]> git.ipfire.org Git - thirdparty/gcc.git/blame - gcc/ChangeLog
projects / thirdparty / gcc.git / blame
? search:
summary | shortlog | log | commit | commitdiff | tree
blob | blame (incremental) | history | HEAD
re PR libmudflap/58230 (multiple test fail in german language version)
[thirdparty/gcc.git] / gcc / ChangeLog
CommitLineData
1ddbbbc0
MM		 12013-10-17 Michael Meissner <meissner@linux.vnet.ibm.com>
2
3 PR target/58673
4 * config/rs6000/rs6000.c (rs6000_legitimate_address_p): Only
5 restrict TImode addresses to single indirect registers if both
6 -mquad-memory and -mvsx-timode are used.
7 (rs6000_output_move_128bit): Use quad_load_store_p to determine if
8 we should emit load/store quad. Remove using %y for quad memory
9 addresses.
10
11 * config/rs6000/rs6000.md (mov<mode>_ppc64, TI/PTImode): Add
12 constraints to allow load/store quad on machines where TImode is
13 not allowed in VSX registers. Use 'n' instead of 'F' constraint
14 for TImode to load integer constants.
15
f541a481
KT		 162013-10-17 Kyrylo Tkachov <kyrylo.tkachov@arm.com>
17
18 * config/aarch64/aarch64.c (aarch64_print_operand): Handle 'c'.
19
51bb310d
MS		 202013-10-17 Marcus Shawcroft <marcus.shawcroft@arm.com>
21
22 * config/aarch64/aarch64.c (aarch64_preferred_reload_class): Adjust
23 handling of STACK_REG.
24
f4bed77b
RB		 252013-10-17 Richard Biener <rguenther@suse.de>
26
27 PR tree-optimization/58143
28 * tree-ssa-loop-im.c (arith_code_with_undefined_signed_overflow):
29 New function.
30 (rewrite_to_defined_overflow): Likewise.
31 (move_computations_dom_walker::before_dom): Rewrite stmts
32 with undefined signed overflow that are not always executed
33 into unsigned arithmetic.
34
d358fb9d
MM		 352013-10-16 Michael Meissner <meissner@linux.vnet.ibm.com>
36
37 PR target/57756
38 * config/rs6000/rs6000.opt (rs6000_isa_flags_explicit): Move the
39 explicit isa flag to be an options variable, instead of using
40 global_options_set. Remove define from rs6000.h.
41 * config/rs6000/rs6000.h (rs6000_isa_flags_explicit): Likewise.
42
43 * config/rs6000/rs6000.c (rs6000_option_override_internal):
44 Initialize rs6000_isa_flags_explicit.
45 (rs6000_function_specific_save): Add gcc_options* parameter, so
46 that the powerpc builds after the 2013-10-15 changes.
47 (rs6000_function_specific_restore): Likewise.
48
c71aed47
DD		 492013-10-16 DJ Delorie <dj@redhat.com>
50
51 * config/rl78/rl78.c (rl78_alloc_address_registers_macax): Verify
52 op is a REG before checking REGNO.
53 (rl78_alloc_physical_registers): Verify pattern is a SET before
54 checking SET_SRC.
55
90d3bd51
BS		 562013-10-16 Bill Schmidt <wschmidt@linux.vnet.ibm.com>
57
58 * gcc/config/rs6000/vector.md (vec_unpacks_hi_v4sf): Correct for
59 endianness.
60 (vec_unpacks_lo_v4sf): Likewise.
61 (vec_unpacks_float_hi_v4si): Likewise.
62 (vec_unpacks_float_lo_v4si): Likewise.
63 (vec_unpacku_float_hi_v4si): Likewise.
64 (vec_unpacku_float_lo_v4si): Likewise.
65
de75c876
BS		 662013-10-16 Bill Schmidt <wschmidt@linux.vnet.ibm.com>
67
68 * config/rs6000/vsx.md (vsx_concat_<mode>): Adjust output for LE.
69 (vsx_concat_v2sf): Likewise.
70
21216998
JG		 712013-10-16 James Greenhalgh <james.greenhalgh@arm.com>
72
73 * config/aarch64/aarch64.md
74 (*mov<mode>_aarch64): Fix output template for DUP (element) Scalar.
75
1ef88893 762013-10-16 Andrew MacLeod <amacleod@redhat.com>
7ab4168e 77
1ef88893 78 PR tree-optimization/58697
7ab4168e 79 * cfgloop.c (get_estimated_loop_iterations_int): Rename from
1ef88893
AM		 80 estimated_loop_iterations_int.
81 (max_stmt_executions_int): Call get_max_loop_iterations_int.
7ab4168e 82 (get_max_loop_iterations_int): New. HWINT version of
1ef88893
AM		 83 get_max_loop_iterations.
84 * cfgloop.h: Add prototypes.
85 * loop-iv.c (find_simple_exit): call get_estimated_loop_iterations_int.
7ab4168e 86 * loop-unroll.c (decide_peel_once_rolling): Call
1ef88893
AM		 87 get_estimated_loop_iterations_int.
88 * tree-ssa-loop-niter.c (estimated_loop_iterations_int): Add back.
89 * tree-ssa-loop-niter.h: Tweak prototypes.
90
c22df64f
DM		 912013-10-16 David Malcolm <dmalcolm@redhat.com>
92
93 * gengtype-parse.c (struct_field_seq): Ignore access-control
94 keywords ("public:" etc).
95
f24bb080
MS		 962013-10-16 Marcus Shawcroft <marcus.shawcroft@arm.com>
97
98 * config/aarch64/aarch64.c (aarch64_regno_regclass): Classify
99 FRAME_POINTER_REGNUM and ARG_POINTER_REGNUM as POINTER_REGS.
100
c281bf26
YR		 1012013-10-16 Yvan Roux <yvan.roux@linaro.org>
102
103 * config/arm/arm.opt (mlra): New option.
104 * config/arm/arm.c (arm_lra_p): New function.
105 (TARGET_LRA_P): Define.
106
5806f481
PM		 1072013-10-16 Paulo Matos <pmatos@broadcom.com>
108
109 * tree-core.h (tree_code_name): Remove.
110 * tree.h (get_tree_code_name): New prototype.
111 * tree.c (tree_code_name): Make static.
112 (get_tree_code_name): New function.
113 (dump_tree_statistics, tree_check_failed, tree_not_check_failed,
114 tree_class_check_failed, tree_range_check_failed,
115 tree_not_class_check_failed, omp_clause_check_failed,
116 tree_contains_struct_check_failed, tree_operand_check_failed): Use new
117 wrapper get_tree_code_name instead of calling tree_code_name directly.
118 * tree-vrp.c (dump_asserts_for): Likewise.
119 * tree-dump.c (dequeue_and_dump): Likewise.
120 * tree-pretty-print.c (do_niy, dump_generic_node): Likewise.
121 * tree-pretty-print.h (pp_unsupported_tree): Likewise.
122 * lto-streamer-out.c (lto_write_tree, DFS_write_tree): Likewise.
123 * tree-ssa-dom.c (print_expr_hash_elt): Likewise.
124 * gimple-pretty-print.c (dump_unary_rhs, dump_binary_rhs,
125 dump_ternary_rhs, dump_gimple_assign, dump_gimple_cond,
126 dump_gimple_omp_for): Likewise.
127 * tree-vect-data-refs.c (vect_create_data_ref_ptr): Likewise.
128 * tree-ssa-pre.c (print_pre_expr): Likewise.
129 * ipa-prop.c (ipa_print_node_jump_functions_for_edge): Likewise.
130 * print-tree.c (print_node_brief, print_node): Likewise.
131 * gimple.c (gimple_check_failed): Likewise.
132 * lto-streamer.c (lto_tag_name, print_lto_report): Likewise.
133 * config/frv/frv.c (frv_init_cumulative_args): Likewise.
134 * config/mep/mep.c (mep_validate_vliw): Likewise.
135 * config/iq2000/iq2000.c (init_cumulative_args): Likewise.
136 * config/rs6000/rs6000.c (init_cumulative_args): Likewise.
137
82c9df56
GG		 1382013-10-16 Ganesh Gopalasubramanian <Ganesh.Gopalasubramanian@amd.com>
139
140 * config/i386/i386.c (ix86_option_override_internal): Enable FMA4
141 for AMD bdver3.
142
b82d0df9
HPN		 1432013-10-16 Hans-Peter Nilsson <hp@axis.com>
144
145 * config/cris/t-elfmulti (MULTILIB_OPTIONS, MULTILIB_DIRNAMES)
146 (MULTILIB_MATCHES): Add multilib for -march=v8.
147
bf7b5747
ST		 1482013-10-15 Sriraman Tallam <tmsriram@google.com>
149
150 PR target/57756
151 * optc-save-gen.awk: Add extra parameter to the save and restore
152 target calls.
153 * opth-gen.awk: Generate new TARGET_* macros to accept a parameter.
154 * tree.c (build_optimization_node): New parameter. Add extra parameter
155 to call to cl_optimization_save.
156 (build_target_option_node): New parameter. Add extra parameter
157 to call to cl_target_option_save.
158 * tree.h (build_optimization_node): New parameter.
159 (build_target_option_node): New parameter.
160 * c-family/c-common.c (handle_optimize_attribute): Fix calls to
161 build_optimization_node and build_target_option_node.
162 * c-family/c-pragma.c (handle_pragma_optimize): Ditto.
163 (handle_pragma_push_options): Ditto.
164 * toplev.c (process_options): Ditto.
165 * opts.c (init_options_struct): Check for opts_set non-null.
166 * target.def (target_option.save): New parameter.
167 (target_option.restore): New parameter.
168 * tm.texi: Generate.
169 * config/i386/i386-c.c (ix86_target_macros_internal): Ditto.
170 (ix86_pragma_target_parse): Ditto.
171 * config/i386/i386-protos.h (ix86_valid_target_attribute_tree): New
172 parameters.
173 * config/rs6000/rs6000.c (rs6000_option_override_internal): Fix calls
174 to build_optimization_node and build_target_option_node.
175 (rs6000_valid_attribute_p): Ditto.
176 (rs6000_pragma_target_parse): Ditto.
177 * config/i386/i386.opt (x_ix86_target_flags_explicit): New TargetSave
178 data.
179 * config/i386/i386.h:
180 TARGET_64BIT_P: New Macro
181 TARGET_MMX_P: New Macro.
182 TARGET_3DNOW_P: New Macro.
183 TARGET_3DNOW_A_P: New Macro.
184 TARGET_SSE_P: New Macro.
185 TARGET_SSE2_P: New Macro.
186 TARGET_SSE3_P: New Macro.
187 TARGET_SSSE3_P: New Macro.
188 TARGET_SSE4_1_P: New Macro.
189 TARGET_SSE4_2_P: New Macro.
190 TARGET_AVX_P: New Macro.
191 TARGET_AVX2_P: New Macro.
192 TARGET_AVX512F_P: New Macro.
193 TARGET_AVX512PF_P: New Macro.
194 TARGET_AVX512ER_P: New Macro.
195 TARGET_AVX512CD_P: New Macro.
196 TARGET_FMA_P: New Macro.
197 TARGET_SSE4A_P: New Macro.
198 TARGET_FMA4_P: New Macro.
199 TARGET_XOP_P: New Macro.
200 TARGET_LWP_P: New Macro.
201 TARGET_ABM_P: New Macro.
202 TARGET_BMI_P: New Macro.
203 TARGET_BMI2_P: New Macro.
204 TARGET_LZCNT_P: New Macro.
205 TARGET_TBM_P: New Macro.
206 TARGET_POPCNT_P: New Macro.
207 TARGET_SAHF_P: New Macro.
208 TARGET_MOVBE_P: New Macro.
209 TARGET_CRC32_P: New Macro.
210 TARGET_AES_P: New Macro.
211 TARGET_PCLMUL_P: New Macro.
212 TARGET_CMPXCHG16B_P: New Macro.
213 TARGET_FSGSBASE_P: New Macro.
214 TARGET_RDRND_P: New Macro.
215 TARGET_F16C_P: New Macro.
216 TARGET_RTM_P: New Macro.
217 TARGET_HLE_P: New Macro.
218 TARGET_RDSEED_P: New Macro.
219 TARGET_PRFCHW_P: New Macro.
220 TARGET_ADX_P: New Macro.
221 TARGET_FXSR_P: New Macro.
222 TARGET_XSAVE_P: New Macro.
223 TARGET_XSAVEOPT_P: New Macro.
224 TARGET_LP64_P: New Macro.
225 TARGET_X32_P: New Macro.
226 TARGET_FPMATH_DEFAULT_P: New Macro.
227 TARGET_FLOAT_RETURNS_IN_80387_P: New Macro.
228 * config/i386/i386.c (ix86_option_override_internal): New parameters.
229 opts and opts_set.
230 Change ix86_tune_string to access opts->x_ix86_tune_string.
231 Change ix86_isa_flags to access opts->x_ix86_isa_flags.
232 Change ix86_arch_string to access opts->x_ix86_arch_string.
233 Change ix86_stringop_alg to access opts->x_ix86_stringop_alg.
234 Change ix86_pmode to access opts->x_ix86_pmode.
235 Change ix86_abi to access opts->x_ix86_abi.
236 Change ix86_cmodel to access opts->x_ix86_cmodel.
237 Change ix86_asm_dialect to access opts->x_ix86_asm_dialect.
238 Change ix86_isa_flags_explicit to access
239 opts->x_ix86_isa_flags_explicit.
240 Change ix86_dump_tunes to access opts->x_ix86_dump_tunes.
241 Change ix86_regparm to access opts->x_ix86_regparm.
242 Change ix86_branch_cost to access opts->x_ix86_branch_cost.
243 Change ix86_preferred_stack_boundary_arg to access
244 opts->x_ix86_preferred_stack_boundary_arg.
245 Change ix86_force_align_arg_pointer to access
246 opts->x_ix86_force_align_arg_pointer.
247 Change ix86_incoming_stack_boundar_arg to access
248 opts->x_ix86_incoming_stack_boundar_arg.
249 Change ix86_fpmath to access opts->x_ix86_fpmath.
250 Change ix86_veclibabi_type to access opts->x_ix86_veclibabi_type.
251 Change ix86_recip_name to access opts->x_ix86_recip_name.
252 Change ix86_stack_protector_guard to access
253 opts->x_ix86_stack_protector_guard.
254 Change ix86_tune_memcpy_strategy to access
255 opts->x_ix86_tune_memcpy_strategy.
256 Change ix86_tune_memset_strategy to access
257 opts->x_ix86_tune_memset_strategy.
258 Change global_options to access opts.
259 Change global_options_set to access opts_set.
260 Change TARGET_64BIT to TARGET_64BIT_P (opts->...)
261 Change TARGET_MMX to TARGET_MMX_P (opts->...)
262 Change TARGET_3DNOW to TARGET_3DNOW_P (opts->...)
263 Change TARGET_3DNOW_A to TARGET_3DNOW_A_P (opts->...)
264 Change TARGET_SSE to TARGET_SSE_P (opts->...)
265 Change TARGET_SSE2 to TARGET_SSE2_P (opts->...)
266 Change TARGET_SSE3 to TARGET_SSE3_P (opts->...)
267 Change TARGET_SSSE3 to TARGET_SSSE3_P (opts->...)
268 Change TARGET_SSE4_1 to TARGET_SSE4_1_P (opts->...)
269 Change TARGET_SSE4_2 to TARGET_SSE4_2_P (opts->...)
270 Change TARGET_AVX to TARGET_AVX_P (opts->...)
271 Change TARGET_AVX2 to TARGET_AVX2_P (opts->...)
272 Change TARGET_AVX512F to TARGET_AVX512F_P (opts->...)
273 Change TARGET_AVX512PF to TARGET_AVX512PF_P (opts->...)
274 Change TARGET_AVX512ER to TARGET_AVX512ER_P (opts->...)
275 Change TARGET_AVX512CD to TARGET_AVX512CD_P (opts->...)
276 Change TARGET_FMA to TARGET_FMA_P (opts->...)
277 Change TARGET_SSE4A to TARGET_SSE4A_P (opts->...)
278 Change TARGET_FMA4 to TARGET_FMA4_P (opts->...)
279 Change TARGET_XOP to TARGET_XOP_P (opts->...)
280 Change TARGET_LWP to TARGET_LWP_P (opts->...)
281 Change TARGET_ABM to TARGET_ABM_P (opts->...)
282 Change TARGET_BMI to TARGET_BMI_P (opts->...)
283 Change TARGET_BMI2 to TARGET_BMI2_P (opts->...)
284 Change TARGET_LZCNT to TARGET_LZCNT_P (opts->...)
285 Change TARGET_TBM to TARGET_TBM_P (opts->...)
286 Change TARGET_POPCNT to TARGET_POPCNT_P (opts->...)
287 Change TARGET_SAHF to TARGET_SAHF_P (opts->...)
288 Change TARGET_MOVBE to TARGET_MOVBE_P (opts->...)
289 Change TARGET_CRC32 to TARGET_CRC32_P (opts->...)
290 Change TARGET_AES to TARGET_AES_P (opts->...)
291 Change TARGET_PCLMUL to TARGET_PCLMUL_P (opts->...)
292 Change TARGET_CMPXCHG16B to TARGET_CMPXCHG16B_P (opts->...)
293 Change TARGET_FSGSBASE to TARGET_FSGSBASE_P (opts->...)
294 Change TARGET_RDRND to TARGET_RDRND_P (opts->...)
295 Change TARGET_F16C to TARGET_F16C_P (opts->...)
296 Change TARGET_RTM to TARGET_RTM_P (opts->...)
297 Change TARGET_HLE to TARGET_HLE_P (opts->...)
298 Change TARGET_RDSEED to TARGET_RDSEED_P (opts->...)
299 Change TARGET_PRFCHW to TARGET_PRFCHW_P (opts->...)
300 Change TARGET_ADX to TARGET_ADX_P (opts->...)
301 Change TARGET_FXSR to TARGET_FXSR_P (opts->...)
302 Change TARGET_XSAVE to TARGET_XSAVE_P (opts->...)
303 Change TARGET_XSAVEOPT to TARGET_XSAVEOPT_P (opts->...)
304 Change TARGET_LP64 to TARGET_LP64_P (opts->...)
305 Change TARGET_X32 to TARGET_X32_P (opts->...)
306 Change TARGET_FPMATH_DEFAULT to TARGET_FPMATH_DEFAULT_P (opts->...)
307 Change TARGET_FLOAT_RETURNS_IN_80387 to
308 TARGET_FLOAT_RETURNS_IN_80387_P (opts->...)
309 (ix86_function_specific_save): New parameter. Use opts-> fields
310 to replace global fields.
311 (ix86_function_specific_restore): Ditto.
312 (ix86_valid_target_attribute_inner_p): New parameters.
313 Fix recursive call.
314 Fix call to ix86_handle_option and set_option.
315 (ix86_valid_target_attribute_tree): New parameters.
316 Change global_options to access opts.
317 Change global_options_set to access opts_set.
318 Fix call to ix86_valid_target_attribute_inner_p.
319 Change ix86_tune_string to access opts->x_ix86_tune_string.
320 Change ix86_arch_string to access opts->x_ix86_arch_string.
321 Change ix86_fpmath to access opts->x_ix86_fpmath
322 Fix call to ix86_option_override_internal.
323 Fix call to ix86_add_new_builtins.
324 Fix calls to build_optimization_node and build_target_option_node.
325 (ix86_valid_target_attribute_p): Remove access to global_options.
326 Use new gcc_options structure func_options.
327 Fix call to ix86_valid_target_attribute_tree.
328 Fix call to build_optimization_node.
329 (get_builtin_code_for_version): Fix call to
330 ix86_valid_target_attribute_tree.
331
459260ec
DM		 3322013-10-15 David Malcolm <dmalcolm@redhat.com>
333
334 * Makefile.in (PICFLAG): New.
335 (enable_host_shared): New.
336 (INTERNAL_CFLAGS): Use PICFLAG.
337 (LIBIBERTY): Use pic build of libiberty.a if configured with
338 --enable-host-shared.
339 * configure.ac: Add --enable-host-shared, setting up new
340 PICFLAG variable.
341 * configure: Regenerate.
342 * doc/install.texi (--enable-shared): Add note contrasting it
343 with...
344 (--enable-host-shared): New option.
345
16022041
RS		 3462013-10-15 Richard Biener <rguenther@suse.de>
347
348 * tree-tailcall.c (find_tail_calls): Don't use tail-call recursion
349 for built-in functions.
350
b114bfb4
ZC		 3512013-10-15 Zhenqiang Chen <zhenqiang.chen@arm.com>
352
353 * tree-ssa-reassoc.c: Include rtl.h and tm_p.h.
354 (optimize_range_tests_1): New function,
355 extracted from optimize_range_tests.
356 (optimize_range_tests_xor): Similarly.
357 (optimize_range_tests_diff): New function.
358 (optimize_range_tests): Use optimize_range_tests_1.
359
69b8f2f9
CH		 3602013-10-15 Cong Hou <congh@google.com>
361
7ab4168e
MS		 362 * tree-vect-loop.c (vect_is_simple_reduction_1): Relax the
363 requirement of the reduction pattern so that one operand of the
69b8f2f9
CH		 364 reduction operation can come from outside of the loop.
365
8ff5a4e4
JG		 3662013-10-15 James Greenhalgh <james.greenhalgh@arm.com>
367
368 * config/arm/neon-schedgen.ml: Remove.
369 * config/arm/cortex-a9-neon.md: Remove comment regarding
370 neon-schedgen.ml.
371
52b69ad2
JG		 3722013-10-15 James Greenhalgh <james.greenhalgh@arm.com>
373
374 * config/arm/types: Remove old neon types.
375
e7c6a556
JG		 3762013-10-15 James Greenhalgh <james.greenhalgh@arm.com>
377
378 * config/arm/cortex-a7.md
379 (cortex_a7_neon_type): New.
380 (cortex_a7_neon_mul): Update for new types.
381 (cortex_a7_neon_mla): Likewise.
382 (cortex_a7_neon): Likewise.
383
795dd8d5
JG		 3842013-10-15 James Greenhalgh <james.greenhalgh@arm.com>
385
386 * config/arm/cortex-a15-neon.md
387 (cortex_a15_neon_type): New,
388
389 (cortex_a15_neon_int_1): Remove.
390 (cortex_a15_neon_int_2): Likewise.
391 (cortex_a15_neon_int_3): Likewise.
392 (cortex_a15_neon_int_4): Likewise.
393 (cortex_a15_neon_int_5): Likewise.
394 (cortex_a15_neon_vqneg_vqabs): Likewise.
395 (cortex_a15_neon_vmov): Likewise.
396 (cortex_a15_neon_vaba): Likewise.
397 (cortex_a15_neon_vaba_qqq): Likewise.
398 (cortex_a15_neon_mul_ddd_8_16_qdd_16_8_long_32_16_long): Likewise.
399 (cortex_a15_neon_mul_qqq_8_16_32_ddd_32): Likewise.
400 (cortex_a15_neon_mul_qdd_64_32_long_qqd_16_ddd_32_scalar_64_32_long_scalar):
401 Likewise.
402 (cortex_a15_neon_mla_ddd_8_16_qdd_16_8_long_32_16_long): Likewise.
403 (cortex_a15_neon_mla_qqq_8_16): Likewise.
404 (cortex_a15_neon_mla_ddd_32_qqd_16_ddd_32_scalar): Likewise.
405 (cortex_a15_neon_mla_qqq_32_qqd_32_scalar): Likewise.
406 (cortex_a15_neon_mul_ddd_16_scalar_32_16_long_scalar): Likewise.
407 (cortex_a15_neon_mul_qqd_32_scalar): Likewise.
408 (cortex_a15_neon_mla_ddd_16_scalar_qdd_32_16_long_scalar): Likewise.
409 (cortex_a15_neon_shift_1): Likewise.
410 (cortex_a15_neon_shift_2): Likewise.
411 (cortex_a15_neon_shift_3): Likewise.
412 (cortex_a15_neon_vshl_ddd): Likewise.
413 (cortex_a15_neon_vqshl_vrshl_vqrshl_qqq): Likewise.
414 (cortex_a15_neon_vsra_vrsra): Likewise.
415 (cortex_a15_neon_fp_vmla_ddd_scalar): Likewise.
416 (cortex_a15_neon_fp_vmla_qqq_scalar): Likewise.
417 (cortex_a15_neon_bp_3cycle): Likewise.
418 (cortex_a15_neon_ldm_2): Likewise.
419 (cortex_a15_neon_stm_2): Likewise.
420 (cortex_a15_neon_mcr): Likewise.
421 (cortex_a15_neon_mrc): Likewise.
422 (cortex_a15_neon_fp_vadd_ddd_vabs_dd): Likewise.
423 (cortex_a15_neon_fp_vadd_qqq_vabs_qq): Likewise.
424 (cortex_a15_neon_fp_vmul_ddd): Likewise.
425 (cortex_a15_neon_fp_vmul_qqd): Likewise.
426 (cortex_a15_neon_fp_vmla_ddd): Likewise.
427 (cortex_a15_neon_fp_vmla_qqq): Likewise.
428 (cortex_a15_neon_fp_vmla_ddd_scalar): Likewise.
429 (cortex_a15_neon_fp_vmla_qqq_scalar): Likewise.
430 (cortex_a15_neon_fp_vrecps_vrsqrts_ddd): Likewise.
431 (cortex_a15_neon_fp_vrecps_vrsqrts_qqq): Likewise.
432 (cortex_a15_neon_bp_simple): Likewise.
433 (cortex_a15_neon_bp_2cycle): Likewise.
434 (cortex_a15_neon_bp_3cycle): Likewise.
435 (cortex_a15_neon_vld1_1_2_regs): Likewise.
436 (cortex_a15_neon_vld1_3_4_regs): Likewise.
437 (cortex_a15_neon_vld2_2_regs_vld1_vld2_all_lanes): Likewise.
438 (cortex_a15_neon_vld2_4_regs): Likewise.
439 (cortex_a15_neon_vld3_vld4): Likewise.
440 (cortex_a15_neon_vst1_1_2_regs_vst2_2_regs): Likewise.
441 (cortex_a15_neon_vst1_3_4_regs): Likewise.
442 (cortex_a15_neon_vst2_4_regs_vst3_vst4): Rename to...
443 (cortex_a15_neon_vst2_4_regs_vst3): ...This, update for new attributes.
444 (cortex_a15_neon_vst3_vst4): Rename to...
445 (cortex_a15_neon_vst4): This, update for new attributes.
446 (cortex_a15_neon_vld1_vld2_lane): Update for new attributes.
447 (cortex_a15_neon_vld3_vld4_lane): Likewise.
448 (cortex_a15_neon_vst1_vst2_lane): Likewise.
449 (cortex_a15_neon_vst3_vst4_lane): Likewise.
450 (cortex_a15_neon_vld3_vld4_all_lanes): Likewise.
451 (cortex_a15_neon_ldm_2): Likewise.
452 (cortex_a15_neon_stm_2): Likewise.
453 (cortex_a15_neon_mcr): Likewise.
454 (cortex_a15_neon_mcr_2_mcrr): Likewise.
455 (cortex_a15_neon_mrc): Likewise.
456 (cortex_a15_neon_mrrc): Likewise.
457
458 (cortex_a15_neon_abd): New.
459 (cortex_a15_neon_abd_q): Likewise.
460 (cortex_a15_neon_aba): Likewise.
461 (cortex_a15_neon_aba_q): Likewise.
462 (cortex_a15_neon_acc): Likewise.
463 (cortex_a15_neon_acc_q): Likewise.
464 (cortex_a15_neon_arith_basic): Likewise.
465 (cortex_a15_neon_arith_complex): Likewise.
466 (cortex_a15_neon_multiply): Likewise.
467 (cortex_a15_neon_multiply_q): Likewise.
468 (cortex_a15_neon_mla): Likewise.
469 (cortex_a15_neon_mla_q): Likewise.
470 (cortex_a15_neon_sat_mla_long): Likewise.
471 (cortex_a15_neon_shift_acc): Likewise.
472 (cortex_a15_neon_shift_imm_basic): Likewise.
473 (cortex_a15_neon_shift_imm_complex): Likewise.
474 (cortex_a15_neon_shift_reg_basic): Likewise.
475 (cortex_a15_neon_shift_reg_basic_q): Likewise.
476 (cortex_a15_neon_shift_reg_complex): Likewise.
477 (cortex_a15_neon_shift_reg_complex_q): Likewise.
478 (cortex_a15_neon_fp_negabs): Likewise
479 (cortex_a15_neon_fp_arith): Likewise
480 (cortex_a15_neon_fp_arith_q): Likewise
481 (cortex_a15_neon_fp_cvt_int): Likewise
482 (cortex_a15_neon_fp_cvt_int_q): Likewise
483 (cortex_a15_neon_fp_cvt_16): Likewise
484 (cortex_a15_neon_fp_mul): Likewise
485 (cortex_a15_neon_fp_mul_q): Likewise
486 (cortex_a15_neon_fp_mla): Likewise
487 (cortex_a15_neon_fp_mla_q): Likewise
488 (cortex_a15_neon_fp_recps_rsqrte): Likewise.
489 (cortex_a15_neon_fp_recps_rsqrte_q): Likewise.
490 (cortex_a15_neon_bitops): Likewise.
491 (cortex_a15_neon_bitops_q): Likewise.
492 (cortex_a15_neon_from_gp): Likewise.
493 (cortex_a15_neon_from_gp_q): Likewise.
494 (cortex_a15_neon_tbl3_tbl4): Likewise.
495 (cortex_a15_neon_zip_q): Likewise.
496 (cortex_a15_neon_to_gp): Likewise.
497 (cortex_a15_neon_load_a): Likewise.
498 (cortex_a15_neon_load_b): Likewise.
499 (cortex_a15_neon_load_c): Likewise.
500 (cortex_a15_neon_load_d): Likewise.
501 (cortex_a15_neon_load_e): Likewise.
502 (cortex_a15_neon_load_f): Likewise.
503 (cortex_a15_neon_store_a): Likewise.
504 (cortex_a15_neon_store_b): Likewise.
505 (cortex_a15_neon_store_c): Likewise.
506 (cortex_a15_neon_store_d): Likewise.
507 (cortex_a15_neon_store_e): Likewise.
508 (cortex_a15_neon_store_f): Likewise.
509 (cortex_a15_neon_store_g): Likewise.
510 (cortex_a15_neon_store_h): Likewise.
511 (cortex_a15_vfp_to_from_gp): Likewise.
512
bc5e57e3
JG		 5132013-10-15 James Greenhalgh <james.greenhalgh@arm.com>
514
515 * config/arm/cortex-a9-neon.md (cortex_a9_neon_type): New.
516
517 (cortex_a9_neon_vshl_ddd): Remove.
518 (cortex_a9_neon_vst3_vst4): Likewise.
519 (cortex_a9_neon_vld3_vld4_all_lanes): Likewise.
520
521 (cortex_a9_neon_bit_ops_q): New.
522
523 (cortex_a9_neon_int_1): Use cortex_a8_neon_type.
524 (cortex_a9_neon_int_2): Likewise.
525 (cortex_a9_neon_int_3): Likewise.
526 (cortex_a9_neon_int_4): Likewise.
527 (cortex_a9_neon_int_5): Likewise.
528 (cortex_a9_neon_vqneg_vqabs): Likewise.
529 (cortex_a9_neon_vmov): Likewise.
530 (cortex_a9_neon_vaba): Likewise.
531 (cortex_a9_neon_vaba_qqq): Likewise.
532 (cortex_a9_neon_shift_1): Likewise.
533 (cortex_a9_neon_shift_2): Likewise.
534 (cortex_a9_neon_shift_3): Likewise.
535 (cortex_a9_neon_vqshl_vrshl_vqrshl_qqq): Likewise.
536 (cortex_a9_neon_vsra_vrsra): Likewise.
537 (cortex_a9_neon_mul_ddd_8_16_qdd_16_8_long_32_16_long): Likewise.
538 (cortex_a9_neon_mul_qqq_8_16_32_ddd_32): Likewise.
539 (cortex_a9_neon_mul_qdd_64_32_long_qqd_16_ddd_32_scalar_64_32_long_scalar):
540 Likewise.
541 (cortex_a9_neon_mla_ddd_8_16_qdd_16_8_long_32_16_long): Likewise.
542 (cortex_a9_neon_mla_qqq_8_16): Likewise.
543 (cortex_a9_neon_mla_ddd_32_qqd_16_ddd_32_scalar_qdd_64_32_long_scalar_qdd_64_32_long):
544 Likewise.
545 (cortex_a9_neon_mla_qqq_32_qqd_32_scalar): Likewise.
546 (cortex_a9_neon_mul_ddd_16_scalar_32_16_long_scalar): Likewise.
547 (cortex_a9_neon_mul_qqd_32_scalar): Likewise.
548 (cortex_a9_neon_mla_ddd_16_scalar_qdd_32_16_long_scalar): Likewise.
549 (cortex_a9_neon_fp_vadd_ddd_vabs_dd): Likewise.
550 (cortex_a9_neon_fp_vadd_qqq_vabs_qq): Likewise.
551 (cortex_a9_neon_fp_vsum): Likewise.
552 (cortex_a9_neon_fp_vmul_ddd): Likewise.
553 (cortex_a9_neon_fp_vmul_qqd): Likewise.
554 (cortex_a9_neon_fp_vmla_ddd): Likewise.
555 (cortex_a9_neon_fp_vmla_qqq): Likewise.
556 (cortex_a9_neon_fp_vmla_ddd_scalar): Likewise.
557 (cortex_a9_neon_fp_vmla_qqq_scalar): Likewise.
558 (cortex_a9_neon_fp_vrecps_vrsqrts_ddd): Likewise.
559 (cortex_a9_neon_fp_vrecps_vrsqrts_qqq): Likewise.
560 (cortex_a9_neon_bp_simple): Likewise.
561 (cortex_a9_neon_bp_2cycle): Likewise.
562 (cortex_a9_neon_bp_3cycle): Likewise.
563 (cortex_a9_neon_ldr): Likewise.
564 (cortex_a9_neon_str): Likewise.
565 (cortex_a9_neon_vld1_1_2_regs): Likewise.
566 (cortex_a9_neon_vld1_3_4_regs): Likewise.
567 (cortex_a9_neon_vld2_2_regs_vld1_vld2_all_lanes): Likewise.
568 (cortex_a9_neon_vld2_4_regs): Likewise.
569 (cortex_a9_neon_vld3_vld4): Likewise.
570 (cortex_a9_neon_vld1_vld2_lane): Likewise.
571 (cortex_a9_neon_vld3_vld4_lane): Likewise.
572 (cortex_a9_neon_vld3_vld4_all_lanes): Likewise.
573 (cortex_a9_neon_vst1_1_2_regs_vst2_2_regs): Likewise.
574 (cortex_a9_neon_vst1_3_4_regs): Likewise.
575 (cortex_a9_neon_vst2_4_regs_vst3_vst4): Likewise.
576 (cortex_a9_neon_vst1_vst2_lane): Likewise.
577 (cortex_a9_neon_vst3_vst4_lane): Likewise.
578 (cortex_a9_neon_mcr): Likewise.
579 (cortex_a9_neon_mcr_2_mcrr): Likewise.
580 (cortex_a9_neon_mrc): Likewise.
581 (cortex_a9_neon_mrrc): Likewise.
582
9deda8cb
JG		 5832013-10-15 James Greenhalgh <james.greenhalgh@arm.com>
584
585 * config/arm/cortex-a8-neon.md (cortex_a8_neon_type): New.
586
587 (cortex_a8_neon_vshl_ddd): Remove.
588 (cortex_a8_neon_vst3_vst4): Likewise.
589 (cortex_a8_neon_vld3_vld4_all_lanes): Likewise.
590
591 (cortex_a8_neon_bit_ops_q): New.
592
593 (cortex_a8_neon_int_1): Use cortex_a8_neon_type.
594 (cortex_a8_neon_int_2): Likewise..
595 (cortex_a8_neon_int_3): Likewise.
596 (cortex_a8_neon_int_5): Likewise.
597 (cortex_a8_neon_vqneg_vqabs): Likewise.
598 (cortex_a8_neon_int_4): Likewise.
599 (cortex_a8_neon_vaba): Likewise.
600 (cortex_a8_neon_vaba_qqq): Likewise.
601 (cortex_a8_neon_shift_1): Likewise.
602 (cortex_a8_neon_shift_2): Likewise.
603 (cortex_a8_neon_shift_3): Likewise.
604 (cortex_a8_neon_vqshl_vrshl_vqrshl_qqq): Likewise.
605 (cortex_a8_neon_vsra_vrsra): Likewise.
606 (cortex_a8_neon_mul_ddd_8_16_qdd_16_8_long_32_16_long): Likewise.
607 (cortex_a8_neon_mul_qqq_8_16_32_ddd_32): Likewise.
608 (cortex_a8_neon_mul_qdd_64_32_long_qqd_16_ddd_32_scalar_64_32_long_scalar):
609 Likewise.
610 (cortex_a8_neon_mla_ddd_8_16_qdd_16_8_long_32_16_long): Likewise.
611 (cortex_a8_neon_mla_qqq_8_16): Likewise.
612 (cortex_a8_neon_mla_ddd_32_qqd_16_ddd_32_scalar_qdd_64_32_long_scalar_qdd_64_32_long):
613 Likewise.
614 (cortex_a8_neon_mla_qqq_32_qqd_32_scalar): Likewise.
615 (cortex_a8_neon_mul_ddd_16_scalar_32_16_long_scalar): Likewise.
616 (cortex_a8_neon_mul_qqd_32_scalar): Likewise.
617 (cortex_a8_neon_mla_ddd_16_scalar_qdd_32_16_long_scalar): Likewise.
618 (cortex_a8_neon_fp_vadd_ddd_vabs_dd): Likewise.
619 (cortex_a8_neon_fp_vadd_qqq_vabs_qq): Likewise.
620 (cortex_a8_neon_fp_vsum): Likewise.
621 (cortex_a8_neon_fp_vmul_ddd): Likewise.
622 (cortex_a8_neon_fp_vmul_qqd): Likewise.
623 (cortex_a8_neon_fp_vmla_ddd): Likewise.
624 (cortex_a8_neon_fp_vmla_qqq): Likewise.
625 (cortex_a8_neon_fp_vmla_ddd_scalar): Likewise.
626 (cortex_a8_neon_fp_vmla_qqq_scalar): Likewise.
627 (cortex_a8_neon_fp_vrecps_vrsqrts_ddd): Likewise.
628 (cortex_a8_neon_fp_vrecps_vrsqrts_qqq): Likewise.
629 (cortex_a8_neon_bp_simple): Likewise.
630 (cortex_a8_neon_bp_2cycle): Likewise.
631 (cortex_a8_neon_bp_3cycle): Likewise.
632 (cortex_a8_neon_ldr): Likewise.
633 (cortex_a8_neon_str): Likewise.
634 (cortex_a8_neon_vld1_1_2_regs): Likewise.
635 (cortex_a8_neon_vld1_3_4_regs): Likewise.
636 (cortex_a8_neon_vld2_2_regs_vld1_vld2_all_lanes): Likewise.
637 (cortex_a8_neon_vld2_4_regs): Likewise.
638 (cortex_a8_neon_vld3_vld4): Likewise.
639 (cortex_a8_neon_vld1_vld2_lane): Likewise.
640 (cortex_a8_neon_vld3_vld4_lane): Likewise.
641 (cortex_a8_neon_vst1_1_2_regs_vst2_2_regs): Likewise.
642 (cortex_a8_neon_vst1_3_4_regs): Likewise.
643 (cortex_a8_neon_vst2_4_regs_vst3_vst4): Likewise.
644 (cortex_a8_neon_vst1_vst2_lane): Likewise.
645 (cortex_a8_neon_vst3_vst4_lane): Likewise.
646 (cortex_a8_neon_mcr): Likewise.
647 (cortex_a8_neon_mcr_2_mcrr): Likewise.
648 (cortex_a8_neon_mrc): Likewise.
649 (cortex_a8_neon_mrrc): Likewise.
650
0f686aa9
JG		 6512013-10-15 James Greenhalgh <james.greenhalgh@arm.com>
652
653 * config/aarch64/iterators.md (Vetype): Add SF and DF modes.
654 (fp): New.
655 * config/aarch64/aarch64-simd.md (neon_type): Remove.
656 (aarch64_simd_dup<mode>): Add "type" attribute.
657 (aarch64_dup_lane<mode>): Likewise.
658 (aarch64_dup_lane_<vswap_width_name><mode>): Likewise.
659 (*aarch64_simd_mov<mode>): Likewise.
660 (aarch64_simd_mov_from_<mode>low): Likewise.
661 (aarch64_simd_mov_from_<mode>high): Likewise.
662 (orn<mode>3): Likewise.
663 (bic<mode>3): Likewise.
664 (add<mode>3): Likewise.
665 (sub<mode>3): Likewise.
666 (mul<mode>3): Likewise.
667 (*aarch64_mul3_elt<mode>): Likewise.
668 (*aarch64_mul3_elt_<vswap_width_name><mode>): Likewise.
669 (*aarch64_mul3_elt_to_128df): Likewise.
670 (*aarch64_mul3_elt_to_64v2df): Likewise.
671 (neg<mode>2): Likewise.
672 (abs<mode>2): Likewise.
673 (abd<mode>_3): Likewise.
674 (aba<mode>_3): Likewise.
675 (fabd<mode>_3): Likewise.
676 (*fabd_scalar<mode>3): Likewise.
677 (and<mode>3): Likewise.
678 (ior<mode>3): Likewise.
679 (xor<mode>3): Likewise.
680 (one_cmpl<mode>2): Likewise.
681 (aarch64_simd_vec_set<mode>): Likewise.
682 (aarch64_simd_lshr<mode>): Likewise.
683 (aarch64_simd_ashr<mode>): Likewise.
684 (aarch64_simd_imm_shl<mode>): Likewise.
685 (aarch64_simd_reg_sshl<mode): Likewise.
686 (aarch64_simd_reg_shl<mode>_unsigned): Likewise.
687 (aarch64_simd_reg_shl<mode>_signed): Likewise.
688 (aarch64_simd_vec_setv2di): Likewise.
689 (aarch64_simd_vec_set<mode>): Likewise.
690 (aarch64_mla<mode>): Likewise.
691 (*aarch64_mla_elt<mode>): Likewise.
692 (*aarch64_mla_elt_<vswap_width_name><mode>): Likewise.
693 (aarch64_mls<mode>): Likewise.
694 (*aarch64_mls_elt<mode>): Likewise.
695 (*aarch64_mls_elt_<vswap_width_name><mode>): Likewise.
696 (<su><maxmin><mode>3): Likewise.
697 (move_lo_quad_<mode>): Likewise.
698 (aarch64_simd_move_hi_quad_<mode>): Likewise.
699 (aarch64_simd_vec_pack_trunc_<mode>): Likewise.
700 (vec_pack_trunc_<mode>): Likewise.
701 (aarch64_simd_vec_unpack<su>_lo_<mode>): Likewise.
702 (aarch64_simd_vec_unpack<su>_hi_<mode>): Likewise.
703 (*aarch64_<su>mlal_lo<mode>): Likewise.
704 (*aarch64_<su>mlal_hi<mode>): Likewise.
705 (*aarch64_<su>mlsl_lo<mode>): Likewise.
706 (*aarch64_<su>mlsl_hi<mode>): Likewise.
707 (*aarch64_<su>mlal<mode>): Likewise.
708 (*aarch64_<su>mlsl<mode>): Likewise.
709 (aarch64_simd_vec_<su>mult_lo_<mode>): Likewise.
710 (aarch64_simd_vec_<su>mult_hi_<mode>): Likewise.
711 (add<mode>3): Likewise.
712 (sub<mode>3): Likewise.
713 (mul<mode>3): Likewise.
714 (div<mode>3): Likewise.
715 (neg<mode>2): Likewise.
716 (abs<mode>2): Likewise.
717 (fma<mode>4): Likewise.
718 (*aarch64_fma4_elt<mode>): Likewise.
719 (*aarch64_fma4_elt_<vswap_width_name><mode>): Likewise.
720 (*aarch64_fma4_elt_to_128df): Likewise.
721 (*aarch64_fma4_elt_to_64v2df): Likewise.
722 (fnma<mode>4): Likewise.
723 (*aarch64_fnma4_elt<mode>): Likewise.
724 (*aarch64_fnma4_elt_<vswap_width_name><mode>
725 (*aarch64_fnma4_elt_to_128df): Likewise.
726 (*aarch64_fnma4_elt_to_64v2df): Likewise.
727 (<frint_pattern><mode>2): Likewise.
728 (l<fcvt_pattern><su_optab><VDQF:mode><fcvt_target>2): Likewise.
729 (<optab><fcvt_target><VDQF:VDQF:mode>2): Likewise.
730 (vec_unpacks_lo_v4sf): Likewise.
731 (aarch64_float_extend_lo_v2df): Likewise.
732 (vec_unpacks_hi_v4sf): Likewise.
733 (aarch64_float_truncate_lo_v2sf): Likewise.
734 (aarch64_float_truncate_hi_v4sf): Likewise.
735 (aarch64_vmls<mode>): Likewise.
736 (<su><maxmin><mode>3): Likewise.
737 (<maxmin_uns><mode>3): Likewise.
738 (reduc_<sur>plus_<mode>): Likewise.
739 (reduc_<sur>plus_v2di): Likewise.
740 (reduc_<sur>plus_v2si): Likewise.
741 (reduc_<sur>plus_<mode>): Likewise.
742 (aarch64_addpv4sf): Likewise.
743 (clz<mode>2): Likewise.
744 (reduc_<maxmin_uns>_<mode>): Likewise.
745 (reduc_<maxmin_uns>_v2di): Likewise.
746 (reduc_<maxmin_uns>_v2si): Likewise.
747 (reduc_<maxmin_uns>_<mode>): Likewise.
748 (reduc_<maxmin_uns>_v4sf): Likewise.
749 (aarch64_simd_bsl<mode>_internal): Likewise.
750 (*aarch64_get_lane_extend<GPI:mode><VDQQH:mode>): Likewise.
751 (*aarch64_get_lane_zero_extendsi<mode>): Likewise.
752 (aarch64_get_lane<mode>): Likewise.
753 (*aarch64_combinez<mode>): Likewise.
754 (aarch64_combine<mode>): Likewise.
755 (aarch64_simd_combine<mode>): Likewise.
756 (aarch64_<ANY_EXTEND:su><ADDSUB:optab>l<mode>_hi_internal): Likewise.
757 (aarch64_<ANY_EXTEND:su><ADDSUB:optab>l<mode>_lo_internal): Likewise.
758 (aarch64_<ANY_EXTEND:su><ADDSUB:optab>l<mode>): Likewise.
759 (aarch64_<ANY_EXTEND:su><ADDSUB:optab>w<mode>): Likewise.
760 (aarch64_<ANY_EXTEND:su><ADDSUB:optab>w2<mode>_internal): Likewise.
761 (aarch64_<sur>h<addsub><mode>): Likewise.
762 (aarch64_<sur><addsub>hn<mode>): Likewise.
763 (aarch64_<sur><addsub>hn2<mode>): Likewise.
764 (aarch64_pmul<mode>): Likewise.
765 (aarch64_<su_optab><optab><mode>): Likewise.
766 (aarch64_<sur>qadd<mode>): Likewise.
767 (aarch64_sqmovun<mode>): Likewise.
768 (aarch64_<sur>qmovn<mode>): Likewise.
769 (aarch64_s<optab><mode>): Likewise.
770 (aarch64_sq<r>dmulh<mode>): Likewise.
771 (aarch64_sq<r>dmulh_lane<mode>): Likewise.
772 (aarch64_sq<r>dmulh_laneq<mode>): Likewise.
773 (aarch64_sq<r>dmulh_lane<mode>): Likewise.
774 (aarch64_sqdml<SBINQOPS:as>l<mode>): Likewise.
775 (aarch64_sqdml<SBINQOPS:as>l_lane<mode>_internal): Likewise.
776 (aarch64_sqdml<SBINQOPS:as>l_lane<mode>_internal): Likewise.
777 (aarch64_sqdml<SBINQOPS:as>l_n<mode>): Likewise.
778 (aarch64_sqdml<SBINQOPS:as>l2<mode>_internal): Likewise.
779 (aarch64_sqdml<SBINQOPS:as>l2_lane<mode>_internal): Likewise.
780 (aarch64_sqdml<SBINQOPS:as>l2_n<mode>_internal): Likewise.
781 (aarch64_sqdmull<mode>): Likewise.
782 (aarch64_sqdmull_lane<mode>_internal): Likewise.
783 (aarch64_sqdmull_n<mode>): Likewise.
784 (aarch64_sqdmull2<mode>_internal): Likewise.
785 (aarch64_sqdmull2_lane<mode>_internal): Likewise.
786 (aarch64_sqdmull2_n<mode>_internal): Likewise.
787 (aarch64_<sur>shl<mode>): Likewise.
788 (aarch64_<sur>q<r>shl<mode>
789 (aarch64_<sur>shll_n<mode>): Likewise.
790 (aarch64_<sur>shll2_n<mode>): Likewise.
791 (aarch64_<sur>shr_n<mode>): Likewise.
792 (aarch64_<sur>sra_n<mode>): Likewise.
793 (aarch64_<sur>s<lr>i_n<mode>): Likewise.
794 (aarch64_<sur>qshl<u>_n<mode>): Likewise.
795 (aarch64_<sur>q<r>shr<u>n_n<mode>): Likewise.
796 (aarch64_cm<optab><mode>): Likewise.
797 (aarch64_cm<optab>di): Likewise.
798 (aarch64_cm<optab><mode>): Likewise.
799 (aarch64_cm<optab>di): Likewise.
800 (aarch64_cmtst<mode>): Likewise.
801 (aarch64_cmtstdi): Likewise.
802 (aarch64_cm<optab><mode>): Likewise.
803 (*aarch64_fac<optab><mode>): Likewise.
804 (aarch64_addp<mode>): Likewise.
805 (aarch64_addpdi): Likewise.
806 (sqrt<mode>2): Likewise.
807 (vec_load_lanesoi<mode>): Likewise.
808 (vec_store_lanesoi<mode>): Likewise.
809 (vec_load_lanesci<mode>): Likewise.
810 (vec_store_lanesci<mode>): Likewise.
811 (vec_load_lanesxi<mode>): Likewise.
812 (vec_store_lanesxi<mode>): Likewise.
813 (*aarch64_mov<mode>): Likewise.
814 (aarch64_ld2<mode>_dreg): Likewise.
815 (aarch64_ld2<mode>_dreg): Likewise.
816 (aarch64_ld3<mode>_dreg): Likewise.
817 (aarch64_ld3<mode>_dreg): Likewise.
818 (aarch64_ld4<mode>_dreg): Likewise.
819 (aarch64_ld4<mode>_dreg): Likewise.
820 (aarch64_tbl1<mode>): Likewise.
821 (aarch64_tbl2v16qi): Likewise.
822 (aarch64_combinev16qi): Likewise.
823 (aarch64_<PERMUTE:perm_insn><PERMUTE:perm_hilo><mode>): Likewise.
824 (aarch64_st2<mode>_dreg): Likewise.
825 (aarch64_st2<mode>_dreg): Likewise.
826 (aarch64_st3<mode>_dreg): Likewise.
827 (aarch64_st3<mode>_dreg): Likewise.
828 (aarch64_st4<mode>_dreg): Likewise.
829 (aarch64_st4<mode>_dreg): Likewise.
830 (*aarch64_simd_ld1r<mode>): Likewise.
831 (aarch64_frecpe<mode>): Likewise.
832 (aarch64_frecp<FRECP:frecp_suffix><mode>): Likewise.
833 (aarch64_frecps<mode>): Likewise.
834
f7379e5e
JG		 8352013-10-15 James Greenhalgh <james.greenhalgh@arm.com>
836
837 * config/arm/iterators.md (V_elem_ch): New.
838 (q): Likewise.
839 (VQH_type): Likewise.
840 * config/arm/arm.md (is_neon_type): New.
841 (conds): Use is_neon_type.
842 (anddi3_insn): Update type attribute.
843 (xordi3_insn): Likewise.
844 (one_cmpldi2): Likewise.
845 * gcc/config/arm/vfp.md (movhf_vfp_neon): Update type attribute.
846 * gcc/config/arm/neon.md (neon_mov): Update type attribute.
847 (*movmisalign<mode>_neon_store): Likewise.
848 (*movmisalign<mode>_neon_load): Likewise.
849 (vec_set<mode>_internal): Likewise.
850 (vec_set<mode>_internal): Likewise.
851 (vec_setv2di_internal): Likewise.
852 (vec_extract<mode>): Likewise.
853 (vec_extract<mode>): Likewise.
854 (vec_extractv2di): Likewise.
855 (*add<mode>3_neon): Likewise.
856 (adddi3_neon): Likewise.
857 (*sub<mode>3_neon): Likewise.
858 (subdi3_neon): Likewise.
859 (fma<VCVTF:mode>4): Likewise.
860 (fma<VCVTF:mode>4_intrinsic): Likewise.
861 (*fmsub<VCVTF:mode>4): Likewise.
862 (fmsub<VCVTF:mode>4_intrinsic): Likewise.
863 (neon_vrint<NEON_VRINT:nvrint_variant><VCVTF:mode>): Likewise.
864 (ior<mode>3): Likewise.
865 (and<mode>3): Likewise.
866 (orn<mode>3_neon): Likewise.
867 (orndi3_neon): Likewise.
868 (bic<mode>3_neon): Likewise.
869 (bicdi3_neon): Likewise.
870 (xor<mode>3): Likewise.
871 (one_cmpl<mode>2): Likewise.
872 (abs<mode>2): Likewise.
873 (neg<mode>2): Likewise.
874 (negdi2_neon): Likewise.
875 (*umin<mode>3_neon): Likewise.
876 (*umax<mode>3_neon): Likewise.
877 (*smin<mode>3_neon): Likewise.
878 (*smax<mode>3_neon): Likewise.
879 (vashl<mode>3): Likewise.
880 (vashr<mode>3_imm): Likewise.
881 (vlshr<mode>3_imm): Likewise.
882 (ashl<mode>3_signed): Likewise.
883 (ashl<mode>3_unsigned): Likewise.
884 (neon_load_count): Likewise.
885 (ashldi3_neon_noclobber): Likewise.
886 (ashldi3_neon): Likewise.
887 (signed_shift_di3_neon): Likewise.
888 (unsigned_shift_di3_neon): Likewise.
889 (ashrdi3_neon_imm_noclobber): Likewise.
890 (lshrdi3_neon_imm_noclobber): Likewise.
891 (<shift>di3_neon): Likewise.
892 (widen_ssum<mode>3): Likewise.
893 (widen_usum<mode>3): Likewise.
894 (quad_halves_<code>v4si): Likewise.
895 (quad_halves_<code>v4sf): Likewise.
896 (quad_halves_<code>v8hi): Likewise.
897 (quad_halves_<code>v16qi): Likewise.
898 (reduc_splus_v2di): Likewise.
899 (neon_vpadd_internal<mode>): Likewise.
900 (neon_vpsmin<mode>): Likewise.
901 (neon_vpsmax<mode>): Likewise.
902 (neon_vpumin<mode>): Likewise.
903 (neon_vpumax<mode>): Likewise.
904 (*ss_add<mode>_neon): Likewise.
905 (*us_add<mode>_neon): Likewise.
906 (*ss_sub<mode>_neon): Likewise.
907 (*us_sub<mode>_neon): Likewise.
908 (neon_vadd<mode>_unspec): Likewise.
909 (neon_vaddl<mode>): Likewise.
910 (neon_vaddw<mode>): Likewise.
911 (neon_vhadd<mode>): Likewise.
912 (neon_vqadd<mode>): Likewise.
913 (neon_vaddhn<mode>): Likewise.
914 (neon_vmul<mode>): Likewise.
915 (neon_vfms<VCVTF:mode>): Likewise.
916 (neon_vmlal<mode>): Likewise.
917 (neon_vmls<mode>): Likewise.
918 (neon_vmlsl<mode>): Likewise.
919 (neon_vqdmulh<mode>): Likewise.
920 (neon_vqdmlal<mode>): Likewise.
921 (neon_vqdmlsl<mode>): Likewise.
922 (neon_vmull<mode>): Likewise.
923 (neon_vqdmull<mode>): Likewise.
924 (neon_vsub<mode>_unspec): Likewise.
925 (neon_vsubl<mode>): Likewise.
926 (neon_vsubw<mode>): Likewise.
927 (neon_vqsub<mode>): Likewise.
928 (neon_vhsub<mode>): Likewise.
929 (neon_vsubhn<mode>): Likewise.
930 (neon_vceq<mode>): Likewise.
931 (neon_vcge<mode>): Likewise.
932 (neon_vcgeu<mode>): Likewise.
933 (neon_vcgt<mode>): Likewise.
934 (neon_vcgtu<mode>): Likewise.
935 (neon_vcle<mode>): Likewise.
936 (neon_vclt<mode>): Likewise.
937 (neon_vcage<mode>): Likewise.
938 (neon_vcagt<mode>): Likewise.
939 (neon_vtst<mode>): Likewise.
940 (neon_vabd<mode>): Likewise.
941 (neon_vabdl<mode>): Likewise.
942 (neon_vaba<mode>): Likewise.
943 (neon_vabal<mode>): Likewise.
