Remove.
22358 (struct mem_ref): Make accesses_in_loop a vec of a vec of
22359 aggregate mem_ref_loc.
22360 (free_mem_ref_locs): Inline into ...
22361 (memref_free): ... this and adjust.
22362 (mem_ref_alloc): Adjust.
22363 (mem_ref_locs_alloc): Remove.
22364 (record_mem_ref_loc): Adjust.
22365 (get_all_locs_in_loop): Rewrite into ...
22366 (for_all_locs_in_loop): ... this iterator.
22367 (rewrite_mem_ref_loc): New functor.
22368 (rewrite_mem_refs): Use for_all_locs_in_loop.
22369 (sm_set_flag_if_changed): New functor.
22370 (execute_sm_if_changed_flag_set): Use for_all_locs_in_loop.
22371 (ref_always_accessed): New functor.
22372 (ref_always_accessed_p): Use for_all_locs_in_loop.
22373
6f37411d
MG
223742013-03-21 Marc Glisse <marc.glisse@inria.fr>
22375
22376 * tree-pass.h (PROP_gimple_lvec): New.
22377 * passes.c (dump_properties): Handle PROP_gimple_lvec.
22378 (init_optimization_passes): Move pass_lower_vector.
22379 * tree-vect-generic.c (gate_expand_vector_operations_ssa): Test
22380 PROP_gimple_lvec.
22381 (pass_lower_vector): Provide PROP_gimple_lvec.
22382 (pass_lower_vector_ssa): Likewise.
22383 * cfgexpand.c (pass_expand): Require PROP_gimple_lvec.
22384
511ef689
MW
223852013-03-21 Mark Wielaard <mjw@redhat.com>
22386
22387 * dwarf2out.c (size_of_aranges): Skip DECL_IGNORED_P functions.
22388
cc1df30b
UB
223892013-03-21 Uros Bizjak <ubizjak@gmail.com>
22390
22391 * config/i386/i386.md (*movdi_internal): Disparage slightly
22392 all MMX moves to/from memory. Use Yi instead of x for SSE-MMX
22393 conversion alternatives.
22394
c845cfe1
JJ
223952013-03-21 Jakub Jelinek <jakub@redhat.com>
22396
37e99116
JJ
22397 PR middle-end/48087
22398 * diagnostic.def (DK_WERROR): New kind.
22399 * diagnostic.h (werrorcount): Define.
22400 * diagnostic.c (diagnostic_report_diagnostic): For DK_WARNING
22401 promoted to DK_ERROR, increment DK_WERROR counter instead of
22402 DK_ERROR counter.
22403 * toplev.c (toplev_main): Call print_ignored_options even if
22404 just werrorcount is non-zero. Exit with FATAL_EXIT_CODE
22405 even if just werrorcount is non-zero.
22406
c845cfe1
JJ
22407 PR debug/55608
22408 * dwarf2out.c (tree_add_const_value_attribute): Call ggc_free (array)
22409 on failure.
22410 (resolve_one_addr): Fail if referenced STRING_CST hasn't been written.
22411 (string_cst_pool_decl): New function.
22412 (optimize_one_addr_into_implicit_ptr): New function.
22413 (resolve_addr_in_expr): Optimize DWARF location expression
22414 DW_OP_addr DW_OP_stack_value where DW_OP_addr refers to some variable
22415 which doesn't live in memory, but has DW_AT_location or
22416 DW_AT_const_value, or refers to a string literal, into
22417 DW_OP_GNU_implicit_pointer.
22418 (optimize_location_into_implicit_ptr): New function.
22419 (resolve_addr): If removing DW_AT_location of a variable because
22420 it was DW_OP_addr of address of the variable, but the variable doesn't
22421 live in memory, try to emit const value attribute for the initializer.
22422
08e0cda6
MG
224232013-03-21 Marc Glisse <marc.glisse@inria.fr>
22424
22425 * tree.h (VECTOR_TYPE_P): New macro.
22426 (VECTOR_INTEGER_TYPE_P, VECTOR_FLOAT_TYPE_P, FLOAT_TYPE_P,
22427 TYPE_MODE): Use it.
22428 * fold-const.c (fold_cond_expr_with_comparison): Use build_zero_cst.
22429 VEC_COND_EXPR cannot be lvalues.
22430 (fold_ternary_loc) <VEC_COND_EXPR>: Merge with the COND_EXPR case.
22431
d08633b4
MG
224322013-03-21 Marc Glisse <marc.glisse@inria.fr>
22433
22434 * simplify-rtx.c (simplify_binary_operation_1) <VEC_CONCAT>:
22435 Restrict the transformation to equal modes.
22436
e6647190
RB
224372013-03-21 Richard Biener <rguenther@suse.de>
22438
22439 PR tree-optimization/39326
22440 * tree-ssa-loop-im.c (UNANALYZABLE_MEM_ID): New define.
22441 (MEM_ANALYZABLE): Adjust.
22442 (record_mem_ref_loc): Move bitmap ops ...
22443 (gather_mem_refs_stmt): ... here. Use the shared mem-ref for
22444 unanalyzable refs, do not record locations for it.
22445 (analyze_memory_references): Allocate ref zero as shared
22446 unanalyzable ref.
22447 (refs_independent_p): Do not test for unanalyzed mems here.
22448 (ref_indep_loop_p_1): Special-case disambiguation against
22449 the unanalyzed ref.
cc1df30b 22450 (ref_indep_loop_p): Assert we are not queried for the unanalyzed mem.
e6647190 22451
65074f54
CL
224522013-03-21 Christophe Lyon <christophe.lyon@linaro.org>
22453
22454 * config/arm/arm-protos.h (tune_params): Add
22455 prefer_neon_for_64bits field.
22456 * config/arm/arm.c (prefer_neon_for_64bits): New variable.
22457 (arm_slowmul_tune): Default prefer_neon_for_64bits to false.
22458 (arm_fastmul_tune, arm_strongarm_tune, arm_xscale_tune): Ditto.
22459 (arm_9e_tune, arm_v6t2_tune, arm_cortex_tune): Ditto.
22460 (arm_cortex_a15_tune, arm_cortex_a5_tune): Ditto.
22461 (arm_cortex_a9_tune, arm_v6m_tune, arm_fa726te_tune): Ditto.
22462 (arm_option_override): Handle -mneon-for-64bits new option.
22463 * config/arm/arm.h (TARGET_PREFER_NEON_64BITS): New macro.
22464 (prefer_neon_for_64bits): Declare new variable.
22465 * config/arm/arm.md (arch): Rename neon_onlya8 and neon_nota8 to
cc1df30b 22466 avoid_neon_for_64bits and neon_for_64bits. Remove onlya8 and nota8.
65074f54
CL
22467 (arch_enabled): Handle new arch types. Remove support for onlya8
22468 and nota8.
22469 (one_cmpldi2): Use new arch names.
bc5faa5b 22470 (zero_extend<mode>di2, extend<mode>di2): Ditto.
65074f54
CL
22471 * config/arm/arm.opt (mneon-for-64bits): Add option.
22472 * config/arm/neon.md (adddi3_neon, subdi3_neon, iordi3_neon)
22473 (anddi3_neon, xordi3_neon, ashldi3_neon, <shift>di3_neon): Use
22474 neon_for_64bits instead of nota8 and avoid_neon_for_64bits instead
22475 of onlya8.
22476 * doc/invoke.texi (-mneon-for-64bits): Document.
22477
5a2d2a79
RB
224782013-03-21 Richard Biener <rguenther@suse.de>
22479
22480 PR tree-optimization/39326
22481 * tree-ssa-loop-im.c (bb_loop_postorder): New global static.
22482 (sort_bbs_in_loop_postorder_cmp): New function.
cc1df30b 22483 (gather_mem_refs_in_loops): Assign mem-ref IDs in loop postorder.
5a2d2a79 22484
5abe1e05
RB
224852013-03-21 Richard Biener <rguenther@suse.de>
22486
22487 * tree-vect-data-refs.c (vect_update_interleaving_chain): Remove.
22488 (vect_insert_into_interleaving_chain): Likewise.
22489 (vect_drs_dependent_in_basic_block): Inline ...
22490 (vect_slp_analyze_data_ref_dependence): ... here. New function,
22491 split out from ...
22492 (vect_analyze_data_ref_dependence): ... here. Simplify.
22493 (vect_check_interleaving): Simplify.
22494 (vect_analyze_data_ref_dependences): Likewise. Split out ...
22495 (vect_slp_analyze_data_ref_dependences): ... this new function.
22496 (dr_group_sort_cmp): New function.
22497 (vect_analyze_data_ref_accesses): Compute data-reference groups
22498 here instead of in vect_analyze_data_ref_dependence. Use
22499 a more efficient algorithm.
22500 * tree-vect-slp.c (vect_slp_analyze_bb_1): Use
22501 vect_slp_analyze_data_ref_dependences. Call
22502 vect_analyze_data_ref_accesses earlier.
22503 * tree-vect-loop.c (vect_analyze_loop_2): Likewise.
22504 * tree-vectorizer.h (vect_analyze_data_ref_dependences): Adjust.
22505 (vect_slp_analyze_data_ref_dependences): New prototype.
22506
cad1735b
RB
225072013-03-21 Richard Biener <rguenther@suse.de>
22508
22509 * tree-ssa-loop-im.c (can_sm_ref_p): Do not test whether
22510 ref is stored in the loop.
22511 (find_refs_for_sm): Walk only over all stores.
22512 (store_motion_loop): Allocate from lim_bitmap_obstack.
22513 (store_motion): Likewise.
22514
141310ef
RB
225152013-03-21 Richard Biener <rguenther@suse.de>
22516
22517 * tree-vect-loop-manip.c (slpeel_tree_peel_loop_to_edge):
22518 Update virtual SSA form.
22519
5022315a
RO
225202013-03-21 Rainer Orth <ro@CeBiTec.Uni-Bielefeld.DE>
22521
22522 * configure.ac (gcc_cv_ld_eh_frame_ciev3): New test.
22523 * configure: Regenerate.
22524 * config.in: Regenerate.
22525 * config/sol2.c (solaris_override_options): Only enforce DWARF 2
22526 if !HAVE_LD_EH_FRAME_CIEV3.
22527
50d4421c
RB
225282013-03-21 Richard Biener <rguenther@suse.de>
22529
22530 * tree-cfg.c (verify_expr_no_block): New function.
22531 (verify_expr_location_1): Verify that neither DECL_DEBUG_EXPR
22532 nor DECL_VALUE_EXPR have locations with associated blocks.
22533 * tree-ssa-live.c (clear_unused_block_pointer_1): Remove.
22534 (clear_unused_block_pointer): Remove code dealing with
22535 blocks in DECL_DEBUG_EXPR locations.
22536
839b422f
RB
225372013-03-21 Richard Biener <rguenther@suse.de>
22538
22539 * tree.h (DECL_DEBUG_EXPR_IS_FROM): Rename to ...
22540 (DECL_HAS_DEBUG_EXPR_P): ... this. Guard properly.
22541 * tree.c (copy_node_stat): Do not copy DECL_HAS_DEBUG_EXPR_P.
22542 * dwarf2out.c (add_var_loc_to_decl): Use DECL_HAS_DEBUG_EXPR_P
22543 instead of DECL_DEBUG_EXPR_IS_FROM.
22544 * gimplify.c (gimplify_modify_expr): Likewise.
22545 * tree-cfg.c (verify_expr_location_1): Likewise.
22546 * tree-complex.c (create_one_component_var): Likewise.
22547 * tree-sra.c (create_access_replacement): Likewise.
22548 * tree-ssa-live.c (clear_unused_block_pointer_1): Likewise.
22549 (clear_unused_block_pointer): Likewise.
22550 * tree-streamer-in.c (unpack_ts_decl_common_value_fields): Likewise.
22551 * tree-streamer-out.c (pack_ts_decl_common_value_fields): Likewise.
22552 * var-tracking.c (var_debug_decl): Likewise.
22553 (track_expr_p): Likewise.
22554 * tree-inline.c (add_local_variables): Likewise. Set
22555 DECL_HAS_DEBUG_EXPR_P after copying it.
22556 * tree-diagnostic.c (default_tree_printer): Use DECL_HAS_DEBUG_EXPR_P
22557 instead of DECL_DEBUG_EXPR_IS_FROM. Guard properly.
22558
fe04878d
UB
225592013-03-21 Uros Bizjak <ubizjak@gmail.com>
22560
22561 PR bootstrap/56656
22562 * configure.ac (HAVE_AS_IX86_INTERUNIT_MOVQ): New test.
22563 * configure: Regenerate.
22564 * config.in: Regenerate.
22565 * config/i386/i386.md (*movdf_internal): Use
22566 HAVE_AS_IX86_INTERUNIT_MOVQ to handle broken assemblers that require
22567 movd instead of movq mnemonic for interunit moves.
22568 (*movdi_internal): Ditto.
22569
bd059b26 225702013-03-21 Naveen H.S <Naveen.Hurugalawadi@caviumnetworks.com>
fa2e9a58
N
22571
22572 * config/aarch64/aarch64-simd.md (simd_fabd): New Attribute.
22573 (abd<mode>_3): New pattern.
22574 (aba<mode>_3): New pattern.
22575 (fabd<mode>_3): New pattern.
22576
bd059b26 225772013-03-21 Naveen H.S <Naveen.Hurugalawadi@caviumnetworks.com>
50ce6f88
N
22578
22579 * config/aarch64/aarch64-elf.h (REGISTER_PREFIX): Remove.
22580 * config/aarch64/aarch64.c (aarch64_print_operand): Remove all
22581 occurrence of REGISTER_PREFIX as its empty string.
22582
4f1f78b9
JL
225832013-03-20 Jeff Law <law@redhat.com>
22584
22585 * tree-ssa-dom.c (record_equivalences_from_incoming_edge): Record
22586 addititional equivalences for equality comparisons between an SSA_NAME
22587 and a constant where the SSA_NAME was set from a widening conversion.
22588
327a1118
WL
225892013-03-20 Walter Lee <walt@tilera.com>
22590
22591 * config/tilegx/sync.md (atomic_test_and_set): New pattern.
22592
813e0036
UB
225932013-03-20 Uros Bizjak <ubizjak@gmail.com>
22594
22595 * config/i386/i386.md (*movoi_internal_avx): Emit insn template
22596 depending on type attribute.
22597 (*movti_internal): Ditto.
22598 (*movtf_internal): Ditto.
22599 (*movxf_internal): Ditto.
22600 (*movdf_internal): Ditto.
22601 (*movsf_internal): Ditto.
22602
7cf34aae
UB
226032013-03-20 Uros Bizjak <ubizjak@gmail.com>
22604
22605 * config/i386/i386.md (*movti_internal): Set prefix attribute to
22606 maybe_vex for sselog1 and ssemov types.
22607 (*movdi_internal): Reorder operand constraints.
22608 (*movsi_internal): Ditto. Set prefix attribute to
22609 maybe_vex for sselog1 and ssemov types.
22610 (*movtf_internal): Set prefix attribute to maybe_vex
22611 for sselog1 and ssemov types.
22612 (*movdf_internal): Ditto. Set prefix_data16 attribute for
22613 DImode ssemov types. Reorder operand constraints.
22614 (*movsf_internal): Set type of alternatives 3,4 to imov. Set prefix
22615 attribute to maybe_vex for sselog1 and ssemov types. Set prefix_data16
22616 attribute for SImode ssemov types. Reorder operand constraints.
22617
19321415
MJ
226182013-03-20 Martin Jambor <mjambor@suse.cz>
22619
22620 * params.def (PARAM_IPA_CP_ARRAY_INDEX_HINT_BONUS): New parameter.
22621 * ipa-cp.c (hint_time_bonus): Add abonus for known array indices.
22622
12211b99 226232013-03-20 Pat Haugen <pthaugen@us.ibm.com>
c6d5ff83
MM
22624
22625 * config/rs6000/predicates.md (indexed_address, update_address_mem
22626 update_indexed_address_mem): New predicates.
22627 * config/rs6000/vsx.md (vsx_extract_<mode>_zero): Set correct "type"
22628 attribute for load/store instructions.
22629 * config/rs6000/dfp.md (movsd_store): Likewise.
22630 (movsd_load): Likewise.
22631 * config/rs6000/rs6000.md (zero_extend<mode>di2_internal1): Likewise.
22632 (unnamed HI->DI extend define_insn): Likewise.
22633 (unnamed SI->DI extend define_insn): Likewise.
22634 (unnamed QI->SI extend define_insn): Likewise.
22635 (unnamed QI->HI extend define_insn): Likewise.
22636 (unnamed HI->SI extend define_insn): Likewise.
22637 (unnamed HI->SI extend define_insn): Likewise.
22638 (extendsfdf2_fpr): Likewise.
22639 (movsi_internal1): Likewise.
22640 (movsi_internal1_single): Likewise.
22641 (movhi_internal): Likewise.
22642 (movqi_internal): Likewise.
22643 (movcc_internal1): Correct mnemonic for stw insn. Set correct "type"
22644 attribute for load/store instructions.
22645 (mov<mode>_hardfloat): Set correct "type" attribute for load/store
22646 instructions.
22647 (mov<mode>_softfloat): Likewise.
22648 (mov<mode>_hardfloat32): Likewise.
22649 (mov<mode>_hardfloat64): Likewise.
22650 (mov<mode>_softfloat64): Likewise.
22651 (movdi_internal32): Likewise.
22652 (movdi_internal64): Likewise.
22653 (probe_stack_<mode>): Likewise.
22654
226552013-03-20 Michael Meissner <meissner@linux.vnet.ibm.com>
22656
22657 * config/rs6000/vector.md (VEC_R): Add 32-bit integer, binary
22658 floating point, and decimal floating point to reload iterator.
22659
22660 * config/rs6000/constraints.md (wl constraint): New constraints to
22661 return FLOAT_REGS if certain options are used to reduce the number
22662 of separate patterns that exist in the file.
22663 (wx constraint): Likewise.
22664 (wz constraint): Likewise.
22665
22666 * config/rs6000/rs6000.c (rs6000_debug_reg_global): If
22667 -mdebug=reg, print wg, wl, wx, and wz constraints.
22668 (rs6000_init_hard_regno_mode_ok): Initialize new constraints.
22669 Initialize the reload functions for 64-bit binary/decimal floating
22670 point types.
22671 (reg_offset_addressing_ok_p): If we are on a power7 or later, use
22672 LFIWZX and STFIWX to load/store 32-bit decimal types, and don't
22673 create the buffer on the stack to overcome not having a 32-bit
22674 load and store.
22675 (rs6000_emit_move): Likewise.
22676 (rs6000_secondary_memory_needed_rtx): Likewise.
22677 (rs6000_alloc_sdmode_stack_slot): Likewise.
22678 (rs6000_preferred_reload_class): On VSX, we can create SFmode 0.0f
22679 via xxlxor, just like DFmode 0.0.
22680
22681 * config/rs6000/rs6000.h (TARGET_NO_SDMODE_STACK): New macro,
22682 define as 1 if we are running on a power7 or newer.
22683 (enum r6000_reg_class_enum): Add new constraints.
22684
22685 * config/rs6000/dfp.md (movsd): Delete, combine with binary
22686 floating point moves in rs6000.md. Combine power6x (mfpgpr) moves
22687 with other moves by using conditional constraits (wg). Use LFIWZX
bd059b26 22688 and STFIWX for loading SDmode on power7. Use xxlxor to create 0.0f.
c6d5ff83
MM
22689 (movsd splitter): Likewise.
22690 (movsd_hardfloat): Likewise.
22691 (movsd_softfloat): Likewise.
22692
22693 * config/rs6000/rs6000.md (FMOVE32): New iterators to combine
22694 binary and decimal floating point moves.
22695 (fmove_ok): New attributes to combine binary and decimal floating
22696 point moves, and to combine power6x (mfpgpr) moves along normal
22697 floating moves.
22698 (real_value_to_target): Likewise.
22699 (f32_lr): Likewise.
22700 (f32_lm): Likewise.
22701 (f32_li): Likewise.
22702 (f32_sr): Likewise.
22703 (f32_sm): Likewise.
22704 (f32_si): Likewise.
22705 (movsf): Combine binary and decimal floating point moves. Combine
22706 power6x (mfpgpr) moves with other moves by using conditional
bd059b26 22707 constraits (wg). Use LFIWZX and STFIWX for loading SDmode on power7.
c6d5ff83
MM
22708 (mov<mode> for SFmode/SDmode); Likewise.
22709 (SFmode/SDmode splitters): Likewise.
22710 (movsf_hardfloat): Likewise.
22711 (mov<mode>_hardfloat for SFmode/SDmode): Likewise.
22712 (movsf_softfloat): Likewise.
22713 (mov<mode>_softfloat for SFmode/SDmode): Likewise.
22714
22715 * doc/md.texi (PowerPC and IBM RS6000 constraints): Document wl,
22716 wx and wz constraints.
22717
22718 * config/rs6000/constraints.md (wg constraint): New constraint to
22719 return FLOAT_REGS if -mmfpgpr (power6x) was used.
22720
22721 * config/rs6000/rs6000.h (enum r6000_reg_class_enum): Add wg
22722 constraint.
22723
22724 * config/rs6000/rs6000.c (rs6000_debug_reg_global): If
22725 -mdebug=reg, print wg, wl, wx, and wz constraints.
22726 (rs6000_init_hard_regno_mode_ok): Initialize new constraints.
22727 Initialize the reload functions for 64-bit binary/decimal floating
22728 point types.
22729 (reg_offset_addressing_ok_p): If we are on a power7 or later, use
22730 LFIWZX and STFIWX to load/store 32-bit decimal types, and don't
22731 create the buffer on the stack to overcome not having a 32-bit
22732 load and store.
22733 (rs6000_emit_move): Likewise.
22734 (rs6000_secondary_memory_needed_rtx): Likewise.
22735 (rs6000_alloc_sdmode_stack_slot): Likewise.
22736 (rs6000_preferred_reload_class): On VSX, we can create SFmode 0.0f
22737 via xxlxor, just like DFmode 0.0.
22738
c6d5ff83
MM
22739 * config/rs6000/dfp.md (movdd): Delete, combine with binary
22740 floating point moves in rs6000.md. Combine power6x (mfpgpr) moves
22741 with other moves by using conditional constraits (wg). Use LFIWZX
22742 and STFIWX for loading SDmode on power7.
22743 (movdd splitters): Likewise.
22744 (movdd_hardfloat32): Likewise.
22745 (movdd_softfloat32): Likewise.
22746 (movdd_hardfloat64_mfpgpr): Likewise.
22747 (movdd_hardfloat64): Likewise.
22748 (movdd_softfloat64): Likewise.
22749
22750 * config/rs6000/rs6000.md (FMOVE64): New iterators to combine
22751 64-bit binary and decimal floating point moves.
22752 (FMOVE64X): Likewise.
22753 (movdf): Combine 64-bit binary and decimal floating point moves.
22754 Combine power6x (mfpgpr) moves with other moves by using
22755 conditional constraits (wg).
22756 (mov<mode> for DFmode/DDmode): Likewise.
22757 (DFmode/DDmode splitters): Likewise.
22758 (movdf_hardfloat32): Likewise.
22759 (mov<mode>_hardfloat32 for DFmode/DDmode): Likewise.
22760 (movdf_softfloat32): Likewise.
22761 (movdf_hardfloat64_mfpgpr): Likewise.
22762 (movdf_hardfloat64): Likewise.
22763 (mov<mode>_hardfloat64 for DFmode/DDmode): Likewise.
22764 (movdf_softfloat64): Likewise.
22765 (mov<mode>_softfloat64 for DFmode/DDmode): Likewise.
22766 (reload_<mode>_load): Move to later in the file so they aren't in
22767 the middle of the floating point move insns.
22768 (reload_<mode>_store): Likewise.
22769
22770 * doc/md.texi (PowerPC and IBM RS6000 constraints): Document wg
22771 constraint.
22772
22773 * config/rs6000/rs6000.c (rs6000_debug_reg_global): Print out wg
22774 constraint if -mdebug=reg.
bd059b26
UB
22775 (rs6000_initi_hard_regno_mode_ok): Enable wg constraint if -mfpgpr.
22776 Enable using dd reload support if needed.
c6d5ff83
MM
22777
22778 * config/rs6000/dfp.md (movtd): Delete, combine with 128-bit
22779 binary and decimal floating point moves in rs6000.md.
22780 (movtd_internal): Likewise.
22781
22782 * config/rs6000/rs6000.md (FMOVE128): Combine 128-bit binary and
22783 decimal floating point moves.
22784 (movtf): Likewise.
22785 (movtf_internal): Likewise.
22786 (mov<mode>_internal, TDmode/TFmode): Likewise.
22787 (movtf_softfloat): Likewise.
22788 (mov<mode>_softfloat, TDmode/TFmode): Likewise.
22789
22790 * config/rs6000/rs6000.md (movdi_mfpgpr): Delete, combine with
22791 movdi_internal64, using wg constraint for move direct operations.
22792 (movdi_internal64): Likewise.
22793
22794 * config/rs6000/rs6000.c (rs6000_debug_reg_global): Print
22795 MODES_TIEABLE_P for selected modes. Print the numerical value of
22796 the various virtual registers. Use GPR/FPR first/last values,
22797 instead of hard coding the register numbers. Print which modes
22798 have reload functions registered.
bd059b26
UB
22799 (rs6000_option_override_internal): If -mdebug=reg, trace the options
22800 settings before/after setting cpu, target and subtarget settings.
22801 (rs6000_secondary_reload_trace): Improve the RTL dump for -mdebug=addr
22802 and for secondary reload failures in rs6000_secondary_reload_inner.
c6d5ff83
MM
22803 (rs6000_secondary_reload_fail): Likewise.
22804 (rs6000_secondary_reload_inner): Likewise.
22805
22806 * config/rs6000/rs6000.md (FIRST_GPR_REGNO): Add convenience
22807 macros for first/last GPR and FPR registers.
22808 (LAST_GPR_REGNO): Likewise.
22809 (FIRST_FPR_REGNO): Likewise.
22810 (LAST_FPR_REGNO): Likewise.
22811
22812 * config/rs6000/vector.md (mul<mode>3): Use the combined macro
22813 VECTOR_UNIT_ALTIVEC_OR_VSX_P instead of separate calls to
22814 VECTOR_UNIT_ALTIVEC_P and VECTOR_UNIT_VSX_P.
22815 (vcond<mode><mode>): Likewise.
22816 (vcondu<mode><mode>): Likewise.
22817 (vector_gtu<mode>): Likewise.
22818 (vector_gte<mode>): Likewise.
22819 (xor<mode>3): Don't allow logical operations on TImode in 32-bit
bd059b26 22820 to prevent the compiler from converting DImode operations to TImode.
c6d5ff83
MM
22821 (ior<mode>3): Likewise.
22822 (and<mode>3): Likewise.
22823 (one_cmpl<mode>2): Likewise.
22824 (nor<mode>3): Likewise.
22825 (andc<mode>3): Likewise.
22826
22827 * config/rs6000/constraints.md (wt constraint): New constraint
22828 that returns VSX_REGS if TImode is allowed in VSX registers.
22829
22830 * config/rs6000/predicates.md (easy_fp_constant): 0.0f is an easy
22831 constant under VSX.
22832
22833 * config/rs6000/rs6000-modes.def (PTImode): Define, PTImode is
22834 similar to TImode, but it is restricted to being in the GPRs.
22835
22836 * config/rs6000/rs6000.opt (-mvsx-timode): New switch to allow
22837 TImode to occupy a single VSX register.
22838
22839 * config/rs6000/rs6000-cpus.def (ISA_2_6_MASKS_SERVER): Default to
22840 -mvsx-timode for power7/power8.
22841 (power7 cpu): Likewise.
22842 (power8 cpu): Likewise.
22843
22844 * config/rs6000/rs6000.c (rs6000_hard_regno_nregs_internal): Make
22845 sure that TFmode/TDmode take up two registers if they are ever
22846 allowed in the upper VSX registers.
22847 (rs6000_hard_regno_mode_ok): If -mvsx-timode, allow TImode in VSX
22848 registers.
22849 (rs6000_init_hard_regno_mode_ok): Likewise.
22850 (rs6000_debug_reg_global): Add debugging for PTImode and wt
22851 constraint. Print if LRA is turned on.
22852 (rs6000_option_override_internal): Give an error if -mvsx-timode
22853 and VSX is not enabled.
22854 (invalid_e500_subreg): Handle PTImode, restricting it to GPRs. If
22855 -mvsx-timode, restrict TImode to reg+reg addressing, and PTImode
22856 to reg+offset addressing. Use PTImode when checking offset
22857 addresses for validity.
22858 (reg_offset_addressing_ok_p): Likewise.
22859 (rs6000_legitimate_offset_address_p): Likewise.
22860 (rs6000_legitimize_address): Likewise.
22861 (rs6000_legitimize_reload_address): Likewise.
22862 (rs6000_legitimate_address_p): Likewise.
22863 (rs6000_eliminate_indexed_memrefs): Likewise.
22864 (rs6000_emit_move): Likewise.
22865 (rs6000_secondary_reload): Likewise.
22866 (rs6000_secondary_reload_inner): Handle PTImode. Allow 64-bit
22867 reloads to fpr registers to continue to use reg+offset addressing,
22868 but 64-bit reloads to altivec registers need reg+reg addressing.
22869 Drop test for PRE_MODIFY, since VSX loads/stores no longer support
22870 it. Treat LO_SUM like a PLUS operation.
22871 (rs6000_secondary_reload_class): If type is 64-bit, prefer to use
bd059b26 22872 FLOAT_REGS instead of VSX_RGS to allow use of reg+offset addressing.
c6d5ff83
MM
22873 (rs6000_cannot_change_mode_class): Do not allow TImode in VSX
22874 registers to share a register with a smaller sized type, since VSX
22875 puts scalars in the upper 64-bits.
22876 (print_operand): Add support for PTImode.
22877 (rs6000_register_move_cost): Use VECTOR_MEM_VSX_P instead of
22878 VECTOR_UNIT_VSX_P to catch types that can be loaded in VSX
22879 registers, but don't have arithmetic support.
22880 (rs6000_memory_move_cost): Add test for VSX.
22881 (rs6000_opt_masks): Add -mvsx-timode.
22882
22883 * config/rs6000/vsx.md (VSm): Change to use 64-bit aligned moves
22884 for TImode.
22885 (VSs): Likewise.
22886 (VSr): Use wt constraint for TImode.
22887 (VSv): Drop TImode support.
22888 (vsx_movti): Delete, replace with versions for 32-bit and 64-bit.
22889 (vsx_movti_64bit): Likewise.
22890 (vsx_movti_32bit): Likewise.
22891 (vec_store_<mode>): Use VSX iterator instead of vector iterator.
22892 (vsx_and<mode>3): Delete use of '?' constraint on inputs, just put
22893 one '?' on the appropriate output constraint. Do not allow TImode
22894 logical operations on 32-bit systems.
22895 (vsx_ior<mode>3): Likewise.
22896 (vsx_xor<mode>3): Likewise.
22897 (vsx_one_cmpl<mode>2): Likewise.
22898 (vsx_nor<mode>3): Likewise.
22899 (vsx_andc<mode>3): Likewise.
22900 (vsx_concat_<mode>): Likewise.
22901 (vsx_xxpermdi_<mode>): Fix thinko for non V2DF/V2DI modes.
22902
22903 * config/rs6000/rs6000.h (MASK_VSX_TIMODE): Map from
22904 OPTION_MASK_VSX_TIMODE.
22905 (enum rs6000_reg_class_enum): Add RS6000_CONSTRAINT_wt.
22906 (STACK_SAVEAREA_MODE): Use PTImode instead of TImode.
22907
22908 * config/rs6000/rs6000.md (INT mode attribute): Add PTImode.
22909 (TI2 iterator): New iterator for TImode, PTImode.
22910 (wd mode attribute): Add values for vector types.
bd059b26
UB
22911 (movti_string): Replace TI move operations with operations for TImode
22912 and PTImode. Add support for TImode being allowed in VSX registers.
c6d5ff83
MM
22913 (mov<mode>_string, TImode/PTImode): Likewise.
22914 (movti_ppc64): Likewise.
22915 (mov<mode>_ppc64, TImode/PTImode): Likewise.
22916 (TI mode splitters): Likewise.
22917
22918 * doc/md.texi (PowerPC and IBM RS6000 constraints): Document wt
22919 constraint.
22920
1fc5eced
MG
229212013-03-20 Marc Glisse <marc.glisse@inria.fr>
22922
22923 PR tree-optimization/56355
22924 * fold-const.c (tree_binary_nonnegative_warnv_p) <MULT_EXPR>:
22925 Also handle integers with undefined overflow.
22926
22c4c869
CM
229272013-03-20 Catherine Moore <clm@codesourcery.com>
22928 Maciej W. Rozycki <macro@codesourcery.com>
22929 Tom de Vries <tom@codesourcery.com>
12211b99 22930 Nathan Sidwell <nathan@codesourcery.com>
22c4c869
CM
22931 Iain Sandoe <iain@codesourcery.com>
22932 Nathan Froyd <froydnj@codesourcery.com>
12211b99 22933 Chao-ying Fu <fu@mips.com>
22c4c869
CM
22934
22935 * doc/extend.texi: (micromips, nomicromips, nocompression):
7cf34aae 22936 Document new function attributes.
22c4c869
CM
22937 * doc/invoke.texi (minterlink-compressed, mmicromips,
22938 m14k, m14ke, m14kec): Document new options.
22939 (minterlink-mips16): Update documentation.
22940 * doc/md.texi (ZC, ZD): Document new constraints.
22941 * configure.ac (gcc_cv_as_micromips): Check if linker
22942 supports the .set micromips directive.
22943 * configure: Regenerate.
22944 * config.in: Regenerate.
22945 * config/mips/mips-tables.opt: Regenerate.
22946 * config/mips/micromips.md: New file.
22947 * constraints.md (ZC, ZD): New constraints.
22948 * config/mips/predicates.md (movep_src_register): New predicate.
22949 (movep_src_operand): New predicate.
22950 (non_volatile_mem_operand): New predicate.
22951 * config/mips/mips.md (multimem): New type.
22952 (length): Differentiate between 17-bit and 18-bit branch offsets.
22953 (MOVEP1, MOVEP2): New mode iterator.
33159866 22954 (mov_<load>l): Use ZC constraint.
22c4c869
CM
22955 (mov_<load>r): Likewise.
22956 (mov_<store>l): Likewise.
22957 (mov_<store>r): Likewise.
22958 (*branch_equality<mode>_inverted): Add microMIPS support.
22959 (*branch_equality<mode>): Likewise.
22960 (*jump_absolute): Likewise.
22961 (indirect_jump_<mode>): Likewise.
22962 (tablejump_<mode>): Likewise.
22963 (<optab>_internal): Likewise.
22964 (sibcall_internal): Likewise.
22965 (sibcall_value_internal): Likewise.
22966 (prefetch): Use constraint ZD.
22967 * config/mips/mips.opt (minterlink-compressed): New option.
22968 (minterlink-mips16): Now an alias for minterlink-compressed.
22969 (mmicromips): New option.
22970 * config/mips/sync.md (sync_compare_and_swap<mode>): Use ZR constraint.
22971 (compare_and_swap_12): Likewise.
22972 (sync_add<mode>): Likewise.
22973 (sync_<optab>_12): Likewise.
22974 (sync_old_<optab>_12): Likewise.
22975 (sync_new_<optab>_12): Likewise.
22976 (sync_nand_12): Likewise.
22977 (sync_old_nand_12): Likewise.
22978 (sync_new_nand_12): Likewise.
22979 (sync_sub<mode>): Likewise.
22980 (sync_old_add<mode>): Likewise.
22981 (sync_old_sub<mode>): Likewise.
22982 (sync_new_add<mode>): Likewise.
22983 (sync_new_sub<mode>): Likewise.
22984 (sync_<optab><mode>): Likewise.
22985 (sync_old_<optab><mode>): Likewise.
22986 (sync_new_<optab><mode>): Likewise.
22987 (sync_nand<mode>): Likewise.
22988 (sync_old_nand<mode>): Likewise.
22989 (sync_new_nand<mode>): Likewise.
22990 (sync_lock_test_and_set<mode>): Likewise.
22991 (test_and_set_12): Likewise.
22992 (atomic_compare_and_swap<mode>): Likewise.
22993 (atomic_exchange<mode>_llsc): Likewise.
22994 (atomic_fetch_add<mode>_llsc): Likewise.
22995 * config/mips/mips-cpus.def (m14kc, m14k): New processors.
22996 * config/mips/mips-protos.h (umips_output_save_restore): New prototype.
22997 (umips_save_restore_pattern_p): Likewise.
22998 (umips_load_store_pair_p): Likewise.
22999 (umips_output_load_store_pair): Likewise.
23000 (umips_movep_target_p): Likewise.
23001 (umips_12bit_offset_address_p): Likewise.
23002 * config/mips/mips.c (MIPS_MAX_FIRST_STEP): Update for microMIPS.
23003 (mips_base_mips16): Rename this...
23004 (mips_base_compression_flags): ...to this. Update all uses.
23005 (mips_attribute_table): Add micromips, nomicromips and nocompression.
23006 (mips_mips16_decl_p): Delete.
23007 (mips_nomips16_decl_p): Delete.
33159866 23008 (mips_get_compress_on_flags): New function.
22c4c869
CM
23009 (mips_get_compress_off_flags): New function.
23010 (mips_get_compress_mode): New function.
23011 (mips_get_compress_on_name): New function.
23012 (mips_get_compress_off_name): New function.
23013 (mips_insert_attributes): Support multiple compression types.
23014 (mips_merge_decl_attributes): Likewise.
23015 (umips_12bit_offset_address_p): New function.
23016 (mips_start_function_definition): Emit .set micromips directive.
23017 (mips_call_may_need_jalx_p): New function.
23018 (mips_function_ok_for_sibcall): Add microMIPS support.
23019 (mips_print_operand_punctuation): Support short delay slots and
23020 compact jumps.
23021 (umips_swm_mask, umips_swm_encoding): New.
23022 (umips_build_save_restore): New function.
23023 (mips_for_each_saved_gpr_and_fpr): Add microMIPS support.
23024 (was_mips16_p): Remove.
23025 (old_compression_mode): New.
23026 (mips_set_compression_mode): New function.
23027 (mips_set_current_function): Add microMIPS support.
23028 (mips_option_override): Likewise.
23029 (umips_save_restore_pattern_p): New function.
23030 (umips_output_save_restore): New function.
23031 (umips_load_store_pair_p_1): New function.
23032 (umips_load_store_pair_p): New function.
23033 (umips_output_load_store_pair_1): New function.
23034 (umips_output_load_store_pair): New function.
23035 (umips_movep_target_p) New function.
23036 (mips_prepare_pch_save): Add microMIPS support.
23037 * config/mips/mips.h (TARGET_COMPRESSION): New.
23038 (TARGET_CPU_CPP_BUILTINS): Update macro
23039 to use new compression flags and to support microMIPS.
23040 (MIPS_ISA_LEVEL_SPEC): Add m14k processors.
23041 (MIPS_ARCH_FLOAT_SPEC): Likewise.
23042 (ISA_HAS_LWXS): Include TARGET_MICROMIPS.
23043 (ISA_HAS_LOAD_DELAY): Exclude TARGET_MICROMIPS.
23044 (ASM_SPEC): Support mmicromips and mno-micromips.
23045 (M16STORE_REG_P): New macro.
23046 (MIPS_CALL): Support TARGET_MICROMIPS.
23047 (MICROMIPS_J): New macro.
23048 (mips_base_mips16): Rename this...
23049 (mips_base_compression_flags): ...to this.
23050 (UMIPS_12BIT_OFFSET_P): New macro.
23051 * config/mips/t-sde: (MULTILIB_OPTIONS): Add microMIPS.
23052 (MULTILIB_DIRNAMES): Likewise.
92a8d7a7
RB
230532013-03-20 Richard Biener <rguenther@suse.de>
23054
23055 PR tree-optimization/56661
23056 * tree-ssa-sccvn.c (visit_use): Only value-number calls if
23057 the result does not have to be distinct.
23058
54714c68
RB
230592013-03-20 Richard Biener <rguenther@suse.de>
23060
23061 * tree-inline.c (copy_tree_body_r): Sync MEM_REF code with
23062 remap_gimple_op_r.
23063
cca1130d 230642013-03-20 Bill Schmidt <wschmidt@linux.vnet.ibm.com>
12211b99 23065 Steven Bosscher <steven@gcc.gnu.org>
cca1130d
BS
23066
23067 PR rtl-optimization/56605
23068 * loop-iv.c (implies_p): Handle equal RTXs and subregs.
23069
2a930377
UB
230702013-03-20 Uros Bizjak <ubizjak@gmail.com>
23071
23072 PR bootstrap/56656
23073 * config/i386/i386.md (*movdi_internal): Handle broken assemblers
23074 that require movd instead of movq.
23075
d6d305fe
RB
230762013-03-20 Richard Biener <rguenther@suse.de>
23077
23078 * tree-ssa-structalias.c (struct variable_info): Add pointer
23079 to the first field of an aggregate with sub-vars. Make
23080 this and the pointer to the next subfield its ID.
23081 (vi_next): New function.
23082 (nothing_id, anything_id, readonly_id, escaped_id, nonlocal_id,
23083 storedanything_id, integer_id): Increment by one.
23084 (new_var_info, get_call_vi, lookup_call_clobber_vi,
23085 get_call_clobber_vi): Adjust.
23086 (solution_set_expand): Simplify and speedup.
23087 (solution_set_add): Inline into ...
23088 (set_union_with_increment): ... this. Adjust accordingly.
23089 (do_sd_constraint): Likewise.
23090 (do_ds_constraint): Likewise.
23091 (do_complex_constraint): Simplify.
23092 (build_pred_graph): Adjust.
23093 (solve_graph): Likewise. Simplify and speedup.
23094 (get_constraint_for_ssa_var, get_constraint_for_ptr_offset,
23095 get_constraint_for_component_ref, get_constraint_for_1,
23096 first_vi_for_offset, first_or_preceding_vi_for_offset,
23097 create_function_info_for, create_variable_info_for_1,
23098 create_variable_info_for, intra_create_variable_infos): Adjust.
23099 (init_base_vars): Push NULL for ID zero.
23100 (compute_points_to_sets): Adjust.
23101
a271b42d
RB
231022013-03-20 Richard Biener <rguenther@suse.de>
23103
23104 * cfgloop.c (verify_loop_structure): Streamline and avoid
23105 ICEing on corrupt loop tree.
23106 * graph.c (draw_cfg_nodes_for_loop): Avoid ICEing on corrupt
23107 loop tree.
23108
ebd65954
RB
231092013-03-20 Richard Biener <rguenther@suse.de>
23110
23111 * tree-vect-loop-manip.c (slpeel_can_duplicate_loop_p): Do not
23112 check whether an SSA update is needed.
23113
4547b7ee
RS
231142013-03-20 Richard Sandiford <rdsandiford@googlemail.com>
23115
23116 * config/mips/constraints.md (T): Rename to...
23117 (Yf): ...this.
23118 (U): Rename to...
23119 (Yd): ...this.
23120 * config/mips/mips.md (*movdi_64bit, *movdi_64bit_mips16)
23121 (*mov<mode>_internal, *mov<mode>_mips16): Update accordingly.
23122
0b8cdc58
IB
231232013-03-19 Ian Bolton <ian.bolton@arm.com>
23124
23125 * config/aarch64/aarch64.md (*sub<mode>3_carryin): New pattern.
23126 (*subsi3_carryin_uxtw): Likewise.
23127
5977a10d
IB
231282013-03-19 Ian Bolton <ian.bolton@arm.com>
23129
23130 * config/aarch64/aarch64.md (*ror<mode>3_insn): New pattern.
23131 (*rorsi3_insn_uxtw): Likewise.
23132
bd83ff2c
IB
231332013-03-19 Ian Bolton <ian.bolton@arm.com>
23134
23135 * config/aarch64/aarch64.md (*extr<mode>5_insn): New pattern.
23136 (*extrsi5_insn_uxtw): Likewise.
23137
ba65123c
RB
231382013-03-19 Richard Biener <rguenther@suse.de>
23139
23140 PR tree-optimization/56273
23141 * passes.c (init_optimization_passes): Move second VRP after DOM.
23142
2eac0476
UB
231432013-03-19 Uros Bizjak <ubizjak@gmail.com>
23144
23145 * config/i386/i386.md (*movti_internal): Merge from
23146 *movti_internal_rex64 and *movti_internal_sse. Use x64 isa attribute.
23147 (*movdi_internal): Merge with *movdi_internal_rex64. Use x64 and
23148 nox64 isa attributes.
23149
6e55eda7
RB
231502013-03-18 Richard Biener <rguenther@suse.de>
23151
23152 * tree-ssa-structalias.c (find): Use gcc_checking_assert.
23153 (unite): Likewise.
23154 (merge_node_constraints): Likewise.
23155 (build_succ_graph): Likewise.
23156 (valid_graph_edge): Inline into single caller.
23157 (unify_nodes): Likewise. Use bitmap_set_bit return value
23158 and cache varinfo.
23159 (scc_visit): Fix formatting and variable use.
23160 (do_sd_constraint): Use gcc_checking_assert.
23161 (do_ds_constraint): Likewise.
23162 (do_complex_constraint): Likewise.
23163 (condense_visit): Likewise. Cleanup.
23164 (dump_pred_graph): New function.
23165 (perform_var_substitution): Dump the pred-graph before
23166 variable substitution.
23167 (find_equivalent_node): Use gcc_checking_assert.
23168 (rewrite_constraints): Guard checking loop with ENABLE_CHECKING.
23169
4bdd44c4
RB
231702013-03-18 Richard Biener <rguenther@suse.de>
23171
23172 * tree-vect-loop-manip.c (vect_create_cond_for_alias_checks):
23173 Remove cond_expr_stmt_list argument and do not gimplify the
23174 built expression.
23175 (vect_loop_versioning): Adjust.
23176 * tree-vect-data-refs.c (vect_create_addr_base_for_vector_ref):
23177 Cleanup to use less temporaries.
23178 (vect_create_data_ref_ptr): Cleanup.
23179
38c56a5b
JJ
231802013-03-18 Jakub Jelinek <jakub@redhat.com>
23181
23182 PR tree-optimization/56635
23183 * fold-const.c (operand_equal_p): For MEM_REF and TARGET_MEM_REF,
23184 require types_compatible_p types.
23185
20b2e6a0
NC
231862013-03-18 Nick Clifton <nickc@redhat.com>
23187
a6178a25
NC
23188 * config/stormy16/stormy16.c (xstormy16_expand_prologue): Remove
23189 spurious backslash.
23190
20b2e6a0
NC
23191 * config/mn10300/mn10300.c (mn10300_get_live_callee_saved_regs):
23192 Add missing line to comment describing function.
23193
92608d0e
RB
231942013-03-18 Richard Biener <rguenther@suse.de>
23195
23196 PR tree-optimization/56210
23197 * tree-ssa-structalias.c (find_func_aliases_for_builtin_call):
23198 Handle string / character search functions.
23199 * tree-ssa-alias.c (ref_maybe_used_by_call_p_1): Likewise.
23200
31348d52
RB
232012013-03-18 Richard Biener <rguenther@suse.de>
23202
23203 PR middle-end/56483
23204 * cfgexpand.c (expand_gimple_cond): Inline gimple_cond_single_var_p
23205 and implement properly.
23206 * gimple.h (gimple_cond_single_var_p): Remove.
23207
fcac74a1
RB
232082013-03-18 Richard Biener <rguenther@suse.de>
23209
23210 * tree-data-ref.h (find_data_references_in_loop): Declare.
23211 * tree-data-ref.c (get_references_in_stmt): Use a stack
23212 vector pre-allocated in the callers.
23213 (find_data_references_in_stmt): Adjust.
23214 (graphite_find_data_references_in_stmt): Likewise.
23215 (create_rdg_vertices): Likewise.
23216 (find_data_references_in_loop): Export.
23217 * tree-vect-data-refs.c (vect_analyze_data_ref_dependences):
23218 Compute dependences here...
23219 (vect_analyze_data_refs): ...not here. When we encounter
23220 a non-vectorizable data reference in basic-block vectorization
23221 truncate the data reference vector. Do not bother to
23222 fixup data-dependence information for gather loads.
23223 * tree-vect-slp.c (vect_slp_analyze_bb_1): Check the number
23224 of data references, as reported.
23225
0d5a1b56
RB
232262013-03-18 Richard Biener <rguenther@suse.de>
23227
23228 PR tree-optimization/3713
23229 * tree-ssa-sccvn.c (visit_copy): Simplify. Always propagate
23230 has_constants and expr.
23231 (stmt_has_constants): Properly valueize SSA names when deciding
23232 whether the stmt has constants.
23233
789c34e3
RB
232342013-03-18 Richard Biener <rguenther@suse.de>
23235
23236 * tree-ssa-loop-manip.c (find_uses_to_rename): Do not scan the
23237 whole function when there is nothing to do.
23238 * tree-ssa-loop.c (pass_vectorize): Remove TODO_update_ssa.
23239 * tree-vectorizer.c (vectorize_loops): Update virtual and
23240 loop-closed SSA once.
23241 * tree-vect-loop.c (vect_transform_loop): Do not update SSA here.
23242
076b4605
RB
232432013-03-18 Richard Biener <rguenther@suse.de>
23244
23245 PR middle-end/56113
23246 * domwalk.c (bb_postorder): New global static.
23247 (cmp_bb_postorder): New function.
23248 (walk_dominator_tree): Replace scheme imposing an order for
23249 visiting dominator sons by one sorting them at the time they
23250 are pushed on the stack.
23251
bdb01696
RB
232522013-03-18 Richard Biener <rguenther@suse.de>
23253
23254 PR tree-optimization/39326
23255 * tree-ssa-loop-im.c (refs_independent_p): Exploit symmetry.
23256 (struct mem_ref): Replace mem member with ao_ref typed member.
23257 (MEM_ANALYZABLE): Adjust.
23258 (memref_eq): Likewise.
23259 (mem_ref_alloc): Likewise.
23260 (gather_mem_refs_stmt): Likewise.
23261 (mem_refs_may_alias_p): Use the ao_ref to query the alias oracle.
23262 (execute_sm_if_changed_flag_set): Adjust.
23263 (execute_sm): Likewise.
23264 (ref_always_accessed_p): Likewise.
23265 (refs_independent_p): Likewise.
23266 (can_sm_ref_p): Likewise.
23267
12d2dc5e
JJ
232682013-03-18 Jakub Jelinek <jakub@redhat.com>
23269
23270 PR c/56566
23271 * tree.c (tree_int_cst_min_precision): For integer_zerop (value)
23272 return 1 even for !unsignedp.
23273
286fb677
UB
232742013-03-17 Uros Bizjak <ubizjak@gmail.com>
23275
23276 * config/i386/i386.md (isa): Add x64 and nox64.
23277 (enabled): Define x64 for TARGET_64BIT and nox64 for !TARGET_64BIT.
23278 (*pushtf): Enable *roF alternative for x64 isa only.
23279 (*pushxf): Merge with *pushxf_nointeger. Use Yx*r constraint. Set
23280 mode attribute of integer alternatives to DImode for TARGET_64BIT.
23281 (*pushdf): Merge with *pushdf_rex64. Use x64 and nox64 isa attributes.
23282 (*movtf_internal): Merge from *movtf_internal_rex64 and
23283 *movtf_internal_sse. Use x64 and nox64 isa attributes.
23284 (*movxf_internal): Merge with *movxf_internal_rex64. Use x64 and
23285 nox64 isa attributes.
23286 (*movdf_internal): Merge with *movdf_internal_rex64. Use x64 and
23287 nox64 isa attributes.
23288 * config/i386/constraints.md (Yd): Do not set for TARGET_64BIT.
23289
88b97037
UB
232902013-03-17 Uros Bizjak <ubizjak@gmail.com>
23291
23292 * config/alpha/alpha.c (TARGET_LRA_P): New define.
23293
9f4f1735
JJ
232942013-03-17 Jakub Jelinek <jakub@redhat.com>
23295
23296 PR target/56640
23297 * config/arm/arm.h (REG_CLASS_NAMES): Add "SFP_REG" and "AFP_REG"
23298 class names. Remove trailing comma after "ALL_REGS".
23299
48f4a6fa
JH
233002013-03-16 Jan Hubicka <jh@suse.cz>
23301
23302 * cgraph.h (cgraph_get_create_real_symbol_node): Declare.
23303 * cgraph.c (cgraph_get_create_real_symbol_node): New function.
23304 * cgrpahbuild.c: Use cgraph_get_create_real_symbol_node instead
23305 of cgraph_get_create_node.
23306 * ipa-prop.c (ipa_make_edge_direct_to_target): Likewise.
23307
98e81123
JM
233082013-03-16 Jason Merrill <jason@redhat.com>
23309
23310 PR debug/49090
23311 * dwarf2out.c (gen_generic_params_dies): Indicate default arguments
23312 with DW_AT_default_value.
23313
8a298c04
JJ
233142013-03-16 Jakub Jelinek <jakub@redhat.com>
23315
23316 * BASE-VER: Set to 4.9.0.
23317
4323afa0
AK
233182013-03-14 Andi Kleen <ak@linux.intel.com>
23319
23320 PR target/56619
23321 * doc/extend.texi: Document __ATOMIC_HLE_ACQUIRE,
23322 __ATOMIC_HLE_RELEASE. Document __builtin_ia32 TSX intrincs.
23323 Document _x* TSX intrinsics.
23324
b3c0d469
JJ
233252013-03-14 Edgar E. Iglesias <edgar.iglesias@xilinx.com>
23326 David Holsgrove <david.holsgrove@xilinx.com>
8cc9a5a5
EI
23327
23328 * configure.ac: Add MicroBlaze TLS support detection.
23329 * configure: Regenerate.
88b97037
UB
23330 * config/microblaze/microblaze-protos.h
23331 (microblaze_cannot_force_const_mem, microblaze_tls_referenced_p,
23332 symbol_mentioned_p, label_mentioned_p): Add prototypes.
23333 * config/microblaze/microblaze.c (microblaze_address_type): Add
23334 ADDRESS_TLS and tls_reloc address types.
8cc9a5a5
EI
23335 (microblaze_address_info): Add tls_reloc.
23336 (TARGET_HAVE_TLS): Define.
23337 (get_tls_get_addr, microblaze_tls_symbol_p, microblaze_tls_operand_p_1,
23338 microblaze_tls_referenced_p, microblaze_cannot_force_const_mem,
88b97037
UB
23339 symbol_mentioned_p, label_mentioned_p, tls_mentioned_p,
23340 load_tls_operand, microblaze_call_tls_get_addr,
23341 microblaze_legitimize_tls_address): New functions.
8cc9a5a5
EI
23342 (microblaze_classify_unspec): Handle UNSPEC_TLS.
23343 (get_base_reg): Use microblaze_tls_symbol_p.
23344 (microblaze_classify_address): Handle TLS.
88b97037
UB
23345 (microblaze_legitimate_pic_operand): Use symbol_mentioned_p,
23346 label_mentioned_p and microblaze_tls_referenced_p.
8cc9a5a5
EI
23347 (microblaze_legitimize_address): Handle TLS.
23348 (microblaze_address_insns): Handle ADDRESS_TLS.
23349 (pic_address_needs_scratch): Handle TLS.
23350 (print_operand_address): Handle TLS.
23351 (microblaze_expand_prologue): Check TLS_NEEDS_GOT.
23352 (microblaze_expand_move): Handle TLS.
88b97037
UB
23353 (microblaze_legitimate_constant_p): Check
23354 microblaze_cannot_force_const_mem and microblaze_tls_symbol_p.
8cc9a5a5
EI
23355 (TARGET_CANNOT_FORCE_CONST_MEM): Define.
23356 * config/microblaze/microblaze.h (TLS_NEEDS_GOT): Define
23357 (PIC_OFFSET_TABLE_REGNUM): Set.
23358 * config/microblaze/linux.h (TLS_NEEDS_GOT): Define.
23359 * config/microblaze/microblaze.md (UNSPEC_TLS): Define.
23360 (addsi3, movsi_internal2, movdf_internal): Update constraints
23361 * config/microblaze/predicates.md (arith_plus_operand): Define
88b97037
UB
23362 (move_operand): Redefine as move_src_operand,
23363 check microblaze_tls_referenced_p.
8cc9a5a5 23364
d803a491
IB
233652013-03-14 Ian Bolton <ian.bolton@arm.com>
23366
23367 * config/aarch64/aarch64.md: (*and<mode>3nr_compare0): Use CC_NZ.
23368 (*and_<SHIFT:optab><mode>3nr_compare0): Likewise.
23369
a8504f22
IB
233702013-03-14 Ian Bolton <ian.bolton@arm.com>
23371
23372 * config/aarch64/aarch64.c (aarch64_select_cc_mode): Return correct
23373 CC mode for AND.
23374
df2dfaea
JJ
233752013-03-14 Jakub Jelinek <jakub@redhat.com>
23376
fbd28bc3
JJ
23377 PR tree-optimization/53265
23378 * common.opt (Waggressive-loop-optimizations): New option.
23379 * tree-ssa-loop-niter.c: Include tree-pass.h.
23380 (do_warn_aggressive_loop_optimizations): New function.
23381 (record_estimate): Call it. Don't add !is_exit bounds to loop->bounds
23382 if number_of_latch_executions returned constant.
23383 (estimate_numbers_of_iterations_loop): Call number_of_latch_executions
23384 early. If number_of_latch_executions returned constant, set
23385 nb_iterations_upper_bound back to it.
23386 * cfgloop.h (struct loop): Add warned_aggressive_loop_optimizations
23387 field.
23388 * Makefile.in (tree-ssa-loop-niter.o): Depend on $(TREE_PASS_H).
23389 * doc/invoke.texi (-Wno-aggressive-loop-optimizations): Document.
23390
df2dfaea
JJ
23391 * config/aarch64/t-aarch64-linux (MULTARCH_DIRNAME): Remove.
23392 (MULTILIB_OSDIRNAMES): Set.
23393 * genmultilib: If defaultosdirname doesn't start with :: , set
23394 defaultosdirname2 instead, clear it and emit two . multilib_raw
23395 entries instead of just one.
23396
ee0d2b68
KK
233972013-03-14 Kaz Kojima <kkojima@gcc.gnu.org>
23398
23399 * config/sh/linux.h (TARGET_DEFAULT): Remove MASK_USERMODE.
23400 (SUBTARGET_OVERRIDE_OPTIONS): Set TARGET_USERMODE as default.
23401 * config/sh/netbsd-elf.h (TARGET_DEFAULT): Remove MASK_USERMODE.
23402 (SUBTARGET_OVERRIDE_OPTIONS): New.
23403
decc676e
OE
234042013-03-13 Oleg Endo <olegendo@gcc.gnu.org>
23405
23406 PR target/49880
23407 * config/sh/sh.opt (FPU_SINGLE_ONLY): New mask.
23408 (musermode): Convert to Var(TARGET_USERMODE).
23409 * config/sh/sh.h (SELECT_SH2A_SINGLE_ONLY, SELECT_SH4_SINGLE_ONLY,
23410 MASK_ARCH): Add MASK_FPU_SINGLE_ONLY.
23411 * config/sh/sh.c (sh_option_override): Use
23412 TARGET_FPU_DOUBLE || TARGET_FPU_SINGLE_ONLY for call-fp case.
23413 * config/sh/sh.md (udivsi3_i1, divsi3_i1): Remove ! TARGET_SH4
23414 condition.
23415 (udivsi3_i4, divsi3_i4): Use TARGET_FPU_DOUBLE condition instead of
23416 TARGET_SH4.
23417 (udivsi3_i4_single, divsi3_i4_single): Use
23418 TARGET_FPU_SINGLE_ONLY || TARGET_FPU_DOUBLE instead of TARGET_HARD_SH4.
23419
f4b719c7
DK
234202013-03-13 Dave Korn <dave.korn.cygwin@....>
23421
23422 * config/i386/cygwin.h (SHARED_LIBGCC_SPEC): Make shared libgcc the
23423 default setting.
23424
c40eced0
RB
234252013-03-13 Richard Biener <rguenther@suse.de>
23426
23427 PR tree-optimization/56608
23428 * tree-vect-slp.c (vect_schedule_slp): Do not remove scalar
23429 calls when vectorizing basic-blocks.
23430
1bfa5973
JJ
234312013-03-13 Jakub Jelinek <jakub@redhat.com>
23432
23433 PR plugins/45078
23434 * config.gcc: On arm, mips, sh and sparc add vxworks-dummy.h to
23435 tm_file.
23436
6fcf5434
JJ
234372013-03-12 Jakub Jelinek <jakub@redhat.com>
23438
23439 * doc/invoke.texi (-Waddr-space-convert): Move into the table earlier.
23440
4f38fa8c
JH
234412013-03-11 Jan Hubicka <jh@suse.cz>
23442
23443 PR lto/56557
23444 * lto-streamer-out.c (output_symbol_p): Skip references from
23445 constructors of external variables.
23446
c5c5ba89
JH
234472013-03-11 Jan Hubicka <jh@suse.cz>
23448
23449 PR middle-end/56571
23450 * valtrack.c (cleanup_auto_inc_dec): Unshare clobbers originating
23451 from pseudos.
23452 * emit-rtl.c (verify_rtx_sharing): Likewise.
23453 (copy_insn_1): Likewise.
23454 * rtl.c (copy_rtx): Likewise.
23455
c2a939b1
GJL
234562013-03-11 Georg-Johann Lay <avr@gjlay.de>
23457
23458 PR target/56591
23459 * config/avr/avr.c (avr_print_operand): Add space after '%c' in
23460 output_operand_lossage message.
88b97037 23461
3c3279fb
RE
234622013-03-11 Richard Earnshaw <rearnsha@arm.com>
23463
23464 PR target/56470
23465 * arm.c (shift_op): Validate RTL pattern on the fly.
23466 (arm_print_operand, case 'S'): Don't use shift_operator to validate
23467 the RTL.
23468
aef5ef9d
JDA
234692013-03-10 John David Anglin <dave.anglin@nrc-cnrc.gc.ca>
23470
23471 PR target/56347
23472 * config/pa/pa.md (call_value): Check for calls to powf and direct to
23473 new call patterns that clobber %fr12.
23474 (call_val_powf, call_val_powf_pic, call_val_powf_64bit): New insn,
23475 split and postreload patterns.
23476 * config/pa/pa.c (pa_conditional_register_usage): Revert marking
23477 registers %fr12 and %fr12R as call used.
23478
456610d3
SB
234792013-03-09 Steven Bosscher <steven@gcc.gnu.org>
23480
23481 * dse.c (delete_dead_store_insn): Respect TDF_DETAILS.
23482 (canon_address, record_store, replace_read, check_mem_read_rtx,
23483 scan_insn, dse_step1, dse_step2_init, dse_step2_spill,
23484 dse_step4, dse_step5_nospill, dse_step5_spill, dse_step6,
23485 rest_of_handle_dse): Likewise.
23486
4b1baac8
RS
234872013-03-09 Richard Sandiford <rdsandiford@googlemail.com>
23488
23489 PR middle-end/56524
23490 * tree.h (tree_optimization_option): Rename target_optabs to optabs.
23491 Add base_optabs.
23492 (TREE_OPTIMIZATION_OPTABS): Update after previous field change.
23493 (TREE_OPTIMIZATION_BASE_OPTABS): New macro.
23494 (save_optabs_if_changed): Replace with...
23495 (init_tree_optimization_optabs): ...this.
23496 * optabs.c (save_optabs_if_changed): Rename to...
23497 (init_tree_optimization_optabs): ...this. Take the optimization node
23498 as argument. Do nothing if the base optabs are already correct.
23499 Reuse the existing TREE_OPTIMIZATION_OPTABS memory if we need
23500 to recompute optabs.
23501 * function.h (function): Remove optabs field.
23502 * function.c (invoke_set_current_function_hook): Call
23503 init_tree_optimization_optabs. Use the result to initialize
23504 this_fn_optabs.
23505
b7a78683
AH
235062013-02-27 Aldy Hernandez <aldyh@redhat.com>
23507
23508 * trans-mem.c (expand_transaction): Do not set PR_INSTRUMENTEDCODE
23509 if GTMA_HAS_NO_INSTRUMENTATION.
23510 (generate_tm_state): Keep GTMA_HAS_NO_INSTRUMENTATION bit.
23511 (ipa_tm_transform_transaction): Set GTMA_HAS_NO_INSTRUMENTATION.
23512 * gimple.h (GTMA_HAS_NO_INSTRUMENTATION): Define.
23513 * gimple-pretty-print.c (dump_gimple_transaction): Handle
23514 GTMA_HAS_NO_INSTRUMENTATION.
23515
6384c29b
JJ
235162013-03-08 Jakub Jelinek <jakub@redhat.com>
23517
23518 * config/gnu-user.h (LIBTSAN_EARLY_SPEC): Don't link against
23519 libasan_preinit.o.
23520
ecd4f20a
MP
235212013-03-08 Marek Polacek <polacek@redhat.com>
23522 Jakub Jelinek <jakub@redhat.com>
23523
23524 PR tree-optimization/56478
23525 * predict.c (is_comparison_with_loop_invariant_p): Change the
23526 type of loop_step to tree.
23527 (predict_loops): Adjust.
23528 (predict_iv_comparison): Perform the computations on double_ints.
23529
64366d35
RB
235302013-03-08 Richard Biener <rguenther@suse.de>
23531
23532 PR tree-optimization/56570
23533 * tree-cfg.c (verify_expr_location_1): Verify locations for
23534 DECL_DEBUG_EXPR.
23535 * tree-sra.c (create_access_replacement): Strip locations
23536 from DECL_DEBUG_EXPRs.
23537
a9d5a059
RB
235382013-03-08 Richard Biener <rguenther@suse.de>
23539
23540 * tree-inline.c (expand_call_inline): Do not associate
23541 a BLOCK with the location in BLOCK_SOURCE_LOCATION.
23542 * tree-cfg.c (verify_location): Verify BLOCK_SOURCE_LOCATION.
23543
b7aa4e9a
RB
235442013-03-08 Richard Biener <rguenther@suse.de>
23545
23546 * tree-ssa-ter.c (is_replaceable_p): Do not TER across location
23547 or block changes with -Og. Fix for location / block encoding
23548 changes and PHI arguments with locations.