944 (neon_vmax<mode>): Likewise.
945 (neon_vmin<mode>): Likewise.
946 (neon_vpaddl<mode>): Likewise.
947 (neon_vpadal<mode>): Likewise.
948 (neon_vpmax<mode>): Likewise.
949 (neon_vpmin<mode>): Likewise.
950 (neon_vrecps<mode>): Likewise.
951 (neon_vrsqrts<mode>): Likewise.
952 (neon_vqabs<mode>): Likewise.
953 (neon_vqneg<mode>): Likewise.
954 (neon_vcls<mode>): Likewise.
955 (clz<mode>2): Likewise.
956 (popcount<mode>2): Likewise.
957 (neon_vrecpe<mode>): Likewise.
958 (neon_vrsqrte<mode>): Likewise.
959 (neon_vget_lane<mode>_sext_internal): Likewise.
960 (neon_vget_lane<mode>_zext_internal): Likewise.
961 (neon_vdup_n<mode>): Likewise.
962 (neon_vdup_n<mode>): Likewise.
963 (neon_vdup_nv2di): Likewise.
964 (neon_vdup_lane<mode>_interal): Likewise.
965 (*neon_vswp<mode>): Likewise.
966 (neon_vcombine<mode>): Likewise.
967 (float<mode><V_cvtto>2): Likewise.
968 (floatuns<mode><V_cvtto>2): Likewise.
969 (fix_trunc<mode><V_cvtto>2): Likewise.
970 (fixuns_trunc<mode><V_cvtto>2
971 (neon_vcvt<mode>): Likewise.
972 (neon_vcvt<mode>): Likewise.
973 (neon_vcvtv4sfv4hf): Likewise.
974 (neon_vcvtv4hfv4sf): Likewise.
975 (neon_vcvt_n<mode>): Likewise.
976 (neon_vcvt_n<mode>): Likewise.
977 (neon_vmovn<mode>): Likewise.
978 (neon_vqmovn<mode>): Likewise.
979 (neon_vqmovun<mode>): Likewise.
980 (neon_vmovl<mode>): Likewise.
981 (neon_vmul_lane<mode>): Likewise.
982 (neon_vmul_lane<mode>): Likewise.
983 (neon_vmull_lane<mode>): Likewise.
984 (neon_vqdmull_lane<mode>): Likewise.
985 (neon_vqdmulh_lane<mode>): Likewise.
986 (neon_vqdmulh_lane<mode>): Likewise.
987 (neon_vmla_lane<mode>): Likewise.
988 (neon_vmla_lane<mode>): Likewise.
989 (neon_vmlal_lane<mode>): Likewise.
990 (neon_vqdmlal_lane<mode>): Likewise.
991 (neon_vmls_lane<mode>): Likewise.
992 (neon_vmls_lane<mode>): Likewise.
993 (neon_vmlsl_lane<mode>): Likewise.
994 (neon_vqdmlsl_lane<mode>): Likewise.
995 (neon_vext<mode>): Likewise.
996 (neon_vrev64<mode>): Likewise.
997 (neon_vrev32<mode>): Likewise.
998 (neon_vrev16<mode>): Likewise.
999 (neon_vbsl<mode>_internal): Likewise.
1000 (neon_vshl<mode>): Likewise.
1001 (neon_vqshl<mode>): Likewise.
1002 (neon_vshr_n<mode>): Likewise.
1003 (neon_vshrn_n<mode>): Likewise.
1004 (neon_vqshrn_n<mode>): Likewise.
1005 (neon_vqshrun_n<mode>): Likewise.
1006 (neon_vshl_n<mode>): Likewise.
1007 (neon_vqshl_n<mode>): Likewise.
1008 (neon_vqshlu_n<mode>): Likewise.
1009 (neon_vshll_n<mode>): Likewise.
1010 (neon_vsra_n<mode>): Likewise.
1011 (neon_vsri_n<mode>): Likewise.
1012 (neon_vsli_n<mode>): Likewise.
1013 (neon_vtbl1v8qi): Likewise.
1014 (neon_vtbl2v8qi): Likewise.
1015 (neon_vtbl3v8qi): Likewise.
1016 (neon_vtbl4v8qi): Likewise.
1017 (neon_vtbl1v16qi): Likewise.
1018 (neon_vtbl2v16qi): Likewise.
1019 (neon_vcombinev16qi): Likewise.
1020 (neon_vtbx1v8qi): Likewise.
1021 (neon_vtbx2v8qi): Likewise.
1022 (neon_vtbx3v8qi): Likewise.
1023 (neon_vtbx4v8qi): Likewise.
1024 (*neon_vtrn<mode>_insn): Likewise.
1025 (*neon_vzip<mode>_insn): Likewise.
1026 (*neon_vuzp<mode>_insn): Likewise.
1027 (neon_vld1<mode>): Likewise.
1028 (neon_vld1_lane<mode>): Likewise.
1029 (neon_vld1_lane<mode>): Likewise.
1030 (neon_vld1_dup<mode>): Likewise.
1031 (neon_vld1_dup<mode>): Likewise.
1032 (neon_vld1_dupv2di): Likewise.
1033 (neon_vst1<mode>): Likewise.
1034 (neon_vst1_lane<mode>): Likewise.
1035 (neon_vst1_lane<mode>): Likewise.
1036 (neon_vld2<mode>): Likewise.
1037 (neon_vld2<mode>): Likewise.
1038 (neon_vld2_lane<mode>): Likewise.
1039 (neon_vld2_lane<mode>): Likewise.
1040 (neon_vld2_dup<mode>): Likewise.
1041 (neon_vst2<mode>): Likewise.
1042 (neon_vst2<mode>): Likewise.
1043 (neon_vst2_lane<mode>): Likewise.
1044 (neon_vst2_lane<mode>): Likewise.
1045 (neon_vld3<mode>): Likewise.
1046 (neon_vld3qa<mode>): Likewise.
1047 (neon_vld3qb<mode>): Likewise.
1048 (neon_vld3_lane<mode>): Likewise.
1049 (neon_vld3_lane<mode>): Likewise.
1050 (neon_vld3_dup<mode>): Likewise.
1051 (neon_vst3<mode>): Likewise.
1052 (neon_vst3qa<mode>): Likewise.
1053 (neon_vst3qb<mode>): Likewise.
1054 (neon_vst3_lane<mode>): Likewise.
1055 (neon_vst3_lane<mode>): Likewise.
1056 (neon_vld4<mode>): Likewise.
1057 (neon_vld4qa<mode>): Likewise.
1058 (neon_vld4qb<mode>): Likewise.
1059 (neon_vld4_lane<mode>): Likewise.
1060 (neon_vld4_lane<mode>): Likewise.
1061 (neon_vld4_dup<mode>): Likewise.
1062 (neon_vst4<mode>): Likewise.
1063 (neon_vst4qa<mode>): Likewise.
1064 (neon_vst4qb<mode>): Likewise.
1065 (neon_vst4_lane<mode>): Likewise.
1066 (neon_vst4_lane<mode>): Likewise.
1067 (neon_vec_unpack<US>_lo_<mode>): Likewise.
1068 (neon_vec_unpack<US>_hi_<mode>): Likewise.
1069 (neon_vec_<US>mult_lo_<mode>): Likewise.
1070 (neon_vec_<US>mult_hi_<mode>): Likewise.
1071 (neon_vec_<US>shiftl_<mode>): Likewise.
1072 (neon_unpack<US>_<mode>): Likewise.
1073 (neon_vec_<US>mult_<mode>): Likewise.
1074 (vec_pack_trunc_<mode>): Likewise.
1075 (neon_vec_pack_trunc_<mode>): Likewise.
1076 (neon_vabd<mode>_2): Likewise.
1077 (neon_vabd<mode>_3): Likewise.
1078
a9e66678
JG		 10792013-10-15 James Greenhalgh <james.greenhalgh@arm.com>
1080
1081 * config/aarch64/aarch64.md (movtf_aarch64): Update type attribute.
1082 (load_pair): Update type attribute.
1083 (store_pair): Update type attribute.
1084 * config/aarch64/iterators.md (q): New.
1085
20445ed6
JG		 10862013-10-15 James Greenhalgh <james.greenhalgh@arm.com>
1087
1088 * config/arm/types.md: Add new types for Neon insns.
1089
afb4ac68
AI		 10902013-10-15 Alexander Ivchenko <alexander.ivchenko@intel.com>
1091 Maxim Kuznetsov <maxim.kuznetsov@intel.com>
1092 Sergey Lega <sergey.s.lega@intel.com>
1093 Anna Tikhonova <anna.tikhonova@intel.com>
1094 Ilya Tocar <ilya.tocar@intel.com>
1095 Andrey Turetskiy <andrey.turetskiy@intel.com>
1096 Ilya Verbin <ilya.verbin@intel.com>
1097 Kirill Yukhin <kirill.yukhin@intel.com>
1098 Michael Zolotukhin <michael.v.zolotukhin@intel.com>
1099
1100 * config/i386/sse.md (unspec): Add UNSPEC_RCP14, UNSPEC_RSQRT14,
1101 UNSPEC_FIXUPIMM, UNSPEC_SCALEF, UNSPEC_GETEXP, UNSPEC_GETMANT,
1102 UNSPEC_EXP2, UNSPEC_RCP28, UNSPEC_RSQRT28.
1103 (rcp14<mode>): New.
1104 (srcp14<mode>): Ditto.
1105 (rsqrt14<mode>): Ditto.
1106 (rsqrt14<mode>): Ditto.
1107 (avx512f_vmscalef<mode>): Ditto.
1108 (avx512f_scalef<mode>): Ditto.
1109 (avx512f_getexp<mode>): Ditto.
1110 (avx512f_sgetexp<mode>): Ditto.
1111 (avx512f_fixupimm<mode>): Ditto.
1112 (avx512f_sfixupimm<mode>): Ditto.
1113 (avx512f_rndscale<mode>): Ditto.
1114 (*avx512er_exp2<mode>): Ditto.
1115 (*avx512er_rcp28<mode>): Ditto.
1116 (avx512er_rsqrt28<mode>): Ditto.
1117 (avx512f_getmant<mode>): Ditto.
1118 (avx512f_getmant<mode>): Ditto.
1119 (avx512f_rndscale<mode>): Fix formatting.
1120
a1b499a3
JL		 11212013-10-15 Martin Jambor <mjambor@suse.cz>
1122
1123 * ipa-utils.h (ipa_edge_within_scc): Declare.
1124 * ipa-cp.c (edge_within_scc): Moved...
1125 * ipa-utils.c (ipa_edge_within_scc): ...here. Updated all callers.
afb4ac68 1126
2e2206fa
AI		 11272013-10-15 Alexander Ivchenko <alexander.ivchenko@intel.com>
1128 Maxim Kuznetsov <maxim.kuznetsov@intel.com>
1129 Sergey Lega <sergey.s.lega@intel.com>
1130 Anna Tikhonova <anna.tikhonova@intel.com>
1131 Ilya Tocar <ilya.tocar@intel.com>
1132 Andrey Turetskiy <andrey.turetskiy@intel.com>
1133 Ilya Verbin <ilya.verbin@intel.com>
1134 Kirill Yukhin <kirill.yukhin@intel.com>
1135 Michael Zolotukhin <michael.v.zolotukhin@intel.com>
1136
1137 * config/i386/predicates.md (const_8_to_15_operand): New.
1138 (const_16_to_31_operand): Ditto.
1139 * config/i386/sse.md (V8FI): New.
1140 (V16FI): Ditto.
1141 (reduc_splus_v8df): Ditto.
1142 (reduc_splus_v16sf): Ditto.
1143 (avx512f_vextract<shuffletype>32x4_1): Ditto.
1144 (vec_extract_hi_<mode>): Ditto.
1145 (avx512f_vinsert<shuffletype>32x4_1): Ditto.
1146 (vec_set_lo_<mode>): Ditto.
1147 (vec_set_hi_<mode>): Ditto.
1148 (avx512f_shuf_<shuffletype>64x2_1): Ditto.
1149 (avx512f_shuf_<shuffletype>32x4_1): Ditto.
1150 (avx512f_pshufd_1): Ditto.
1151 (avx512f_broadcast<mode>): Ditto.
1152 (avx512f_broadcast<mode>): Ditto.
1153 (define_split): Split vec_extract_lo into move.
1154 (ssequartermode): Ditto.
1155 (ssedoublemode): Extened with wider modes.
1156 (vec_extract_lo_<mode>): Ditto.
1157
0fe65b75
AI		 11582013-10-15 Alexander Ivchenko <alexander.ivchenko@intel.com>
1159 Maxim Kuznetsov <maxim.kuznetsov@intel.com>
1160 Sergey Lega <sergey.s.lega@intel.com>
1161 Anna Tikhonova <anna.tikhonova@intel.com>
1162 Ilya Tocar <ilya.tocar@intel.com>
1163 Andrey Turetskiy <andrey.turetskiy@intel.com>
1164 Ilya Verbin <ilya.verbin@intel.com>
1165 Kirill Yukhin <kirill.yukhin@intel.com>
1166 Michael Zolotukhin <michael.v.zolotukhin@intel.com>
1167
1168 * config/i386/predicates.md (register_or_constm1_operand): New.
1169 * config/i386/sse.md (unspec): Add UNSPEC_UNSIGNED_PCMP, UNSPEC_TESTM,
1170 UNSPEC_TESTNM, UNSPEC_VTERNLOG, UNSPEC_ALIGN, UNSPEC_CONFLICT,
1171 UNSPEC_MASKED_EQ, UNSPEC_MASKED_GT, UNSPEC_GATHER_PREFETCH,
1172 UNSPEC_SCATTER_PREFETCH
1173 (VI48_512): New.
1174 (avx512f_ucmp<mode>3): Ditto.
1175 (avx512f_vternlog<mode>): Ditto.
1176 (avx512f_align<mode>): Ditto.
1177 (<shift_insn><mode>3): Ditto.
1178 (avx512f_<rotate>v<mode>): Ditto.
1179 (avx512f_<rotate><mode>): Ditto.
1180 (avx512f_eq<mode>3): Ditto.
1181 (avx512f_eq<mode>3_1): Ditto.
1182 (avx512f_gt<mode>3): Ditto.
1183 (avx512f_testm<mode>3): Ditto.
1184 (avx512f_testnm<mode>3): Ditto.
1185 (avx512pf_gatherpf<mode>): Ditto.
1186 (*avx512pf_gatherpf<mode>_mask): Ditto.
1187 (*avx512pf_gatherpf<mode>): Ditto.
1188 (avx512pf_scatterpf<mode>): Ditto.
1189 (*avx512pf_scatterpf<mode>_mask): Ditto.
1190 (*avx512pf_scatterpf<mode>): Ditto.
1191 (avx512f_vec_dup_gpr<mode>): Ditto.
1192 (clz<mode>2): Ditto.
1193 (conflict<mode>): Ditto.
1194 (REDUC_SMINMAX_MODE): Extened with wider modes.
1195 (reduc_<code>_<mode>): Ditto.
1196 (vlshr<mode>3): Ditto.
1197 (vashl<mode>3): Ditto.
1198
ab931c71
AI		 11992013-10-15 Alexander Ivchenko <alexander.ivchenko@intel.com>
1200 Maxim Kuznetsov <maxim.kuznetsov@intel.com>
1201 Sergey Lega <sergey.s.lega@intel.com>
1202 Anna Tikhonova <anna.tikhonova@intel.com>
1203 Ilya Tocar <ilya.tocar@intel.com>
1204 Andrey Turetskiy <andrey.turetskiy@intel.com>
1205 Ilya Verbin <ilya.verbin@intel.com>
1206 Kirill Yukhin <kirill.yukhin@intel.com>
1207 Michael Zolotukhin <michael.v.zolotukhin@intel.com>
1208
1209 * config/i386/sse.md (unspec): Added UNSPEC_VPERMI2, UNSPEC_VPERMT2,
1210 UNSPEC_SCATTER.
1211 (VI48F_512): New.
1212 (avx512fmaskmode): Ditto.
1213 (bcstscalarsuff): Ditto.
1214 (avx512f_blendm<mode>): Ditto.
1215 (cmp_imm_predicate): Ditto.
1216 (avx512f_cmp<mode>3): Ditto.
1217 (avx512f_vec_dup<mode>): Ditto.
1218 (avx512f_vec_dup_mem<mode>): Ditto.
1219 (avx512f_vpermi2var<mode>3): Ditto.
1220 (avx512f_vpermt2var<mode>3): Ditto.
1221 (vec_init<mode>): Ditto.
1222 (avx512f_gathersi<mode>): Ditto.
1223 (*avx512f_gathersi<mode>): Ditto.
1224 (*avx512f_gathersi<mode>_2): Ditto.
1225 (avx512f_gatherdi<mode>): Ditto.
1226 (*avx512f_gatherdi<mode>): Ditto.
1227 (*avx512f_gatherdi<mode>_2): Ditto.
1228 (avx512f_scattersi<mode>): Ditto.
1229 (*avx512f_scattersi<mode>): Ditto.
1230 (avx512f_scatterdi<mode>): Ditto.
1231 (*avx512f_scatterdi<mode>): Ditto.
1232 (sseintprefix): Extened with wider modes.
1233 (VEC_GATHER_IDXSI): Ditto.
1234 (VEC_GATHER_IDXDI): Ditto.
1235 (VEC_GATHER_SRCDI): Ditto.
1236
c9acb877
MGD		 12372013-10-15 Matthew Gretton-Dann <matthew.gretton-dann@arm.com>
1238 Ramana Radhakrishnan <ramana.radhakrishnan@arm.com>
1239
1240 * config/arm/t-aprofile: New file.
1241 * config.gcc: Handle --with-multilib-list option.
1242
ecbda484
BS		 12432013-10-15 Bernd Schmidt <bernds@codesourcery.com>
1244
1245 * reload1.c (reloads_unique_chain_p): Ensure that r1 is the input for
1246 r2.
1247
24f161fd
RB		 12482013-10-15 Richard Biener <rguenther@suse.de>
1249
1250 * tree-loop-distribution.c (build_empty_rdg): Inline into
1251 single user.
1252 (rdg_flag_vertex): Inline into single user.
1253 (rdg_flag_vertex_and_dependent): Likewise.
1254 (build_rdg_partition_for_vertex): Remove processed bitmap.
1255 (rdg_build_partitions): Simplify.
1256
5de989ed
RB		 12572013-10-15 Richard Biener <rguenther@suse.de>
1258
1259 * tree-ssa-forwprop.c (forward_propagate_addr_expr_1):
1260 Restructure forwarding through conversions and copies to
1261 avoid performing copy-propagation the wrong way. Adjust
1262 recursion invocations.
1263 (forward_propagate_addr_expr): Add argument stating if we
1264 are recursing from a single-use.
1265 (ssa_forward_propagate_and_combine): Adjust.
1266
47e0da37
DM		 12672013-10-14 David Malcolm <dmalcolm@redhat.com>
1268
1269 * dumpfile.h (gcc::dump_manager): New class, to hold state
1270 relating to dumpfile management.
1271 (get_dump_file_name): Remove in favor of method of dump_manager.
1272 (dump_initialized_p): Likewise.
1273 (dump_start): Likewise.
1274 (dump_finish): Likewise.
1275 (dump_switch_p): Likewise.
1276 (dump_register): Likewise.
1277 (get_dump_file_info): Likewise.
1278 * context.c (gcc::context::context): Construct the dump_manager
1279 instance.
1280 * context.h (gcc::context::get_dumps): New.
1281 (gcc::context::m_dumps): New.
1282 * coverage.c (coverage_init): Port to dump_manager API.
1283 * dumpfile.c (extra_dump_files): Convert to field of
1284 gcc::dump_manager.
1285 (extra_dump_files_in_use): Likewise.
1286 (extra_dump_files_alloced): Likewise.
1287 (gcc::dump_manager::dump_manager): New.
1288 (dump_register): Convert to...
1289 (gcc::dump_manager::dump_register): ...method, replacing
1290 function-static next_dump with m_next_dump field.
1291 (get_dump_file_info): Convert to...
1292 (gcc::dump_manager::get_dump_file_info): ...method.
1293 (get_dump_file_name): Convert to...
1294 (gcc::dump_manager::get_dump_file_name): ...method.
1295 (dump_start): Convert to...
1296 (gcc::dump_manager::dump_start): ...method.
1297 (dump_finish): Convert to...
1298 (gcc::dump_manager::dump_finish): ...method.
1299 (dump_begin): Replace body with...
1300 (gcc::dump_manager::dump_begin): ...new method.
1301 (dump_phase_enabled_p): Convert to...
1302 (gcc::dump_manager::dump_phase_enabled_p): ...method.
1303 (dump_phase_enabled_p): Convert to...
1304 (gcc::dump_manager::dump_phase_enabled_p): ...method.
1305 (dump_initialized_p): Convert to...
1306 (gcc::dump_manager::dump_initialized_p): ...method.
1307 (dump_flag_name): Replace body with...
1308 (gcc::dump_manager::dump_flag_name): ...new method.
1309 (dump_enable_all): Convert to...
1310 (gcc::dump_manager::dump_enable_all): ...new method.
1311 (opt_info_enable_passes): Convert to...
1312 (gcc::dump_manager::opt_info_enable_passes): ...new method.
1313 (dump_switch_p_1): Convert to...
1314 (gcc::dump_manager::dump_switch_p_1): ...new method.
1315 (dump_switch_p): Convert to...
1316 (gcc::dump_manager::dump_switch_p): ...new method.
1317 (opt_info_switch_p): Port to dump_manager API.
1318 (enable_rtl_dump_file): Likewise.
1319 * opts-global.c (handle_common_deferred_options): Port to new
1320 dump_manager API.
1321 * passes.c (pass_manager::finish_optimization_passes): Likewise.
1322 (pass_manager::register_one_dump_file): Likewise.
1323 (pass_manager::register_pass): Likewise.
1324 (pass_init_dump_file): Likewise.
1325 (pass_fini_dump_file): Likewise.
1326 * statistics.c (statistics_early_init): Likewise.
1327
83fd5d11
RB		 13282013-10-14 Richard Biener <rguenther@suse.de>
1329
1330 * gimple.c (gimple_canonical_types, canonical_type_hash_cache,
1331 iterative_hash_canonical_type, gimple_canonical_type_hash,
1332 gimple_canonical_types_compatible_p, gimple_canonical_type_eq,
1333 gimple_register_canonical_type, print_gimple_types_stats,
1334 free_gimple_type_tables): Move to lto/lto.c
1335 (gt-gimple.h): Do not include.
1336 * gimple.h (gimple_register_canonical_type,
1337 print_gimple_types_stats, free_gimple_type_tables): Remove.
1338 * Makefile.in (GTFILES): Remove gimple.c.
1339
a6dbd0f7
TS		 13402013-10-14 Travis Snoozy <quandary@remstate.com>
1341
1342 PR target/58716
1343 * config/msp430/msp430.c (msp430_option_override): Correct thinko
1344 scanning for msp430x targets.
1345
b5ccb9ed
EB		 13462013-10-14 Eric Botcazou <ebotcazou@adacore.com>
1347
1348 PR bootstrap/58509
1349 * config/sparc/sparc-protos.h (widen_mem_for_ldd_peep): Declare.
1350 (registers_ok_for_ldd_peep): Move around.
1351 * config/sparc/sparc.c (widen_mem_for_ldd_peep): New.
1352 * config/sparc/sparc.md (widening peepholes): Use it.
1353
fd7e5084
RB		 13542013-10-14 Richard Biener <rguenther@suse.de>
1355
1356 PR middle-end/58712
1357 PR middle-end/55358
1358 * gimple.c (iterative_hash_canonical_type): Make sure to
1359 record the hash into the correct hashtable slot.
1360
3246afaa
EB		 13612013-10-13 Eric Botcazou <ebotcazou@adacore.com>
1362
1363 PR rtl-optimization/58662
1364 * combine.c (try_combine): Take into account death nodes on I2 when
1365 splitting a PARALLEL of two independent SETs. Fix dump message.
1366
5d30dc5b
OE		 13672013-10-12 Oleg Endo <olegendo@gcc.gnu.org>
1368
1369 PR target/51244
1370 * config/sh/sh_treg_combine.cc: New SH specific RTL pass.
1371 * config.gcc (SH extra_objs): Add sh_ifcvt.o.
1372 * config/sh/t-sh (sh_treg_combine.o): New entry.
1373 * config/sh/sh.c (sh_fixed_condition_code_regs): New function that
1374 implements the target hook TARGET_FIXED_CONDITION_CODE_REGS.
1375 (register_sh_passes): New function. Register sh_treg_combine pass.
1376 (sh_option_override): Invoke it.
1377 (sh_canonicalize_comparison): Handle op0_preserve_value.
1378 * sh.md (*cbranch_t"): Do not try to optimize missed test and branch
1379 opportunities. Canonicalize branch condition.
1380 (nott): Allow only if pseudos can be created for non-SH2A.
1381
37d1c5db
L		 13822013-10-12 H.J. Lu <hongjiu.lu@intel.com>
1383
1384 PR target/58690
1385 * config/i386/i386.c (ix86_copy_addr_to_reg): New function.
1386 (ix86_expand_movmem): Replace copy_addr_to_reg with
1387 ix86_copy_addr_to_reg.
1388 (ix86_expand_setmem): Likewise.
1389
b86a70ff
AM		 13902013-10-12 Alexander Monakov <amonakov@ispras.ru>
1391
1392 * config/i386/i386.c (ix86_expand_sse_compare_and_jump): Use mode
1393 provided by ix86_fp_compare_mode instead of CCFPUmode.
1394
6e1d5961
JG		 13952013-10-12 James Greenhalgh <james.greenhalgh@arm.com>
1396
1397 * config/aarch64/arm_neon.h
1398 (vtbx<1,3>_<psu>8): Fix register constriants.
1399
ef3cfba2
JL		 14002013-10-11 Jeff Law <law@redhat.com>
1401
1402 PR tree-optimization/58640
adcfd489
UB		 1403 * tree-ssa-threadupdate.c (mark_threaded_blocks): Truncate jump
1404 threading paths that cross over two loop entry points.
ef3cfba2 1405
6e8b7d9c
BS		 14062013-10-11 Bill Schmidt <wschmidt@linux.vnet.ibm.com>
1407
1408 * config/rs6000/vsx.md (*vsx_le_perm_load_v2di): Generalize to
1409 handle vector float as well.
1410 (*vsx_le_perm_load_v4si): Likewise.
1411 (*vsx_le_perm_store_v2di): Likewise.
1412 (*vsx_le_perm_store_v4si): Likewise.
1413
f200869a
BS		 14142013-10-11 Bill Schmidt <wschmidt@linux.vnet.ibm.com>
1415
1416 * config/rs6000/vector.md (vec_realign_load<mode>): Generate vperm
1417 directly to circumvent subtract from splat{31} workaround.
1418 * config/rs6000/rs6000-protos.h (altivec_expand_vec_perm_le): New
1419 prototype.
1420 * config/rs6000/rs6000.c (altivec_expand_vec_perm_le): New.
1421 * config/rs6000/altivec.md (define_c_enum "unspec"): Add
1422 UNSPEC_VPERM_X and UNSPEC_VPERM_UNS_X.
1423 (altivec_vperm_<mode>): Convert to define_insn_and_split to
1424 separate big and little endian logic.
1425 (*altivec_vperm_<mode>_internal): New define_insn.
1426 (altivec_vperm_<mode>_uns): Convert to define_insn_and_split to
1427 separate big and little endian logic.
1428 (*altivec_vperm_<mode>_uns_internal): New define_insn.
1429 (vec_permv16qi): Add little endian logic.
1430
8d28e3fc
MG		 14312013-10-11 Marc Glisse <marc.glisse@inria.fr>
1432
1433 * doc/extend.texi (returns_nonnull): Remove arguments.
1434
3c87b77b
AI		 14352013-10-11 Alexander Ivchenko <alexander.ivchenko@intel.com>
1436 Maxim Kuznetsov <maxim.kuznetsov@intel.com>
1437 Sergey Lega <sergey.s.lega@intel.com>
1438 Anna Tikhonova <anna.tikhonova@intel.com>
1439 Ilya Tocar <ilya.tocar@intel.com>
1440 Andrey Turetskiy <andrey.turetskiy@intel.com>
1441 Ilya Verbin <ilya.verbin@intel.com>
1442 Kirill Yukhin <kirill.yukhin@intel.com>
1443 Michael Zolotukhin <michael.v.zolotukhin@intel.com>
1444
1445 * config/i386/sse.md (VI48F_256_512): New.
1446 (avx2_permvar<mode>): Change to ...
1447 (<avx2_avx512f>_permvar<mode>): This.
1448
50e60d7d
AI		 14492013-10-11 Alexander Ivchenko <alexander.ivchenko@intel.com>
1450 Maxim Kuznetsov <maxim.kuznetsov@intel.com>
1451 Sergey Lega <sergey.s.lega@intel.com>
1452 Anna Tikhonova <anna.tikhonova@intel.com>
1453 Ilya Tocar <ilya.tocar@intel.com>
1454 Andrey Turetskiy <andrey.turetskiy@intel.com>
1455 Ilya Verbin <ilya.verbin@intel.com>
1456 Kirill Yukhin <kirill.yukhin@intel.com>
1457 Michael Zolotukhin <michael.v.zolotukhin@intel.com>
1458
1459 * config/i386/i386.c (bdesc_args): Change corresponding pattern for
1460 __builtin_ia32_cvtps2dq, __builtin_ia32_cvtps2dq256.
1461 * config/i386/sse.md (VI4_AVX): New.
1462 (sf2simodelower): Ditto.
1463 (sse2_cvtps2dq): Change to ...
1464 (<sse2_avx_avx512f>_fix_notrunc<sf2simodelower><mode>): This.
1465
f62ce24f
AI		 14662013-10-11 Alexander Ivchenko <alexander.ivchenko@intel.com>
1467 Maxim Kuznetsov <maxim.kuznetsov@intel.com>
1468 Sergey Lega <sergey.s.lega@intel.com>
1469 Anna Tikhonova <anna.tikhonova@intel.com>
1470 Ilya Tocar <ilya.tocar@intel.com>
1471 Andrey Turetskiy <andrey.turetskiy@intel.com>
1472 Ilya Verbin <ilya.verbin@intel.com>
1473 Kirill Yukhin <kirill.yukhin@intel.com>
1474 Michael Zolotukhin <michael.v.zolotukhin@intel.com>
1475
1476 * config/i386/sse.md (V_512): New.
1477 (VI_512): Ditto.
1478 (vcond<V_512:mode><VF_512:mode>): Ditto.
1479 (vcond<V_512:mode><VI_512:mode>): Ditto.
1480 (vcondu<V_512:mode><VI_512:mode>): Ditto.
1481
558d9f79
AI		 14822013-10-11 Alexander Ivchenko <alexander.ivchenko@intel.com>
1483 Maxim Kuznetsov <maxim.kuznetsov@intel.com>
1484 Sergey Lega <sergey.s.lega@intel.com>
1485 Anna Tikhonova <anna.tikhonova@intel.com>
1486 Ilya Tocar <ilya.tocar@intel.com>
1487 Andrey Turetskiy <andrey.turetskiy@intel.com>
1488 Ilya Verbin <ilya.verbin@intel.com>
1489 Kirill Yukhin <kirill.yukhin@intel.com>
1490 Michael Zolotukhin <michael.v.zolotukhin@intel.com>
1491
1492 * config/i386/i386.c (ix86_rtx_costs): Enable fma for TARGET_AVX512F.
1493 * config/i386/sse.md (FMAMODEM): Changed modes and conditions.
1494 (FMAMODE): Ditto.
1495 (fma<mode>4): Removed condition.
1496 (fms<mode>4): Ditto.
1497 (fnma<mode>4): Ditto.
1498 (fnms<mode>4): Ditto.
1499 (fma4i_fmadd_<mode>): Ditto.
1500 (*fma_fmadd_<mode>): Ditto.
1501 (*fma_fmsub_<mode>): Ditto.
1502 (*fma_fnmadd_<mode>): Ditto.
1503 (*fma_fnmsub_<mode>): Ditto.
1504 (fmaddsub_<mode>): Allow for TARGET_AVX512F.
1505 (*fma_fmaddsub_<mode>): Ditto.
1506 (*fma_fmsubadd_<mode>): Ditto.
1507 (*fmai_fmadd_<mode>): Ditto.
1508 (*fmai_fmsub_<mode>): Ditto.
1509 (*fmai_fnmadd_<mode>): Ditto.
1510 (*fmai_fnmsub_<mode>): Ditto.
1511
e8d08206
AI		 15122013-10-11 Alexander Ivchenko <alexander.ivchenko@intel.com>
1513 Maxim Kuznetsov <maxim.kuznetsov@intel.com>
1514 Sergey Lega <sergey.s.lega@intel.com>
1515 Anna Tikhonova <anna.tikhonova@intel.com>
1516 Ilya Tocar <ilya.tocar@intel.com>
1517 Andrey Turetskiy <andrey.turetskiy@intel.com>
1518 Ilya Verbin <ilya.verbin@intel.com>
1519 Kirill Yukhin <kirill.yukhin@intel.com>
1520 Michael Zolotukhin <michael.v.zolotukhin@intel.com>
1521
1522 * config/i386/sse.md (VI248_AVX2_8_AVX512F): New.
1523 (VI124_256): Changed to ...
1524 (VI124_256_48_512): This.
1525 (ssepackmode): Extended with wider modes.
1526 (<code><mode>3): Changed iterator.
1527 (*avx2_<code><mode>3): Ditto.
1528 (vec_pack_trunc_<mode>): Ditto.
1529
16821545
AI		 15302013-10-11 Alexander Ivchenko <alexander.ivchenko@intel.com>
1531 Maxim Kuznetsov <maxim.kuznetsov@intel.com>
1532 Sergey Lega <sergey.s.lega@intel.com>
1533 Anna Tikhonova <anna.tikhonova@intel.com>
1534 Ilya Tocar <ilya.tocar@intel.com>
1535 Andrey Turetskiy <andrey.turetskiy@intel.com>
1536 Ilya Verbin <ilya.verbin@intel.com>
1537 Kirill Yukhin <kirill.yukhin@intel.com>
1538 Michael Zolotukhin <michael.v.zolotukhin@intel.com>
1539
1540 * config/i386/sse.md (VI124_AVX2_48_AVX512F): New.
1541 (VI8F_256_512): Ditto.
1542 (abs<mode>2): Changed iterator.
1543 (avx2_perm<mode>): Changed to ...
1544 (<avx2_avx512f>_perm<mode>): This.
1545 (avx2_perm<mode>_1): Changed to ...
1546 (<avx2_avx512f>_perm<mode>_1): This.
1547
5348cff8
AI		 15482013-10-11 Alexander Ivchenko <alexander.ivchenko@intel.com>
1549 Maxim Kuznetsov <maxim.kuznetsov@intel.com>
1550 Sergey Lega <sergey.s.lega@intel.com>
1551 Anna Tikhonova <anna.tikhonova@intel.com>
1552 Ilya Tocar <ilya.tocar@intel.com>
1553 Andrey Turetskiy <andrey.turetskiy@intel.com>
1554 Ilya Verbin <ilya.verbin@intel.com>
1555 Kirill Yukhin <kirill.yukhin@intel.com>
1556 Michael Zolotukhin <michael.v.zolotukhin@intel.com>
1557
1558 * config/i386/sse.md (VI48_AVX512F): New.
1559 (VI48_AVX2): Changed to ...
1560 (VI48_AVX2_48_AVX512F): This.
1561 (avx2_ashrv<mode>): Changed to ...
1562 (<avx2_avx512f>_ashrv<mode>): This.
1563 (avx2_<shift_insn>v<mode>): Changed to ...
1564 (<avx2_avx512f>_<shift_insn>v<mode>): This.
1565
f5f41d88
AI		 15662013-10-11 Alexander Ivchenko <alexander.ivchenko@intel.com>
1567 Maxim Kuznetsov <maxim.kuznetsov@intel.com>
1568 Sergey Lega <sergey.s.lega@intel.com>
1569 Anna Tikhonova <anna.tikhonova@intel.com>
1570 Ilya Tocar <ilya.tocar@intel.com>
1571 Andrey Turetskiy <andrey.turetskiy@intel.com>
1572 Ilya Verbin <ilya.verbin@intel.com>
1573 Kirill Yukhin <kirill.yukhin@intel.com>
1574 Michael Zolotukhin <michael.v.zolotukhin@intel.com>
1575
1576 * config/i386/sse.md (VI4_AVX512F): New.
1577 (VI8_AVX2_AVX512F): Ditto.
1578 (mul<mode>3): Extended with wider modes.
1579 (*<sse4_1_avx2>_mul<mode>3): Ditto.
1580 (mul<mode>3): Ditto.
1581 (vec_widen_<s>mult_odd_<mode>): Ditto.
1582
3bdf6340
AI		 15832013-10-11 Alexander Ivchenko <alexander.ivchenko@intel.com>
1584 Maxim Kuznetsov <maxim.kuznetsov@intel.com>
1585 Sergey Lega <sergey.s.lega@intel.com>
1586 Anna Tikhonova <anna.tikhonova@intel.com>
1587 Ilya Tocar <ilya.tocar@intel.com>
1588 Andrey Turetskiy <andrey.turetskiy@intel.com>
1589 Ilya Verbin <ilya.verbin@intel.com>
1590 Kirill Yukhin <kirill.yukhin@intel.com>
1591 Michael Zolotukhin <michael.v.zolotukhin@intel.com>
1592
1593 * config/i386/sse.md (VI2_AVX512F): New.
1594 (VI124_AVX512F): Ditto.
1595 (sseunpackmode): Extended with wider modes.
1596 (sseunpackfltmode): Ditto.
1597 (vec_unpacks_float_hi_<mode>): Ditto.
1598 (vec_unpacks_float_lo_<mode>): Ditto.
1599 (vec_unpacku_float_hi_<mode>): Ditto.
1600 (vec_unpacku_float_lo_<mode>): Ditto.
1601 (vec_unpacks_lo_<mode>): Ditto.
1602 (vec_unpacks_hi_<mode>): Ditto.
1603 (vec_unpacku_lo_<mode>): Ditto.
1604 (vec_unpacku_hi_<mode>): Ditto.
1605
67f783cb
AI		 16062013-10-11 Alexander Ivchenko <alexander.ivchenko@intel.com>
1607 Maxim Kuznetsov <maxim.kuznetsov@intel.com>
1608 Sergey Lega <sergey.s.lega@intel.com>
1609 Anna Tikhonova <anna.tikhonova@intel.com>
1610 Ilya Tocar <ilya.tocar@intel.com>
1611 Andrey Turetskiy <andrey.turetskiy@intel.com>
1612 Ilya Verbin <ilya.verbin@intel.com>
1613 Kirill Yukhin <kirill.yukhin@intel.com>
1614 Michael Zolotukhin <michael.v.zolotukhin@intel.com>
1615
1616 * config/i386/i386.md (multdiv): New.
1617 (multdiv_mnemonic): Ditto.
1618 * config/i386/sse.md (<sse>_vmmul<mode>3): Changed to...
1619 (<sse>_vm<multdiv_mnemonic><mode>3): This.
1620 (<sse>_vmdiv<mode>3): Removed.
1621
ec5e777c
AI		 16222013-10-11 Alexander Ivchenko <alexander.ivchenko@intel.com>
1623 Maxim Kuznetsov <maxim.kuznetsov@intel.com>
1624 Sergey Lega <sergey.s.lega@intel.com>
1625 Anna Tikhonova <anna.tikhonova@intel.com>
1626 Ilya Tocar <ilya.tocar@intel.com>
1627 Andrey Turetskiy <andrey.turetskiy@intel.com>
1628 Ilya Verbin <ilya.verbin@intel.com>
1629 Kirill Yukhin <kirill.yukhin@intel.com>
1630 Michael Zolotukhin <michael.v.zolotukhin@intel.com>
1631
1632 * config/i386/sse.md (V): Extended with wider modes.
1633 (VF2): Ditto.
1634 (ssehalfvecmode): Ditto.
1635 (i128): Ditto.
1636 (ssepackfltmode): Ditto.
1637 (avx_vec_concat<mode>): Ditto.
1638 (V_256_512): New iterator.
1639 (VF2_512_256): Ditto.
1640 (si2dfmode): New attribute.
1641 (si2dfmodelower): Ditto.
1642 (sf2dfmode): Ditto.
1643 (concat_tg_mode): Ditto.
1644 (floatv4siv4df2): Changed to ...
1645 (float<si2dfmodelower><mode>2): This.
1646 (avx_cvtps2pd256): Changed to ...
1647 (<sse2_avx_avx512f>_cvtps2pd<avxsizesuffix>): This.
1648 (vec_pack_trunc_v4df): Changed to ...
1649 (vec_pack_trunc_<mode>): This.
1650 (avx_vpermil<mode>): Changed to ...
1651 (<sse2_avx_avx512f>_vpermil<mode>): This.
1652 (<fixsuffix>fix_truncv8dfv8si2): New.
1653 (vec_pack_sfix_trunc_v8df): Ditto.
1654 (avx512f_rndscale<mode>): Ditto.
1655 (avx512f_roundpd512): Ditto.
1656 (vec_pack_ufix_trunc_<mode>): Updated iterator.
1657
a9ccbba2
AI		 16582013-10-11 Alexander Ivchenko <alexander.ivchenko@intel.com>
1659 Maxim Kuznetsov <maxim.kuznetsov@intel.com>
1660 Sergey Lega <sergey.s.lega@intel.com>
1661 Anna Tikhonova <anna.tikhonova@intel.com>
1662 Ilya Tocar <ilya.tocar@intel.com>
1663 Andrey Turetskiy <andrey.turetskiy@intel.com>
1664 Ilya Verbin <ilya.verbin@intel.com>
1665 Kirill Yukhin <kirill.yukhin@intel.com>
1666 Michael Zolotukhin <michael.v.zolotukhin@intel.com>
1667
1668 * config/i386/i386.md (any_fix): New iterator.
1669 (fixsuffix): New attribute.
1670 * config/i386/sse.md (VF1): Extened with wider modes.
1671 (VI): Ditto.
1672 (VI_AVX2): Ditto.
1673 (VI8): Ditto.
1674 (sseintvecmodelower): Ditto.
1675 (ssescalarmode): Ditto.
1676 (ssescalarnum): Ditto.
1677 (VF1_128_256): New.
1678 (ssexmmmode): Ditto.
1679 (<fixsuffix>fix_truncv16sfv16si2): Ditto.
1680 (<sse>_rcp<mode>2): Change iterator.
1681 (rsqrt<mode>2): Ditto.
1682 (<sse>_rsqrt<mode>2): Ditto.
1683 (avx2_vec_dup<mode>): Ditto.
1684 (<sse4_1>_round<ssemodesuffix>_sfix<avxsizesuffix>): Ditto.
1685 (round<mode>2_sfix): Ditto.
1686 (avx2_pbroadcast<mode>): Ditto.
1687 (*andnot<mode>3): Handle XI mode.
1688 (*<code><mode>3): Ditto.
1689 (AVXTOSSEMODE): Removed.
1690 (avx_vpermil<mode>): Changed to ...
1691 (<sse2_avx_avx512f>_vpermil<mode>): This.
1692
2b1ebb0c
AI		 16932013-10-11 Alexander Ivchenko <alexander.ivchenko@intel.com>
1694 Maxim Kuznetsov <maxim.kuznetsov@intel.com>
1695 Sergey Lega <sergey.s.lega@intel.com>
1696 Anna Tikhonova <anna.tikhonova@intel.com>
1697 Ilya Tocar <ilya.tocar@intel.com>
1698 Andrey Turetskiy <andrey.turetskiy@intel.com>
1699 Ilya Verbin <ilya.verbin@intel.com>
1700 Kirill Yukhin <kirill.yukhin@intel.com>
1701 Michael Zolotukhin <michael.v.zolotukhin@intel.com>
1702
1703 * config/i386/sse.md (<sse2>_movnt<mode>): Update constraint to "v".
1704 (<sse>_comi): Ditto.
1705 (<sse>_ucomi): Ditto.
1706 (sse_cvtss2siq_2): Ditto.
1707 (sse2_cvtsd2si): Ditto.
1708 (sse2_cvtsd2siq): Ditto.
1709 (sse2_cvttsd2si): Ditto.
1710 (sse2_cvttsd2siq): Ditto.
1711 (<shift_insn><mode>3): Ditto.
1712 (sse2_cvtsi2sdq): Update constraint and prefix.
1713 (sse_cvtsi2ss): Update prefix.
1714 (sse_cvtsi2ssq): Ditto.
1715
d89124ed
JJ		 17162013-10-11 Jakub Jelinek <jakub@redhat.com>
1717
1718 * tree-vrp.c (infer_nonnull_range): Use is_gimple_call,
1719 ignore internal calls.
1720
e6ad28c3
RB		 17212013-10-11 Richard Biener <rguenther@suse.de>
1722
1723 * tree-pretty-print.c (dump_generic_node): Allow to dump
1724 both (D) and (ab) for SSA_NAMEs. Mark INTEGER_CSTs with
1725 (OVF) if TREE_OVERFLOW is set.
1726
d1227447
TS		 17272013-10-11 Thomas Schwinge <thomas@codesourcery.com>
1728
1ac4b8e4
TS		 1729 * tree.h (OMP_CLAUSE_CODE): Remove duplicate definition.
1730
7d05cebb
TS		 1731 * gimple.c: GIMPLE statements have subcodes, not sub-codes.
1732 * gimple.h: Likewise.
1733
adcfd489 1734 * doc/generic.texi (OpenMP): OMP_CLAUSE_* are subcodes, not sub-codes.
6545f0b3 1735
1a79fb8a
TS		 1736 * doc/generic.texi (Adding new DECL node types): Explain *_CHECK
1737 macros.
1738
3afd2873
TS		 1739 * doc/gimple.texi (is_gimple_omp): Move into the correct section.
1740
adcfd489 1741 * acinclude.m4 (gcc_GAS_FLAGS): Add more gcc_cv_as_flags overrides.
d1227447
TS		 1742 * configure: Regenerate.
1743
acf0174b
JJ		 17442013-10-11 Jakub Jelinek <jakub@redhat.com>
1745
1746 * tree-pretty-print.c (dump_omp_clause): Handle OMP_CLAUSE__LOOPTEMP_
1747 and new OpenMP 4.0 clauses, handle UDR OMP_CLAUSE_REDUCTION,
1748 formatting fixes, use pp_colon instead of pp_character (..., ':'),
1749 similarly pp_right_paren.
1750 (dump_generic_node): Handle OMP_DISTRIBUTE, OMP_TEAMS,
1751 OMP_TARGET_DATA, OMP_TARGET, OMP_TARGET_UPDATE, OMP_TASKGROUP,
1752 allow OMP_FOR_INIT to be NULL, handle OMP_ATOMIC_SEQ_CST.
1753 * tree.c (omp_clause_num_ops, omp_clause_code_name): Add OpenMP 4.0
1754 clauses.
1755 (omp_declare_simd_clauses_equal,
1756 omp_remove_redundant_declare_simd_attrs): New functions.
1757 (attribute_value_equal): Use omp_declare_simd_clauses_equal.
1758 (walk_tree_1): Handle new OpenMP 4.0 clauses.
1759 * tree.h (OMP_LOOP_CHECK): Define.
1760 (OMP_FOR_BODY, OMP_FOR_CLAUSES, OMP_FOR_INIT, OMP_FOR_COND,
1761 OMP_FOR_INCR, OMP_FOR_PRE_BODY): Use it.
1762 (OMP_TASKGROUP_BODY, OMP_TEAMS_BODY, OMP_TEAMS_CLAUSES,
1763 OMP_TARGET_DATA_BODY, OMP_TARGET_DATA_CLAUSES, OMP_TARGET_BODY,
1764 OMP_TARGET_CLAUSES, OMP_TARGET_UPDATE_CLAUSES, OMP_CLAUSE_SIZE,
1765 OMP_ATOMIC_SEQ_CST, OMP_CLAUSE_DEPEND_KIND, OMP_CLAUSE_MAP_KIND,
1766 OMP_CLAUSE_MAP_ZERO_BIAS_ARRAY_SECTION, OMP_CLAUSE_PROC_BIND_KIND,
1767 OMP_CLAUSE_REDUCTION_OMP_ORIG_REF, OMP_CLAUSE_ALIGNED_ALIGNMENT,
1768 OMP_CLAUSE_NUM_TEAMS_EXPR, OMP_CLAUSE_THREAD_LIMIT_EXPR,
1769 OMP_CLAUSE_DEVICE_ID, OMP_CLAUSE_DIST_SCHEDULE_CHUNK_EXPR,
1770 OMP_CLAUSE_SIMDLEN_EXPR): Define.
1771 (OMP_CLAUSE_DECL): Change range up to OMP_CLAUSE__LOOPTEMP_.
1772 (omp_remove_redundant_declare_simd_attrs): New prototype.
1773 * gimple.def (GIMPLE_OMP_TASKGROUP, GIMPLE_OMP_TARGET,
1774 GIMPLE_OMP_TEAMS): New codes.
1775 (GIMPLE_OMP_RETURN): Use GSS_OMP_ATOMIC_STORE instead of GSS_BASE.
1776 * omp-low.c (struct omp_context): Add cancel_label and cancellable
1777 fields.
1778 (target_nesting_level): New variable.
1779 (extract_omp_for_data): Handle GF_OMP_FOR_KIND_DISTRIBUTE and
1780 OMP_CLAUSE_DIST_SCHEDULE. Don't fallback to library implementation
1781 for collapse > 1 static schedule unless ordered.
1782 (get_ws_args_for): Add par_stmt argument. Handle combined loops.
1783 (determine_parallel_type): Adjust get_ws_args_for caller.
1784 (install_var_field): Handle mask & 4 for double indirection.
1785 (scan_sharing_clauses): Ignore shared clause on teams construct.
1786 Handle OMP_CLAUSE__LOOPTEMP_ and new OpenMP 4.0 clauses.
1787 (create_omp_child_function): If inside target or declare target
adcfd489 1788 constructs, set "omp declare target" attribute on the child function.
acf0174b
JJ		 1789 (find_combined_for): New function.
1790 (scan_omp_parallel): Handle combined loops.
1791 (scan_omp_target, scan_omp_teams): New functions.
1792 (check_omp_nesting_restrictions): Check new OpenMP 4.0 nesting
1793 restrictions and set ctx->cancellable for cancellable constructs.
1794 (scan_omp_1_stmt): Call check_omp_nesting_restrictions also on
1795 selected builtin calls. Handle GIMPLE_OMP_TASKGROUP,
1796 GIMPLE_OMP_TARGET, GIMPLE_OMP_TEAMS.
adcfd489 1797 (build_omp_barrier): Add lhs argument, return gimple rather than tree.
acf0174b
JJ		 1798 (omp_clause_aligned_alignment): New function.
1799 (lower_rec_simd_input_clauses): Only call SET_DECL_VALUE_EXPR
1800 on decls.
1801 (lower_rec_input_clauses): Add FD argument. Ignore shared clauses
1802 on teams constructs. Handle user defined reductions and new
1803 OpenMP 4.0 clauses.
1804 (lower_reduction_clauses): Don't set placeholder to address of ref
1805 if it has already the right type.
1806 (lower_send_clauses): Handle OMP_CLAUSE__LOOPTEMP_.
1807 (expand_parallel_call): Use the new non-_start suffixed builtins,
1808 handle OMP_CLAUSE_PROC_BIND, don't call the outlined function
1809 and GOMP_parallel_end after the call.
1810 (expand_task_call): Handle OMP_CLAUSE_DEPEND.
1811 (expand_omp_for_init_counts): Handle combined loops.
1812 (expand_omp_for_init_vars): Add inner_stmt argument, handle combined
1813 loops.
1814 (expand_omp_for_generic): Likewise. Use GOMP_loop_end_cancel at the
1815 end of cancellable loops.
1816 (expand_omp_for_static_nochunk, expand_omp_for_static_chunk):
1817 Likewise. Handle collapse > 1 loops.
1818 (expand_omp_simd): Handle combined loops.
1819 (expand_omp_for): Add inner_stmt argument, adjust callers of
1820 expand_omp_for* functions, use expand_omp_for_static*chunk even
1821 for collapse > 1 unless ordered.
1822 (expand_omp_sections): Use GOMP_sections_end_cancel at the end
1823 of cancellable sections.
1824 (expand_omp_single): Remove need_barrier variable, just rely on
1825 gimple_omp_return_nowait_p. Adjust build_omp_barrier caller.
1826 (expand_omp_synch): Allow GIMPLE_OMP_TASKGROUP and GIMPLE_OMP_TEAMS.
1827 (expand_omp_atomic_load, expand_omp_atomic_store,
1828 expand_omp_atomic_fetch_op): Handle gimple_omp_atomic_seq_cst_p.
1829 (expand_omp_target): New function.
1830 (expand_omp): Handle combined loops. Handle GIMPLE_OMP_TASKGROUP,
1831 GIMPLE_OMP_TEAMS, GIMPLE_OMP_TARGET.
1832 (build_omp_regions_1): Immediately close region for
1833 GF_OMP_TARGET_KIND_UPDATE.
1834 (maybe_add_implicit_barrier_cancel): New function.
1835 (lower_omp_sections): Adjust lower_rec_input_clauses caller. Handle
1836 cancellation.
1837 (lower_omp_single): Likewise. Add clobber after the barrier.
1838 (lower_omp_taskgroup): New function.
1839 (lower_omp_for): Handle combined loops. Adjust
1840 lower_rec_input_clauses caller. Handle cancellation.
1841 (lower_depend_clauses): New function.
1842 (lower_omp_taskreg): Lower depend clauses. Adjust
1843 lower_rec_input_clauses caller. Add clobber after the call. Handle
1844 cancellation.
1845 (lower_omp_target, lower_omp_teams): New functions.
1846 (lower_omp_1): Handle cancellation. Handle GIMPLE_OMP_TASKGROUP,
1847 GIMPLE_OMP_TARGET, GIMPLE_OMP_TEAMS and GOMP_barrier, GOMP_cancel
1848 and GOMP_cancellation_point calls.
1849 (lower_omp): Fold stmts inside of target region.
1850 (diagnose_sb_1, diagnose_sb_2): Handle GIMPLE_OMP_TASKGROUP,
1851 GIMPLE_OMP_TARGET and GIMPLE_OMP_TEAMS.
1852 * builtin-types.def (DEF_FUNCTION_TYPE_8): Document.
1853 (BT_FN_VOID_OMPFN_PTR_UINT,
1854 BT_FN_VOID_OMPFN_PTR_UINT_LONG_LONG_LONG,
1855 BT_FN_VOID_OMPFN_PTR_UINT_LONG_LONG_LONG_LONG,
1856 BT_FN_VOID_OMPFN_PTR_OMPCPYFN_LONG_LONG_BOOL_UINT): Remove.
1857 (BT_FN_VOID_OMPFN_PTR_UINT_UINT_UINT,
1858 BT_FN_VOID_OMPFN_PTR_UINT_LONG_LONG_LONG_UINT,
1859 BT_FN_VOID_OMPFN_PTR_UINT_LONG_LONG_LONG_LONG_UINT,
1860 BT_FN_BOOL_INT, BT_FN_BOOL_INT_BOOL, BT_FN_VOID_UINT_UINT,
1861 BT_FN_VOID_INT_PTR_SIZE_PTR_PTR_PTR,
1862 BT_FN_VOID_INT_OMPFN_PTR_SIZE_PTR_PTR_PTR,
1863 BT_FN_VOID_OMPFN_PTR_OMPCPYFN_LONG_LONG_BOOL_UINT_PTR): New.
1864 * tree-ssa-alias.c (ref_maybe_used_by_call_p_1,
1865 call_may_clobber_ref_p_1): Handle BUILT_IN_GOMP_BARRIER_CANCEL,
1866 BUILT_IN_GOMP_TASKGROUP_END, BUILT_IN_GOMP_LOOP_END_CANCEL,
1867 BUILT_IN_GOMP_SECTIONS_END_CANCEL. Don't handle
1868 BUILT_IN_GOMP_PARALLEL_END.
1869 * gimple-low.c (lower_stmt): Handle GIMPLE_OMP_TASKGROUP,
1870 GIMPLE_OMP_TARGET and GIMPLE_OMP_TEAMS.
1871 * gimple-pretty-print.c (dump_gimple_omp_for): Handle
1872 GF_OMP_FOR_KIND_DISTRIBUTE.
1873 (dump_gimple_omp_target, dump_gimple_omp_teams): New functions.
1874 (dump_gimple_omp_block): Handle GIMPLE_OMP_TASKGROUP.
1875 (dump_gimple_omp_return): Print lhs if it has any.
1876 (dump_gimple_omp_atomic_load, dump_gimple_omp_atomic_store): Handle
1877 gimple_omp_atomic_seq_cst_p.
1878 (pp_gimple_stmt_1): Handle GIMPLE_OMP_TASKGROUP, GIMPLE_OMP_TARGET
1879 and GIMPLE_OMP_TEAMS.
1880 * langhooks.c (lhd_omp_mappable_type): New function.
1881 * tree-vectorizer.c (struct simd_array_to_simduid): Fix up comment.
1882 * langhooks.h (struct lang_hooks_for_types): Add omp_mappable_type
1883 hook.