23549
c4c2f9fa
SB
235502013-03-07 Steven Bosscher <steven@gcc.gnu.org>
23551
23552 * bitmap.c (struct bitmap_descriptor_d): Use unsigned HOST_WIDEST_INT
23553 for all counters.
23554 (struct output_info): Likewise.
23555 (register_overhead): Remove bad gcc_assert.
23556 (bitmap_find_bit): If there is only a single bitmap element, do not
23557 count a miss as a search.
23558 (print_statistics): Update for counter type changes.
23559 (dump_bitmap_statistics): Likewise. Print headers such that they
23560 are properly lined up with the printed counters.
23561
5bf6606a
JJ
235622013-03-07 Jakub Jelinek <jakub@redhat.com>
23563
23564 PR tree-optimization/56559
23565 * tree-ssa-reassoc.c (zero_one_operation): When looking at rhs2,
23566 check that it has only a single use.
23567
2c653d46
RB
235682013-03-07 Richard Biener <rguenther@suse.de>
23569
23570 * doc/invoke.texi (fwhole-program): Discourage use in combination
23571 with -flto.
23572
a72d8780
JJ
235732013-03-06 Jakub Jelinek <jakub@redhat.com>
23574
01a454df
JJ
23575 * config/arm/t-arm (TM_H, OPTIONS_H_EXTRA): Add arm-cores.def.
23576
c1781047
JJ
23577 PR tree-optimization/56539
23578 * tree-tailcall.c (adjust_return_value_with_ops): Use GSI_SAME_STMT
23579 instead of GSI_CONTINUE_LINKING as last argument to
23580 force_gimple_operand_gsi. Adjust function comment.
23581
9772c47a
JJ
23582 * config/aarch64/t-aarch64 (TM_H, OPTIONS_H_EXTRA): Add
23583 aarch64-cores.def.
23584
a72d8780
JJ
23585 PR middle-end/56548
23586 * expr.c (expand_cond_expr_using_cmove): When expanding cmove in
23587 promoted mode, convert the result back to the original mode.
23588
fa5556de
RB
235892013-03-06 Richard Biener <rguenther@suse.de>
23590
23591 PR middle-end/56294
23592 * tree-into-ssa.c (insert_phi_nodes_for): Add dumping.
23593 (insert_updated_phi_nodes_compare_uids): New function.
23594 (update_ssa): Sort symbols_to_rename after UID before
23595 traversing it to insert PHI nodes.
23596
010403d1
RB
235972013-03-06 Richard Biener <rguenther@suse.de>
23598
23599 PR middle-end/50494
23600 * tree-vect-data-refs.c (vect_can_force_dr_alignment_p):
23601 Do not adjust alignment of DECL_IN_CONSTANT_POOL decls.
23602
23603 Revert
23604 2013-02-13 Richard Biener <rguenther@suse.de>
23605
23606 PR lto/50494
23607 * varasm.c (output_constant_def_1): Get the decl representing
23608 the constant as argument.
23609 (output_constant_def): Wrap output_constant_def_1.
23610 (make_decl_rtl): Use output_constant_def_1 with the decl
23611 representing the constant.
23612 (build_constant_desc): Optionally re-use a decl already
23613 representing the constant.
23614 (tree_output_constant_def): Adjust.
23615
3c27ce4c
JY
236162013-03-06 Joey Ye <joey.ye@arm.com>
23617
23618 PR lto/50293
23619 * gcc.c (convert_white_space): New function.
23620 (main): Handles white space in function name.
23621
8f6d1c86
OE
236222013-03-06 Oleg Endo <olegendo@gcc.gnu.org>
23623
23624 PR target/56529
23625 * config/sh/sh.c (sh_option_override): Check for TARGET_DYNSHIFT
23626 instead of TARGET_SH2 for call-table case. Do not set sh_div_strategy
23627 to SH_DIV_CALL_TABLE for TARGET_SH2.
23628 * config.gcc (sh_multilibs): Add m2 and m2a to sh*-*-linux* multilib
23629 list.
23630 * doc/invoke.texi (SH options): Document mdiv= call-div1, call-fp,
23631 call-table options.
23632
798a209f
SA
236332013-03-05 Sterling Augustine <saugustine@google.com>
23634 Cary Coutant <ccoutant@google.com>
23635
23636 PR debug/55364
23637 * dwarf2out.c (resolve_addr): Don't call
23638 remove_loc_list_addr_table_entries a second time for the same
23639 expression.
23640
6cfa417f 236412013-03-05 Jakub Jelinek <jakub@redhat.com>
ee61ea38 23642
6cfa417f
JJ
23643 PR debug/56510
23644 * cfgexpand.c (expand_debug_parm_decl): Call copy_rtx on incoming.
23645 (avoid_complex_debug_insns): New function.
23646 (expand_debug_locations): Call it.
23647
23648 PR rtl-optimization/56484
23649 * ifcvt.c (noce_process_if_block): If else_bb is NULL, avoid extending
23650 lifetimes of hard registers on small register class machines.
23651
236522013-03-05 David Holsgrove <david.holsgrove@xilinx.com>
23653
23654 * config/microblaze/microblaze-protos.h: Rename
ee61ea38 23655 microblaze_is_interrupt_handler to microblaze_is_interrupt_variant.
6cfa417f 23656 * config/microblaze/microblaze.c (microblaze_attribute_table): Add
ee61ea38
DH
23657 fast_interrupt.
23658 (microblaze_fast_interrupt_function_p): New function.
23659 (microblaze_is_interrupt_handler): Rename to
23660 microblaze_is_interrupt_variant and add fast_interrupt check.
23661 (microblaze_must_save_register): Use microblaze_is_interrupt_variant.
23662 (save_restore_insns): Likewise.
23663 (compute_frame_size): Likewise.
23664 (microblaze_function_prologue): Add FAST_INTERRUPT_NAME.
23665 (microblaze_globalize_label): Likewise.
6cfa417f
JJ
23666 * config/microblaze/microblaze.h: Define FAST_INTERRUPT_NAME.
23667 * config/microblaze/microblaze.md: Use wrapper
ee61ea38
DH
23668 microblaze_is_interrupt_variant.
23669
731aefac
KT
236702013-03-05 Kai Tietz <ktietz@redhat.com>
23671
23672 * sdbout.c (sdbout_one_type): Switch to current function's section
23673 supporting cold/hot.
23674
a72d8780 236752013-03-05 David Holsgrove <david.holsgrove@xilinx.com>
e34fd5a7
DH
23676
23677 * doc/invoke.texi (MicroBlaze): Add -mbig-endian, -mlittle-endian,
23678 -mxl-reorder.
23679
334e71e8
JJ
236802013-03-05 Jakub Jelinek <jakub@redhat.com>
23681
0b50e654
JJ
23682 PR middle-end/56461
23683 * ggc-common.c (gt_pch_save): For ENABLE_VALGRIND_CHECKING,
23684 if VALGRIND_GET_VBITS is defined, temporarily make object
23685 memory all defined, and restore previous valgrind addressability
23686 and definability afterwards. Free this_object at the end.
23687
4ccf8f43
JJ
23688 PR middle-end/56461
23689 * lra.c (lra): Call lra_clear_live_ranges if live_p,
23690 right before calling lra_create_live_ranges, also call it
23691 when clearing live_p. Only call lra_clear_live_ranges
23692 at the end if live_p.
23693
334e71e8
JJ
23694 PR middle-end/56461
23695 * sched-deps.c (delete_dep_node): Free DEP_REPLACE.
23696
9ca966ca
RB
236972013-03-05 Richard Biener <rguenther@suse.de>
23698
23699 PR tree-optimization/56521
23700 * tree-ssa-sccvn.c (set_value_id_for_result): Always initialize
23701 value-id.
23702
d273b176
SB
237032013-03-05 Steven Bosscher <steven@gcc.gnu.org>
23704
23705 PR c++/55135
23706 * except.h (remove_unreachable_eh_regions): New prototype.
23707 * except.c (remove_eh_handler_splicer): New function, split out
23708 of remove_eh_handler.
23709 (remove_eh_handler): Use remove_eh_handler_splicer. Add comment
23710 warning about running it on many EH regions one at a time.
23711 (remove_unreachable_eh_regions_worker): New function, walk the
23712 EH tree in depth-first order and remove non-marked regions.
23713 (remove_unreachable_eh_regions): New function.
23714 * tree-eh.c (mark_reachable_handlers): New function, split out
23715 from remove_unreachable_handlers.
23716 (remove_unreachable_handlers): Use mark_reachable_handlers and
23717 remove_unreachable_eh_regions.
23718 (remove_unreachable_handlers_no_lp): Use mark_reachable_handlers
23719 and remove_unreachable_eh_regions.
23720
5e7f1aef
RB
237212013-03-05 Richard Biener <rguenther@suse.de>
23722
23723 PR middle-end/56525
23724 * loop-init.c (fix_loop_structure): Remove loops in two stages,
23725 not freeing them until the end.
23726
f276b762
AK
237272013-03-05 Andreas Krebbel <Andreas.Krebbel@de.ibm.com>
23728
23729 * config/s390/s390.h: Define DWARF2_ASM_LINE_DEBUG_INFO.
23730
12a43ab8
RB
237312013-03-05 Richard Biener <rguenther@suse.de>
23732
23733 PR tree-optimization/56270
23734 * tree-vect-slp.c (vect_schedule_slp): Clear vectorized stmts
23735 of loads after scheduling an SLP instance.
23736
93675444
JJ
237372013-03-05 Jakub Jelinek <jakub@redhat.com>
23738
db4138e3
JJ
23739 * Makefile.in (dg_target_exps): Add aarch64.exp, epiphany.exp and
23740 tic6x.exp.
23741 (check_gcc_parallelize): Run guality.exp as a separate job from
23742 vect.exp with unsorted.exp and $(dg_target_exps) separately from
23743 struct-layout-1.exp with stackalign.exp.
23744
dd3d1ec0
JJ
23745 * alias.c (init_alias_analysis): Clear reg_known_equiv_p bitmap.
23746
b8d381a3
JJ
23747 PR middle-end/56461
23748 * tree-vect-slp.c (vect_supported_load_permutation_p): Free
23749 load_index sbitmap even if some bit in it isn't set.
23750
b4f9786b
JJ
23751 PR middle-end/56461
23752 * tree-ssa-loop-niter.c (bb_queue): Remove typedef.
23753 (discover_iteration_bound_by_body_walk): Change queues to
23754 vec<vec<basic_block> > and queue to vec<basic_block>. Fix up
23755 spelling in comment. Call safe_push on queues[bound_index] directly.
23756 Release queues[queue_index] in every iteration unconditionally.
23757 Release bounds vector.
23758
93675444
JJ
23759 PR middle-end/56461
23760 * tree-vect-stmts.c (free_stmt_vec_info_vec): Call
23761 free_stmt_vec_info on any left-over stmt_vec_info in the vector.
23762 * tree-vect-loop.c (vect_create_epilog_for_reduction): Release
23763 inner_phis vector.
23764
3e492e9c
RB
237652013-03-05 Richard Biener <rguenther@suse.de>
23766
23767 PR lto/56515
23768 * tree-inline.c (remap_blocks_to_null): New function.
23769 (expand_call_inline): When expanding a call stmt without
23770 an associated block inline remap all callee blocks to NULL.
23771
a5d1569a
JJ
237722013-03-05 Jakub Jelinek <jakub@redhat.com>
23773
86efb5cd
JJ
23774 PR rtl-optimization/56494
23775 * simplify-rtx.c (simplify_truncation): If C is narrower than A,
23776 optimize (truncate:A (subreg:B (truncate:C X) 0)) into
23777 (subreg:A (truncate:C X) 0) instead of (truncate:A X).
23778
85f5dbea
JJ
23779 PR middle-end/56461
23780 * sel-sched-ir.c (free_sched_pools): Release
23781 succs_info_pool.stack[succs_info_pool.max_top] vectors too
23782 if succs_info_pool.max_top isn't -1.
23783
a5d1569a
JJ
23784 PR bootstrap/56509
23785 * opts.c (opts_obstack, opts_concat): Moved to...
23786 * opts-common.c (opts_obstack, opts_concat): ... here.
23787
4432aa6c
JJ
237882013-03-04 Jakub Jelinek <jakub@redhat.com>
23789
23790 PR middle-end/56461
23791 * diagnostic.c (diagnostic_append_note): Save and restore old prefix.
23792
f3a1fb91
MJ
237932013-03-04 Martin Jambor <mjambor@suse.cz>
23794
23795 * tree-dfa.c (get_or_create_ssa_default_def): Use parameter fn in
23796 all appropriate places.
23797
22110e6c
EB
237982013-01-04 Eric Botcazou <ebotcazou@adacore.com>
23799
23800 PR tree-optimization/56424
23801 * ipa-split.c (split_function): Do not set the RSO flag if result is
23802 not by reference and its type is a register type.
23803
a72d8780 238042013-03-04 David Holsgrove <david.holsgrove@xilinx.com>
91c4e421 23805
88b97037 23806 * config/microblaze/microblaze.c (microblaze_valid_pic_const): New
91c4e421 23807 (microblaze_legitimate_pic_operand): Likewise
88b97037 23808 * config/microblaze/microblaze.h (LEGITIMATE_PIC_OPERAND_P): calls
91c4e421
DH
23809 new function microblaze_legitimate_pic_operand
23810 * config/microblaze/microblaze-protos.h
23811 (microblaze_legitimate_pic_operand): Declare.
23812
a72d8780 238132013-03-04 Edgar E. Iglesias <edgar.iglesias@gmail.com>
82f8f53a 23814
a72d8780 23815 * config/microblaze/predicates.md (call_insn_simple_operand):
82f8f53a 23816 New predicate for supported rtx code types.
a72d8780 23817 * config/microblaze/microblaze.md (call_internal1): Use
82f8f53a
EI
23818 call_insn_simple_operand predicate.
23819
541d9ac8
JJ
238202013-03-04 Jakub Jelinek <jakub@redhat.com>
23821
5eb010bc
JJ
23822 PR middle-end/56461
23823 * tree-loop-distribution.c (ldist_gen): Call partition_free after each
23824 partitions.ordered_remove.
23825
30862efc
JJ
23826 PR middle-end/56461
23827 * tree-vect-stmts.c (vectorizable_conversion): Don't call
23828 vec_oprnds0.create (1) for modifier == NONE.
23829
8930f723
JJ
23830 PR middle-end/56461
23831 * tree-vect-stmts.c (vectorizable_shift): Don't call create methods
23832 on vec_oprnds0 or vec_oprnds1 before loop, only call it on
23833 vec_oprnds1 right before pushing anything to it for
23834 scalar_shift_arg.
23835
541d9ac8
JJ
23836 PR middle-end/56461
23837 * tree-vect-loop.c (destroy_loop_vec_info): For !clean_stmts, just
23838 set nbbs to 0 instead of having separate code path.
23839 (vect_analyze_loop_form): Call destroy_loop_vec_info with true
23840 instead of false as last argument if returning NULL.
23841
7aa7f2e3
SL
238422013-03-03 Sandra Loosemore <sandra@codesourcery.com>
23843
23844 * target.def (TARGET_OPTION_VALID_ATTRIBUTE_P): Update comments;
23845 the attribute is now called "target" instead of "option".
23846 (TARGET_OPTION_PRAGMA_PARSE): Likewise, for the pragma.
23847 * doc/tm.texi.in (Target Attributes): Likewise document the correct
23848 attribute/pragma name for TARGET_OPTION_VALID_P and
23849 TARGET_OPTION_PRAGMA_PARSE. Also copy-edit and correct markup.
23850 * doc/tm.texi: Regenerated.
23851
8930f723 238522013-03-02 David Holsgrove <david.holsgrove@xilinx.com>
cb8a1637 23853
8930f723 23854 * config/microblaze/microblaze.c:
cb8a1637 23855 Check mcpu, pcmp requirement and set TARGET_REORDER to 0 if not met.
8930f723
JJ
23856 * config/microblaze/microblaze.h: Add -mxl-reorder to
23857 DRIVER_SELF_SPECS.
23858 * config/microblaze/microblaze.md: New bswapsi2 and bswaphi2.
23859 instructions emitted if TARGET_REORDER.
88b97037
UB
23860 * config/microblaze/microblaze.opt: New option -mxl-reorder set to 1
23861 or 0 for -m/-mno case, but initialises as 2 to detect default use case
8930f723 23862 separately.
cb8a1637 23863
4dc1d68c
XDL
238642013-03-01 Xinliang David Li <davidxl@google.com>
23865
23866 * tree-ssa-uninit.c (compute_control_dep_chain): Limit post-dom
23867 walk length.
23868
689eaba3
JJ
238692013-03-01 Jakub Jelinek <jakub@redhat.com>
23870
9d676bf6
JJ
23871 PR middle-end/56461
23872 * tree-ssa-loop-ivcanon.c (tree_estimate_loop_size): Release path
23873 vector even when returning true. Fix up function comment formatting.
23874
caff7edf
JJ
23875 PR middle-end/56461
23876 * ira-build.c (ira_loop_nodes_count): New variable.
23877 (create_loop_tree_nodes): Initialize it.
23878 (finish_loop_tree_nodes): Use it instead of looking at current_loops.
23879
b6b9227d
JJ
23880 PR middle-end/56461
23881 * tree-vect-data-refs.c (vect_permute_store_chain): Avoid using copy
23882 method on dr_chain and result_chain.
23883 * tree-vect-stmts.c (vectorizable_store): Only call
23884 result_chain.create if j == 0.
23885
689eaba3
JJ
23886 PR middle-end/56461
23887 * tree-vect-stmts.c (vect_create_vectorized_promotion_stmts): Call
23888 vec_oprnds0->release (); rather than vec_oprnds0->truncate (0)
23889 before overwriting it.
23890
40bf31ed
TB
238912013-03-01 Tobias Burnus <burnus@net-b.de>
23892
23893 * doc/extended.texi (C Extensions): Change order in @menu
23894 to match @node.
23895 (Other MIPS Built-in Functions): Move last MIPS entry before
23896 "picoChip Built-in Functions".
23897 (SH Built-in Functions): Move after RX Built-in Functions.
23898 * doc/gcc.texi (Introduction): Change order in @menu
23899 to match @node.
23900 * doc/md.texi (Constraints): Ditto.
23901 * gty.texi (Type Information): Ditto.
23902 (User-provided marking routines for template types): Make
23903 subsection.
23904 * doc/invoke.texi (AArch64 Options): Move before
23905 "Adapteva Epiphany Options".
23906
e664c61c
KS
239072013-02-28 Konstantin Serebryany <konstantin.s.serebryany@gmail.com>
23908 Jakub Jelinek <jakub@redhat.com>
23909
23910 PR sanitizer/56454
23911 * asan.c (gate_asan): Lookup no_sanitize_address instead of
23912 no_address_safety_analysis attribute.
23913 * doc/extend.texi (no_address_safety_attribute): Rename to
23914 no_sanitize_address attribute, mention no_address_safety_analysis
23915 attribute as deprecated alias.
23916
37b5ec8f
JJ
239172013-02-28 Jakub Jelinek <jakub@redhat.com>
23918
23919 PR middle-end/56461
23920 * tree-vectorizer.h (vect_get_slp_defs): Change 3rd argument
23921 type to vec<vec<tree> > *.
23922 * tree-vect-slp.c (vect_get_slp_defs): Likewise. Change vec_defs
23923 to be vec<tree> instead of vec<tree> *, set vec_defs
23924 to vNULL and call vec_defs.create (number_of_vects), adjust other
23925 uses of vec_defs.
23926 * tree-vect-stmts.c (vect_get_vec_defs, vectorizable_call,
23927 vectorizable_condition): Adjust vect_get_slp_defs callers.
23928
ba96cdfb
JG
239292013-02-28 James Greenhalgh <james.greenhalgh@arm.com>
23930
23931 * config/aarch64/aarch64.c
23932 (aarch64_float_const_representable): Remove unused variable.
23933
6f549691
JG
239342013-02-28 James Greenhalgh <james.greenhalgh@arm.com>
23935
23936 * config/aarch64/aarch64.c (aarch64_mangle_type): Make static.
23937
af55e82d
JG
239382013-02-28 James Greenhalgh <james.greenhalgh@arm.com>
23939
23940 * config/aarch64/aarch64-builtins.c
23941 (aarch64_init_simd_builtins): Make static.
23942
1df3f464
JG
239432013-02-28 James Greenhalgh <james.greenhalgh@arm.com>
23944
23945 * config/aarch64/aarch64.c
23946 (aarch64_simd_make_constant): Make static.
23947
f8f42513
MJ
239482013-02-28 Martin Jambor <mjambor@suse.cz>
23949
23950 * tree-sra.c (load_assign_lhs_subreplacements): Do not put replacements
23951 with no initialization to the RHS of debug statements.
23952
b48b3fc4
MJ
239532013-02-28 Martin Jambor <mjambor@suse.cz>
23954
23955 PR tree-optimization/56294
23956 * tree-sra.c (analyze_access_subtree): Create replacement declarations.
23957 Adjust dumping.
23958 (get_access_replacement): Do not call create_access_replacement.
23959 Assert a replacement exists.
23960 (get_repl_default_def_ssa_name): Create the replacement declaration
23961 itself.
23962
c3ae224c
RR
239632013-02-28 Ramana Radhakrishnan <ramana.radhakrishnan@arm.com>
23964
23965 * config/arm/arm.c (arm_output_mi_thunk): Call final_start_function and
23966 final_end_function.
23967
45fa0eef
MP
239682013-02-28 Marek Polacek <polacek@redhat.com>
23969
23970 PR rtl-optimization/56466
23971 * loop-unroll.c (unroll_and_peel_loops): Call fix_loop_structure
23972 if we're changing a loop.
23973 (peel_loops_completely): Likewise.
23974
502c067d
PC
239752013-02-28 Paolo Carlini <paolo.carlini@oracle.com>
23976
23977 PR c++/55813
23978 * doc/invoke.texi ([-Wctor-dtor-privacy]): Complete.
23979
ff24afc8
GJL
239802013-02-28 Georg-Johann Lay <avr@gjlay.de>
23981
23982 PR target/56445
23983 * config/avr/avr.c (avr_init_builtins): Use 'n' instead of empty
23984 macro parameters with: FX_FTYPE_FX, FX_FTYPE_FX_INT, INT_FTYPE_FX,
23985 INTX_FTYPE_FX, FX_FTYPE_INTX.
23986 * config/avr/builtins.def: Adjust respective DEF_BUILTIN.
23987
c7b1fc1b
GJL
239882013-02-28 Georg-Johann Lay <avr@gjlay.de>
23989
23990 * avr/avr-mcus.def (ata5272, ata5505, attiny1634, ata6285)
23991 (ata6286, atmega8a, atmega48pa, ata5790, ata5790n, ata5795)
23992 (atmega164pa, atmega165pa, atmega168pa, atmega16hva, atmega16hvb)
23993 (atmega16hvbrevb, atmega16m1, atmega16u4, atmega26hvg, atmega32a)
23994 (atmega32a, atmega3250pa, atmega3290pa, atmega32c1, atmega32m1)
23995 (atmega32u4, atmega32u6, atmega64a, atmega6490a, atmega6490p)
23996 (atmega64c1, atmega64m1, atmega64rfa2, atmega64rfr2, atmega32hvb)
23997 (atmega32hvbrevb, atmega16hva2, atmega48hvf, at90pwm161)
23998 (atmega128a, atmega1284, atmxt112sl, atmxt224, atmxt224e)
23999 (atmxt336s, atxmega16a4u, atxmega16c4, atxmega32a4u, atxmega32c4)
24000 (atxmega32e5, atxmega64a3u, atxmega64a4u, atxmega64b1, atxmega64b3)
24001 (atxmega64c3, atxmega64d4, atxmega128a3u, atxmega128b1)
24002 (atxmega128b3, atxmega128c3, atxmega128d4, atmxt540s, atmxt540sreva)
24003 (atxmega192a3u, atxmega192c3, atxmega256a3u, atxmega256c3)
24004 (atxmega384c3, atxmega384d3, atxmega128a4u): New AVR_MCU.
24005 (avrxmega6): Increase max flash segments from 5 to 6.
24006 * config/avr/t-multilib: Regenerate.
24007 * config/avr/avr-tables.opt: Regenerate.
24008 * doc/avr-mmcu.texi: Regenerate.
24009
4a0e3cfe
GJL
240102013-02-28 Georg-Johann Lay <avr@gjlay.de>
24011
24012 * config/avr/avr.h (device_to_arch): Rename to device_to_ld.
24013 (avr_device_to_arch): Rename to avr_device_to_ld.
24014 (avr_device_to_as): New prototype.
24015 (EXTRA_SPEC_FUNCTIONS): Add device_to_as.
24016 (ASM_SPEC): Use device_to_as to get -mmcu= and -mno-skip-bug=.
24017 * config/avr/driver-avr.c (avr_device_to_as): New.
24018 (avr_device_to_arch): Rename to avr_device_to_ld.
24019
97785e52
JJ
240202013-02-27 Jakub Jelinek <jakub@redhat.com>
24021
3f292312
JJ
24022 PR middle-end/56461
24023 * tree-vect-data-refs.c (vect_permute_load_chain): Avoid using copy
24024 method on dr_chain and result_chain.
24025
a344216b
JJ
24026 PR middle-end/56461
24027 * tree-ssa-loop-niter.c (maybe_lower_iteration_bound): Call
24028 pointer_set_destroy on not_executed_last_iteration.
24029
f121ad02 24030 PR middle-end/56461
88b97037 24031 * tree-vect-loop.c (vectorizable_reduction): Release vect_defs vector.
f121ad02 24032
307f83a3
JJ
24033 PR middle-end/56461
24034 * ipa-pure-const.c (propagate): Use FOR_EACH_FUNCTION instead of
24035 FOR_EACH_DEFINED_FUNCTION when freeing state.
24036
e19624ee
JJ
24037 PR middle-end/56461
24038 * df-scan.c (df_insn_delete): Use df_scan_free_mws_vec before
24039 pool_free.
24040 (df_insn_rescan_debug_internal): Use df_scan_free_mws_vec before
24041 overwriting it.
24042
90e709fd
JJ
24043 PR middle-end/56461
24044 * ipa-cp.c (decide_whether_version_node): Call vec_free on
24045 known_aggs[i].items and release known_aggs vector.
24046
97785e52
JJ
24047 PR middle-end/56461
24048 * ipa-reference.c (propagate): Free node_info even for alias nodes.
24049
240502013-02-27 Edgar E. Iglesias <edgar.iglesias@gmail.com>
bbb3a9e2 24051
97785e52
JJ
24052 * config/microblaze/microblaze.c (microblaze_emit_compare):
24053 Use xor for EQ/NE comparisions.
24054 * config/microblaze/microblaze.md (cstoresf4): Add constraints
24055 (cbranchsf4): Adjust operator to comparison_operator.
bbb3a9e2 24056
6fa5e0ed
JJ
240572013-02-27 Jakub Jelinek <jakub@redhat.com>
24058
24059 PR middle-end/56461
24060 * tree-flow.h (edge_var_map_vector): Change into va_heap, vl_embed
24061 vector.
24062 * tree-ssa.c (redirect_edge_var_map_add): Use vec_safe_reserve and
24063 vec_safe_push, always update *slot.
24064 (redirect_edge_var_map_clear): Use vec_free.
24065 (redirect_edge_var_map_dup): Use vec_safe_copy and vec_safe_reserve.
24066 (free_var_map_entry): Use vec_free.
24067 * tree-cfgcleanup.c (remove_forwarder_block_with_phi): Use
24068 FOR_EACH_VEC_SAFE_ELT instead of FOR_EACH_VEC_ELT.
24069
436a956a
AB
240702013-02-27 Andrey Belevantsev <abel@ispras.ru>
24071
24072 PR middle-end/45472
436a956a
AB
24073 * sel-sched-ir.c (merge_expr): Also change vinsn of merged expr
24074 when the may_trap_p bit of the exprs being merged differs.
436a956a
AB
24075 Reorder tests for speculativeness in the logical and operator.
24076
f5c2caca
JJ
240772013-02-27 Jakub Jelinek <jakub@redhat.com>
24078
0fcb564b
JJ
24079 * incpath.c (add_standard_paths): Use reconcat instead of concat
24080 where appropriate and avoid leaking memory.
24081
dc357798
JJ
24082 * opts.h: Include obstack.h.
24083 (opts_concat): New prototype.
24084 (opts_obstack): New declaration.
24085 * opts.c (opts_concat): New function.
24086 (opts_obstack): New variable.
24087 (init_options_struct): Call gcc_init_obstack on opts_obstack.
24088 (finish_options): Use opts_concat instead of concat
24089 and XOBNEWVEC instead of XNEWVEC.
24090 * opts-common.c (generate_canonical_option, decode_cmdline_option,
24091 generate_option): Likewise.
24092 * Makefile.in (OPTS_H): Depend on $(OBSTACK_H).
24093 * lto-wrapper.c (main): Call gcc_init_obstack on opts_obstack.
24094
f5c2caca
JJ
24095 PR target/56455
24096 * stmt.c (expand_switch_as_decision_tree_p): If flag_pic
24097 and ASM_OUTPUT_ADDR_DIFF_ELT isn't defined, return true.
24098
d0163673
JJ
240992013-02-26 Jakub Jelinek <jakub@redhat.com>
24100
24101 PR middle-end/56461
24102 * lra-spills.c (lra_spill): Free spill_hard_reg at the end.
24103
b5ad2b8e
JR
241042013-02-26 Joern Rennecke <joern.rennecke@embecosm.com>
24105
24106 * config/arm/arm.c (const_ok_for_dimode_op): Back out last change.
24107 (arm_block_move_unaligned_straight): Likewise.
24108 (arm_adjust_block_mem): Likewise.
24109
9b639e2c
JR
241102013-02-26 Joern Rennecke <joern.rennecke@embecosm.com>
24111
24112 PR target/48901
24113 * config/lm32/lm32.c (gen_int_relational): Remove unused variables
24114 temp, cond and label.
24115 * config/lm32/lm32.md (ashlsi3): Remove unused variable one.
24116
e13a0ccb
JR
24117 PR target/52500
24118 * config/c6x/c6x.c (dbx_register_map): Change to unsigned.
24119 * config/c6x/c6x.h (dbx_register_map): Update declaration.
24120
fbe4f171
JR
24121 PR target/52501
24122 * config/cr16/cr16-protos.h: Move end of RTX_CODE guard below end
24123 of prologue/epilogue functions.
24124
ae006543
JR
24125 PR target/52550
24126 * config/tilegx/tilegx.c (tilegx_expand_prologue):
24127 Remove unused variable cfa_offset.
24128 * config/tilepro/tilepro.c (tilepro_expand_prologue): Likewise.
24129
c81369fa
JR
24130 PR target/54639
24131 * config/mn10300/mn10300.c (mn10300_expand_epilogue): Avoid offset
24132 type promotion to unsigned.
24133
f8a8fea7
JR
24134 PR target/54640
24135 * config/arm/arm.c (const_ok_for_dimode_op): Make code consistent
24136 for HOST_WIDE_INT of 32 bit / same size as int.
24137 (arm_block_move_unaligned_straight): Likewise.
24138 (arm_adjust_block_mem): Likewise.
24139
f8be5169
JR
24140 PR target/54662
24141 * config/mep/t-mep (mep-pragma.o): Use ALL_COMPILERFLAGS instead of
24142 ALL_CFLAGS.
24143
f1ad3354
MP
241442013-02-26 Marek Polacek <polacek@redhat.com>
24145
24146 PR tree-optimization/56426
88b97037 24147 * tree-ssa-loop.c (tree_ssa_loop_init): Always call scev_initialize.
f1ad3354 24148
a6af0f42
RB
241492013-02-26 Richard Biener <rguenther@suse.de>
24150
24151 PR target/56444
24152 * config/mn10300/mn10300.c (mn10300_scan_for_setlb_lcc): Remove
24153 unused variable loops.
24154
aca43c6c
JJ
241552013-02-26 Jakub Jelinek <jakub@redhat.com>
24156
cecbe5d9
JJ
24157 PR tree-optimization/56448
24158 * fold-const.c (operand_equal_p) <case tcc_reference>: Don't look at
24159 TREE_SIDE_EFFECTS if flags contain OEP_CONSTANT_ADDRESS_OF.
24160 Clear OEP_CONSTANT_ADDRESS_OF from flags before recursing on second or
24161 later operands of the references, or even first operand for
24162 INDIRECT_REF, TARGET_MEM_REF or MEM_REF.
24163
aca43c6c
JJ
24164 PR tree-optimization/56443
24165 * tree-vect-stmts.c (get_vectype_for_scalar_type_and_size): For
24166 overaligned types, pass TYPE_UNSIGNED (scalar_type) as second argument
24167 to type_for_mode langhook.
24168
53e2e141
MT
241692013-02-25 Matt Turner <mattst88@gmail.com>
24170
24171 * doc/invoke.texi: Document r4700.
24172
259ee451
RB
241732013-02-25 Richard Biener <rguenther@suse.de>
24174
24175 PR tree-optimization/56175
24176 * tree-ssa-forwprop.c (hoist_conversion_for_bitop_p): New predicate,
24177 split out from ...