1884 * gimplify.c (enum gimplify_omp_var_data): Add GOVD_MAP,
1885 GOVD_ALIGNED and GOVD_MAP_TO_ONLY.
1886 (enum omp_region_type): Add ORT_TEAMS, ORT_TARGET_DATA and
1887 ORT_TARGET.
1888 (struct gimplify_omp_ctx): Add combined_loop field.
1889 (gimplify_call_expr, gimplify_modify_expr): Don't call fold_stmt
1890 on stmts inside of target region.
1891 (is_gimple_stmt): Return true for OMP_DISTRIBUTE and OMP_TASKGROUP.
1892 (omp_firstprivatize_variable): Handle GOVD_MAP, GOVD_ALIGNED,
1893 ORT_TARGET and ORT_TARGET_DATA.
1894 (omp_add_variable): Avoid checks on readding var for GOVD_ALIGNED.
1895 Handle GOVD_MAP.
1896 (omp_notice_threadprivate_variable): Complain about threadprivate
1897 variables in target region.
1898 (omp_notice_variable): Complain about vars with non-mappable type
1899 in target region. Handle ORT_TEAMS, ORT_TARGET and ORT_TARGET_DATA.
1900 (omp_check_private): Ignore ORT_TARGET* regions.
1901 (gimplify_scan_omp_clauses, gimplify_adjust_omp_clauses_1,
1902 gimplify_adjust_omp_clauses): Handle new OpenMP 4.0 clauses.
1903 (find_combined_omp_for): New function.
1904 (gimplify_omp_for): Handle gimplification of combined loops.
1905 (gimplify_omp_workshare): Gimplify also OMP_TARGET, OMP_TARGET_DATA,
1906 OMP_TEAMS.
1907 (gimplify_omp_target_update): New function.
1908 (gimplify_omp_atomic): Handle OMP_ATOMIC_SEQ_CST.
1909 (gimplify_expr): Handle OMP_DISTRIBUTE, OMP_TARGET, OMP_TARGET_DATA,
1910 OMP_TARGET_UPDATE, OMP_TEAMS, OMP_TASKGROUP.
1911 (gimplify_body): If fndecl has "omp declare target" attribute, add
1912 implicit ORT_TARGET context around it.
1913 * tree.def (OMP_DISTRIBUTE, OMP_TEAMS, OMP_TARGET_DATA, OMP_TARGET,
1914 OMP_TASKGROUP, OMP_TARGET_UPDATE): New tree codes.
1915 * tree-nested.c (convert_nonlocal_reference_stmt,
1916 convert_local_reference_stmt, convert_gimple_call): Handle
1917 GIMPLE_OMP_TARGET, GIMPLE_OMP_TEAMS and GIMPLE_OMP_TASKGROUP.
1918 * omp-builtins.def (BUILT_IN_GOMP_TASK): Use
1919 BT_FN_VOID_OMPFN_PTR_OMPCPYFN_LONG_LONG_BOOL_UINT_PTR
1920 instead of BT_FN_VOID_OMPFN_PTR_OMPCPYFN_LONG_LONG_BOOL_UINT.
1921 (BUILT_IN_GOMP_TARGET, BUILT_IN_GOMP_TARGET_DATA,
1922 BUILT_IN_GOMP_TARGET_END_DATA, BUILT_IN_GOMP_TARGET_UPDATE,
1923 BUILT_IN_GOMP_TEAMS, BUILT_IN_BARRIER_CANCEL,
1924 BUILT_IN_GOMP_LOOP_END_CANCEL,
1925 BUILT_IN_GOMP_SECTIONS_END_CANCEL, BUILT_IN_OMP_GET_TEAM_NUM,
1926 BUILT_IN_OMP_GET_NUM_TEAMS, BUILT_IN_GOMP_TASKGROUP_START,
1927 BUILT_IN_GOMP_TASKGROUP_END, BUILT_IN_GOMP_PARALLEL_LOOP_STATIC,
1928 BUILT_IN_GOMP_PARALLEL_LOOP_DYNAMIC,
1929 BUILT_IN_GOMP_PARALLEL_LOOP_GUIDED,
1930 BUILT_IN_GOMP_PARALLEL_LOOP_RUNTIME, BUILT_IN_GOMP_PARALLEL,
1931 BUILT_IN_GOMP_PARALLEL_SECTIONS, BUILT_IN_GOMP_CANCEL,
1932 BUILT_IN_GOMP_CANCELLATION_POINT): New built-ins.
1933 (BUILT_IN_GOMP_PARALLEL_LOOP_STATIC_START,
1934 BUILT_IN_GOMP_PARALLEL_LOOP_DYNAMIC_START,
1935 BUILT_IN_GOMP_PARALLEL_LOOP_GUIDED_START,
1936 BUILT_IN_GOMP_PARALLEL_LOOP_RUNTIME_START,
1937 BUILT_IN_GOMP_PARALLEL_START, BUILT_IN_GOMP_PARALLEL_END,
1938 BUILT_IN_GOMP_PARALLEL_SECTIONS_START): Remove.
1939 * tree-inline.c (remap_gimple_stmt, estimate_num_insns):
1940 Handle GIMPLE_OMP_TARGET, GIMPLE_OMP_TEAMS and GIMPLE_OMP_TASKGROUP.
1941 * gimple.c (gimple_build_omp_taskgroup, gimple_build_omp_target,
1942 gimple_build_omp_teams): New functions.
1943 (walk_gimple_op): Handle GIMPLE_OMP_TARGET, GIMPLE_OMP_TEAMS and
1944 GIMPLE_OMP_TASKGROUP. Walk optional lhs on GIMPLE_OMP_RETURN.
1945 (walk_gimple_stmt, gimple_copy): Handle GIMPLE_OMP_TARGET,
1946 GIMPLE_OMP_TEAMS and GIMPLE_OMP_TASKGROUP.
1947 * gimple.h (enum gf_mask): GF_OMP_FOR_KIND_DISTRIBUTE,
1948 GF_OMP_FOR_COMBINED, GF_OMP_FOR_COMBINED_INTO,
1949 GF_OMP_TARGET_KIND_MASK, GF_OMP_TARGET_KIND_REGION,
1950 GF_OMP_TARGET_KIND_DATA, GF_OMP_TARGET_KIND_UPDATE,
1951 GF_OMP_ATOMIC_SEQ_CST): New.
1952 (gimple_build_omp_taskgroup, gimple_build_omp_target,
1953 gimple_build_omp_teams): New prototypes.
1954 (gimple_has_substatements): Handle GIMPLE_OMP_TARGET,
1955 GIMPLE_OMP_TEAMS and GIMPLE_OMP_TASKGROUP.
1956 (gimple_omp_subcode): Use GIMPLE_OMP_TEAMS instead of
1957 GIMPLE_OMP_SINGLE as end of range.
1958 (gimple_omp_return_set_lhs, gimple_omp_return_lhs,
1959 gimple_omp_return_lhs_ptr, gimple_omp_atomic_seq_cst_p,
1960 gimple_omp_atomic_set_seq_cst, gimple_omp_for_combined_p,
1961 gimple_omp_for_set_combined_p, gimple_omp_for_combined_into_p,
1962 gimple_omp_for_set_combined_into_p, gimple_omp_target_clauses,
1963 gimple_omp_target_clauses_ptr, gimple_omp_target_set_clauses,
1964 gimple_omp_target_kind, gimple_omp_target_set_kind,
1965 gimple_omp_target_child_fn, gimple_omp_target_child_fn_ptr,
1966 gimple_omp_target_set_child_fn, gimple_omp_target_data_arg,
1967 gimple_omp_target_data_arg_ptr, gimple_omp_target_set_data_arg,
1968 gimple_omp_teams_clauses, gimple_omp_teams_clauses_ptr,
1969 gimple_omp_teams_set_clauses): New inlines.
1970 (CASE_GIMPLE_OMP): Add GIMPLE_OMP_TARGET, GIMPLE_OMP_TEAMS
1971 and GIMPLE_OMP_TASKGROUP.
adcfd489 1972 * tree-core.h (enum omp_clause_code): Add new OpenMP 4.0 clause codes.
acf0174b
JJ		 1973 (enum omp_clause_depend_kind, enum omp_clause_map_kind,
1974 enum omp_clause_proc_bind_kind): New.
1975 (union omp_clause_subcode): Add depend_kind, map_kind and
1976 proc_bind_kind fields.
1977 * tree-cfg.c (make_edges): Handle GIMPLE_OMP_TARGET,
1978 GIMPLE_OMP_TEAMS and GIMPLE_OMP_TASKGROUP.
1979 * langhooks-def.h (lhd_omp_mappable_type): New prototype.
1980 (LANG_HOOKS_OMP_MAPPABLE_TYPE): Define.
1981 (LANG_HOOKS_FOR_TYPES_INITIALIZER): Add it.
1982
67fa7880
TJ		 19832013-10-10 Teresa Johnson <tejohnson@google.com>
1984
1985 * predict.c (tree_estimate_probability): Add new parameter
1986 for estimate_bb_frequencies.
1987 (estimate_bb_frequencies): Add new parameter to force estimation.
1988 (rebuild_frequencies): When max frequency in function is small,
1989 recompute counts from frequencies.
1990 * predict.h (estimate_bb_frequencies): New parameter.
1991
b591a8b7
DM		 19922013-10-10 David Malcolm <dmalcolm@redhat.com>
1993
1994 * ipa-inline.c (ipa_inline): Fix leak of "order" when
1995 optimizations are disabled.
1996
782f0db2
DM		 19972013-10-10 David Malcolm <dmalcolm@redhat.com>
1998
1999 * coverage.c (coverage_finish): Fix leak of da_file_name.
2000
13ef00fa
JH		 20012013-10-10 Jan Hubicka <jh@suse.cz>
2002
2003 * config/i386/x86-tune.def: Enable X86_TUNE_SSE_TYPELESS_STORES
2004 for generic, enable X86_TUNE_SSE_LOAD0_BY_PXOR for Bulldozer,
2005 Bobcat and generic.
2006
a1d50386
JJ		 20072013-10-10 Jakub Jelinek <jakub@redhat.com>
2008
2009 PR middle-end/58670
2010 * stmt.c (expand_asm_operands): Add FALLTHRU_BB argument,
2011 if any labels are in FALLTHRU_BB, use a special label emitted
2012 immediately after the asm goto insn rather than label_rtx
2013 of the LABEL_DECL.
2014 (expand_asm_stmt): Adjust caller.
2015 * cfgrtl.c (commit_one_edge_insertion): Force splitting of
2016 edge if the last insn in predecessor is a jump with single successor,
2017 but it isn't simplejump_p.
2018
e2c2fde2
RB		 20192013-10-10 Richard Biener <rguenther@suse.de>
2020
2021 PR tree-optimization/58656
2022 * tree-ssa-pre.c (phi_translate): Do not cache failed
2023 translations.
2024
de6a5608
AM		 20252013-10-10 Andrew MacLeod <amacleod@redhat.com>
2026
adcfd489
UB		 2027 * gimplify.c: Include expr.h and tm_p.h for targets with special
2028 va-arg padding requirements.
de6a5608 2029
b184c8f1
AM		 20302013-10-10 Andrew MacLeod <amacleod@redhat.com>
2031
2032 * tree-flow.h: Move some prototypes to gimple.h.
2033 (gimple_fold_indirect_ref): Move prototype to gimple-fold.h.
2034 * gimple.h: Relocate some prototypes from tree-flow.h
2035 * builtins.c (std_gimplify_va_arg_expr, build_va_arg_indirect_ref):
2036 Move to gimplify.c.
2037 * gimplify.c (gimple_fold_indirect_ref): Move to gimple-fold.c.
2038 (build_va_arg_indirect_ref): Relocate and make static.
2039 (std_gimplify_va_arg_expr): Relocate here.
2040 * gimple-fold.c (gimple_fold_indirect_ref): Relocate here.
2041 * gimple-fold.h (gimple_fold_indirect_ref): Add prototype.
2042
13b72c22
AK		 20432013-10-10 Andreas Krebbel <Andreas.Krebbel@de.ibm.com>
2044
2045 * doc/md.texi: Document the mnemonic attribute.
2046
5db40447
AK		 20472013-10-10 Andreas Krebbel <Andreas.Krebbel@de.ibm.com>
2048
2049 PR target/57377
2050 * gensupport.c (gen_mnemonic_attr): Handle (set (attr x) y) and
2051 (set_attr_alternative x ...) when searching for user defined
2052 mnemonic attribute.
2053
ed018d4b
AM		 20542013-10-10 Andrew MacLeod <amacleod@redhat.com>
2055
2056 * config/aplha/alpha.c: Add gimple-ssa.h to include list.
2057
4d5b5e9f
ER		 20582013-10-09 Easwaran Raman <eraman@google.com>
2059
2060 * params.def (PARAM_MIN_SIZE_FOR_STACK_SHARING): New param...
2061 * cfgexpand.c (defer_stack_allocation): ...use here
2062 * doc/invoke.texi: Add documentation for min-size-for-stack-sharing.
2063
c9ef86a1
ZC		 20642013-10-09 Zhenqiang Chen <zhenqiang.chen@arm.com>
2065
2066 * tree-ssa-phiopts.c (rhs_is_fed_for_value_replacement): New function.
2067 (operand_equal_for_value_replacement): New function, extracted from
2068 value_replacement and enhanced to catch more cases.
2069 (value_replacement): Use operand_equal_for_value_replacement.
2070
2431114f
AM		 20712013-10-09 Andrew MacLeod <amacleod@redhat.com>
2072
adcfd489 2073 * loop-doloop.c (doloop_modify, doloop_optimize): Use
2431114f
AM		 2074 get_max_loop_iterations.
2075
1dd2a9a0
KT		 20762013-10-09 Kyrylo Tkachov <kyrylo.tkachov@arm.com>
2077
2078 * config/arm/aarch-common.c (arm_early_load_addr_dep):
2079 Place comment above function.
2080
c1bf2a39
AM		 20812013-10-09 Andrew MacLeod <amacleod@redhat.com>
2082
2083 * tree-flow.h: Remove all remaining prototypes, enums and structs that
2084 are not related to tree-cfg.c.
2085 * tree-ssa-address.h: New file. Relocate prototypes.
2086 * tree-ssa-address.c: (struct mem_address): Relocate from tree-flow.h.
2087 (addr_for_mem_ref): New. Combine call to get_address_description and
2088 return addr_for_mem_ref.
2089 * expr.c (expand_expr_real_1): Use new addr_for_mem_ref routine.
2090 * tree-ssa-live.h: Adjust prototypes.
2091 * passes.c: Include tree-ssa-live.h.
2092 * gimple-pretty-print.h (gimple_dump_bb): Add prototype.
2093 * graphite.c (graphite_transform_loops): Make static.
2094 (graphite_transforms, gate_graphite_transforms, pass_data_graphite,
adcfd489 2095 make_pass_graphite, pass_data_graphite_transforms,
c1bf2a39
AM		 2096 make_pass_graphite_transforms): Relocate here from tree-ssa-loop.c.
2097 * ipa-pure-const.c (warn_function_noreturn): Make static.
2098 (execute_warn_function_noreturn, gate_warn_function_noreturn,
2099 class pass_warn_function_noreturn, make_pass_warn_function_noreturn):
2100 Relocate from tree-cfg.c
2101 * tree-cfg.c (tree_node_can_be_shared, gimple_empty_block_p): Make
2102 static.
2103 (execute_warn_function_noreturn, gate_warn_function_noreturn,
2104 class pass_warn_function_noreturn, make_pass_warn_function_noreturn):
2105 Move to ipa-pure-const.c.
adcfd489
UB		 2106 (execute_fixup_cfg, class pass_fixup_cfg, make_pass_fixup_cfg):
2107 Relocate from tree-optimize.c.
c1bf2a39
AM		 2108 * tree-optimize.c (execute_fixup_cfg, class pass_fixup_cfg,
2109 make_pass_fixup_cfg): Move to tree-cfg.c.
2110 * tree-chrec.h: (enum ev_direction): Relocate here from tree-flow.h.
2111 Relocate some prototypes.
2112 * tree-data-ref.h (tree_check_data_deps) Add prototype.
2113 * tree-dump.c (dump_function_to_file): Remove prototype.
2114 Add tree-flow.h to the include file.
2115 * tree-dump.h: Remove prototype.
2116 * tree-parloops.h: New File. Add prototypes.
2117 * tree-parloops.c (gate_tree_parallelize_loops, tree_parallelize_loops,
2118 pass_data_parallelize_loops, make_pass_parallelize_loops): Relocate
2119 from tree-ssa-loop.c.
2120 * tree-predcom.c (run_tree_predictive_commoning,
2121 gate_tree_predictive_commoning, pass_data_predcom, make_pass_predcom):
2122 Relocate here from tree-ssa-loop.c.
adcfd489 2123 * tree-ssa-dom.c (tree_ssa_dominator_optimize) Don't call
c1bf2a39
AM		 2124 ssa_name_values.release ().
2125 * tree-ssa-threadedge.h: New File. Relocate prototypes here.
2126 (ssa_name_values): Relocate from tree-flow.h.
2127 * tree-ssa.h: Include tree-ssa-threadedge.h and tree-ssa-address.h.
2128 * tree-ssa-loop.c (run_tree_predictive_commoning,
2129 gate_tree_predictive_commoning, pass_data_predcom, make_pass_predcom,
2130 graphite_transforms, gate_graphite_transforms, pass_data_graphite,
2131 make_pass_graphite, pass_data_graphite_transforms,
2132 make_pass_graphite_transforms, gate_tree_parallelize_loops,
2133 tree_parallelize_loops, pass_data_parallelize_loops,
2134 make_pass_parallelize_loops): Move to other files.
2135 * tree-vectorizer.h (lpeel_tree_duplicate_loop_to_edge_cfg): Prototype
2136 moved here.
2137 * tree.h: Remove prototypes from tree-address.c.
2138
cc524fc7
AM		 21392013-10-09 Andrew MacLeod <amacleod@redhat.com>
2140
2141 * tree-flow.h (tm_restart_node, gimple_df): Move to gimple-ssa.h.
2142 (struct int_tree_map): Move to tree-hasher.h
2143 (SCALE, LABEL, PERCENT): Move to gimple.h
2144 * tree-flow-inline.h: Delete. Move functions to other files.
2145 (unmodifiable_var_p, ref_contains_array_ref): Unused, so delete.
2146 * gimple-ssa.h (tm_restart_node, gimple_df): Relocate from tree-flow.h.
2147 (gimple_in_ssa_p, gimple_vop): Relocate from tree-flow-inline.h
2148 * gimple.h (imple_stmt_max_uid, set_gimple_stmt_max_uid,
2149 inc_gimple_stmt_max_uid, get_lineno): Relocate from tree-flow-inline.h.
2150 (SCALE, LABEL, PERCENT): Relocate from tree-flow.h
2151 * tree-hasher.h: Don't include tree-flow.h.
2152 (struct int_tree_map): Relocate from tree-flow.h.
2153 * tree-sra.c (contains_view_convert_expr_p): Relocate from
2154 tree-flow-inline.h and make static.
adcfd489
UB		 2155 * tree-ssa-alias.h (ranges_overlap_p): Relocate from
2156 tree-flow-inline.h.
cc524fc7
AM		 2157 * tree-ssa-operands.c (gimple_ssa_operands): Relocate from
2158 tree-flow-inline.h and make static.
2159 * tree.h (is_global_var, may_be_aliased): Relocate from
2160 tree-flow-inline.h.
2161 * Makefile.in (GTFILES): Remove tree-flow.h and add gimple-ssa.h.
2162 * value-prof.c: No longer include tree-flow-inline.h.
2163 * tree-switch-conversion.c: No longer include tree-flow-inline.h.
2164
71343877
AM		 21652013-10-09 Andrew MacLeod <amacleod@redhat.com>
2166
2167 * tree-flow.h: Move some protoypes. Include new tree-ssa-loop.h.
2168 (struct affine_iv, struct tree_niter_desc): Move to tree-ssa-loop.h.
2169 (enum move_pos): Move to tree-ssa-loop-im.h
2170 * cfgloop.h: Move some prototypes.
2171 (gcov_type_to_double_int): relocate from tree-ssa-loop.niter.c.
2172 * tree-flow-inline.h (loop_containing_stmt): Move to tree-ssa-loop.h.
2173 * tree-ssa-loop.h: New File. Include other tree-ssa-loop-*.h files.
2174 (struct affine_iv, struct tree_niter_desc): Relocate from tree-flow.h.
2175 (loop_containing_stmt): Relocate from tree-flow-inline.h.
2176 * tree-ssa-loop-ch.c: (do_while_loop_p): Make static.
2177 * tree-ssa-loop-im.c (for_each_index): Move to tree-ssa-loop.c.
2178 (enum move_pos): Relocate here.
2179 (lsm_tmp_name_add, gen_lsm_tmp_name, get_lsm_tmp_name): Move to
2180 tree-ssa-loop.c.
2181 (execute_sm_if_changed_flag_set): Change get_lsm_tmp_name call.
adcfd489
UB		 2182 (tree_ssa_loop_im, gate_tree_ssa_loop_im, pass_data_lim,
2183 make_pass_lim): Relocate here from tree-ssa-loop.c.
2184 * tree-ssa-loop-ivcanon.c (tree_num_loop_insns): Move to
71343877
AM		 2185 tree-ssa-loop.c.
2186 (loop_edge_to_cancel, unloop_loops): Make static.
2187 (tree_ssa_loop_ivcanon, gate_tree_ssa_loop_ivcanon, pass_data_iv_canon,
2188 make_pass_iv_canon): Relocate from tree-ssa-loop.c.
2189 (tree_complete_unroll, gate_tree_complete_unroll,
2190 pass_data_complete_unroll, make_pass_complete_unroll): Relocate here.
2191 (tree_complete_unroll_inner, gate_tree_complete_unroll_inner,
2192 pass_data_complete_unrolli, make_pass_complete_unrolli): Relocate here.
2193 * tree-ssa-loop-ivopts.c: Remove local prototypes.
2194 (stmt_invariant_in_loop_p): Remove unused function.
2195 * tree-ssa-loop-ivopts.h: New file. Add prototypes.
2196 * tree-ssa-loop-manip.h: New file. Add prototypes.
2197 * tree-ssa-loop-niter.c (record_niter_bound): Move to cfgloop.c.
2198 (gcov_type_to_double_int): Move to cfgloop.h.
2199 (double_int_cmp, bound_index,
2200 estimate_numbers_of_iterations_loop): Make static.
2201 (estimated_loop_iterations): Factor out get_estimated_loop_iterations.
2202 (max_loop_iterations): Factor out get_max_loop_iterations.
2203 (estimated_loop_iterations_int, max_stmt_executions_int): Move to
2204 cfgloop.c.
2205 * tree-ssa-loop-niter.h: New file. Add prototypes.
2206 * tree-ssa-loop-prefetch.c (tree_ssa_loop_prefetch,
2207 gate_tree_ssa_loop_prefetch, pass_data_loop_prefetch,
2208 make_pass_loop_prefetch): Relocate from tree-ssa-loop.c.
2209 * tree-ssa-loop-unswitch.c (tree_ssa_loop_unswitch,
2210 gate_tree_ssa_loop_unswitch, pass_data_tree_unswitch,
2211 make_pass_tree_unswitch): Relocate from tree-ssa-loop.c.
2212 * tree-ssa-loop.c (tree_ssa_loop_im, gate_tree_ssa_loop_im,
2213 pass_data_lim, make_pass_lim): Move to tree-ssa-loop-im.c.
2214 (tree_ssa_loop_unswitch, gate_tree_ssa_loop_unswitch,
2215 pass_data_tree_unswitch, make_pass_tree_unswitch): Move.
2216 (tree_ssa_loop_ivcanon, gate_tree_ssa_loop_ivcanon, pass_data_iv_canon,
2217 make_pass_iv_canon, tree_complete_unroll, gate_tree_complete_unroll,
2218 pass_data_complete_unroll, make_pass_complete_unroll,
2219 tree_complete_unroll_inner, gate_tree_complete_unroll_inner,
2220 pass_data_complete_unrolli, make_pass_complete_unrolli): Move to
2221 tree-ssa-loop-ivcanon.c.
2222 (tree_ssa_loop_prefetch, gate_tree_ssa_loop_prefetch,
2223 pass_data_loop_prefetch, make_pass_loop_prefetch): Move to
2224 tree-ssa-loop-prefetch.c.
2225 (for_each_index, lsm_tmp_name_add, gen_lsm_tmp_name): Relocate from
2226 tree-ssa-loop-im.c.
2227 (get_lsm_tmp_name): Relocate and add suffix parameter.
2228 (tree_num_loop_insns): Relocate from tree-ssa-ivcanon.c.
2229 * tree-scalar-evolution.h (simple_iv): Don't use affive_iv typedef.
2230 * cfgloop.c (record_niter_bound, estimated_loop_iterations_int,
2231 max_stmt_executions_int): Move from tree-ssa-loop-niter.c.
adcfd489 2232 (get_estimated_loop_iterations): Factor out accessor from
71343877 2233 estimated_loop_iterations in tree-ssa-loop-niter.c.
adcfd489
UB		 2234 (get_max_loop_iterations): Factor out accessor from
2235 _max_loop_iterations in tree-ssa-niter.c.
71343877
AM		 2236 * loop-unroll.c (decide_unroll_constant_iterations,
2237 decide_unroll_runtime_iterations, decide_peel_simple,
2238 decide_unroll_stupid): Use new get_* accessors.
2239
826cacfe
MG		 22402013-10-09 Marc Glisse <marc.glisse@inria.fr>
2241
2242 PR tree-optimization/20318
2243 * doc/extend.texi (returns_nonnull): New function attribute.
2244 * fold-const.c (tree_expr_nonzero_warnv_p): Look for returns_nonnull
2245 attribute.
2246 * tree-vrp.c (gimple_stmt_nonzero_warnv_p): Likewise.
2247 (stmt_interesting_for_vrp): Accept all GIMPLE_CALL.
2248
378f8976
EB		 22492013-10-09 Eric Botcazou <ebotcazou@adacore.com>
2250
2251 PR middle-end/58570
2252 * tree-ssa-alias.c (nonoverlapping_component_refs_of_decl_p): Return
2253 false if both components are bitfields.
2254
0fe04f5c
AV		 22552013-10-09 Alex Velenko <Alex.Velenko@arm.com>
2256
2257 * config/aarch64/arm_neon.h (vclz_s8, vclz_s16, vclz_s32)
2258 (vclzq_s8, vclzq_s16, vclzq_s32, vclz_u8, vclz_u16, vclz_u32)
2259 (vclzq_u8, vclzq_u16, vclzq_u32): Replace ASM with C.
2260 * config/aarch64/aarch64.h
2261 (CLZ_DEFINED_VALUE_AT_ZERO): Macro fixed for clz.
2262 * config/aarch64/aarch64-simd-builtins.def
2263 (VAR1 (UNOP, clz, 0, v4si)): Replaced with iterator.
2264
bed9bae4
AV		 22652013-10-09 Alex Velenko <Alex.Velenko@arm.com>
2266
2267 * config/aarch64/arm_neon.h (vadd_f64, vsub_f64): Implementation added.
2268
7f3d8b19
AV		 22692013-10-09 Alex Velenko <Alex.Velenko@arm.com>
2270
2271 * config/aarch64/arm_neon.h (vdiv_f64): Added.
2272
7df625a6
AV		 22732013-10-09 Alex Velenko <Alex.Velenko@arm.com>
2274
2275 * config/aarch64/arm_neon.h (vneg_f32): Asm replaced with C.
2276 (vneg_f64): New intrinsic.
2277 (vneg_s8): Asm replaced with C.
2278 (vneg_s16): Likewise.
2279 (vneg_s32): Likewise.
2280 (vneg_s64): New intrinsic.
2281 (vnegq_f32): Asm replaced with C.
2282 (vnegq_f64): Likewise.
2283 (vnegq_s8): Likewise.
2284 (vnegq_s16): Likewise.
2285 (vnegq_s32): Likewise.
2286 (vnegq_s64): Likewise.
2287
2b86fca7
RL		 22882013-10-09 Renlin Li <Renlin.Li@arm.com>
2289
2290 * config/arm/arm.c (arm_output_mi_thunk): Use plus_constant.
2291
e179df83
AK		 22922013-10-09 Andreas Krebbel <Andreas.Krebbel@de.ibm.com>
2293
2294 * config/s390/s390.c (s390_register_info_stdarg_fpr): Remove
2295 packed stack special handling.
2296 (s390_frame_info, s390_emit_prologue, s390_emit_epilogue): Switch
2297 back to fixed stack slots for FPRs saved due to stdarg.
2298
74129172
AK		 22992013-10-09 Andreas Krebbel <Andreas.Krebbel@de.ibm.com>
2300
2301 * config/s390/s390.c (s390_frame_info): Restructure function.
2302
6455a49e
AK		 23032013-10-09 Andreas Krebbel <Andreas.Krebbel@de.ibm.com>
2304
2305 * config/s390/s390.c (struct s390_frame_layout): New field
2306 gpr_save_slots.
2307 (cfun_save_arg_fprs_p, cfun_gpr_save_slot): New macros.
2308 (s390_reg_clobbered_rtx, s390_regs_ever_clobbered): Change type of
2309 regs_ever_clobbered to char*.
adcfd489
UB		 2310 (s390_regs_ever_clobbered): Check crtl->saves_all_registers instead
2311 of cfun->has_nonlocal_label. Ignore frame related restore INSNs.
6455a49e
AK		 2312 (s390_register_info): Enable FPR save slots. Move/Copy some
2313 functionality into ...
2314 (s390_register_info_gprtofpr, s390_register_info_stdarg_fpr)
2315 (s390_register_info_stdarg_gpr, s390_optimize_register_info): New
2316 function.
2317 (s390_frame_info): Do gpr slot allocation here now. stdarg does
2318 not imply a stack frame.
2319 (s390_init_frame_layout): Remove variable clobbered_regs.
2320 (s390_update_register_info): Remove function.
2321 (s390_hard_regno_rename_ok): Call-saved regs without a save slot
2322 cannot be used for register renaming.
2323 (s390_hard_regno_scratch_ok): New function.
2324 (TARGET_HARD_REGNO_SCRATCH_OK): Define target hook.
2325 (s390_initial_elimination_offset): Change offset calculation of
2326 the return address pointer.
adcfd489 2327 (save_gprs): Deal with only r6 being saved from the call-saved regs.
6455a49e 2328 (restore_gprs): Set frame related flag.
adcfd489 2329 (s390_save_gprs_to_fprs, s390_restore_gprs_from_fprs): New functions.
6455a49e
AK		 2330 (s390_emit_prologue): Call s390_register_info instead of
2331 s390_update_frame_layout. Call s390_save_gprs_to_fprs.
2332 (s390_emit_epilogue): Call s390_restore_gprs_from_fprs.
2333 (s390_optimize_prologue): Call s390_optimize_register_info.
2334 Try to remove also FPR slot save/restore INSNs. Remove frame
2335 related flags from restore INSNs.
2336
59d96342
DD		 23372013-10-08 DJ Delorie <dj@redhat.com>
2338
0dc7adad
DD		 2339 * config/rl78/rl78-expand.md (movqi): use operands[] not operandN.
2340 (movhi): Likewise.
2341
59d96342
DD		 2342 * config/rl78/rl78.c (rl78_print_operand_1): Change %c to %C to
2343 avoid conflict with the MI use of %c.
2344 * config/rl78/rl78-real.md: change %c to %C throughout.
2345 * config/rl78/rl78-virt.md: Likewise.
2346
a81169d8
JH		 23472013-10-08 Jan Hubicka <jh@suse.cz>
2348
2349 * config/i386/i386.c (ix86_option_override_internal): Switch
2350 to SSE math for -ffast-math when target ISA supports SSE2.
2351
481d1b81
AM		 23522013-10-08 Andrew MacLeod <amacleod@redhat.com>
2353
2354 * tree-flow.h: Remove some prototypes.
2355 * tree.h: Remove some protypes, add a couple.
2356 * tree.c (using_eh_for_cleanups_flag, using_eh_for_cleanups,
2357 using_eh_for_cleanups_p): Add interface routines for front ends.
2358 * tree-eh.h: New file. Add protoptyes.
2359 * tree-eh.c (using_eh_for_cleanups_p, using_eh_for_cleanups): Delete.
2360 (add_stmt_to_eh_lp_fn): Make static.
2361 (lower_try_finally): Use new using_eh_for_cleanups_p.
2362 * emit-rtl.c: Include tree-eh.h.
2363 * gimple.h: Include tree-eh.h.
2364
0e6a0e48
MG		 23652013-10-08 Marc Glisse <marc.glisse@inria.fr>
2366
2367 PR tree-optimization/58480
2368 * tree-vrp.c (infer_nonnull_range): New function.
2369 (infer_value_range): Call infer_nonnull_range.
2370
ef6179d1
DC		 23712013-10-08 Dehao Chen <dehao@google.com>
2372
2373 PR tree-optimization/58619
2374 * tree-inline.c (copy_phis_for_bb): Combine location data
2375 only if non-null.
2376
07d964d5
ZC		 23772013-10-08 Zhenqiang Chen <zhenqiang.chen@linaro.org>
2378
2379 PR target/58423
2380 * config/arm/arm.c (arm_emit_ldrd_pop): Attach
2381 RTX_FRAME_RELATED_P on INSN.
2382
19e9b2a3
BS		 23832013-10-07 Bill Schmidt <wschmidt@linux.vnet.ibm.com>
2384
2385 * config/rs6000/rs6000.c (altivec_expand_vec_perm_const_le): New.
2386 (altivec_expand_vec_perm_const): Call it.
2387
0cf68694
BS		 23882013-10-07 Bill Schmidt <wschmidt@linux.vnet.ibm.com>
2389
2390 * config/rs6000/vector.md (mov<mode>): Emit permuted move
2391 sequences for LE VSX loads and stores at expand time.
2392 * config/rs6000/rs6000-protos.h (rs6000_emit_le_vsx_move): New
2393 prototype.
2394 * config/rs6000/rs6000.c (rs6000_const_vec): New.
2395 (rs6000_gen_le_vsx_permute): New.
2396 (rs6000_gen_le_vsx_load): New.
2397 (rs6000_gen_le_vsx_store): New.
2398 (rs6000_gen_le_vsx_move): New.
2399 * config/rs6000/vsx.md (*vsx_le_perm_load_v2di): New.
2400 (*vsx_le_perm_load_v4si): New.
2401 (*vsx_le_perm_load_v8hi): New.
2402 (*vsx_le_perm_load_v16qi): New.
2403 (*vsx_le_perm_store_v2di): New.
2404 (*vsx_le_perm_store_v4si): New.
2405 (*vsx_le_perm_store_v8hi): New.
2406 (*vsx_le_perm_store_v16qi): New.
2407 (*vsx_xxpermdi2_le_<mode>): New.
2408 (*vsx_xxpermdi4_le_<mode>): New.
2409 (*vsx_xxpermdi8_le_V8HI): New.
2410 (*vsx_xxpermdi16_le_V16QI): New.
2411 (*vsx_lxvd2x2_le_<mode>): New.
2412 (*vsx_lxvd2x4_le_<mode>): New.
2413 (*vsx_lxvd2x8_le_V8HI): New.
2414 (*vsx_lxvd2x16_le_V16QI): New.
2415 (*vsx_stxvd2x2_le_<mode>): New.
2416 (*vsx_stxvd2x4_le_<mode>): New.
2417 (*vsx_stxvd2x8_le_V8HI): New.
2418 (*vsx_stxvd2x16_le_V16QI): New.
2419
9520e1eb
RL		 24202013-10-07 Renlin Li <Renlin.Li@arm.com>
2421
2422 * config/arm/arm-cores.def (cortex-a53): Use cortex tuning.
2423
36c0bd4f
AK		 24242013-10-07 Andreas Krebbel <Andreas.Krebbel@de.ibm.com>
2425
2426 * config/s390/s390.c (s390_register_info): Make the call-saved FPR
2427 loop to work also for 31bit ABI.
2428 Save the stack pointer for frame_size > 0.
2429
ee163e72
AK		 24302013-10-07 Andreas Krebbel <Andreas.Krebbel@de.ibm.com>
2431
2432 * config/s390/s390.md ("tbegin", "tbegin_nofloat", "tbegin_retry")
2433 ("tbegin_retry_nofloat", "tend", "tabort", "tx_assist"): Remove
2434 constraint letters from expanders.
2435 ("tbegin_retry", "tbegin_retry_nofloat"): Change predicate of the
2436 retry count to general_operand.
2437 ("tabort"): Give operand 0 a mode.
2438 ("tabort_1"): Add mode and constraint letter for operand 0.
2439 * doc/extend.texi: Fix protoype of __builtin_non_tx_store.
2440
45d99234
JL		 24412013-10-04 Jeff Law <law@redhat.com>
2442
2443 * tree-ssa-threadedge.c: Fix some trailing whitespace problems.
2444
adcfd489
UB		 2445 * tree-ssa-threadedge.c (thread_through_normal_block): Broken
2446 out of ...
45d99234
JL		 2447 (thread_across_edge): Here. Call it.
2448
7c327f7b
CC		 24492013-10-04 Cary Coutant <ccoutant@google.com>
2450
2451 * dwarf2out.c (dw_sra_loc_expr): Release addr_table entries when
adcfd489 2452 discarding a location list expression (or a piece of one).
7c327f7b 2453
6867e128
JH		 24542013-10-03 Jan Hubicka <jh@suse.cz>
2455
65efa7e7
UB		 2456 * config/i386/i386.c (ix86_issue_rate): Pentium4, Nocona has issue
2457 rate of 2. Core2, Corei7 and Haswell has issue rate of 4.
3700bce3 2458 (ix86_adjust_cost): Remove Atom case; fix core2/corei7/Haswell case.
6867e128 2459
c7f36d55
JH		 24602013-10-03 Jan Hubicka <jh@suse.cz>
2461
65efa7e7 2462 * config/i386/i386.c (ix86_option_override_internal): Do not enable
c7f36d55
JH		 2463 accumulate-outgoing-args when producing unwind info.
2464
4f0bee4c
WM		 24652013-10-03 Wei Mi <wmi@google.com>
2466
2467 * lra-constraints.c (insert_move_for_subreg): New function
2468 extracted from simplify_operand_subreg.
2469 (simplify_operand_subreg): Add reload for paradoxical subreg.
2470
111c3f39
RX		 24712013-10-03 Rong Xu <xur@google.com>
2472
65efa7e7
UB		 2473 * ipa-inline-analysis.c (find_foldable_builtin_expect): Find
2474 the candidate of builtin_expect such that we should fix the
2475 size/time estimation.
2476 (estimate_function_body_sizes): Do the acutally size/time fix-up
2477 for builtin_expect.
111c3f39 2478
942df739
RX		 24792013-10-03 Rong Xu <xur@google.com>
2480
65efa7e7
UB		 2481 * predict.c (tree_predict_by_opcode): Get the probability
2482 for builtin_expect from param builtin_expect_probability.
2483 * params.def (BUILTIN_EXPECT_PROBABILITY): New parameter.
2484 * predict.def (PRED_BUILTIN_EXPECT_RELAXED): Fix comments.
2485 * doc/invoke.texi: Add documentation for builtin-expect-probability.
942df739 2486
2284b034
MG		 24872013-10-03 Marc Glisse <marc.glisse@inria.fr>
2488
2489 PR c++/19476
2490 * common.opt (fcheck-new): Moved from c.opt. Make it 'Common'.
2491 * calls.c (alloca_call_p): Use get_callee_fndecl.
2492 * fold-const.c (tree_expr_nonzero_warnv_p): Handle operator new.
2493 * tree-vrp.c (gimple_stmt_nonzero_warnv_p, stmt_interesting_for_vrp):
2494 Likewise.
2495 (vrp_visit_stmt): Remove duplicated code.
2496
0609bdf2
MM		 24972013-10-03 Michael Meissner <meissner@linux.vnet.ibm.com>
2498
2499 * config/rs6000/rs6000-builtin.def (XSRDPIM): Use floatdf2,
2500 ceildf2, btruncdf2, instead of vsx_* name.
2501
2502 * config/rs6000/vsx.md (vsx_add<mode>3): Change arithmetic
2503 iterators to only do V2DF and V4SF here. Move the DF code to
2504 rs6000.md where it is combined with SF mode. Replace <VSv> with
2505 just 'v' since only vector operations are handled with these insns
2506 after moving the DF support to rs6000.md.
2507 (vsx_sub<mode>3): Likewise.
2508 (vsx_mul<mode>3): Likewise.
2509 (vsx_div<mode>3): Likewise.
2510 (vsx_fre<mode>2): Likewise.
2511 (vsx_neg<mode>2): Likewise.
2512 (vsx_abs<mode>2): Likewise.
2513 (vsx_nabs<mode>2): Likewise.
2514 (vsx_smax<mode>3): Likewise.
2515 (vsx_smin<mode>3): Likewise.
2516 (vsx_sqrt<mode>2): Likewise.
2517 (vsx_rsqrte<mode>2): Likewise.
2518 (vsx_fms<mode>4): Likewise.
2519 (vsx_nfma<mode>4): Likewise.
2520 (vsx_copysign<mode>3): Likewise.
2521 (vsx_btrunc<mode>2): Likewise.
2522 (vsx_floor<mode>2): Likewise.
2523 (vsx_ceil<mode>2): Likewise.
2524 (vsx_smaxsf3): Delete scalar ops that were moved to rs6000.md.
2525 (vsx_sminsf3): Likewise.
2526 (vsx_fmadf4): Likewise.
2527 (vsx_fmsdf4): Likewise.
2528 (vsx_nfmadf4): Likewise.
2529 (vsx_nfmsdf4): Likewise.
2530 (vsx_cmpdf_internal1): Likewise.
2531
2532 * config/rs6000/rs6000.h (TARGET_SF_SPE): Define macros to make it
2533 simpler to select whether a target has SPE or traditional floating
2534 point support in iterators.
2535 (TARGET_DF_SPE): Likewise.
2536 (TARGET_SF_FPR): Likewise.
2537 (TARGET_DF_FPR): Likewise.
2538 (TARGET_SF_INSN): Macros to say whether floating point support
2539 exists for a given operation for expanders.
2540 (TARGET_DF_INSN): Likewise.
2541
2542 * config/rs6000/rs6000.c (Ftrad): New mode attributes to allow
2543 combining of SF/DF mode operations, using both traditional and VSX
2544 registers.
2545 (Fvsx): Likewise.
2546 (Ff): Likewise.
2547 (Fv): Likewise.
2548 (Fs): Likewise.
2549 (Ffre): Likewise.
2550 (FFRE): Likewise.
2551 (abs<mode>2): Combine SF/DF modes using traditional floating point
2552 instructions. Add support for using the upper DF registers with
2553 VSX support, and SF registers with power8-vector support. Update
2554 expanders for operations supported by both the SPE and traditional
2555 floating point units.
2556 (abs<mode>2_fpr): Likewise.
2557 (nabs<mode>2): Likewise.
2558 (nabs<mode>2_fpr): Likewise.
2559 (neg<mode>2): Likewise.
2560 (neg<mode>2_fpr): Likewise.
2561 (add<mode>3): Likewise.
2562 (add<mode>3_fpr): Likewise.
2563 (sub<mode>3): Likewise.
2564 (sub<mode>3_fpr): Likewise.
2565 (mul<mode>3): Likewise.
2566 (mul<mode>3_fpr): Likewise.
2567 (div<mode>3): Likewise.
2568 (div<mode>3_fpr): Likewise.
2569 (sqrt<mode>3): Likewise.
2570 (sqrt<mode>3_fpr): Likewise.
2571 (fre<Fs>): Likewise.
2572 (rsqrt<mode>2): Likewise.
2573 (cmp<mode>_fpr): Likewise.
2574 (smax<mode>3): Likewise.
2575 (smin<mode>3): Likewise.
2576 (smax<mode>3_vsx): Likewise.
2577 (smin<mode>3_vsx): Likewise.
2578 (negsf2): Delete SF operations that are merged with DF.
2579 (abssf2): Likewise.
2580 (addsf3): Likewise.
2581 (subsf3): Likewise.
2582 (mulsf3): Likewise.
2583 (divsf3): Likewise.
2584 (fres): Likewise.
2585 (fmasf4_fpr): Likewise.
2586 (fmssf4_fpr): Likewise.
2587 (nfmasf4_fpr): Likewise.
2588 (nfmssf4_fpr): Likewise.
2589 (sqrtsf2): Likewise.
2590 (rsqrtsf_internal1): Likewise.
2591 (smaxsf3): Likewise.
2592 (sminsf3): Likewise.
2593 (cmpsf_internal1): Likewise.
2594 (copysign<mode>3_fcpsgn): Add VSX/power8-vector support.
2595 (negdf2): Delete DF operations that are merged with SF.
2596 (absdf2): Likewise.
2597 (nabsdf2): Likewise.
2598 (adddf3): Likewise.
2599 (subdf3): Likewise.
2600 (muldf3): Likewise.
2601 (divdf3): Likewise.
2602 (fred): Likewise.
2603 (rsqrtdf_internal1): Likewise.
2604 (fmadf4_fpr): Likewise.
2605 (fmsdf4_fpr): Likewise.
2606 (nfmadf4_fpr): Likewise.
2607 (nfmsdf4_fpr): Likewise.
2608 (sqrtdf2): Likewise.
2609 (smaxdf3): Likewise.
2610 (smindf3): Likewise.
2611 (cmpdf_internal1): Likewise.
2612 (lrint<mode>di2): Use TARGET_<MODE>_FPR macro.
2613 (btrunc<mode>2): Delete separate expander, and combine with the
2614 insn and add VSX instruction support. Use TARGET_<MODE>_FPR.
2615 (btrunc<mode>2_fpr): Likewise.
2616 (ceil<mode>2): Likewise.
2617 (ceil<mode>2_fpr): Likewise.
2618 (floor<mode>2): Likewise.
2619 (floor<mode>2_fpr): Likewise.
2620 (fma<mode>4_fpr): Combine SF and DF fused multiply/add support.
2621 Add support for using the upper registers with VSX and
2622 power8-vector. Move insns to be closer to the define_expands. On
2623 VSX systems, prefer the traditional form of FMA over the VSX
2624 version, since the traditional form allows the target not to
2625 overlap with the inputs.
2626 (fms<mode>4_fpr): Likewise.
2627 (nfma<mode>4_fpr): Likewise.
2628 (nfms<mode>4_fpr): Likewise.
2629
5bea0c6c 26302013-10-03 Kyrylo Tkachov <kyrylo.tkachov@arm.com>
65efa7e7 2631 Richard Earnshaw <richard.earnshaw@arm.com>
5bea0c6c
KT		 2632
2633 * config/arm/aarch-common-protos.h (struct alu_cost_table): New.
2634 (struct mult_cost_table): Likewise.
2635 (struct mem_cost_table): Likewise.
2636 (struct fp_cost_table): Likewise.
2637 (struct vector_cost_table): Likewise.
2638 (cpu_cost_table): Likewise.
2639 * config/arm/arm.opt (mold-rts-costs): New option.
2640 (mnew-generic-costs): Likewise.
2641 * config/arm/arm.c (generic_extra_costs): New table.
2642 (cortexa15_extra_costs): Likewise.
2643 (arm_slowmul_tune): Use NULL as new costs.
2644 (arm_fastmul_tune): Likewise.
2645 (arm_strongarm_tune): Likewise.
2646 (arm_xscale_tune): Likewise.
2647 (arm_9e_tune): Likewise.
2648 (arm_v6t2_tune): Likewise.
2649 (arm_cortex_a5_tune): Likewise.
2650 (arm_cortex_a9_tune): Likewise.
2651 (arm_v6m_tune): Likewise.
2652 (arm_fa726te_tune): Likewise.
2653 (arm_cortex_a15_tune): Use cortex15_extra_costs.
2654 (arm_cortex_tune): Use generict_extra_costs.
2655 (shifter_op_p): New function.
2656 (arm_unspec_cost): Likewise.
2657 (LIBCALL_COST): Define.
2658 (arm_new_rtx_costs): New function.
2659 (arm_rtx_costs): Use arm_new_rtx_costs when core-specific
2660 table is available. Use old costs otherwise unless mnew-generic-costs
2661 is specified.
2662 * config/arm/arm-protos.h (tune_params): Add insn_extra_cost field.
2663 (cpu_cost_table): Declare.
2664
24c56925
MS		 26652013-10-03 Marcus Shawcroft <marcus.shawcroft@arm.com>
2666
2667 PR target/58460
2668 * config/aarch64/aarch64.md (*adds_mul_imm_<mode>)
2669 (*subs_mul_imm_<mode>)
2670 (*add_<shift>_<mode>, *add_<shift>_si_uxtw,*add_mul_imm_<mode>)
2671 (*sub_<shift>_<mode>)
2672 (*sub_<shift>_si_uxtw,*sub_mul_imm_<mode>, *sub_mul_imm_si_uxtw):
2673 Remove k constraint.
2674
2ab8f063
IB		 26752013-10-03 Ian Bolton <ian.bolton@arm.com>
2676
2677 * config/aarch64/aarch64.c (aarch64_secondary_reload): Remove legacy
2678 code.
2679 * config/aarch64/aarch64.md (reload_sp_immediate): Likewise.
2680
79221839
TJ		 26812013-10-02 Teresa Johnson <tejohnson@google.com>
2682
2683 * predict.c (probably_never_executed): New function.
2684 (probably_never_executed_bb_p): Invoke probably_never_executed.
2685 (probably_never_executed_edge_p): Ditto.
2686 * bb-reorder.c (find_rarely_executed_basic_blocks_and_crossing_edges):
2687 Treat profile insanities conservatively.
2688
d441afe0
JDA		 26892013-10-02 John David Anglin <danglin@gcc.gnu.org>
2690
2691 * config.gcc (hppa*64*-*-linux*): Don't add pa/t-linux to tmake_file.
2692
80f466c4
VM		 26932013-10-02 Vladimir Makarov <vmakarov@redhat.com>
2694
2695 * lra-constraints.c (process_alt_operand): Calculate scratch_p and
2696 use it. Use smaller increase for scratch. Don't increase reject
2697 for early clobber scratch.
2698 * lra-eliminations.c (eliminate_regs_in_insn): Remove all insns
2699 setting eliminated regs except setting fp from hfp.
2700 (lra_eliminate): Check lra_insn_recog_data on NULL.
2701
6e228b4b
MM		 27022013-10-02 Michael Meissner <meissner@linux.vnet.ibm.com>
2703
2704 PR target/58587
2705 * config/rs6000/rs6000-cpus.def (ISA_2_6_MASKS_SERVER): Turn off
65efa7e7 2706 setting -mvsx-timode by default until the underlying problem is fixed.
6e228b4b
MM		 2707 (RS6000_CPU, power7 defaults): Likewise.
2708
11b54a5a
UB		 27092013-10-02 Uros Bizjak <ubizjak@gmail.com>
2710
2711 * config/x-linux (host-linux.o): Remove header dependencies.
2712 Use $(COMPILE) and $(POSTCOMPILE).
2713 * config/t-linux-android (linux-android.o): Ditto.
2714
bbc02b69
UB		 27152013-10-02 Uros Bizjak <ubizjak@gmail.com>
2716
2717 * Makefile.in (expmed.o-warn): Remove.
2718
440917de
AM		 27192013-10-02 Andrew MacLeod <amacleod@redhat.com>
2720
2721 * graphite-scop-detection.c: Include tree-ssa-propagate,h.
2722 * graphite-sese-to-poly.c: Include tree-ssa-propagate.h.
2723
99206ca9
TJ		 27242013-10-02 Teresa Johnson <tejohnson@google.com>
2725
2726 * dojump.c (do_jump_1): Divide probability between
2727 both conditions of a TRUTH_ANDIF_EXPR/TRUTH_ORIF_EXPR.
2728
56e82b14
TT		 27292013-10-02 Tom Tromey <tromey@redhat.com>
2730
2731 * Makefile.in (DRIVER_DEFINES): Use $(if), not $(and).
2732
3d9c733e
AM		 27332013-10-02 Andrew MacLeod <amacleod@redhat.com>
2734
2735 * tree-flow.h: Remove some prototypes.
2736 * tree-ssa-dce.c (mark_virtual_operand_for_renaming,
2737 mark_virtual_phi_result_for_renaming): Move to tree-into-ssa.c.
2738 * tree-into-ssa.c (mark_virtual_operand_for_renaming,
2739 mark_virtual_phi_result_for_renaming): Relocate here.
2740 * tree-into-ssa.h: Add prototypes.
bbc02b69 2741 * tree-ssa-phiopt.c: (tree_ssa_phiopt_worker) Use
3d9c733e
AM		 2742 single_pred_before_succ_order.
2743 (blocks_in_phiopt_order): Rename and move to cfganal.c.
2744 (nonfreeing_call_p) Move to gimple.c.
2745 * cfganal.c (single_pred_before_succ_order): Move and renamed from
2746 tree-ssa-phiopt.c.
2747 * basic-block.h (single_pred_before_succ_order): Add prototype.
2748 * gimple.c (nonfreeing_call_p): Relocate here.
2749 * gimple.h: Add prototype.
2750 * tree-ssa-ifcombine.c: Include tree-ssa-phiopt.h.
2751 * tree-ssa-dom.h: New file. Relocate prototypes here.
2752 * tree-ssa.h: Include tree-ssa-dom.h.
2753
78cedfb1
UB		 27542013-10-02 Uros Bizjak <ubizjak@gmail.com>
2755
2756 * config/i386/x-i386 (driver-i386.o): Remove header dependencies.
2757 Use $(COMPILE) and $(POSTCOMPILE).
2758
2759 * config/alpha/x-alpha (driver-alpha.o): Ditto.
2760
744730a4
AM		 27612013-10-02 Andrew MacLeod <amacleod@redhat.com>
2762
2763 * tree-flow.h: Remove some prototypes.
2764 * gimple-fold.h: Add prototypes from gimple.h and tree-flow.h.
2765 * tree-ssa-propagate.h: Relocate prototypes from tree-flow.h.
2766 * tree-ssa-copy.c (may_propagate*, propagate_value, replace_exp,
2767 propagate_tree_value*): Move from here to...
2768 * tree-ssa-propagate.c (may_propagate*, propagate_value, replace_exp,
2769 propagate_tree_value*): Relocate here.
2770 * tree-ssa-propagate.h: Relocate prototypes from tree-flow.h.
2771 * gimple.h: Include gimple-fold.h, move prototypes into gimple-fold.h.
2772 * gimple-fold.c: Remove gimple-fold.h from include list.
2773 * tree-vrp.c: Remove gimple-fold.h from include list.
2774 * tree-ssa-sccvn.c: Remove gimple-fold.h from include list.
2775 * tree-ssa-ccp.c: Remove gimple-fold.h from include list.
2776 * tree-scalar-evolution.c: Add tree-ssa-propagate.h to include list.
2777 * tree-ssa-pre.c: Add tree-ssa-propagate.h to include list.
2778 * sese.c: Add tree-ssa-propagate.h to include list.
2779
826a536d
RB		 27802013-10-02 Richard Biener <rguenther@suse.de>
2781
2782 * tree-loop-distribution.c: Include tree-vectorizer.h for
2783 find_loop_location.
2784 (enum partition_kind): Remove PKIND_REDUCTION.
78cedfb1 2785 (struct partition_s): Remove has_writes member, add reduction_p member.
826a536d
RB		 2786 (partition_alloc): Adjust.
2787 (partition_builtin_p): Likewise.
2788 (partition_has_writes): Remove.
2789 (partition_reduction_p): New function.
2790 (partition_merge_into): Likewise.
2791 (generate_code_for_partition): Commonize builtin partition
2792 handling tail.
2793 (rdg_cannot_recompute_vertex_p): Remove.
2794 (already_processed_vertex_p): Likewise.
2795 (rdg_flag_vertex): Do not set has_writes.
2796 (classify_partition): Adjust.
2797 (rdg_build_partitions): Do not set has_writes, treat all
2798 partitions as useful.
78cedfb1 2799 (distribute_loop): Record number of library calls generated. Adjust.
826a536d
RB		 2800 (tree_loop_distribution): Report number of loops and library
2801 calls generated as opt-info.
2802
4b403ece
AM		 28032013-10-02 Andrew MacLeod <amacleod@redhat.com>
2804
2805 * tree-flow.h: Include new .h files. Move prototypes.
2806 * tree-cfgcleanup.h: New file. Add prototypes from tree-flow.h.
2807 * tree-dfa.h: New File. Add prototypes from tree-flow.h.
2808 (get_addr_base_and_unit_offset_1) Move from tree-flow-inline.h.
2809 * tree-pretty-print.h: Add prototypes from tree-flow.h.
2810 * tree-into-ssa.h: New File. Add prototypes from tree-flow.h.
2811 ({debug|dump}*): Move debugging prototypes out of tree-into-ssa.c.
2812 * tree-into-ssa.c ({debug|dump}*): Move prototypes to header file.
2813 * tree.h (get_ref_base_and_extent): Move prototype out.
78cedfb1 2814 * tree-flow-inline.h (get_addr_base_and_unit_offset_1): Move to
4b403ece
AM		 2815 tree-dfa.h.
2816 * gimple-low.h: New File. Add prototypes from tree-flow.h.
2817 * gimple-low.c (try_catch_may_fallthru, block_may_fallthru): Move to...
2818 * tree.c (try_catch_may_fallthru, block_may_fallthru): Here.
2819 * tree-scalar-evolution.c: Include tree.h.
2820 * sese.c: Include tree.h.
2821 * dumpfile.c: Move gimple-pretty-print.h include after tree.h.
2822 * dwarf2out.c: Include tree-dfa.h.
2823 * tree-chrec.c: Include tree.h.
2824 * tree-data-ref.c: Include tree.h.
2825
1d2151c6
YZ		 28262013-10-02 Yufeng Zhang <yufeng.zhang@arm.com>
2827
2828 * gimple-ssa-strength-reduction.c (backtrace_base_for_ref):
2829 Fix whitespace.
2830
441ad147
RO		 28312013-10-02 Rainer Orth <ro@CeBiTec.Uni-Bielefeld.DE>
2832
2833 * config/t-sol2 (sol2-c.o): Remove header dependencies.
2834 Use $(COMPILE) and $(POSTCOMPILE).
2835 (sol2-cxx.o): Likewise.
2836 (sol2-stubs.o): Likewise.
2837 (sol2.o): Likewise.
2838 * config/x-solaris (host-solaris.o): Likewise.
2839
2840 * config/sparc/t-sparc (sparc.o): Remove.
2841 (sparc-c.o): Remove header dependencies.
2842 Use $(COMPILE) and $(POSTCOMPILE).
2843 * config/sparc/x-sparc: Likewise.
2844
2deaf8b0
JR		 28452013-10-02 Joern Rennecke <joern.rennecke@embecosm.com>
2846
2847 * config/arc/arc-opts.h: Add 2013 to Copyright years.
2848 * config/arc/arc700.md: Likewise.
2849 * config/arc/arc-modes.def: Likewise.
2850 * config/arc/arc-simd.h: Likewise.
2851 * config/arc/t-arc-uClibc: Likewise.
2852 * config/arc/t-arc-newlib: Likewise.
2853
f6fe771a
RL		 28542013-10-02 Renlin Li <renlin.li@arm.com>
2855
78cedfb1
UB		 2856 * config/aarch64/aarch64.c (aarch64_expand_prologue): Use
2857 plus_constant.
f6fe771a
RL		 2858 (aarch64_expand_epilogue): Likewise.
2859
0916f876
YZ		 28602013-10-02 Bill Schmidt <wschmidt@linux.vnet.ibm.com>
2861 Yufeng Zhang <yufeng.zhang@arm.com>
2862
2863 * gimple-ssa-strength-reduction.c (legal_cast_p_1): Forward
2864 declaration.
2865 (backtrace_base_for_ref): Call get_unwidened with 'base_in' if
2866 'base_in' represent a conversion and legal_cast_p_1 holds; set
2867 'base_in' with the returned value from get_unwidened.
2868
ebfcd719
KT		 28692013-10-02 Kyrylo Tkachov <kyrylo.tkachov@arm.com>
2870
2871 * config/arm/arm.c (arm_legitimize_reload_address): Explain why
2872 plus_constant is not used.
2873
157ca3e9
WM		 28742013-10-01 Wei Mi <wmi@google.com>
2875
78cedfb1 2876 * config/i386/x86-tune.def (DEF_TUNE): Remove m_CORE_ALL.
157ca3e9
WM		 2877 * config/i386/i386.md: Add define_peephole2 to
2878 break partial reg stall for cvtss2sd/cvtsd2ss.
2879
cd4dd8f0
JR		 28802013-10-01 Joern Rennecke <joern.rennecke@embecosm.com>
2881
2882 * config/arc/arc.c (pass_arc_ifcvt::clone):
78cedfb1 2883 Update for ctxt_ -> m_ctxt change.
cd4dd8f0 2884
1465e5cf
JL		 28852013-10-01 Jeff Law <law@redhat.com>
2886
2887 * tree-ssa-threadupdate.c (struct redirection_data): Delete
2888 outgoing_edge and intermediate_edge fields. Instead store the path.
2889 (redirection_data::hash): Hash on the last edge's destination index.
2890 (redirection_data::equal): Check the entire thread path.
2891 (lookup_redirectio_data): Corresponding changes.
2892 (create_edge_and_update_destination_phis): Likewise.
2893 (thread_single_edge): Likewise.
2894
25c606cb 28952013-10-01 Joern Rennecke <joern.rennecke@embecosm.com>
0ccbc132 2896 Diego Novillo <dnovillo@google.com>
25c606cb
JR		 2897
2898 * config/arc/simdext.md (UNSPEC_ARC_SIMD_VLD32WH): Delete.
2899 (UNSPEC_ARC_SIMD_VLD32WL): Likewise.
2900 (vld32wh_insn, vld32wl_insn): Delete commented-out old
2901 versions of these patterns.
2902
0ccbc132
JR		 2903 * doc/extend.texi (long_call/medium_call/short_call): Typo fix.
2904 (__builtin_arc_aligned): Likewise.
2905
f55d4a20
JR		 2906 * config/arc/arc.md: Expand adc_0 comment stating the intended
2907 purpose and why it isn't ready.
2908 Replace commented out call_value_via_label_mixed with a
2909 plain comment about bl_s.
2910
5719867d 2911 * config/arc/arc.c (stdio.h): Don't include directly.
5719867d
JR		 2912 (arc_expand_epilogue): Remove [0]: Remove fp_restored_p.
2913 Remove if (1) condition.
2914 (arc_encode_section_info): Fix comment.
2915
6462fab0
JR		 29162013-10-01 Joern Rennecke <joern.rennecke@embecosm.com>
2917
2918 * config/arc/arc.c (arc_conditional_register_usage):
2919 Use ARC_FIRST_SIMD_VR_REG / ARC_LAST_SIMD_VR_REG.
2920 Also set reg_alloc_order for DMA config regs.
2921
5d5f6720
JR		 29222013-10-01 Joern Rennecke <joern.rennecke@embecosm.com>
2923 Jeremy Bennett <jeremy.bennett@embecosm.com>
2924
2925 * doc/install.texi (--with-cpu): Mention ARC.
2926 (arc-*-elf32): New paragraph.
2927 (arc-linux-uclibc): Likewise.
2928 * doc/md.texi (Machine Constraints): Add ARC part.
2929 * doc/invoke.texi: (menu): Add ARC Options.
2930 (Machine Dependent Options) <ARC Options>: Add synopsis.
2931 (node ARC Options): Add.
2932 * doc/extend.texi (long_call / short_call attribute): Add ARC.
2933 (ARC Built-in Functions): New section defining
2934 generic ARC built-in functions.
2935 (ARC SIMD Built-in Functions): New section defining SIMD specific
2936 built-in functions.
2937 (Declaring Attributes of Functions): Extended
2938 description of short_call and long_call attributes for ARC and
2939 added index entries.
2940
526b7aee
SV		 29412013-10-01 Saurabh Verma <saurabh.verma@codito.com>
2942 Ramana Radhakrishnan <ramana.radhakrishnan@codito.com>
2943 Joern Rennecke <joern.rennecke@embecosm.com>
2944 Muhammad Khurram Riaz <khurram.riaz@arc.com>
2945 Brendan Kehoe <brendan@zen.org>
2946 Michael Eager <eager@eagercon.com>
2947 Simon Cook <simon.cook@embecosm.com>
2948 Jeremy Bennett <jeremy.bennett@embecosm.com>
2949
2950 * config/arc, common/config/arc: New directories.
2951
53426f6c
JR		 29522013-10-01 Joern Rennecke <joern.rennecke@embecosm.com>
2953 Brendan Kehoe <brendan@zen.org>
2954 Simon Cook <simon.cook@embecosm.com>
2955
2956 * config.gcc (arc*-*-elf*, arc*-*-linux-uclibc*): New configurations.
2957
1df9f5a9
AM		 29582013-10-01 Andrew MacLeod <amacleod@redhat.com>
2959
2960 * tree-ssa-live.h (coalesce_ssa_name): Move Prototype to...
2961 * tree-ssa-coalesce.h: New. Move prototype to here.
2962 * tree-outof-ssa.h: Include tree-ssa-coalesce.h.
2963 * tree-ssa-coalesce.c: Include tree-outof-ssa.h.
2964 (gimple_can_coalesce_p): Move to...
2965 * gimple.c (gimple_can_coalesce_p): Here.
2966
80560f95
AM		 29672013-10-01 Andrew MacLeod <amacleod@redhat.com>
2968
2969 * tree-into-ssa.c (enum need_phi_state): Relocate from tree-flow.h.
2970 (dump_decl_set): Move to gimple.c.
2971 * gimple.h: Don't include tree-ssa-operands.h.
2972 (dump_decl_set): Add prototype.
2973 (gimple_vuse_op, gimple_vdef_op, update_stmt, update_stmt_if_modified):
2974 Move to gimple-ssa.h.
2975 (phi_ssa_name_p, phi_nodes, phi_nodes_ptr, gimple_phi_arg_def,
2976 gimple_phi_arg_def_ptr, gimple_phi_arg_edge, gimple_phi_arg_location,
2977 gimple_phi_arg_location_from_edge, gimple_phi_arg_set_location,
2978 gimple_phi_arg_has_location): Relocate from tree-flow-inline.h
2979 * gimple.c (walk_stmt_load_store_ops): Use gimple_phi_arg_def rather
2980 than PHI_ARG_DEF.
2981 (dump_decl_set): Relocate here.
2982 * gimple-ssa.h: New file.
2983 (gimple_vuse_op, gimple_vdef_op, update_stmt, update_stmt_if_modified):
2984 Relocate from gimple.h.
2985 * tree-cfg.c (has_zero_uses_1, single_imm_use_1): Move to...
2986 * tree-ssa-operands.c (swap_ssa_operands): Rename from
2987 swap_tree_operands and remove non-ssa path.
2988 (has_zero_uses_1, single_imm_use_1): Relocate from tree-cfg.c.
2989 * tree-ssa-reassoc.c (linearize_expr_tree, repropagate_negates): Use
2990 swap_ssa_operands.
2991 * tree-vect-loop.c (destroy_loop_vec_info, vect_is_slp_reduction,
2992 vect_is_simple_reduction_1): Use swap_ssa_operands.
2993 * tree-flow.h: Move various prototypes to tree-phinodes.h.
2994 (enum need_phi_state): Move to tree-into-ssa.c.
2995 (struct immediate_use_iterator_d, FOR_EACH_IMM_*,
2996 BREAK_FROM_IMM_USE_STMT): Move to ssa-iterators.h.
2997 (swap_tree_operands): Rename and move prototype to tree-ssa-operands.h.
2998 * tree-flow-inline.h (delink_imm_use, link_imm_use_to_list,
2999 link_imm_use, set_ssa_use_from_ptr, link_imm_use_stmt, relink_imm_use,
3000 relink_imm_use_stmt, end_readonly_imm_use_p, first_readonly_imm_use,
3001 next_readonly_imm_use, has_zero_uses, has_single_use, single_imm_use,
3002 num_imm_uses): Move to ssa-iterators.h.
3003 (get_use_from_ptr, get_def_from_ptr): Move to tree-ssa-operands.h
7ab4168e 3004 (gimple_phi_arg_imm_use_ptr, phi_arg_index_from_use): Move to
80560f95
AM		 3005 tree-phinodes.h.
3006 (op_iter_done, op_iter_next_def, op_iter_next_tree,
3007 clear_and_done_ssa_iter, op_iter_init, op_iter_init_use,
3008 op_iter_init_def, op_iter_init_tree, single_ssa_tree_operand,
3009 single_ssa_use_operand, single_ssa_def_operand, zero_ssa_operands,
3010 num_ssa_operands, delink_stmt_imm_use, single_phi_def,
3011 op_iter_init_phiuse, op_iter_init_phidef, end_imm_use_stmt_p,
3012 end_imm_use_stmt_traverse, move_use_after_head, link_use_stmts_after,
3013 first_imm_use_stmt, next_imm_use_stmt, first_imm_use_on_stmt,
3014 end_imm_use_on_stmt_p, next_imm_use_on_stmt): Move to ssa-iterators.h.
3015 (gimple_phi_arg_def, gimple_phi_arg_def_ptr, gimple_phi_arg_edge,
3016 gimple_phi_arg_location, gimple_phi_arg_location_from_edge,
3017 gimple_phi_arg_set_location, gimple_phi_arg_has_location, phi_nodes,
3018 phi_nodes_ptr, phi_ssa_name_p): Move to gimple.h.
3019 (set_phi_nodes): Move to tree-phinodes.h.
3020 * tree-ssa-operands.h (enum ssa_op_iter_type,
3021 struct ssa_operand_iterator_d, SSA_OP*, FOR_EACH_SSA*, SINGLE_SSA*,
3022 ZERO_SSA_OPERANDS, NUM_SSA_OPERANDS): Move to ssa-iterators.h.
3023 (dump_decl_set): Remove prototype.
3024 (get_use_from_ptr, get_def_from_ptr): Relocate from tree-flow.h.
3025 * tree-phinodes.h: New file. Move some prototypes from tree-flow.h.
3026 (set_phi_nodes): Relocate from tree-flow-inline.h.
3027 (gimple_phi_arg_imm_use_ptr, phi_arg_index_from_use): Relocate from
3028 tree-flow-inline.h
3029 * tree-ssa.h: Add tree-phinodes.h, gimple-ssa.h, ssa-iterators.h to
3030 include list. Temporarily add gimple.h to include list.
3031 * ssa-iterators.h: New file.
3032 (struct immediate_use_iterator_d, FOR_EACH_IMM_*,
3033 BREAK_FROM_IMM_USE_STMT): Relocate from tree-flow.h.
3034 (enum ssa_op_iter_type, struct ssa_operand_iterator_d, SSA_OP*,
3035 FOR_EACH_SSA*, SINGLE_SSA*, ZERO_SSA_OPERANDS, NUM_SSA_OPERANDS):
3036 Relocate from tree-ssa-operands.h.
3037 (delink_imm_use, link_imm_use_to_list, link_imm_use,
3038 set_ssa_use_from_ptr, link_imm_use_stmt, relink_imm_use,
3039 relink_imm_use_stmt, end_readonly_imm_use_p, first_readonly_imm_use,
3040 next_readonly_imm_use, has_zero_uses, has_single_use, single_imm_use,
3041 num_imm_uses, get_use_from_ptr, get_def_from_ptr,
3042 phi_arg_index_from_use, op_iter_done, op_iter_next_def,
3043 op_iter_next_tree, clear_and_done_ssa_iter, op_iter_init,
3044 op_iter_init_use, op_iter_init_def, op_iter_init_tree,
3045 single_ssa_tree_operand, single_ssa_use_operand, single_ssa_def_operand,
3046 zero_ssa_operands, num_ssa_operands, delink_stmt_imm_use,
3047 single_phi_def, op_iter_init_phiuse, op_iter_init_phidef,
3048 end_imm_use_stmt_p, end_imm_use_stmt_traverse, move_use_after_head,
3049 link_use_stmts_after, first_imm_use_stmt, next_imm_use_stmt,
3050 first_imm_use_on_stmt, end_imm_use_on_stmt_p, next_imm_use_on_stmt):
3051 Relocate from tree-flow-inline.h.
3052 * tree-outof-ssa.h: Change _SSAEXPAND_H macro to GCC_TREE_OUTOF_SSA_H.
3053
8da00d65
VP		 30542013-10-01 Vidya Praveen <vidyapraveen@arm.com>
3055
3056 * aarch64-simd.md
3057 (aarch64_<ANY_EXTEND:su><ADDSUB:optab>l2<mode>_internal): Rename to ...
3058 (aarch64_<ANY_EXTEND:su><ADDSUB:optab>l<mode>_hi_internal): ... this;
3059 Insert '\t' to output template.
3060 (aarch64_<ANY_EXTEND:su><ADDSUB:optab>l<mode>_lo_internal): New.
3061 (aarch64_saddl2<mode>, aarch64_uaddl2<mode>): Modify to call
3062 gen_aarch64_<ANY_EXTEND:su><ADDSUB:optab>l<mode>_hi_internal() instead.
3063 (aarch64_ssubl2<mode>, aarch64_usubl2<mode>): Ditto.
3064
e2ebe1c2
UB		 30652013-10-01 Uros Bizjak <ubizjak@gmail.com>
3066
3067 * doc/install.texi (Host/target specific installation notes for GCC):
3068 Put @anchor before @heading.
3069 * doc/gcc.texi (titlepage): Use @uref and http:// prefix for website.
3070 Use @email for email addresses.
3071
aee2d611
JL		 30722013-10-01 Jeff Law <law@redhat.com>
3073
3074 * tree-ssa-threadedge.c (thread_across_edge): Make path a pointer to
3075 a vec. Only delete the path if we create one without successfully
3076 registering a jump thread.
3077 * tree-ssa-threadupdate.h (register_jump_thread): Pass in path vector
3078 as a pointer.
3079 * tree-ssa-threadupdate.c (threaded_edges): Remove. No longer used
3080 (paths): New vector of jump threading paths.
3081 (THREAD_TARGET, THREAD_TARGET2): Remove accessor macros.
3082 (THREAD_PATH): New accessor macro for the entire thread path.
3083 (lookup_redirection_data): Get intermediate and final outgoing edge
3084 from the thread path.
3085 (create_edge_and_update_destination_phis): Copy the threading path.
3086 (ssa_fix_duplicate_block_edges): Get edges and block types from the
3087 jump threading path.
3088 (ssa_redirect_edges): Get edges and block types from the jump threading
3089 path. Free the path vector.
3090 (thread_block): Get edges from the jump threading path. Look at the
3091 entire path to see if we thread to a loop exit. If we cancel a jump
3092 thread request, then free the path vector.
3093 (thread_single_edge): Get edges and block types from the jump threading
3094 path. Free the path vector.
3095 (thread_through_loop_header): Get edges and block types from the jump
3096 threading path. Free the path vector.
3097 (mark_threaded_blocks): Iterate over the vector of paths and store
3098 the path on the appropriate edge. Get edges and block types from the
3099 jump threading path.
3100 (mark_threaded_blocks): Get edges and block types from the jump
3101 threading path. Free the path vector.
3102 (thread_through_all_blocks): Use the vector of paths rather than
3103 a vector of 3-edge sets.
3104 (register_jump_thread): Accept pointer to a path vector rather
3105 than the path vector itself. Store the path vector for later use.
3106 Simplify.
3107
966f97ac 31082013-10-01 Jakub Jelinek <jakub@redhat.com>
8da00d65 3109 Andreas Krebbel <Andreas.Krebbel@de.ibm.com>
966f97ac
JJ		 3110
3111 PR target/58574
3112 * config/s390/s390.c (s390_split_branches): Modify check for table
3113 jump insns.
3114 (s390_chunkify_start): Rearrange table jump insn check in order to
3115 deal with compare and branch insns correctly.
3116
3a323a38
KV		 31172013-10-01 Kugan Vivekanandarajah <kuganv@linaro.org>
3118
3119 PR target/58578
3120 Revert
3121 2013-04-05 Greta Yorsh <Greta.Yorsh@arm.com>
3122 * config/arm/arm.md (arm_ashldi3_1bit): define_insn into
3123 define_insn_and_split.
3124 (arm_ashrdi3_1bit,arm_lshrdi3_1bit): Likewise.
3125 (shiftsi3_compare): New pattern.
3126 (rrx): New pattern.
3127 * config/arm/unspecs.md (UNSPEC_RRX): New.
3128
90444831
AM		 31292013-10-01 Alan Modra <amodra@gmail.com>
3130
3131 * stmt.c (expand_asm_operands): Revert part of 2013-09-24 special
3132 casing inout operands.
3133
818625cf
RB		 31342013-10-01 Richard Biener <rguenther@suse.de>
3135
3136 PR tree-optimization/58553
3137 * tree-loop-distribution.c (struct partition_s): Add niter member.
3138 (classify_partition): Populate niter member for the partition
3139 and properly identify whether the relevant store happens before
3140 or after the loop exit.
3141 (generate_memset_builtin): Use niter member from the partition.
3142 (generate_memcpy_builtin): Likewise.
3143
30f641cd
RS		 31442013-09-30 Richard Sandiford <rdsandiford@googlemail.com>
3145
3146 * vec.h (vec_prefix, vec): Prefix member names with "m_".
3147 * vec.c (vec_prefix::calculate_allocation): Update accordingly.
3148
65d3284b
RS		 31492013-09-30 Richard Sandiford <rdsandiford@googlemail.com>
3150
3151 * basic-block.h (edge_list): Prefix member names with "m_".
3152 * context.h (context): Likewise.
3153 * domwalk.h (dom_walker): Likewise.
3154 * gengtype-state.c (s_expr_writer, state_writer): Likewise.
3155 * graphite-sese-to-poly.c (sese_dom_walker): Likewise.
3156 * hash-table.h (hash_table): Likewise.
3157 * machmode.h (bit_field_mode_iterator): Likewise.
3158 * pass_manager.h (pass_list): Likewise.
3159 * tree-into-ssa.c (mark_def_dom_walker): Likewise.
3160 * tree-pass.h (pass_data): Likewise.
3161 * tree-ssa-dom.c (dom_opt_dom_walker): Likewise.
3162 * tree-ssa-phiopt.c (nontrapping_dom_walker): Likewise,
3163 * tree-ssa-uncprop.c (uncprop_dom_walker): Likewise.
3164 * asan.c (pass_data_asan): Update accordingly.
3165 * cfganal.c (control_dependences::find_control_dependence): Likewise.
3166 (control_dependences::control_dependences): Likewise.
3167 (control_dependences::~control_dependences): Likewise.
3168 (control_dependences::~control_dependences): Likewise.
3169 (control_dependences::get_edges_dependent_on): Likewise.
3170 * cgraphbuild.c (pass_data_rebuild_cgraph_edges::clone): Likewise.
3171 (pass_data_remove_cgraph_callee_edges::clone): Likewise.
3172 * context.c (gcc::context::context): Likewise.
3173 * cprop.c (pass_rtl_cprop::clone): Likewise.
3174 * domwalk.c (dom_walker::walk): Likewise.
3175 * ipa-inline-analysis.c (pass_inline_parameters::clone): Likewise.
3176 * ipa-pure-const.c (pass_local_pure_const::clone): Likewise.
3177 * mode-switching.c (pass_mode_switching::clone): Likewise.
3178 * passes.c (opt_pass::opt_pass): Likewise.
3179 (pass_manager::pass_manager): Likewise.
3180 * predict.c (pass_strip_predict_hints::clone): Likewise.
3181 * recog.c (pass_data pass_data_peephole2::clone): Likewise.
3182 (pass_split_all_insns::clone): Likewise.
3183 * stor-layout.c (bit_field_mode_iterator::bit_field_mode_iterator):
3184 Likewise.
3185 (bit_field_mode_iterator::next_mode): Likewise.
3186 (bit_field_mode_iterator::prefer_smaller_modes): Likewise.
3187 * tree-cfg.c (pass_split_crit_edges::clone): Likewise.
3188 * tree-cfgcleanup.c (pass_merge_phi::clone): Likewise.
3189 * tree-complex.c (pass_lower_complex::clone): Likewise.
3190 * tree-eh.c (pass_cleanup_eh::clone): Likewise.
3191 * tree-object-size.c (pass_object_sizes::clone): Likewise.
3192 * tree-optimize.c (pass_fixup_cfg::clone): Likewise.
3193 * tree-ssa-ccp.c (pass_data_ccp::clone): Likewise.
3194 (pass_fold_builtins::clone): Likewise.
3195 * tree-ssa-copy.c (pass_data_copy_prop::clone): Likewise.
3196 * tree-ssa-copyrename.c (pass_rename_ssa_copies::clone): Likewise.
3197 * tree-ssa-dce.c (pass_dce::clone, pass_dce_loop::clone): Likewise.
3198 (pass_cd_dce::clone): Likewise.
3199 * tree-ssa-dom.c (pass_dominator::clone): Likewise.
3200 (pass_phi_only_cprop::clone): Likewise.
3201 * tree-ssa-dse.c (pass_dse::clone): Likewise.
3202 * tree-ssa-forwprop.c (pass_forwprop::clone): Likewise.
3203 * tree-ssa-loop.c (pass_lim::clone): Likewise.
3204 * tree-ssa-phiopt.c (pass_phiopt::clone): Likewise.
3205 * tree-ssa-pre.c (pass_fre::clone): Likewise.
3206 * tree-ssa-reassoc.c (pass_reassoc::clone): Likewise.
3207 * tree-ssa-uninit.c (pass_late_warn_uninitialized::clone): Likewise.
3208 * tree-tailcall.c (pass_tail_recursion::clone): Likewise.
3209 * tree-vect-generic.c (pass_lower_vector_ssa::clone): Likewise.
3210 * tree-vrp.c (pass_vrp::clone): Likewise.
3211 * tsan.c (pass_tsan::clone): Likewise.
3212
f66d0891
JJ		 32132013-09-30 Jakub Jelinek <jakub@redhat.com>
3214
d2a365a8
JJ		 3215 PR middle-end/58564
3216 * fold-const.c (tree_unary_nonnegative_warnv_p): Use
3217 INTEGRAL_TYPE_P (t) instead of TREE_CODE (t) == INTEGER_TYPE.
3218
f66d0891
JJ		 3219 PR middle-end/58564
3220 * fold-const.c (fold_ternary_loc): For A < 0 : <sign bit of A> : 0
3221 optimization, punt if sign_bit_p looked through any zero extension.
3222
05357ac3
TJ		 32232013-09-30 Teresa Johnson <tejohnson@google.com>
3224
3225 * tree-ssa-threadupdate.c (ssa_fix_duplicate_block_edges):
3226 Update redirected out edge count in joiner case.
3227 (ssa_redirect_edges): Common the joiner and non-joiner cases
3228 so that joiner case gets profile updates.
3229
ca406576
RB		 32302013-09-30 Richard Biener <rguenther@suse.de>
3231
3232 PR tree-optimization/58554
e2ebe1c2
UB		 3233 * tree-loop-distribution.c (classify_partition): Require
3234 unconditionally executed stores for memcpy and memset recognition.
ca406576
RB		 3235 (tree_loop_distribution): Calculate dominance info.
3236
92d649c4
VK		 32372013-09-30 Venkataramanan Kumar <venkataramanan.kumar@linaro.org>
3238
3239 * config/aarch64/aarch64.h (MCOUNT_NAME): Define.
3240 (NO_PROFILE_COUNTERS): Likewise.
3241 (PROFILE_HOOK): Likewise.
3242 (FUNCTION_PROFILER): Likewise.
3243 * config/aarch64/aarch64.c (aarch64_function_profiler): Remove.
3244
bd9534e2
IS		 32452013-09-30 Iain Sandoe <iain@codesourcery.com>
3246
3247 * config/rs6000/darwin.md (load_macho_picbase_si): Wrap machopic
3248 calls and defines in TARGET_MACHO conditional.
3249 (load_macho_picbase_di): Likewise.
3250 (reload_macho_picbase): Likewise.
3251 (reload_macho_picbase_si): Likewise.
3252 (reload_macho_picbase_di): Likewise.
3253 (nonlocal_goto_receiver): Likewise.
3254
4494fbc9
NC		 32552013-09-30 Nick Clifton <nickc@redhat.com>
3256
3257 * config/msp430/msp430.c (msp430x_names): New array. Lists MCUs
3258 that use the MSP430X ISA.
3259 (msp430_option_override): Scan -mmcu command line option for any
3260 MCU name that supports the MSP430X ISA.
3261 * config/msp430/t-msp430 (MULTILIB_MATCHES): Add matches for known
3262 -mmcu options which enable the MSP430X ISA.
3263
791d9044
RB		 32642013-09-30 Richard Biener <rguenther@suse.de>
3265
3266 PR middle-end/58532
3267 * tree-cfg.c (make_abnormal_goto_edges): Skip debug statements
3268 before looking for setjmp-like calls.
3269
e1c5c877
IS		 32702013-09-29 Iain Sandoe <iain@codesourcery.com>
3271
3272 PR target/10901
3273 * config/darwin-protos.h (machopic_get_function_picbase): New.
3274 * config/darwin.c (machopic_get_function_picbase): New.
3275 * config/rs6000/darwin.md (load_macho_picbase_si): Update picbase
3276 label for a new func. (load_macho_picbase_di): Likewise.
3277 (reload_macho_picbase): New expand.
3278 (reload_macho_picbase_si): New insn.
3279 (reload_macho_picbase_di): New insn.
3280 (nonlocal_goto_receiver): New define and split.
3281 * config/rs6000/rs6000.md (unspec enum): Add UNSPEC_RELD_MPIC.
3282 (unspecv enum): Add UNSPECV_NLGR.
3283
749278c5
IS		 32842013-09-29 Iain Sandoe <iain@codesourcery.com>
3285
3286 * config/rs6000/rs6000.c (rs6000_init_dwarf_reg_sizes_extra): Ensure
3287 that altivec registers are correctly sized on Darwin.
3288
2c43a51c
IS		 32892013-09-29 Iain Sandoe <iain@codesourcery.com>
3290
3291 * config/t-darwin (darwin.o, darwin-c.o, darwin-f.o,
3292 darwin-driver.o): Use COMPILE and POSTCOMPILE.
3293 * config/x-darwin (host-darwin.o): Likewise.
3294 * config/i386/x-darwin (host-i386-darwin.o): Likewise.
3295 * config/rs6000/x-darwin (host-ppc-darwin.o): Likewise.
3296 * config/rs6000/x-darwin64 (host-ppc64-darwin.o): Likewise.
3297
7a5911d3
UB		 32982013-09-29 Uros Bizjak <ubizjak@gmail.com>
3299
3300 * doc/invoke.texi: Fix usage of @tie{} command.
3301
3f67a73d
EB		 33022013-09-29 Eric Botcazou <ebotcazou@adacore.com>
3303
3304 * config/sparc/sync.md: Add peephole for consecutive memory barriers.
3305
9ac2f538
JH		 33062013-09-28 Jan Hubicka <jh@suse.cz>
3307
3308 * config/i386/x86-tune.def: Add documentation for each of the options;
3309 add whitespace.
3310
b0ff06da
JH		 33112013-09-28 Jan Hubicka <jh@suse.cz>
3312
7a5911d3
UB		 3313 * x86-tune.def (X86_TUNE_SSE_UNALIGNED_LOAD_OPTIMAL): Enable for
3314 generic.
b0ff06da
JH		 3315 (X86_TUNE_SSE_UNALIGNED_STORE_OPTIMAL): Likewise.
3316 (X86_TUNE_FOUR_JUMP_LIMIT): Drop for generic and buldozer.
3317 (X86_TUNE_PAD_RETURNS): Drop for buldozer chips.
3318 (X86_TUNE_AVOID_VECTOR_DECODE): Drop for generic.
3319 (X86_TUNE_REASSOC_FP_TO_PARALLEL): Enable for generic.
3320
c3284718
RS		 33212013-09-28 Richard Sandiford <rdsandiford@googlemail.com>
3322
3323 * alloc-pool.c, asan.c, auto-inc-dec.c, basic-block.h, bb-reorder.c,
3324 bitmap.c, bitmap.h, bt-load.c, builtins.c, calls.c, cfgcleanup.c,
3325 cfgexpand.c, cfghooks.c, cfgloop.c, cfgloopmanip.c, cfgrtl.c, cgraph.c,
3326 cgraph.h, cgraphbuild.c, cgraphclones.c, cgraphunit.c, collect2.c,
3327 combine-stack-adj.c, combine.c, compare-elim.c, context.c, context.h,
3328 cprop.c, cse.c, cselib.c, dbxout.c, dce.c, defaults.h, df-core.c,
3329 df-problems.c, df-scan.c, df.h, diagnostic.c, double-int.c, dse.c,
3330 dumpfile.c, dwarf2asm.c, dwarf2cfi.c, dwarf2out.c, emit-rtl.c,
3331 errors.c, except.c, expmed.c, expr.c, file-find.c, final.c,
3332 fixed-value.c, fold-const.c, function.c, fwprop.c, gcc-ar.c, gcc.c,
3333 gcov-io.c, gcov-io.h, gcov.c, gcse.c, genattr-common.c, genattr.c,
3334 genattrtab.c, genautomata.c, genconfig.c, genemit.c, genextract.c,
3335 genflags.c, gengenrtl.c, gengtype-state.c, gengtype.c, genmodes.c,
3336 genopinit.c, genoutput.c, genpeep.c, genpreds.c, genrecog.c,
3337 gensupport.c, ggc-common.c, ggc-page.c, gimple-fold.c, gimple-low.c,
3338 gimple-pretty-print.c, gimple-ssa-strength-reduction.c, gimple.c,
3339 gimple.h, godump.c, graphite-clast-to-gimple.c,
3340 graphite-optimize-isl.c, graphite-poly.h, graphite-sese-to-poly.c,
3341 graphite.c, haifa-sched.c, hash-table.c, hash-table.h, hwint.c,
3342 hwint.h, ifcvt.c, incpath.c, init-regs.c, input.h, intl.c, intl.h,
3343 ipa-cp.c, ipa-devirt.c, ipa-inline-analysis.c, ipa-inline.c,
3344 ipa-profile.c, ipa-pure-const.c, ipa-reference.c, ipa-split.c,
3345 ipa-utils.c, ipa.c, ira-build.c, ira.c, jump.c, loop-doloop.c,
3346 loop-init.c, loop-invariant.c, loop-iv.c, lower-subreg.c, lto-cgraph.c,
3347 lto-streamer-in.c, lto-streamer-out.c, lto-wrapper.c, mcf.c,
3348 mode-switching.c, modulo-sched.c, omp-low.c, optabs.c, opts.c,
3349 pass_manager.h, passes.c, plugin.c, postreload-gcse.c, postreload.c,
3350 predict.c, prefix.c, pretty-print.c, print-rtl.c, print-tree.c,
3351 profile.c, read-md.c, real.c, real.h, recog.c, ree.c, reg-stack.c,
3352 regcprop.c, reginfo.c, regmove.c, regrename.c, regs.h, regstat.c,
3353 reload1.c, reorg.c, rtl.c, rtl.h, rtlanal.c, sbitmap.c, sched-rgn.c,
3354 sdbout.c, sel-sched-ir.c, sel-sched.c, sparseset.c, stack-ptr-mod.c,
3355 statistics.c, stmt.c, stor-layout.c, store-motion.c, streamer-hooks.h,
3356 system.h, target-hooks-macros.h, targhooks.c, targhooks.h, toplev.c,
3357 tracer.c, trans-mem.c, tree-browser.c, tree-call-cdce.c, tree-cfg.c,
3358 tree-cfgcleanup.c, tree-complex.c, tree-data-ref.c, tree-data-ref.h,
3359 tree-eh.c, tree-emutls.c, tree-flow.h, tree-if-conv.c, tree-into-ssa.c,
3360 tree-iterator.c, tree-loop-distribution.c, tree-mudflap.c,
3361 tree-nested.c, tree-nomudflap.c, tree-nrv.c, tree-object-size.c,
3362 tree-optimize.c, tree-pass.h, tree-pretty-print.c, tree-profile.c,
3363 tree-scalar-evolution.c, tree-sra.c, tree-ssa-ccp.c,
3364 tree-ssa-coalesce.c, tree-ssa-copy.c, tree-ssa-copyrename.c,
3365 tree-ssa-dce.c, tree-ssa-dom.c, tree-ssa-dse.c, tree-ssa-forwprop.c,
3366 tree-ssa-ifcombine.c, tree-ssa-live.c, tree-ssa-loop-ch.c,
3367 tree-ssa-loop-im.c, tree-ssa-loop-ivopts.c, tree-ssa-loop-prefetch.c,
3368 tree-ssa-loop.c, tree-ssa-math-opts.c, tree-ssa-operands.c,
3369 tree-ssa-phiopt.c, tree-ssa-phiprop.c, tree-ssa-pre.c,
3370 tree-ssa-reassoc.c, tree-ssa-sink.c, tree-ssa-strlen.c,
3371 tree-ssa-structalias.c, tree-ssa-threadedge.c, tree-ssa-threadupdate.c,
3372 tree-ssa-uncprop.c, tree-ssa-uninit.c, tree-ssa.c, tree-ssanames.c,
3373 tree-stdarg.c, tree-switch-conversion.c, tree-tailcall.c,
3374 tree-vect-data-refs.c, tree-vect-generic.c, tree-vect-loop-manip.c,
3375 tree-vect-stmts.c, tree-vectorizer.c, tree-vectorizer.h, tree-vrp.c,
3376 tree.c, tree.h, tsan.c, tsystem.h, value-prof.c, var-tracking.c,
3377 varasm.c, vec.h, vmsdbgout.c, vtable-verify.c, web.c: Add missing
3378 whitespace before "(".
3379
c6285bd7
SL		 33802013-09-28 Sandra Loosemore <sandra@codesourcery.com>
3381
3382 * expr.h (extract_bit_field): Remove packedp parameter.
3383 * expmed.c (extract_fixed_bit_field): Remove packedp parameter
3384 from forward declaration.
3385 (store_split_bit_field): Remove packedp arg from calls to
3386 extract_fixed_bit_field.
3387 (extract_bit_field_1): Remove packedp parameter and packedp
3388 argument from recursive calls and calls to extract_fixed_bit_field.
3389 (extract_bit_field): Remove packedp parameter and corresponding
3390 arg to extract_bit_field_1.
3391 (extract_fixed_bit_field): Remove packedp parameter. Remove code
3392 to issue warnings.
3393 (extract_split_bit_field): Remove packedp arg from call to
3394 extract_fixed_bit_field.
3395 * expr.c (emit_group_load_1): Adjust calls to extract_bit_field.
3396 (copy_blkmode_from_reg): Likewise.
3397 (copy_blkmode_to_reg): Likewise.
3398 (read_complex_part): Likewise.
3399 (store_field): Likewise.
3400 (expand_expr_real_1): Likewise.
3401 * calls.c (store_unaligned_arguments_into_pseudos): Adjust call
3402 to extract_bit_field.
3403 * config/tilegx/tilegx.c (tilegx_expand_unaligned_load): Adjust
3404 call to extract_bit_field.
3405 * config/tilepro/tilepro.c (tilepro_expand_unaligned_load): Adjust
3406 call to extract_bit_field.
3407 * doc/invoke.texi (Code Gen Options): Remove mention of warnings
3408 and special packedp behavior from -fstrict-volatile-bitfields
3409 documentation.
3410
522d4efc
JBG		 34112013-09-27 Jan-Benedict Glaw <jbglaw@lug-owl.de>
3412
3413 * lra-eliminations.c (init_elim_table): Guard value_p.
3414
19cdb489
MM		 34152013-09-27 Michael Meissner <meissner@linux.vnet.ibm.com>
3416
3417 * config/rs6000/rs6000.c (rs6000_hard_regno_mode_ok): Allow
3418 DFmode, DImode, and SFmode in the upper VSX registers based on the
3419 -mupper-regs-{df,sf} flags. Fix wu constraint to be ALTIVEC_REGS
3420 if -mpower8-vector. Combine -mvsx-timode handling with the rest
3421 of the VSX register handling.
3422
3423 * config/rs6000/rs6000.md (f32_lv): Use %x0 for VSX regsters.
3424 (f32_sv): Likewise.
3425 (zero_extendsidi2_lfiwzx): Add support for loading into the
3426 Altivec registers with -mpower8-vector. Use wu/wv constraints to
3427 only do VSX memory options on Altivec registers.
3428 (extendsidi2_lfiwax): Likewise.
3429 (extendsfdf2_fpr): Likewise.
3430 (mov<mode>_hardfloat, SF/SD modes): Likewise.
3431 (mov<mode>_hardfloat32, DF/DD modes): Likewise.
3432 (mov<mode>_hardfloat64, DF/DD modes): Likewise.
3433 (movdi_internal64): Likewise.
3434
d6d11272
XDL		 34352013-09-27 Xinliang David Li <davidxl@google.com>
3436
3437 * opts.c (finish_options): Adjust parameters
3438 according to vect cost model.
3439 (common_handle_option): Set dynamic vect cost
3440 model for FDO.
3441 targhooks.c (default_add_stmt_cost): Compute stmt cost
3442 unconditionally.
3443 * tree-vect-loop.c (vect_estimate_min_profitable_iters):
3444 Use helper function.
3445 * tree-vectorizer.h (unlimited_cost_model): New function.
3446 * tree-vect-slp.c (vect_slp_analyze_bb_1): Use helper function.
3447 * tree-vect-data-refs.c (vect_peeling_hash_insert): Use helper
3448 function.
3449 (vect_enhance_data_refs_alignment): Ditto.
3450 * flag-types.h: New enum.
3451 * common/config/i386/i386-common.c (ix86_option_init_struct):
3452 No need to initialize vect_cost_model flag.
3453 * config/i386/i386.c (ix86_add_stmt_cost): Compute stmt cost
3454 unconditionally.
3455
ac1857a3
DN		 34562013-09-27 Diego Novillo <dnovillo@google.com>
3457
3458 * gimple.h (enum ssa_mode): Remove.
3459
0d0e2af6
PM		 34602013-09-27 Paulo Matos <pmatos@broadcom.com>
3461
3462 * cfgloop.h (number_of_loops): Fix typo in check for null.
3463
09dc585e
JJ		 34642013-09-27 Jakub Jelinek <jakub@redhat.com>
3465
3466 PR middle-end/58551
3467 * tree-cfg.c (move_sese_region_to_fn): Also move loops that
3468 are children of outermost saved_cfun's loop, and set it up to
3469 be moved to dest_cfun's outermost loop. Fix up num_nodes adjustments
3470 if loop != loop0 and SESE region contains bbs that belong to loop0.
3471
ec5a3504
RS		 34722013-09-27 Richard Sandiford <rdsandiford@googlemail.com>
3473
3474 * rtlanal.c (must_be_base_p, must_be_index_p): Delete.
3475 (binary_scale_code_p, get_base_term, get_index_term): New functions.
3476 (set_address_segment, set_address_base, set_address_index)
3477 (set_address_disp): Accept the argument unconditionally.
3478 (baseness): Remove must_be_base_p and must_be_index_p checks.
3479 (decompose_normal_address): Classify as much as possible in the
3480 main loop.
3481
f91aec98
RS		 34822013-09-27 Richard Sandiford <rdsandiford@googlemail.com>
3483
3484 * cse.c (count_reg_usage): Handle INT_LIST.
3485 * lra-eliminations.c (lra_eliminate_regs_1): Likewise.
3486 * reginfo.c (reg_scan_mark_refs): Likewise.
3487 * reload1.c (eliminate_regs_1): Likewise.
3488
a9195970
IS		 34892013-09-27 Iain Sandoe <iain@codesourcery.com>
3490
3491 PR middle-end/58547
3492 * rtlanal.c (lsb_bitfield_op_p): Make both parts of the comparison
3493 signed.
3494
e6f1c509
RB		 34952013-09-27 Richard Biener <rguenther@suse.de>
3496
3497 PR tree-optimization/58459
3498 * tree-ssa-forwprop.c (forward_propagate_addr_expr): Remove
3499 restriction not propagating into loops.
3500
84f48495
FW		 35012013-09-26 Florian Weimer <fw@deneb.enyo.de>
3502
3503 * tree-ssa.h (walk_use_def_chains_fn, walk_use_def_chains): Delete.
3504 * tree-ssa.c (walk_use_def_chains_1, walk_use_def_chains): Delete.
3505 * doc/tree-ssa.texi (Walking use-def chains): Delete.
3506
74fc8b8a
RB		 35072013-09-26 Richard Biener <rguenther@suse.de>
3508
7a5911d3 3509 * tree-into-ssa.c (rewrite_into_ssa): Make more SSA names to anonymous.
74fc8b8a 3510
b4ada065
RB		 35112013-09-26 Richard Biener <rguenther@suse.de>
3512
3513 * alias.h (component_uses_parent_alias_set): Rename to ...
3514 (component_uses_parent_alias_set_from): ... this.
3515 * alias.c (component_uses_parent_alias_set): Rename to ...
3516 (component_uses_parent_alias_set_from): ... this and return
3517 the desired parent.
3518 (reference_alias_ptr_type_1): Use the result from
3519 component_uses_parent_alias_set_from instead of stripping
3520 components one at a time.
3521 * emit-rtl.c (set_mem_attributes_minus_bitpos): Adjust.
3522
78bca40d
AM		 35232013-09-26 Andrew MacLeod <amacleod@redhat.com>
3524
7a5911d3
UB		 3525 * tree-ssa-live.h (find_replaceable_exprs, dump_replaceable_exprs):
3526 Move prototypes to...
78bca40d
AM		 3527 * tree-ssa-ter.h: New File. Move prototypes here.
3528 * tree-flow.h (stmt_is_replaceable_p): Remove prototype.
3529 * tree-outof-ssa.h: New. Rename ssaexpand.h, include tree-ssa-ter.h.
3530 * tree-outof-ssa.c (ssa_is_replaceable_p): New. Refactor common bits
7a5911d3 3531 from is_replaceable_p.
78bca40d
AM		 3532 * tree-ssa-ter.c (is_replaceable_p, stmt_is_replaceable_p): Delete.
3533 (ter_is_replaceable_p): New. Use new refactored ssa_is_replaceable_p.
3534 (process_replaceable): Use ter_is_replaceable_p.
3535 (find_replaceable_in_bb): Use ter_is_replaceable_p.
3536 * expr.c (stmt_is_replaceable_p): Relocate from tree-ssa-ter.c. Use
3537 newly refactored ssa_is_replaceable_p.
3538 * cfgexpand.c: Include tree-outof-ssa.h.
3539 * ssaexpand.h: Delete.
3540
ff2a63a7
AM		 35412013-09-26 Andrew MacLeod <amacleod@redhat.com>
3542
3543 * gimple.c (gimple_replace_lhs): Move to tree-ssa.c and rename.
3544 (struct count_ptr_d, count_ptr_derefs, count_uses_and_derefs): Move to
3545 tree-ssa.c
3546 (create_gimple_tmp): Delete.
3547 (get_expr_type, build_assign, build_type_cast): Move to...
3548 * gimple-builder.c: New File.
3549 (get_expr_type): Relocate from gimple.c.
3550 (build_assign, build_type_cast): Change to only create ssanames.
3551 * gimple.h: Move prototypes to...
3552 * gimple-builder.h: New File. Here.
3553 * tree-ssa.h: And here.
3554 * tree-ssa.c (struct count_ptr_d, count_ptr_derefs,
3555 count_uses_and_derefs): Relocate from gimple.c.
3556 (gimple_replace_ssa_lhs): Renamed gimple_replace_ssa from gimple.c
3557 * tree-ssa-reassoc.c (repropagate_negates): Use gimple_replace_ssa_lhs.
3558 * tree-ssa-math-opts (execute_cse_reciprocals): Use
3559 gimple_replace_ssa_lhs.
3560 * asan.c: Include gimple-builder.h.
3561 * Makefile.in: Add gimple-builder.o.
3562
a2544177
RB		 35632013-09-26 Richard Biener <rguenther@suse.de>
3564
3565 * tree-ssa-live.c (var_map_base_init): Handle SSA names with
3566 DECL_IGNORED_P base VAR_DECLs like anonymous SSA names.
3567 (loe_visit_block): Use gcc_checking_assert.
3568 * tree-ssa-coalesce.c (create_outofssa_var_map): Use
3569 gimple_assign_ssa_name_copy_p.
3570 (gimple_can_coalesce_p): Adjust according to the var_map_base_init
3571 change.
3572
6d1ca9a0
DE		 35732013-09-26 David Edelsohn <dje.gcc@gmail.com>
3574
3575 * config/rs6000/t-rs6000 (rs6000.o): Remove.
3576 (rs6000-c.o): Use COMPILE and POSTCOMPILE.
3577
46e950db
RB		 35782013-09-26 Richard Biener <rguenther@suse.de>
3579
3580 PR tree-optimization/58539
3581 * tree-vect-loop.c (vect_create_epilog_for_reduction): Honor
3582 the fact that debug statements are not taking part in loop-closed
3583 SSA construction.
3584
40ada30a
NC		 35852013-09-26 Nick Clifton <nickc@redhat.com>
3586
3587 * config/msp430/msp430.c (msp430_expand_epilogue): Fix compile
3588 time warning message.
3589 (msp430_print_operand_raw): Delete unused letter parameter.
3590 (TARGET_PRINT_OPERAND_ADDRESS): Define.
3591 (msp430_print_operand_address): New function.
3592 (msp430_print_operand): Move address printing code from here to
3593 new function.
3594 * config/msp430/msp430.md (movsipsi2): Add comment in generated
3595 assembler.
3596 (zero_extendpsisi2): Likewise.
3597 (extendpsisi2): New pattern.
3598 (andneghi3): New pattern.
3599
38e8f663
YR		 36002013-09-26 Yvan Roux <yvan.roux@linaro.org>
3601
3602 * config/aarch64/aarch64.opt (mlra): New option.
3603 * config/aarch64/aarch64.c (aarch64_lra_p): New function.
3604 (TARGET_LRA_P): Define.
3605
4167e8d4
EB		 36062013-09-26 Eric Botcazou <ebotcazou@adacore.com>
3607
3608 * expr.c (expand_assignment): Remove obsolete comment.
3609
5254eac4
JL		 36102013-09-25 Jeff Law <law@redhat.com>
3611
3612 * tree-flow.h (thread_through_all_blocks): Prototype moved into
3613 tree-ssa-threadupdate.h.
3614 (register_jump_thread): Similarly.
3615 * tree-ssa-threadupdate.h: New header file.
3616 * tree-ssa-dom.c: Include tree-ssa-threadupdate.h.
3617 * tree-vrp.c: Likewise.
3618 * tree-ssa-threadedge.c: Include tree-ssa-threadupdate.h.
3619 (thread_around_empty_blocks): Change type of path vector argument to
3620 an edge,type pair from just an edge. Initialize both elements when
3621 appending to a jump threading path. Tweak references to elements
3622 appropriately.
3623 (thread_across_edge): Similarly. Release memory for the elements
3624 as needed.
3625 * tree-ssa-threadupdate.c: Include tree-ssa-threadupdate.h.
3626 (dump_jump_thread_path): New function broken out from
3627 register_jump_thread.
3628 (register_jump_thread): Use dump_jump_thread_path. Change type of
3629 path vector entries. Search the path for NULL edges and dump
3630 the path if one is found. Tweak the conversion of path to 3-edge
3631 form to use the block copy type information embedded in the path.
3632
20f114a3
YR		 36332013-09-25 Yvan Roux <yvan.roux@linaro.org>
3634
3635 * lra.c (update_inc_notes): Remove all REG_DEAD and REG_UNUSED notes.
3636
3936bafc
YR		 36372013-09-25 Yvan Roux <yvan.roux@linaro.org>
3638 Vladimir Makarov <vmakarov@redhat.com>
3639
3640 * rtlanal.c (lsb_bitfield_op_p): New predicate for bitfield operations
3641 from the least significant bit.
3642 (strip_address_mutations): Add bitfield operations handling.
3643 (must_be_index_p): Add shifting and rotate operations handling.
3644 (set_address_base): Use must_be_base_p predicate.
4167e8d4 3645 (set_address_index): Use must_be_index_p predicate.
3936bafc 3646
b86f6e9e
AI		 36472013-09-25 Alexander Ivchenko <alexander.ivchenko@intel.com>
3648 Maxim Kuznetsov <maxim.kuznetsov@intel.com>
3649 Sergey Lega <sergey.s.lega@intel.com>
3650 Anna Tikhonova <anna.tikhonova@intel.com>
3651 Ilya Tocar <ilya.tocar@intel.com>
3652 Andrey Turetskiy <andrey.turetskiy@intel.com>
3653 Ilya Verbin <ilya.verbin@intel.com>
3654 Kirill Yukhin <kirill.yukhin@intel.com>
3655 Michael Zolotukhin <michael.v.zolotukhin@intel.com>
3656
3657 * config/i386/i386.c (ix86_avx256_split_vector_move_misalign):
3658 Use new names.
3659 (ix86_expand_vector_move_misalign): Support new unaligned load and
3660 stores and use new names.
3661 (CODE_FOR_sse2_storedqu): Rename to ...
3662 (CODE_FOR_sse2_storedquv16qi): ... this.
3663 (CODE_FOR_sse2_loaddqu): Rename to ...
3664 (CODE_FOR_sse2_loaddquv16qi): ... this.
3665 (CODE_FOR_avx_loaddqu256): Rename to ...
3666 (CODE_FOR_avx_loaddquv32qi): ... this.
3667 (CODE_FOR_avx_storedqu256): Rename to ...
3668 (CODE_FOR_avx_storedquv32qi): ... this.
3669 * config/i386/i386.md (fpint_logic): New.
3670 * config/i386/sse.md (VMOVE): Extend for AVX512.
3671 (VF): Ditto.
3672 (VF_128_256): New.
3673 (VF_512): Ditto.
3674 (VI_UNALIGNED_LOADSTORE): Ditto.
3675 (sse2_avx_avx512f): Ditto.
3676 (sse2_avx2): Extend for AVX512.
3677 (sse4_1_avx2): Ditto.
3678 (avx2_avx512f): New.
3679 (sse): Extend for AVX512.
3680 (sse2): Ditto.
3681 (sse4_1): Ditto.
3682 (avxsizesuffix): Ditto.
3683 (sseintvecmode): Ditto.
3684 (ssePSmode): Ditto.
3685 (<sse>_loadu<ssemodesuffix><avxsizesuffix>): Ditto.
3686 (<sse>_storeu<ssemodesuffix><avxsizesuffix>): Ditto.
3687 (<sse2>_loaddqu<avxsizesuffix>): Extend for AVX512 and rename to ...
3688 (<sse2_avx_avx512f>_loaddqu<mode>): ... this.
3689 (<sse2>_storedqu<avxsizesuffix>): Extend for AVX512 and rename to ...
3690 (<sse2_avx_avx512f>_storedqu<mode): ... this.
3691 (<sse>_movnt<mode>): Replace constraint "x" with "v".
3692 (STORENT_MODE): Extend for AVX512.
3693 (*absneg<mode>2): Replace constraint "x" with "v".
3694 (*mul<mode>3): Ditto.
3695 (*ieee_smin<mode>3): Ditto.
3696 (*ieee_smax<mode>3): Ditto.
3697 (avx_cmp<mode>3): Replace VF with VF_128_256.
3698 (*<sse>_maskcmp<mode>3_comm): Ditto.
3699 (<sse>_maskcmp<mode>3): Ditto.
3700 (<sse>_andnot<mode>3): Extend for AVX512.
3701 (<code><mode>3, anylogic): Replace VF with VF_128_256.
3702 (<code><mode>3, fpint_logic): New.
3703 (*<code><mode>3): Extend for AVX512.
3704 (avx512flogicsuff): New.
3705 (avx512f_<logic><mode>): Ditto.
3706 (<sse>_movmsk<ssemodesuffix><avxsizesuffix>): Replace VF with
3707 VF_128_256.
3708 (<sse4_1>_blend<ssemodesuffix><avxsizesuffix>): Ditto.
3709 (<sse4_1>_blendv<ssemodesuffix><avxsizesuffix>): Ditto.
3710 (<sse4_1>_dp<ssemodesuffix><avxsizesuffix>): Ditto.
3711 (avx_vtest<ssemodesuffix><avxsizesuffix>): Ditto.
3712 (<sse4_1>_round<ssemodesuffix><avxsizesuffix>): Ditto.
3713 (xop_vpermil2<mode>3): Ditto.
3714 (*avx_vpermilp<mode>): Extend for AVX512 and rename to ...
3715 (*<sse2_avx_avx512f>_vpermilp<mode>): ... this.
3716 (avx_vpermilvar<mode>3): Extend for AVX512 and rename to ...
3717 (<sse2_avx_avx512f>_vpermilvar<mode>3): ... this.
3718
4d44d03c
TT		 37192013-09-25 Tom Tromey <tromey@redhat.com>
3720
3721 * Makefile.in (PARTITION_H, LTO_SYMTAB_H, COMMON_TARGET_DEF_H)
3722 (RTL_ERROR_H, TRANS_MEM_H, COVERAGE_H, DEMANGLE_H, ALIAS_H)
3723 (SCHED_INT_H, SEL_SCHED_IR_H, SEL_SCHED_DUMP_H, VALTRACK_H, DDG_H)
3724 (GGC_INTERNAL_H, DECNUM_H, BACKTRACE_H, MKDEPS_H, TREE_HASHER_H)
3725 (TREE_SSA_LIVE_H, SSAEXPAND_H, DWARF2OUT_H, SCEV_H, OMEGA_H)
3726 (TREE_DATA_REF_H, IRA_INT_H, LRA_INT_H, DBGCNT_H, DATA_STREAMER_H)
3727 (GIMPLE_STREAMER_H, TREE_STREAMER_H, STREAMER_HOOKS_H)
3728 (TREE_VECTORIZER_H, IPA_INLINE_H, GSTAB_H, LIBFUNCS_H)
3729 (GRAPHITE_HTAB_H): Remove.
3730
86aaf68d
TT		 37312013-09-25 Tom Tromey <tromey@redhat.com>
3732
3733 * config/mcore/t-mcore (CROSS_FLOAT_H): Remove.
3734
c987deb8
TT		 37352013-09-25 Tom Tromey <tromey@redhat.com>
3736
3737 * config/t-glibc (glibc-c.o): Use COMPILE and POSTCOMPILE.
3738
9439747e
TT		 37392013-09-25 Tom Tromey <tromey@redhat.com>
3740
3741 * config/i386/t-i386 (i386.o): Remove.