24178 (simplify_bitwise_binary): ... here. Also guard the conversion
24179 of (type) X op CST to (type) (X op ((type-x) CST)) with it.
24180
6d65e8f1
CM
241812013-02-25 Catherine Moore <clm@codesourcery.com>
24182
24183 Revert:
24184 2013-02-24 Catherine Moore <clm@codesourcery.com>
24185 Maciej W. Rozycki <macro@codesourcery.com>
24186 Tom de Vries <tom@codesourcery.com>
a72d8780 24187 Nathan Sidwell <nathan@codesourcery.com>
6d65e8f1
CM
24188 Iain Sandoe <iain@codesourcery.com>
24189 Nathan Froyd <froydnj@codesourcery.com>
a72d8780 24190 Chao-ying Fu <fu@mips.com>
88b97037 24191
6d65e8f1 24192 * doc/extend.texi: (micromips, nomicromips, nocompression):
88b97037 24193 Document new function attributes.
6d65e8f1
CM
24194 * doc/invoke.texi (minterlink-compressed, mmicromips,
24195 m14k, m14ke, m14kec): Document new options.
24196 (minterlink-mips16): Update documentation.
24197 * doc/md.texi (ZC, ZD): Document new constraints.
24198 * configure.ac (gcc_cv_as_micromips): Check if linker
24199 supports the .set micromips directive.
24200 * configure: Regenerate.
24201 * config.in: Regenerate.
24202 * config/mips/mips-tables.opt: Regenerate.
24203 * config/mips/micromips.md: New file.
24204 * constraints.md (ZC, AD): New constraints.
24205 * config/mips/predicates.md (movep_src_register): New predicate.
24206 (movep_src_operand): New predicate.
24207 (non_volatile_mem_operand): New predicate.
24208 * config/mips/mips.md (multimem): New type.
24209 (length): Differentiate between 17-bit and 18-bit branch offsets.
24210 (MOVEP1, MOVEP2): New mode iterator.
24211 (mov_<load>l): Use ZC constraint.
24212 (mov_<load>r): Likewise.
24213 (mov_<store>l): Likewise.
24214 (mov_<store>r): Likewise.
24215 (*branch_equality<mode>_inverted): Add microMIPS support.
24216 (*branch_equality<mode>): Likewise.
24217 (*jump_absolute): Likewise.
24218 (indirect_jump_<mode>): Likewise.
24219 (tablejump_<mode>): Likewise.
24220 (<optab>_internal): Likewise.
24221 (sibcall_internal): Likewise.
24222 (sibcall_value_internal): Likewise.
24223 (prefetch): Use constraint ZD.
24224 * config/mips/mips.opt (minterlink-compressed): New option.
24225 (minterlink-mips16): Now an alias for minterlink-compressed.
24226 (mmicromips): New option.
24227 * config/mips/sync.md (sync_compare_and_swap<mode>): Use ZR constraint.
24228 (compare_and_swap_12): Likewise.
24229 (sync_add<mode>): Likewise.
24230 (sync_<optab>_12): Likewise.
24231 (sync_old_<optab>_12): Likewise.
24232 (sync_new_<optab>_12): Likewise.
24233 (sync_nand_12): Likewise.
24234 (sync_old_nand_12): Likewise.
24235 (sync_new_nand_12): Likewise.
24236 (sync_sub<mode>): Likewise.
24237 (sync_old_add<mode>): Likewise.
24238 (sync_old_sub<mode>): Likewise.
24239 (sync_new_add<mode>): Likewise.
24240 (sync_new_sub<mode>): Likewise.
24241 (sync_<optab><mode>): Likewise.
24242 (sync_old_<optab><mode>): Likewise.
24243 (sync_new_<optab><mode>): Likewise.
24244 (sync_nand<mode>): Likewise.
24245 (sync_old_nand<mode>): Likewise.
24246 (sync_new_nand<mode>): Likewise.
24247 (sync_lock_test_and_set<mode>): Likewise.
24248 (test_and_set_12): Likewise.
24249 (atomic_compare_and_swap<mode>): Likewise.
24250 (atomic_exchange<mode>_llsc): Likewise.
24251 (atomic_fetch_add<mode>_llsc): Likewise.
24252 * config/mips/mips-cpus.def (m14kc, m14k): New processors.
24253 * config/mips/mips-protos.h (umips_output_save_restore): New prototype.
24254 (umips_save_restore_pattern_p): Likewise.
24255 (umips_load_store_pair_p): Likewise.
24256 (umips_output_load_store_pair): Likewise.
24257 (umips_movep_target_p): Likewise.
24258 (umips_12bit_offset_address_p): Likewise.
24259 * config/mips/mips.c (MIPS_MAX_FIRST_STEP): Update for microMIPS.
24260 (mips_base_mips16): Rename this...
24261 (mips_base_compression_flags): ...to this. Update all uses.
24262 (mips_attribute_table): Add micromips, nomicromips and nocompression.
24263 (mips_mips16_decl_p): Delete.
24264 (mips_nomips16_decl_p): Delete.
24265 (mips_get_compress_on_flags): New function.
24266 (mips_get_compress_off_flags): New function.
24267 (mips_get_compress_mode): New function.
24268 (mips_get_compress_on_name): New function.
24269 (mips_get_compress_off_name): New function.
24270 (mips_insert_attributes): Support multiple compression types.
24271 (mips_merge_decl_attributes): Likewise.
24272 (umips_12bit_offset_address_p): New function.
24273 (mips_start_function_definition): Emit .set micromips directive.
24274 (mips_call_may_need_jalx_p): New function.
24275 (mips_function_ok_for_sibcall): Add microMIPS support.
24276 (mips_print_operand_punctuation): Support short delay slots and
24277 compact jumps.
24278 (umips_swm_mask, umips_swm_encoding): New.
24279 (umips_build_save_restore): New function.
24280 (mips_for_each_saved_gpr_and_fpr): Add microMIPS support.
24281 (was_mips16_p): Remove.
24282 (old_compression_mode): New.
24283 (mips_set_compression_mode): New function.
24284 (mips_set_current_function): Add microMIPS support.
24285 (mips_option_override): Likewise.
24286 (umips_save_restore_pattern_p): New function.
24287 (umips_output_save_restore): New function.
24288 (umips_load_store_pair_p_1): New function.
24289 (umips_load_store_pair_p): New function.
24290 (umips_output_load_store_pair_1): New function.
24291 (umips_output_load_store_pair): New function.
24292 (umips_movep_target_p) New function.
24293 (mips_prepare_pch_save): Add microMIPS support.
24294 * config/mips/mips.h (TARGET_COMPRESSION): New.
24295 (TARGET_CPU_CPP_BUILTINS): Update macro
24296 to use new compression flags and to support microMIPS.
24297 (MIPS_ISA_LEVEL_SPEC): Add m14k processors.
24298 (MIPS_ARCH_FLOAT_SPEC): Likewise.
24299 (ISA_HAS_LWXS): Include TARGET_MICROMIPS.
24300 (ISA_HAS_LOAD_DELAY): Exclude TARGET_MICROMIPS.
24301 (ASM_SPEC): Support mmicromips and mno-micromips.
24302 (M16STORE_REG_P): New macro.
24303 (MIPS_CALL): Support TARGET_MICROMIPS.
24304 (MICROMIPS_J): New macro.
24305 (mips_base_mips16): Rename this...
24306 (mips_base_compression_flags): ...to this.
24307 (UMIPS_12BIT_OFFSET_P): New macro.
24308 * config/mips/t-sde: (MULTILIB_OPTIONS): Add microMIPS.
24309 (MULTILIB_DIRNAMES): Likewise.
24310
5e5df392
TV
243112013-02-25 Tom de Vries <tom@codesourcery.com>
24312
24313 PR rtl-optimization/56131
24314 * insn-notes.def (INSN_NOTE_BASIC_BLOCK): Update comment.
24315 * cfgrtl.c (delete_insn): Don't reorder NOTE_INSN_DELETED_LABEL and
24316 NOTE_INSN_BASIC_BLOCK if BLOCK_FOR_INSN == NULL.
24317
3fdb53c1
TB
243182013-02-25 Tobias Burnus <burnus@net-b.de>
24319
24320 * doc/invoke.texi (-fsanitize=): Move from optimization
24321 to debugging options.
24322
ed358aea
AB
243232013-02-25 Andrey Belevantsev <abel@ispras.ru>
24324
24325 * sched-deps.c (sched_analyze_insn): Fix typo in comment.
24326
f45e9053
AB
243272013-02-25 Andrey Belevantsev <abel@ispras.ru>
24328 Alexander Monakov <amonakov@ispras.ru>
24329
24330 PR middle-end/56077
24331 * sched-deps.c (sched_analyze_insn): When reg_pending_barrier,
24332 flush pending lists also on non-jumps. Adjust comment.
24333
6941b508
CM
243342013-02-24 Catherine Moore <clm@codesourcery.com>
24335 Maciej W. Rozycki <macro@codesourcery.com>
24336 Tom de Vries <tom@codesourcery.com>
a72d8780 24337 Nathan Sidwell <nathan@codesourcery.com>
6941b508
CM
24338 Iain Sandoe <iain@codesourcery.com>
24339 Nathan Froyd <froydnj@codesourcery.com>
a72d8780 24340 Chao-ying Fu <fu@mips.com>
6941b508
CM
24341
24342 * doc/extend.texi: (micromips, nomicromips, nocompression):
88b97037 24343 Document new function attributes.
6941b508
CM
24344 * doc/invoke.texi (minterlink-compressed, mmicromips,
24345 m14k, m14ke, m14kec): Document new options.
24346 (minterlink-mips16): Update documentation.
24347 * doc/md.texi (ZC, ZD): Document new constraints.
24348 * configure.ac (gcc_cv_as_micromips): Check if linker
24349 supports the .set micromips directive.
24350 * configure: Regenerate.
24351 * config.in: Regenerate.
24352 * config/mips/mips-tables.opt: Regenerate.
24353 * config/mips/micromips.md: New file.
24354 * constraints.md (ZC, AD): New constraints.
24355 * config/mips/predicates.md (movep_src_register): New predicate.
24356 (movep_src_operand): New predicate.
24357 (non_volatile_mem_operand): New predicate.
24358 * config/mips/mips.md (multimem): New type.
24359 (length): Differentiate between 17-bit and 18-bit branch offsets.
24360 (MOVEP1, MOVEP2): New mode iterator.
3fdb53c1 24361 (mov_<load>l): Use ZC constraint.
6941b508
CM
24362 (mov_<load>r): Likewise.
24363 (mov_<store>l): Likewise.
24364 (mov_<store>r): Likewise.
24365 (*branch_equality<mode>_inverted): Add microMIPS support.
24366 (*branch_equality<mode>): Likewise.
24367 (*jump_absolute): Likewise.
24368 (indirect_jump_<mode>): Likewise.
24369 (tablejump_<mode>): Likewise.
24370 (<optab>_internal): Likewise.
24371 (sibcall_internal): Likewise.
24372 (sibcall_value_internal): Likewise.
24373 (prefetch): Use constraint ZD.
24374 * config/mips/mips.opt (minterlink-compressed): New option.
24375 (minterlink-mips16): Now an alias for minterlink-compressed.
24376 (mmicromips): New option.
24377 * config/mips/sync.md (sync_compare_and_swap<mode>): Use ZR constraint.
24378 (compare_and_swap_12): Likewise.
24379 (sync_add<mode>): Likewise.
24380 (sync_<optab>_12): Likewise.
24381 (sync_old_<optab>_12): Likewise.
24382 (sync_new_<optab>_12): Likewise.
24383 (sync_nand_12): Likewise.
24384 (sync_old_nand_12): Likewise.
24385 (sync_new_nand_12): Likewise.
24386 (sync_sub<mode>): Likewise.
24387 (sync_old_add<mode>): Likewise.
24388 (sync_old_sub<mode>): Likewise.
24389 (sync_new_add<mode>): Likewise.
24390 (sync_new_sub<mode>): Likewise.
24391 (sync_<optab><mode>): Likewise.
24392 (sync_old_<optab><mode>): Likewise.
24393 (sync_new_<optab><mode>): Likewise.
24394 (sync_nand<mode>): Likewise.
24395 (sync_old_nand<mode>): Likewise.
24396 (sync_new_nand<mode>): Likewise.
24397 (sync_lock_test_and_set<mode>): Likewise.
24398 (test_and_set_12): Likewise.
24399 (atomic_compare_and_swap<mode>): Likewise.
24400 (atomic_exchange<mode>_llsc): Likewise.
24401 (atomic_fetch_add<mode>_llsc): Likewise.
24402 * config/mips/mips-cpus.def (m14kc, m14k): New processors.
24403 * config/mips/mips-protos.h (umips_output_save_restore): New prototype.
24404 (umips_save_restore_pattern_p): Likewise.
24405 (umips_load_store_pair_p): Likewise.
24406 (umips_output_load_store_pair): Likewise.
24407 (umips_movep_target_p): Likewise.
24408 (umips_12bit_offset_address_p): Likewise.
24409 * config/mips/mips.c (MIPS_MAX_FIRST_STEP): Update for microMIPS.
24410 (mips_base_mips16): Rename this...
24411 (mips_base_compression_flags): ...to this. Update all uses.
24412 (mips_attribute_table): Add micromips, nomicromips and nocompression.
24413 (mips_mips16_decl_p): Delete.
24414 (mips_nomips16_decl_p): Delete.
3fdb53c1 24415 (mips_get_compress_on_flags): New function.
6941b508
CM
24416 (mips_get_compress_off_flags): New function.
24417 (mips_get_compress_mode): New function.
24418 (mips_get_compress_on_name): New function.
24419 (mips_get_compress_off_name): New function.
24420 (mips_insert_attributes): Support multiple compression types.
24421 (mips_merge_decl_attributes): Likewise.
24422 (umips_12bit_offset_address_p): New function.
24423 (mips_start_function_definition): Emit .set micromips directive.
24424 (mips_call_may_need_jalx_p): New function.
24425 (mips_function_ok_for_sibcall): Add microMIPS support.
24426 (mips_print_operand_punctuation): Support short delay slots and
24427 compact jumps.
24428 (umips_swm_mask, umips_swm_encoding): New.
24429 (umips_build_save_restore): New function.
24430 (mips_for_each_saved_gpr_and_fpr): Add microMIPS support.
24431 (was_mips16_p): Remove.
24432 (old_compression_mode): New.
24433 (mips_set_compression_mode): New function.
24434 (mips_set_current_function): Add microMIPS support.
24435 (mips_option_override): Likewise.
24436 (umips_save_restore_pattern_p): New function.
24437 (umips_output_save_restore): New function.
24438 (umips_load_store_pair_p_1): New function.
24439 (umips_load_store_pair_p): New function.
24440 (umips_output_load_store_pair_1): New function.
24441 (umips_output_load_store_pair): New function.
24442 (umips_movep_target_p) New function.
24443 (mips_prepare_pch_save): Add microMIPS support.
24444 * config/mips/mips.h (TARGET_COMPRESSION): New.
24445 (TARGET_CPU_CPP_BUILTINS): Update macro
24446 to use new compression flags and to support microMIPS.
24447 (MIPS_ISA_LEVEL_SPEC): Add m14k processors.
24448 (MIPS_ARCH_FLOAT_SPEC): Likewise.
24449 (ISA_HAS_LWXS): Include TARGET_MICROMIPS.
24450 (ISA_HAS_LOAD_DELAY): Exclude TARGET_MICROMIPS.
24451 (ASM_SPEC): Support mmicromips and mno-micromips.
24452 (M16STORE_REG_P): New macro.
24453 (MIPS_CALL): Support TARGET_MICROMIPS.
24454 (MICROMIPS_J): New macro.
24455 (mips_base_mips16): Rename this...
24456 (mips_base_compression_flags): ...to this.
24457 (UMIPS_12BIT_OFFSET_P): New macro.
24458 * config/mips/t-sde: (MULTILIB_OPTIONS): Add microMIPS.
24459 (MULTILIB_DIRNAMES): Likewise.
24460
cdebbc6d
JJ
244612013-02-24 Jakub Jelinek <jakub@redhat.com>
24462
24463 PR target/52555
24464 * target-globals.c (save_target_globals): For init_reg_sets and
24465 target_reinit remporarily set this_fn_optabs to this_target_optabs.
24466
18c63565
JG
244672013-02-22 James Grennahlgh <james.greenhalgh@arm.com>
24468
24469 * config/aarch64/aarch64-simd-builtins.def: Add copyright header.
24470 * config/aarch64/t-aarch64
24471 (aarch64-builtins.o): Depend on aarch64-simd-builtins.def.
24472
2194f7a2
VM
244732013-02-22 Vladimir Makarov <vmakarov@redhat.com>
24474
24475 PR inline-asm/56148
24476 * lra-constraints.c (process_alt_operands): Reload operand
24477 conflicting with earlier clobber only if no more other conflicting
24478 operands.
24479
7d613735
JJ
244802013-02-22 Jakub Jelinek <jakub@redhat.com>
24481
24482 PR sanitizer/56393
24483 * config/gnu-user.h (LIBASAN_EARLY_SPEC): Link in libasan_preinit.o
24484 if not linking a shared library.
24485
ac8d93a7
SL
244862013-02-22 Seth LaForge <sethml@google.com>
24487
24488 * config.gcc (arm*-*-eabi*): Treat arm*eb as big-endian.
24489
e0237780
GY
244902013-02-22 Greta Yorsh <Greta.Yorsh@arm.com>
24491
3fdb53c1
TB
24492 * config/arm/arm.md (split for extendsidi): Update condition.
24493 (zero_extend<mode>di2,extend<mode>di2): Add an alternative.
24494 * config/arm/iterators.md (qhs_extenddi_cstr): Likewise.
24495 (qhs_zextenddi_cstr): Likewise.
e0237780 24496
d7fde18c
JJ
244972013-02-21 Jakub Jelinek <jakub@redhat.com>
24498
be63b77d
JJ
24499 PR middle-end/56420
24500 * expmed.c (EXACT_POWER_OF_2_OR_ZERO_P): Do subtraction in uhwi, to
24501 avoid signed wrapping.
24502 (expand_mult): Handle properly multiplication by
24503 ((dword_type) -1) << (BITS_PER_WORD - 1). Improve multiplication by
24504 ((dword_type) 1) << (BITS_PER_WORD - 1). Avoid undefined behavior
24505 in the compiler if coeff is HOST_WIDE_INT_MIN.
24506 (expand_divmod): Don't make ext_op1 static, change it's type to
24507 uhwi. Avoid undefined behavior in -INTVAL (op1).
24508
d7fde18c
JJ
24509 PR rtl-optimization/50339
24510 * lower-subreg.h (struct lower_subreg_choices): Add splitting_ashiftrt
24511 field.
24512 * lower-subreg.c (compute_splitting_shift): Handle ASHIFTRT.
24513 (compute_costs): Call compute_splitting_shift also for ASHIFTRT
24514 into splitting_ashiftrt field.
24515 (find_decomposable_shift_zext, resolve_shift_zext): Handle also
24516 ASHIFTRT.
24517 (dump_choices): Fix up printing LSHIFTRT choices, print ASHIFTRT
24518 choices.
24519
6aad4455
AH
245202013-02-20 Aldy Hernandez <aldyh@redhat.com>
24521
24522 PR middle-end/56108
24523 * trans-mem.c (execute_tm_mark): Do not expand transactions that
24524 are sure to go irrevocable.
24525
38fe784d
HPN
245262013-02-21 Hans-Peter Nilsson <hp@axis.com>
24527
24528 * doc/rtl.texi (vec_concat, vec_duplicate): Mention that
24529 scalars are valid operands.
24530
0fd44da3
MJ
245312013-02-21 Martin Jambor <mjambor@suse.cz>
24532
24533 PR tree-optimization/56310
24534 * ipa-cp.c (agg_replacements_to_vector): New parameter index, copy
24535 only matching indices and non-negative final offsets.
24536 (intersect_aggregates_with_edge): Pass src_idx to
24537 agg_replacements_to_vector. Pass src_idx insstead of index to
24538 intersect_with_agg_replacements.
24539
7a92038b
MJ
245402013-02-21 Martin Jambor <mjambor@suse.cz>
24541
24542 * ipa-cp.c (good_cloning_opportunity_p): Dump the real threshold
24543 instead of hard-wired defaults.
24544
c0da9c37
MR
245452013-02-21 Maciej W. Rozycki <macro@codesourcery.com>
24546
24547 * doc/invoke.texi (MIPS Options): Update documentation of the
24548 floating-point multiply-accumulate instruction restrictions.
24549
d247ea0c 245502013-02-21 Kostya Serebryany <kcc@google.com>
7df59255
KS
24551
24552 * config/i386/i386.c (ix86_asan_shadow_offset): Use 0x7fff8000 as
24553 asan_shadow_offset on x86_64 linux.
24554
22deefcb
RB
245552013-02-21 Richard Biener <rguenther@suse.de>
24556
24557 PR tree-optimization/56415
24558 Revert
24559 2013-02-11 Richard Biener <rguenther@suse.de>
24560
24561 PR tree-optimization/56273
24562 * tree-vrp.c (simplify_cond_using_ranges): Disable for the
24563 first VRP run.
24564
7bcc6e75
JJ
245652013-02-21 Jakub Jelinek <jakub@redhat.com>
24566
a0ad148f
JJ
24567 PR bootstrap/56258
24568 * doc/invoke.texi (-fdump-rtl-pro_and_epilogue): Use @item
24569 instead of @itemx.
24570
7bcc6e75
JJ
24571 PR inline-asm/56405
24572 * expr.c (expand_expr_real_1) <case TARGET_MEM_REF, MEM_REF>: Don't
24573 use movmisalign or extract_bit_field for EXPAND_MEMORY modifier.
24574
a0a7b611
JH
245752013-02-20 Jan Hubicka <jh@suse.cz>
24576
24577 PR tree-optimization/56265
88b97037
UB
24578 * ipa-prop.c (ipa_make_edge_direct_to_target): Fixup callgraph
24579 when target is referenced for first time.
a0a7b611 24580
c0e50f72
RB
245812013-02-20 Richard Biener <rguenther@suse.de>
24582
24583 * tree-call-cdce.c (tree_call_cdce): Do not remove unused locals.
24584 * tree-ssa-forwprop.c (ssa_forward_propagate_and_combine): Likewise.
24585 * tree-ssa-dce.c (perform_tree_ssa_dce): Likewise.
24586 * tree-ssa-copyrename.c (copy_rename_partition_coalesce): Do
24587 not return anything.
24588 (rename_ssa_copies): Do not remove unused locals.
24589 * tree-ssa-ccp.c (do_ssa_ccp): Likewise.
88b97037 24590 * tree-ssanames.c (pass_release_ssa_names): Remove unused locals first.
c0e50f72
RB
24591 * passes.c (execute_function_todo): Do not schedule unused locals
24592 removal if cleanup_tree_cfg did something.
24593 * tree-ssa-live.c (remove_unused_locals): Dump statistics
24594 about the number of removed locals.
24595
a52ca739
RB
245962013-02-20 Richard Biener <rguenther@suse.de>
24597
24598 PR tree-optimization/56398
88b97037 24599 * tree-vect-loop-manip.c (adjust_debug_stmts): Skip SSA default defs.
a52ca739 24600
ec9c9d1b
MJ
246012013-02-20 Martin Jambor <mjambor@suse.cz>
24602
24603 PR tree-optimization/55334
24604 * ipa-cp.c (initialize_node_lattices): Disable IPA-CP through and to
24605 restricted pointers to arrays.
24606
e91c8ed6 246072013-02-20 Richard Biener <rguenther@suse.de>
88b97037 24608 Jakub Jelinek <jakub@redhat.com>
e91c8ed6
RB
24609
24610 PR tree-optimization/56396
24611 * tree-ssa-ccp.c (n_const_val): New static variable.
24612 (get_value): Return NULL for SSA names we don't have a lattice
24613 entry for.
24614 (ccp_initialize): Initialize n_const_val.
24615 * tree-ssa-copy.c (n_copy_of): New static variable.
24616 (init_copy_prop): Initialize n_copy_of.
24617 (get_value): Return NULL_TREE for SSA names we don't have a
24618 lattice entry for.
24619
3d916479
MJ
246202013-02-20 Martin Jambor <mjambor@suse.cz>
24621
24622 * ipa-cp.c (initialize_node_lattices): Fix dumping condition.
24623
71a86758
RB
246242013-02-20 Richard Biener <rguenther@suse.de>
24625
24626 * genpreds.c (write_lookup_constraint): Do not compare first
24627 letter of the constraint again.
24628
79836a12
RB
246292013-02-20 Richard Biener <rguenther@suse.de>
24630
24631 * tree-ssa-loop-ivopts.c (alloc_use_cost_map): Use bitmap_count_bits
24632 and ceil_log2.
24633 (get_use_iv_cost): Terminate hashtable walk when coming across
24634 an empty entry.
24635
bbe4fb2c
IZ
246362013-02-20 Igor Zamyatin <igor.zamyatin@intel.com>
24637
24638 * config/i386/i386.c (initial_ix86_tune_features): Turn on fp
24639 reassociation for avx2 targets.
24640
a72d8780 246412012-02-19 Edgar E. Iglesias <edgar.iglesias@gmail.com>
79836a12 24642
c77f83d5
EI
24643 * config/microblaze/microblaze.c: microblaze_has_clz = 0
24644 Add version check for v8.10.a to enable microblaze_has_clz
88b97037 24645 * config/microblaze/microblaze.h: Add TARGET_HAS_CLZ as combined
c77f83d5
EI
24646 version and TARGET_PATTERN_COMPARE check
24647 * config/microblaze/microblaze.md: New clzsi2 instruction
3fdb53c1 24648
a72d8780 246492012-02-19 Edgar E. Iglesias <edgar.iglesias@gmail.com>
959434e6 24650
a72d8780 24651 * config/microblaze/microblaze.md (call_value_intern): Check symbol is
959434e6
EI
24652 function before branching.
24653
242387fa
AB
246542012-02-19 Andrey Belevantsev <abel@ispras.ru>
24655
24656 * sel-sched-dump.c (dump_insn_rtx_flags): Explicitly set
24657 DUMP_INSN_RTX_UID.
24658 (dump_insn_rtx_1): Pass PATTERN (insn) to str_pattern_slim.
24659
b4979ab9
AB
246602012-02-19 Andrey Belevantsev <abel@ispras.ru>
24661
24662 PR middle-end/55889
b4979ab9
AB
24663 * sel-sched.c: Include ira.h.
24664 (implicit_clobber_conflict_p): New function.
24665 (moveup_expr): Use it.
88b97037 24666 * Makefile.in (sel-sched.o): Depend on ira.h.
b4979ab9 24667
24d63016
RB
246682013-02-19 Richard Biener <rguenther@suse.de>
24669
24670 PR tree-optimization/56384
24671 * tree-ssa-sccvn.h (struct vn_phi_s): Add type member.
24672 (vn_hash_type): Split out from ...
24673 (vn_hash_constant_with_type): ... here.
24674 * tree-ssa-sccvn.c (vn_phi_compute_hash): Use vn_hash_type.
24675 (vn_phi_eq): Compare types from vn_phi_s structure.
24676 (vn_phi_lookup): Populate vn_phi_s type.
24677 (vn_phi_insert): Likewise.
24678
a475fd3d
JJ
246792013-02-19 Jakub Jelinek <jakub@redhat.com>
24680
47cc28f5
JJ
24681 PR tree-optimization/56350
24682 * tree-vect-loop.c (vectorizable_reduction): If orig_stmt, return false
24683 if haven't found reduction or nested cycle operand, rather than
24684 asserting we must find it.
24685
a475fd3d
JJ
24686 PR tree-optimization/56381
24687 * tree-ssa-pre.c (create_expression_by_pieces): Fix up last argument
24688 to fold_build3.
24689
246902013-02-18 Aldy Hernandez <aldyh@redhat.com>
135204dd
AH
24691 Jakub Jelinek <jakub@redhat.com>
24692
24693 PR target/52555
24694 * genopinit.c (raw_optab_handler): Use this_fn_optabs.
24695 (swap_optab_enable): Same.
24696 (init_all_optabs): Use argument instead of global.
88b97037 24697 * tree.h (struct tree_optimization_option): New field target_optabs.
135204dd
AH
24698 * expr.h (init_all_optabs): Add argument to prototype.
24699 (TREE_OPTIMIZATION_OPTABS): New.
24700 (save_optabs_if_changed): Protoize.
24701 * optabs.h: Declare this_fn_optabs.
24702 * optabs.c (save_optabs_if_changed): New.
24703 Declare this_fn_optabs.
24704 (init_optabs): Add argument to init_all_optabs() call.
24705 * function.c (invoke_set_current_function_hook): Handle per
24706 function optabs.
24707 * function.h (struct function): New field optabs.
24708 * config/mips/mips.c (mips_set_mips16_mode): Handle when
24709 optimization_current_node has changed.
24710 * target-globals.h (save_target_globals_default_opts): Protoize.
24711 * target-globals.c (save_target_globals_default_opts): New.
24712
3f587ca3
JDA
247132013-02-18 John David Anglin <dave.anglin@nrc-cnrc.gc.ca>
24714
24715 PR target/56347
24716 * config/pa/pa.c (pa_conditional_register_usage): On HP-UX, mark
24717 registers %fr12 and %fr12R as call used.
ceaca33e
JDA
24718
24719 PR target/56214
24720 * config/pa/predicates.md (base14_operand): Except for BLKmode, QImode
88b97037
UB
24721 and HImode, require all displacements to be an integer multiple of
24722 their mode size.
ceaca33e
JDA
24723 * config/pa/pa.c (pa_legitimate_address_p): For REG+BASE addresses,
24724 only allow QImode and HImode when reload is in progress and strict is
24725 true. Likewise for symbolic addresses. Use base14_operand to check
24726 displacements in REG+BASE addresses.
24727
fe0b4796
RB
247282013-02-18 Richard Biener <rguenther@suse.de>
24729
24730 PR tree-optimization/56366
24731 * tree-vect-loop.c (get_initial_def_for_induction): Properly
24732 handle sign-conversion of outer-loop initial induction value.
24733
6aaf596b
RB
247342013-02-18 Richard Biener <rguenther@suse.de>
24735
73db8ff1 24736 PR middle-end/56349
6aaf596b
RB
24737 * cfghooks.c (merge_blocks): If we merge a latch into another
24738 block adjust references to it.
24739 * cfgloop.c (flow_loops_find): Reset latch before recomputing it.
24740 (verify_loop_structure): Verify that a recorded latch is in fact
24741 a latch.
24742
5e97dfb6
RB
247432013-02-18 Richard Biener <rguenther@suse.de>
24744
24745 PR tree-optimization/56321
24746 * tree-ssa-reassoc.c (propagate_op_to_single_use): Properly
24747 order SSA name release and virtual operand unlinking.
24748
825527e8
EI
247492013-02-17 Edgar E. Iglesias <edgar.iglesias@gmail.com>
24750
24751 * config/microblaze/microblaze.md (save_stack_block): Define.
24752 (restore_stack_block): Likewise.
24753
debd11d9
EI
247542013-02-16 Edgar E. Iglesias <edgar.iglesias@gmail.com>
24755
24756 * config/microblaze/linux.h (TARGET_SUPPORTS_PIC): Define as 1.
24757 * config/microblaze/microblaze.h (TARGET_SUPPORTS_PIC): Define as 1.
24758 * config/microblaze/microblaze.c (microblaze_option_override):
24759 Bail out early for PIC modes when target does not support PIC.
24760
8ec77be0
EI
247612013-02-16 Edgar E. Iglesias <edgar.iglesias@gmail.com>
24762
88b97037 24763 * config/microblaze/microblaze.c (microblaze_asm_trampoline_template):
8ec77be0
EI
24764 Replace with a microblaze version.
24765 (microblaze_trampoline_init): Adapt for microblaze.
a752b80f 24766 * config/microblaze/microblaze.h (TRAMPOLINE_SIZE): Adapt for
8ec77be0
EI
24767 microblaze.
24768
b41288b3
JJ
247692013-02-16 Jakub Jelinek <jakub@redhat.com>
24770 Dodji Seketeli <dodji@redhat.com>
24771
24772 PR asan/56330
88b97037 24773 * asan.c (get_mem_refs_of_builtin_call): White space and style cleanup.
b41288b3
JJ
24774 (instrument_mem_region_access): Do not forget to always put
24775 instrumentation of the of 'base' and 'base + len' in a "if (len !=
24776 0) statement, even for cases where either 'base' or 'base + len'
24777 are not instrumented -- because they have been previously
24778 instrumented. Simplify the logic by putting all the statements
24779 instrument 'base + len' inside a sequence, and then insert that
24780 sequence right before the current insertion point. Then, to
24781 instrument 'base + len', just get an iterator on that statement.
24782 And do not forget to update the pointer to iterator the function
24783 received as argument.
24784
47918951
VM
247852013-02-15 Vladimir Makarov <vmakarov@redhat.com>
24786
24787 PR rtl-optimization/56348
24788 * lra-assigns.c (reload_pseudo_compare_func): Prefer bigger pseudos.
24789
3fb7c699
SB
247902013-02-15 Steven Bosscher <steven@gcc.gnu.org>
24791
24792 * graph.c (start_graph_dump): Print dumpfile base as digraph label.
24793 (clean_graph_dump_file): Pass base to start_graph_dump.