3742 (i386-c.o): Use COMPILE and POSTCOMPILE.
3743
aefc31a1
TT		 37442013-09-25 Tom Tromey <tromey@redhat.com>
3745
3746 * Makefile.in ($(out_object_file)): Use COMPILE and POSTCOMPILE.
3747
b6541edc
TT		 37482013-09-25 Tom Tromey <tromey@redhat.com>
3749
3750 * Makefile.in (graph.o, sbitmap.o, sparseset.o, gcc-ar.o)
3751 (gcc-ranlib.o, gcc-nm.o, collect2.o, collect2-aix.o, tlink.o)
3752 (lto-wrapper.o, default-c.o, attribs.o, incpath.o, prefix.o)
3753 (gcc.o, options.o, options-save.o, version.o, gtype-desc.o)
3754 (trans-mem.o, ggc-common.o, ggc-page.o, ggc-none.o, stringpool.o)
3755 (convert.o, double-int.o, lto-compress.o, data-streamer-in.o)
3756 (data-streamer-out.o, data-streamer.o, gimple-streamer-in.o)
3757 (gimple-streamer-out.o, tree-streamer.o, tree-streamer-in.o)
3758 (tree-streamer-out.o, streamer-hooks.o, lto-cgraph.o)
3759 (lto-streamer-in.o, lto-streamer-out.o, lto-section-in.o)
3760 (lto-section-out.o, lto-opts.o, lto-streamer.o, langhooks.o)
3761 (test-dump.o, tree.o, tree-dump.o, tree-inline.o, print-tree.o)
3762 (stor-layout.o, asan.o, tsan.o, ubsan.o, tree-ssa-tail-merge.o)
3763 (tree-ssa-structalias.o, tree-ssa-uninit.o, tree-ssa.o)
3764 (tree-into-ssa.o, tree-ssa-ter.o, tree-ssa-coalesce.o)
3765 (tree-outof-ssa.o, tree-ssa-dse.o, tree-ssa-forwprop.o)
3766 (tree-ssa-phiprop.o, tree-ssa-ifcombine.o, tree-ssa-phiopt.o)
3767 (tree-nrv.o, tree-ssa-copy.o, tree-ssa-propagate.o)
3768 (tree-ssa-dom.o, tree-ssa-uncprop.o, tree-ssa-threadedge.o)
3769 (tree-ssa-threadupdate.o, tree-ssanames.o, tree-phinodes.o)
3770 (domwalk.o, tree-ssa-live.o, tree-ssa-copyrename.o)
3771 (tree-ssa-pre.o, tree-ssa-sccvn.o)
3772 (gimple-ssa-strength-reduction.o, tree-vrp.o, tree-cfg.o)
3773 (tree-cfgcleanup.o, tree-tailcall.o, tree-ssa-sink.o)
3774 (tree-nested.o, tree-if-conv.o, tree-iterator.o, tree-dfa.o)
3775 (tree-ssa-operands.o, tree-eh.o, tree-ssa-loop.o)
3776 (tree-ssa-loop-unswitch.o, tree-ssa-address.o)
3777 (tree-ssa-loop-niter.o, tree-ssa-loop-ivcanon.o)
3778 (tree-ssa-loop-ch.o, tree-ssa-loop-prefetch.o, tree-predcom.o)
3779 (tree-ssa-loop-ivopts.o, tree-affine.o, tree-ssa-loop-manip.o)
3780 (tree-ssa-loop-im.o, tree-ssa-math-opts.o, tree-ssa-alias.o)
3781 (tree-ssa-reassoc.o, tree-optimize.o, gimplify.o)
3782 (gimple-iterator.o, gimple-fold.o, gimple-low.o, omp-low.o)
3783 (tree-browser.o, omega.o, tree-chrec.o, tree-scalar-evolution.o)
3784 (tree-data-ref.o, sese.o, graphite.o, graphite-blocking.o)
3785 (graphite-clast-to-gimple.o, graphite-dependences.o)
3786 (graphite-interchange.o, graphite-poly.o)
3787 (graphite-scop-detection.o, graphite-sese-to-poly.o)
3788 (graphite-optimize-isl.o, tree-vect-loop.o)
3789 (tree-vect-loop-manip.o, tree-vect-patterns.o, tree-vect-slp.o)
3790 (tree-vect-stmts.o, tree-vect-data-refs.o, tree-vectorizer.o)
3791 (vtable-verify.o, tree-loop-distribution.o, tree-parloops.o)
3792 (tree-stdarg.o, tree-object-size.o, internal-fn.o, gimple.o)
3793 (gimple-pretty-print.o, tree-mudflap.o, tree-nomudflap.o)
3794 (tree-pretty-print.o, tree-diagnostic.o, fold-const.o)
3795 (diagnostic.o, diagnostic-color.o, opts.o, opts-global.o)
3796 (opts-common.o, targhooks.o, common/common-targhooks.o, input.o)
3797 (toplev.o, hwint.o, passes.o, plugin.o, main.o, host-default.o)
3798 (rtl-error.o, rtl.o, print-rtl.o, rtlanal.o, varasm.o, function.o)
3799 (statistics.o, stmt.o, except.o, expr.o, dojump.o, builtins.o)
3800 (calls.o, expmed.o, explow.o, optabs.o, dbxout.o, debug.o)
3801 (sdbout.o, dwarf2out.o, dwarf2cfi.o, dwarf2asm.o, vmsdbgout.o)
3802 (xcoffout.o, godump.o, emit-rtl.o, real.o, realmpfr.o, dfp.o)
3803 (fixed-value.o, jump.o, simplify-rtx.o, symtab.o, cgraph.o)
3804 (cgraphunit.o, cgraphclones.o, cgraphbuild.o, varpool.o, ipa.o)
3805 (ipa-profile.o, ipa-devirt.o, ipa-prop.o, ipa-ref.o, ipa-cp.o)
3806 (ipa-split.o, ipa-inline.o, ipa-inline-analysis.o)
3807 (ipa-inline-transform.o, ipa-utils.o, ipa-reference.o)
3808 (ipa-pure-const.o, coverage.o, cselib.o, cse.o, dce.o, dumpfile.o)
3809 (dse.o, fwprop.o, web.o, ree.o, cprop.o, gcse.o, store-motion.o)
3810 (resource.o, lcm.o, mode-switching.o, tree-ssa-dce.o)
3811 (tree-call-cdce.o, tree-ssa-ccp.o, tree-ssa-strlen.o, tree-sra.o)
3812 (tree-switch-conversion.o, tree-complex.o, tree-emutls.o)
3813 (tree-vect-generic.o, df-core.o, df-problems.o, df-scan.o)
3814 (regstat.o, valtrack.o, var-tracking.o, profile.o, mcf.o)
3815 (tree-profile.o, value-prof.o, loop-doloop.o, alloc-pool.o)
3816 (auto-inc-dec.o, cfg.o, cfghooks.o, cfgexpand.o, cfgrtl.o)
3817 (cfganal.o, cfgbuild.o, cfgcleanup.o, cfgloop.o, cfgloopanal.o)
3818 (graphds.o, loop-iv.o, loop-invariant.o, cfgloopmanip.o)
3819 (loop-init.o, loop-unswitch.o, loop-unroll.o, dominance.o)
3820 (et-forest.o, combine.o, reginfo.o, bitmap.o, vec.o, hash-table.o)
3821 (reload.o, reload1.o, rtlhooks.o, postreload.o, postreload-gcse.o)
3822 (caller-save.o, bt-load.o, reorg.o, alias.o, stack-ptr-mod.o)
3823 (init-regs.o, ira-build.o, ira-costs.o, ira-conflicts.o)
3824 (ira-color.o, ira-emit.o, ira-lives.o, ira.o, lra.o)
3825 (lra-assigns.o, lra-coalesce.o, lra-constraints.o)
3826 (lra-eliminations.o, lra-lives.o, lra-spills.o, regmove.o)
3827 (combine-stack-adj.o, compare-elim.o, ddg.o, modulo-sched.o)
3828 (haifa-sched.o, sched-deps.o, sched-rgn.o, sched-ebb.o)
3829 (sched-vis.o, sel-sched.o, sel-sched-dump.o, sel-sched-ir.o)
3830 (final.o, recog.o, reg-stack.o, sreal.o, predict.o, lists.o)
3831 (bb-reorder.o, tracer.o, timevar.o, regcprop.o, regrename.o)
3832 (ifcvt.o, params.o, pointer-set.o, hooks.o, pretty-print.o)
3833 (errors.o, dbgcnt.o, lower-subreg.o, target-globals.o)
3834 (hw-doloop.o, file-find.o, context.o, $(common_out_object_file))
3835 (insn-attrtab.o, insn-automata.o, insn-dfatab.o, insn-emit.o)
3836 (insn-enums.o, insn-extract.o, insn-latencytab.o, insn-modes.o)
3837 (insn-opinit.o, insn-output.o, insn-peep.o, insn-preds.o)
3838 (insn-recog.o, intl.o, cppbuiltin.o, cppdefault.o, gcov.o)
3839 (gcov-dump.o): Remove.
3840 (default-c.o): Use COMPILE and POSTCOMPILE.
3841 (CFLAGS-gcc.o): New variable.
3842 ($(common_out_object_file)): Use COMPILE and POSTCOMPILE.
3843
8c796d19
TT		 38442013-09-25 Tom Tromey <tromey@redhat.com>
3845
3846 * Makefile.in (c-family/cppspec.o, c-family/c-common.o)
3847 (c-family/c-cppbuiltin.o, c-family/c-dump.o, c-family/c-format.o)
3848 (c-family/c-gimplify.o, c-family/c-lex.o, c-family/c-omp.o)
3849 (c-family/c-opts.o, c-family/c-pch.o, c-family/c-ppoutput.o)
3850 (c-family/c-pragma.o, c-family/c-pretty-print.o)
3851 (c-family/c-semantics.o, c-family/c-ada-spec.o)
3852 (c-family/array-notation-common.o, c-family/stub-objc.o)
3853 (c-family/c-ubsan.o): Remove.
3854
0a6c2227
TT		 38552013-09-25 Tom Tromey <tromey@redhat.com>
3856
3857 * Makefile.in (C_TREE_H): Reference c/c-tree.h.
3858
f3bc55f0
TT		 38592013-09-25 Tom Tromey <tromey@redhat.com>
3860
3861 * Makefile.in (DRIVER_DEFINES): Use $(and), not shell code,
3862 to add -DENABLE_SHARED_LIBGCC.
3863 (gcc.o): Don't use subshell.
3864
c11c795e
TT		 38652013-09-25 Tom Tromey <tromey@redhat.com>
3866
3867 * Makefile.in (OUTPUT_OPTION): Define as "-o $@".
3868 * configure.ac: Don't invoke AM_PROG_CC_C_O.
3869 (NO_MINUS_C_MINUS_O, OUTPUT_OPTION): Don't subst.
3870 * configure, config.in: Rebuild.
3871
043378c3
TT		 38722013-09-25 Tom Tromey <tromey@redhat.com>
3873
3874 * Makefile.in (CCDEPMODE, DEPDIR, depcomp, COMPILE.base)
3875 (COMPILE, POSTCOMPILE): New variables.
3876 (.cc.o .c.o): Use COMPILE, POSTCOMPILE.
3877 (DEPFILES): New variable.
3878 Include ".Po" files.
3879 * configure.ac: Add checks for dependency checking.
3880 * configure, aclocal.m4: Regenerate.
3881
eab34643
TT		 38822013-09-25 Tom Tromey <tromey@redhat.com>
3883
3884 * Makefile.in (ALL_HOST_BACKEND_OBJS): Add lto-wrapper.o.
7a5911d3 3885 ($(ALL_HOST_OBJS)): Move order-only dependency to end of file.
eab34643 3886
d2db5e91
TT		 38872013-09-25 Tom Tromey <tromey@redhat.com>
3888
3889 * Makefile.in (generated_files): Add options.h,
3890 target-hooks-def.h, insn-opinit.h,
3891 common/common-target-hooks-def.h, pass-instances.def,
3892 c-family/c-target-hooks-def.h.
3893
af33629e
JL		 38942013-09-25 Jeff Law <law@redhat.com>
3895
3896 * tree-ssa-threadedge.c (thread_across_edge): Use foo.last () rather
3897 than foo[foo.length () - 1] to access last member in a vec.
3898 * tree-ssa-threadupdate.c (register_jump_thread): Similarly.
3899
ed0d3051
RB		 39002013-09-25 Richard Biener <rguenther@suse.de>
3901
3902 PR middle-end/58521
3903 * tree.c (iterative_hash_expr): Remove MEM_REF special handling.
3904
123485ca
JH		 39052013-09-25 Jan Hubicka <jh@suse.cz>
3906
3907 * cgraph.c (cgraph_resolve_speculation): Use semantical equivalency
3908 test.
3909
d70b0f1f
MP		 39102013-09-25 Marek Polacek <polacek@redhat.com>
3911
3912 PR sanitizer/58420
3913 * ubsan.c (ubsan_type_descriptor): Handle IDENTIFIER_NODEs
3914 when determining the type name.
3915
e3f02798
OE		 39162013-09-24 Oleg Endo <olegendo@gcc.gnu.org>
3917
3918 * config/sh/sh.md: Fix formatting.
3919
7a5911d3 39202013-09-24 Xinliang David Li <davidxl@google.com>
4f17aa0b 3921
7a5911d3
UB		 3922 * tree-vect-data-refs.c (vect_enhance_data_refs_alignment): Check
3923 max peel iterations parameter.
4f17aa0b
XDL		 3924 * param.def: New parameter.
3925 * doc/invoke.texi: Document New parameter.
3926
0498471b
CL		 39272013-09-24 Christophe Lyon <christophe.lyon@linaro.org>
3928
3929 * gimple-pretty-print.c: Various whitespace tweaks.
3930 * tree-core.h: Likewise.
3931 * tree-pretty-print.c: Likewise.
3932 * tree-ssa-alias.c: Likewise.
3933 * tree-ssa-copy.c: Likewise.
3934 * tree-ssanames.c: Likewise.
3935 * tree-ssanames.h: Likewise.
3936 * tree-vrp.c: Likewise.
3937
5a8c2b57
AM		 39382013-09-24 Alan Modra <amodra@gmail.com>
3939
3940 PR middle-end/57134
3941 PR middle-end/57586
7a5911d3
UB		 3942 * stmt.c (expand_asm_operands): Call expand_expr with EXPAND_MEMORY
3943 for output operands that disallow regs. Don't use EXPAND_WRITE on
3944 inout operands.
5a8c2b57 3945
f40333af
RB		 39462013-09-24 Richard Biener <rguenther@suse.de>
3947
3948 PR middle-end/58513
3949 * tree.c (reference_alias_ptr_type): Move ...
3950 * alias.c (reference_alias_ptr_type): ... here and implement
3951 in terms of the new reference_alias_ptr_type_1.
3952 (ref_all_alias_ptr_type_p): New helper.
3953 (get_deref_alias_set_1): Drop flag_strict_aliasing here,
3954 use ref_all_alias_ptr_type_p.
3955 (get_deref_alias_set): Add flag_strict_aliasing check here.
3956 (reference_alias_ptr_type_1): New function, split out from ...
3957 (get_alias_set): ... here.
3958 (alias_ptr_types_compatible_p): New function.
3959 * alias.h (reference_alias_ptr_type): Declare.
3960 (alias_ptr_types_compatible_p): Likewise.
3961 * tree.h (reference_alias_ptr_type): Remove.
3962 * fold-const.c (operand_equal_p): Use alias_ptr_types_compatible_p
3963 to compare MEM_REF alias types.
3964
583e8bf5
RB		 39652013-09-24 Richard Biener <rguenther@suse.de>
3966
3967 * tree-vrp.c (vrp_finalize): Check for SSA name presence.
3968
5e8586d7
MM		 39692013-09-23 Michael Meissner <meissner@linux.vnet.ibm.com>
3970
19c3e797
MM		 3971 * config/rs6000/rs6000.c (rs6000_vector_reload): Delete, combine
3972 reload helper function arrays into a single array reg_addr.
3973 (reload_fpr_gpr): Likewise.
3974 (reload_gpr_vsx): Likewise.
3975 (reload_vsx_gpr): Likewise.
3976 (struct rs6000_reg_addr): Likewise.
3977 (reg_addr): Likewise.
3978 (rs6000_debug_reg_global): Change rs6000_vector_reload,
3979 reload_fpr_gpr, reload_gpr_vsx, reload_vsx_gpr uses to reg_addr.
3980 (rs6000_init_hard_regno_mode_ok): Likewise.
3981 (rs6000_secondary_reload_direct_move): Likewise.
3982 (rs6000_secondary_reload): Likewise.
3983
5e8586d7
MM		 3984 * config/rs6000/rs6000.h (enum r6000_reg_class_enum): Add new
3985 constraints: wu, ww, and wy. Repurpose wv constraint added during
3986 power8 changes. Put wg constraint in alphabetical order.
3987
3988 * config/rs6000/rs6000.opt (-mvsx-scalar-float): New debug switch
3989 for future work to add ISA 2.07 VSX single precision support.
3990 (-mvsx-scalar-double): Change default from -1 to 1, update
3991 documentation comment.
3992 (-mvsx-scalar-memory): Rename debug switch to -mupper-regs-df.
3993 (-mupper-regs-df): New debug switch to control whether DF values
3994 can go in the traditional Altivec registers.
3995 (-mupper-regs-sf): New debug switch to control whether SF values
3996 can go in the traditional Altivec registers.
3997
3998 * config/rs6000/rs6000.c (rs6000_debug_reg_global): Print wu, ww,
3999 and wy constraints.
4000 (rs6000_init_hard_regno_mode_ok): Use ssize_t instead of int for
4001 loop variables. Rename -mvsx-scalar-memory to -mupper-regs-df.
4002 Add new constraints, wu/ww/wy. Repurpose wv constraint.
4003 (rs6000_debug_legitimate_address_p): Print if we are running
4004 before, during, or after reload.
4005 (rs6000_secondary_reload): Add a comment.
4006 (rs6000_opt_masks): Add -mupper-regs-df, -mupper-regs-sf.
4007
4008 * config/rs6000/constraints.md (wa constraint): Sort w<x>
4009 constraints. Update documentation string.
4010 (wd constraint): Likewise.
4011 (wf constraint): Likewise.
4012 (wg constraint): Likewise.
4013 (wn constraint): Likewise.
4014 (ws constraint): Likewise.
4015 (wt constraint): Likewise.
4016 (wx constraint): Likewise.
4017 (wz constraint): Likewise.
4018 (wu constraint): New constraint for ISA 2.07 SFmode scalar
4019 instructions.
4020 (ww constraint): Likewise.
4021 (wy constraint): Likewise.
4022 (wv constraint): Repurpose ISA 2.07 constraint that we not used in
4023 the previous submissions.
4024 * doc/md.texi (PowerPC and IBM RS6000): Likewise.
4025
e5af9ddd
RS		 40262013-09-23 Richard Sandiford <rdsandiford@googlemail.com>
4027
4028 * doc/rtl.texi (REG_NOTES): Say that int_list can also be used.
4029 (REG_BR_PROB): Say that the probability is stored in an int_list.
4030 * reg-notes.def: Update commentary to mention INT_LIST.
4031 * rtl.def (EXPR_LIST, INSN_LIST): Capitalize comments.
4032 (INT_LIST): New rtx.
4033 * rtl.h (add_int_reg_note, add_shallow_copy_of_reg_note): Declare.
4034 * rtlanal.c (int_reg_note_p): New function.
4035 (alloc_reg_note): Assert that the note does not have an int argument.
4036 (add_int_reg_note, add_shallow_copy_of_reg_note): New functions.
4037 * combine.c (distribute_notes): Use add_shallow_copy_of_rtx.
4038 * cse.c (cse_process_notes_1): Expect REG_EQUAL to be an EXPR_LIST
4039 rather than an INSN_LIST. Handle INT_LIST.
4040 * ifcvt.c (cond_exec_process_insns): Take the probability as an int
4041 rather than an rtx. Use gen_rtx_INT_LIST to create a REG_BR_PROB note.
4042 (cond_exec_process_if_block): Use XINT to extract REG_BR_PROB values.
4043 Manipulate them as ints rather than rtxes.
4044 * reg-stack.c (subst_asm_stack_regs): Only handle EXPR_LIST notes.
4045 * regmove.c (copy_src_to_dest): Likewise.
4046 * sched-vis.c (print_insn_with_notes): Handle INT_LIST.
4047
4048 * config/i386/winnt.c (i386_pe_seh_unwind_emit): Sink pat assignment
4049 into the cases that need it.
4050 * config/arm/arm.c (arm_unwind_emit): Likewise.
4051
4052 * asan.c (asan_clear_shadow): Use add_int_reg_note for REG_BR_PROB.
4053 * emit-rtl.c (try_split, emit_copy_of_insn_after): Likewise.
4054 * loop-doloop.c (add_test, doloop_modify): Likewise.
4055 * loop-unswitch.c (compare_and_jump_seq): Likewise.
4056 * optabs.c (emit_cmp_and_jump_insn_1): Likewise.
4057 * predict.c (combine_predictions_for_insn): Likewise.
4058 * print-rtl.c (print_rtx): Handle INT_LIST.
4059 * config/aarch64/aarch64.c (aarch64_emit_unlikely_jump): Likewise.
4060 * config/alpha/alpha.c (emit_unlikely_jump): Likewise.
4061 * config/arm/arm.c (emit_unlikely_jump): Likewise.
4062 * config/i386/i386.c (ix86_expand_split_stack_prologue): Likewise.
4063 (ix86_split_fp_branch, predict_jump): Likewise.
4064 * config/rs6000/rs6000.c (emit_unlikely_jump): Likewise.
4065 * config/sh/sh.c (expand_cbranchsi4): Likewise.
4066 * config/spu/spu.c (ea_load_store_inline): Likewise.
4067
4068 * cfgbuild.c (compute_outgoing_frequencies): Use XINT to access the
4069 value of a REG_BR_PROB note.
4070 * cfgrtl.c (force_nonfallthru_and_redirect): Likewise.
4071 (update_br_prob_note, rtl_verify_edges, purge_dead_edges): Likewise.
4072 * emit-rtl.c (try_split): Likewise.
4073 * predict.c (br_prob_note_reliable_p): Likewise.
4074 (invert_br_probabilities, combine_predictions_for_insn): Likewise.
4075 * reorg.c (mostly_true_jump): Likewise.
4076 * config/bfin/bfin.c (cbranch_predicted_taken_p): Likewise.
4077 * config/frv/frv.c (frv_print_operand_jump_hint): Likewise.
4078 * config/i386/i386.c (ix86_print_operand): Likewise.
4079 * config/ia64/ia64.c (ia64_print_operand): Likewise.
4080 * config/mmix/mmix.c (mmix_print_operand): Likewise.
4081 * config/rs6000/rs6000.c (output_cbranch): Likewise.
4082 * config/s390/s390.c (s390_expand_tbegin): Likewise.
4083 * config/sh/sh.c (sh_print_operand, sh_adjust_cost): Likewise.
4084 * config/sparc/sparc.c (output_cbranch): Likewise.
4085 * config/spu/spu.c (get_branch_target): Likewise.
4086 * config/tilegx/tilegx.c (cbranch_predicted_p): Likewise.
4087 * config/tilepro/tilepro.c (cbranch_predicted_p): Likewise.
4088
450ad0cd
JH		 40892013-09-23 Jan Hubicka <jh@suse.cz>
4090
4091 * ipa-cp.c (ipa_get_indirect_edge_target_1): Add sanity check
4092 for ipa-devirt.
4093 * ipa-utils.h (possible_polymorphic_call_target_p): New function.
4094 * ipa-devirt.c (possible_polymorphic_call_target_p): Be tolerant
4095 of external calls
4096 * gimple-fold.c: Include ipa-utils.h and gimple-pretty-print.h
4097 (gimple_fold_call): Dump inconsistent devirtualizations; add
4098 sanity check for type based devirtualizations.
4099 * ipa-prop.c: Include ipa-utils.h
4100 (ipa_intraprocedural_devirtualization): Add sanity check.
4101 (try_make_edge_direct_virtual_call): Likewise.
4102
cefce769
EB		 41032013-09-23 Eric Botcazou <ebotcazou@adacore.com>
4104
4105 * tree-ssa-ccp.c (insert_clobber_before_stack_restore): Recurse on copy
4106 assignment statements.
4107
a895a2b8
KV		 41082013-09-23 Kugan Vivekanandarajah <kuganv@linaro.org>
4109
4110 * gimple-pretty-print.c (dump_ssaname_info): New function.
4111 (dump_gimple_phi): Call it.
4112 (pp_gimple_stmt_1): Likewise.
4113 * tree-core.h (tree_ssa_name): New union ssa_name_info_type field.
4114 (range_info_def): Declare.
4115 * tree-pretty-print.c (pp_double_int): New function.
4116 (dump_generic_node): Call it.
4117 * tree-pretty-print.h (pp_double_int): Declare.
4118 * tree-ssa-alias.c (dump_alias_info): Check pointer type.
4119 * tree-ssanames.h (range_info_def): New structure.
4120 (value_range_type): Move definition here.
4121 (set_range_info, value_range_type, duplicate_ssa_name_range_info):
4122 Declare.
4123 * tree-ssanames.c (make_ssa_name_fn): Check pointer type at
4124 initialization.
4125 (set_range_info): New function.
4126 (get_range_info): Likewise.
4127 (duplicate_ssa_name_range_info): Likewise.
4128 (duplicate_ssa_name_fn): Check pointer type and call
4129 duplicate_ssa_name_range_info.
4130 * tree-ssa-copy.c (fini_copy_prop): Likewise.
4131 * tree-vrp.c (value_range_type): Remove definition, now in
4132 tree-ssanames.h.
7a5911d3 4133 (vrp_finalize): Call set_range_info to update value range of SSA_NAMEs.
a895a2b8
KV		 4134 * tree.h (SSA_NAME_PTR_INFO): Macro changed to access via union.
4135 (SSA_NAME_RANGE_INFO): New macro.
4136
984af6ac
RB		 41372013-09-23 Richard Biener <rguenther@suse.de>
4138
4139 PR tree-optimization/58464
4140 * tree-ssa-pre.c (phi_trans_lookup): Remove.
7a5911d3 4141 (phi_trans_add): Change to add conditionally on being not present.
984af6ac
RB		 4142 (phi_translate_1): Remove recursion detection here.
4143 (phi_translate): Pre-seed the cache with NULL to catch
4144 recursion here in a more generic way.
4145 (bitmap_find_leader): Adjust comment.
4146 (get_representative_for): Dump value-numbers.
4147 (create_expression_by_pieces): Likewise.
4148 (insert_into_preds_of_block): Likewise.
4149
0e26cf79
CB		 41502013-09-23 Christian Bruel <christian.bruel@st.com>
4151
4152 PR target/58475
4153 * config/sh/sh.md (movsf_ie): Allow fpul_operand.
4154 * config/sh/predicate.md (arith_reg_operand): Disallow FPUL_REG.
4155
4d0e5b54
JG		 41562013-09-23 James Greenhalgh <james.greenhalgh@arm.com>
4157
4158 Revert r202780:
4159 2013-09-20 Renlin Li <renlin.li@arm.com>
4160
7a5911d3
UB		 4161 * config/aarch64/aarch64.c (aarch64_expand_prologue): Use
4162 plus_constant.
4d0e5b54
JG		 4163 (aarch64_expand_epilogue): Likewise.
4164 (aarch64_legitimize_reload_address): Likewise.
4165
ca081cc8
EB		 41662013-09-22 Eric Botcazou <ebotcazou@adacore.com>
4167
4168 * gimplify.c (gimplify_asm_expr): Reset the TREE_CHAIN of clobbers to
4169 NULL_TREE before pushing them onto the vector. Likewise for labels.
4170
8fc53a5f
EB		 41712013-09-21 Eric Botcazou <ebotcazou@adacore.com>
4172
4173 * config/ia64/predicates.md (ia64_cbranch_operator): Accept unordered
4174 comparison operators when -fno-trapping-math is in effect.
4175 * config/ia64/ia64.c (ia64_expand_compare): Add support for unordered
4176 comparison operators in TFmode and assert that unsupported operators
4177 cannot reach here.
4178 (ia64_print_operand): Likewise.
4179
41802013-09-21 Jan Hubicka <jh@suse.cz>
0ca6c49f
JH		 4181
4182 * x86-tune.def (partial_reg_stall): Disable for CoreI7 and newer.
4183 (sse_typeless_stores): Enable for core
4184 (sse_load0_by_pxor): Likewise.
4185 (four_jump_limit): Disable for core.
4186 (pad_returns): Likewise.
4187 (avoid_vector_decode): Likewise.
4188 (fuse_cmp_and_branch): Enable for cores.
4189 * i386.c (x86_accumulate_outgoing_args): Disable for cores.
4190
578e51e6
JDA		 41912013-09-20 John David Anglin <danglin@gcc.gnu.org>
4192
96ce28a3
JDA		 4193 PR middle-end/56791
4194 * config/pa/pa.c (pa_option_override): Disable auto increment and
4195 decrement instructions until reload is completed.
4196
578e51e6
JDA		 4197 * config/pa/pa-linux.h (TARGET_OS_CPP_BUILTINS): Define
4198 __GCC_HAVE_SYNC_COMPARE_AND_SWAP_1, __GCC_HAVE_SYNC_COMPARE_AND_SWAP_2,
4199 and __GCC_HAVE_SYNC_COMPARE_AND_SWAP_4.
4200
6fcd3a13 42012013-09-20 DJ Delorie <dj@redhat.com>
7a5911d3 4202 Nick Clifton <nickc@redhat.com>
6fcd3a13
DD		 4203
4204 * config/rl78/rl78.c: Various whitespace and comment tweaks.
4205 (need_to_save): Save bank 0 on interrupts.
4206 (characterize_address): Strip far address wrappers.
4207 (rl78_as_legitimate_address): Likewise.
4208 (transcode_memory_rtx): Likewise.
4209 (rl78_peep_movhi_p): Disable this peephole after devirt.
4210 (rl78_propogate_register_origins): Forget all origins when a
4211 CLOBBER is seen.
4212 * config/rl78/rl78-virt.md: Various whitespace tweaks.
4213 * config/rl78/rl78-real.md: Various whitespace tweaks. Additional
4214 peephole2's.
4215 * config/rl78/rl78.md (sel_rb): Disable for G10 just in case.
4216 * config/rl78/rl78-expand.md (movqi): Check for subregs of consts.
4217 * config/rl78/rl78.h (LINK_SPEC): Pass -gc-sections unless
4218 relocating.
4219 * config/rl78/constraints.md: Various whitespace and paren tweaks.
4220
4a08db33
JDA		 42212013-09-20 John David Anglin <danglin@gcc.gnu.org>
4222
4223 * config/pa/pa.md: In "scc" insn patterns, change output template to
4224 handle const0_rtx in reg_or_0_operand operands.
4225
7691132c
JBG		 42262013-09-20 Martin Husemann <martin@NetBSD.org>
4227
4228 PR target/56875
4229 * config/vax/vax.c (vax_output_int_move): Use D format specifier.
4230 * config/vax/vax.md (ashldi3, <unnamed>): Ditto.
4231
a3cc13cc
RB		 42322013-09-20 Richard Biener <rguenther@suse.de>
4233
4234 PR middle-end/58484
4235 * tree-scalar-evolution.c (struct scev_info_str): Shrink by
4236 remembering SSA name version and block index.
4237 (new_scev_info_str): Adjust.
4238 (hash_scev_info): Likewise. Also hash the block index.
4239 (eq_scev_info): Adjust.
4240 (find_var_scev_info): Likewise.
4241 (struct instantiate_cache_entry): Remove.
4242 (struct instantiate_cache_type): Use a htab to map name, block
4243 to chrec.
4244 (instantiate_cache_type::~instantiate_cache_type): Adjust.
4245 (get_instantiated_value_entry): Likewise.
4246 (hash_idx_scev_info, eq_idx_scev_info): New functions.
4247 (instantiate_scev_name): Adjust.
4248
925f3871
JL		 42492013-09-20 Jeff Law <law@redhat.com>
4250
4251 * tree-ssa-dom.c (record_temporary_equivalences): Add comment.
4252
2888c331
YZ		 42532013-09-20 Yufeng Zhang <yufeng.zhang@arm.com>
4254
4255 * config/aarch64/aarch64-builtins.c (aarch64_simd_expand_args):
4256 Call aarch64_simd_expand_args to update op[argc].
4257
0a811e96
BS		 42582013-09-20 Basile Starynkevitch <basile@starynkevitch.net>
4259
7a5911d3
UB		 4260 * plugin.c (parse_plugin_arg_opt): Accept equal sign inside
4261 plugin argument.
0a811e96 4262
0078f462
BS		 42632013-09-20 Basile Starynkevitch <basile@starynkevitch.net>
4264
7a5911d3
UB		 4265 * gengtype.c (file_rules): Added rule for *.cc files.
4266 (get_output_file_with_visibility): Give fatal message when no
4267 rules found.
0078f462 4268
d20e48b0
RL		 42692013-09-20 Renlin Li <renlin.li@arm.com>
4270
4271 * config/aarch64/aarch64.c (aarch64_expand_prologue): Use plus_constant.
4272 (aarch64_expand_epilogue): Likewise.
4273 (aarch64_legitimize_reload_address): Likewise.
4274
59bb1544
BE		 42752013-09-20 Bernd Edlinger <bernd.edlinger@hotmail.de>
4276
4277 PR middle-end/57748
4278 * expr.c (expand_assignment): Remove misalignp code path.
4279
a5430019
MP		 42802013-09-20 Marek Polacek <polacek@redhat.com>
4281
4282 PR sanitizer/58413
4283 * ubsan.c (get_ubsan_type_info_for_type): Use TYPE_SIZE instead of
4284 TYPE_PRECISION. Add asserts.
4285
9fed7f3a
RB		 42862013-09-20 Richard Biener <rguenther@suse.de>
4287
4288 PR tree-optimization/58453
4289 * tree-loop-distribution.c (distribute_loop): Apply the cost
4290 model for -ftree-loop-distribute-patterns, too.
4291
a820c834
RB		 42922013-09-20 Richard Biener <rguenther@suse.de>
4293
4294 PR middle-end/58473
4295 * tree-chrec.h (build_polynomial_chrec): Use gcc_checking_assert,
4296 make type comparison less strict.
4297
3cbe17f7
AM		 42982013-09-20 Alan Modra <amodra@gmail.com>
4299
4300 * configure: Regenerate.
4301 * aclocal.m4: Regenerate.
4302
87fccdbb
MP		 43032013-09-20 Marek Polacek <polacek@redhat.com>
4304
4305 PR other/58467
4306 * doc/extend.texi: Document that attribute used is meant to be used
4307 on variables with static storage duration.
4308
8928eff3
JJ		 43092013-09-19 Jakub Jelinek <jakub@redhat.com>
4310
4311 PR tree-optimization/58472
4312 * tree-vect-stmts.c (vectorizable_store, vectorizable_load): For
4313 simd_lane_access set inv_p = false.
4314 * omp-low.c (lower_rec_input_clauses): Set TREE_NO_WARNING on
4315 the simduid magic VAR_DECL.
4316
b7b28c53
JH		 43172013-09-19 Jan Hubicka <jh@suse.cz>
4318
4319 * i386.c (generic_memcpy, generic_memset): Fix 32bit template.
4320
83ae86f5
JL		 43212013-09-17 Jeff Law <law@redhat.com>
4322
4323 * tree-ssa-dom.c (record_temporary_equivalences): New function
4324 split out of dom_opt_dom_walker::after_dom_children.
4325 (dom_opt_dom_walker::thread_across_edge): Move common code
4326 in here from dom_opt_dom_walker::after_dom_children.
4327 (dom_opt_dom_walker::after_dom_children): Corresponding simplifictions.
4328
9d532162
JH		 43292013-09-19 Jan Hubicka <jh@suse.cz>
4330
4331 * i386.h (TARGET_GENERIC32, TARGET_GENERIC64): Remove.
4332 (TARGET_GENERIC): Use PROCESOR_GENERIC
4333 (enum processor_type): Unify generic32 and 64.
4334 * i386.md (cpu): Likewise.
4335 * x86-tune.def (use_leave): Enable for generic32.
4336 (avoid_vector_decode, slow_imul_imm32_mem, slow_imul_imm8): Likewise.
4337 * athlon.md: Change generic64 to generic in all occurences.
4338 * i386-c.c (ix86_target_macros_internal): Unify generic64 and 32.
4339 (ix86_target_macros_internal): Likewise.
4340 * driver-i386.c (host_detect_local_cpu): Likewise.
7a5911d3
UB		 4341 * i386.c (generic64_memcpy, generic64_memset, generic64_cost): Rename
4342 to ..
9d532162
JH		 4343 (generic_memcpy, generic_memset, generic_cost): This one.
4344 (generic32_memcpy, generic32_memset, generic32_cost): Remove.
4345 (m_GENERIC32, m_GENERIC64): Remove.
4346 (m_GENERIC): Turn into one flag.
4347 (processor_target): Unify generic tunnings.
4348 (ix86_option_override_internal): Replace generic32/64 by generic.
4349 (ix86_issue_rate): Likewise.
4350 (ix86_adjust_cost): Likewise.
4351
188c7d00
JH		 43522013-09-19 Jan Hubicka <jh@suse.cz>
4353
4354 * cgraph.c (cgraph_create_edge_1): Avoid uninitialized read
4355 of speculative flag.
4356
4befd127
JJ		 43572013-09-19 Jakub Jelinek <jakub@redhat.com>
4358
4359 * omp-low.c (expand_omp_sections): Always pass len - 1 to
4360 GOMP_sections_start, even if !exit_reachable.
4361
8a26ad39
VM		 43622013-09-18 Vladimir Makarov <vmakarov@redhat.com>
4363
4364 * lra-constraints.c (need_for_all_save_p): Use macro
4365 HARD_REGNO_CALL_PART_CLOBBERED.
4366 * lra-lives.c (check_pseudos_live_through_calls): Use the macro to
4367 set up pseudo conflict hard regs.
4368
09487185
MM		 43692013-09-18 Michael Meissner <meissner@linux.vnet.ibm.com>
4370
4371 PR target/58452
4372 * config/rs6000/paired.md (movmisalignv2sf): Fix to allow memory
4373 operands.
4374
080cbf9e
VM		 43752013-09-18 Vladimir Makarov <vmakarov@redhat.com>
4376
4377 PR rtl-optimization/58438
4378 * lra.c (lra): Clear lra_optional_reload_pseudos in upper loop.
4379 * lra-constraints.c (undo_optional_reloads): Keep optional reloads
4380 from previous subpasses.
4381
a0f70fcb
RE		 43822013-09-18 Richard Earnshaw <rearnsha@arm.com>
4383
4384 * arm.c (arm_get_frame_offsets): Validate architecture supports
4385 LDRD/STRD before accepting the tuning preference.
4386 (arm_expand_prologue): Likewise.
4387 (arm_expand_epilogue): Likewise.
4388
0547c9b6
RB		 43892013-09-18 Richard Biener <rguenther@suse.de>
4390
4391 PR tree-optimization/58417
4392 * tree-chrec.c (chrec_fold_plus_1): Assert that we do not
4393 have chrecs with symbols defined in the loop as operands.
4394 (chrec_fold_multiply): Likewise.
4395 * tree-scalar-evolution.c (interpret_rhs_expr): Instantiate
4396 parameters before folding binary operations.
4397 (struct instantiate_cache_entry_hasher): Remove.
4398 (struct instantiate_cache_type): Use a pointer-map.
4399 (instantiate_cache_type::instantiate_cache_type): New function.
4400 (instantiate_cache_type::get): Likewise.
4401 (instantiate_cache_type::set): Likewise.
4402 (instantiate_cache_type::~instantiate_cache_type): Adjust.
4403 (get_instantiated_value_entry): Likewise.
4404 (global_cache): New global.
4405 (instantiate_scev_r, instantiate_scev_poly, instantiate_scev_binary,
4406 instantiate_array_ref, instantiate_scev_convert, instantiate_scev_3,
4407 instantiate_scev_2, instantiate_scev_1): Do not pass along cache.
4408 (instantiate_scev_name): Adjust.
4409 (instantiate_scev): Construct global instead of local cache.
4410 (resolve_mixers): Likewise.
4411
309dc1aa
DM		 44122013-09-18 Daniel Morris <danielm@ecoscentric.com>
4413 Paolo Carlini <paolo.carlini@oracle.com>
4414
4415 PR c++/58458
4416 * doc/implement-cxx.texi: Fix references to the C++ standards.
4417
995a1b4a
JJ		 44182013-09-18 Jakub Jelinek <jakub@redhat.com>
4419
4420 * omp-low.c (copy_var_decl): Copy DECL_ATTRIBUTES.
4421 * tree-vect-data-refs.c (vect_analyze_data_refs): For
4422 simd_lane_access drs, update also DR_ALIGNED_TO.
4423
ce6923c5
MP		 44242013-09-18 Marek Polacek <polacek@redhat.com>
4425
4426 PR sanitizer/58411
4427 * doc/extend.texi: Document no_sanitize_undefined attribute.
4428 * builtins.c (fold_builtin_0): Don't sanitize function if it has the
4429 no_sanitize_undefined attribute.
4430
d30d00a2
NC		 44312013-09-18 Nick Clifton <nickc@redhat.com>
4432
4433 * config/msp430/msp430.h (ASM_SPEC): Pass -md on to the assembler.
4434 (ASM_DECLARE_FUNCTION_NAME): Define.
4435
4d9192b5
TS		 44362013-09-17 Trevor Saunders <tsaunders@mozilla.com>
4437
4438 * compare-elim.c (find_comparison_dom_walker): New class
4439 (find_comparisons_in_bb): Rename to
4440 find_comparison_dom_walker::before_dom_children
4441 (find_comparisons): Adjust
4442 * domwalk.c (walk_dominator_tree): Rename to dom_walker::walk, and
4443 adjust.
4444 (init_walk_dominator_tree, fini_walk_dominator_tree): Remove
4445 * domwalk.h (dom_walk_data): Convert it To a class dom_walker.
4446 (init_walk_dominator_tree): Remove declaration.
4447 (fini_walk_dominator_tree): Remove declaration.
4448 * fwprop.c (single_def_use_dom_walker): New class
4449 (single_def_use_enter_block): Convert to
4450 single_def_use_dom_walker::before_dom_children.
4451 (single_def_use_leave_block): Convert to
4452 single_def_use_dom_walker::after_dom_children.
4453 (build_single_def_use_links): Adjust.
4454 * gimple-ssa-strength-reduction.c (find_candidates_dom_walker): New
4455 class.
4456 (find_candidates_in_block): Convert to
4457 find_candidates_dom_walker::before_dom_children.
4458 (execute_strength_reduction): Adjust.
4459 * graphite-sese-to-poly.c (struct bsc, build_sese_conditions): Remove.
4460 (sese_dom_walker): New class.
4461 (sese_dom_walker::sese_dom_walker): New constructor.
4462 (sese_dom_walker::~sese_dom_walker): New destructor.
4463 (build_sese_conditions_before): Convert to
4464 sese_dom_walker::before_dom_children.
4465 (build_sese_conditions_after): Convert to
4466 sese_dom_walker::after_dom_children.
4467 (build_poly_scop): Adjust
4468 * tree-into-ssa.c (rewrite_dom_walker): New class
4469 (rewrite_enter_block): Convert to
4470 rewrite_dom_walker::before_dom_children.
4471 (rewrite_leave_block): Convert to
4472 rewrite_dom_walker::after_dom_children.
4473 (rewrite_update_dom_walker): New class.
4474 (rewrite_update_enter_block): Convert to
4475 rewrite_update_dom_walker::before_dom_children.
4476 (rewrite_update_leave_block): Convert to
4477 rewrite_update_dom_walker::after_dom_children.
4478 (rewrite_blocks, rewrite_into_ssa): Adjust.
4479 (mark_def_dom_walker): New class.
4480 (mark_def_dom_walker::mark_def_dom_walker): New constructor.
4481 (mark_def_dom_walker::~mark_def_dom_walker): New destructor.
4482 (mark_def_sites_blocks): Convert to
4483 mark_def_dom_walker::before_dom_children.
4484 (mark_def_site_blocks): Remove.
4485 * tree-ssa-dom.c (dom_opt_dom_walker): New class.
4486 (tree_ssa_dominator_optimize): Adjust.
4487 (dom_thread_across_edge): Convert to method
4488 dom_opt_dom_walker::thread_across_edge.
4489 (dom_opt_enter_block): Convert to member function
4490 dom_opt_dom_walker::before_dom_children.
4491 (dom_opt_leave_block): Convert to member function
4492 dom_opt_dom_walker::after_dom_children.
4493 * tree-ssa-dse.c (dse_dom_walker): New class.
4494 (dse_enter_block): Convert to member function
4495 dse_dom_walker::before_dom_children.
4496 (tree_ssa_dse): Adjust.
4497 * tree-ssa-loop-im.c (invariantness_dom_walker): New class.
4498 (determine_invariantness_stmt): Convert to method
4499 invariantness_dom_walker::before_dom_children.
4500 (determine_invariantness): Remove
4501 (move_computations_dom_walker): New class.
4502 (move_computations_stmt): Convert to method
4503 move_computations_dom_walker::before_dom_children.
4504 (move_computations, tree_ssa_lim): Adjust.
4505 * tree-ssa-phiopt.c (nontrapping_dom_walker): new class
4506 (nt_init_block): Make method
4507 notrappping_dom_walker::before_dom_children.
4508 (nt_fini_block): Make
4509 method nontrapping_dom_walker::after_dom_children.
4510 (get_non_trapping): Adjust.
4511 * tree-ssa-pre.c (eliminate_dom_walker): New class.
4512 (eliminate_bb): Make method eliminate_dom_walker::before_dom_children.
4513 (eliminate_leave_block): Make method.
4514 eliminate_dom_walker::after_dom_children.
4515 (eliminate): Adjust
4516 * tree-ssa-strlen.c (strlen_dom_walker): New class.
4517 (strlen_enter_block): Make method
4518 strlen_dom_walker::before_dom_children.
4519 (strlen_leave_block): Make
4520 method strlen_dom_walker::after_dom_children.
4521 (tree_ssa_strlen): Adjust.
4522 * tree-ssa-uncprop.c (uncprop_dom_walker): New class.
4523 (tree_ssa_uncprop): Adjust.
4524 (uncprop_leave_block): Make method
4525 uncprop_dom_walker::after_dom_children.
4526 (uncprop_leave_block): Make method
4527 uncprop_dom_walker::before_dom_children.
4528
3b3cc26b
BC		 45292013-09-18 Bin Cheng <bin.cheng@arm.com>
4530
7a5911d3
UB		 4531 * config/arm/arm.c (thumb1_reorg): Search for flag setting insn before
4532 branch in same basic block. Check both src and dest of the move insn.
3b3cc26b 4533
b5ba3a9b
NC		 45342013-09-17 Nick Clifton <nickc@redhat.com>
4535
4536 * config/rl78/rl78-real.md (bf): New pattern.
4537 (bt): New pattern.
4538 * config/rl78/rl78.c (rl78_print_operand_1): Handle %B.
4539 (rl78_print_operand): Do not put a # before a %B.
4540 * config/rl78/rl78.opt: Tweak doc strings.
4541
09d5094b
DD		 45422013-09-17 DJ Delorie <dj@redhat.com>
4543
4544 * config/rl78/constraints.md (Wcv): Allow up to $r31.
4545 * config/rl78/rl78.c (rl78_asm_file_start: Likewise.
4546 (rl78_option_override): Likewise, if -mallregs.
4547 (is_virtual_register): Likewise.
4548 * config/rl78/rl78.h (reg_class): Extend VREGS to $r31.
4549 (REGNO_OK_FOR_BASE_P): Likewise.
4550 * config/rl78/rl78.opt (-mallregs): New.
4551
2ba1dca1
NC		 45522013-09-17 Nick Clifton <nickc@redhat.com>
4553
4554 * config/rl78/rl78.c (need_to_save): Change return type to bool.
4555 For interrupt functions: save all call clobbered registers if the
4556 interrupt handler is not a leaf function.
4557 (rl78_expand_prologue): Always recompute the frame information.
4558 For interrupt functions: only select bank 0 if one of the bank 0
4559 registers is going to be psuhed.
4560
2e7c3f21
DD		 45612013-09-17 DJ Delorie <dj@redhat.com>
4562
4563 * config/rl78/constraints.md: For each W* constraint, rename to C*
4564 and create a W* constraint that checks for an optional ES: prefix
4565 pattern also.
4566 * config/rl78/rl78.md (UNS_ES_ADDR): New.
4567 (es_addr): New. Used to wrap far addresses.
4568 * config/rl78/rl78-protos.h (rl78_es_addr): New.
4569 (rl78_es_base): New.
4570 * config/rl78/rl78.c (rl78_as_legitimate_address): Accept "unspec"
4571 wrapped far addresses.
4572 (rl78_print_operand_1): Unwrap far addresses before processing.
4573 (rl78_lo16): Wrap far addresses in unspecs.
4574 (rl78_es_addr): New.
4575 (rl78_es_base): New.
4576 (insn_ok_now): Check for not-yet-wrapped far addresses.
4577 (transcode_memory_rtx): Properly re-wrap far addresses.
4578
5c26a69a
SH		 45792013-09-17 Sebastian Huber <sebastian.huber@embedded-brains.de>
4580
4581 * config/sparc/t-rtems: Add leon3 multilibs.
4582
181f5f3e
CH		 45832013-09-17 Cong Hou <congh@google.com>
4584
4585 * tree-vect-patterns.c (vect_recog_dot_prod_pattern): Fix a bug
4586 when checking the dot production pattern. The type of rhs operand
4587 of multiply is now checked correctly.
4588
8d34e421
JL		 45892013-09-17 Jeff Law <law@redhat.com>
4590
4591 * tree-ssa-dom.c (cprop_into_successor_phis): Also propagate
4592 edge implied equivalences into successor phis.
4593 * tree-ssa-threadupdate.c (phi_args_equal_on_edges): Moved into
4594 here from tree-ssa-threadedge.c.
4595 (mark_threaded_blocks): When threading through a joiner, if both
4596 successors of the joiner's clone reach the same block, verify the
4597 PHI arguments are equal. If not, cancel the jump threading request.
4598 * tree-ssa-threadedge.c (phi_args_equal_on_edges): Moved into
4599 tree-ssa-threadupdate.c
4600 (thread_across_edge): Don't check PHI argument equality when
4601 threading through joiner block here.
7a5911d3 4602
c152901f
AM		 46032013-09-17 Andrew MacLeod <amacleod@redhat.com>
4604
4605 * tree-flow.h (ssa_undefined_value_p): Remove prototype.
4606 * tree-ssa.c (ssa_undefined_value_p): Move pass independent parts here.
7a5911d3
UB		 4607 (warn_uninit, warn_uninitialized_vars,
4608 execute_early_warn_uninitialized, make_pass_early_warn_uninitialized):
4609 Move to tree-ssa-uninit.c.
78cedfb1 4610 * tree-ssa-uninit.c (ssa_undefined_value_p): Move to tree-ssa.c.
7a5911d3 4611 (has_undefined_value_p): New. Pass dependant parts of
c152901f
AM		 4612 ssa_undefined_value_p.
4613 (uninit_undefined_value_p): Use has_undefined_value_p.
7a5911d3
UB		 4614 (warn_uninit, warn_uninitialized_vars,
4615 execute_early_warn_uninitialized, make_pass_early_warn_uninitialized):
4616 Move from tree-ssa.c.
4617 * tree-ssa.h: Adjust prototypes.
c152901f 4618
b631d45a
JH		 46192013-09-17 Jan Hubicka <jh@suse.cz>
4620
4621 PR middle-end/58332
4622 * cif-code.def (FUNCTION_NOT_OPTIMIZED): New CIF code.
4623 * ipa-inline.c (can_inline_edge_p): Do not downgrade
4624 FUNCTION_NOT_OPTIMIZED.
4625 * ipa-inline-analysis.c (compute_inline_parameters): Function
4626 not optimized is not inlinable unless it is alwaysinline.
4627 (inline_analyze_function): Force calls in not optimized
4628 function not inlinable.
4629
5b79657a
JH		 46302013-09-17 Jan Hubicka <jh@suse.cz>
4631
4632 PR middle-end/58329
4633 * ipa-devirt.c (ipa_devirt): Be ready for symtab_nonoverwritable_alias
4634 to return NULL.
4635 * ipa.c (function_and_variable_visibility): Likewise.
4636 * ipa-profile.c (ipa_profile): Likewise.
4637
27442c24
BE		 46382013-09-17 Bernd Edlinger <bernd.edlinger@hotmail.de>
4639
4640 PR ipa/58398
4641 * cgraph.c (cgraph_function_body_availability): Check for ifunc
4642 attribute, and don't inline the resolver in this case.
4643
753b9bc3
TJ		 46442013-09-17 Teresa Johnson <tejohnson@google.com>
4645
4646 * coverage.c (get_coverage_counts): Add missing newline.
4647
e3d3cfb4
KT		 46482013-09-17 Kyrylo Tkachov <kyrylo.tkachov@arm.com>
4649
4650 PR tree-optimization/58088
4651 * fold-const.c (mask_with_trailing_zeros): New function.
4652 (fold_binary_loc): Make sure we don't recurse infinitely
4653 when the X in (X & C1) | C2 is a tree of the form (Y * K1) & K2.
4654 Use mask_with_trailing_zeros where appropriate.
4655
9a0ac98f
YR		 46562013-09-17 Yuri Rumyantsev <ysrumyan@gmail.com>
4657
4658 * config/i386/i386.c (distance_agu_use_in_bb) : Proper initialization
4659 of 'prev' var to get better distance estimation.
4660
78bbd765
EB		 46612013-09-17 Eric Botcazou <ebotcazou@adacore.com>
4662
4663 * tree-inline.h (struct copy_body_data): Add transform_parameter.
4664 * tree-inline.c (is_parameter_of): New predicate.
4665 (remap_gimple_op_r): Do not propagate TREE_THIS_NOTRAP on MEM_REF if
4666 a parameter has been remapped.
4667 (copy_tree_body_r): Likewise on INDIRECT_REF and MEM_REF.
4668 (optimize_inline_calls): Initialize transform_parameter.
4669 (copy_gimple_seq_and_replace_locals): Likewise.
4670 (tree_function_versioning): Likewise.
4671 (maybe_inline_call_in_expr): Likewise.
4672
cad055a4
NC		 46732013-09-17 Nick Clifton <nickc@redhat.com>
4674
4675 * config/msp430/msp430-protos.h: Add prototypes for new functions.
4676 * config/msp430/msp430.c (msp430_preserve_reg_p): Add support for
4677 interrupt handlers.
4678 (is_attr_func): New function.
4679 (msp430_is_interrupt_func): New function.
4680 (is_naked_func): New function.
4681 (is_reentrant_func): New function.
4682 (is_critical_func): New function.
4683 (msp430_start_function): Add annotations for function attributes.
4684 (msp430_attr): New function.
4685 (msp430_attribute_table): New.
4686 (msp430_function_section): New function.
4687 (TARGET_ASM_FUNCTION_SECTION): Define.
4688 (msp430_builtin): New enum.
4689 (msp430_init_builtins): New function.
4690 (msp430_builtin_devl): New function.
4691 (msp430_expand_builtin): New function.
4692 (TARGET_INIT_BUILTINS): Define.
4693 (TARGET_EXPAND_BUILTINS): Define.
4694 (TARGET_BUILTIN_DECL): Define.
4695 (msp430_expand_prologue): Add support for naked, interrupt,
4696 critical and reentrant functions.
4697 (msp430_expand_epilogue): Likewise.
4698 (msp430_print_operand): Handle 'O' character.
4699 * config/msp430/msp430.h (TARGET_CPU_CPP_BUILTINS): Define
4700 NO_TRAMPOLINES.
4701 * config/msp430/msp430.md (unspec): Add UNS_DINT, UNS_EINT,
4702 UNS_PUSH_INTR, UNS_POP_INTR, UNS_BIC_SR, UNS_BIS_SR.
4703 (pushm): Use a 'n' rather than an 'i' constraint.
4704 (msp_return): Add generation of the interrupt return instruction.
4705 (disable_interrupts): New pattern.
4706 (enable_interrupts): New pattern.
4707 (push_intr_state): New pattern.
4708 (pop_intr_state): New pattern.
4709 (bic_SR): New pattern.
4710 (bis_SR): New pattern.
4711 * doc/extend.texi: Document MSP430 function attributes and builtin
4712 functions.
4713
deb6c11a
RB		 47142013-09-17 Richard Biener <rguenther@suse.de>
4715
4716 PR tree-optimization/58432
4717 * tree-loop-distribution.c (tree_loop_distribution): Also
4718 scan PHIs for outside loop uses and seed a partition from them.
4719
c068654b
BC		 47202013-09-17 Bin Cheng <bin.cheng@arm.com>
4721
4722 * gimple-ssa-strength-reduction.c (backtrace_base_for_ref): New.
4723 (restructure_reference): Call backtrace_base_for_ref.
4724
567969e4
AM		 47252013-09-17 Alan Modra <amodra@gmail.com>
4726
4727 PR target/57589
7a5911d3
UB		 4728 * config/rs6000/driver-rs6000.c (elf_platform): Revert 2013-06-11
4729 patch.
567969e4 4730
5c0029de
DD		 47312013-09-16 DJ Delorie <dj@redhat.com>
4732
4733 * config/rl78/rl78.c (rl78_asm_file_start): Specify alternate
4734 vregs location for RL78/G10.
4735 (rl78_expand_prologue): Avoid SEL on G10.
4736 (rl78_expand_epilogue): Likewise.
4737 (rl78_peep_movhi_p): Can't move a constant to memory in HImode.
4738 * config/rl78/rl78.h (TARGET_CPU_CPP_BUILTINS): Define
4739 __RL78_G10__ when appropriate.
4740 (ASM_SPEC): Pass -mg10 along to the assembler.
4741 * config/rl78/rl78.md (sel_rb): Disable for G10.
4742 * config/rl78/rl78.opt: Add -mg10 option.
4743 * config/rl78/t-rl78: Add -mg10 multilib.
4744
ea0f3e87
XDL		 47452013-09-16 Xinliang David Li <davidxl@google.com>
4746
4747 * tree-if-conv.c (main_tree_if_conversion): Check new flag.
4748 * omp-low.c (omp_max_vf): Ditto.
4749 (expand_omp_simd): Ditto.
4750 * tree-vectorizer.c (vectorize_loops): Ditto.
4751 (gate_vect_slp): Ditto.
4752 (gate_increase_alignment): Ditto.
4753 * tree-ssa-pre.c (inhibit_phi_insertion): Ditto.
4754 * tree-ssa-loop.c (gate_tree_vectorize): Ditto.
4755 (gate_tree_vectorize): Name change.
4756 (tree_vectorize): Ditto.
4757 (pass_vectorize::gate): Call new function.
4758 (pass_vectorize::execute): Ditto.
7a5911d3 4759 * opts.c: O3 default setting change.
ea0f3e87
XDL		 4760 (finish_options): Check new flag.
4761 * doc/invoke.texi: Document new flags.
4762 * common.opt: New flags.
4763
4f3da2e5
AS		 47642013-09-16 Andreas Schwab <schwab@linux-m68k.org>
4765
4766 * doc/tm.texi.in (Cond Exec Macros): Remove node.
4767 (Condition Code): Don't reference it.
4768 * doc/tm.texi: Regenerate.
4769
aa940f7c
VM		 47702013-09-16 Vladimir Makarov <vmakarov@redhat.com>
4771
4772 PR middle-end/58418
4773 * lra-constraints.c (undo_optional_reloads): Consider all optional
4774 reload even if it did not get a hard reg.
4775
e645e942
TJ		 47762013-09-16 Teresa Johnson <tejohnson@google.com>
4777
4778 * dumpfile.c (dump_loc): Remove newline emission.
4779 * tree-vect-data-refs.c (vect_lanes_optab_supported_p): Add newline
4780 emission to dump_printf_loc calls where missing.
4781 (vect_mark_for_runtime_alias_test): Ditto.
4782 (vect_analyze_data_ref_dependence): Ditto.
4783 (vect_analyze_data_ref_dependences): Ditto.
4784 (vect_slp_analyze_data_ref_dependence): Ditto.
4785 (vect_slp_analyze_data_ref_dependences): Ditto.
4786 (vect_compute_data_ref_alignment): Ditto.
4787 (vect_update_misalignment_for_peel): Ditto.
4788 (vect_verify_datarefs_alignment): Ditto.
4789 (vector_alignment_reachable_p): Ditto.
4790 (vect_get_data_access_cost): Ditto.
4791 (vect_enhance_data_refs_alignment): Ditto.
4792 (vect_find_same_alignment_drs): Ditto.
4793 (vect_analyze_data_refs_alignment): Ditto.
4794 (vect_analyze_group_access): Ditto.
4795 (vect_analyze_data_ref_access): Ditto.
4796 (vect_analyze_data_ref_accesses): Ditto.
4797 (vect_prune_runtime_alias_test_list): Ditto.
4798 (vect_analyze_data_refs): Ditto.
4799 (vect_create_addr_base_for_vector_ref): Ditto.
4800 (vect_create_data_ref_ptr): Ditto.
4801 (vect_grouped_store_supported): Ditto.
4802 (vect_grouped_load_supported): Ditto.
4803 * value-prof.c (check_counter): Ditto.
4804 (check_ic_target): Ditto.
4805 * tree-vect-patterns.c (vect_recog_dot_prod_pattern): Ditto.
4806 (vect_recog_widen_mult_pattern): Ditto.
4807 (vect_recog_widen_sum_pattern): Ditto.
4808 (vect_recog_over_widening_pattern): Ditto.
4809 (vect_recog_widen_shift_pattern): Ditto.
4810 (vect_recog_rotate_pattern): Ditto.
4811 (vect_recog_vector_vector_shift_pattern): Ditto.
4812 (vect_recog_divmod_pattern): Ditto.
4813 (vect_recog_mixed_size_cond_pattern): Ditto.
4814 (vect_recog_bool_pattern): Ditto.
4815 (vect_pattern_recog_1): Ditto.
4816 (vect_pattern_recog): Ditto.
4817 * tree-vect-loop.c (vect_determine_vectorization_factor): Ditto.
4818 (vect_is_simple_iv_evolution): Ditto.
4819 (vect_analyze_scalar_cycles_1): Ditto.
4820 (vect_get_loop_niters): Ditto.
4821 (vect_analyze_loop_1): Ditto.
4822 (vect_analyze_loop_form): Ditto.
4823 (vect_analyze_loop_operations): Ditto.
4824 (vect_analyze_loop_2): Ditto.
4825 (vect_analyze_loop): Ditto.
4826 (report_vect_op): Ditto.
4827 (vect_is_slp_reduction): Ditto.
4828 (vect_is_simple_reduction_1): Ditto.
4829 (vect_get_known_peeling_cost): Ditto.
4830 (vect_estimate_min_profitable_iters): Ditto.
4831 (vect_model_reduction_cost): Ditto.
4832 (vect_model_induction_cost): Ditto.
4833 (get_initial_def_for_induction): Ditto.
4834 (vect_create_epilog_for_reduction): Ditto.
4835 (vectorizable_reduction): Ditto.
4836 (vectorizable_induction): Ditto.
4837 (vectorizable_live_operation): Ditto.
4838 (vect_loop_kill_debug_uses): Ditto.
4839 (vect_transform_loop): Ditto.
4840 * tree-vect-stmts.c (vect_mark_relevant): Ditto.
4841 (vect_stmt_relevant_p): Ditto.
4842 (process_use): Ditto.
4843 (vect_mark_stmts_to_be_vectorized): Ditto.
4844 (vect_model_simple_cost): Ditto.
4845 (vect_model_promotion_demotion_cost): Ditto.
4846 (vect_model_store_cost): Ditto.
4847 (vect_get_store_cost): Ditto.
4848 (vect_model_load_cost): Ditto.
4849 (vect_get_load_cost): Ditto.
4850 (vect_init_vector_1): Ditto.
4851 (vect_get_vec_def_for_operand): Ditto.
4852 (vect_finish_stmt_generation): Ditto.
4853 (vectorizable_call): Ditto.
4854 (vectorizable_conversion): Ditto.
4855 (vectorizable_assignment): Ditto.
4856 (vectorizable_shift): Ditto.
4857 (vectorizable_operation): Ditto.
4858 (vectorizable_store): Ditto.
4859 (vectorizable_load): Ditto.
4860 (vectorizable_condition): Ditto.
4861 (vect_analyze_stmt): Ditto.
4862 (vect_transform_stmt): Ditto.
4863 (vect_is_simple_use): Ditto.
4864 * tree-vect-loop-manip.c (slpeel_make_loop_iterate_ntimes): Ditto.
4865 (vect_can_advance_ivs_p): Ditto.
4866 (vect_update_ivs_after_vectorizer): Ditto.
4867 (vect_do_peeling_for_loop_bound): Ditto.
4868 (vect_gen_niters_for_prolog_loop): Ditto.
4869 (vect_update_inits_of_drs): Ditto.
4870 (vect_create_cond_for_alias_checks): Ditto.
4871 * tree-vect-slp.c (vect_get_and_check_slp_defs): Ditto.
4872 (vect_build_slp_tree_1): Ditto.
4873 (vect_supported_load_permutation_p): Ditto.
4874 (vect_analyze_slp_instance): Ditto.
4875 (vect_analyze_slp): Ditto.
4876 (vect_make_slp_decision): Ditto.
4877 (vect_detect_hybrid_slp): Ditto.
4878 (vect_bb_vectorization_profitable_p): Ditto.
4879 (vect_slp_analyze_bb_1): Ditto.
4880 (vect_update_slp_costs_according_to_vf): Ditto.
4881 (vect_get_mask_element): Ditto.
4882 (vect_transform_slp_perm_load): Ditto.
4883 (vect_schedule_slp_instance): Ditto.
4884 (vect_schedule_slp): Ditto.
4885 (vect_slp_transform_bb): Ditto.
4886 * profile.c (read_profile_edge_counts): Ditto.
4887 (compute_branch_probabilities): Ditto.
4888 * coverage.c (get_coverage_counts): Ditto.
4889
46f851f3
DN		 48902013-09-16 Diego Novillo <dnovillo@google.com>
4891
7a5911d3 4892 * tree-core.h: Add missing comment lines from refactoring of tree.h.
46f851f3 4893
1632a686
JH		 48942013-09-16 Jan Hubicka <jh@suse.cz>
4895
4896 * gimple-fold.c (can_refer_decl_in_current_unit_p): Do not accept
7a5911d3 4897 abstract functions; for static functions check the presence of body.
1632a686 4898
828e70c1
JG		 48992013-09-16 James Greenhalgh <james.greenhalgh@arm.com>
4900
4901 * config/aarch64/aarch64-simd-builtins.def (fma): New.
4902 * config/aarch64/aarch64-simd.md
4903 (aarch64_mla_elt<mode>): New.
4904 (aarch64_mla_elt_<vswap_width_name><mode>): Likewise.
4905 (aarch64_mls_elt<mode>): Likewise.
4906 (aarch64_mls_elt_<vswap_width_name><mode>): Likewise.
4907 (aarch64_fma4_elt<mode>): Likewise.
4908 (aarch64_fma4_elt_<vswap_width_name><mode>): Likewise.
4909 (aarch64_fma4_elt_to_128v2df): Likewise.
4910 (aarch64_fma4_elt_to_64df): Likewise.
4911 (fnma<mode>4): Likewise.
4912 (aarch64_fnma4_elt<mode>): Likewise.
4913 (aarch64_fnma4_elt_<vswap_width_name><mode>): Likewise.
4914 (aarch64_fnma4_elt_to_128v2df): Likewise.
4915 (aarch64_fnma4_elt_to_64df): Likewise.
4916 * config/aarch64/iterators.md (VDQSF): New.
4917 * config/aarch64/arm_neon.h
4918 (vfm<as><sdq>_lane<q>_f<32, 64>): Convert to C implementation.
4919 (vml<sa><q>_lane<q>_<fsu><16, 32, 64>): Likewise.
4920
779aea46
JG		 49212013-09-16 James Greenhalgh <james.greenhalgh@arm.com>
4922
4923 * config/aarch64/aarch64-simd.md (aarch64_mul3_elt<mode>): New.
4924 (aarch64_mul3_elt_<vswap_width_name><mode>): Likewise.
4925 (aarch64_mul3_elt_to_128df): Likewise.
4926 (aarch64_mul3_elt_to_64v2df): Likewise.
4927 * config/aarch64/iterators.md (VEL): Also handle DFmode.
4928 (VMUL): New.
4929 (VMUL_CHANGE_NLANES) Likewise.
4930 (h_con): Likewise.
4931 (f): Likewise.
4932 * config/aarch64/arm_neon.h
4933 (vmul<q>_lane<q>_<suf><16,32,64>): Convert to C implementation.
4934
a407a750
JG		 49352013-09-16 James Greenhalgh <james.greenhalgh@arm.com>
4936
4937 * config/aarch64/arm_neon.h
4938 (vcvtx_high_f32_f64): Fix parameters.
4939
f77e2d2d 49402013-09-16 Jan-Benedict Glaw <jbglaw@lug-owl.de>
a5bd71fa 4941 Uros Bizjak <ubizjak@gmail.com>
f77e2d2d 4942
a5bd71fa 4943 * config/alpha.c: Include tree-ssanames.h.
f77e2d2d 4944
36875e8f
RB		 49452013-09-16 Richard Biener <rguenther@suse.de>
4946
4947 * tree-loop-distribution.c (enum rdg_dep_type): Add control_dd.
4948 (dot_rdg_1): Handle control_dd.
4949 (create_edge_for_control_dependence): New function.
4950 (create_rdg_edges): Add control dependences if asked for.
4951 (build_rdg): Likewise.
4952 (generate_loops_for_partition): If there are not necessary
4953 control stmts remove all their dependencies.
4954 (collect_condition_stmts, rdg_flag_loop_exits): Remove.
4955 (distribute_loop): Pass on control dependences.
4956 (tree_loop_distribution): Compute control dependences and remove
4957 restriction on number of loop nodes.
4958
96e24d49
JJ		 49592013-09-16 Jakub Jelinek <jakub@redhat.com>
4960
4961 * ipa-prop.c (ipa_compute_jump_functions_for_edge): Return early
4962 for internal calls.
4963
40dbb05c
RS		 49642013-09-16 Richard Sandiford <rdsandiford@googlemail.com>
4965
4966 * cse.c (try_const_anchors): Punt on CC modes.
4967
97191ad0
JBG		 49682013-09-15 Jan-Benedict Glaw <jbglaw@lug-owl.de>
4969
4970 * config/vax/constraints.md (T): Add missing CONSTANT_P check.
4971
403a3fb7
JDA		 49722013-09-14 John David Anglin <danglin@gcc.gnu.org>
4973
4974 PR target/58382
4975 * config/pa/pa.c (pa_expand_prologue): Change mode in gen_rtx_POST_INC
4976 calls to word_mode.
4977
fb5610fb
IS		 49782013-09-14 Iain Sandoe <iain@codesourcery.com>
4979
4980 PR target/48094
4981 * config/darwin.c (darwin_objc2_section): Note if ObjC Metadata is
f77e2d2d
JBG		 4982 seen.
4983 (darwin_objc1_section): Likewise.
fb5610fb
IS		 4984 (darwin_file_end): Emit Image Info section when required.
4985
afdec9bd
JH		 49862013-09-14 Jan Hubicka <jh@suse.cz>
4987
4988 * tree-into-ssa.c (gate_into_ssa): New.
4989 (pass_data_build_ssa): Use it.
4990 * cgraph.h (expand_thunk): Update prototype.
4991 * cgraphunit.c (analyze_function): Expand thunks early.
4992 (expand_thunk): Fix DECL_CONTEXT of reust_decl;
4993 build proper cgraph; set in_ssa_p; clear bogus TREE_ASM_WRITTEN;
4994 set lowered flag; do not add new function.
4995 (assemble_thunks_and_aliases): Update.
4996 * tree-ssa.c (gate_init_datastructures): New gate.
4997 (pass_data_init_datastructures): Use it.
4998
492d1e0a
IS		 49992013-09-14 Iain Sandoe <iain@codesourcery.com>
5000
5001 PR target/58269
f77e2d2d 5002 * config/i386/i386.c (ix86_function_arg_regno_p): Make Darwin use the
492d1e0a
IS		 5003 xmm register set described in the psABI.
5004
d13c6b70
EG		 50052013-09-13 Evgeny Gavrin <e.gavrin@samsung.com>
5006
f77e2d2d
JBG		 5007 * dwarf2out.c (should_emit_struct_debug): Add check
5008 for type_decl variable is not NULL.
d13c6b70 5009
88b0e79e
JC		 50102013-09-13 Jacek Caban <jacek@codeweavers.com>
5011
5012 * config.gcc: Use new winnt-c.c target hooks
5013 * config/t-winnt: New file
5014 * config/winnt-c.c: New file
5015 * doc/tm.texi.in: Document new hook
5016 * doc/tm.texi: Regenerated
5017
19ba6aab
JH		 50182013-09-13 Jan Hubicka <jh@suse.cz>
5019
5020 PR middle-end/58094
5021 * ipa-inline.c (check_callers): New function.
5022 (check_caller_edge): Remove.
5023 (want_inline_function_to_all_callers_p): Also permit alises that are
5024 called dirrectly.
5025 (inline_to_all_callers): Terminate the walk when devirtualization
5026 introduce new calls.
5027
a93c18c8
JH		 50282013-09-13 Jan Hubicka <jh@suse.cz>
5029
5030 * ipa-inline-analysis.c (struct growth_data): Add node.
5031 (do_estimate_growth_1): Fix detection of recursion.
5032
f3b331d1
JJ		 50332013-09-13 Jakub Jelinek <jakub@redhat.com>
5034
5035 PR tree-optimization/58392
5036 * tree-cfg.c (move_sese_region_to_fn): Rename loop variable
5037 to avoid shadowing of outer loop variable. If
5038 saved_cfun->has_simduid_loops or saved_cfun->has_force_vect_loops,
5039 replace_by_duplicate_decl simduid of loops that have it set and
5040 set dest_cfun->has_simduid_loops and/or
5041 dest_cfun->has_force_vect_loops.
5042 * omp-low.c (build_outer_var_ref): Call maybe_lookup_decl_in_outer_ctx
5043 instead of maybe_lookup_decl.
5044 * tree-inline.c (copy_loops): Change blocks_to_copy argument to id.
5045 Use id->blocks_to_copy instead of blocks_to_copy. Adjust recursive
5046 call. Copy over force_vect and copy and remap simduid. Set
5047 cfun->has_simduid_loops and/or cfun->has_force_vect_loops.
5048 (copy_cfg_body): Remove blocks_to_copy argument. Use
5049 id->blocks_to_copy instead of blocks_to_copy. Adjust copy_loops
5050 caller. Don't set cfun->has_simduid_loops and/or
5051 cfun->has_force_vect_loops here.
5052 (copy_body): Remove blocks_to_copy argument. Adjust copy_cfg_body
5053 caller.
5054 (expand_call_inline, tree_function_versioning): Adjust copy_body
5055 callers.
5056
ae6d0907
MJ		 50572013-09-13 Martin Jambor <mjambor@suse.cz>
5058
5059 PR bootstrap/58388
5060 * ipa-prop.c (try_make_edge_direct_simple_call): Be less strict in
5061 the assert if the edge was a speculative one.
5062
2fd5894f
RB		 50632013-09-13 Richard Biener <rguenther@suse.de>
5064
5065 * tree-data-ref.h (known_dependences_p): Move here ...
5066 * tree-loop-distribution.c (known_dependences_p): ... from here.
5067 (dump_rdg_component, debug_rdg_component): Remove.
5068 (dump_rdg): Adjust.
5069 (generate_loops_for_partition): Use gimple_uid instead of
5070 relying on matching stmt visit order.
5071 (rdg_build_partitions): Take starting stmt vector.
5072 (ldist_gen): Merge into ...
5073 (distribute_loop): ... this function. Do not compute starting
5074 vertices vector.
5075 * tree-cfg.c (gimple_duplicate_bb): Copy UID for PHIs.
5076
86eb4bd7
KT		 50772013-09-13 Kyrylo Tkachov <kyrylo.tkachov@arm.com>
5078
5079 * config/arm/arm.md (arm_cmpsi_insn): Split rI alternative.
5080 Set type attribute correctly. Set predicable_short_it attribute.
5081 (cmpsi_shiftsi): Remove %? from output template.
5082
83a95546
RB		 50832013-09-13 Richard Biener <rguenther@suse.de>
5084
5085 * tree-loop-distribution.c (struct rdg_component,
5086 rdg_defs_used_in_other_loops_p, free_rdg_components,
5087 rdg_build_components): Remove.
5088 (stmts_from_loop): Do not record virtual PHIs.
5089 (generate_loops_for_partition): Skip virtual PHIs.
5090 (build_rdg_partition_for_component): Rename to ...
5091 (build_rdg_partition_for_vertex): ... this and adjust.
5092 (rdg_build_partitions): Take a vector of starting vertices
5093 instead of components. Remove unnecessary leftover handling.
5094 (ldist_gen): Do not build components or record other stores.
5095 (distribute_loop): Do not distribute loops containing stmts
5096 with side-effects.
5097
c0ecf7f1
CB		 50982013-09-13 Christian Bruel <christian.bruel@st.com>
5099
5100 PR target/58314
5101 * config/sh/sh.md (mov<mode>_reg_reg): Allow memory reloads.
5102
34f56efc
KT		 51032013-09-13 Kai Tietz <ktietz@redhat.com>
5104
740f2bf3
KT		 5105 * config.gcc: Separate cases for mingw and cygwin targets,
5106 and add 64-bit cygwin target case.
5107
34f56efc
KT		 5108 * config/i386/winnt-cxx.c (i386_pe_type_dllexport_p): Don't
5109 dll-export inline-functions.
5110 * config/i386/winnt.c (i386_pe_determine_dllexport_p): Likewise.
5111
5112
51132013-09-13 Jeff Law <law@redhat.com>
823b5d6b
JL		 5114
5115 PR middle-end/58387
5116 Revert:
5117 2013-09-06 Jeff Law <law@redhat.com>
5118
5119 * tree-ssa-dom.c (cprop_into_successor_phis): Also propagate
5120 edge implied equivalences into successor phis.
5121
aadb52ad
DD		 51222013-09-12 DJ Delorie <dj@redhat.com>
5123
5124 * config/rl78/rl78-virt.md: Change from | to \; for asm line
5125 separators.
5126
90ae7019
BM		 51272013-09-12 Brooks Moses <bmoses@google.com>
5128
5129 PR driver/42955
5130 * Makefile.in: Do not install driver binaries in $(target)/bin.
5131
17b2557c
DD		 51322013-09-12 DJ Delorie <dj@redhat.com>
5133
c5215a9b
DD		 5134 * config/rl78/rl78.opt (mrelax): New.
5135 * config/rl78/rl78.h (ASM_SPEC): New, pass on -mrelax to gas.
5136 * config/rl78/rl78.h (LINK_SPEC): New, pass on -mrelax to ld.
5137
17b2557c
DD		 5138 * config/rl78/rl78.c (rl78_expand_prologue): Use AX to copy
5139 between SP and FP.
5140 (rl78_expand_epilogue): Likewise.
5141
ecafcf05
VM		 51422013-09-12 Vladimir Makarov <vmakarov@redhat.com>
5143
5144 PR middle-end/58335
84f6d9fa 5145 * lra-eliminations.c (remove_reg_equal_offset_note): New.
ecafcf05
VM		 5146 (eliminate_regs_in_insn): Rewrite frame pointer to hard frame
5147 pointer elimination with using remove_reg_equal_offset_note.
5148
8369f38a
DD		 51492013-09-12 DJ Delorie <dj@redhat.com>
5150
f6a83b4a
DD		 5151 * config/msp430/: New port.
5152 * config.gcc (msp430): Added.
5153 * doc/invoke.texi: Document MSP430 options.
5154 * doc/install.texi: Document msp430-elf
5155 * doc/md.texi: Document msp430-elf
5156 * doc/contrib.texi: Document msp430-elf
5157
8369f38a
DD		 5158 * cfgexpand.c (expand_debug_expr): Avoid sign-extending SImode to
5159 PSImode.
5160
a854f856
MJ		 51612013-09-12 Martin Jambor <mjambor@suse.cz>
5162
5163 PR ipa/58389
5164 * ipa-prop.c (remove_described_reference): Give up if the edge in the
5165 reference descriptor is NULL.
5166 (ipa_edge_removal_hook): If owning a reference descriptor, set its
5167 edge to NULL.
5168
b12ebd96
AM		 51692013-09-12 Andrew MacLeod <amacleod@redhat.com>
5170
5171 * tree-flow.h (FREE_SSANAMES): Move to tree-ssanames.c
5172 (SSANAMES, MODIFIED_NORETURN_CALLS, DEFAULT_DEFS, ptr_info_def,
5173 num_ssa_names, ssa_name): Move to tree-ssanames.h + prototypes.
5174 * tree-flow-inline.h (make_ssa_name, copy_ssa_name, duplicate_ssa_name,
5175 make_temp_ssa_name): move to tree-ssanames.h
5176 * tree-ssa-alias.h: Move prototype.
5177 * tree-ssa.h: Include tree-ssanames.h.
5178 * tree-ssanames.c (FREE_SSANAMES): Move to here.
5179 * tree-ssanames.h: New. Move items from tree-flow*.h
5180 * Makefile.in (tree-ssanames.h): Add to tree-ssanames.o and GTFILES.
5181
e1fb4ad3
RB		 51822013-09-12 Richard Biener <rguenther@suse.de>
5183
5184 PR tree-optimization/58404
5185 * tree-ssa-forwprop.c (forward_propagate_addr_expr_1): Also
5186 propagate non-invariant addresses into dereferences wrapped
5187 in component references.
5188
154c9188
RB		 51892013-09-12 Richard Biener <rguenther@suse.de>
5190
5191 PR tree-optimization/58402
5192 * passes.def: Move pass_late_warn_uninitialized later.
5193
7a300452
AM		 51942013-09-12 Andrew MacLeod <amacleod@redhat.com>
5195
5196 * tree-ssa.h: New. Move content from tree-flow.h and
5197 tree-flow-inline.h.
5198 * tree-flow.h (_edge_var_map, edge_var_map_vector): Move to tree-ssa.h.
5199 Move prototypes belonging to tree-ssa.c.
f77e2d2d 5200 * tree-flow-inline.h (redirect_edge_var_map_def,
7a300452
AM		 5201 redirect_edge_var_map_result, redirect_edge_var_map_location): Move to
5202 tree-ssa.h.
5203 * gimple.h: Adjust prototypes.
f77e2d2d
JBG		 5204 * tree-ssa.c (useless_type_conversion_p, types_compatible_p): Move
5205 to...
5206 * gimple.c (useless_type_conversion_p, types_compatible_p): Here.
7a300452
AM		 5207 * tree.h: Move prototype to tree-ssa.h.
5208 * gengtype.c (open_base_files): Replace tree-flow.h with tree-ssa.h.
5209 * Makefile.in: (TREE_SSA_H, TREE_FLOW_H): Adjust dependencies.
5210 * alias.c, asan.c, builtins.c, calls.c, cfgexpand.c, cfghooks.c,
5211 cfgloop.c, cfgloopmanip.c, cgraph.c, cgraphbuild.c, cgraphclones.c,
5212 cgraphunit.c, dse.c, except.c, expr.c, final.c, fold-const.c,
5213 ggc-page.c, gimple-fold.c, gimple-iterator.c, gimple-low.c,
5214 gimple-pretty-print.c, gimple-ssa-strength-reduction.c,
5215 gimple-streamer-in.c, gimple-streamer-out.c, gimple.c, gimplify.c,
f77e2d2d
JBG		 5216 graphite-blocking.c, graphite-clast-to-gimple.c,
5217 graphite-dependences.c, graphite-interchange.c,
5218 graphite-optimize-isl.c, graphite-poly.c, graphite-scop-detection.c,
5219 graphite-sese-to-poly.c, graphite.c, ipa-cp.c, ipa-inline-analysis.c,
5220 ipa-inline-transform.c, ipa-inline.c, ipa-prop.c, ipa-pure-const.c,
5221 ipa-reference.c, ipa-split.c, ipa-utils.c,
7a300452
AM		 5222 loop-init.c, lto-cgraph.c, lto-section-in.c, lto-section-out.c,
5223 lto-streamer-in.c, lto-streamer-out.c, lto-streamer.c, omp-low.c,
5224 passes.c, predict.c, print-tree.c, profile.c, sese.c, targhooks.c,
5225 tracer.c, trans-mem.c, tree-call-cdce.c, tree-cfg.c, tree-cfgcleanup.c,
5226 tree-chrec.c, tree-complex.c, tree-data-ref.c, tree-dfa.c, tree-eh.c,
5227 tree-emutls.c, tree-if-conv.c, tree-inline.c, tree-into-ssa.c,
5228 tree-loop-distribution.c, tree-mudflap.c, tree-nested.c, tree-nrv.c,
5229 tree-object-size.c, tree-optimize.c, tree-outof-ssa.c, tree-parloops.c,
5230 tree-phinodes.c, tree-predcom.c, tree-pretty-print.c, tree-profile.c,
5231 tree-scalar-evolution.c, tree-sra.c, tree-ssa*.c, tree-stdarg.c,
5232 tree-streamer-in.c, tree-switch-conversion.c, tree-tailcall.c,
5233 tree-vect-data-refs.c, tree-vect-generic.c, tree-vect-loop-manip.c,
5234 tree-vect-loop.c, tree-vect-patterns.c, tree-vect-slp.c,
5235 tree-vect-stmts.c, tree-vectorizer.c, tree-vrp.c, tsan.c,
5236 value-prof.c, var-tracking.c,
5237 varpool.c, vtable-verify.c: Replace tree-flow.h with tree-ssa.h
5238
97463b2b
RB		 52392013-09-12 Richard Biener <rguenther@suse.de>
5240
5241 PR tree-optimization/58396
5242 * tree-loop-distribution.c (create_rdg_edges): Free unused DDRs.
5243 (build_rdg): Take a loop-nest parameter, fix memleaks.
5244 (distribute_loop): Compute loop-nest here and pass it to build_rdg.
5245
80ecd89d
YR		 52462013-09-12 Yuri Rumyantsev <ysrumyan@gmail.com>
5247
5248 * config/i386/x86-tune.def: Turn on X86_TUNE_AVOID_MEM_OPND_FOR_CMOVE
5249 for SLM.
5250
893e85fa
CM		 52512013-09-12 Cameron McInally <cameron.mcinally@nyu.edu>
5252
5253 * doc/extend.texi: Fix errors in x86 FMA builtin naming.
5254 The FMA instruction names should have a 'v' prefix.
5255
174ec470
RB		 52562013-09-12 Richard Biener <rguenther@suse.de>
5257
5258 * tree-loop-distribution.c (dot_rdg_1): Make graph prettier.
5259 (dot_rdg): Use popen instead of system in optional code.
5260 (remaining_stmts, upstream_mem_writes): Remove global bitmaps.
5261 (already_processed_vertex_p): Adjust.
5262 (has_anti_or_output_dependence, predecessor_has_mem_write,
5263 mark_nodes_having_upstream_mem_writes, has_upstream_mem_writes,
5264 rdg_flag_uses): Remove.
5265 (rdg_flag_vertex): Simplify.
5266 (rdg_flag_vertex_and_dependent): Rely on a correct RDG and
5267 remove recursion.
5268 (build_rdg_partition_for_component): Process the first vertex
5269 of a component only.
5270 (ldist_gen): Do not compute remaining_stmts or upstream_mem_writes.
5271
38c89763
AM		 52722013-09-12 Alan Modra <amodra@gmail.com>
5273
5274 * config/rs6000/rs6000.c (toc_relative_expr_p): Use add_cint_operand.
5275
3e3d9d17
DD		 52762013-09-11 DJ Delorie <dj@redhat.com>
5277 Nick Clifton <nickc@redhat.com>
5278
5279 * config/rl78/predicates.md (rl78_cmp_operator_signed): New.
5280 (rl78_stack_based_mem): New.
5281 * config/rl78/constraints.md (Iv08): New.
5282 (Iv16): New.
5283 (Iv24): New.
5284 (Is09): New.
5285 (Is17): New.
5286 (Is25): New.
5287 (ISsi): New.
5288 (IShi): New.
5289 (ISqi): New.
5290 * config/rl78/rl78-expand.md (movqi): Reject more SUBREG operands.
5291 (movhi): Likewise.
5292 (movsi): Change from expand to insn-and-split.
5293 (ashrsi3): Clobber AX.
5294 (lshrsi3): New.
5295 (ashlsi3): New.
5296 (cbranchsi4): New.
5297 * config/rl78/rl78.md (CC_REG): Fix.
5298 (addsi3): Allow memory and immediate operands.
5299 (addsi3_internal): Split into...
5300 (addsi3_internal_virt): ...new, and ...
5301 (addsi3_internal_real): ...new.
5302 (subsi): New.
5303 (subsi3_internal_virt): New.
5304 (subsi3_internal_real): New.
5305 (mulsi3): Add memory operand.
5306 (mulsi3_rl78): Likewise.
5307 (mulsi3_g13): Likewise.
5308 * config/rl78/rl78-real.md (cbranchqi4_real_signed): New.
5309 (cbranchqi4_real): Add more constraint options.
5310 (cbranchhi4_real): Expand pattern.
5311 (cbranchhi4_real_signed): New.
5312 (cbranchhi4_real_inverted): New.
5313 (cbranchsi4_real_lt): New.
5314 (cbranchsi4_real_ge): New.
5315 (cbranchsi4_real_signed): New.
5316 (cbranchsi4_real): New.
5317 (peephole2): New.
893e85fa
CM		 5318 * config/rl78/rl78-virt.md (ashrsi3_virt): Add custom cases for
5319 constant shifts.
3e3d9d17
DD		 5320 (lshrsi3_virt): Likewise.
5321 (ashlsi3_virt): Likewise.
5322 (cbranchqi4_virt_signed): New.
5323 (cbranchhi4_virt_signed): New.
5324 (cbranchsi4_virt): New.
5325 * config/rl78/rl78.c: Whitespace fixes throughout.
5326 (move_elim_pass): New.
5327 (pass_data_rl78_move_elim): New.
5328 (pass_rl78_move_elim): New.
5329 (make_pass_rl78_move_elim): New.
5330 (rl78_devirt_info): Run devirt earlier.
5331 (rl78_move_elim_info): New.
5332 (rl78_asm_file_start): Register it.
5333 (rl78_split_movsi): New.
5334 (rl78_as_legitimate_address): Allow virtual base registers when
5335 appropriate.
5336 (rl78_addr_space_convert): Remove spurious debug stuff.
5337 (rl78_print_operand_1): Add z,s,S,r,E modifiers.
5338 (rl78_print_operand): More cases for not printing '#'.
5339 (rl78_expand_compare): Remove most of the logic.
5340 (content_memory): New.
5341 (clear_content_memory): New.
5342 (get_content_index): New.
5343 (get_content_name): New.
5344 (display_content_memory): New.
5345 (update_content): New.
5346 (record_content): New.
5347 (already_contains): New.
5348 (insn_ok_now): Re-recog insns with virtual registers.
5349 (add_postponed_content_update): New.
5350 (process_postponed_content_update): New.
5351 (gen_and_emit_move): New.
893e85fa
CM		 5352 (transcode_memory_rtx): Record new location content.
5353 Use gen_and_emit_move.
3e3d9d17
DD		 5354 (force_into_acc): New.
5355 (move_to_acc): Use gen_and_emit_move.
5356 (move_from_acc): Likewise.
5357 (move_acc_to_reg): Likewise.
5358 (move_to_x): Likewise.
5359 (move_to_hl): Likewise.
5360 (move_to_de): Likewise.
5361 (rl78_alloc_physical_registers_op1): Record location content.
5362 (has_constraint): New.
5363 (rl78_alloc_physical_registers_op2): Record location content.
5364 Optimize use of HL.
5365 (rl78_alloc_physical_registers_ro1): Likewise.
5366 (rl78_alloc_physical_registers_cmp): Likewise.
5367 (rl78_alloc_physical_registers_umul): Likewise.
5368 (rl78_alloc_address_registers_macax): New.
5369 (rl78_alloc_physical_registers): Initialize and set location
5370 content memory as needed.
5371 (rl78_reorg): Make sure split2 is called.
5372 (rl78_rtx_costs): New.
5373
088845a5
RS		 53742013-09-11 Richard Sandiford <rdsandiford@googlemail.com>
5375
5376 * simplify-rtx.c (simplify_unary_operation_1): Use simplify_gen_binary
5377 for (not (neg ...)) and (neg (not ...)) cases.
5378
5c2961cf
RB		 53792013-09-11 Richard Biener <rguenther@suse.de>
5380
5381 PR middle-end/58377
5382 * passes.def: Split critical edges before late uninit warning passes.
5383 * tree-cfg.c (pass_split_crit_edges): Implement clone method.
5384
427e6a14
JJ		 53852013-09-11 Jakub Jelinek <jakub@redhat.com>
5386
5387 PR tree-optimization/58385
5388 * fold-const.c (build_range_check): If both low and high are NULL,
5389 use omit_one_operand_loc to preserve exp side-effects.
5390
df45c0cc
KT		 53912013-09-11 Kyrylo Tkachov <kyrylo.tkachov@arm.com>
5392
5393 * config/arm/arm.md (arm_shiftsi3): New alternative l/l/M.
5394
80ab0b19
RB		 53952013-09-11 Richard Biener <rguenther@suse.de>
5396
5397 * tree-data-ref.c (dump_rdg_vertex, debug_rdg_vertex,
5398 dump_rdg_component, debug_rdg_component, dump_rdg, debug_rdg,
5399 dot_rdg_1, dot_rdg, rdg_vertex_for_stmt, create_rdg_edge_for_ddr,
5400 create_rdg_edges_for_scalar, create_rdg_edges, create_rdg_vertices,
5401 stmts_from_loop, known_dependences_p, build_empty_rdg,
5402 build_rdg, free_rdg, rdg_defs_used_in_other_loops_p): Move ...
5403 * tree-loop-distribution.c: ... here.
5404 * tree-data-ref.h (struct rdg_vertex, RDGV_STMT, RDGV_DATAREFS,
5405 RDGV_HAS_MEM_WRITE, RDGV_HAS_MEM_READS, RDG_STMT, RDG_DATAREFS,
5406 RDG_MEM_WRITE_STMT, RDG_MEM_READS_STMT, enum rdg_dep_type,
5407 struct rdg_edge, RDGE_TYPE, RDGE_LEVEL, RDGE_RELATION): Move ...
5408 * tree-loop-distribution.c: ... here.
5409 * tree-loop-distribution.c: Include gimple-pretty-print.h.
5410 (struct partition_s): Add loops member.
5411 (partition_alloc, partition_free, rdg_flag_uses, rdg_flag_vertex,
5412 rdg_flag_vertex_and_dependent, rdg_flag_loop_exits,
5413 build_rdg_partition_for_component, rdg_build_partitions): Adjust.
5414
85a77221
AI		 54152013-09-11 Alexander Ivchenko <alexander.ivchenko@intel.com>
5416 Maxim Kuznetsov <maxim.kuznetsov@intel.com>
5417 Sergey Lega <sergey.s.lega@intel.com>
5418 Anna Tikhonova <anna.tikhonova@intel.com>
5419 Ilya Tocar <ilya.tocar@intel.com>
5420 Andrey Turetskiy <andrey.turetskiy@intel.com>
5421 Ilya Verbin <ilya.verbin@intel.com>
5422 Kirill Yukhin <kirill.yukhin@intel.com>
5423 Michael Zolotukhin <michael.v.zolotukhin@intel.com>
5424
5425 * config/i386/constraints.md (k): New.
5426 (Yk): Ditto.
5427 * config/i386/i386.c (const regclass_map): Add new mask registers.
5428 (dbx_register_map): Ditto.
5429 (dbx64_register_map): Ditto.
5430 (svr4_dbx_register_map): Ditto.
5431 (ix86_conditional_register_usage): Squash mask registers if AVX512F is
5432 disabled.
5433 (ix86_preferred_reload_class): Disable constants for mask registers.
5434 (ix86_secondary_reload): Do spill of mask register using 32-bit insn.
5435 (ix86_hard_regno_mode_ok): Support new mask registers.
5436 (x86_order_regs_for_local_alloc): Ditto.
5437 * config/i386/i386.h (FIRST_PSEUDO_REGISTER): Update.
5438 (FIXED_REGISTERS): Add new mask registers.
5439 (CALL_USED_REGISTERS): Ditto.
5440 (REG_ALLOC_ORDER): Ditto.
5441 (VALID_MASK_REG_MODE): New.
5442 (FIRST_MASK_REG): Ditto.
5443 (LAST_MASK_REG): Ditto.
5444 (reg_class): Add MASK_EVEX_REGS, MASK_REGS.
5445 (MAYBE_MASK_CLASS_P): New.
5446 (REG_CLASS_NAMES): Add MASK_EVEX_REGS, MASK_REGS.
5447 (REG_CLASS_CONTENTS): Ditto.
5448 (MASK_REGNO_P): New.
5449 (ANY_MASK_REG_P): Ditto.
5450 (HI_REGISTER_NAMES): Add new mask registers.
893e85fa
CM		 5451 * config/i386/i386.md (MASK0_REG, MASK1_REG, MASK2_REG, MASK3_REG,
5452 MASK4_REG, MASK5_REG, MASK6_REG, MASK7_REG): Constants for new
5453 mask registers.
85a77221
AI		 5454 (attribute "type"): Add mskmov, msklog.
5455 (attribute "length_immediate"): Support them.
5456 (attribute "memory"): Ditto.
5457 (attribute "prefix_0f"): Ditto.
5458 (*movhi_internal): Support new mask registers.
5459 (*movqi_internal): Ditto.
5460 (define_split): Split out clobber pattern is a logic
5461 insn on mask registers.
5462 (*k<logic><mode>): New.
5463 (*andhi_1): Extend to support mask regs.
5464 (*andqi_1): Extend to support mask regs.
5465 (kandn<mode>): New.
5466 (define_split): Split and-not to and and not if operands
5467 are not mask regs.
5468 (*<code><mode>_1): Separate HI mode to new pattern...
5469 (*<code>hi_1): This.
5470 (*<code>qi_1): Extend to support mask regs.
5471 (kxnor<mode>): New.
5472 (kortestzhi): Ditto.
5473 (kortestchi): Ditto.
5474 (kunpckhi): Ditto.
5475 (*one_cmpl<mode>2_1): Remove HImode and handle it...
5476 (*one_cmplhi2_1): ...Here, now with mask registers support.
5477 (*one_cmplqi2_1): Support new mask registers.
893e85fa
CM		 5478 (HI/QImode arithmetics splitter): Don't split if mask registers
5479 are used.
85a77221
AI		 5480 (HI/QImode not splitter): Ditto.
5481 * config/i386/predicated.md (mask_reg_operand): New.
5482 (general_reg_operand): Ditto.
5483
7e570821
AI		 54842013-09-11 Alexander Ivchenko <alexander.ivchenko@intel.com>
5485
5486 * doc/invoke.texi: Document fxsr, xsave and xsaveopt options.
5487 * doc/extend.texi: Document fxsr, xsave and xsaveopt builtins.
5488
9e1376e9
JL		 54892013-09-10 Jeff Law <law@redhat.com>
5490
5491 PR tree-optimization/58380
5492 * tree-ssa-threadupdate.c (thread_block): Recognize another case
5493 of threading through a buried loop header.
5494
5495 * tree-ssa-threadedge.c (thread_around_empty_blocks): Correct
5496 return value for single successor case.
5497
5bbcb888
JH		 54982013-09-10 Jan Hubicka <jh@suse.cz>
5499
5500 * ipa-devirt.c (ipa_devirt): Enable with LTO.
5501
d742ff4b
RE		 55022013-09-10 Richard Earnshaw <rearnsha@arm.com>
5503
5504 PR target/58361
5505 * arm/vfp.md (combine_vcvt_f32_<FCVTI32typename>): Fix pattern to
5506 support conditional execution.
5507 (combine_vcvt_f64_<FCVTI32typename>): Likewise.
5508
b0681c9e
VM		 55092013-09-10 Vladimir Makarov <vmakarov@redhat.com>
5510
5511 * lra.c (lra): Clear lra_optional_reload_pseudos before every
5512 constraint pass.
f77e2d2d
JBG		 5513 * lra-constraints.c (curr_insn_transform): Switch on optional reloads.
5514 Check destination too to check move insn.
b0681c9e 5515 (undo_optional_reloads): Add check that the original peudo did not
f77e2d2d
JBG		 5516 changed its allocation and the optional reload was inherited on last
5517 inheritance pass. Break loop after deciding to keep optional reload.
b0681c9e
VM		 5518 (lra_undo_inheritance): Add check that inherited pseudo still in
5519 memory.
5520
c9b1c957
JG		 55212013-09-10 James Greenhalgh <james.greenhalgh@arm.com>
5522
5523 * config/aarch64/aarch64.md (generic_sched): New.
5524 * config/aarch64/aarch64-generic.md (load): Make conditional
5525 on generic_sched attribute.
5526 (nonload): Likewise.
5527
82d618d3
JH		 55282013-09-10 Jan Hubicka <jh@suse.cz>
5529
5530 * lto-cgraph.c: Include ipa-utils.h.
5531 (compute_ltrans_boundary): Also add possible targets into the boundary.
5532
668b6894
JH		 55332013-09-10 Jan Hubicka <jh@suse.cz>
5534
5535 * gimple-fold.c (gimple_get_virt_method_for_binfo): Pass real
5536 VAR_DECL of vtable rather than full expression.
5537
4f90d3e0
JH		 55382013-09-10 Jan Hubicka <jh@suse.cz>
5539 Paolo Carlini <paolo.carlini@oracle.com>
5540
5541 * cgraphunit.c (analyze_functions): Save input_location, set it
5542 to UNKNOWN_LOCATION and restore it at the end.
5543
39e87baf
MJ		 55442013-09-10 Martin Jambor <mjambor@suse.cz>
5545
5546 * ipa-cp.c (propagate_constants_topo): Do not ignore SCC
5547 represented by a thunk.
5548
581aedec
JL		 55492013-09-10 Jeff Law <law@redhat.com>
5550
5551 PR tree-optimization/58343
5552 * tree-ssa-threadupdate.c (thread_block): Identify and disable
5553 jump threading requests through loop headers buried in the middle
5554 of a jump threading path.
5555
5556 * tree-ssa-threadedge.c (thread_around_empty_blocks): Fix thinko
5557 in return value/type.
5558
84cf4ab6
JJ		 55592013-09-10 Jakub Jelinek <jakub@redhat.com>
5560
5561 PR rtl-optimization/58365
5562 * cfgcleanup.c (merge_memattrs): Also clear MEM_READONLY_P
5563 resp. MEM_NOTRAP_P if they differ, or set MEM_VOLATILE_P if
5564 it differs.
5565
447dd906
RB		 55662013-09-10 Richard Biener <rguenther@suse.de>
5567
5568 * tree-data-ref.h (build_rdg): Drop all parameters but loop.
5569 * tree-data-ref.c (create_rdg_vertices): Collect all data
5570 references, signal failure to the caller, use data-ref API.
893e85fa
CM		 5571 (build_rdg): Compute data references only once. Maintain lifetime
5572 of data references and data dependences from within RDG.
447dd906
RB		 5573 (free_rdg): Free dependence relations.
5574 * tree-loop-distribution.c (rdg_flag_uses): Drop weird code
5575 inventing extra dependences.
5576 (distribute_loop): Update for RDG API changes.
5577
f14a7324
KT		 55782013-09-10 Kai Tietz <ktietz@redhat.com>
5579
5580 * doc/invoke.texi (fms-extensions): Document changed
5581 behavior for ms-abi targets.
5582 * config/i386/i386.c (ix86_option_override_internal):
5583 Set default value of option -fms-extension for ms-abi targets.
5584
ff131f26
MZ		 55852013-09-10 Michael Zolotukhin <michael.v.zolotukhin@gmail.com>
5586
5587 * config/i386/i386.c (ix86_expand_movmem): Fix epilogue generation.
5588
33115a2c
AM		 55892013-09-10 Alan Modra <amodra@gmail.com>
5590
5591 PR target/58330
5592 * config/rs6000/rs6000.md (bswapdi2_64bit): Disable for volatile mems.
5593
b6872a13
AM		 55942013-09-10 Alan Modra <amodra@gmail.com>
5595
5596 * config/rs6000/predicates.md (add_cint_operand): New.
5597 (reg_or_add_cint_operand, small_toc_ref): Use add_cint_operand.
5598 * config/rs6000/rs6000.md (largetoc_high_plus): Restrict offset
5599 using add_cint_operand.
5600 (largetoc_high_plus_aix): Likewise.
5601
28fd0ba2
JJ		 56022013-09-09 Jakub Jelinek <jakub@redhat.com>
5603
5604 PR tree-optimization/58364
5605 * tree-ssa-reassoc.c (init_range_entry): For BIT_NOT_EXPR on
5606 BOOLEAN_TYPE, only invert in_p and continue with arg0 if
5607 the current range can't be an unconditional true or false.
5608
552d99fe
JG		 56092013-09-09 James Greenhalgh <james.greenhalgh@arm.com>
5610
5611 * config/aarch64/arm_neon.h (vrsqrte_f64): Fix parameter type.
5612
f43245d1
UB		 56132013-09-09 Uros Bizjak <ubizjak@gmail.com>
5614
5615 * ipa-prop.c (ipa_modify_call_arguments): Initialize deref_align.
5616
c9b0866a
PC		 56172013-09-09 Paolo Carlini <paolo.carlini@oracle.com>
5618
5619 PR c++/43452
5620 * doc/invoke.texi (-Wdelete-incomplete): Document it.
5621
78d8b9f0
IB		 56222013-09-09 Ian Bolton <ian.bolton@arm.com>
5623
5624 * config/aarch64/aarch64.c (aarch64_preferred_reload_class): Return
5625 NO_REGS for immediate that can't be moved directly into FP_REGS.
5626
274b2532
KT		 56272013-09-09 Kyrylo Tkachov <kyrylo.tkachov@arm.com>
5628
5629 * config/aarch64/aarch64.c (aarch64_select_cc_mode): Return CC_SWP for
5630 comparison with negated operand.
c9b0866a
PC		 5631 * config/aarch64/aarch64.md (compare_neg<mode>): Match canonical
5632 RTL form.
274b2532 5633
467a3558
RB		 56342013-09-09 Richard Biener <rguenther@suse.de>
5635
5636 PR middle-end/58326
5637 * cfgloopmanip.c (fix_bb_placements): When fixing the placement
5638 of a subloop record all its block as affecting loop-closed SSA form.
5639
088c5368
RS		 56402013-09-09 Richard Sandiford <rdsandiford@googlemail.com>
5641
5642 * expmed.c (lshift_value): Take an unsigned HOST_WIDE_INT instead
5643 of an rtx/bitpos pair.
5644 (store_fixed_bit_field): Update accordingly.
5645
69db2d57
RS		 56462013-09-09 Richard Sandiford <rdsandiford@googlemail.com>
5647
5648 * asan.c (asan_emit_stack_protection): Use gen_int_mode instead of
5649 GEN_INT.
5650 * builtins.c (expand_errno_check): Likewise.
5651 * dwarf2cfi.c (init_return_column_size): Likewise.
5652 * except.c (sjlj_mark_call_sites): Likewise.
5653 * expr.c (move_by_pieces_1, store_by_pieces_2): Likewise.
5654 * lra-constraints.c (emit_inc): Likewise.
5655 * ree.c (combine_set_extension): Likewise.
5656 * regmove.c (fixup_match_2): Likewise.
5657 * reload1.c (inc_for_reload): Likewise.
5658
69a59f0f
RS		 56592013-09-09 Richard Sandiford <rdsandiford@googlemail.com>
5660
5661 * combine.c (simplify_set, expand_field_assignment, extract_left_shift)
5662 (force_to_mode, simplify_shift_const_1, simplify_comparison):
5663 Use gen_int_mode with the mode of the associated simplify_* call.
5664 * explow.c (probe_stack_range, anti_adjust_stack_and_probe): Likewise.
5665 * expmed.c (expand_shift_1): Likewise.
5666 * function.c (instantiate_virtual_regs_in_insn): Likewise.
5667 * loop-iv.c (iv_number_of_iterations): Likewise.
5668 * loop-unroll.c (unroll_loop_runtime_iterations): Likewise.
5669 * simplify-rtx.c (simplify_binary_operation_1): Likewise.
5670
2f1cd2eb
RS		 56712013-09-09 Richard Sandiford <rdsandiford@googlemail.com>
5672
5673 * asan.c (asan_clear_shadow): Use gen_int_mode with the mode
5674 of the associated expand_* call.
5675 (asan_emit_stack_protection): Likewise.
5676 * builtins.c (round_trampoline_addr): Likewise.
5677 * explow.c (allocate_dynamic_stack_space, probe_stack_range): Likewise.
5678 * expmed.c (expand_smod_pow2, expand_sdiv_pow2, expand_divmod)
5679 (emit_store_flag): Likewise.
5680 * expr.c (emit_move_resolve_push, push_block, emit_single_push_insn_1)
5681 (emit_push_insn, optimize_bitfield_assignment_op, expand_expr_real_1):
5682 Likewise.
5683 * function.c (instantiate_virtual_regs_in_insn): Likewise.
5684 * ifcvt.c (noce_try_store_flag_constants): Likewise.
5685 * loop-unroll.c (unroll_loop_runtime_iterations): Likewise.
5686 * modulo-sched.c (generate_prolog_epilog): Likewise.
5687 * optabs.c (expand_binop, widen_leading, expand_doubleword_clz)
5688 (expand_ctz, expand_ffs, expand_unop): Likewise.
5689
4789c0ce
RS		 56902013-09-09 Richard Sandiford <rdsandiford@googlemail.com>
5691
5692 * alias.c (addr_side_effect_eval): Use gen_int_mode with the mode
5693 of the associated gen_rtx_* call.
5694 * caller-save.c (init_caller_save): Likewise.
5695 * combine.c (find_split_point, make_extraction): Likewise.
5696 (make_compound_operation): Likewise.
5697 * dwarf2out.c (mem_loc_descriptor): Likewise.
5698 * explow.c (plus_constant, probe_stack_range): Likewise.
5699 * expmed.c (expand_mult_const): Likewise.
5700 * expr.c (emit_single_push_insn_1, do_tablejump): Likewise.
5701 * reload1.c (init_reload): Likewise.
5702 * valtrack.c (cleanup_auto_inc_dec): Likewise.
5703 * var-tracking.c (adjust_mems): Likewise.
5704 * modulo-sched.c (sms_schedule): Likewise, but use gen_rtx_GT
5705 rather than gen_rtx_fmt_ee.
5706
a811f7cb
JH		 57072013-09-09 Jan Hubicka <jh@suse.cz>
5708
5709 PR middle-end/58294
5710 * value-prof.c (gimple_ic): Copy also abnormal edges.
5711
dcad1dd3
RS		 57122013-09-09 Richard Sandiford <rdsandiford@googlemail.com>
5713
5714 * asan.c (asan_shadow_cst): Use gen_int_mode.
5715
daf5c770
JH		 57162013-09-08 Jan Hubicka <jh@suse.cz>
5717
5718 * ipa-profile.c: Add toplevel comment.
5719 (ipa_propagate_frequency_1): Be more conservative when profile is read.
5720 (contains_hot_call_p): New function.
5721 (ipa_propagate_frequency): Set frequencies based on counts when
5722 profile is read.
5723 * predict.c (compute_function_frequency): Use PROFILE_READ gueard for
5724 profile; do not tamper with profile after inlining if it is read.
5725
a12cd2db
JH		 57262013-09-08 Jan Hubicka <jh@suse.cz>
5727
5728 * ipa-prop.c (try_make_edge_direct_simple_call): Do not special case
5729 speculative edges.
5730
856a1588
JH		 57312013-09-08 Jan Hubicka <jh@suse.cz>
5732
5733 * ipa.c (walk_polymorphic_call_targets): Fix redirection before IPA
5734 summary generation.
5735
4f4b0b73
JL		 57362013-09-08 Jeff Law <law@redhat.com>
5737
5738 PR bootstrap/58340
5739 * tree-ssa-threadedge.c (thread_across_edge): Fix initialization
5740 of 'found'.
5741
cc4029ee
AK		 57422013-09-08 Andi Kleen <ak@linux.intel.com>
5743
5744 * tree-inline.c (estimate_num_insns): Limit asm cost to 1000.
5745
7b395ddd
JH		 57462013-09-08 Jan Hubicka <jh@suse.cz>
5747
5748 * ipa.c (walk_polymorphic_call_targets): Fix inliner summary update.
5749
c34c46dd
RS		 57502013-09-08 Richard Sandiford <rdsandiford@googlemail.com>
5751
5752 * ira.c (update_equiv_regs): Only call set_paradoxical_subreg
5753 for non-debug insns.
5754 * lra.c (new_insn_reg): Take the containing insn as a parameter.
5755 Only modify lra_reg_info[].biggest_mode if it's non-debug insn.
5756 (collect_non_operand_hard_regs, add_regs_to_insn_regno_info): Update
5757 accordingly.
5758
3462aa02
JH		 57592013-09-08 Jan Hubicka <jh@suse.cz>
5760
5761 * cgraphunit.c (walk_polymorphic_call_targets): Permit 0 possible
5762 targets and devirtualize to BUILT_IN_UNREACHABLE.
5763 * timevar.def (TV_IPA_UNREACHABLE): New timevar.
5764 * ipa.c (walk_polymorphic_call_targets): New function.
5765 (symtab_remove_unreachable_nodes): Use it; do not keep all virtual
5766 functions; use the new timevar.
5767 * ipa-devirt.c (maybe_record_node): Do not insert static nodes that
5768 was removed from the program.
5769 (record_binfo): If BINFO corresponds to an anonymous namespace, we may
5770 not consider it in the walk when its vtable is dead.
5771 (possible_polymorphic_call_targets_1): Pass anonymous flag to
5772 record_binfo.
5773 (devirt_variable_node_removal_hook): New function.
5774 (possible_polymorphic_call_targets): Also register
5775 devirt_variable_node_removal_hook.