24794
576fe41a
RH
247952013-02-14 Richard Henderson <rth@redhat.com>
24796
24797 PR target/55941
24798 * lower-subreg.c (simple_move): Check dest mode instead of src mode.
24799
bb3accfa
SB
248002013-02-14 Steven Bosscher <steven@gcc.gnu.org>
24801
24802 * collect2-aix.h: Define F_LOADONLY.
24803
03143140
RB
248042013-02-14 Richard Biener <rguenther@suse.de>
24805
24806 PR lto/50494
24807 * varasm.c (output_constant_def_1): Get the decl representing
24808 the constant as argument.
24809 (output_constant_def): Wrap output_constant_def_1.
24810 (make_decl_rtl): Use output_constant_def_1 with the decl
24811 representing the constant.
24812 (build_constant_desc): Optionally re-use a decl already
24813 representing the constant.
24814 (tree_output_constant_def): Adjust.
24815
8fb06726
DS
248162013-02-14 Dodji Seketeli <dodji@redhat.com>
24817
24818 Fix an asan crash
24819 * asan.c (instrument_builtin_call): Really put the length of the
24820 second source argument into src1_len.
24821
c4bfe8bf
JJ
248222013-02-13 Jakub Jelinek <jakub@redhat.com>
24823
24824 * asan.c (create_cond_insert_point): Add create_then_fallthru_edge
24825 argument. If it is false, don't create edge from then_bb to
24826 fallthru_bb.
24827 (insert_if_then_before_iter): Pass true to it.
24828 (build_check_stmt): Pass false to it.
24829 (transform_statements): Flush hash table only on extended basic
24830 block boundaries, rather than at the beginning of every bb.
24831 Don't flush hash table on nonfreeing_call_p calls.
24832 * tree-flow.h (nonfreeing_call_p): New prototype.
24833 * tree-ssa-phiopt.c (nonfreeing_call_p): No longer static.
24834
7afe2801
DM
248352013-02-13 David S. Miller <davem@davemloft.net>
24836
24837 * expmed.c (expand_shift_1): Only strip scalar integer subregs.
24838
70cc3288
VM
248392013-02-13 Vladimir Makarov <vmakarov@redhat.com>
24840
24841 PR target/56184
24842 * ira.c (max_regno_before_ira): Move from ...
24843 (ira): ... here.
24844 (fix_reg_equiv_init): Use max_regno_before_ira instead of
24845 vec_safe_length.
24846
6422242b
JJ
248472013-02-13 Jakub Jelinek <jakub@redhat.com>
24848
24849 * config/i386/i386.c (ix86_asan_shadow_offset): Revert last change.
24850
976a81ee
RB
248512013-02-13 Richard Biener <rguenther@suse.de>
24852
24853 PR lto/56295
24854 * gimple-streamer-out.c (output_gimple_stmt): Undo wrapping
24855 globals in MEM_REFs.
24856
c1874a87
RB
248572013-02-13 Richard Biener <rguenther@suse.de>
24858
24859 * loop-init.c (loop_optimizer_init): Clear loop state when
24860 re-initializing preserved loops.
24861 * loop-unswitch.c (unswitch_single_loop): Return whether
24862 we unswitched the loop. Do not verify loop state here.
88b97037 24863 (unswitch_loops): When we unswitched a loop discover new loops.
c1874a87 24864
b4ab7d34
KS
248652013-02-13 Kostya Serebryany <kcc@google.com>
24866
88b97037
UB
24867 * config/i386/i386.c: Use 0x7fff8000 as asan_shadow_offset
24868 on x86_64 linux.
b4ab7d34
KS
24869 * sanitizer.def: Rename __asan_init to __asan_init_v1.
24870
bdcbe80c
DS
248712013-02-12 Dodji Seketeli <dodji@redhat.com>
24872
24873 Avoid instrumenting duplicated memory access in the same basic block
24874 * Makefile.in (asan.o): Add new dependency on hash-table.h
24875 * asan.c (struct asan_mem_ref, struct mem_ref_hasher): New types.
24876 (asan_mem_ref_init, asan_mem_ref_get_end, get_mem_ref_hash_table)
24877 (has_stmt_been_instrumented_p, empty_mem_ref_hash_table)
24878 (free_mem_ref_resources, has_mem_ref_been_instrumented)
24879 (has_stmt_been_instrumented_p, update_mem_ref_hash_table)
24880 (get_mem_ref_of_assignment): New functions.
24881 (get_mem_refs_of_builtin_call): Extract from
24882 instrument_builtin_call and tweak a little bit to make it fit with
24883 the new signature.
24884 (instrument_builtin_call): Use the new
24885 get_mem_refs_of_builtin_call. Use gimple_call_builtin_p instead
24886 of is_gimple_builtin_call.
24887 (instrument_derefs, instrument_mem_region_access): Insert the
24888 instrumented memory reference into the hash table.
24889 (maybe_instrument_assignment): Renamed instrument_assignment into
24890 this, and change it to advance the iterator when instrumentation
24891 actually happened and return true in that case. This makes it
24892 homogeneous with maybe_instrument_assignment, and thus give a
24893 chance to callers to be more 'regular'.
24894 (transform_statements): Clear the memory reference hash table
24895 whenever we enter a new BB, when we cross a function call, or when
24896 we are done transforming statements. Use
24897 maybe_instrument_assignment instead of instrumentation. No more
24898 need to special case maybe_instrument_assignment and advance the
24899 iterator after calling it; it's now handled just like
24900 maybe_instrument_call. Update comment.
24901
4861a1f7
RB
249022013-02-13 Richard Biener <rguenther@suse.de>
24903
24904 * config/mn10300/mn10300.c (mn10300_scan_for_setlb_lcc):
24905 Fix loop discovery code.
24906
1c86bd80
VM
249072013-02-12 Vladimir Makarov <vmakarov@redhat.com>
24908
24909 PR inline-asm/56148
24910 * lra-constraints.c (process_alt_operands): Match early clobber
70cc3288 24911 operand with itself. Check conflicts with earlyclobber only if
1c86bd80
VM
24912 the operand is not reloaded. Prefer to reload conflicting operand
24913 if earlyclobber and matching operands are the same.
24914
7cbda518
RB
249152013-02-12 Richard Biener <rguenther@suse.de>
24916
24917 PR lto/56297
24918 * lto-streamer-out.c (write_symbol): Do not output symbols
24919 for hard register variables.
24920
e68a4ef6
GJL
249212013-02-12 Georg-Johann Lay <avr@gjlay.de>
24922
24923 PR target/54222
24924 * config/avr/avr-dimode.md (umulsidi3, mulsidi3): New expanders.
24925 (umulsidi3_insn, mulsidi3_insn): New insns.
24926
a72d8780 249272013-02-12 Christophe Lyon <christophe.lyon@linaro.org>
2597da22
CL
24928
24929 * config/arm/arm-protos.h (struct cpu_vec_costs): New struct type.
24930 (struct tune_params): Add vec_costs field.
24931 * config/arm/arm.c (arm_builtin_vectorization_cost)
24932 (arm_add_stmt_cost): New functions.
24933 (TARGET_VECTORIZE_BUILTIN_VECTORIZATION_COST)
24934 (TARGET_VECTORIZE_ADD_STMT_COST): Define.
24935 (arm_default_vec_cost): New struct of type cpu_vec_costs.
24936 (arm_slowmul_tune, arm_fastmul_tune, arm_strongarm_tune)
24937 (arm_xscale_tune, arm_9e_tune, arm_v6t2_tune, arm_cortex_tune)
24938 (arm_cortex_a15_tune, arm_cortex_a5_tune, arm_cortex_a9_tune)
24939 (arm_v6m_tune, arm_fa726te_tune): Define new vec_costs field.
24940
43320568
RB
249412013-02-12 Richard Biener <rguenther@suse.de>
24942
24943 PR lto/56295
24944 * gimple-streamer-in.c (input_gimple_stmt): Strip MEM_REFs off
24945 decls again if possible.
24946
a011aa39
RB
249472013-02-12 Richard Biener <rguenther@suse.de>
24948
24949 PR middle-end/56288
24950 * tree-ssa.c (verify_ssa_name): Fix check, move
24951 SSA_NAME_IN_FREE_LIST check up.
24952
6da26889
JJ
249532013-02-12 Jakub Jelinek <jakub@redhat.com>
24954 Steven Bosscher <steven@gcc.gnu.org>
24955
24956 PR rtl-optimization/56151
24957 * optabs.c (add_equal_note): Don't return 0 if target is a MEM,
24958 equal to op0 or op1, and last_insn pattern is CODE operation
24959 with MEM dest and one of the operands matches that MEM.
24960
f80e0faf
ST
249612013-02-11 Sriraman Tallam <tmsriramgoogle.com>
24962
24963 * doc/extend.texi: Document Function Multiversioning and "default"
24964 parameter string to target attribute.
24965 * config/i386/i386.c (get_builtin_code_for_version): Return 0 if
24966 target attribute parameter is "default".
24967 (ix86_compare_version_priority): Remove checks for target attribute.
24968 (ix86_mangle_function_version_assembler_name): Change error to sorry.
24969 Remove check for target attribute equal to NULL. Add assert.
24970 (ix86_generate_version_dispatcher_body): Change error to sorry.
24971
6c59ffd1
IS
249722013-02-11 Iain Sandoe <iain@codesourcery.com>
24973 Jack Howarth <howarth@bromo.med.uc.edu>
24974 Patrick Marlier <patrick.marlier@gmail.com>
24975
24976 PR libitm/55693
24977 * config/darwin.h: Replace ENDFILE_SPEC with TM_DESTRUCTOR and
24978 define ENDFILE_SPEC as TM_DESTRUCTOR.
24979 * config/i386/darwin.h (ENDFILE_SPEC): Use TM_DESTRUCTOR.
24980
a72d8780 249812013-02-11 Alexander Potapenko <glider@google.com>
6a52ccb3
AP
24982 Jack Howarth <howarth@bromo.med.uc.edu>
24983 Jakub Jelinek <jakub@redhat.com>
24984
24985 PR sanitizer/55617
24986 * config/darwin.c (cdtor_record): Rename ctor_record.
24987 (sort_cdtor_records): Rename sort_ctor_records.
24988 (finalize_dtors): New routine to sort destructors by
24989 priority before use in assemble_integer.
24990 (machopic_asm_out_destructor): Use finalize_dtors if needed.
24991
b63fe007
UB
249922013-02-11 Uros Bizjak <ubizjak@gmail.com>
24993
24994 PR rtl-optimization/56275
24995 * simplify-rtx.c (avoid_constant_pool_reference): Check that
24996 offset is non-negative and less than cmode size before
24997 calling simplify_subreg.
24998
8e89b5b5
RB
249992013-02-11 Richard Biener <rguenther@suse.de>
25000
25001 PR tree-optimization/56264
25002 * cfgloop.h (fix_loop_structure): Adjust prototype.
25003 * loop-init.c (fix_loop_structure): Return the number of
25004 newly discovered loops.
25005 * tree-cfgcleanup.c (repair_loop_structures): When new loops
25006 are discovered, do a full loop-closed SSA rewrite.
25007
b4a4b56d
RB
250082013-02-11 Richard Biener <rguenther@suse.de>
25009
25010 PR tree-optimization/56273
25011 * tree-vrp.c (simplify_cond_using_ranges): Disable for the
25012 first VRP run.
25013 (check_array_ref): Fix missing newline in dumps.
25014 (search_for_addr_array): Likewise.
25015
0c885229
DE
250162013-02-09 David Edelsohn <dje.gcc@gmail.com>
25017
25018 * config/rs6000/aix61.h (OS_MISSING_ALTIVEC): Undefine.
25019
59ac9a55
JJ
250202013-02-09 Jakub Jelinek <jakub@redhat.com>
25021
25022 PR target/56256
25023 * config/rs6000/rs6000.h (ASSEMBLER_DIALECT): Define.
25024
25bb0bb5
VM
250252013-02-08 Vladimir Makarov <vmakarov@redhat.com>
25026
25027 PR rtl-optimization/56246
0c885229 25028 * lra-constraints.c (simplify_operand_subreg): Try to reuse
25bb0bb5
VM
25029 reload pseudo.
25030 * lra.c (lra): Clear lra_optional_reload_pseudos only when all
25031 constraints are satisfied.
25032
a698cc03
JL
250332013-02-08 Jeff Law <law@redhat.com>
25034
25035 PR debug/53948
25036 * emit-rtl.c (reg_is_parm_p): New function.
25037 * regs.h (reg_is_parm_p): New prototype.
25038 * ira-conflicts.c (ira_build_conflicts): Allow parameters in
25039 callee-clobbered registers.
25040
e1122ddd
MM
250412013-02-08 Michael Meissner <meissner@linux.vnet.ibm.com>
25042
25043 PR target/56043
25044 * config/rs6000/rs6000.c (rs6000_builtin_vectorized_libmass):
25045 If there is no implicit builtin declaration, just return NULL.
25046
19c5f6e6
UB
250472013-02-08 Uros Bizjak <ubizjak@gmail.com>
25048
25049 * config/i386/sse.md (FMAMODEM): New mode iterator.
25050 (fma<mode>4, fms<mode>4, fnma<mode>4, fnms<mode>4): Use FMAMODEM
25051 mode iterator. Do not use TARGET_SSE_MATH in insn constraint.
25052
2480f2ca 250532013-02-08 Uros Bizjak <ubizjak@gmail.com>
76ef61fb 25054
2480f2ca
UB
25055 * config/i386/gnu-user.h (TARGET_CAN_SPLIT_STACK): Define only
25056 when HAVE_GAS_CFI_PERSONALITY_DIRECTIVE is set.
25057 * config/i386/gnu-user64.h (TARGET_CAN_SPLIT_STACK): Ditto.
25058
250592013-02-08 Edgar E. Iglesias <edgar.iglesias@gmail.com>
25060
25061 * config.gcc (microblaze*-linux*): Add TARGET_BIG_ENDIAN_DEFAULT.
25062 (microblaze*-*-elf): Likewise.
25063 * config/microblaze/linux.h: Add -mbig-endian / -mlittle-endian to
25064 LINK_SPEC.
25065 * config/microblaze/microblaze-c.c: Add builtin defines for
25066 _LITTLE_ENDIAN and _BIG_ENDIAN.
25067 * config/microblaze/microblaze.h: Add TARGET_ENDIAN_DEFAULT and
25068 add to TARGET_DEFAULT flags.
76ef61fb 25069 Expand ASM_SPEC and LINK_SPEC.
2480f2ca
UB
25070 Update BYTES_BIG_ENDIAN and WORDS_BIG_ENDIAN.
25071 * config/microblaze/microblaze.md: Update extendsidi2 and
25072 movdi_internal instructions to use low-order / high-order reg
25073 print_operands.
25074 * config/microblaze/microblaze.opt: Add mbig-endian and mlittle-endian
25075 options and inversemask / mask of LITTLE_ENDIAN.
25076 * config/microblaze/t-microblaze: Expand multilib options to
25077 include mlittle-endian (le) and update exceptions patterns.
25078
600a5961
JJ
250792013-02-08 Jakub Jelinek <jakub@redhat.com>
25080
5df81313
JJ
25081 PR rtl-optimization/56195
25082 * lra-constraints.c (get_reload_reg): Don't reuse regs
25083 if they have smaller mode than requested, if they have
25084 wider mode than requested, try to return a SUBREG.
25085
600a5961
JJ
25086 PR tree-optimization/56250
25087 * fold-const.c (extract_muldiv_1) <case NEGATE_EXPR>: Don't optimize
25088 if type is unsigned and code isn't MULT_EXPR.
25089
ff544649
GJL
250902013-02-08 Georg-Johann Lay <avr@gjlay.de>
25091
25092 PR tree-optimization/56064
25093 * fixed-value.c (fixed_from_double_int): Sign/zero extend payload
25094 bits according to mode.
25095 * fixed-value.h (fixed_from_double_int)
25096 (const_fixed_from_double_int): Adjust comments.
25097
e45cde98
RB
250982013-02-08 Richard Biener <rguenther@suse.de>
25099
25100 PR lto/56231
25101 * lto-streamer.h (struct data_in): Remove current_file, current_line
25102 and current_col members.
25103 * lto-streamer-out.c (lto_output_location): Stream changed bits
25104 en-block for efficiency.
25105 * lto-streamer-in.c (clear_line_info): Remove.
25106 (lto_input_location): Cache current file, line and column
25107 globally via local statics. Read changed bits en-block.
25108 (input_function): Do not call clear_line_info.
25109 (lto_read_body): Likewise.
25110 (lto_input_toplevel_asms): Likewise.
25111
c1ca73d8
MM
251122013-02-08 Michael Matz <matz@suse.de>
25113
25114 PR tree-optimization/52448
25115 * tree-ssa-phiopt.c (struct name_to_bb): Add phase member.
25116 (nt_call_phase): New static.
25117 (add_or_mark_expr): Only mark accesses with newer phase than any
25118 call seen.
25119 (nonfreeing_call_p): New.
25120 (nt_init_block): Update nt_call_phase, mark blocks as visited.
25121 (nt_fini_block): Keep blocks marked as visited.
25122 (get_non_trapping): Initialize nt_call_phase, and reset aux pointer.
25123
57548aa2
RB
251242013-02-08 Richard Biener <rguenther@suse.de>
25125
25126 * ira.c (ira): Free broken dominator information.
25127
8e10366f
UB
251282013-02-08 Uros Bizjak <ubizjak@gmail.com>
25129
25130 * config/i386/i386.c (ix86_spill_class): Use INTEGER_CLASS_P macro.
25131
f64fb0fa
MP
251322013-02-08 Marek Polacek <polacek@redhat.com>
25133
8e10366f 25134 * cfgloop.c (verify_loop_structure): Add more checking of headers.
f64fb0fa 25135
0375167b
RB
251362013-02-08 Richard Biener <rguenther@suse.de>
25137
25138 PR middle-end/56181
25139 * cfgloop.h (flow_loops_find): Adjust.
25140 (bb_loop_header_p): Declare.
25141 * cfgloop.c (bb_loop_header_p): New function split out from ...
25142 (flow_loops_find): ... here. Adjust function signature,
25143 support incremental loop structure update.
25144 (verify_loop_structure): Cleanup. Verify a loop is a loop.
25145 * cfgloopmanip.c (fix_loop_structure): Move ...
25146 * loop-init.c (fix_loop_structure): ... here.
25147 (apply_loop_flags): Split out from ...
25148 (loop_optimizer_init): ... here.
25149 (fix_loop_structure): Use apply_loop_flags. Use flow_loops_find
25150 in incremental mode, only remove dead loops here.
25151
85d768f3
GJL
251522013-02-08 Georg-Johann Lay <avr@gjlay.de>
25153
25154 PR target/54222
25155 * config/avr/avr.md (unspec) <UNSPEC_ROUND>: Add.
25156 * config/avr/avr-fixed.md (ALL4QA, ALL124QA): New mode iterators.
25157 (round<mode>3, round<mode>3_const): New expanders for fixed-mode.
25158 (*round<mode>3.libgcc): New insns for fixed-modes.
25159 * config/avr/builtins.def (ABSxx): Use a non-NULL LIBNAME.
25160 (ROUNDxx, COUNTLSxx, BITSxx, xxBITS): New DEF_BUILTINs.
25161 (ROUNDFX, COUNTLSFX, ABSFX): New DEF_BUILTINs.
25162 * config/avr/stdfix.h (absFX, bitsFX, FXbits): Remove inline
25163 implementations. Define to __builtin_avr_absFX,
25164 __builtin_avr_bitsFX, __builtin_avr_FXbits, respectively.
25165 (roundFX, countlsFX): Define to __builtin_avr_roundFX,
25166 __builtin_avr_countlsFX, respectively.
25167 * config/avr/avr-c.c (target.h): Include it.
25168 (enum avr_builtin_id): New enum.
25169 (avr_resolve_overloaded_builtin): New static function.
25170 (avr_register_target_pragmas): Use it to set
25171 targetm.resolve_overloaded_builtin.
25172 * config/avr/avr.c (avr_init_builtins): Supply myriads of local
25173 tree nodes used by DEF_BUILTIN.
25174 (avr_expand_builtin) <AVR_BUILTIN_ROUNDxx>: Sanity-check them.
25175 (avr_fold_builtin) <AVR_BUILTIN_BITSxx>: Fold to VIEW_COVERT_EXPR.
25176 <AVR_BUILTIN_xxBITS>: Same.
25177
661bc682
RB
251782013-02-08 Richard Biener <rguenther@suse.de>
25179
25180 * cfgloop.c (verify_loop_structure): Properly handle
25181 a loop exiting to another loop header.
25182 * ira-int.h (ira_loops): Remove.
25183 * ira.c (ira_loops): Remove.
25184 (ira): Use loop_optimizer_init and loop_optimizer_finalize.
25185 (do_reload): Use loop_optimizer_finalize.
25186 * ira-build.c (create_loop_tree_nodes): Use get_loops and
25187 number_of_loops to access the loop tree.
25188 (more_one_region_p): Likewise.
25189 (finish_loop_tree_nodes): Likewise.
25190 (rebuild_regno_allocno_maps): Likewise.
25191 (mark_loops_for_removal): Likewise.
25192 (mark_all_loops_for_removal): Likewise.
25193 (remove_unnecessary_regions): Likewise.
25194 (ira_build): Likewise.
25195 * ira-emit.c (setup_entered_from_non_parent_p): Likewise.
25196
0d5049b2
RB
251972013-02-08 Richard Biener <rguenther@suse.de>
25198
25199 * Makefile.in (tree-tailcall.o): Add $(CFGLOOP_H) dependency.
25200 * ipa-pure-const.c (analyze_function): Avoid calling
25201 mark_irreducible_loops twice.
8e10366f 25202 * tree-tailcall.c (tree_optimize_tail_calls_1): Mark loops for fixup.
0d5049b2 25203
499f32e8
DM
252042013-02-07 David S. Miller <davem@davemloft.net>
25205
25206 * dwarf2out.c (based_loc_descr): Perform leaf register remapping
25207 on 'reg'.
25208 * var-tracking.c (vt_add_function_parameter): Test the presence of
25209 HAVE_window_save properly and do not remap argument registers when
25210 we have a leaf function.
25211
6edc3e32
UB
252122013-02-07 Uros Bizjak <ubizjak@gmail.com>
25213
25214 PR bootstrap/56227
25215 * ggc-page.c (ggc_print_statistics): Use HOST_LONG_LONG_FORMAT
25216 instead of "ll".
25217 * config/i386/i386.c (ix86_print_operand): Ditto.
25218
5306401f
VM
252192013-02-07 Vladimir Makarov <vmakarov@redhat.com>
25220
6edc3e32 25221 * lra-constraints.c (process_alt_operands): Fix recently added comment.
5306401f 25222
027ece11
VM
252232013-02-07 Vladimir Makarov <vmakarov@redhat.com>
25224
25225 PR rtl-optimization/56225
25226 * lra-constraints.c (process_alt_operands): Check that reload hard
25227 reg can hold value for strict_low_part.
25228
f980dfdb
JJ
252292013-02-07 Jakub Jelinek <jakub@redhat.com>
25230
25231 PR debug/56154
25232 * dwarf2out.c (dwarf2_debug_hooks): Set end_function hook to
25233 dwarf2out_end_function.
25234 (in_first_function_p, maybe_at_text_label_p,
25235 first_loclabel_num_not_at_text_label): New variables.
25236 (dwarf2out_var_location): In the first function find out
6edc3e32 25237 lowest loclabel_num N where .LVLN is known not to be equal to .Ltext0.
f980dfdb
JJ
25238 (find_empty_loc_ranges_at_text_label, dwarf2out_end_function): New
25239 functions.
25240
73dd3123
EB
252412013-02-07 Eric Botcazou <ebotcazou@adacore.com>
25242
25243 PR rtl-optimization/56178
25244 * cse.c (cse_insn): Do not create a REG_EQUAL note if the source is a
25245 SUBREG of a register. Tidy up related block of code.
25246 * fwprop.c (forward_propagate_and_simplify): Do not create a REG_EQUAL
25247 note if the source is a register or a SUBREG of a register.
25248
e3936f47
JJ
252492013-02-07 Jakub Jelinek <jakub@redhat.com>
25250
25251 PR target/56228
25252 * config/rs6000/rs6000.md (ptrm): New mode attr.
25253 (call_indirect_aix<ptrsize>, call_indirect_aix<ptrsize>_nor11,
25254 call_value_indirect_aix<pttrsize>,
25255 call_value_indirect_aix<pttrsize>_nor11): Use <ptrm> instead of
25256 m in constraints.
25257
d96d674b
MH
252582013-02-07 Michael Haubenwallner <michael.haubenwallner@salomon.at>
25259
25260 * collect2.c (main): Set aix64_flag for -G and -bsvr4 too, disable
25261 if -bnortl. Convert to strcmp and strncmp.
25262
921f2dee
AM
252632013-02-07 Alan Modra <amodra@gmail.com>
25264
25265 PR target/54009
25266 * config/rs6000/rs6000.c (mem_operand_gpr): Check that LO_SUM
25267 addresses won't wrap when offsetting.
25268 (rs6000_secondary_reload): Provide secondary reloads needed for
25269 wrapping LO_SUM addresses.
25270
d09c7dba
TS
252712013-02-06 Thomas Schwinge <thomas@codesourcery.com>
25272
25273 * config/gnu.h (GNU_USER_TARGET_OS_CPP_BUILTINS): Never define
25274 MACH, just __MACH__.
25275
a44bbd48
RB
252762013-02-06 Richard Biener <rguenther@suse.de>
25277
25278 * tracer.c (tracer): Mark loops with LOOPS_NEED_FIXUP
25279 instead of calling fix_loop_structure.
25280
6d840d99
JJ
252812013-02-06 Jakub Jelinek <jakub@redhat.com>
25282
25283 PR middle-end/56217
25284 * omp-low.c (use_pointer_for_field): Return false if
25285 lower_send_shared_vars doesn't generate any copy-out code.
25286
0f33baa9
TV
252872013-02-06 Tom de Vries <tom@codesourcery.com>
25288
25289 PR rtl-optimization/56131
25290 * cfgrtl.c (delete_insn): Use NOTE_BASIC_BLOCK instead of BLOCK_FOR_INSN
25291 to get the bb of a NOTE_INSN_BASIC_BLOCK. Handle the case that the bb
25292 of the label is NULL. Add comment.
25293
11f1e3ab
JJ
252942013-02-05 Jakub Jelinek <jakub@redhat.com>
25295
31502f9f
JJ
25296 * tree.h (struct tree_decl_with_vis): Remove thread_local field.
25297
a74db9bd
JJ
25298 PR sanitizer/55374
25299 * config/gnu-user.h (LIBTSAN_EARLY_SPEC): Define.
25300 (STATIC_LIBTSAN_LIBS): Likewise.
25301 * gcc.c (ADD_STATIC_LIBTSAN_LIBS, LIBTSAN_EARLY_SPEC): Define.
25302 (LIBTSAN_SPEC): Add ADD_STATIC_LIBTSAN_LIBS, if LIBTSAN_EARLY_SPEC
25303 is defined, don't add anything else beyond that.
25304 (SANITIZER_EARLY_SPEC, SANITIZER_SPEC): Define.
25305 (LINK_COMMAND_SPEC): Use them.
25306
11f1e3ab
JJ
25307 PR tree-optimization/56205
25308 * tree-stdarg.c (check_all_va_list_escapes): Return true if
25309 there are any PHI nodes that set non-va_list_escape_vars SSA_NAME
25310 and some va_list_escape_vars SSA_NAME appears in some PHI argument.
25311
14ac6aa2
RB
253122013-02-05 Richard Biener <rguenther@suse.de>
25313
25314 PR tree-optimization/53342
25315 PR tree-optimization/53185
25316 * tree-vectorizer.h (vect_check_strided_load): Remove.
25317 * tree-vect-data-refs.c (vect_enhance_data_refs_alignment): Do
25318 not disallow peeling for vectorized strided loads.
25319 (vect_check_strided_load): Make static and simplify.
25320 (vect_analyze_data_refs): Adjust.
25321 * tree-vect-stmts.c (vectorizable_load): Handle peeled loops
25322 correctly when vectorizing strided loads.
25323
6f22445a
RB
253242013-02-05 Richard Biener <rguenther@suse.de>
25325
25326 * doc/install.texi: Refer to ISL, not PPL.
25327
39f9719e
JH
253282013-02-05 Jan Hubicka <jh@suse.cz>
25329
ec4224ac
JH
25330 PR tree-optimization/55789
25331 * params.def (PARAM_EARLY_INLINER_MAX_ITERATIONS): Drop to 1.
25332
253332013-02-05 Jan Hubicka <jh@suse.cz>
25334
25335 PR tree-optimization/55789
39f9719e
JH
25336 * cgraphclones.c (cgraph_remove_node_and_inline_clones): Remove
25337 the dead call anyway.
25338
956d3b33
EB
253392013-02-05 Eric Botcazou <ebotcazou@adacore.com>
25340
25341 PR sanitizer/55374
25342 * config/gnu-user.h (LIBASAN_EARLY_SPEC): Add missing guard.
25343
253442013-02-04 Alexander Potapenko <glider@google.com>
25345 Jack Howarth <howarth@bromo.med.uc.edu>
ae526fe5
AP
25346 Jakub Jelinek <jakub@redhat.com>
25347
25348 PR sanitizer/55617
25349 * config/darwin.c (sort_ctor_records): Stabilized qsort
25350 on constructor priority by using original position.
25351 (finalize_ctors): New routine to sort constructors by
25352 priority before use in assemble_integer.
25353 (machopic_asm_out_constructor): Use finalize_ctors if needed.
25354
7ac3af38
JJ
253552013-02-04 Jakub Jelinek <jakub@redhat.com>
25356
25357 PR libstdc++/54314
25358 * config/i386/winnt.c (i386_pe_assemble_visibility): Don't warn
25359 about visibility on artificial decls.
25360 * config/sol2.c (solaris_assemble_visibility): Likewise.
25361
152689dc
KT
253622013-02-04 Kai Tietz <ktietz@redhat.com>
25363
25364 PR target/56186
25365 * config/i386/i386.c (function_value_ms_64): Add additional valtype
25366 argument and improve checking of return-argument types for 16-byte
25367 modes.
25368 (ix86_function_value_1): Add additional valtype argument on call
25369 of function_value_64.
25370 (return_in_memory_ms_64): Sync 16-byte sized mode handling with
25371 handling infunction_value_64 function.
25372
2b5987b5
MGD
253732013-02-04 Matthew Gretton-Dann <matthew.gretton-dann@linaro.org>
25374
7ac3af38 25375 * reload.c (subst_reloads): Fix DEBUG_RELOAD build issue.
2b5987b5 25376
14597080
RB
253772013-02-04 Richard Biener <rguenther@suse.de>
25378
25379 PR tree-optimization/56188
25380 * tree-ssa-structalias.c (label_visit): Consider case with
25381 initially non-empty points-to set.
25382 (perform_var_substitution): Dump node mapping and clean up.
25383
ed73881e
RG
253842013-02-04 Richard Guenther <rguenther@suse.de>
25385
25386 PR lto/56168
25387 * lto-symtab.c (lto_symtab_merge_decls_1): Make non-builtin
25388 node prevail as last resort.
25389 (lto_symtab_merge_decls): Remove guard on LTRANS here.
6edc3e32 25390 (lto_symtab_prevailing_decl): Builtins are their own prevailing decl.
ed73881e 25391
139a0707
RB
253922013-02-04 Richard Biener <rguenther@suse.de>
25393
25394 PR tree-optimization/56113
25395 * tree-ssa-structalias.c (equiv_class_lookup, equiv_class_add):
25396 Merge into ...
25397 (equiv_class_lookup_or_add): ... this.
25398 (label_visit): Adjust and fix error in previous patch.
25399 (perform_var_substitution): Adjust.
25400
50fe8924
OE
254012013-02-03 Oleg Endo <olegendo@gcc.gnu.org>
25402
25403 * config/sh/divtab.c: Fix formatting and comments throughout the file.
25404 * config/sh/sh4-300.md: Likewise.
25405 * config/sh/sh4a.md: Likewise.
25406 * config/sh/constraints.md: Likewise.
25407 * config/sh/sh.md: Likewise.
25408 * config/sh/netbsd-elf.h: Likewise.
25409 * config/sh/predicates.md: Likewise.
25410 * config/sh/sh-protos.h: Likewise.
25411 * config/sh/ushmedia.h: Likewise.
25412 * config/sh/linux.h: Likewise.
25413 * config/sh/sh.c: Likewise.
25414 * config/sh/superh.h: Likewise.
25415 * config/sh/elf.h: Likewise.
25416 * config/sh/sh4.md: Likewise.
25417 * config/sh/sh.h: Likewise.
25418
1a04ac2b
JDA
254192013-02-03 John David Anglin <dave.anglin@nrc-cnrc.gc.ca>
25420
25421 * config/pa/constraints.md: Adjust unused letters. Change "T"
25422 constraint to match_test floating_point_store_memory_operand().
25423 * config/pa/predicates.md (reg_plus_base_memory_operand): New.
25424 (base14_operand): New.
25425 (floating_point_store_memory_operand): New.
25426 (integer_store_memory_operand): Revise to use base14_operand and
25427 reg_plus_base_memory_operand.