5776 (ipa_devirt): Do not do non-speculative devirtualization.
5777 (gate_ipa_devirt): One execute if devirtualizing speculatively.
5778
26e5b0fd
JH		 57792013-09-08 Jan Hubicka <jh@suse.cz>
5780
5781 * cgraph.h (varpool_node_hook, varpool_node_hook_list,
5782 varpool_add_node_removal_hook, varpool_add_variable_insertion_hook,
5783 varpool_remove_variable_insertion_hook): Declare.
5784 * varpool.c (varpool_node_hook_list): New structure.
5785 (first_varpool_node_removal_hook,
5786 first_varpool_variable_insertion_hook): New variables.
5787 (varpool_add_node_removal_hook, varpool_remove_node_removal_hook,
5788 varpool_call_node_removal_hooks, varpool_add_variable_insertion_hook,
5789 varpool_remove_variable_insertion_hook,
5790 varpool_call_variable_insertion_hooks): New functions.
5791 (varpool_remove_node): Use it.
5792
97f1f314 57932013-09-08 Paolo Carlini <paolo.carlini@oracle.com>
6df2ca09 5794
97f1f314
PC		 5795 PR c++/54941
5796 * diagnostic.c (diagnostic_build_prefix): When s.file is
5797 "<built-in>" don't output line and column numbers.
6df2ca09 5798
bcb650cb
JH		 57992013-09-06 Jan Hubicka <jh@suse.cz>
5800
5801 * cgraphunit.c (expand_thunk): Get body before touching arguments.
5802 * lto-streamer-out.c: Stream thunks, too.
5803 * lto-streamer-in.c (input_function): Pop cfun here
5804 (lto_read_body): Instead of here.
5805
87f3fea8
CT		 58062013-09-06 Caroline Tice <cmtice@google.com>
5807
5808 * doc/install.texi: Add documentation for the --enable-vtable-verify
5809 and the --disable-libvtv configure options.
5810
5e8d9b45
JL		 58112013-09-06 Jeff Law <law@redhat.com>
5812
5813 * tree-ssa-dom.c (cprop_into_successor_phis): Also propagate
5814 edge implied equivalences into successor phis.
5815
3a001aff
JR		 58162013-09-06 Joern Rennecke <joern.rennecke@embecosm.com>
5817
5818 * resource.c (mark_referenced_resources): Handle COND_EXEC.
5819
58202013-09-06 Claudiu Zissulescu <claziss@synopsys.com>
c42091c1
CZ		 5821
5822 * resource.c (mark_target_live_regs): Compute resources taking
5823 into account if a call is predicated or not.
5824
2c641f8b
EB		 58252013-09-06 Eric Botcazou <ebotcazou@adacore.com>
5826
5827 * toplev.c (output_stack_usage): Be prepared for suffixes created by
5828 the compiler in the function names.
5829
a81b0a3d
JH		 58302013-09-06 Jan Hubicka <jh@suse.cz>
5831
5832 PR middle-end/58094
5833 * ipa-inline.c (has_caller_p): New function.
5834 (want_inline_function_to_all_callers_p): Use it.
5835 (sum_callers, inline_to_all_callers): Break out from ...
5836 (ipa_inline): ... here.
5837
68e4ca7e
JH		 58382013-09-06 Jan Hubicka <jh@suse.cz>
5839
2c641f8b
EB		 5840 * config/i386/i386.c (ix86_hard_regno_mode_ok): AVX modes are valid
5841 only when AVX is enabled.
68e4ca7e 5842
3c73e2fe
JG		 58432013-09-06 James Greenhalgh <james.greenhalgh@arm.com>
5844
5845 * config/aarch64/aarch64.md
5846 (*movtf_aarch64): Use neon_<ls>dm_2 as type where v8type
5847 is fpsimd_<load/store>2.
5848 (load_pair<mode>): Likewise.
5849 (store_pair<mode>): Likewise.
5850
8c48eecd
JG		 58512013-09-06 James Greenhalgh <james.greenhalgh@arm.com>
5852
5853 * config/arm/types.md (type): Add "mrs" type.
5854 * config/aarch64/aarch64.md
5855 (aarch64_load_tp_hard): Make type "mrs".
5856 * config/arm/arm.md
5857 (load_tp_hard): Make type "mrs".
5858 * config/arm/cortex-a15.md: Update with new attributes.
5859 * config/arm/cortex-a5.md: Update with new attributes.
5860 * config/arm/cortex-a53.md: Update with new attributes.
5861 * config/arm/cortex-a7.md: Update with new attributes.
5862 * config/arm/cortex-a8.md: Update with new attributes.
5863 * config/arm/cortex-a9.md: Update with new attributes.
5864 * config/arm/cortex-m4.md: Update with new attributes.
5865 * config/arm/cortex-r4.md: Update with new attributes.
5866 * config/arm/fa526.md: Update with new attributes.
5867 * config/arm/fa606te.md: Update with new attributes.
5868 * config/arm/fa626te.md: Update with new attributes.
5869 * config/arm/fa726te.md: Update with new attributes.
5870
9ea187a5
JG		 58712013-09-06 James Greenhalgh <james.greenhalgh@arm.com>
5872
5873 * config/aarch64/aarch64.md
5874 (*movti_aarch64): Use "multiple" for type where v8type is "move2".
5875 (*movtf_aarch64): Likewise.
5876 * config/arm/arm.md
5877 (thumb1_movdi_insn): Use "multiple" for type where more than one
5878 instruction is used for a move.
5879 (*arm32_movhf): Likewise.
5880 (*thumb_movdf_insn): Likewise.
5881
292b89b3
JG		 58822013-09-06 James Greenhalgh <james.greenhalgh@arm.com>
5883
5884 * config/arm/types.md (type): Rename fcpys to fmov.
5885 * config/arm/vfp.md
5886 (*arm_movsi_vfp): Rename type fcpys as fmov.
5887 (*thumb2_movsi_vfp): Likewise
5888 (*movhf_vfp_neon): Likewise
5889 (*movhf_vfp): Likewise
5890 (*movsf_vfp): Likewise
5891 (*thumb2_movsf_vfp): Likewise
5892 (*movsfcc_vfp): Likewise
5893 (*thumb2_movsfcc_vfp): Likewise
5894 * config/aarch64/aarch64-simd.md
5895 (move_lo_quad_<mode>): Replace type mov_reg with fmovs.
5896 * config/aarch64/aarch64.md
5897 (*movsi_aarch64): Replace type mov_reg with fmovs.
5898 (*movdi_aarch64): Likewise
5899 (*movsf_aarch64): Likewise
5900 (*movdf_aarch64): Likewise
5901 * config/arm/arm.c
5902 (cortexa7_older_only): Rename TYPE_FCPYS to TYPE_FMOV.
5903 * config/arm/iwmmxt.md
5904 (*iwmmxt_movsi_insn): Rename type fcpys as fmov.
5905 * config/arm/arm1020e.md: Update with new attributes.
5906 * config/arm/cortex-a15-neon.md: Update with new attributes.
5907 * config/arm/cortex-a5.md: Update with new attributes.
5908 * config/arm/cortex-a53.md: Update with new attributes.
5909 * config/arm/cortex-a7.md: Update with new attributes.
5910 * config/arm/cortex-a8-neon.md: Update with new attributes.
5911 * config/arm/cortex-a9.md: Update with new attributes.
5912 * config/arm/cortex-m4-fpu.md: Update with new attributes.
5913 * config/arm/cortex-r4f.md: Update with new attributes.
5914 * config/arm/marvell-pj4.md: Update with new attributes.
5915 * config/arm/vfp11.md: Update with new attributes.
5916
6a24a83e
JG		 59172013-09-06 James Greenhalgh <james.greenhalgh@arm.com>
5918
5919 * config/aarch64/aarch64.md
5920 (*madd<mode>): Fix type attribute.
5921 (*maddsi_uxtw): Likewise.
5922 (*msub<mode>): Likewise.
5923 (*msubsi_uxtw): Likewise.
5924 (<su_optab>maddsidi4): Likewise.
5925 (<su_optab>msubsidi4): Likewise.
5926
b86923f0
JG		 59272013-09-06 James Greenhalgh <james.greenhalgh@arm.com>
5928
5929 * config/arm/types.md: Split fdiv<sd> as fsqrt<sd>, fdiv<sd>.
5930 * config/arm/arm.md (core_cycles): Remove fdiv.
5931 * config/arm/vfp.md:
5932 (*sqrtsf2_vfp): Update for attribute changes.
5933 (*sqrtdf2_vfp): Likewise.
5934 * config/aarch64/aarch64.md:
5935 (sqrt<mode>2): Update for attribute changes.
5936 * config/arm/arm1020e.md: Update with new attributes.
5937 * config/arm/cortex-a15-neon.md: Update with new attributes.
5938 * config/arm/cortex-a5.md: Update with new attributes.
5939 * config/arm/cortex-a53.md: Update with new attributes.
5940 * config/arm/cortex-a7.md: Update with new attributes.
5941 * config/arm/cortex-a8-neon.md: Update with new attributes.
5942 * config/arm/cortex-a9.md: Update with new attributes.
5943 * config/arm/cortex-m4-fpu.md: Update with new attributes.
5944 * config/arm/cortex-r4f.md: Update with new attributes.
5945 * config/arm/marvell-pj4.md: Update with new attributes.
5946 * config/arm/vfp11.md: Update with new attributes.
5947
7b49c9e1
JG		 59482013-09-06 James Greenhalgh <james.greenhalgh@arm.com>
5949
5950 * config/arm/types.md
5951 (type): Split f_cvt as f_cvt, f_cvtf2i, f_cvti2f.
5952 * config/aarch64/aarch64.md
5953 (l<fcvt_pattern><su_optab><GPF:mode><GPI:mode>2): Update with
5954 new attributes.
5955 (fix_trunc<GPF:mode><GPI:mode>2): Likewise.
5956 (fixuns_trunc<GPF:mode><GPI:mode>2): Likewise.
5957 (float<GPI:mode><GPF:mode>2): Likewise.
5958 * config/arm/vfp.md
5959 (*truncsisf2_vfp): Update with new attributes.
5960 (*truncsidf2_vfp): Likewise.
5961 (fixuns_truncsfsi2): Likewise.
5962 (fixuns_truncdfsi2): Likewise.
5963 (*floatsisf2_vfp): Likewise.
5964 (*floatsidf2_vfp): Likewise.
5965 (floatunssisf2): Likewise.
5966 (floatunssidf2): Likewise.
5967 (*combine_vcvt_f32_<FCVTI32typename>): Likewise.
5968 (*combine_vcvt_f64_<FCVTI32typename>): Likewise.
5969 * config/arm/arm1020e.md: Update with new attributes.
5970 * config/arm/cortex-a15-neon.md: Update with new attributes.
5971 * config/arm/cortex-a5.md: Update with new attributes.
5972 * config/arm/cortex-a53.md: Update with new attributes.
5973 * config/arm/cortex-a7.md: Update with new attributes.
5974 * config/arm/cortex-a8-neon.md: Update with new attributes.
5975 * config/arm/cortex-a9.md: Update with new attributes.
5976 * config/arm/cortex-m4-fpu.md: Update with new attributes.
5977 * config/arm/cortex-r4f.md: Update with new attributes.
5978 * config/arm/marvell-pj4.md: Update with new attributes.
5979 * config/arm/vfp11.md: Update with new attributes.
5980
e7df8af8
JG		 59812013-09-06 James Greenhalgh <james.greenhalgh@arm.com>
5982
5983 * config/aarch64/arm_neon.h
5984 (vqtbl<1,2,3,4><q>_s8): Fix control vector parameter type.
5985 (vqtbx<1,2,3,4><q>_s8): Likewise.
5986
594726e4
JG		 59872013-09-06 James Greenhalgh <james.greenhalgh@arm.com>
5988
f43245d1
UB		 5989 * config/arm/types.md: Add "no_insn", "multiple" and "untyped" types.
5990 * config/arm/arm-fixed.md: Add type attribute to all insn patterns.
594726e4
JG		 5991 (add<mode>3): Add type attribute.
5992 (add<mode>3): Likewise.
5993 (usadd<mode>3): Likewise.
5994 (ssadd<mode>3): Likewise.
5995 (sub<mode>3): Likewise.
5996 (sub<mode>3): Likewise.
5997 (ussub<mode>3): Likewise.
5998 (sssub<mode>3): Likewise.
5999 (ssmulsa3): Likewise.
6000 (usmulusa3): Likewise.
6001 (arm_usatsihi): Likewise.
6002 * config/arm/vfp.md
6003 (*movdi_vfp): Add types for all instructions.
6004 (*movdi_vfp_cortexa8): Likewise.
6005 (*movhf_vfp_neon): Likewise.
6006 (*movhf_vfp): Likewise.
6007 (*movdf_vfp): Likewise.
6008 (*thumb2_movdf_vfp): Likewise.
6009 (*thumb2_movdfcc_vfp): Likewise.
6010 * config/arm/arm.md: Add type attribute to all insn patterns.
6011 (*thumb1_adddi3): Add type attribute.
6012 (*arm_adddi3): Likewise.
6013 (*adddi_sesidi_di): Likewise.
6014 (*adddi_zesidi_di): Likewise.
6015 (*thumb1_addsi3): Likewise.
6016 (addsi3_compare0): Likewise.
6017 (*addsi3_compare0_scratch): Likewise.
6018 (*compare_negsi_si): Likewise.
6019 (cmpsi2_addneg): Likewise.
6020 (*addsi3_carryin_<optab>): Likewise.
6021 (*addsi3_carryin_alt2_<optab>): Likewise.
6022 (*addsi3_carryin_clobercc_<optab>): Likewise.
6023 (*subsi3_carryin): Likewise.
6024 (*subsi3_carryin_const): Likewise.
6025 (*subsi3_carryin_compare): Likewise.
6026 (*subsi3_carryin_compare_const): Likewise.
6027 (*arm_subdi3): Likewise.
6028 (*thumb_subdi3): Likewise.
6029 (*subdi_di_zesidi): Likewise.
6030 (*subdi_di_sesidi): Likewise.
6031 (*subdi_zesidi_di): Likewise.
6032 (*subdi_sesidi_di): Likewise.
6033 (*subdi_zesidi_ze): Likewise.
6034 (thumb1_subsi3_insn): Likewise.
6035 (*arm_subsi3_insn): Likewise.
6036 (*anddi3_insn): Likewise.
6037 (*anddi_zesidi_di): Likewise.
6038 (*anddi_sesdi_di): Likewise.
6039 (*ne_zeroextracts): Likewise.
6040 (*ne_zeroextracts): Likewise.
6041 (*ite_ne_zeroextr): Likewise.
6042 (*ite_ne_zeroextr): Likewise.
6043 (*anddi_notdi_di): Likewise.
6044 (*anddi_notzesidi): Likewise.
6045 (*anddi_notsesidi): Likewise.
6046 (andsi_notsi_si): Likewise.
6047 (thumb1_bicsi3): Likewise.
6048 (*iordi3_insn): Likewise.
6049 (*iordi_zesidi_di): Likewise.
6050 (*iordi_sesidi_di): Likewise.
6051 (*thumb1_iorsi3_insn): Likewise.
6052 (*xordi3_insn): Likewise.
6053 (*xordi_zesidi_di): Likewise.
6054 (*xordi_sesidi_di): Likewise.
6055 (*arm_xorsi3): Likewise.
6056 (*andsi_iorsi3_no): Likewise.
6057 (*smax_0): Likewise.
6058 (*smax_m1): Likewise.
6059 (*arm_smax_insn): Likewise.
6060 (*smin_0): Likewise.
6061 (*arm_smin_insn): Likewise.
6062 (*arm_umaxsi3): Likewise.
6063 (*arm_uminsi3): Likewise.
6064 (*minmax_arithsi): Likewise.
6065 (*minmax_arithsi_): Likewise.
6066 (*satsi_<SAT:code>): Likewise.
6067 (arm_ashldi3_1bit): Likewise.
6068 (arm_ashrdi3_1bit): Likewise.
6069 (arm_lshrdi3_1bit): Likewise.
6070 (*arm_negdi2): Likewise.
6071 (*thumb1_negdi2): Likewise.
6072 (*arm_negsi2): Likewise.
6073 (*thumb1_negsi2): Likewise.
6074 (*negdi_extendsid): Likewise.
6075 (*negdi_zero_extend): Likewise.
6076 (*arm_abssi2): Likewise.
6077 (*thumb1_abssi2): Likewise.
6078 (*arm_neg_abssi2): Likewise.
6079 (*thumb1_neg_abss): Likewise.
6080 (one_cmpldi2): Likewise.
6081 (extend<mode>di2): Likewise.
6082 (*compareqi_eq0): Likewise.
6083 (*arm_extendhisi2addsi): Likewise.
6084 (*arm_movdi): Likewise.
6085 (*thumb1_movdi_insn): Likewise.
6086 (*arm_movt): Likewise.
6087 (*thumb1_movsi_insn): Likewise.
6088 (pic_add_dot_plus_four): Likewise.
6089 (pic_add_dot_plus_eight): Likewise.
6090 (tls_load_dot_plus_eight): Likewise.
6091 (*thumb1_movhi_insn): Likewise.
6092 (*thumb1_movsf_insn): Likewise.
6093 (*movdf_soft_insn): Likewise.
6094 (*thumb_movdf_insn): Likewise.
6095 (cbranchsi4_insn): Likewise.
6096 (cbranchsi4_scratch): Likewise.
6097 (*negated_cbranchsi4): Likewise.
6098 (*tbit_cbranch): Likewise.
6099 (*tlobits_cbranch): Likewise.
6100 (*tstsi3_cbranch): Likewise.
6101 (*cbranchne_decr1): Likewise.
6102 (*addsi3_cbranch): Likewise.
6103 (*addsi3_cbranch_scratch): Likewise.
6104 (*arm_cmpdi_insn): Likewise.
6105 (*arm_cmpdi_unsig): Likewise.
6106 (*arm_cmpdi_zero): Likewise.
6107 (*thumb_cmpdi_zero): Likewise.
6108 (*deleted_compare): Likewise.
6109 (*mov_scc): Likewise.
6110 (*mov_negscc): Likewise.
6111 (*mov_notscc): Likewise.
6112 (*cstoresi_eq0_thumb1_insn): Likewise.
6113 (cstoresi_nltu_thumb1): Likewise.
6114 (cstoresi_ltu_thu): Likewise.
6115 (thumb1_addsi3_addgeu): Likewise.
6116 (*arm_jump): Likewise.
6117 (*thumb_jump): Likewise.
6118 (*check_arch2): Likewise.
6119 (arm_casesi_internal): Likewise.
6120 (thumb1_casesi_dispatch): Likewise.
6121 (*arm_indirect_jump): Likewise.
6122 (*thumb1_indirect_jump): Likewise.
6123 (nop): Likewise.
6124 (*and_scc): Likewise.
6125 (*ior_scc): Likewise.
6126 (*compare_scc): Likewise.
6127 (*cond_move): Likewise.
6128 (*cond_arith): Likewise.
6129 (*cond_sub): Likewise.
6130 (*cmp_ite0): Likewise.
6131 (*cmp_ite1): Likewise.
6132 (*cmp_and): Likewise.
6133 (*cmp_ior): Likewise.
6134 (*ior_scc_scc): Likewise.
6135 (*ior_scc_scc_cmp): Likewise.
6136 (*and_scc_scc): Likewise.
6137 (*and_scc_scc_cmp): Likewise.
6138 (*and_scc_scc_nod): Likewise.
6139 (*negscc): Likewise.
6140 (movcond_addsi): Likewise.
6141 (movcond): Likewise.
6142 (*ifcompare_plus_move): Likewise.
6143 (*if_plus_move): Likewise.
6144 (*ifcompare_move_plus): Likewise.
6145 (*if_move_plus): Likewise.
6146 (*ifcompare_arith_arith): Likewise.
6147 (*if_arith_arith): Likewise.
6148 (*ifcompare_arith_move): Likewise.
6149 (*if_arith_move): Likewise.
6150 (*ifcompare_move_arith): Likewise.
6151 (*if_move_arith): Likewise.
6152 (*ifcompare_move_not): Likewise.
6153 (*if_move_not): Likewise.
6154 (*ifcompare_not_move): Likewise.
6155 (*if_not_move): Likewise.
6156 (*ifcompare_shift_move): Likewise.
6157 (*if_shift_move): Likewise.
6158 (*ifcompare_move_shift): Likewise.
6159 (*if_move_shift): Likewise.
6160 (*ifcompare_shift_shift): Likewise.
6161 (*ifcompare_not_arith): Likewise.
6162 (*ifcompare_arith_not): Likewise.
6163 (*if_arith_not): Likewise.
6164 (*ifcompare_neg_move): Likewise.
6165 (*if_neg_move): Likewise.
6166 (*ifcompare_move_neg): Likewise.
6167 (*if_move_neg): Likewise.
6168 (prologue_thumb1_interwork): Likewise.
6169 (*cond_move_not): Likewise.
6170 (*sign_extract_onebit): Likewise.
6171 (*not_signextract_onebit): Likewise.
6172 (stack_tie): Likewise.
6173 (align_4): Likewise.
6174 (align_8): Likewise.
6175 (consttable_end): Likewise.
6176 (consttable_1): Likewise.
6177 (consttable_2): Likewise.
6178 (consttable_4): Likewise.
6179 (consttable_8): Likewise.
6180 (consttable_16): Likewise.
6181 (*thumb1_tablejump): Likewise.
6182 (prefetch): Likewise.
6183 (force_register_use): Likewise.
6184 (thumb_eh_return): Likewise.
6185 (load_tp_hard): Likewise.
6186 (load_tp_soft): Likewise.
6187 (tlscall): Likewise.
6188 (*arm_movtas_ze): Likewise.
6189 (*arm_rev): Likewise.
6190 (*arm_revsh): Likewise.
6191 (*arm_rev16): Likewise.
6192 * config/arm/thumb2.md
6193 (*thumb2_smaxsi3): Likewise.
6194 (*thumb2_sminsi3): Likewise.
6195 (*thumb32_umaxsi3): Likewise.
6196 (*thumb2_uminsi3): Likewise.
6197 (*thumb2_negdi2): Likewise.
6198 (*thumb2_abssi2): Likewise.
6199 (*thumb2_neg_abss): Likewise.
6200 (*thumb2_movsi_insn): Likewise.
6201 (tls_load_dot_plus_four): Likewise.
6202 (*thumb2_movhi_insn): Likewise.
6203 (*thumb2_mov_scc): Likewise.
6204 (*thumb2_mov_negs): Likewise.
6205 (*thumb2_mov_negs): Likewise.
6206 (*thumb2_mov_nots): Likewise.
6207 (*thumb2_mov_nots): Likewise.
6208 (*thumb2_movsicc_): Likewise.
6209 (*thumb2_movsfcc_soft_insn): Likewise.
6210 (*thumb2_indirect_jump): Likewise.
6211 (*thumb2_and_scc): Likewise.
6212 (*thumb2_ior_scc): Likewise.
6213 (*thumb2_ior_scc_strict_it): Likewise.
6214 (*thumb2_cond_move): Likewise.
6215 (*thumb2_cond_arith): Likewise.
6216 (*thumb2_cond_ari): Likewise.
6217 (*thumb2_cond_sub): Likewise.
6218 (*thumb2_negscc): Likewise.
6219 (*thumb2_movcond): Likewise.
6220 (thumb2_casesi_internal): Likewise.
6221 (thumb2_casesi_internal_pic): Likewise.
6222 (*thumb2_alusi3_short): Likewise.
6223 (*thumb2_mov<mode>_shortim): Likewise.
6224 (*thumb2_addsi_short): Likewise.
6225 (*thumb2_subsi_short): Likewise.
6226 (thumb2_addsi3_compare0): Likewise.
6227 (*thumb2_cbz): Likewise.
6228 (*thumb2_cbnz): Likewise.
6229 (*thumb2_one_cmplsi2_short): Likewise.
6230 (*thumb2_negsi2_short): Likewise.
6231 (*orsi_notsi_si): Likewise.
6232 * config/arm/arm1020e.md: Update with new attributes.
6233 * config/arm/arm1026ejs.md: Update with new attributes.
6234 * config/arm/arm1136jfs.md: Update with new attributes.
6235 * config/arm/arm926ejs.md: Update with new attributes.
6236 * config/arm/cortex-a15.md: Update with new attributes.
6237 * config/arm/cortex-a5.md: Update with new attributes.
6238 * config/arm/cortex-a53.md: Update with new attributes.
6239 * config/arm/cortex-a7.md: Update with new attributes.
6240 * config/arm/cortex-a8.md: Update with new attributes.
6241 * config/arm/cortex-a9.md: Update with new attributes.
6242 * config/arm/cortex-m4.md: Update with new attributes.
6243 * config/arm/cortex-r4.md: Update with new attributes.
6244 * config/arm/fa526.md: Update with new attributes.
6245 * config/arm/fa606te.md: Update with new attributes.
6246 * config/arm/fa626te.md: Update with new attributes.
6247 * config/arm/fa726te.md: Update with new attributes.
6248
1c83b673
JG		 62492013-09-06 James Greenhalgh <james.greenhalgh@arm.com>
6250
6251 * config/aarch64/aarch64-simd.md
6252 (aarch64_sqdml<SBINQOPS:as>l_n<mode>_internal): Use
6253 <vwx> iterator to ensure correct register choice.
6254 (aarch64_sqdml<SBINQOPS:as>l2_n<mode>_internal): Likewise.
6255 (aarch64_sqdmull_n<mode>): Likewise.
6256 (aarch64_sqdmull2_n<mode>_internal): Likewise.
6257 * config/aarch64/arm_neon.h
6258 (vml<as><q>_lane<q>_<su>16): Use 'x' constraint for element vector.
6259 (vml<as><q>_n_<su>16): Likewise.
6260 (vml<as>l_high_lane<q>_<su>16): Likewise.
6261 (vml<as>l_high_n_<su>16): Likewise.
6262 (vml<as>l_lane<q>_<su>16): Likewise.
6263 (vml<as>l_n_<su>16): Likewise.
6264 (vmul<q>_lane<q>_<su>16): Likewise.
6265 (vmul<q>_n_<su>16): Likewise.
6266 (vmull_lane<q>_<su>16): Likewise.
6267 (vmull_n_<su>16): Likewise.
6268 (vmull_high_lane<q>_<su>16): Likewise.
6269 (vmull_high_n_<su>16): Likewise.
6270 (vqrdmulh<q>_n_s16): Likewise.
6271
bb1ae543
TB		 62722013-09-06 Tejas Belagod <tejas.belagod@arm.com>
6273
6274 * config/aarch64/arm_neon.h: Fix all vdup<bhsd_lane<q> intrinsics to
6275 have the correct lane parameter.
6276
f23c0742
RB		 62772013-09-06 Richard Biener <rguenther@suse.de>
6278
6279 * cfganal.c (control_dependences::~control_dependences):
6280 Properly free all of the vector.
6281
c1d49770
KY		 62822013-09-06 Kirill Yukhin <kirill.yukhin@intel.com>
6283
6284 PR target/58269
6285 * config/i386/i386.c (ix86_conditional_register_usage):
6286 Proper initialize extended SSE registers.
6287
6da47f52
JH		 62882013-09-06 Jan Hubicka <jh@suse.cz>
6289
6290 PR tree-optimization/58311
6291 * ipa-devirt.c (gate_ipa_devirt): Only execute when optimizing.
6292
62932013-09-06 Jan Hubicka <jh@suse.cz>
6294
6295 * Makefile.in (tree-sra.o): Update dependencies.
6296 * tree-sra.c: Include ipa-utils.h
6297 (scan_function): Use recursive_call_p.
6298 (has_caller_p): New function.
6299 (cgraph_for_node_and_aliases): Count also callers of aliases.
6300
fc11f321
JH		 63012013-09-06 Jan Hubicka <jh@suse.cz>
6302
6303 PR middle-end/58094
6304 * cgraph.h (symtab_semantically_equivalent_p): Declare.
6305 * tree-tailcall.c: Include ipa-utils.h.
6306 (find_tail_calls): Use it.
6307 * ipa-pure-const.c (check_call): Likewise.
6308 * ipa-utils.c (recursive_call_p): New function.
6309 * ipa-utils.h (recursive_call_p): Dclare.
6310 * symtab.c (symtab_nonoverwritable_alias): Fix formatting.
6311 (symtab_semantically_equivalent_p): New function.
6312 * Makefile.in (tree-tailcall.o): Update dependencies.
6313
9a6af450
EB		 63142013-09-06 Eric Botcazou <ebotcazou@adacore.com>
6315
6316 * ipa-split.c (split_function): Set DECL_NO_INLINE_WARNING_P on the
6317 non-inlinable part.
6318
35f32ad4
RB		 63192013-09-06 Richard Biener <rguenther@suse.de>
6320
6321 * lto-streamer.h (lto_global_var_decls): Remove.
6322 * Makefile.in (OBJS): Remove lto-symtab.o.
6323 (lto-symtab.o): Remove.
6324 (GTFILES): Remove lto-symtab.c
6325 * lto-symtab.c: Move to lto/
6326
d12a76f3
AK		 63272013-09-06 Andreas Krebbel <Andreas.Krebbel@de.ibm.com>
6328
6329 * config/s390/s390.md (UNSPEC_FPINT_FLOOR, UNSPEC_FPINT_BTRUNC)
6330 (UNSPEC_FPINT_ROUND, UNSPEC_FPINT_CEIL, UNSPEC_FPINT_NEARBYINT)
6331 (UNSPEC_FPINT_RINT): New constant definitions.
6332 (FPINT, fpint_name, fpint_roundingmode): New integer iterator
6333 definition with 2 attributes.
6334 ("<FPINT:fpint_name><BFP:mode>2", "rint<BFP:mode>2")
6335 ("<FPINT:fpint_name><DFP:mode>2", "rint<DFP:mode>2"): New pattern
6336 definitions.
6337
63382013-09-06 Andreas Krebbel <Andreas.Krebbel@de.ibm.com>
6339
f43245d1 6340 * config/s390/s390.md: Add "bcr_flush" value to mnemonic attribute.
d12a76f3
AK		 6341 ("mem_thread_fence_1"): Use bcr 14,0 for z196 and later.
6342 Set the mnemonic attribute to "bcr_flush". Set the "z196prop"
6343 attribute to "z196_alone".
6344 * config/s390/2827.md: Add "bcr_flush" to "ooo_groupalone" and
6345 "zEC12_simple".
6346
c8e9d8c3
RB		 63472013-09-06 Richard Biener <rguenther@suse.de>
6348
6349 * basic-block.h (class control_dependences): New.
6350 * tree-ssa-dce.c (control_dependence_map): Remove.
6351 (cd): New global.
6352 (EXECUTE_IF_CONTROL_DEPENDENT): Remove.
6353 (set_control_dependence_map_bit, clear_control_dependence_bitmap,
6354 find_pdom, find_control_dependence, find_all_control_dependences):
6355 Move to cfganal.c.
f43245d1
UB		 6356 (mark_control_dependent_edges_necessary,
6357 find_obviously_necessary_stmts, propagate_necessity, tree_dce_init,
6358 tree_dce_done, perform_tree_ssa_dce): Adjust.
c8e9d8c3
RB		 6359 * cfganal.c (set_control_dependence_map_bit,
6360 clear_control_dependence_bitmap, find_pdom, find_control_dependence,
6361 find_all_control_dependences): Move from tree-ssa-dce.c and
6362 implement as methods of control_dependences class.
6363 (control_dependences::control_dependences): New.
6364 (control_dependences::~control_dependences): Likewise.
6365 (control_dependences::get_edges_dependent_on): Likewise.
6366 (control_dependences::get_edge): Likewise.
6367
7d2ba471
JH		 63682013-09-04 Jan Hubicka <jh@suse.cz>
6369
6370 * tree.c (types_same_for_odr): Drop overactive check.
6371 * ipa-devirt.c (hash_type_name): Likewise.
6372
e18412fc
JH		 63732013-09-04 Jan Hubicka <jh@suse.cz>
6374
6375 * cgraphunit.c (walk_polymorphic_call_targets): Break out from ...
6376 (analyze_functions): ... here.
6377
d352b245
JH		 63782013-09-04 Jan Hubicka <jh@suse.cz>
6379
6380 PR middle-end/58201
6381 * cgraphunit.c (analyze_functions): Clear AUX fields
6382 after processing; initialize assembler name has.
6383
770da076
JL		 63842013-09-05 Jeff Law <law@redhat.com>
6385
6386 * tree-ssa-threadedge.c (thread_around_empty_blocks): Renamed
6387 from thread_around_empty_block. Record threading path into PATH.
6388 Recurse if threading through the initial block is successful.
6389 (thread_across_edge): Corresponding changes to slightly simplify.
6390
fe6f68e2
JG		 63912013-09-05 James Greenhalgh <james.greenhalgh@arm.com>
6392
6393 * config/aarch64/aarch64.md
6394 (type): Remove frecpe, frecps, frecpx.
6395 (aarch64_frecp<FRECP:frecp_suffix><mode>): Move to aarch64-simd.md,
6396 fix to be a TARGET_SIMD instruction.
6397 (aarch64_frecps): Remove.
6398 * config/aarch64/aarch64-simd.md
6399 (aarch64_frecp<FRECP:frecp_suffix><mode>): New, moved from aarch64.md
6400 (aarch64_frecps<mode>): Handle all float/vector of float modes.
6401
6e4150e1
JG		 64022013-09-05 James Greenhalgh <james.greenhalgh@arm.com>
6403 Sofiane Naci <sofiane.naci@arm.com>
6404
f43245d1 6405 * config/arm/types.md (define_attr "type"): Expand "arlo_imm"
6e4150e1 6406 into "adr", "alu_imm", "alus_imm", "logic_imm", "logics_imm".
f43245d1
UB		 6407 Expand "arlo_reg" into "adc_reg", "adc_imm", "adcs_reg", "adcs_imm",
6408 "alu_ext", "alu_reg", "alus_ext", "alus_reg", "bfm", "csel",
6409 "logic_reg", "logics_reg", "rev". Expand "arlo_shift" into
6410 "alu_shift_imm", "alus_shift_imm", "logic_shift_imm",
6411 "logics_shift_imm". Expand "arlo_shift_reg" into "alu_shift_reg",
6412 "alus_shift_reg", "logic_shift_reg", "logics_shift_reg". Expand "clz"
6413 into "clz, "rbit". Rename "shift" to "shift_imm".
6e4150e1 6414 * config/arm/arm.md (define_attr "core_cycles"): Update for attribute
f43245d1 6415 changes. Update for attribute changes all occurrences of arlo_* and
6e4150e1
JG		 6416 shift* types.
6417 * config/arm/arm-fixed.md: Update for attribute changes
6418 all occurrences of arlo_* types.
6419 * config/arm/thumb2.md: Update for attribute changes all occurrences
6420 of arlo_* types.
6421 * config/arm/arm.c (xscale_sched_adjust_cost): (rtx insn, rtx
6422 (cortexa7_older_only): Likewise.
6423 (cortexa7_younger): Likewise.
6424 * config/arm/arm1020e.md (1020alu_op): Update for attribute changes.
6425 (1020alu_shift_op): Likewise.
6426 (1020alu_shift_reg_op): Likewise.
6427 * config/arm/arm1026ejs.md (alu_op): Update for attribute changes.
6428 (alu_shift_op): Likewise.
6429 (alu_shift_reg_op): Likewise.
f43245d1 6430 * config/arm/arm1136jfs.md (11_alu_op): Update for attribute changes.
6e4150e1
JG		 6431 (11_alu_shift_op): Likewise.
6432 (11_alu_shift_reg_op): Likewise.
6433 * config/arm/arm926ejs.md (9_alu_op): Update for attribute changes.
6434 (9_alu_shift_reg_op): Likewise.
6435 * config/arm/cortex-a15.md (cortex_a15_alu): Update for
6436 attribute changes.
6437 (cortex_a15_alu_shift): Likewise.
6438 (cortex_a15_alu_shift_reg): Likewise.
6439 * config/arm/cortex-a5.md (cortex_a5_alu): Update for
6440 attribute changes.
6441 (cortex_a5_alu_shift): Likewise.
f43245d1
UB		 6442 * config/arm/cortex-a53.md (cortex_a53_alu): Update for
6443 attribute changes.
6e4150e1 6444 (cortex_a53_alu_shift): Likewise.
f43245d1
UB		 6445 * config/arm/cortex-a7.md (cortex_a7_alu_imm): Update for
6446 attribute changes.
6e4150e1
JG		 6447 (cortex_a7_alu_reg): Likewise.
6448 (cortex_a7_alu_shift): Likewise.
f43245d1
UB		 6449 * config/arm/cortex-a8.md (cortex_a8_alu): Update for
6450 attribute changes.
6e4150e1
JG		 6451 (cortex_a8_alu_shift): Likewise.
6452 (cortex_a8_alu_shift_reg): Likewise.
f43245d1 6453 * config/arm/cortex-a9.md (cortex_a9_dp): Update for attribute changes.
6e4150e1 6454 (cortex_a9_dp_shift): Likewise.
f43245d1
UB		 6455 * config/arm/cortex-m4.md (cortex_m4_alu): Update for
6456 attribute changes.
6e4150e1
JG		 6457 * config/arm/cortex-r4.md
6458 (cortex_r4_alu): Update for attribute changes.
6459 (cortex_r4_mov): Likewise.
6460 (cortex_r4_alu_shift_reg): Likewise.
f43245d1 6461 * config/arm/fa526.md (526_alu_op): Update for attribute changes.
6e4150e1 6462 (526_alu_shift_op): Likewise.
f43245d1
UB		 6463 * config/arm/fa606te.md (606te_alu_op): Update for attribute changes.
6464 * config/arm/fa626te.md (626te_alu_op): Update for attribute changes.
6e4150e1 6465 (626te_alu_shift_op): Likewise.
f43245d1 6466 * config/arm/fa726te.md (726te_alu_op): Update for attribute changes.
6e4150e1
JG		 6467 (726te_alu_shift_op): Likewise.
6468 (726te_alu_shift_reg_op): Likewise.
6469 * config/arm/fmp626.md (mp626_alu_op): Update for attribute changes.
6470 (mp626_alu_shift_op): Likewise.
6471 * config/arm/marvell-pj4.md (pj4_alu): Update for attribute changes.
6472 (pj4_alu_conds): Likewise.
6473 (pj4_shift): Likewise.
6474 (pj4_shift_conds): Likewise.
6475 (pj4_alu_shift): Likewise.
6476 (pj4_alu_shift_conds): Likewise.
6477 * config/aarch64/aarch64.md: Update for attribute change
6478 all occurrences of arlo_* and shift* types.
6479
e6523306
MS		 64802013-09-05 Mike Stump <mikestump@comcast.net>
6481
6482 * tree.h: Move documentation for tree_function_decl to tree-core.h
6483 with the declaration.
6484
80a18161
PB		 64852013-09-05 Peter Bergner <bergner@vnet.ibm.com>
6486
6487 PR target/58139
6488 * reginfo.c (choose_hard_reg_mode): Scan through all mode classes
6489 looking for widest mode.
6490
6994430a
EB		 64912013-09-05 Eric Botcazou <ebotcazou@adacore.com>
6492
6493 * config.gcc (*-*-vxworks*): Do not override an existing extra_objs.
6494
e67f39f7
RB		 64952013-09-05 Richard Biener <rguenther@suse.de>
6496
6497 PR tree-optimization/58137
6498 * tree-vect-stmts.c (get_vectype_for_scalar_type_and_size):
6499 Do not create vectors of pointers.
6500 * tree-vect-loop.c (get_initial_def_for_induction): Use proper
6501 types for the components of the vector initializer.
6502 * tree-cfg.c (verify_gimple_assign_binary): Remove special-casing
6503 allowing pointer vectors with PLUS_EXPR/MINUS_EXPR.
6504
568cda29
MJ		 65052013-09-05 Martin Jambor <mjambor@suse.cz>
6506
6507 * ipa-prop.c (remove_described_reference): Accept missing references,
6508 return false if that hppens, otherwise return true.
6509 (cgraph_node_for_jfunc): New function.
6510 (try_decrement_rdesc_refcount): Likewise.
6511 (try_make_edge_direct_simple_call): Use them.
6512 (ipa_edge_removal_hook): Remove references from rdescs.
6513 (ipa_edge_duplication_hook): Clone rdescs and their references
6514 when the new edge has the same caller as the old one.
6515 * cgraph.c (cgraph_resolve_speculation): Remove speculative
6516 reference before removing any edges.
6517
5a200acb
RE		 65182013-09-05 Richard Earnshaw <rearnsha@arm.com>
6519
6520 * arm.c (thumb2_emit_strd_push): Rewrite to use pre-decrement on
6521 initial store.
6522 * thumb2.md (thumb2_storewb_parisi): New pattern.
6523
5922847b
YZ		 65242013-09-05 Yufeng Zhang <yufeng.zhang@arm.com>
6525
6526 * config/aarch64/aarch64-option-extensions.def: Add
6527 AARCH64_OPT_EXTENSION of 'crc'.
6528 * config/aarch64/aarch64.h (AARCH64_FL_CRC): New define.
6529 (AARCH64_ISA_CRC): Ditto.
6530 * doc/invoke.texi (-march and -mcpu feature modifiers): Add
6531 description of the CRC extension.
6532
996746aa
AI		 65332013-09-05 Alexander Ivchenko <alexander.ivchenko@intel.com>
6534
6535 * config/rs6000/linux64.h: Define OPTION_BIONIC and OPTION_UCLIBC.
6536 * config/rs6000/linux.h: Ditto.
6537 * alpha/linux.h: Ditto.
6538 * config/bfin/uclinux.h: Define TARGET_LIBC_HAS_FUNCTION as
6539 no_c99_libc_has_function.
6540 * config/c6x/uclinux-elf.h: Ditto.
6541 * config/lm32/uclinux-elf.h: Ditto.
6542 * config/m68k/uclinux.h: Ditto.
6543 * config/moxie/uclinux.h: Ditto.
6544 * config.gcc (bfin*-linux-uclibc*): Add t-linux-android to tmake_file.
6545 (crisv32-*-linux*, cris-*-linux*): Ditto.
6546 * config/bfin/bfin.c: Include "tm_p.h".
6547
c370d99f
RB		 65482013-09-05 Richard Biener <rguenther@suse.de>
6549
6550 * tree-vect-loop.c (vect_analyze_loop_operations): Properly
6551 check for a definition without a basic-block.
6552
003bb7f3
JG		 65532013-09-05 James Greenhalgh <james.greenhalgh@arm.com>
6554 Sofiane Naci <sofiane.naci@arm.com>
6555
6556 * config/aarch64/aarch64.md
6557 (*movti_aarch64): Rename r_2_f and f_2_r.
6558 (*movsf_aarch64): Likewise.
6559 (*movdf_aarch64): Likewise.
6560 (*movtf_aarch64): Likewise.
6561 (aarch64_movdi_<mode>low): Likewise.
6562 (aarch64_movdi_<mode>high): Likewise.
6563 (aarch64_mov<mode>high_di): Likewise.
6564 (aarch64_mov<mode>low_di): Likewise.
6565 (aarch64_movtilow_tilow): Likewise.
6566 * config/arm/arm.md (attribute "neon_type"): Delete. Move attribute
6567 values to config/arm/types.md
6568 (attribute "conds"): Update for attribute change.
6569 (anddi3_insn): Likewise.
6570 (iordi3_insn): Likewise.
6571 (xordi3_insn): Likewise.
6572 (one_cmpldi2): Likewise.
6573 * config/arm/types.md (type): Add Neon types.
6574 * config/arm/neon.md (neon_mov<mode>): Remove "neon_type" attribute,
6575 use "type" attribute.
6576 (movmisalign<mode>_neon_store): Likewise.
6577 (movmisalign<mode>_neon_load): Likewise.
6578 (vec_set<mode>_internal): Likewise.
6579 (vec_setv2di_internal): Likewise.
6580 (vec_extract<mode>): Likewise.
6581 (vec_extractv2di): Likewise.
6582 (add<mode>3_neon): Likewise.
6583 (adddi3_neon): Likewise.
6584 (sub<mode>3_neon): Likewise.
6585 (subdi3_neon): Likewise.
6586 (mul<mode>3_neon): Likewise.
6587 (mul<mode>3add<mode>_neon): Likewise.
6588 (mul<mode>3neg<mode>add<mode>_neon): Likewise.
6589 (fma<VCVTF:mode>4)): Likewise.
6590 (fma<VCVTF:mode>4_intrinsic): Likewise.
6591 (fmsub<VCVTF:mode>4)): Likewise.
6592 (fmsub<VCVTF:mode>4_intrinsic): Likewise.
6593 (neon_vrint<NEON_VRINT:nvrint_variant><VCVTF:mode>): Likewise.
6594 (ior<mode>3): Likewise.
6595 (and<mode>3): Likewise.
6596 (anddi3_neon): Likewise.
6597 (orn<mode>3_neon): Likewise.
6598 (orndi3_neon): Likewise.
6599 (bic<mode>3_neon): Likewise.
6600 (bicdi3_neon): Likewise.
6601 (xor<mode>3): Likewise.
6602 (one_cmpl<mode>2): Likewise.
6603 (abs<mode>2): Likewise.
6604 (neg<mode>2): Likewise.
6605 (umin<mode>3_neon): Likewise.
6606 (umax<mode>3_neon): Likewise.
6607 (smin<mode>3_neon): Likewise.
6608 (smax<mode>3_neon): Likewise.
6609 (vashl<mode>3): Likewise.
6610 (vashr<mode>3_imm): Likewise.
6611 (vlshr<mode>3_imm): Likewise.
6612 (ashl<mode>3_signed): Likewise.
6613 (ashl<mode>3_unsigned): Likewise.
6614 (neon_load_count): Likewise.
6615 (ashldi3_neon_noclobber): Likewise.
6616 (signed_shift_di3_neon): Likewise.
6617 (unsigned_shift_di3_neon): Likewise.
6618 (ashrdi3_neon_imm_noclobber): Likewise.
6619 (lshrdi3_neon_imm_noclobber): Likewise.
6620 (widen_ssum<mode>3): Likewise.
6621 (widen_usum<mode>3): Likewise.
6622 (quad_halves_<code>v4si): Likewise.
6623 (quad_halves_<code>v4sf): Likewise.
6624 (quad_halves_<code>v8hi): Likewise.
6625 (quad_halves_<code>v16qi): Likewise.
6626 (reduc_splus_v2di): Likewise.
6627 (neon_vpadd_internal<mode>): Likewise.
6628 (neon_vpsmin<mode>): Likewise.
6629 (neon_vpsmax<mode>): Likewise.
6630 (neon_vpumin<mode>): Likewise.
6631 (neon_vpumax<mode>): Likewise.
6632 (ss_add<mode>_neon): Likewise.
6633 (us_add<mode>_neon): Likewise.
6634 (ss_sub<mode>_neon): Likewise.
6635 (us_sub<mode>_neon): Likewise.
6636 (neon_vadd<mode>_unspec): Likewise.
6637 (neon_vaddl<mode>): Likewise.
6638 (neon_vaddw<mode>): Likewise.
6639 (neon_vhadd<mode>): Likewise.
6640 (neon_vqadd<mode>): Likewise.
6641 (neon_vaddhn<mode>): Likewise.
6642 (neon_vmul<mode>): Likewise.
6643 (neon_vmla<mode>): Likewise.
6644 (neon_vmlal<mode>): Likewise.
6645 (neon_vmls<mode>): Likewise.
6646 (neon_vmlsl<mode>): Likewise.
6647 (neon_vqdmulh<mode>): Likewise.
6648 (neon_vqdmlal<mode>): Likewise.
6649 (neon_vqdmlsl<mode>): Likewise.
6650 (neon_vmull<mode>): Likewise.
6651 (neon_vqdmull<mode>): Likewise.
6652 (neon_vsub<mode>_unspec): Likewise.
6653 (neon_vsubl<mode>): Likewise.
6654 (neon_vsubw<mode>): Likewise.
6655 (neon_vqsub<mode>): Likewise.
6656 (neon_vhsub<mode>): Likewise.
6657 (neon_vsubhn<mode>): Likewise.
6658 (neon_vceq<mode>): Likewise.
6659 (neon_vcge<mode>): Likewise.
6660 (neon_vcgeu<mode>): Likewise.
6661 (neon_vcgt<mode>): Likewise.
6662 (neon_vcgtu<mode>): Likewise.
6663 (neon_vcle<mode>): Likewise.
6664 (neon_vclt<mode>): Likewise.
6665 (neon_vcage<mode>): Likewise.
6666 (neon_vcagt<mode>): Likewise.
6667 (neon_vtst<mode>): Likewise.
6668 (neon_vabd<mode>): Likewise.
6669 (neon_vabdl<mode>): Likewise.
6670 (neon_vaba<mode>): Likewise.
6671 (neon_vabal<mode>): Likewise.
6672 (neon_vmax<mode>): Likewise.
6673 (neon_vmin<mode>): Likewise.
6674 (neon_vpaddl<mode>): Likewise.
6675 (neon_vpadal<mode>): Likewise.
6676 (neon_vpmax<mode>): Likewise.
6677 (neon_vpmin<mode>): Likewise.
6678 (neon_vrecps<mode>): Likewise.
6679 (neon_vrsqrts<mode>): Likewise.
6680 (neon_vqabs<mode>): Likewise.
6681 (neon_vqneg<mode>): Likewise.
6682 (neon_vcls<mode>): Likewise.
6683 (clz<mode>2): Likewise.
6684 (popcount<mode>2): Likewise.
6685 (neon_vrecpe): Likewise.
6686 (neon_vrsqrte): Likewise.
6687 (neon_vget_lane<mode>_sext_internal): Likewise.
6688 (neon_vget_lane<mode>_zext_internal): Likewise.
6689 (neon_vdup_n<mode>): Likewise.
6690 (neon_vdup_nv2di): Likewise.
6691 (neon_vdpu_lane<mode>_internal): Likewise.
6692 (neon_vswp<mode>): Likewise.
6693 (float<mode><V_cvtto>2): Likewise.
6694 (floatuns<mode><V_cvtto>2): Likewise.
6695 (fix_trunc<mode><V_cvtto>)2): Likewise
6696 (fixuns_trunc<mode><V_cvtto)2): Likewise.
6697 (neon_vcvt<mode>): Likewise.
6698 (neon_vcvtv4sfv4hf): Likewise.
6699 (neon_vcvtv4hfv4sf): Likewise.
6700 (neon_vcvt_n<mode>): Likewise.
6701 (neon_vmovn<mode>): Likewise.
6702 (neon_vqmovn<mode>): Likewise.
6703 (neon_vqmovun<mode>): Likewise.
6704 (neon_vmovl<mode>): Likewise.
6705 (neon_vmul_lane<mode>): Likewise.
6706 (neon_vmull_lane<mode>): Likewise.
6707 (neon_vqdmull_lane<mode>): Likewise.
6708 (neon_vqdmulh_lane<mode>): Likewise.
6709 (neon_vmla_lane<mode>): Likewise.
6710 (neon_vmlal_lane<mode>): Likewise.
6711 (neon_vqdmlal_lane<mode>): Likewise.
6712 (neon_vmls_lane<mode>): Likewise.
6713 (neon_vmlsl_lane<mode>): Likewise.
6714 (neon_vqdmlsl_lane<mode>): Likewise.
6715 (neon_vext<mode>): Likewise.
6716 (neon_vrev64<mode>): Likewise.
6717 (neon_vrev32<mode>): Likewise.
6718 (neon_vrev16<mode>): Likewise.
6719 (neon_vbsl<mode>_internal): Likewise.
6720 (neon_vshl<mode>): Likewise.
6721 (neon_vqshl<mode>): Likewise.
6722 (neon_vshr_n<mode>): Likewise.
6723 (neon_vshrn_n<mode>): Likewise.
6724 (neon_vqshrn_n<mode>): Likewise.
6725 (neon_vqshrun_n<mode>): Likewise.
6726 (neon_vshl_n<mode>): Likewise.
6727 (neon_vqshl_n<mode>): Likewise.
6728 (neon_vqshlu_n<mode>): Likewise.
6729 (neon_vshll_n<mode>): Likewise.
6730 (neon_vsra_n<mode>): Likewise.
6731 (neon_vsri_n<mode>): Likewise.
6732 (neon_vsli_n<mode>): Likewise.
6733 (neon_vtbl1v8qi): Likewise.
6734 (neon_vtbl2v8qi): Likewise.
6735 (neon_vtbl3v8qi): Likewise.
6736 (neon_vtbl4v8qi): Likewise.
6737 (neon_vtbx1v8qi): Likewise.
6738 (neon_vtbx2v8qi): Likewise.
6739 (neon_vtbx3v8qi): Likewise.
6740 (neon_vtbx4v8qi): Likewise.
6741 (neon_vtrn<mode>_internal): Likewise.
6742 (neon_vzip<mode>_internal): Likewise.
6743 (neon_vuzp<mode>_internal): Likewise.
6744 (neon_vld1<mode>): Likewise.
6745 (neon_vld1_lane<mode>): Likewise.
6746 (neon_vld1_dup<mode>): Likewise.
6747 (neon_vld1_dupv2di): Likewise.
6748 (neon_vst1<mode>): Likewise.
6749 (neon_vst1_lane<mode>): Likewise.
6750 (neon_vld2<mode>): Likewise.
6751 (neon_vld2_lane<mode>): Likewise.
6752 (neon_vld2_dup<mode>): Likewise.
6753 (neon_vst2<mode>): Likewise.
6754 (neon_vst2_lane<mode>): Likewise.
6755 (neon_vld3<mode>): Likewise.
6756 (neon_vld3qa<mode>): Likewise.
6757 (neon_vld3qb<mode>): Likewise.
6758 (neon_vld3_lane<mode>): Likewise.
6759 (neon_vld3_dup<mode>): Likewise.
6760 (neon_vst3<mode>): Likewise.
6761 (neon_vst3qa<mode>): Likewise.
6762 (neon_vst3qb<mode>): Likewise.
6763 (neon_vst3_lane<mode>): Likewise.
6764 (neon_vld4<mode>): Likewise.
6765 (neon_vld4qa<mode>): Likewise.
6766 (neon_vld4qb<mode>): Likewise.
6767 (neon_vld4_lane<mode>): Likewise.
6768 (neon_vld4_dup<mode>): Likewise.
6769 (neon_vst4<mode>): Likewise.
6770 (neon_vst4qa<mode>): Likewise.
6771 (neon_vst4qb<mode>): Likewise.
6772 (neon_vst4_lane<mode>): Likewise.
6773 (neon_vec_unpack<US>_lo_<mode>): Likewise.
6774 (neon_vec_unpack<US>_hi_<mode>): Likewise.
6775 (neon_vec_<US>mult_lo_<mode>): Likewise.
6776 (neon_vec_<US>mult_hi_<mode>): Likewise.
6777 (neon_vec_<US>shiftl_<mode>): Likewise.
6778 (neon_unpack<US>_<mode>): Likewise.
6779 (neon_vec_<US>mult_<mode>): Likewise.
6780 (vec_pack_trunc_<mode>): Likewise.
6781 (neon_vec_pack_trunk_<mode>): Likewise.
6782 (neon_vabd<mode>_2): Likewise.
6783 (neon_vabd<mode>_3): Likewise.
6784 * config/arm/vfp.md (arm_movsi_vfp): Update for attribute changes.
6785 (thumb2_movsi_vfp): Likewise.
6786 (movdi_vfp): Likewise.
6787 (movdi_vfp_cortexa8): Likewise.
6788 (movhf_vfp_neon): Likewise.
6789 (movhf_vfp): Likewiwse.
6790 (movsf_vfp): Likewiwse.
6791 (thumb2_movsf_vfp): Likewiwse.
6792 (movdf_vfp): Likewise.
6793 (thumb2_movdf_vfp): Likewise.
6794 (movsfcc_vfp): Likewise.
6795 (thumb2_movsfcc_vfp): Likewise.
6796 (movdfcc_vfp): Likewise.
6797 (thumb2_movdfcc_vfp): Likewise.
6798 * config/arm/arm.c (cortexa7_older_only): Update for attribute change.
6799 * config/arm/arm1020e.md (v10_c2v): Update for attribute change.
6800 (v10_v2c): Likewise.
6801 * config/arm/cortex-a15-neon.md (cortex_a15_neon_int_1): Update for
6802 attribute change.
6803 (cortex_a15_neon_int_2): Likewise.
6804 (cortex_a15_neon_int_3): Likewise.
6805 (cortex_a15_neon_int_4): Likewise.
6806 (cortex_a15_neon_int_5): Likewise.
6807 (cortex_a15_neon_vqneg_vqabs): Likewise.
6808 (cortex_a15_neon_vmov): Likewise.
6809 (cortex_a15_neon_vaba): Likewise.
6810 (cortex_a15_neon_vaba_qqq): Likewise.
6811 (cortex_a15_neon_mul_ddd_8_16_qdd_16_8_long_32_16_long): Likewise.
6812 (cortex_a15_neon_mul_qqq_8_16_32_ddd_32): Likewise.
6813 (cortex_a15_neon_mul_qdd_64_32_long_qqd_16_ddd_32_\
6814 scalar_64_32_long_scalar): Likewise.
6815 (cortex_a15_neon_mla_ddd_8_16_qdd_16_8_long_32_16_long): Likewise.