25428 (move_dest_operand): Allow symbolic_memory_operands.
25429 (symbolic_memory_operand): Check for LO_SOM.
25430 (symbolic_operand): Change default case to break.
25431 * config/pa/pa.md: Remove unamed DFmode and SFmode patterns to force
25432 CONST_DOUBLE values to be reloaded by putting them into memory when
25433 the destination is a floating point register.
25434 (movdf): Remove code to handle CONST_DOUBLE.
25435 (movsf): Likewise.
25436 (reload_indf_r1): New.
25437 (reload_insf_r1): New.
25438 Consistently use "Q" and "T" constraints with integer and floating
25439 point move instructions, respectively.
25440 (movdi): Remove FAIL.
25441 Change predicate for source operand unamed DImode move from
25442 general_operand to move_src_operand.
25443 (umulsidi3): Change predicate for destination operand to
25444 register_operand.
25445 Likewise for similar unamed patterns.
25446 * config/pa/pa-protos.h (pa_legitimize_reload_address): Declare.
25447 * config/pa/pa.c (pa_symbolic_expression_p): Remove extra parenthesis.
25448 (hppa_legitimize_address): Simplify mask calculation.
25449 (pa_emit_move_sequence): Revised handling of secondary reloads from
25450 REG+D addresses for floating point loads and stores. Directly handle
25451 loading CONST0_RTX (mode) to a floating point register.
25452 (pa_secondary_reload): Handle reloading DF and SFmode constant values
25453 to floating point registers. Don't restrict secondary reloads to
25454 floating point registers to integer modes. Revise some comments and
25455 cleanup some code.
25456 (TARGET_LEGITIMATE_ADDRESS_P): Define.
25457 (pa_legitimate_address_p): New.
25458 (pa_legitimize_reload_address): New.
25459 * config/pa/pa.h (STRICT_REG_OK_FOR_INDEX_P): New.
25460 (STRICT_REG_OK_FOR_BASE_P): New.
25461 (GO_IF_LEGITIMATE_ADDRESS): Delete. Update some related comments.
25462 (LEGITIMIZE_RELOAD_ADDRESS): Revise to use pa_legitimize_reload_address.
25463
3bf65591
DE
254642013-02-03 David Edelsohn <dje.gcc@gmail.com>
25465 Andrew Dixie <andrewd@gentrack.com>
25466
25467 * collect2.c (GCC_CHECK_HDR): Do not scan objects with F_LOADONLY
25468 flag set.
25469
c0a8a3e6
RS
254702013-02-03 Richard Sandiford <rdsandiford@googlemail.com>
25471
25472 * expmed.c (extract_bit_field_1): Pass the full width of the
25473 structure to get_best_reg_extraction_insn.
25474
99113dff
DE
254752013-02-01 David Edelsohn <dje.gcc@gmail.com>
25476
25477 PR target/54601
25478 * configure.ac (use_cxa_atexit): Add AIX.
25479 * configure: Regenerate.
25480
25481 * config/rs6000/aix61.h (STARTFILE_SPEC): Add crtcxa.o.
25482
ddd84654
JJ
254832013-02-01 Jakub Jelinek <jakub@redhat.com>
25484
25485 PR debug/54793
25486 * final.c (need_profile_function): New variable.
25487 (final_start_function): Drop ATTRIBUTE_UNUSED from first argument.
25488 If first of NOTE_INSN_BASIC_BLOCK or NOTE_INSN_FUNCTION_BEG
25489 is only preceeded by NOTE_INSN_VAR_LOCATION or NOTE_INSN_DELETED
25490 notes, targetm.asm_out.function_prologue doesn't emit anything,
25491 HAVE_prologue and profiler should be emitted before prologue,
25492 set need_profile_function instead of emitting it.
25493 (final_scan_insn): If need_profile_function, emit
25494 profile_function on the first NOTE_INSN_BASIC_BLOCK or
25495 NOTE_INSN_FUNCTION_BEG note.
25496
254972013-02-01 Richard Henderson <rth@redhat.com>
749af8ee
RH
25498
25499 * config/rs6000/rs6000.md (smulditi3): New.
25500 (umulditi3): New.
25501
ff2a9d88
RH
25502 * config/alpha/alpha.md (umulditi3): New.
25503
14d52b90
DE
255042013-02-01 David Edelsohn <dje.gcc@gmail.com>
25505
25506 * config/rs6000/xcoff.h (ASM_OUTPUT_ALIGNED_COMMON): Use floor_log2.
25507 (ASM_OUTPUT_ALIGNED_LOCAL): New.
25508
8c7ca45c
RB
255092013-02-01 Richard Biener <rguenther@suse.de>
25510
25511 PR tree-optimization/56113
25512 * tree-ssa-structalias.c (label_visit): Reduce work for
25513 single-predecessor nodes.
25514
9f419393
EB
255152013-02-01 Eric Botcazou <ebotcazou@adacore.com>
25516
25517 * fold-const.c (make_range_step) <TRUTH_NOT_EXPR>: Bail out if the
25518 range isn't testing for zero.
25519
c5f4be84
SB
255202013-01-31 Steven Bosscher <steven@gcc.gnu.org>
25521
25522 PR middle-end/56113
25523 * fwprop.c (fwprop_init): Set up loops without CFG modifications.
25524
dbdbd982
NC
255252013-01-31 Hiroyuki Ono <hiroyuki.ono.jc@renesas.com>
25526 Nick Clifton <nickc@redhat.com>
25527
25528 * config/v850/constraints.md (Q): Define as a memory constraint.
25529 * config/v850/predicates.md (label_ref_operand): New predicate.
25530 (e3v5_shift_operand): New predicate.
25531 (ior_operator): New predicate.
25532 * config/v850/t-v850: Add e3v5 multilib.
25533 * config/v850/v850-protos.h (v850_adjust_insn_length): Prototype.
25534 (v850_gen_movdi): Prototype.
25535 * config/v850/v850.c: Add support for e3v5 architecture.
25536 Rename all uses of TARGET_V850E || TARGET_V850E2_ALL to
25537 TARGET_V850E_UP.
25538 (construct_save_jarl): Add e3v5 long JARL support.
25539 (v850_adjust_insn_length): New function. Adjust length of call
25540 insns when using e3v5 instructions.
25541 (v850_gen_movdi): New function: Generate instructions to move a
25542 DImode value.
25543 * config/v850/v850.h (TARGET_CPU_v850e3v5): Define.
25544 (CPP_SPEC): Define __v850e3v5__ as appropriate.
25545 (TARGET_USE_FPU): Enable for e3v5.
25546 (CONST_OK_FOR_W): New macro.
25547 (ADJUST_INSN_LENGTH): Define.
25548 * config/v850/v850.md (UNSPEC_LOOP): Define.
25549 (attr cpu): Add v850e3v5.
25550 Rename all uses of TARGET_V850E2 to TARGET_V850E2V3_UP.
25551 (movdi): New pattern.
25552 (movdi_internal): New pattern.
25553 (cbranchsf4): Conditionalize on TARGET_USE_FPU.
25554 (cbranchdf4): Conditionalize on TARGET_USE_FPU.
25555 (cstoresf4): Likewise.
25556 (cstoredf4): Likewise.
25557 (insv): New pattern.
25558 (rotlso3_a): New pattern.
25559 (rotlsi3_b): New pattern
25560 (rotlsi3_v850e3v5): New pattern.
25561 (doloop_begin): New pattern.
25562 (fix_loop_counter): New pattern.
25563 (doloop_end): New pattern.
25564 (branch_normal): Add e3v5 long branch support.
25565 (branch_invert): Likewise.
25566 (branch_z_normal): Likewise.
25567 (branch_z_invert): Likewise.
25568 (branch_nz_normal): Likewise.
25569 (branch_nz_invert): Likewise.
25570 (call_internal_short): Add e3v5 register-indirect JARL support.
25571 (call_internal_long): Likewise.
25572 (call_value_internal_short): Likewise.
25573 (call_value_internal_long): Likewise.
25574 * config/v850/v850.opt (mv850e3v5, mv850e2v4): New options.
25575 (mloop): New option.
25576 * config.gcc: Add support for configuring v840e3v5 target.
6edc3e32 25577 * doc/invoke.texi: Document new v850 specific command line options.
dbdbd982 25578
73861a41
PK
255792013-01-31 Paul Koning <ni1d@arrl.net>
25580
25581 PR debug/55059
25582 PR debug/54508
25583 * dwarf2out.c (prune_unused_types_mark): Mark all of parent's
25584 children if parent is a class.
25585 (prune_unused_types_prune): Don't add DW_AT_declaration.
25586
e44978dc
RB
255872013-01-31 Richard Biener <rguenther@suse.de>
25588
25589 PR tree-optimization/56157
25590 * tree-vect-slp.c (vect_get_slp_defs): More thoroughly try to
25591 match up operand with SLP child.
25592
90b10dec
JM
255932013-01-31 Jason Merrill <jason@redhat.com>
25594
04d2dadd 25595 PR debug/54410
90b10dec
JM
25596 * dwarf2out.c (gen_struct_or_union_type_die): Always schedule template
25597 parameters the first time.
25598 (gen_scheduled_generic_parms_dies): Check completeness here.
25599
6e616110
RB
256002013-01-31 Richard Biener <rguenther@suse.de>
25601
25602 PR middle-end/53073
25603 * common.opt (faggressive-loop-optimizations): New flag,
25604 enabled by default.
25605 * doc/invoke.texi (faggressive-loop-optimizations): Document.
25606 * tree-ssa-loop-niter.c (estimate_numbers_of_iterations_loop): Guard
25607 infer_loop_bounds_from_undefined by it.
25608
636f59cf
RB
256092013-01-31 Richard Biener <rguenther@suse.de>
25610
25611 PR tree-optimization/56150
25612 * tree-ssa-loop-manip.c (find_uses_to_rename_stmt): Do not
25613 visit virtual operands.
25614 (find_uses_to_rename_bb): Likewise.
25615
256162013-01-31 Richard Biener <rguenther@suse.de>
fcfa87ac
RB
25617
25618 PR tree-optimization/56150
25619 * tree-ssa-tail-merge.c (gimple_equal_p): Properly handle
25620 mixed store non-store stmts.
25621
32887460
JJ
256222013-01-30 Jakub Jelinek <jakub@redhat.com>
25623
e60e09a0
JJ
25624 PR sanitizer/55374
25625 * gcc.c (LIBASAN_SPEC): Define just to ADD_STATIC_LIBASAN_LIBS if
25626 LIBASAN_EARLY_SPEC is defined.
25627 (LIBASAN_EARLY_SPEC): Define to empty string if not already defined.
25628 (LINK_COMMAND_SPEC): Add LIBASAN_EARLY_SPEC for -fsanitize=address,
25629 before %o.
25630 * config/gnu-user.h (LIBASAN_EARLY_SPEC): Define.
25631
32887460
JJ
25632 PR c++/55742
25633 * config/i386/i386.c (ix86_valid_target_attribute_inner_p): Diagnose
25634 invalid args instead of ICEing on it.
25635 (ix86_valid_target_attribute_tree): Return error_mark_node if
25636 ix86_valid_target_attribute_inner_p failed.
25637 (ix86_valid_target_attribute_p): Return false only if
25638 ix86_valid_target_attribute_tree returned error_mark_node. Allow
25639 target("default") attribute.
25640 (sorted_attr_string): Change argument from const char * to tree,
25641 merge in all target attribute arguments rather than just one.
25642 Formatting fix. Use XNEWVEC instead of xmalloc and XDELETEVEC
25643 instead of free. Avoid using strcat.
25644 (ix86_mangle_function_version_assembler_name): Mangle
25645 target("default") as if no target attribute is present. Adjust
25646 sorted_attr_string caller. Avoid leaking memory. Use XNEWVEC
25647 instead of xmalloc and XDELETEVEC instead of free.
25648 (ix86_function_versions): Don't return true if one of the decls
25649 doesn't have target attribute. If they don't and one of the decls
25650 is DECL_FUNCTION_VERSIONED, report an error. Adjust
25651 sorted_attr_string caller. Use XDELETEVEC instead of free.
25652 (ix86_supports_function_versions): Remove.
25653 (make_name): Fix up formatting.
25654 (make_dispatcher_decl): Remove resolver_name and its initialization.
25655 Avoid leaking memory.
25656 (is_function_default_version): Return true if there is
25657 target("default") attribute rather than no target attribute at all.
25658 (make_resolver_func): Avoid leaking memory.
25659 (ix86_generate_version_dispatcher_body): Likewise.
25660 (TARGET_OPTION_SUPPORTS_FUNCTION_VERSIONS): Remove.
25661 * target.def (supports_function_versions): Remove.
25662 * doc/tm.texi.in (SUPPORTS_FUNCTION_VERSIONS): Remove.
25663 * doc/tm.texi: Regenerated.
25664
73cca0cc
VM
256652013-01-30 Vladimir Makarov <vmakarov@redhat.com>
25666
25667 PR rtl-optimization/56144
25668 * lra-constraints.c (get_reload_reg): Don't reuse reload pseudo
25669 for values with side effects.
25670
11452e7b
RB
256712013-01-30 Richard Biener <rguenther@suse.de>
25672
25673 * sparseset.h (sparseset_bit_p): Use gcc_checking_assert.
25674 (sparseset_pop): Likewise.
25675 * cfganal.c (compute_idf): Likewise. Increase work-stack size
25676 to be able to use quick_push in the worker loop.
25677
01cb1ef5
MP
256782013-01-30 Marek Polacek <polacek@redhat.com>
25679
25680 * cfgcleanup.c (cleanup_cfg): Don't mark affected BBs.
25681
e892936e
RB
256822013-01-30 Richard Biener <rguenther@suse.de>
25683
25684 PR lto/56147
6edc3e32 25685 * lto-symtab.c (lto_symtab_merge_decls_1): Guard DECL_BUILT_IN check.
e892936e 25686
cc06c01d
GJL
256872013-01-30 Georg-Johann Lay <avr@gjlay.de>
25688
25689 PR tree-optimization/56064
25690 * fixed-value.c (fixed_from_double_int): New function.
25691 * fixed-value.h (fixed_from_double_int): New prototype.
25692 (const_fixed_from_double_int): New static inline function.
25693 * fold-const.c (native_interpret_fixed): New static function.
25694 (native_interpret_expr) <FIXED_POINT_TYPE>: Use it.
25695 (can_native_interpret_type_p) <FIXED_POINT_TYPE>: Return true.
25696 (native_encode_fixed): New static function.
25697 (native_encode_expr) <FIXED_CST>: Use it.
25698 (native_interpret_int): Move double_int worker code to...
25699 * double-int.c (double_int::from_buffer): ...this new static method.
25700 * double-int.h (double_int::from_buffer): Prototype it.
25701
d394a308
RB
257022013-01-30 Richard Biener <rguenther@suse.de>
25703
25704 * tree-ssa-structalias.c (final_solutions, final_solutions_obstack):
25705 New pointer-map and obstack.
25706 (init_alias_vars): Allocate pointer-map and obstack.
25707 (delete_points_to_sets): Free them.
25708 (find_what_var_points_to): Cache result.
25709 (find_what_p_points_to): Adjust for changed interface of
25710 find_what_var_points_to.
25711 (compute_points_to_sets): Likewise.
25712 (ipa_pta_execute): Likewise.
25713
20804d96
RO
257142013-01-30 Rainer Orth <ro@CeBiTec.Uni-Bielefeld.DE>
25715
25716 * configure.ac (HAVE_AS_SPARC_NOBITS): New test.
25717 * configure: Regenerate.
25718 * config.in: Regenerate.
25719 * config/sparc/sparc.c (sparc_solaris_elf_asm_named_section): Emit
25720 #nobits/#progbits if supported.
25721
a7ad88a2
OE
257222013-01-29 Oleg Endo <olegendo@gcc.gnu.org>
25723
25724 PR target/56121
25725 * config/sh/sh.md (bclr_m2a, bset_m2a, bst_m2a, bld_m2a, bldsign_m2a,
25726 bld_reg, *bld_regqi, band_m2a, bandreg_m2a, bor_m2a, borreg_m2a,
25727 bxor_m2a, bxorreg_m2a): Add satisfies_constraint_K03 condition.
25728
7af79f92
GY
257292013-01-29 Greta Yorsh <Greta.Yorsh@arm.com>
25730
91bfca59
OE
25731 * config/arm/cortex-a7.md (cortex_a7_neon, cortex_a7_all): Remove.
25732 (cortex_a7_idiv): Use cortex_a7_both instead of cortex_a7_all.
7af79f92 25733
753bcf7b
GY
257342013-01-29 Greta Yorsh <Greta.Yorsh@arm.com>
25735
91bfca59
OE
25736 * config/arm/arm.c (cortexa7_younger): Return true for TYPE_CALL.
25737 * config/arm/cortex-a7.md (cortex_a7_call): Update required units.
753bcf7b 25738
8cbc2ea8
GY
257392013-01-29 Greta Yorsh <Greta.Yorsh@arm.com>
25740
91bfca59
OE
25741 * config/arm/arm-protos.h (arm_mac_accumulator_is_result): New
25742 declaration.
25743 * config/arm/arm.c (arm_mac_accumulator_is_result): New function.
25744 * config/arm/cortex-a7.md: New bypasses using
25745 arm_mac_accumulator_is_result.
8cbc2ea8 25746
697a3325
GY
257472013-01-29 Greta Yorsh <Greta.Yorsh@arm.com>
25748
91bfca59 25749 * config/arm/cortex-a7.md (cortex_a7_neon_mul): New reservation.
697a3325 25750 (cortex_a7_neon_mla): Likewise.
91bfca59
OE
25751 (cortex_a7_fpfmad): New reservation.
25752 (cortex_a7_fpmacs): Use ffmas and update required units.
25753 (cortex_a7_fpmuld): Update required units and latency.
25754 (cortex_a7_fpmacd): Likewise.
25755 (cortex_a7_fdivs, cortex_a7_fdivd): Likewise.
25756 (cortex_a7_neon). Likewise.
25757 (bypass) Update participating units.
697a3325 25758
29637783
GY
257592013-01-29 Greta Yorsh <Greta.Yorsh@arm.com>
25760
91bfca59
OE
25761 * config/arm/arm.md (type): Add ffmas and ffmad to "type" attribute.
25762 * config/arm/vfp.md (fma,fmsub,fnmsub,fnmadd): Change type
25763 from fmac to ffma.
25764 * config/arm/vfp11.md (vfp_farith): Use ffmas.
25765 (vfp_fmul): Use ffmad.
25766 * config/arm/cortex-r4f.md (cortex_r4_fmacs): Use ffmas.
25767 (cortex_r4_fmacd): Use ffmad.
25768 * config/arm/cortex-m4-fpu.md (cortex_m4_fmacs): Use ffmas.
25769 * config/arm/cortex-a9.md (cortex_a9_fmacs): Use ffmas.
25770 (cortex_a9_fmacd): Use ffmad.
25771 * config/arm/cortex-a8-neon.md (cortex_a8_vfp_macs): Use ffmas.
25772 (cortex_a8_vfp_macd): Use ffmad.
25773 * config/arm/cortex-a5.md (cortex_a5_fpmacs): Use ffmas.
25774 (cortex_a5_fpmacd): Use ffmad.
25775 * config/arm/cortex-a15-neon.md (cortex_a15_vfp_macs) Use ffmas.
25776 (cortex_a15_vfp_macd): Use ffmad.
25777 * config/arm/arm1020e.md (v10_fmul): Use ffmas and ffmad.
29637783 25778
2ee8a2d5
JM
257792013-01-29 Jason Merrill <jason@redhat.com>
25780
25781 PR libstdc++/54314
25782 * varasm.c (default_assemble_visibility): Don't warn about
25783 visibility on artificial decls.
25784
78d087bc
RB
257852013-01-29 Richard Biener <rguenther@suse.de>
25786
25787 PR tree-optimization/56113
25788 * tree-ssa-structalias.c (equiv_class_lookup): Also return
25789 the bitmap leader.
25790 (label_visit): Free duplicate bitmaps and record the leader instead.
25791 (perform_var_substitution): Adjust.
25792
83ba4d6f
RB
257932013-01-29 Richard Biener <rguenther@suse.de>
25794
25795 PR tree-optimization/55270
25796 * tree-ssa-dom.c (eliminate_degenerate_phis): If we changed
25797 the CFG, schedule loops for fixup.
25798
66dfe13f
NC
257992013-01-29 Nick Clifton <nickc@redhat.com>
25800
25801 * config/rl78/rl78.c (rl78_regno_mode_code_ok_for_base_p): Allow
25802 SP_REG.
25803
5a579c3b
LE
258042013-01-28 Leif Ekblad <leif@rdos.net>
25805
25806 * config.gcc (i[34567]86-*-rdos*, x86_64-*-rdos*): New targets.
25807 * config/i386/i386.h (TARGET_RDOS): New macro.
25808 (DEFAULT_LARGE_SECTION_THRESHOLD): New macro.
25809 * config/i386/i386.c (ix86_option_override_internal): For 64bit
25810 TARGET_RDOS, set ix86_cmodel to CM_MEDIUM_PIC and flag_pic to 1.
25811 * config/i386/i386.opt (mlarge-data-threshold): Initialize to
25812 DEFAULT_LARGE_SECTION_THRESHOLD.
25813 * config/i386/i386.md (R14_REG, R15_REG): New constants.
25814 * config/i386/rdos.h: New file.
25815 * config/i386/rdos64.h: New file.
25816
51e44392
BS
258172013-01-28 Bernd Schmidt <bernds@codesourcery.com>
25818
25819 PR other/54814
25820 * reload.c (find_valid_class_1): Use in_hard_reg_set_p instead of
25821 TEST_HARD_REG_BIT.
25822
db1fb332
JJ
258232013-01-28 Jakub Jelinek <jakub@redhat.com>
25824
25825 PR rtl-optimization/56117
25826 * sched-deps.c (sched_analyze_2) <case PREFETCH>: For use_cselib
25827 call cselib_lookup_from_insn on the MEM before calling
25828 add_insn_mem_dependence.
25829
16917761
RB
258302013-01-28 Richard Biener <rguenther@suse.de>
25831
25832 * tree-inline.c (remap_gimple_stmt): Do not assing a BLOCK
25833 to a stmt that didn't have one.
25834 (copy_phis_for_bb): Likewise for PHI arguments.
25835 (copy_debug_stmt): Likewise for debug stmts.
25836
b9fc0497
RB
258372013-01-28 Richard Biener <rguenther@suse.de>
25838
25839 PR tree-optimization/56034
6edc3e32 25840 * tree-loop-distribution.c (enum partition_kind): Add PKIND_REDUCTION.
b9fc0497
RB
25841 (partition_builtin_p): Adjust.
25842 (generate_code_for_partition): Handle PKIND_REDUCTION. Assert
25843 it is the last partition.
25844 (rdg_flag_uses): Check SSA_NAME_IS_DEFAULT_DEF before looking
25845 up the vertex for the definition.
25846 (classify_partition): Classify whether a partition is a
25847 PKIND_REDUCTION, thus has uses outside of the loop.
25848 (ldist_gen): Inherit PKIND_REDUCTION when merging partitions.
25849 Merge all PKIND_REDUCTION partitions into the last partition.
25850 (tree_loop_distribution): Seed partitions from reductions as well.
25851
aa710d25
JJ
258522013-01-28 Jakub Jelinek <jakub@redhat.com>
25853
0bfbca58
JJ
25854 PR tree-optimization/56125
25855 * tree-ssa-math-opts.c (gimple_expand_builtin_pow): Don't optimize
25856 pow(x,c) into sqrt(x) * powi(x, n/2) or
25857 1.0 / (sqrt(x) * powi(x, abs(n/2))) if c is an integer or when
25858 optimizing for size.
25859 Don't optimize pow(x,c) into powi(x, n/3) * powi(cbrt(x), n%3) or
25860 1.0 / (powi(x, abs(n)/3) * powi(cbrt(x), abs(n)%3)) if 2c is an
25861 integer.
25862
aa710d25
JJ
25863 PR tree-optimization/56094
25864 * gimplify.c (force_gimple_operand_1): Temporarily set input_location
25865 to UNKNOWN_LOCATION while gimplifying expr.
25866
77dc5297
UB
258672013-01-27 Uros Bizjak <ubizjak@gmail.com>
25868
25869 PR target/56114
25870 * config/i386/i386.md (*movabs<mode>_1): Add square brackets around
25871 operand 0 in movabs insn template for -masm=intel asm alternative.
25872 (*movabs<mode>_2): Ditto for operand 1.
25873
0bfbca58 258742013-01-26 David Holsgrove <david.holsgrove@xilinx.com>
fbf0cf90
DH
25875
25876 PR target/54663
25877 * config.gcc (microblaze*-linux*): Add tmake_file to allow building
25878 of microblaze-c.o
25879
0bfbca58 258802013-01-26 Edgar E. Iglesias <edgar.iglesias@gmail.com>
6f14eef2
EI
25881
25882 * config.gcc (microblaze*-*-*): Rename microblaze*-*-elf, update
25883 tm_file.
25884
0bfbca58 258852013-01-25 Naveen H.S <Naveen.Hurugalawadi@caviumnetworks.com>
706b2314
N
25886
25887 * config/aarch64/aarch64.c (TARGET_FIXED_CONDITION_CODE_REGS):
25888 Undef to avoid warning.
25889
478f60f9
MH
258902013-01-25 Michael Haubenwallner <michael.haubenwallner@salomon.at>
25891
25892 * configure.ac (gcc_cv_ld_static_dynamic): Define for AIX native ld.
25893 * configure: Regenerate.
25894
d7fa6ee2
JJ
258952013-01-25 Jakub Jelinek <jakub@redhat.com>
25896
25897 PR tree-optimization/56098
25898 * tree-ssa-phiopt.c (nt_init_block): Don't call add_or_mark_expr
25899 for stmts with volatile ops.
25900 (cond_store_replacement): Don't optimize if assign has volatile ops.
25901 (cond_if_else_store_replacement_1): Don't optimize if either
25902 then_assign or else_assign have volatile ops.
25903 (hoist_adjacent_loads): Don't optimize if either def1 or def2 have
25904 volatile ops.
25905
f8fe87bd
GJL
259062013-01-25 Georg-Johann Lay <avr@gjlay.de>
25907
25908 * doc/invoke.texi (AVR Built-in Macros): Document __XMEGA__.
25909
28398d0d
GJL
259102013-01-25 Georg-Johann Lay <avr@gjlay.de>
25911
25912 * doc/extend.texi (Example of asm with clobbered asm reg): Fix
25913 missing ':' in asm example.
25914
b7d7d917
TB
259152013-01-25 Tejas Belagod <tejas.belagod@arm.com>
25916
25917 * config/aarch64/aarch64-simd-builtins.def: Separate sq<r>dmulh_lane
25918 entries into lane and laneq entries.
77dc5297
UB
25919 * config/aarch64/aarch64-simd.md (aarch64_sq<r>dmulh_lane<mode>):
25920 Remove AdvSIMD scalar modes.
b7d7d917
TB
25921 (aarch64_sq<r>dmulh_laneq<mode>): New.
25922 (aarch64_sq<r>dmulh_lane<mode>): New RTL pattern for Scalar AdvSIMD
25923 modes.
25924 * config/aarch64/arm_neon.h: Fix all the vq<r>dmulh_lane* intrinsics'
25925 builtin implementations to relfect changes in RTL in aarch64-simd.md.
25926 * config/aarch64/iterators.md (VCOND): New.
25927 (VCONQ): New.
25928
556f9906
GJL
259292013-01-25 Georg-Johann Lay <avr@gjlay.de>
25930
25931 PR target/54222
25932 * config/avr/builtins.def (DEF_BUILTIN): Add LIBNAME argument.
25933 Add NULL LIBNAME argument to existing definitions.
25934 (ABSHR, ABSR, ABSLR, ABSLLR, ABSHK, ABSK, ABSLK, ABSLLK): New.
25935 * config/avr/avr-c.c (DEF_BUILTIN): Add LIBNAME argument.
25936 * config/avr/avr.c (DEF_BUILTIN): Same.
25937 (avr_init_builtins): Pass down LIBNAME to add_builtin_function.
25938 (avr_expand_builtin): Expand to a vanilla call if a libgcc
25939 implementation is available (DECL_ASSEMBLER_NAME is set).
25940 (avr_fold_absfx): New static function.
25941 (avr_fold_builtin): Use it to handle: AVR_BUILTIN_ABSHR,
25942 AVR_BUILTIN_ABSR, AVR_BUILTIN_ABSLR, AVR_BUILTIN_ABSLLR,
25943 AVR_BUILTIN_ABSHK, AVR_BUILTIN_ABSK, AVR_BUILTIN_ABSLK,
25944 AVR_BUILTIN_ABSLLK.
25945 * config/avr/stdfix.h (abshr, absr, abslr, absllr)
25946 (abshk, absk, abslk, absllk): Provide as static inline functions.
25947
1f546bbb
MP
259482013-01-25 Marek Polacek <polacek@redhat.com>
25949
25950 PR tree-optimization/56035
25951 * cfgloopmanip.c (fix_loop_structure): Remove redundant condition.
25952
7e184bd7
UB
259532012-01-24 Uros Bizjak <ubizjak@gmail.com>
25954
25955 * config/i386/i386.md (*movti_internal_rex64): Add (o,e) alternative.
25956 (*movtf_internal_rex64): Add (!o,C) alternative
25957 (*movxf_internal_rex64): Ditto.
25958 (*movdf_internal_rex64): Add (?r,C) and (?m,C) alternatives.
25959
f5ba49ac
SM
259602013-01-24 Shenghou Ma <minux.ma@gmail.com>
25961
25962 * doc/invoke.texi: fix typo.
25963 * doc/objc.texi: fix typo.
25964
a6343728
RS
259652013-01-24 Richard Sandiford <rdsandiford@googlemail.com>
25966
25967 * config/mips/mips.md (*and<mode>3_mips16): Use the "W" constraint
25968 for the first two alternatives.
25969
cd030c07
DN
259702013-01-24 Diego Novillo <dnovillo@google.com>
25971
77dc5297 25972 * Makefile.in (GGC): Remove. Replace all instances with ggc-page.o.
cd030c07
DN
25973 (ggc-zone.o): Remove.
25974 * configure.ac: Remove option --with-gc.
25975 * configure: Re-generate.
25976 * doc/install.texi: Remove documentation for --with-gc.
25977 * gengtype.c (write_enum_defn): Remove. Update all users.
25978 (write_Types_process_field): Remove generation of gt_e_* argument.
25979 (output_type_enum): Remove. Update all users.
25980 (write_enum_defn): Remove. Update all users.
25981 (enum alloc_zone): Remove. Update all users.
77dc5297 25982 (write_splay_tree_allocator_def): Remove generation of gt_e_* argument.
cd030c07
DN
25983 * ggc-common.c (ggc_splay_alloc): Remove first argument.
25984 Update all callers.
25985 (struct ptr_data): Remove field TYPE. Update all users.
77dc5297 25986 (gt_pch_note_object): Remove argument TYPE. Update all users.
cd030c07
DN
25987 * ggc-internal.h (ggc_pch_alloc_object): Remove last argument.
25988 Update all users.
aaf1e810 25989 * ggc-none.c (ggc_alloc_typed_stat): Remove.
cd030c07
DN
25990 (struct alloc_zone): Remove.
25991 (ggc_internal_alloc_zone_stat): Remove.
25992 (ggc_internal_cleared_alloc_zone_stat): Remove.
25993 * ggc-page.c (ggc_alloc_typed_stat): Remove.
77dc5297
UB
25994 (ggc_pch_count_object): Remove last argument. Update all users.
25995 (ggc_pch_alloc_object): Remove last argument. Update all users.
cd030c07
DN
25996 (struct alloc_zone): Remove.
25997 * ggc-zone.c: Remove.
77dc5297 25998 * ggc.h (gt_pch_note_object): Remove last argument. Update all users.
cd030c07
DN
25999 (struct alloc_zone): Remove.
26000 (ggc_alloc_typed_stat): Remove.
26001 (ggc_alloc_typed): Remove.
26002 (ggc_splay_alloc): Remove first argument.
26003 (rtl_zone): Remove. Update all users.
26004 (tree_zone): Remove. Update all users.
26005 (tree_id_zone): Remove. Update all users.
26006 (ggc_internal_zone_alloc_stat): Remove. Update all users.
77dc5297 26007 (ggc_internal_zone_cleared_alloc_stat): Remove. Update all users.
cd030c07 26008 (ggc_internal_zone_vec_alloc_stat): Remove. Update all users.
77dc5297 26009 * tree-ssanames.c: Remove references to zone allocator in comments.
cd030c07 26010
a861ffa4
GJL
260112013-01-24 Georg-Johann Lay <avr@gjlay.de>
26012
26013 * config/avr/avr.c (avr_out_fract): Make register numbers that
26014 might be outside of source operand signed.
26015
593c0ddd
UB
260162013-01-24 Uros Bizjak <ubizjak@gmail.com>
26017
26018 * config/i386/constraints.md (Yf): New constraint.
26019 * config/i386/i386.md (*movdf_internal_rex64): Use Yf*f instead
26020 of f constraint to conditionaly disable x87 register preferences.
26021 (*movdf_internal): Ditto.
26022 (*movsf_internal): Ditto.
26023
e86c0101
SB
260242013-01-24 Steven Bosscher <steven@gcc.gnu.org>
26025
26026 PR inline-asm/55934
26027 * lra-assigns.c (assign_by_spills): Throw away the pattern of asms
26028 that have operands with impossible constraints.
26029 Add a FIXME for a speed-up opportunity.
26030 * lra-constraints.c (process_alt_operands): Verify that a class
26031 selected from constraints on asms is valid for the operand mode.
26032 (curr_insn_transform): Remove incorrect comment.
26033
f6fee35f
DE
260342013-01-23 David Edelsohn <dje.gcc@gmail.com>
26035
26036 * config/rs6000/rs6000.c (rs6000_delegitimize_address): Check that
26037 TOC operand is a valid symbol ref in the constant pool.
26038
260392013-01-23 Edgar E. Iglesias <edgar.iglesias@gmail.com>
c7729184 26040
aaf1e810 26041 * config/microblaze/linux.h: Add TARGET_OS_CPP_BUILTINS
c7729184 26042
dc62d7d1
GJL
260432013-01-23 Georg-Johann Lay <avr@gjlay.de>
26044
26045 PR target/54222
26046 * config/avr/stdfix.h: New file.
26047 * t-avr (stdfix-gcc.h): New rule to build it.
26048 (EXTRA_HEADERS): Set it to install stdfix.h, stdfix-gcc.h.
26049
2660d12d
KS
260502013-01-23 Kostya Serebryany <kcc@google.com>
26051
77dc5297
UB
26052 * config/darwin.h: remove dependency on
26053 CoreFoundation (asan on Mac OS).
2660d12d 26054
a70418fc
JJ
260552013-01-23 Jakub Jelinek <jakub@redhat.com>
26056
26057 PR target/49069
26058 * config/arm/arm.md (cbranchdi4, cstoredi4): Use s_register_operand
26059 instead of cmpdi_operand for first comparison operand.
26060 Don't assert that comparison operands aren't both constants.
26061
47876a2a
JW
260622013-01-22 Jonathan Wakely <jwakely.gcc@gmail.com>
26063
26064 * doc/install.texi (Downloading the Source): Update references to
26065 downloading separate components.
26066
260672013-01-22 Jonathan Wakely <jwakely.gcc@gmail.com>
4ebd56a5
JW
26068
26069 * doc/extend.texi (__int128): Improve grammar.
26070
47876a2a 260712013-01-22 Uros Bizjak <ubizjak@gmail.com>
32fc5b8a
UB
26072
26073 PR target/56028
26074 * config/i386/i386.md (*movti_internal_rex64): Change (o,riF)
26075 alternative to (o,r).
26076 (*movdi_internal_rex64): Remove (!o,n) alternative.
26077 (DImode immediate->memory splitter): Remove.
26078 (DImode immediate->memory peephole2): Remove.
26079 (movtf): Enable for TARGET_64BIT || TARGET_SSE.
26080 (*movtf_internal_rex64): Rename from *movtf_internal. Change (!o,F*r)
26081 alternative to (!o,*r).
26082 (*movtf_internal_sse): New pattern.
26083 (*movxf_internal_rex64): New pattern.
26084 (*movxf_internal): Disable for TARGET_64BIT.
26085 (*movdf_internal_rex64): Remove (!o,F) alternative.
26086
3a984f10
JJ
260872013-01-22 Jakub Jelinek <jakub@redhat.com>
26088
502498d5
JJ
26089 PR middle-end/56074
26090 * dumpfile.c (dump_loc): Only print loc if LOCATION_LOCUS (loc)
26091 isn't UNKNOWN_LOCATION nor BUILTINS_LOCATION.
26092 * tree-vect-loop-manip.c (find_loop_location): Also ignore
26093 stmt locations where LOCATION_LOCUS of the stmt location is
26094 UNKNOWN_LOCATION or BUILTINS_LOCATION.
26095
3a984f10
JJ
26096 PR target/55686
26097 * config/i386/i386.md (UNSPEC_STOS): New.
26098 (strset_singleop, *strsetdi_rex_1, *strsetsi_1, *strsethi_1,
26099 *strsetqi_1): Add UNSPEC_STOS.
26100
fa817f7f
PC
261012013-01-22 Paolo Carlini <paolo.carlini@oracle.com>
26102
26103 PR c++/56067
26104 * doc/invoke.texi: Remove left over -Wsynth example.
26105
8f498c1b
JJ
261062013-01-21 Jakub Jelinek <jakub@redhat.com>
26107
26108 PR tree-optimization/56051
26109 * fold-const.c (fold_binary_loc): Don't fold
26110 X < (cast) (1 << Y) into (X >> Y) != 0 if cast is either
26111 a narrowing conversion, or widening conversion from signed
26112 to unsigned.
26113
47876a2a 261142013-01-21 Uros Bizjak <ubizjak@gmail.com>
35d59da7
UB
26115
26116 PR rtl-optimization/56023
26117 * haifa-sched.c (fix_inter_tick): Do not update ticks of instructions,
26118 dependent on debug instruction.
26119
5b9db1bc
MJ
261202013-01-21 Martin Jambor <mjambor@suse.cz>
26121
26122 PR middle-end/56022
26123 * function.c (allocate_struct_function): Call
26124 invoke_set_current_function_hook earlier.
26125
e8bb7d68
JJ
261262013-01-21 Jakub Jelinek <jakub@redhat.com>
26127
26128 * reload1.c (init_reload): Only initialize reload_obstack
26129 during the first call.
26130
616a4e32
MP
261312013-01-21 Marek Polacek <polacek@redhat.com>
26132
26133 * cfgloop.c (verify_loop_structure): Fix up grammar.
26134
4401981b
YHH
261352013-01-21 Yi-Hsiu Hsu <ahsu@marvell.com>
26136
26137 * config/arm/marvell-pj4.md (pj4_shift_conds, pj4_alu_shift,
26138 pj4_alu_shift_conds, pj4_shift): Handle simple_alu_shift.
26139
8e87740b
RR
261402013-01-21 Ramana Radhakrishnan <ramana.radhakrishnan@arm.com>
26141
26142 PR target/56058
26143 * config/arm/marvell-pj4.md: Update copyright year.
26144 Fix up use of alu to alu_reg and simple_alu_imm.
26145
47876a2a 261462013-01-21 Uros Bizjak <ubizjak@gmail.com>
1f6bc337
UB
26147
26148 * config/i386/i386.md (enabled): Do not disable fma4 for TARGET_FMA.
26149
89d56d79
VM
261502013-01-20 Vladimir Makarov <vmakarov@redhat.com>
26151
26152 PR target/55433
26153 * lra-constraints.c (curr_insn_transform): Don't reuse original
1f6bc337 26154 insn for secondary memory move when memory mode should be different.
89d56d79 26155
fe603553
JDA
261562013-01-20 John David Anglin <dave.anglin@nrc-cnrc.gc.ca>
26157
26158 * config/pa/pa.md (atomic_loaddi, atomic_loaddi_1, atomic_storedi,
26159 atomic_storedi_1): New patterns.
26160
01284895
VK
261612013-01-20 Venkataramanan Kumar <venkataramanan.kumar@amd.com>
26162
26163 btver2 pipeline descriptions.
26164 * config/i386/i386.c: Enable CPU_BTVER2 to use btver2 pipeline
26165 descriptions.
26166 * config/i386/i386.md (btver2_decode): New type attributes.
1f6bc337 26167 * config/i386/sse.md (btver2_decode, btver2_sse_attr): New
01284895
VK
26168 type attributes.
26169 * config/i386/btver2.md: New file describing btver2 pipelines.
26170
5630e3e1
JL
261712013-01-19 Andrew Pinski <apinski@cavium.com>
26172
26173 PR tree-optimization/52631
26174 * tree-ssa-sccvn (visit_use): Before looking up the original
26175 statement, try looking up the simplified expression.
26176
650ae806
AG
261772013-01-19 Anthony Green <green@moxielogic.com>
26178
26179 * config/moxie/moxie.c (moxie_expand_prologue): Set
26180 current_function_static_stack_size.
26181
e300ec2d
JJ
261822013-01-18 Jakub Jelinek <jakub@redhat.com>
26183
26184 PR tree-optimization/56029
26185 * tree-phinodes.c (reserve_phi_args_for_new_edge): Set
26186 gimple_phi_arg_location for the new arg to UNKNOWN_LOCATION.
26187
a3d7af04
SS
261882013-01-18 Sharad Singhai <singhai@google.com>
26189
26190 PR tree-optimization/55995
26191 * dumpfile.c (dump_loc): Print location only if available.
26192 * tree-vectorizer.c (increase_alignment): Intialize vect_location.
26193
66aa7879
VM
261942013-01-18 Vladimir Makarov <vmakarov@redhat.com>
26195
26196 PR target/55433
26197 * lra-constraints.c (curr_insn_transform): Reuse original insn for
26198 secondary memory move.
26199 (inherit_reload_reg): Use rclass instead of cl for
26200 check_secondary_memory_needed_p.
26201
3f0fee7b
JJ
262022013-01-18 Jakub Jelinek <jakub@redhat.com>
26203
26204 PR middle-end/56015
26205 * expr.c (expand_expr_real_2) <case COMPLEX_EXPR>: Handle
1f6bc337 26206 the case where writing real complex part of target modifies op1.
3f0fee7b 26207
70c67693
JG
262082013-01-18 James Greenhalgh <james.greenhalgh@arm.com>
26209
26210 * config/aarch64/aarch64-simd.md
26211 (aarch64_vcond_internal<mode>): Handle unordered cases.
26212 * config/aarch64/iterators.md (v_cmp_result): New.
26213
df8de9b3
YHH
262142013-01-18 Yi-Hsiu Hsu <ahsu@marvell.com>
26215 Ramana Radhakrishnan <ramana.radhakrishnan@arm.com>
26216
26217 * config/arm/marvell-pj4.md: New file.
26218 * config/arm/arm.c (arm_issue_rate): Add marvell_pj4.
26219 * config/arm/arm.md (generic_sched): Add marvell_pj4.
26220 (generic_vfp): Likewise.
26221 * config/arm/arm-cores.def: Add marvell-pj4.
26222 * config/arm/arm-tune.md: Regenerate.
26223 * config/arm/arm-tables.opt: Regenerate.
26224 * config/arm/bpabi.h (BE8_LINK_SPEC): Add marvell_pj4.
26225 * doc/invoke.texi: Document marvell-pj4.
26226
be30c356
TB
262272013-01-18 Tejas Belagod <tejas.belagod@arm.com>
26228
26229 * config/aarch64/arm_neon.h: Map scalar types to standard types.
26230
0bfbca58 262312013-01-18 Alexandre Oliva <aoliva@redhat.com>
af6236c1
AO
26232
26233 PR debug/54114
26234 PR debug/54402
26235 PR debug/49888
26236 * var-tracking.c (negative_power_of_two_p): New.
26237 (global_get_addr_cache, local_get_addr_cache): New.
26238 (get_addr_from_global_cache, get_addr_from_local_cache): New.
26239 (vt_canonicalize_addr): Rewrite using the above. Adjust the
26240 heading comment.
26241 (vt_stack_offset_p): Remove.
26242 (vt_canon_true_dep): Always canonicalize loc's address.
26243 (clobber_overlapping_mems): Make sure we have a MEM.
26244 (local_get_addr_clear_given_value): New.
26245 (val_reset): Clear local cached entries.
26246 (compute_bb_dataflow): Create and release the local cache.
26247 Disable duplicate MEMs clobbering.
26248 (emit_notes_in_bb): Clobber MEMs likewise.
26249 (vt_emit_notes): Create and release the local cache.
26250 (vt_initialize, vt_finalize): Create and release the global
26251 cache, respectively.
1f6bc337 26252 * alias.c (rtx_equal_for_memref_p): Compare operands of ENTRY_VALUEs.
af6236c1 26253
0bfbca58 262542013-01-18 Alexandre Oliva <aoliva@redhat.com>
7ff37ffa
AO
26255
26256 PR libmudflap/53359
26257 * tree-mudflap.c (mudflap_finish_file): Skip deferred decls
26258 not found in the symtab.
26259
0bfbca58 262602013-01-18 Alexandre Oliva <aoliva@redhat.com>
3aa03517 26261
c350ba53 26262 PR debug/56006
3aa03517
AO
26263 PR rtl-optimization/55547
26264 PR rtl-optimization/53827
26265 PR debug/53671
26266 PR debug/49888
26267 * alias.c (offset_overlap_p): New, factored out of...
26268 (memrefs_conflict_p): ... this. Use absolute sizes. Retain
26269 the conservative special case for symbolic constants. Don't
26270 adjust zero sizes on alignment.
26271
c664546f
JL
262722013-01-18 Bernd Schmidt <bernds@codesourcery.com>
26273
26274 PR rtl-optimization/52573
26275 * regrename.c (build_def_use): Ignore REG_DEAD notes if there is a
26276 REG_UNUSED for the same register.
26277
1bd3f750
MP
262782013-01-17 Richard Biener <rguenther@suse.de>
26279 Marek Polacek <polacek@redhat.com>
26280
26281 PR rtl-optimization/55833
26282 * loop-unswitch.c (unswitch_loops): Move loop verification...
26283 (unswitch_single_loop): ...here. Call mark_irreducible_loops.
26284 * cfgloopmanip.c (fix_loop_placement): Add IRRED_INVALIDATED parameter.
26285 Set it to true when we're removing a loop from hierarchy tree in
26286 an irreducible region.
26287 (fix_bb_placements): Adjust caller.
26288 (fix_loop_placements): Likewise.
26289
e52a8b71
GJL
262902013-01-17 Georg-Johann Lay <avr@gjlay.de>
26291
26292 * config/avr/builtins.def (DEF_BUILTIN): Factor out
26293 "__builtin_avr_" from NAME, turn NAME to an uppercase identifier.
26294 Factor out 'CODE_FOR_' from ICODE, use 'nothing' instead of '-1'.
26295 Remove ID. Adjust comments.
26296 * config/avr/avr-c.c (avr_builtin_name): Remove.
26297 (avr_cpu_cpp_builtins): Use DEF_BUILTIN instead of for-loop.
26298 * config/avr/avr.c (avr_tolower): New static function.
26299 (DEF_BUILTIN): Remove parameter ID. Prefix ICODE by 'CODE_FOR_'.
26300 Stringify NAME, prefix it with "__builtin_avr_" and lowercase it.
26301 (avr_expand_builtin): Assert insn_code != CODE_FOR_nothing for
26302 default expansion.
26303
8386a7ea
JH
263042013-01-17 Jan Hubicka <jh@suse.cz>
26305
610fb637 26306 PR tree-optimization/55273
8386a7ea
JH
26307 * loop-iv.c (iv_number_of_iterations): Consider zero iteration case.
26308
47876a2a 263092013-01-17 Uros Bizjak <ubizjak@gmail.com>
8abaebcd
UB
26310
26311 PR target/55981
26312 * config/i386/sync.md (atomic_store<mode>): Always generate SWImode
26313 store through atomic_store<mode>_1.
26314 (atomic_store<mode>_1): Macroize insn using SWI mode iterator.
26315
8222c37e
MJ
263162013-01-17 Martin Jambor <mjambor@suse.cz>
26317
26318 PR tree-optimizations/55264
26319 * ipa-inline-transform.c (can_remove_node_now_p_1): Never return true
26320 for virtual methods.
26321 * ipa.c (symtab_remove_unreachable_nodes): Never return true for
26322 virtual methods before inlining is over.
26323 * cgraph.h (cgraph_only_called_directly_or_aliased_p): Return false for
26324 virtual functions.
26325 * cgraphclones.c (cgraph_create_virtual_clone): Mark clones as
26326 non-virtual.
26327
79f01c76
VM
263282013-01-16 Vladimir Makarov <vmakarov@redhat.com>
26329
26330 PR rtl-optimization/56005
26331 * sched-deps.c (sched_analyze_2): Check deps->readonly for adding
26332 pending reads for prefetch.
26333
d0b6bb1b
IB
263342013-01-16 Ian Bolton <ian.bolton@arm.com>
26335
aaf1e810 26336 * config/aarch64/aarch64.md
d0b6bb1b
IB
26337 (*cstoresi_neg_uxtw): New pattern.
26338 (*cmovsi_insn_uxtw): New pattern.
26339 (*<optab>si3_uxtw): New pattern.
26340 (*<LOGICAL:optab>_<SHIFT:optab>si3_uxtw): New pattern.
26341 (*<optab>si3_insn_uxtw): New pattern.
26342 (*bswapsi2_uxtw): New pattern.
26343
cb9cf03b
RB
263442013-01-16 Richard Biener <rguenther@suse.de>
26345
26346 * tree-inline.c (tree_function_versioning): Remove set but
26347 never used variable.
26348
2cfc56b9
RB
263492013-01-16 Richard Biener <rguenther@suse.de>
26350
26351 PR tree-optimization/55964
26352 * tree-flow.h (rename_variables_in_loop): Remove.
26353 (rename_variables_in_bb): Likewise.
26354 * tree-loop-distribution.c (update_phis_for_loop_copy): Remove.
26355 (copy_loop_before): Adjust and delete update-ssa status.
26356 * tree-vect-loop-manip.c (rename_variables_in_bb): Make static.
8abaebcd 26357 (rename_variables_in_bb): Likewise. Properly walk over predecessors.
2cfc56b9
RB
26358 (rename_variables_in_loop): Remove.
26359 (slpeel_update_phis_for_duplicate_loop): Likewise.
26360 (slpeel_tree_duplicate_loop_to_edge_cfg): Handle nested loops,
26361 use available cfg machinery instead of duplicating it.
26362 Update PHI nodes and perform poor-mans SSA update here.
26363 (slpeel_tree_peel_loop_to_edge): Adjust.
26364
c25a0c60
RB
263652013-01-16 Richard Biener <rguenther@suse.de>
26366
26367 PR tree-optimization/54767
26368 PR tree-optimization/53465
26369 * tree-vrp.c (vrp_meet_1): Revert original fix for PR53465.
26370 (vrp_visit_phi_node): For PHI arguments coming via backedges
26371 drop all symbolical range information.
26372 (execute_vrp): Compute backedges.
26373
04b535af
RB
263742013-01-16 Richard Biener <rguenther@suse.de>
26375
26376 * doc/install.texi: Update CLooG and ISL requirements to
26377 0.18.0 and 0.11.1.
26378
8b0a1e0b
CB
263792013-01-16 Christian Bruel <christian.bruel@st.com>
26380
26381 PR target/55301
26382 * config/sh/sh.c (sh_expand_prologue): Postpone new_stack mem symbol.
26383 (broken_move): Handle UNSPECV_SP_SWITCH_B.
26384 * config/sh/sh.md (sp_switch_1): Use set (reg:SI SP_REG).
26385
263862013-01-16 DJ Delorie <dj@redhat.com>
26387
26388 * config/sh/sh.md (UNSPECV_SP_SWITCH_B): New.
26389 (UNSPECV_SP_SWITCH_E): New.
26390 (sp_switch_1): Change to an unspec.
26391 (sp_switch_2): Change to an unspec. Don't use post-inc when we
26392 replace $r15.
26393
263942013-01-16 Uros Bizjak <ubizjak@gmail.com>
88e784e6
UB
26395
26396 * emit-rtl.c (need_atomic_barrier_p): Mask memory model argument
26397 with MEMMODEL_MASK before comparing with MEMMODEL_* memory types.
26398 * optabs.c (maybe_emit_sync_lock_test_and_set): Ditto.
26399 (expand_mem_thread_fence): Ditto.
26400 (expand_mem_signal_fence): Ditto.
26401 (expand_atomic_load): Ditto.
26402 (expand_atomic_store): Ditto.
26403
0bfbca58 264042013-01-16 Alexandre Oliva <aoliva@redhat.com>
5147bf6a
AO
26405
26406 PR rtl-optimization/55547
26407 PR rtl-optimization/53827
26408 PR debug/53671
26409 PR debug/49888
26410 * alias.c (memrefs_conflict_p): Set sizes to negative after
26411 AND adjustments.
26412
305e3ac1
JJ
264132013-01-15 Jakub Jelinek <jakub@redhat.com>
26414
26415 PR target/55940
26416 * function.c (thread_prologue_and_epilogue_insns): Always
26417 add crtl->drap_reg to set_up_by_prologue.set, even if
26418 stack_realign_drap is false.
26419
f78ac4f2
JBG
264202013-01-15 Jan-Benedict Glaw <jbglaw@lug-owl.de>
26421
26422 * config/vax/vax.md (add<mode>3, sub<mode>3, mul<mode>3, div<mode>3,
26423 and<mode>3, *and<mode>_const_int, ior<mode>3, xor<mode>3, ashrsi3,
26424 *call): Fix indention.
26425
a78a8cc4
TV
264262013-01-15 Tom de Vries <tom@codesourcery.com>
26427
26428 PR target/55876
26429 * optabs.c (widen_operand): Use gen_lowpart instead of gen_rtx_SUBREG.
26430 Update comment.
26431
0e80383f
VM
264322013-01-15 Vladimir Makarov <vmakarov@redhat.com>
26433
305e3ac1 26434 PR rtl-optimization/55153
0e80383f
VM
26435 * sched-deps.c (sched_analyze_2): Add pending reads for prefetch.
26436
207b5956
MJ
264372013-01-15 Martin Jambor <mjambor@suse.cz>
26438
26439 PR tree-optimization/55920
26440 * tree-sra.c (analyze_access_subtree): Do not mark non-removable
26441 accesses as grp_to_be_debug_replaced.
26442
a7818b54
JJ
264432013-01-15 Jakub Jelinek <jakub@redhat.com>
26444
26445 PR tree-optimization/55920
26446 * tree-sra.c (sra_modify_assign): If for lacc->grp_to_be_debug_replaced
26447 there is non-useless type conversion needed from debug rhs to lhs,
26448 use build_debug_ref_for_model and/or VIEW_CONVERT_EXPR.
26449
b0fe107e
JM
264502013-01-15 Joseph Myers <joseph@codesourcery.com>
26451 Mikael Pettersson <mikpe@it.uu.se>
26452
26453 PR target/43961
26454 * config/arm/arm.h (ADDR_VEC_ALIGN): Align SImode jump tables for
26455 Thumb.
26456 (ASM_OUTPUT_CASE_LABEL): Remove.
26457 (ASM_OUTPUT_BEFORE_CASE_LABEL): Define to empty.
26458 * final.c (shorten_branches): Update alignment of labels before
26459 jump tables if CASE_VECTOR_SHORTEN_MODE.
26460
34ab62ee
RB
264612013-01-15 Richard Biener <rguenther@suse.de>
26462
26463 PR bootstrap/55961
26464 * system.h: Do not include gmp.h for building host tools.
26465
783a3a05
RB
264662013-01-15 Richard Biener <rguenther@suse.de>
26467
26468 PR middle-end/55882
26469 * emit-rtl.c (set_mem_attributes_minus_bitpos): Correctly
26470 account for bitpos when computing alignment.
26471
3a579e09
VY
264722013-01-15 Vladimir Yakovlev <vladimir.b.yakovlev@intel.com>
26473
26474 * config/i386/i386-c.c (ix86_target_macros_internal): New case.
26475 (ix86_target_macros_internal): Likewise.
26476
26477 * config/i386/i386.c (m_CORE2I7): Removed.
26478 (m_CORE_HASWELL): New macro.
26479 (m_CORE_ALL): Likewise.
26480 (initial_ix86_tune_features): m_CORE2I7 is replaced by m_CORE_ALL.
26481 (initial_ix86_arch_features): Likewise.
26482 (processor_target_table): Initializations for Core avx2.
26483 (cpu_names): New names "core-avx2".
26484 (ix86_option_override_internal): Changed PROCESSOR_COREI7 by
26485 PROCESSOR_CORE_HASWELL.
26486 (ix86_issue_rate): New case.
26487 (ia32_multipass_dfa_lookahead): Likewise.
26488 (ix86_sched_init_global): Likewise.
26489
26490 * config/i386/i386.h (TARGET_HASWELL): New macro.
26491 (target_cpu_default): New TARGET_CPU_DEFAULT_haswell.
26492 (processor_type): New PROCESSOR_HASWELL.
26493
955f5a07
JJ
264942013-01-15 Jakub Jelinek <jakub@redhat.com>
26495
ff784829
JJ
26496 PR tree-optimization/55955
26497 * tree-vect-loop.c (vectorizable_reduction): Give up early on
26498 *SHIFT_EXPR and *ROTATE_EXPR codes.
26499
955f5a07
JJ
26500 PR tree-optimization/48766
26501 * opts.c (common_handle_option): For -fwrapv disable -ftrapv, for
26502 -ftrapv disable -fwrapv.
26503
aeb8b4e9
GJL
265042013-01-14 Georg-Johann Lay <avr@gjlay.de>
26505
26506 PR target/55974
26507 * config/avr/avr-c.c (avr_cpu_cpp_builtins): Define __FLASH
26508 etc. to 1 and not to __flash.
26509 Use LL suffix for __INT24_MAX__ with -mint8.
26510 Use ULL suffix for __UINT24_MAX__ with -mint8.
26511
1c494c6a
GJL
265122013-01-14 Georg-Johann Lay <avr@gjlay.de>
26513
26514 * config/avr/avr-arch.h
26515 (struct base_arch_s): Use typedef avr_arch_t instead.
26516 (struct arch_info_s): Use typedef avr_arch_info_t instead.
26517 (struct mcu_type_s): Use typedef avr_mcu_t instead.
26518 * config/avr/avr.c: Same.
26519 * config/avr/avr-devices.c: Same.
26520 * config/avr/driver-avr.c: Same.
26521 * config/avr/gen-avr-mmcu-texi.c: Same.
26522 * config/avr/avr-mcus.def: Adjust comment.
26523
a50344cb
TB
265242013-01-14 Tejas Belagod <tejas.belagod@arm.com>
26525
88e784e6
UB
26526 * config/aarch64/aarch64-simd.md (*aarch64_simd_ld1r<mode>): New.
26527 * config/aarch64/iterators.md (VALLDI): New.
a50344cb 26528
47876a2a 265292013-01-14 Uros Bizjak <ubizjak@gmail.com>
e6f0e052
UB
26530 Andi Kleen <ak@linux.intel.com>
26531
26532 PR target/55948
26533 * config/i386/sync.md (atomic_store<mode>_1): New pattern.
26534 (atomic_store<mode>): Call atomic_store<mode>_1 for IX86_HLE_RELEASE
26535 memmodel flag.
26536
00892272
GJL
265372013-01-14 Georg-Johann Lay <avr@gjlay.de>
26538
26539 * config/avr/avr-stdint.h: Remove trailing blanks.
26540 * config/avr/avr-log.h: Same.
26541 * config/avr/avr-arch.h: Same.
26542 * config/avr/avr-devices.c: Same.
26543 * config/avr/avr-dimode.md: Same.
26544 * config/avr/predicates.md: Same.
26545 * config/avr/avr-c.c: Same. And fix typo.
e6f0e052 26546
00892272
GJL
26547 * config/avr/avr-protos.h: Same. And:
26548 (function_arg_regno_p): Rename to avr_function_arg_regno_p.
26549 (init_cumulative_args): Rename to avr_init_cumulative_args.
26550 (expand_prologue): Rename to avr_expand_prologue.
26551 (expand_epilogue): Rename to avr_expand_epilogue.
26552 (adjust_insn_length): Rename to avr_adjust_insn_length.
26553 (notice_update_cc): Rename to avr_notice_update_cc.
26554 (final_prescan_insn): Rename to avr_final_prescan_insn.
26555 * config/avr/avr.c: Same.
26556 * config/avr/avr.h: Same.
26557 * config/avr/avr.md: Remove trailing blanks.
26558 (prologue): Use avr_expand_prologue.
26559 (epilogue, sibcall_epilogue): Use avr_expand_epilogue.
26560
7bb01996
RB
265612013-01-14 Richard Biener <rguenther@suse.de>
26562
26563 * tree-cfg.c (verify_expr_location, verify_expr_location_1,
26564 verify_location, collect_subblocks): New functions.
26565 (verify_gimple_in_cfg): Verify that locations only reference
26566 BLOCKs in the functions BLOCK tree.
26567
2724573f
RB
265682013-01-14 Richard Biener <rguenther@suse.de>
26569
26570 * tree-cfgcleanup.c (remove_forwarder_block): Unshare propagated
26571 PHI argument.
26572 * graphite-sese-to-poly.c (insert_out_of_ssa_copy): Properly
26573 unshare reference.
26574 (insert_out_of_ssa_copy_on_edge): Likewise.
26575 (rewrite_close_phi_out_of_ssa): Likewise.
26576 * tree-ssa.c (insert_debug_temp_for_var_def): Properly unshare
26577 debug expressions.
26578 * tree-ssa-pre.c (insert_into_preds_of_block): Properly unshare
26579 propagated constants.
26580 * tree-cfg.c (tree_node_can_be_shared): Handled component-refs
26581 can not be shared.
26582
9a0bbab6
GJL
265832013-01-14 Georg-Johann Lay <avr@gjlay.de>
26584
26585 * config/avr/avr-modes.def: Add GPL copyright notice.
26586
45805f17
UB
265872013-01-13 Uros Bizjak <ubizjak@gmail.com>
26588
26589 * config/i386/sync.md (mem_thread_fence): Mask operands[0] with
26590 MEMMODEL_MASK to determine memory model.
26591 (atomic_store<mode>): Ditto from operands[2].
26592 * config/i386/i386.c (ix86_memmodel_check): Declare "strong" as bool.
26593
9d60be38
JJ
265942013-01-13 Jakub Jelinek <jakub@redhat.com>
26595
26596 PR fortran/55935
45805f17 26597 * gimple-fold.c (get_symbol_constant_value): Call unshare_expr.
9d60be38
JJ
26598 (fold_gimple_assign): Don't call unshare_expr here.
26599 (fold_ctor_reference): Call unshare_expr.
26600
e7f49d92
TG
266012013-01-13 Terry Guo <terry.guo@arm.com>
26602
aaf1e810
EB
26603 * Makefile.in (s-mlib): New argument MULTILIB_REUSE.
26604 * doc/fragments.texi: Document MULTILIB_REUSE.
26605 * gcc.c (multilib_reuse): New internal spec.
e7f49d92 26606 (set_multilib_dir): Also search multilib from multilib_reuse.
aaf1e810 26607 * genmultilib (tmpmultilib3): Refactor code.
e7f49d92
TG
26608 (tmpmultilib4): Ditto.
26609 (multilib_reuse): New multilib argument.
26610
fbd03a27
RS
266112013-01-13 Richard Sandiford <rdsandiford@googlemail.com>
26612
26613 * Makefile.in: Update copyright.
26614
b3681f13
TV
266152013-01-12 Tom de Vries <tom@codesourcery.com>
26616
26617 PR middle-end/55890
26618 * calls.c (expand_call): Check if arg_nr is valid.
26619
3f287e4b
MM
266202013-01-11 Michael Meissner <meissner@linux.vnet.ibm.com>
26621
26622 * doc/extend.texi (X86 Built-in Functions): Add whitespace in
26623 __builtin_ia32_paddb256 and __builtin_ia32_pavgb256
5bc35113
MM
26624 documentation. Add missing '__' in front of
26625 __builtin_ia32_packssdw256.
3f287e4b 26626
1abcd5eb
AK
266272013-01-11 Andreas Krebbel <Andreas.Krebbel@de.ibm.com>
26628
26629 PR target/55719
26630 * config/s390/s390.c (s390_preferred_reload_class): Do not return
26631 NO_REGS for larl operands.
26632 (s390_reload_larl_operand): Use s390_load_address instead of
26633 emit_move_insn.
26634
980d0812
RB
266352013-01-11 Richard Biener <rguenther@suse.de>
26636
26637 * tree-cfg.c (verify_node_sharing_1): Split out from ...
26638 (verify_node_sharing): ... here.
26639 (verify_gimple_in_cfg): Use verify_node_sharing_1 for walk_tree.
26640
7cb9fd07
EB
266412013-01-11 Eric Botcazou <ebotcazou@adacore.com>
26642
26643 * configure.ac (Tree checking): Set TREECHECKING to yes if enabled.
26644 Substitute TREECHECKING.
26645 * configure: Regenerate.
26646 * Makefile.in (TREECHECKING): New.
26647
47876a2a 266482013-01-11 Richard Guenther <rguenther@suse.de>
c7ab2530
RG
26649
26650 PR tree-optimization/44061
26651 * tree-vrp.c (extract_range_basic): Compute zero as
26652 value-range for __builtin_constant_p of function parameters.
26653
d1e082c2
RS
266542013-01-10 Richard Sandiford <rdsandiford@googlemail.com>
26655
45805f17 26656 Update copyright years.
d1e082c2 26657
f85021b0
VM
266582013-01-10 Vladimir Makarov <vmakarov@redhat.com>
26659
9d60be38 26660 PR rtl-optimization/55672
0160442c
VM
26661 * lra-eliminations.c (mark_not_eliminable): Permit addition with
26662 const to be eliminable.
f85021b0 26663
7a8b1ec4
DE
266642013-01-10 David Edelsohn <dje.gcc@gmail.com>
26665
26666 * configure.ac (HAVE_AS_TLS): Add check for powerpc-ibm-aix.
26667 * configure: Regenerate.
26668
ad2e5b71
RB
266692013-01-10 Richard Biener <rguenther@suse.de>
26670
ddf9322d 26671 * builtins.c (expand_builtin_init_trampoline): Use set_mem_attributes.
ad2e5b71 26672
ffc5b2bb
RB
266732013-01-10 Richard Biener <rguenther@suse.de>
26674
26675 PR bootstrap/55792
26676 * tree-into-ssa.c (rewrite_add_phi_arguments): Do not set
26677 locations for virtual PHI arguments.
26678 (rewrite_update_phi_arguments): Likewise.
26679
e1f674e4
JS
266802013-01-10 Joel Sherrill <joel.sherrill@OARcorp.com>
26681
26682 * config/v850/rtems.h (ASM_SPEC): Pass -m8byte-align and -mgcc-abi
26683 on to assembler.
26684
a57fc743
JJ
266852013-01-10 Jakub Jelinek <jakub@redhat.com>
26686
26687 PR tree-optimization/55921
26688 * tree-complex.c (expand_complex_asm): New function.
26689 (expand_complex_operations_1): Call it for GIMPLE_ASM.
26690
0ff4390d
AK
266912013-01-10 Andreas Krebbel <Andreas.Krebbel@de.ibm.com>
26692
26693 PR target/55718
26694 * config/s390/s390.c (s390_symref_operand_p)
26695 (s390_loadrelative_operand_p): Merge the two functions.
26696 (s390_check_qrst_address, print_operand_address): Add parameters
26697 to s390_loadrelative_operand_p invokation.
26698 (s390_check_symref_alignment): Use s390_loadrelative_operand_p.
26699 (s390_reload_larl_operand, s390_secondary_reload): Use
26700 s390_loadrelative_operand_p instead of s390_symref_operand_p.
26701 (legitimize_pic_address): Handle @GOTENT and @PLT + addend.
26702
6f557e0e
MS
267032013-01-09 Mike Stump <mikestump@comcast.net>
26704
26705 * dse.c (record_store): Remove unnecessary assert.
26706
7770c9e9
JH
267072013-01-09 Jan Hubicka <jh@suse.cz>
26708
26709 PR tree-optimization/55569
26710 * cfgloopmanip.c (scale_loop_profile): Make ITERATION_BOUND gcov_type.
26711 * cfgloop.h (scale_loop_profile): Likewise.
26712
a19b1432
JH
267132013-01-09 Jan Hubicka <jh@suse.cz>
26714
26715 PR lto/45375
ddf9322d
UB
26716 * ipa-inline.c (ipa_inline): Remove extern inlines and virtual
26717 functions.
a19b1432
JH
26718 * cgraphclones.c (cgraph_clone_node): Cpoy also LTO file data.
26719
9a002da8
RS
267202013-01-09 Richard Sandiford <rdsandiford@googlemail.com>
26721
26722 PR middle-end/55114
26723 * expr.h (maybe_emit_group_store): Declare.
26724 * expr.c (maybe_emit_group_store): New function.
26725 * builtins.c (expand_builtin_int_roundingfn): Call it.
26726 (expand_builtin_int_roundingfn_2): Likewise.
26727
511dcace
VM
267282013-01-09 Vladimir Makarov <vmakarov@redhat.com>
26729
e1f2b729 26730 PR rtl-optimization/55829
511dcace
VM
26731 * lra-constraints.c (match_reload): Add code for absent output.
26732 (curr_insn_transform): Add code for reloads of matched inputs
26733 without output.
26734
7b0fe4f4
UB
267352013-01-09 Uros Bizjak <ubizjak@gmail.com>
26736
26737 * config/i386/sse.md (*vec_interleave_highv2df): Change mode
26738 attribute of movddup insn to DF.
26739 (*vec_interleave_lowv2df): Ditto.
26740 (vec_dupv2df): Ditto.
26741
870ca331
JH
267422013-01-09 Jan Hubicka <jh@suse.cz>
26743
26744 PR tree-optimiation/55875
26745 * tree-ssa-loop-niter.c (number_of_iterations_cond): Add
26746 EVERY_ITERATION parameter.
7b0fe4f4 26747 (number_of_iterations_exit): Check if exit is executed every iteration.
870ca331
JH
26748 (idx_infer_loop_bounds): Similarly here.
26749 (n_of_executions_at_most): Simplify
26750 to only test for cases where statement is dominated by the
7b0fe4f4 26751 particular bound; handle correctly the "postdominance" test.
870ca331
JH
26752 (scev_probably_wraps_p): Use max loop iterations info
26753 as a global bound first.
26754
6f575fe4 267552013-01-09 Nguyen Duy Dat <dat.nguyen.yn@rvc.renesas.com>
3cd232b5
NDD
26756 Nick Clifton <nickc@redhat.com>
26757
26758 * config/v850/v850.md (cbranchsf4): New pattern.
26759 (cstoresf4): New pattern.
26760 (cbranchdf4): New pattern.
26761 (cstoredf4): New pattern.
26762 (movsicc): Disallow floating point comparisons.
26763 (cmpsf_le_insn): Fix order of operators.
26764 (cmpsf_lt_insn): Likewise.
26765 (cmpsf_eq_insn): Likewise.
26766 (cmpdf_le_insn): Likewise.
26767 (cmpdf_lt_insn): Likewise.
26768 (cmpdf_eq_insn): Likewise.
26769 (cmpsf_ge_insn): Use LE comparison.
26770 (cmpdf_ge_insn): Likewise.
26771 (cmpsf_gt_insn): Use LT comparison.
26772 (cmpdf_gt_insn): Likewise.
26773 (cmpsf_ne_insn): Delete pattern.
26774 (cmpdf_ne_insn): Delete pattern.
26775 * config/v850/v850.c (v850_gen_float_compare): Use
26776 gen_cmpdf_eq_insn for NE comparison.
26777 (v850_float_z_comparison_operator)
26778 (v850_float_nz_comparison_operator): Move from here ...
26779 * config/v850/predicates.md: ... to here. Move GT and GE
26780 comparisons into v850_float_z_comparison_operator.
26781 * config/v850/v850-protos.h (v850_float_z_comparison_operator):
26782 Delete prototype.
26783 (v850_float_nz_comparison_operator): Likewise.
26784
f0d54148
JDA
267852013-01-09 John David Anglin <dave.anglin@nrc-cnrc.gc.ca>
26786
26787 * config/pa/pa.c (pa_emit_move_sequence): Replace calls to gen_insv
26788 with calls to gen_insvsi/gen_insvdi.
26789
8f01beca
VK
267902013-01-09 Venkataramanan Kumar <venkataramanan.kumar@amd.com>
26791
26792 * config/i386/i386.c (initial_ix86_tune_features): Set up
26793 X86_TUNE_AVX128_OPTIMAL for m_BTVER2.
26794
2aa579ad
JJ
267952013-01-09 Steven Bosscher <steven@gcc.gnu.org>
26796 Jakub Jelinek <jakub@redhat.com>
26797
26798 PR tree-optimization/48189
26799 * predict.c (predict_loops): If max is 0, don't call compare_tree_int.
26800 If nitercst is 0, don't predict the exit edge.
26801
6edc3e32 268022013-01-08 Naveen H.S <Naveen.Hurugalawadi@caviumnetworks.com>
01a3a324
N
26803
26804 * config/aarch64/aarch64.c (aarch64_print_operand): Replace %r
26805 in asm_fprintf with reg_names.
26806 (aarch64_print_operand_address): Likewise.
26807 (aarch64_return_addr): Likewise.
26808 * config/aarch64/aarch64.h (ASM_FPRINTF_EXTENSIONS): Remove.
26809
f6f94d94
JDA
268102013-01-08 John David Anglin <dave.anglin@nrc-cnrc.gc.ca>
26811
26812 * config/pa/pa.h (VAL_U6_BITS_P): Define.
26813 (INT_U6_BITS): Likewise.
26814 * config/pa/predicates.md (uint6_operand): New predicate.
26815 (shift5_operand, shift6_operand): Likewise.
26816 * config/pa/pa.md (lshrsi3, rotrsi3): Use shift5_operand instead of
26817 arith32_operand.
26818 (lshrdi3): Use shift6_operand.
26819 (shrpsi4, shrpdi4): New insn patterns.
26820 (extzv): Delete expander.
26821 (extzvsi, extzvdi): New expanders. Use uint5_operand and uint6_operand
26822 predicates in unamed zero extract patterns. Tighten common constraint.
26823 (extv): Delete expander.
26824 (extvsi, extvdi): New expanders. Use uint5_operand and uint6_operand
26825 predicates in unamed sign extract patterns. Tighten common constraint.
26826 (insv): Delete expander.
26827 (insvsi, insvdi): New expanders. Use uint5_operand and uint6_operand
26828 predicates in unamed insert patterns. Tighten common constraint.
26829 Change uint32_operand predicate to uint6_operand predicate in unamed
26830 DImode pattern to insert constant values of type 1...1xxxx.
26831
36b72910
JH
268322013-01-04 Jan Hubicka <jh@suse.cz>
26833
26834 PR tree-optimization/55823
7b0fe4f4
UB
26835 * ipa-prop.c (update_indirect_edges_after_inlining): Fix ordering
26836 issue.
36b72910 26837
47876a2a 268382013-01-08 Jakub Jelinek <jakub@redhat.com>
c6d851b9
JJ
26839 Uros Bizjak <ubizjak@gmail.com>
26840
26841 PR rtl-optimization/55845
26842 * df-problems.c (can_move_insns_across): Stop scanning at
26843 volatile_insn_p source instruction or give up if
26844 across_from .. across_to range contains any volatile_insn_p
26845 instructions.
26846
4369c11e
TB
268472013-01-08 Tejas Belagod <tejas.belagod@arm.com>
26848
7b0fe4f4
UB
26849 * config/aarch64/aarch64-simd.md (vec_init<mode>): New.
26850 * config/aarch64/aarch64-protos.h (aarch64_expand_vector_init):
26851 Declare.
4369c11e 26852 * config/aarch64/aarch64.c (aarch64_simd_dup_constant,
7b0fe4f4 26853 aarch64_simd_make_constant, aarch64_expand_vector_init): New.
4369c11e 26854
aeb7e7c1
JJ
268552013-01-08 Jakub Jelinek <jakub@redhat.com>
26856
26857 PR fortran/55341
26858 * asan.c (asan_clear_shadow): New function.
26859 (asan_emit_stack_protection): Use it.
26860
a02ad1aa
TB
268612013-01-08 Tejas Belagod <tejas.belagod@arm.com>
26862
26863 * config/aarch64/aarch64-simd.md (aarch64_simd_vec_<su>mult_lo_<mode>,
26864 aarch64_simd_vec_<su>mult_hi_<mode>): Separate instruction and operand
26865 with tab instead of space.
26866
f90d87f5
NC
268672013-01-08 Nick Clifton <nickc@redhat.com>
26868
26869 * config/rl78/rl78.c (rl78_expand_prologue): Always select
26870 register bank 0 at the start of an interrupt handler.
83ffd964
NC
26871 * config/rl78/rl78.md (mulsi3_g13): Correct values for MDBL and
26872 MDBH registers.
f90d87f5 26873
385eb93d
JG
268742013-01-08 James Greenhalgh <james.greenhalgh@arm.com>
26875
26876 * config/aarch64/aarch64-simd.md
26877 (aarch64_simd_bsl<mode>_internal): Add floating-point modes.
26878 (aarch64_simd_bsl): Likewise.
26879 (aarch64_vcond_internal<mode>): Likewise.
26880 (vcond<mode><mode>): Likewise.
26881 (aarch64_cm<cmp><mode>): Fix constraints, add new modes.
26882 * config/aarch64/iterators.md (V_cmp_result): Add V2DF.
26883
4dcd1054
JG
268842013-01-08 James Greenhalgh <james.greenhalgh@arm.com>
26885
26886 * config/aarch64/aarch64-builtins.c
26887 (aarch64_builtin_vectorized_function): Handle sqrt, sqrtf.
26888
4267a4a6
MJ
268892013-01-08 Martin Jambor <mjambor@suse.cz>
26890
26891 PR debug/55579
26892 * tree-sra.c (analyze_access_subtree): Return true also after
26893 potentially creating a debug-only replacement.
26894
5f4e6de3
JJ
268952013-01-08 Jakub Jelinek <jakub@redhat.com>
26896
3138f224
JJ
26897 PR middle-end/55890
26898 * tree-ssa-ccp.c (evaluate_stmt): Use gimple_call_builtin_p.
26899
5f4e6de3
JJ
26900 PR tree-optimization/54120
26901 * tree-vrp.c (range_fits_type_p): Don't allow
26902 src_precision < precision from signed vr to unsigned_p
26903 if vr->min or vr->max is negative.
26904 (simplify_float_conversion_using_ranges): Test can_float_p
26905 against CODE_FOR_nothing.
26906
f3ef18ff
JJ
269072013-01-08 Jakub Jelinek <jakub@redhat.com>
26908 Richard Biener <rguenther@suse.de>
26909
26910 PR middle-end/55851
26911 * fold-const.c (int_binop_types_match_p): Allow all INTEGRAL_TYPE_P
26912 types instead of just INTEGER_TYPE types.
26913
25c210f9
MK
269142013-01-07 Mark Kettenis <kettenis@openbsd.org>
26915
26916 * config/i386/openbsdelf.h (LIBGCC2_HAS_TF_MODE, LIBGCC2_TF_CEXT,
26917 TF_SIZE): Define.
f90d87f5 26918
81826a7b
SE
269192013-01-07 Steve Ellcey <sellcey@mips.com>
26920
26921 PR target/42661
26922 * config/mips/mips.opt: Change mad to mmad to match documentation.
26923
35678514
GJL
269242013-01-07 Georg-Johann Lay <avr@gjlay.de>
26925
26926 PR target/55897
26927 * doc/extend.texi (AVR Named Address Spaces): __memx goes into
26928 .progmemx.data now.
26929
e5669488
GJL
269302013-01-07 Georg-Johann Lay <avr@gjlay.de>
26931
26932 PR target/55897
26933 * config/avr/avr.h (ADDR_SPACE_COUNT): New enum.
26934 (avr_addrspace_t): Add .section_name field.
26935 * config/avr/avr.c (progmem_section): Use ADDR_SPACE_COUNT as
26936 array size.
26937 (avr_addrspace): Same. Initialize .section_name. Remove last
26938 NULL entry. Put __memx into .progmemx.data.
26939 (progmem_section_prefix): Remove.
26940 (avr_asm_init_sections): No need to initialize progmem_section.
26941 (avr_asm_named_section): Use avr_addrspace[].section_name to get
26942 section name prefix.
26943 (avr_asm_select_section): Ditto. And use get_unnamed_section to
26944 retrieve the progmem section.
26945 * avr-c.c (avr_cpu_cpp_builtins): Use ADDR_SPACE_COUNT as loop
26946 boundary to run over avr_addrspace[].
26947 (avr_register_target_pragmas): Ditto.
26948
656e6f37
JJ
269492013-01-06 Jakub Jelinek <jakub@redhat.com>
26950
26951 * varasm.c (output_constant_def_contents): For asan_protect_global
26952 protected strings, adjust DECL_ALIGN if needed, before testing for
26953 anchored symbols.
26954 (place_block_symbol): Adjust size for asan protected STRING_CSTs if
26955 TREE_CONSTANT_POOL_ADDRESS_P. Increase alignment for asan protected
26956 normal decls.
26957 (output_object_block): For asan protected decls, emit asan padding
26958 after their contents.
26959 * asan.c (asan_protect_global): Don't check TREE_ASM_WRITTEN here.
26960 (asan_finish_file): Test it here instead.
26961
6edc3e32
UB
269622013-01-07 Nick Clifton <nickc@redhat.com>
26963 Matthias Klose <doko@debian.org>
26964 Doug Kwan <dougkwan@google.com>
26965 H.J. Lu <hongjiu.lu@intel.com>
b352afba
NC
26966
26967 PR driver/55470
26968 * collect2.c (main): Support -fuse-ld=bfd and -fuse-ld=gold.
26969
26970 * common.opt: Add fuse-ld=bfd and fuse-ld=gold.
26971
26972 * gcc.c (LINK_COMMAND_SPEC): Pass -fuse-ld=* to collect2.
26973
7b0fe4f4 26974 * opts.c (comman_handle_option): Ignore -fuse-ld=bfd and -fuse-ld=gold.
b352afba
NC
26975
26976 * doc/invoke.texi: Document -fuse-ld=bfd and -fuse-ld=gold.
26977
2a095093
GJL
269782013-01-07 Georg-Johann Lay <avr@gjlay.de>
26979
f3b503f4 26980 PR target/54461
2a095093
GJL
26981 * doc/install.texi (Cross-Compiler-Specific Options): Document
26982 --with-avrlibc.
26983
383f9b34
TB
269842013-01-07 Tejas Belagod <tejas.belagod@arm.com>
26985
26986 * config/aarch64/arm_neon.h (vmovn_high_is16, vmovn_high_s32,
26987 vmovn_high_s64, vmovn_high_u16, vmovn_high_u32, vmovn_high_u64,
26988 vqmovn_high_s16, vqmovn_high_s32, vqmovn_high_s64, vqmovn_high_u16,
26989 vqmovn_high_u32, vqmovn_high_u64, vqmovun_high_s16, vqmovun_high_s32,
26990 vqmovun_high_s64): Fix source operand number and update copyright.
26991
3626621a
RB
269922013-01-07 Richard Biener <rguenther@suse.de>
26993
26994 PR middle-end/55890
26995 * gimple.h (gimple_call_builtin_p): New overload.
26996 * gimple.c (validate_call): New function.
26997 (gimple_call_builtin_p): Likewise.
26998 * tree-ssa-structalias.c (find_func_aliases_for_builtin_call):
26999 Use gimple_call_builtin_p.
27000 (find_func_clobbers): Likewise.
27001 * tree-ssa-strlen.c (adjust_last_stmt): Likewise.
27002 (strlen_optimize_stmt): Likewise.
27003
8b2ea410
JG
270042013-01-07 James Greenhalgh <james.greenhalgh@arm.com>
27005
27006 * config/aarch64/arm_neon.h (vld1_dup_*): Make argument const.
27007 (vld1q_dup_*): Likewise.
27008 (vld1_*): Likewise.
27009 (vld1q_*): Likewise.
27010 (vld1_lane_*): Likewise.
27011 (vld1q_lane_*): Likewise.
27012
9713d329
RB
270132013-01-07 Richard Biener <rguenther@suse.de>
27014
27015 * lto-streamer.h (LTO_minor_version): Bump to 2.
27016
3520f7cc
JG
270172013-01-07 James Greenhalgh <james.greenhalgh@arm.com>
27018
27019 * config/aarch64/aarch64-protos.h
27020 (aarch64_const_double_zero_rtx_p): Rename to...
27021 (aarch64_float_const_zero_rtx_p): ...this.
27022 (aarch64_float_const_representable_p): New.
27023 (aarch64_output_simd_mov_immediate): Likewise.
27024 * config/aarch64/aarch64-simd.md (*aarch64_simd_mov<mode>): Refactor
27025 move immediate case.
27026 * config/aarch64/aarch64.c
27027 (aarch64_const_double_zero_rtx_p): Rename to...
27028 (aarch64_float_const_zero_rtx_p): ...this.
27029 (aarch64_print_operand): Allow printing of new constants.
27030 (aarch64_valid_floating_const): New.
27031 (aarch64_legitimate_constant_p): Check for valid floating-point
27032 constants.
27033 (aarch64_simd_valid_immediate): Likewise.
27034 (aarch64_vect_float_const_representable_p): New.
27035 (aarch64_float_const_representable_p): Likewise.
27036 (aarch64_simd_imm_zero_p): Also allow for floating-point 0.0.
27037 (aarch64_output_simd_mov_immediate): New.
27038 * config/aarch64/aarch64.md (*movsf_aarch64): Add new alternative.
27039 (*movdf_aarch64): Likewise.
27040 * config/aarch64/constraints.md (Ufc): New.
27041 (Y): call aarch64_float_const_zero_rtx.
27042 * config/aarch64/predicates.md (aarch64_fp_compare_operand): New.
27043
e167c04d
RB
270442013-01-07 Richard Biener <rguenther@suse.de>
27045
27046 PR tree-optimization/55888
27047 PR tree-optimization/55862
27048 * tree-ssa-pre.c (phi_translate_1): Revert previous change.
27049 (valid_in_sets): Check if a NAME has a leader in AVAIL_OUT,
27050 not if it is contained therein.
27051
0139aaab
GJL
270522013-01-07 Georg-Johann Lay <avr@gjlay.de>
27053
27054 * config/avr/t-avr: Typo.
27055
4a176b23
GJL
270562013-01-07 Georg-Johann Lay <avr@gjlay.de>
27057
27058 PR55243
27059 * config/avr/t-avr: Don't automatically rebuild
27060 $(srcdir)/config/avr/t-multilib
27061 $(srcdir)/config/avr/avr-tables.opt
27062 $(srcdir)/doc/avr-mmcu.texi
27063 (avr-mcus): New phony target to build them on request.
27064 (s-avr-mlib, s-avr-mmcu-texi): Remove.
27065 * avr/avr-mcus.def: Adjust comments.
27066
c7afdc98
UB
270672013-01-07 Uros Bizjak <ubizjak@gmail.com>
27068
27069 * config/i386/i386.c (DEFAULT_PCC_STRUCT_RETURN): Remove.
27070
1ab05c31
RS
270712013-01-06 Richard Sandiford <rdsandiford@googlemail.com>
27072
27073 * file-find.c, file-find.h, realmpfr.c: Add FSF as copyright holder.
27074
488e3acc
RS
270752013-01-06 Richard Sandiford <rdsandiford@googlemail.com>
27076
27077 * config/tilepro/gen-mul-tables.cc: Put copyright on one line.
27078
a89599d2
DE
270792013-01-05 David Edelsohn <dje.gcc@gmail.com>
27080
27081 * config/rs6000/aix53.h (LIB_SPEC): Add -lpthreads when compiling
27082 to generate profiling.
27083 * config/rs6000/aix64.h (LIB_SPEC): Same.
27084
70f09188
AP
270852013-01-04 Andrew Pinski <apinski@cavium.com>
27086
27087 * config/aarch64/aarch64.c (aarch64_fixed_condition_code_regs):
27088 New function.
27089 (TARGET_FIXED_CONDITION_CODE_REGS): Define.
27090
918d445f
UB
270912013-01-04 Uros Bizjak <ubizjak@gmail.com>
27092
27093 * config/i386/i386.c (ix86_legitimize_address): Call convert_to_mode
27094 unconditionally.
27095 (ix86_expand_move): Ditto.
27096 (ix86_zero_extend_to_Pmode): Ditto.
27097 (ix86_expand_call): Ditto.
27098 (ix86_expand_special_args_builtin): Ditto.
27099 (ix86_expand_builtin): Ditto.
27100
361618ec
RB
271012013-01-04 Richard Biener <rguenther@suse.de>
27102
27103 PR tree-optimization/55862
27104 * tree-ssa-pre.c (phi_translate_1): Valueize SSA names after
27105 translating them through PHI nodes.
27106
87eab554
MJ
271072013-01-04 Martin Jambor <mjambor@suse.cz>
27108
27109 PR tree-optimization/55755
27110 * tree-sra.c (sra_modify_assign): Do not check that an access has no
27111 children when trying to avoid producing a VIEW_CONVERT_EXPR.
27112
33879b9f
MP
271132013-01-04 Marek Polacek <polacek@redhat.com>
27114
27115 PR middle-end/55859
27116 * opts.c (default_options_optimization): Clarify error message.
27117
3068819a
RB
271182013-01-04 Richard Biener <rguenther@suse.de>
27119
27120 PR middle-end/55863
27121 * fold-const.c (split_tree): Undo -X - 1 to ~X folding for
27122 reassociation.
27123
0e4ae794
JDA
271242013-01-03 John David Anglin <dave.anglin@nrc-cnrc.gc.ca>
27125
27126 PR target/53789
27127 * config/pa/pa.md (movsi): Revert previous change.
e58d2e48
JDA
27128 * config/pa/pa.c (pa_legitimate_constant_p): Reject all TLS symbol
27129 references.
0e4ae794 27130
65c88cef
RH
271312013-01-03 Richard Henderson <rth@redhat.com>
27132
27133 * config/i386/i386.c (ix86_expand_move): Always assign to op1
27134 after eliminating TLS symbols.
27135
8ac16127
MG
271362013-01-03 Marc Glisse <marc.glisse@inria.fr>
27137
27138 PR bootstrap/50167
27139 * graphite-interchange.c (pdr_stride_in_loop): Use gmp_fprintf.
27140 * graphite-poly.c (debug_gmp_value): Likewise.
27141
bb664f09
UB
271422013-01-03 Uros Bizjak <ubizjak@gmail.com>
27143
27144 PR target/55712
27145 * config/i386/i386-c.c (ix86_target_macros_internal): Depending on
27146 selected code model, define __code_mode_small__, __code_model_medium__,
27147 __code_model_large__, __code_model_32__ or __code_model_kernel__.
27148 * config/i386/cpuid.h (__cpuid, __cpuid_count) [__i386__]: Prefix
27149 xchg temporary register with %k. Declare temporary register as
27150 early clobbered.
27151 [__x86_64__]: For medium and large code models, preserve %rbx register.
27152
bcf1ef00
RB
271532013-01-03 Richard Biener <rguenther@suse.de>
27154
0506634a 27155 * tree-data-ref.c (dump_conflict_function): Use less vertical spacing.
bcf1ef00
RB
27156 (dump_subscript): Adjust.
27157 (finalize_ddr_dependent): Do not dump redundant info.
27158 (analyze_siv_subscript): Adjust.
27159 (subscript_dependence_tester): Likewise.
27160 (compute_affine_dependence): Likewise.
27161
59fd17e3
RB
271622013-01-03 Richard Biener <rguenther@suse.de>
27163
27164 Revert
27165 2013-01-03 Richard Biener <rguenther@suse.de>
27166
27167 PR tree-optimization/55857
27168 * tree-vect-stmts.c (vectorizable_load): Do not setup
27169 re-alignment for invariant loads.
27170
27171 2013-01-02 Richard Biener <rguenther@suse.de>
27172
27173 * tree-vect-stmts.c (vectorizable_load): When vectorizing an
0506634a 27174 invariant load do not generate a vector load from the scalar location.
59fd17e3 27175
595c2679
RB
271762013-01-03 Richard Biener <rguenther@suse.de>
27177
27178 * tree-vect-loop.c (vect_analyze_loop_form): Clarify reason
27179 for not vectorizing.
27180 * tree-vect-data-refs.c (vect_create_addr_base_for_vector_ref): Do
27181 not build INDIRECT_REFs, call get_name once only.
27182 (vect_create_data_ref_ptr): Likewise. Dump base object kind
27183 based on DR_BASE_OBJECT, not DR_BASE_ADDRESS.
27184
90bb7d7a
RB
271852013-01-03 Richard Biener <rguenther@suse.de>
27186
27187 PR tree-optimization/55857
27188 * tree-vect-stmts.c (vectorizable_load): Do not setup
27189 re-alignment for invariant loads.
27190
f09b77ca
RB
271912013-01-03 Richard Biener <rguenther@suse.de>
27192
27193 PR lto/55848
27194 * lto-symtab.c (lto_symtab_merge_decls_1): As last resort, always
27195 prefer a built-in decl.
27196
3461a16e
JJ
271972013-01-03 Jakub Jelinek <jakub@redhat.com>
27198
df8e2b4f
JJ
27199 * gcc.c (process_command): Update copyright notice dates.
27200 * gcov.c (print_version): Likewise.
27201 * gcov-dump.c (print_version): Likewise.
27202
3461a16e
JJ
27203 PR rtl-optimization/55838
27204 * loop-iv.c (iv_number_of_iterations): Call lowpart_subreg on
27205 iv0.step, iv1.step and step.
27206
8b5546d6
JJ
272072013-01-03 Jakub Jelinek <jakub@redhat.com>
27208 Marc Glisse <marc.glisse@inria.fr>
27209
27210 PR tree-optimization/55832
27211 * fold-const.c (fold_binary_loc): For ABS_EXPR<x> >= 0 and
27212 ABS_EXPR<x> < 0 folding use constant_boolean_node instead of
27213 integer_{one,zero}_node.
27214
8ab1d2e9
JJ
272152013-01-03 Jakub Jelinek <jakub@redhat.com>
27216
27217 PR debug/54402
27218 * params.def (PARAM_MAX_VARTRACK_REVERSE_OP_SIZE): New param.
27219 * var-tracking.c (reverse_op): Don't add reverse ops to
27220 VALUEs that have already
0506634a 27221 PARAM_VALUE (PARAM_MAX_VARTRACK_REVERSE_OP_SIZE) or longer locs list.
8ab1d2e9 27222
5b9ad1d4
GP
272232013-01-02 Gerald Pfeifer <gerald@pfeifer.com>
27224
27225 * doc/contrib.texi: Note years as release manager for Mark Mitchell.
27226
e25a6711
TJ
272272013-01-02 Teresa Johnson <tejohnson@google.com>
27228
27229 * dumpfile.c (dump_loc): Print filename with location.
27230 * tree-ssa-loop-ivcanon.c (try_unroll_loop_completely): Use
27231 new location_t parameter to emit complete unroll message with
27232 new dump framework.
27233 (canonicalize_loop_induction_variables): Compute loops location
27234 and pass to try_unroll_loop_completely.
27235 * loop-unroll.c (report_unroll_peel): New function.
27236 (peel_loops_completely): Use new dump format with location
27237 for main dumpfile message, and invoke report_unroll_peel on success.
27238 (decide_unrolling_and_peeling): Ditto.
27239 (decide_peel_once_rolling): Remove old dumpfile message subsumed
27240 by report_unroll_peel.
27241 (decide_peel_completely): Ditto.
27242 (decide_unroll_constant_iterations): Ditto.
27243 (decide_unroll_runtime_iterations): Ditto.
27244 (decide_peel_simple): Ditto.
27245 (decide_unroll_stupid): Ditto.
27246 * cfgloop.c (get_loop_location): New function.
27247 * cfgloop.h (get_loop_location): Declare.
27248
77878621
ST
272492013-01-02 Sriraman Tallam <tmsriram@google.com>
27250
27251 * config/i386/i386.c (fold_builtin_cpu): Remove unnecessary checks for
27252 NULL.
27253
9e65d03e
JDA
272542013-01-02 John David Anglin <dave.anglin@nrc-cnrc.gc.ca>
27255
27256 PR middle-end/55198
27257 * expr.c (expand_expr_real_1): Don't use bitfield extraction for non
27258 BLKmode objects when EXPAND_MEMORY is specified.
27259
6a7da30f
ST
272602013-01-02 Sriraman Tallam <tmsriram@google.com>
27261
27262 * config/i386/i386.c (ix86_get_function_versions_dispatcher): Fix bug
27263 in loop predicate.
27264 (fold_builtin_cpu): Do not share cpu model decls across statements.
27265
e78167f2
JM
272662013-01-02 Jason Merrill <jason@redhat.com>
27267
27268 PR c++/55804
27269 * tree.c (build_array_type_1): Revert earlier change.
27270
8c075fb4
YZ
272712013-01-02 Yufeng Zhang <yufeng.zhang@arm.com>
27272
27273 * config/aarch64/aarch64-cores.def: Add entries for "cortex-a53" and
27274 "cortex-a57".
27275 * config/aarch64/aarch64-tune.md: Re-generate.
27276
0682ed3e
RB
272772013-01-02 Richard Biener <rguenther@suse.de>
27278
27279 * tree-vect-stmts.c (vectorizable_load): When vectorizing an
0506634a 27280 invariant load do not generate a vector load from the scalar location.
0682ed3e 27281
b8f6a302
RB
272822013-01-02 Richard Biener <rguenther@suse.de>
27283
27284 PR bootstrap/55784
27285 * configure.ac: Add $GMPINC to CFLAGS/CXXFLAGS.
27286 * configure: Regenerate.
27287
04b80dbb
RS
272882013-01-02 Richard Sandiford <rdsandiford@googlemail.com>
27289
27290 * builtins.c (expand_builtin_mathfn, expand_builtin_mathfn_2)
27291 (expand_builtin_mathfn_ternary, expand_builtin_mathfn_3)
27292 (expand_builtin_int_roundingfn_2): Keep the original target around
27293 for the fallback case.
27294
635b0b0c
RS
272952013-01-02 Richard Sandiford <rdsandiford@googlemail.com>
27296
27297 * tree-vrp.c (range_fits_type_p): Require the MSB of the double_int
27298 to be clear for sign changes.
27299
42d57399
JH
273002013-01-01 Jan Hubicka <jh@suse.cz>
27301
27302 * ipa-inline-analysis.c: Fix formatting.
27303
5bb6669d
JJ
273042013-01-01 Jakub Jelinek <jakub@redhat.com>
27305
27306 PR tree-optimization/55831
27307 * tree-vect-loop.c (get_initial_def_for_induction): Use
27308 gsi_after_labels instead of gsi_start_bb.
ad41bd84 27309\f
86a2db33 27310Copyright (C) 2013 Free Software Foundation, Inc.
ad41bd84
JM
27311
27312Copying and distribution of this file, with or without modification,
27313are permitted in any medium without royalty provided the copyright
27314notice and this notice are preserved.