6816 (cortex_a15_neon_mla_qqq_8_16): Likewise.
6817 (cortex_a15_neon_mla_ddd_32_qqd_16_ddd_32_scalar_qdd_64_32_\
6818 lotype_qdd_64_32_long): Likewise.
6819 (cortex_a15_neon_mla_qqq_32_qqd_32_scalar): Likewise.
6820 (cortex_a15_neon_mul_ddd_16_scalar_32_16_long_scalar): Likewise.
6821 (cortex_a15_neon_mul_qqd_32_scalar): Likewise.
6822 (cortex_a15_neon_mla_ddd_16_scalar_qdd_32_16_long_scalar): Likewise.
6823 (cortex_a15_neon_shift_1): Likewise.
6824 (cortex_a15_neon_shift_2): Likewise.
6825 (cortex_a15_neon_shift_3): Likewise.
6826 (cortex_a15_neon_vshl_ddd): Likewise.
6827 (cortex_a15_neon_vqshl_vrshl_vqrshl_qqq): Likewise.
6828 (cortex_a15_neon_vsra_vrsra): Likewise.
6829 (cortex_a15_neon_fp_vadd_ddd_vabs_dd): Likewise.
6830 (cortex_a15_neon_fp_vadd_qqq_vabs_qq): Likewise.
6831 (cortex_a15_neon_fp_vmul_ddd): Likewise.
6832 (cortex_a15_neon_fp_vmul_qqd): Likewise.
6833 (cortex_a15_neon_fp_vmla_ddd): Likewise.
6834 (cortex_a15_neon_fp_vmla_qqq): Likewise.
6835 (cortex_a15_neon_fp_vmla_ddd_scalar): Likewise.
6836 (cortex_a15_neon_fp_vmla_qqq_scalar): Likewise.
6837 (cortex_a15_neon_fp_vrecps_vrsqrts_ddd): Likewise.
6838 (cortex_a15_neon_fp_vrecps_vrsqrts_qqq): Likewise.
6839 (cortex_a15_neon_bp_simple): Likewise.
6840 (cortex_a15_neon_bp_2cycle): Likewise.
6841 (cortex_a15_neon_bp_3cycle): Likewise.
6842 (cortex_a15_neon_vld1_1_2_regs): Likewise.
6843 (cortex_a15_neon_vld1_3_4_regs): Likewise.
6844 (cortex_a15_neon_vld2_2_regs_vld1_vld2_all_lanes): Likewise.
6845 (cortex_a15_neon_vld2_4_regs): Likewise.
6846 (cortex_a15_neon_vld3_vld4): Likewise.
6847 (cortex_a15_neon_vst1_1_2_regs_vst2_2_regs): Likewise.
6848 (cortex_a15_neon_vst1_3_4_regs): Likewise.
6849 (cortex_a15_neon_vst2_4_regs_vst3_vst4): Likewise.
6850 (cortex_a15_neon_vst3_vst4): Likewise.
6851 (cortex_a15_neon_vld1_vld2_lane): Likewise.
6852 (cortex_a15_neon_vld3_vld4_lane" 10
6853 (cortex_a15_neon_vst1_vst2_lane): Likewise.
6854 (cortex_a15_neon_vst3_vst4_lane): Likewise.
6855 (cortex_a15_neon_vld3_vld4_all_lanes): Likewise.
6856 (cortex_a15_neon_ldm_2): Likewise.0
6857 (cortex_a15_neon_stm_2): Likewise.
6858 (cortex_a15_neon_mcr): Likewise.
6859 (cortex_a15_neon_mcr_2_mcrr): Likewise.
6860 (cortex_a15_neon_mrc): Likewise.
6861 (cortex_a15_neon_mrrc): Likewise.
6862 * config/arm/cortex-a15.md (cortex_a15_alu): Update for attribute
6863 change.
6864 (cortex_a15_alu_shift): Likewise.
6865 (cortex_a15_alu_shift_reg): Likewise.
6866 (cortex_a15_mult32): Likewise.
6867 (cortex_a15_mult64): Likewise.
6868 (cortex_a15_block): Likewise.
6869 (cortex_a15_branch): Likewise.
6870 (cortex_a15_load1): Likewise.
6871 (cortex_a15_load3): Likewise.
6872 (cortex_a15_store1): Likewise.
6873 (cortex_a15_store3): Likewise.
6874 (cortex_a15_call): Likewise.
f43245d1 6875 * config/arm/cortex-a5.md (cortex_a5_r2f): Update for attribute change.
003bb7f3
JG		 6876 (cortex_a5_f2r): Likewise.
6877 * config/arm/cortex-a53.md (cortex_a53_r2f): Update for attribute
6878 change.
6879 (cortex_a53_f2r): Likewise.
6880 * config/arm/cortex-a7.md
6881 (cortex_a7_branch): Update for attribute change.
6882 (cortex_a7_call): Likewise.
6883 (cortex_a7_alu_imm): Likewise.
6884 (cortex_a7_alu_reg): Likewise.
6885 (cortex_a7_alu_shift): Likewise.
6886 (cortex_a7_mul): Likewise.
6887 (cortex_a7_load1): Likewise.
6888 (cortex_a7_store1): Likewise.
6889 (cortex_a7_load2): Likewise.
6890 (cortex_a7_store2): Likewise.
6891 (cortex_a7_load3): Likewise.
6892 (cortex_a7_store3): Likewise.
6893 (cortex_a7_load4): Likewise.
6894 (cortex_a7_store4): Likewise.
6895 (cortex_a7_fpalu): Likewise.
6896 (cortex_a7_fconst): Likewise.
6897 (cortex_a7_fpmuls): Likewise.
6898 (cortex_a7_neon_mul): Likewise.
6899 (cortex_a7_fpmacs): Likewise.
6900 (cortex_a7_neon_mla: Likewise.
6901 (cortex_a7_fpmuld: Likewise.
6902 (cortex_a7_fpmacd: Likewise.
6903 (cortex_a7_fpfmad: Likewise.
6904 (cortex_a7_fdivs: Likewise.
6905 (cortex_a7_fdivd: Likewise.
6906 (cortex_a7_r2f: Likewise.
6907 (cortex_a7_f2r: Likewise.
6908 (cortex_a7_f_flags: Likewise.
6909 (cortex_a7_f_loads: Likewise.
6910 (cortex_a7_f_loadd: Likewise.
6911 (cortex_a7_f_stores: Likewise.
6912 (cortex_a7_f_stored: Likewise.
6913 (cortex_a7_neon): Likewise.
6914 * config/arm/cortex-a8-neon.md
6915 (cortex_a8_neon_mrc): Update for attribute change.
6916 (cortex_a8_neon_mrrc): Likewise.
6917 (cortex_a8_neon_int_1): Likewise.
6918 (cortex_a8_neon_int_2): Likewise.
6919 (cortex_a8_neon_int_3): Likewise.
6920 (cortex_a8_neon_int_4): Likewise.
6921 (cortex_a8_neon_int_5): Likewise.
6922 (cortex_a8_neon_vqneg_vqabs): Likewise.
6923 (cortex_a8_neon_vmov): Likewise.
6924 (cortex_a8_neon_vaba): Likewise.
6925 (cortex_a8_neon_vaba_qqq): Likewise.
6926 (cortex_a8_neon_vsma): Likewise.
6927 (cortex_a8_neon_mul_ddd_8_16_qdd_16_8_long_32_16_long): Likewise.
6928 (cortex_a8_neon_mul_qqq_8_16_32_ddd_32): Likewise.
f43245d1
UB		 6929 (cortex_a8_neon_mul_qdd_64_32_long_qqd_16_ddd_32_scalar_64_32_\
6930 long_scalar): Likewise.
003bb7f3
JG		 6931 (cortex_a8_neon_mla_ddd_8_16_qdd_16_8_long_32_16_long): Likewise.
6932 (cortex_a8_neon_mla_qqq_8_16): Likewise.
6933 (cortex_a8_neon_mla_ddd_32_qqd_16_ddd_32_scalar_qdd_64_32_\
6934 long_scalar_qdd_64_32_long): Likewise.
6935 (cortex_a8_neon_mla_qqq_32_qqd_32_scalar): Likewise.
6936 (cortex_a8_neon_mul_ddd_16_scalar_32_16_long_scalar): Likewise.
6937 (cortex_a8_neon_mul_qqd_32_scalar): Likewise.
6938 (cortex_a8_neon_mla_ddd_16_scalar_qdd_32_16_long_scalar): Likewise.
6939 (cortex_a8_neon_shift_1): Likewise.
6940 (cortex_a8_neon_shift_2): Likewise.
6941 (cortex_a8_neon_shift_3): Likewise.
6942 (cortex_a8_neon_vshl_ddd): Likewise.
6943 (cortex_a8_neon_vqshl_vrshl_vqrshl_qqq): Likewise.
6944 (cortex_a8_neon_vsra_vrsra): Likewise.
6945 (cortex_a8_neon_fp_vadd_ddd_vabs_dd): Likewise.
6946 (cortex_a8_neon_fp_vadd_qqq_vabs_qq): Likewise.
6947 (cortex_a8_neon_fp_vsum): Likewise.
6948 (cortex_a8_neon_fp_vmul_ddd): Likewise.
6949 (cortex_a8_neon_fp_vmul_qqd): Likewise.
6950 (cortex_a8_neon_fp_vmla_ddd): Likewise.
6951 (cortex_a8_neon_fp_vmla_qqq): Likewise.
6952 (cortex_a8_neon_fp_vmla_ddd_scalar): Likewise.
6953 (cortex_a8_neon_fp_vmla_qqq_scalar): Likewise.
6954 (cortex_a8_neon_fp_vrecps_vrsqrts_ddd): Likewise.
6955 (cortex_a8_neon_fp_vrecps_vrsqrts_qqq): Likewise.
6956 (cortex_a8_neon_bp_simple): Likewise.
6957 (cortex_a8_neon_bp_2cycle): Likewise.
6958 (cortex_a8_neon_bp_3cycle): Likewise.
6959 (cortex_a8_neon_ldr): Likewise.
6960 (cortex_a8_neon_str): Likewise.
6961 (cortex_a8_neon_vld1_1_2_regs): Likewise.
6962 (cortex_a8_neon_vld1_3_4_regs): Likewise.
6963 (cortex_a8_neon_vld2_2_regs_vld1_vld2_all_lanes): Likewise.
6964 (cortex_a8_neon_vld2_4_regs): Likewise.
6965 (cortex_a8_neon_vld3_vld4): Likewise.
6966 (cortex_a8_neon_vst1_1_2_regs_vst2_2_regs): Likewise.
6967 (cortex_a8_neon_vst1_3_4_regs): Likewise.
6968 (cortex_a8_neon_vst2_4_regs_vst3_vst4): Likewise.
6969 (cortex_a8_neon_vst3_vst4): Likewise.
6970 (cortex_a8_neon_vld1_vld2_lane): Likewise.
6971 (cortex_a8_neon_vld3_vld4_lane): Likewise.
6972 (cortex_a8_neon_vst1_vst2_lane): Likewise.
6973 (cortex_a8_neon_vst3_vst4_lane): Likewise.
6974 (cortex_a8_neon_vld3_vld4_all_lanes): Likewise.
6975 (cortex_a8_neon_mcr): Likewise.
6976 (cortex_a8_neon_mcr_2_mcrr): Likewise.
f43245d1 6977 * config/arm/cortex-a8.md (cortex_a8_alu): Update for attribute change.
003bb7f3
JG		 6978 * config/arm/cortex-a9-neon.md (ca9_neon_mrc): Update for attribute
6979 change.
6980 (ca9_neon_mrrc): Likewise.
6981 (cortex_a9_neon_int_1): Likewise.
6982 (cortex_a9_neon_int_2): Likewise.
6983 (cortex_a9_neon_int_3): Likewise.
6984 (cortex_a9_neon_int_4): Likewise.
6985 (cortex_a9_neon_int_5): Likewise.
6986 (cortex_a9_neon_vqneg_vqabs): Likewise.
6987 (cortex_a9_neon_vmov): Likewise.
6988 (cortex_a9_neon_vaba): Likewise.
6989 (cortex_a9_neon_vaba_qqq): Likewise.
6990 (cortex_a9_neon_vsma): Likewise.
6991 (cortex_a9_neon_mul_ddd_8_16_qdd_16_8_long_32_16_long): Likewise.
6992 (cortex_a9_neon_mul_qqq_8_16_32_ddd_32): Likewise.
f43245d1
UB		 6993 (cortex_a9_neon_mul_qdd_64_32_long_qqd_16_ddd_32_scalar_64_32_\
6994 long_scalar): Likewise.
003bb7f3
JG		 6995 (cortex_a9_neon_mla_ddd_8_16_qdd_16_8_long_32_16_long): Likewise.
6996 (cortex_a9_neon_mla_qqq_8_16): Likewise.
6997 (cortex_a9_neon_mla_ddd_32_qqd_16_ddd_32_scalar_qdd_64_32_\
6998 long_scalar_qdd_64_32_long): Likewise.
6999 (cortex_a9_neon_mla_qqq_32_qqd_32_scalar): Likewise.
7000 (cortex_a9_neon_mul_ddd_16_scalar_32_16_long_scalar): Likewise.
7001 (cortex_a9_neon_mul_qqd_32_scalar): Likewise.
7002 (cortex_a9_neon_mla_ddd_16_scalar_qdd_32_16_long_scalar): Likewise.
7003 (cortex_a9_neon_shift_1): Likewise.
7004 (cortex_a9_neon_shift_2): Likewise.
7005 (cortex_a9_neon_shift_3): Likewise.
7006 (cortex_a9_neon_vshl_ddd): Likewise.
7007 (cortex_a9_neon_vqshl_vrshl_vqrshl_qqq): Likewise.
7008 (cortex_a9_neon_vsra_vrsra): Likewise.
7009 (cortex_a9_neon_fp_vadd_ddd_vabs_dd): Likewise.
7010 (cortex_a9_neon_fp_vadd_qqq_vabs_qq): Likewise.
7011 (cortex_a9_neon_fp_vsum): Likewise.
7012 (cortex_a9_neon_fp_vmul_ddd): Likewise.
7013 (cortex_a9_neon_fp_vmul_qqd): Likewise.
7014 (cortex_a9_neon_fp_vmla_ddd): Likewise.
7015 (cortex_a9_neon_fp_vmla_qqq): Likewise.
7016 (cortex_a9_neon_fp_vmla_ddd_scalar): Likewise.
7017 (cortex_a9_neon_fp_vmla_qqq_scalar): Likewise.
7018 (cortex_a9_neon_fp_vrecps_vrsqrts_ddd): Likewise.
7019 (cortex_a9_neon_fp_vrecps_vrsqrts_qqq): Likewise.
7020 (cortex_a9_neon_bp_simple): Likewise.
7021 (cortex_a9_neon_bp_2cycle): Likewise.
7022 (cortex_a9_neon_bp_3cycle): Likewise.
7023 (cortex_a9_neon_ldr): Likewise.
7024 (cortex_a9_neon_str): Likewise.
7025 (cortex_a9_neon_vld1_1_2_regs): Likewise.
7026 (cortex_a9_neon_vld1_3_4_regs): Likewise.
7027 (cortex_a9_neon_vld2_2_regs_vld1_vld2_all_lanes): Likewise.
7028 (cortex_a9_neon_vld2_4_regs): Likewise.
7029 (cortex_a9_neon_vld3_vld4): Likewise.
7030 (cortex_a9_neon_vst1_1_2_regs_vst2_2_regs): Likewise.
7031 (cortex_a9_neon_vst1_3_4_regs): Likewise.
7032 (cortex_a9_neon_vst2_4_regs_vst3_vst4): Likewise.
7033 (cortex_a9_neon_vst3_vst4): Likewise.
7034 (cortex_a9_neon_vld1_vld2_lane): Likewise.
7035 (cortex_a9_neon_vld3_vld4_lane): Likewise.
7036 (cortex_a9_neon_vst1_vst2_lane): Likewise.
7037 (cortex_a9_neon_vst3_vst4_lane): Likewise.
7038 (cortex_a9_neon_vld3_vld4_all_lanes): Likewise.
7039 (cortex_a9_neon_mcr): Likewise.
7040 (cortex_a9_neon_mcr_2_mcrr): Likewise.
7041 * config/arm/cortex-a9.md (cortex_a9_dp): Update for attribute change.
7042 (cortex_a9_fps): Likewise.
7043 * config/arm/cortex-m4-fpu.md (cortex_m4_vmov_2): Update for attribute
7044 change.
7045 (cortex_m4_fmuls): Likewise.
7046 * config/arm/cortex-r4f.md (cortex_r4_mcr): Update for attribute
7047 change.
7048 (cortex_r4_mrc): Likewise.
7049 * config/arm/iterators.md: Update comment referring to neon_type.
f43245d1 7050 * config/arm/iwmmxt.md (iwmmxt_arm_movdi): Update for attribute change.
003bb7f3 7051 (iwmmxt_movsi_insn): Likewise.
f43245d1
UB		 7052 * config/arm/marvell-pj4.md (pj4_vfp_to_core): Update for
7053 attribute change.
003bb7f3
JG		 7054 (pj4_core_to_vfp): Likewise.
7055 * config/arm/neon-schedgen.ml (emit_insn_reservations): Update for
7056 attribute change.
7057 * config/arm/vfp11.md (vfp_fload): Update for attribute change.
7058 (vfp_fstore): Likewise.
7059 * doc/md.texi: Change references to neon_type to refer to type.
7060
fbe0dc78
DS		 70612013-09-04 Dodji Seketeli <dodji@redhat.com>
7062
7063 * tree.h (DECL_BUILT_IN): Fix typo in comment.
7064
c167bc5b
DE		 70652013-09-04 David Edelsohn <dje.gcc@gmail.com>
7066
7067 * config/rs6000/rs6000.h (ASM_OUTPUT_DEF_FROM_DECLS): Only emit
7068 lglobl if not weak.
7069
035cb59f
ER		 70702013-09-04 Easwaran Raman <eraman@google.com>
7071
7072 PR middle-end/57370
7073 * tree-ssa-reassoc.c (get_stmt_uid_with_default): New function,
7074 (build_and_add_sum): Use it.
7075 (appears_later_in_bb): Simplify code.
7076
71191083
TJ		 70772013-09-04 Teresa Johnson <tejohnson@google.com>
7078
7079 * dumpfile.c (dump_finish): Don't close stderr/stdout.
7080
0c488213
JG		 70812013-09-04 James Greenhalgh <james.greenhalgh@arm.com>
7082
7083 * config/aarch64/arm_neon.h (vaddvq_<su>64): Fix return types.
7084
61a74079
JH		 70852013-09-04 Jan Hubicka <jh@suse.cz>
7086
7087 * Makefile.in (ipa-devirt.o): Add dependency on diagnostic.h
7088 * ipa-devirt.c: Include diganostic.h
7089 (odr_type_d): Add types and types_set.
7090 (hash_type_name): Work for types with vtables during LTO.
7091 (odr_hasher::remove): Fix comment; destroy types_set.
7092 (add_type_duplicate): New function,
7093 (get_odr_type): Use it.
7094 (dump_type_inheritance_graph): Dump type duplicates.
7095 * ipa.c (symtab_remove_unreachable_nodes): Build type inheritance
7096 graph.
7097 * tree.c (types_same_for_odr): Give exact answers on types with
7098 virtual tables.
7099
24ba8195
DS		 71002013-09-04 Dodji Seketeli <dodji@redhat.com>
7101
7102 * tree.h (DECL_BUILT_IN, DECL_IS_BUILTIN): Add more comments
7a5911d3 7103 explaining their differences.
24ba8195 7104
0a04869e
SKS		 71052013-09-04 Sandeep Kumar Singh<Sandeep.Singh2@kpitcummins.com>
7106
7107 * config/rx/rx.h: Add option -mcpu for target variants RX100 and RX200.
7108
3b18bc42
JL		 71092013-09-03 Jeff Law <law@redhat.com>
7110
7111 * tree-ssa-threadedge.c (thread_across_edge): Record entire path
7112 when not threading through a joiner block. Pass joiner/no joiner
7113 state to register_jump_thread.
7114 * tree-ssa-threadupdate.c (register_jump_thread): Get joiner/no joiner
7115 state from argument rather than implying on path length.
7116 Dump the entire jump thread path into debugging dump.
7117 * tree-flow.h (register_jump_thread): Update prototype.
7118
ae84e151
XDL		 71192013-08-29 Xinliang David Li <davidxl@google.com>
7120
7121 * tree-vect-data-refs.c (vect_compute_data_ref_alignment):
7122 Remove a trivial gcc_assert.
7123
c716e67f
XDL		 71242013-08-29 Xinliang David Li <davidxl@google.com>
7125
7126 * tree-vect-slp.c (destroy_bb_vec_info): Data ref cleanup.
7127 * tree-vect-loop.c (destroy_bb_vec_info): Ditto.
7128 * tree-vect-data-refs.c (vect_compute_data_ref_alignment):
7129 Delay base decl alignment adjustment.
7130 * tree-vectorizer.c (vect_destroy_datarefs): New function.
7131 * tree-vectorizer.h: New data structure.
7132 (set_dr_misalignment): New function.
7133 (dr_misalignment): Ditto.
7134 * tree-vect-stmts.c (vectorizable_store): Ensure alignment.
7135 (vectorizable_load): Ditto.
7136 (ensure_base_align): New function.
7137 (vectorize_loops): Add dbg_cnt support.
7138 (execute_vect_slp): Ditto.
7139 * dbgcnt.def: New debug counter.
7140 * Makefile: New dependency.
7141
db8800bc
MI		 71422013-09-03 Meador Inge <meadori@codesourcery.com>
7143
7144 Revert:
7145
7146 2013-08-30 Meador Inge <meadori@codesourcery.com>
7147
7148 * tree-vrp.c (check_array_ref): Bail out on zero-length arrays.
7149
0d9b4f5c
DE		 71502013-09-03 David Edelsohn <dje.gcc@gmail.com>
7151
7152 * config/rs6000/rs6000.h (ASM_OUTPUT_DEF_FROM_DECLS): Emit lglobl for
7153 function descriptor.
7154
29e810b9
RB		 71552013-09-03 Richard Biener <rguenther@suse.de>
7156
7157 * tree-affine.c (add_elt_to_tree): Fix association issue,
7158 avoid useless converts and make sure to always return a
7159 properly typed result.
7160
dbfc2894
RB		 71612013-09-03 Richard Biener <rguenther@suse.de>
7162
7163 PR middle-end/57656
7164 * fold-const.c (negate_expr_p): Fix division case.
7165 (negate_expr): Likewise.
7166
fa3bf4ce
RB		 71672013-09-03 Richard Biener <rguenther@suse.de>
7168
7169 PR lto/58285
7170 * tree-streamer-out.c: Include tm.h.
7171 * Makefile.in (tree-streamer-out.o): Depend on $(TM_H).
7172
a64bbb3f
JH		 71732013-09-03 Jan Hubicka <jh@suse.cz>
7174
7175 * tree-profile.c (tree_profiling): Cleanup CFG when done.
7176
93ed07e2
AM		 71772013-09-03 Alan Modra <amodra@gmail.com>
7178
7179 * config.gcc (powerpc*-*-linux*): Add support for little-endian
7180 multilibs to big-endian target and vice versa.
7181 * config/rs6000/t-linux64: Use := assignment on all vars.
7182 (MULTILIB_EXTRA_OPTS): Remove fPIC.
7183 (MULTILIB_OSDIRNAMES): Specify using mapping from multilib_options.
7184 * config/rs6000/t-linux64le: New file.
7185 * config/rs6000/t-linux64bele: New file.
7186 * config/rs6000/t-linux64lebe: New file.
7187
07d6cd64
JH		 71882013-09-02 Jan Hubicka <jh@suse.cz>
7189
7190 * ipa-inline-transform.c (inline_transform): Do not
7191 optimize_inline_calls when not optimizing.
7192
8e4c9a10
JH		 71932013-09-02 Jan Hubicka <jh@suse.cz>
7194
7195 * lto-symtab.c (lto_symtab_merge_symbols): Add comments; merge
7196 duplicated nodes for assembler names.
7197 * symtab.c (symtab_unregister_node): Do not attempt to unlink
7198 hard registers from assembler name hash.
7199
c91061e6
JH		 72002013-09-02 Jan Hubicka <jh@suse.cz>
7201
7202 * ipa-split.c (execute_split_functions): Split externally visible
7203 functions called once.
7204
2fd0985c
MJ		 72052013-09-02 Martin Jambor <mjambor@suse.cz>
7206
7207 PR ipa/58106
7208 * ipa-prop.c (ipa_edge_duplication_hook): Always put new rdesc to the
7209 linked list. When finding the correct duplicate, also consider also
7210 the caller in additon to its inlined_to node.
7211
91bd4114
JG		 72122013-09-02 James Greenhalgh <james.greenhalgh@arm.com>
7213
7214 * config/aarch64/aarch64-simd-builtins.def
7215 (dup_lane_scalar): Remove.
7216 * config/aarch64/aarch64-simd.md
7217 (aarch64_simd_dup): Add 'w->w' alternative.
7218 (aarch64_dup_lane<mode>): Allow for VALL.
7219 (aarch64_dup_lane_scalar<mode>): Remove.
7220 (aarch64_dup_lane_<vswap_width_name><mode>): New.
7221 (aarch64_get_lane_signed<mode>): Add w->w altenative.
7222 (aarch64_get_lane_unsigned<mode>): Likewise.
7223 (aarch64_get_lane<mode>): Likewise.
7224 * config/aarch64/aarch64.c (aarch64_evpc_dup): New.
7225 (aarch64_expand_vec_perm_const_1): Use aarch64_evpc_dup.
7226 * config/aarch64/iterators.md (VSWAP_WIDTH): New.
7227 (VCON): Change container of V2SF.
7228 (vswap_width_name): Likewise.
7229 * config/aarch64/arm_neon.h
7230 (__aarch64_vdup_lane_any): New.
7231 (__aarch64_vdup<q>_lane<q>_<fpsu><8,16,32,64>): Likewise.
7232 (vdup<q>_n_<psuf><8,16,32,64>): Convert to C implementation.
7233 (vdup<q>_lane<q>_<fpsu><8,16,32,64>): Likewise.
7234
d617d2d8
EB		 72352013-09-02 Eric Botcazou <ebotcazou@adacore.com>
7236
7237 PR middle-end/56382
7238 * expr.c (emit_move_complex): Do not move complex FP values as parts if
7239 the source or the destination is a single hard register.
7240
63fdb7be
RB		 72412013-09-02 Richard Biener <rguenther@suse.de>
7242
7243 PR middle-end/57511
7244 * tree-scalar-evolution.c (instantiate_scev_name): Allow
7245 non-linear SCEVs.
7246
78de2333
RB		 72472013-09-02 Richard Biener <rguenther@suse.de>
7248
7249 * tree-affine.c (add_elt_to_tree): Avoid converting all pointer
7250 arithmetic to sizetype.
7251
85ff4ec6
BC		 72522013-09-02 Bin Cheng <bin.cheng@arm.com>
7253
7254 * tree-ssa-loop-ivopts.c (set_autoinc_for_original_candidates):
7255 Find auto-increment use both before and after candidate.
7256
576016fe
MP		 72572013-09-02 Marek Polacek <polacek@redhat.com>
7258
7259 * Makefile.in (ubsan.o): Add $(TM_P_H) dependency.
7260
08f835dc
JH		 72612013-09-01 Jan Hubicka <jh@suse.cz>
7262
7263 * Makefile.in: Add ipa-profile.o
7264 (ipa.o, ipa-devrit.o, ipa-inline-analysis.o): Adjust dependencies.
7265 * cgraph.c (struct cgraph_propagate_frequency_data,
7266 cgraph_propagate_frequency_1, cgraph_propagate_frequency): Move to
7267 ipa-profile.c; replace cgraph_ by ipa_ prefix.
7268 * cgraph.h (cgraph_propagate_frequency): Remove.
f43245d1
UB		 7269 * ipa-inline-analysis.c: Include ipa-utils.h;
7270 drop duplicated cfgloop.h.
08f835dc
JH		 7271 (inline_update_callee_summaries): Update.
7272 * ipa-profile.c: New file.
7273 * ipa-utils.h (ipa_propagate_frequency): Declare.
7274 * ipa.c: Do not include pointer-set.h, hash-table.h, lto-streamer.h,
f43245d1 7275 data-streamer.h, value-prof.h.
08f835dc
JH		 7276 (symtab_remove_unreachable_nodes): Update profile.
7277 (struct histogram_entry, histogram, histogram_pool, histogram_hash,
7278 account_time_size, cmp_counts, dump_histogram,
7279 ipa_profile_generate_summary, ipa_profile_write_summary,
7280 ipa_profile_read_summary, ipa_profile, gate_ipa_profile,
7281 pass_data_ipa_profile, pass_ipa_profile, make_pass_ipa_profile):
f43245d1 7282 Move to ipa-profile.c.
08f835dc 7283
cec8583c
JDA		 72842013-09-01 John David Anglin <danglin@gcc.gnu.org>
7285
7286 * config/pa/pa.md: Allow "const 0" operand 1 in "scc" insns.
7287
bbc9396b
JH		 72882013-09-01 Jan Hubicka <jh@suse.cz>
7289
7290 * common.opt (fdevirtualize-speculatively): New function.
7291 * invoke.texi (fdevirtualize-speculatively): Document.
7292 * ipa-devirt.c: Include ipa-inline.h
7293 (likely_target_p): New function.
7294 (ipa_devirt): New function.
7295 (gate_ipa_devirt): New function.
7296 (pass_data_ipa_devirt): New static var.
7297 (pass_ipa_devirt): Likewise.
7298 (make_pass_ipa_devirt): New function.
7299 * opts.c (default_options): Add OPT_fdevirtualize_speculatively.
7300 (common_handle_option): Disable devirtualization when
7301 value range profiling is available.
7302 * passes.def (pass_ipa_devirt): Add.
7303 * timever.def (TV_IPA_DEVIRT): New timevar.
f43245d1 7304 * tree-pass.h (make_pass_ipa_devirt):
bbc9396b 7305
0cea211e
IS		 73062013-09-01 Iain Sandoe <iain@codesourcery.com>
7307
7308 * config/darwin.h (LINK_COMMAND_SPEC_A): Revise sanitizer specs to
7309 include sanitize(undefined).
7310
5e351e96
DN		 73112013-08-31 Diego Novillo <dnovillo@google.com>
7312
7313 * Makefile.in (TREE_CORE_H): Define.
7314 (TREE_H): Use.
7315 (GTFILES): Add tree-core.h.
7316 * builtins.c (built_in_class_names): Use BUILT_IN_LAST to
7317 size the array.
7318 * tree-core.h: New file.
7319 Move all data structures, enum, typedefs, global
7320 declarations and constants from ...
7321 * tree.h: ... here.
7322
63bf9a90
JH		 73232013-08-31 Jan Hubicka <jh@suse.cz>
7324
7325 * bulitins.c (expand_builtin): Do not early exit for gcov
7326 instrumented functions.
7327
5c0abd6a
MP		 73282013-08-31 Marek Polacek <polacek@redhat.com>
7329
7330 * ubsan.c: Include tm_p.h.
7331
4843f032
JH		 73322013-08-31 Jan Hubicka <jh@suse.cz>
7333
8356c89c
JH		 7334 * gimple-streamer-in.c (input_gimple_stmt): Silence parameter unused
7335 warning.
7336
4843f032
JH		 7337 * cgraph.c (cgraph_get_body): Update call of lto_input_function_body.
7338 * gimple-streamer-in.c (input_gimple_stmt): Move sanity check to ...
5e581212 7339 * tree-cfg.c (verify_gimple_label): ... here.
4843f032
JH		 7340 * ipa-utils.c: Include lto-streamer.h, ipa-inline.h
7341 (ipa_merge_profiles): New function.
7342 * lto-streamer-in.c (lto_read_body): Take node instead of fn_decl.
7343 (lto_input_function_body): Likewise.
5e581212
JH		 7344 * ipa-utils.h (ipa_merge_profiles): Declare.
7345 * lto-streamer.h (lto_input_function_body): Update prototype.
7346 (emit_label_in_global_context_p): Remove.
7347 * lto-symtab.c: Include ipa-utils.h
7348 (lto_cgraph_replace_node): Use ipa_merge_profiles.
4843f032 7349
57292ce9
JH		 73502013-08-31 Jan Hubicka <jh@suse.cz>
7351
7352 * cgraph.c (cgraph_speculative_call_info): Fix ref lookup
7353
b9cbfeeb
JH		 73542013-08-31 Jan Hubicka <jh@suse.cz>
7355
7356 * basic-block.h (apply_scale): Make scale parmeter gcov_type.
7357
41dedebd
UB		 73582013-08-31 Uros Bizjak <ubizjak@gmail.com>
7359
7360 * config/alpha/alpha.c (alpha_emit_conditional_move): Update
7361 "cmp" RTX before signed_comparison_operator check to account
7362 for "code" changes.
7363
06d65050
JH		 73642013-08-30 Jan Hubicka <jh@suse.cz>
7365
41dedebd 7366 * ipa-prop.c (ipa_set_jf_known_type): Check that we add only records.
06d65050
JH		 7367 (detect_type_change_1): Rename to ...
7368 (detect_type_change): ... this one; early return on non-polymorphic
7369 types.
41dedebd 7370 (detect_type_change_ssa): Add comp_type parameter; update
06d65050
JH		 7371 use of detect_type_change.
7372 (compute_complex_assign_jump_func): Add param_type parameter;
7373 update use of detect_type_change_ssa.
7374 (compute_complex_ancestor_jump_func): Likewise.
7375 (ipa_get_callee_param_type): New function.
7376 (ipa_compute_jump_functions_for_edge): Compute parameter type;
7377 update calls to the jump function computation functions.
7378
600b5b1d
TJ		 73792013-08-30 Teresa Johnson <tejohnson@google.com>
7380 Steven Bosscher <steven@gcc.gnu.org>
7381
7382 * cfgrtl.c (fixup_new_cold_bb): New routine.
7383 (commit_edge_insertions): Invoke fixup_partitions.
7384 (find_partition_fixes): New routine.
7385 (fixup_partitions): Ditto.
7386 (verify_hot_cold_block_grouping): Update comments.
7387 (rtl_verify_edges): Invoke find_partition_fixes.
7388 (rtl_verify_bb_pointers): Update comments.
7389 (rtl_verify_bb_layout): Ditto.
7390 * basic-block.h (probably_never_executed_edge_p): Declare.
7391 (fixup_partitions): Ditto.
7392 * cfgcleanup.c (try_optimize_cfg): Invoke fixup_partitions.
7393 * bb-reorder.c (sanitize_hot_paths): New function.
7394 (find_rarely_executed_basic_blocks_and_crossing_edges): Invoke
7395 sanitize_hot_paths.
7396 * predict.c (probably_never_executed_edge_p): New routine.
7397 * cfg.c (check_bb_profile): Add partition insanity warnings.
7398
0a6ea5c9
MI		 73992013-08-30 Meador Inge <meadori@codesourcery.com>
7400
7401 * tree-vrp.c (check_array_ref): Bail out on zero-length arrays.
7402
de5a5fa1
MP		 74032013-08-30 Marek Polacek <polacek@redhat.com>
7404
7405 * Makefile.in (ubsan.o): Add.
7406 (c-family/c-ubsan.o): Add.
7407 (builtins.o): Add ubsan.h dependency.
7408 * ubsan.h: New file.
7409 * ubsan.c: New file.
7410 * common.opt: Add -fsanitize=undefined option.
7411 (flag_sanitize): Add variable.
7412 (fsanitize=): Add option. Add Driver.
7413 (fsanitize=thread): Remove option.
7414 (fsanitize=address): Likewise.
7415 (static-libubsan): New option.
7416 * doc/invoke.texi: Document the new flag and -static-libubsan.
7417 * sanitizer.def (DEF_SANITIZER_BUILTIN): Define.
7418 (BUILT_IN_UBSAN_HANDLE_BUILTIN_UNREACHABLE): Define.
7419 * builtin-attrs.def (ATTR_COLD): Define.
7420 (ATTR_COLD_NOTHROW_LEAF_LIST): Define.
7421 * builtins.def (BUILT_IN_UBSAN_HANDLE_DIVREM_OVERFLOW,
7422 BUILT_IN_UBSAN_HANDLE_SHIFT_OUT_OF_BOUNDS): Define.
7423 * flag-types.h (sanitize_code): New enum.
7424 * opts.c (common_handle_option): Parse command line arguments
7425 of -fsanitize=. Add -fsanitize=unreachable option.
7426 * varasm.c (get_variable_section): Adjust.
7427 (assemble_noswitch_variable): Likewise.
7428 (assemble_variable): Likewise.
7429 (output_constant_def_contents): Likewise.
7430 (categorize_decl_for_section): Likewise.
7431 (place_block_symbol): Likewise.
7432 (output_object_block): Likewise.
7433 * builtins.def: Likewise.
7434 * toplev.c (compile_file): Likewise.
7435 (process_options): Likewise.
7436 * cppbuiltin.c: Likewise.
7437 * tsan.c (tsan_pass): Likewise.
7438 (tsan_gate): Likewise.
7439 (tsan_gate_O0): Likewise.
7440 * cfgexpand.c (partition_stack_vars): Likewise.
7441 (expand_stack_vars): Likewise.
7442 (defer_stack_allocation): Likewise.
7443 (expand_used_vars): Likewise.
7444 * cfgcleanup.c (old_insns_match_p): Likewise.
7445 * asan.c (asan_finish_file): Likewise.
7446 (asan_instrument): Likewise.
7447 (gate_asan): Likewise.
7448 (initialize_sanitizer_builtins): Build BT_FN_VOID_PTR_PTR_PTR.
7449 (ATTR_COLD_NOTHROW_LEAF_LIST): Define.
7450 (asan_global_struct): Use pointer_sized_int_node instead
7451 calling build_nonstandard_integer_type.
7452 (initialize_sanitizer_builtins): Likewise.
7453 (asan_finish_file): Likewise.
7454 * gcc.c: Document %{%:function(args):X}.
7455 (static_spec_functions): Add sanitize.
7456 (handle_spec_function): Add retval_nonnull argument and if non-NULL,
7457 store funcval != NULL there.
7458 (do_spec_1): Adjust handle_spec_function caller.
7459 (handle_braces): Allow %:function(args) as condition.
7460 (sanitize_spec_function): New function.
7461 (ADD_STATIC_LIBUBSAN_LIBS): Define.
7462 (LIBUBSAN_SPEC): Likewise.
7463 (LIBUBSAN_EARLY_SPEC): Likewise.
7464 (SANITIZER_SPEC): Handle libubsan.
7465 (SANITIZER_EARLY_SPEC): Likewise.
7466 * config/darwin.h (LINK_COMMAND_SPEC_A): Use %:sanitize(address)
7467 instead of fsanitize=address.
7468 * config/arm/linux-eabi.h (ASAN_CC1_SPEC): Use %:sanitize(address)
7469 instead of fsanitize=address*.
7470 * builtins.c: Include ubsan.h.
7471 (fold_builtin_0): Instrument __builtin_unreachable.
7472 * config/rs6000/rs6000.h (FRAME_GROWS_DOWNWARD): Use flag_sanitize
7473 instead of flag_asan.
7474 * tree.h (enum tree_index): Add TI_POINTER_SIZED_TYPE.
7475 (pointer_sized_int_node): Define.
41dedebd 7476 * tree.c (build_common_tree_nodes): Initialize pointer_sized_int_node.
de5a5fa1 7477
f07f30cf
MS		 74782013-08-30 Mike Stump <mikestump@comcast.net>
7479
7480 * doc/install.texi (Prerequisites): Note regression in Tcl 8.6
7481 with RE patterns.
7482
8a41354f
JH		 74832013-08-29 Jan Hubicka <jh@suse.cz>
7484
7485 * cgraph.c (cgraph_function_body_availability): Handle weakref
7486 correctly.
7487 * passes.def: Remove pass_fixup_cfg.
7488 * ipa-inline.c (ipa_inline): When not optimizing, do not inline;
7489 track when we need to remove functions.
7490 (gate_ipa_inline): Execute inlining always; add comment why.
7491 (pass_data_ipa_inline): Remove TODO_remove_functions.
7492 * ipa-inline-analysis.c (inline_generate_summary): When not optimizing
7493 do not produce summaries.
7494 * symtab.c (change_decl_assembler_name): Handle renaming of weakrefs.
7495 (symtab_nonoverwritable_alias): Assert we are not called on weakref.
7496 * varpool.c (cgraph_variable_initializer_availability): Fix weakrefs,
7497 constant pool and vtable.
7498
dc8d7a0f
TB		 74992013-08-30 Tejas Belagod <tejas.belagod@arm.com>
7500
41dedebd
UB		 7501 * config/aarch64/arm_neon.h (__AARCH64_UINT64_C, __AARCH64_INT64_C):
7502 New arm_neon.h's internal macros to specify 64-bit constants.
7503 Avoid using stdint.h's macros.
dc8d7a0f 7504
ac7eacd2
JR		 75052013-08-30 Joern Rennecke <joern.rennecke@embecosm.com>
7506
7507 * recog.c (verify_changes): Verify that changes[i].old is non-zero
7508 before applying REG_P.
7509
8b29fd4e
JJ		 75102013-08-30 Jakub Jelinek <jakub@redhat.com>
7511
7512 PR tree-optimization/58277
7513 * tree-ssa-strlen.c (strlen_enter_block): If do_invalidate gave up
7514 after seeing too many stmts with vdef in between dombb and current
7515 bb, invalidate everything.
7516
26d75703
RB		 75172013-08-30 Richard Biener <rguenther@suse.de>
7518
7519 * fold-const.c (fold_single_bit_test): Fix overflow test.
7520
f7e088e7
EB		 75212013-08-30 Eric Botcazou <ebotcazou@adacore.com>
7522
7523 * function.c (assign_parm_setup_reg): For a parameter passed by pointer
7524 and which can live in a register, always retrieve the value on entry.
7525 * var-tracking.c (add_stores): Treat the copy on entry for a parameter
7526 passed by invisible reference specially.
7527 (emit_notes_in_bb) <MO_VAL_USE>: Emit notes before the instruction.
7528 (vt_add_function_parameter): Correctly deal with a parameter passed by
7529 invisible reference.
7530
75312013-08-30 Jan Hubicka <jh@suse.cz>
47500756
JH		 7532
7533 * tree.c (set_call_expr_flags): Fix handling of TM_PURE.
7534
6e8dad05
RB		 75352013-08-30 Richard Biener <rguenther@suse.de>
7536
7537 PR tree-optimization/58228
7538 * tree-vect-data-refs.c (vect_analyze_data_ref_access): Do not
7539 allow invariant loads in nested loop vectorization.
7540
062ef2c8
RB		 75412013-08-30 Richard Biener <rguenther@suse.de>
7542
7543 PR tree-optimization/58223
7544 * tree-loop-distribution.c (has_anti_dependence): Rename to ...
7545 (has_anti_or_output_dependence): ... this and adjust to also
7546 look for output dependences.
7547 (mark_nodes_having_upstream_mem_writes): Adjust.
7548 (rdg_flag_uses): Likewise.
7549
7a764c60
RB		 75502013-08-30 Richard Biener <rguenther@suse.de>
7551
7552 PR tree-optimization/58010
7553 * tree-vect-loop.c (vect_create_epilog_for_reduction): Remove
7554 assert that we have a loop-closed PHI.
7555
82e9d642
JH		 75562013-08-29 Jan Hubicka <jh@suse.cz>
7557
7558 * lto-symtab.c (lto_cgraph_replace_node): Free decl_in_state.
7559 * cgraph.c (cgraph_release_function_body): Free decl_in_state.
7560 * lto-section-in.c (lto_free_function_in_decl_state): New function.
7561 (lto_free_function_in_decl_state_for_node): New function.
7562
9cc1fb4b
XDL		 75632013-08-29 Xinliang David Li <davidxl@google.com>
7564
41dedebd 7565 * loop-unroll.c (report_unroll_peel): Minor message change.
9cc1fb4b
XDL		 7566 * tree-vect-loop-manip.c (vect_do_peeling_for_alignment):
7567 Emit alignment peeling message with default -fopt-info.
7568 (vect_loop_versioning): Emit loop version info message.
41dedebd 7569 * tree-vectorizer.c (vectorize_loops): Minor message change.
9cc1fb4b
XDL		 7570 (execute_vect_slp): Ditto.
7571
440a5082
EB		 75722013-08-29 Eric Botcazou <ebotcazou@adacore.com>
7573
7574 * cgraphclones.c (cgraph_create_virtual_clone): Compute the DECL_NAME
7575 of the clone from the DECL_NAME of the original function.
7576
3fa3690d
OE		 75772013-08-29 Oleg Endo <olegendo@gcc.gnu.org>
7578
7579 * passes.c (register_pass): Add overload.
7580 * tree-pass.h (register_pass): Forward declare it. Add comment.
7581
0170f33c
JH		 75822013-08-29 Jan Hubicka <jh@suse.cz>
7583
41dedebd
UB		 7584 * lto-streamer-out.c (hash_tree): Stream DECL_FINAL_P,
7585 DECL_CXX_CONSTRUCTOR_P, DECL_CXX_DESTRUCTOR_P and TYPE_FINAL_P.
7586 * lto-streamer-in.c (unpack_ts_decl_with_vis_value_fields): Stream
7587 DECL_FINAL_P, DECL_CXX_CONSTRUCTOR_P and DECL_CXX_DESTRUCTOR_P.
0170f33c
JH		 7588 (unpack_ts_type_common_value_fields): Stream TYPE_FINAL_P.
7589 * tree-streamer-out.c (pack_ts_decl_with_vis_value_fields):
41dedebd 7590 Add DECL_FINAL_P, DECL_CXX_CONSTRUCTOR_P and DECL_CXX_DESTRUCTOR_P.
0170f33c
JH		 7591 (pack_ts_type_common_value_fields): Add TYPE_FINAL_P.
7592
103ff0d6
TJ		 75932013-08-29 Teresa Johnson <tejohnson@google.com>
7594
7595 * dumpfile.c (dump_loc): Output column number.
7596 * dumpfile.h (OPTGROUP_OTHER): Add and enable under OPTGROUP_ALL.
7597 * doc/invoke.texi: Document optall -fopt-info flag.
7598 * profile.c (read_profile_edge_counts): Use new dump framework.
7599 (compute_branch_probabilities): Ditto.
7600 * passes.c (pass_manager::register_one_dump_file): Use OPTGROUP_OTHER
7601 when pass not in any opt group.
7602 * pass_manager.h (pass_manager::get_pass_profile): New method.
7603 * value-prof.c (check_counter): Use new dump framework.
7604 (check_ic_target): Ditto.
7605 * coverage.c (get_coverage_counts): Ditto.
7606 (coverage_init): Setup new dump framework.
7607
301bbc16
RB		 76082013-08-29 Richard Biener <rguenther@suse.de>
7609
7610 PR tree-optimization/58246
7611 * tree-ssa-dce.c (mark_aliased_reaching_defs_necessary_1): Properly
7612 handle the dominance check inside a basic-block.
7613
1b275000
RB		 76142013-08-29 Richard Biener <rguenther@suse.de>
7615
7616 PR middle-end/57287
7617 * tree-ssa-copy.c (may_propagate_copy): Allow propagating
7618 of default defs that appear in abnormal PHI nodes.
7619
6b1184ba
RB		 76202013-08-29 Richard Biener <rguenther@suse.de>
7621
7622 PR tree-optimization/57685
7623 * tree-vrp.c (register_edge_assert_for_1): Recurse only for
7624 single-use operands to avoid exponential complexity.
7625
1b08b734
DC		 76262013-08-28 Dehao Chen <dehao@google.com>
7627
7628 * ipa-inline.c (edge_badness): Fix integer underflow.
7629
48a3fa69
UB		 76302013-08-28 Uros Bizjak <ubizjak@gmail.com>
7631
7632 * gtm-builtins.def (_ITM_free): Declare leaf.
7633
215f73e6
JJ		 76342013-08-28 Jakub Jelinek <jakub@redhat.com>
7635
7636 PR target/58067
7637 * config/i386/i386.md (*tls_global_dynamic_64_largepic): New insn.
7638 (*tls_local_dynamic_base_64_largepic): Likewise.
7639 (tls_global_dynamic_64_<mode>, tls_local_dynamic_base_64_<mode>):
7640 Remove predicate from call operand.
7641 * config/i386/i386.c (ix86_tls_get_addr): For -mcmodel=large -fpic
7642 return sum of pic_offset_table_rtx and UNSPEC_PLTOFF of the symbol.
7643
34554d1a
JL		 76442013-08-28 Jeff Law <law@redhat.com>
7645
7646 * tree-ssa-threadedge.c (thread_around_empty_block): Remove
7647 checks for the number of predecessors and successors allowed.
7648 * tree-ssa-threadupdate.c (mark_threaded_blocks): Ignore requests
7649 which require copying a joiner block if there is a request which
7650 is a subpath that requires no joiner block copying.
7651
c01c111b
JH		 76522013-08-28 Jan Hubicka <jh@suse.cz>
7653
7654 * lto-streamer-out.c (DFS_write_tree_body): Drop
7655 BINFO_INHERITANCE_CHAIN, BINFO_SUBVTT_INDEX and BINFO_VPTR_INDEX.
7656 (hash_tree): Do not hash DECL_DEFER_OUTPUT, BINFO_INHERITANCE_CHAIN,
7657 BINFO_SUBVTT_INDEX, BINFO_VPTR_INDEX, DECL_IN_TEXT_SECTION.
7658 * tree-streamer-in.c (unpack_ts_decl_common_value_fields):
7659 Do not read DECL_ERROR_ISSUED.
7660 (unpack_ts_decl_with_vis_value_fields): Do not read
7661 DECL_DEFER_OUTPUT.
48a3fa69
UB		 7662 (lto_input_ts_binfo_tree_pointers): Do not read
7663 BINFO_INHERITANCE_CHAIN, BINFO_SUBVTT_INDEX, BINFO_VPTR_INDEX
c01c111b
JH		 7664 * tree-streamer-out.c (pack_ts_decl_common_value_fields): Do not
7665 write DECL_ERROR_ISSUED..
7666 (pack_ts_decl_with_vis_value_fields): Do not write
7667 DECL_DEFER_OUTPUT.
7668 (write_ts_binfo_tree_pointers): Do not read BINFO_INHERITANCE_CHAIN,
48a3fa69 7669 BINFO_SUBVTT_INDEX, BINFO_VPTR_INDEX.
c01c111b
JH		 7670 * print-tree.c (print_node): Do not print DECL_ERROR_ISSUED.
7671 * tree.h (tree_decl_common): Update comment.
7672 (DECL_ERROR_ISSUED): Remove.
7673
4b128ece
JJ		 76742013-08-28 Jakub Jelinek <jakub@redhat.com>
7675
7676 PR middle-end/58257
7677 * omp-low.c (copy_var_decl): Copy over TREE_NO_WARNING flag.
7678
12211b99 76792013-08-28 Jan Hubicka <jh@suse.cz>
9e57846a
JH		 7680
7681 * builtins.def (free): Declare leaf.
7682
8dce4dbc
DM		 76832013-08-27 David Malcolm <dmalcolm@redhat.com>
7684
7685 * gdbhooks.py: New.
7686 * configure.ac (gdbinit.in): Add import of gcc/gdbhooks.py.
7687 * configure: Regenerate.
7688
b8f6e610
MJ		 76892013-08-27 Martin Jambor <mjambor@suse.cz>
7690
7691 * ipa-prop.h (ipa_pass_through_data): New field type_preserved.
7692 (ipa_ancestor_jf_data): Likewise.
7693 (ipa_get_jf_pass_through_agg_preserved): Fix comment typo.
7694 (ipa_get_jf_pass_through_type_preserved): New function.
7695 (ipa_get_jf_ancestor_agg_preserved): Fix comment typo.
7696 (ipa_get_jf_ancestor_type_preserved): New function.
12211b99 7697 * ipa-cp.c (ipa_get_jf_pass_through_result): Honor type_preserved flag.
b8f6e610
MJ		 7698 (ipa_get_jf_ancestor_result): Likewise.
7699 (propagate_vals_accross_pass_through): Use
7700 ipa_get_jf_pass_through_result to do all the value mappings.
7701 * ipa-prop.c (ipa_print_node_jump_functions_for_edge): Dump the
7702 type_preserved flag.
7703 (ipa_set_jf_cst_copy): New function.
7704 (ipa_set_jf_simple_pass_through): Set the type_preserved flag.
7705 (ipa_set_jf_arith_pass_through): Likewise.
7706 (ipa_set_ancestor_jf): Likewise.
7707 (compute_complex_assign_jump_func): Set type_preserved instead of
7708 punting.
7709 (ipa_compute_jump_functions_for_edge): Likewise.
7710 (combine_known_type_and_ancestor_jfs): Honor type_preserved.
7711 (update_jump_functions_after_inlining): Update type_preserved.
12211b99 7712 Explicitely create jump functions when combining one with pass_through.
b8f6e610
MJ		 7713 (ipa_write_jump_function): Stream the type_preserved flags.
7714 (ipa_read_jump_function): Likewise.
7715
74bf76ed
JJ		 77162013-08-27 Jakub Jelinek <jakub@redhat.com>
7717 Aldy Hernandez <aldyh@redhat.com>
7718
7719 * Makefile.in (omp-low.o): Depend on $(TARGET_H).
7720 * cfgloop.h (struct loop): Add safelen, force_vect, simduid.
7721 * function.h (struct function): Add has_force_vect_loops and
7722 has_simduid_loops.
12211b99 7723 * gimple-pretty-print.c (dump_gimple_omp_for): Handle GF_OMP_FOR_KIND*.
74bf76ed
JJ		 7724 * gimple.c (gimple_build_omp_critical): Add KIND argument and
7725 handle it.
7726 * gimple.def: Update CLAUSES comments.
7727 * gimple.h (enum gf_mask): Add GF_OMP_FOR_KIND_{FOR,SIMD}.
7728 (gimple_build_omp_for): Add argument to prototype.
7729 (gimple_omp_for_kind): New.
7730 (gimple_omp_for_set_kind): New.
7731 * gimplify.c (enum gimplify_omp_var_data): Add GOVD_LINEAR to
7732 GOVD_DATA_SHARE_CLASS.
7733 (enum omp_region_type): Add ORT_SIMD.
7734 (gimple_add_tmp_var): Handle ORT_SIMD.
7735 (gimplify_var_or_parm_decl): Same.
7736 (is_gimple_stmt): Same.
7737 (omp_firstprivatize_variable): Same.
7738 (omp_add_variable): Only use splay_tree_insert if lookup failed.
7739 (omp_notice_variable): Handle ORT_SIMD.
12211b99 7740 (omp_is_private): Add SIMD argument and handle it as well as ORT_SIMD.
74bf76ed
JJ		 7741 (omp_check_private): Handle ORT_SIMD.
7742 (gimplify_scan_omp_clauses): Handle OMP_CLAUSE_LINEAR and
7743 OMP_CLAUSE_SAFELEN.
7744 (gimplify_adjust_omp_clauses_1): Handle GOVD_LINEAR.
7745 Handle OMP_CLAUSE_LASTPRIVATE.
7746 (gimplify_adjust_omp_clauses): Handle OMP_CLAUSE_LINEAR and
7747 OMP_CLAUSE_SAFELEN.
7748 (gimplify_omp_for): Handle OMP_SIMD and OMP_CLAUSE_LINEAR.
7749 (gimplify_expr): Handle OMP_SIMD.
7750 * internal-fn.c (expand_GOMP_SIMD_LANE): New.
7751 (expand_GOMP_SIMD_VF): New.
7752 (expand_GOMP_SIMD_LAST_LANE): New.
7753 * internal-fn.def (GOMP_SIMD_LANE): New.
7754 (GOMP_SIMD_VF): New.
7755 (GOMP_SIMD_LAST_LANE): New.
7756 * omp-low.c: Include target.h.
7757 (extract_omp_for_data): Handle OMP_SIMD, OMP_CLAUSE_LINEAR,
7758 OMP_CLAUSE_SAFELEN.
7759 (check_omp_nesting_restrictions): Same.
7760 (omp_max_vf): New.
7761 (lower_rec_simd_input_clauses): New.
7762 (lower_rec_input_clauses): Handle OMP_SIMD, GF_OMP_FOR_KIND_SIMD,
7763 OMP_CLAUSE_LINEAR.
7764 (lower_lastprivate_clauses): Handle OMP_CLAUSE_LINEAR,
7765 GF_OMP_FOR_KIND_SIMD, OMP_SIMD.
7766 (expand_omp_build_assign): New.
7767 (expand_omp_for_init_counts): New.
7768 (expand_omp_for_init_vars): New.
7769 (extract_omp_for_update_vars): New.
7770 (expand_omp_for_generic): Use expand_omp_for_{init,update}_vars
7771 and rewrite accordingly.
7772 (expand_omp_simd): New.
7773 (expand_omp_for): Use expand_omp_simd.
7774 (lower_omp_for_lastprivate): Unshare vinit when appropriate.
7775 (lower_omp_for): Do not lower the body.
7776 * tree-data-ref (get_references_in_stmt): Allow IFN_GOMP_SIMD_LANE
7777 in their own loops.
7778 * tree-flow.h (find_omp_clause): Remove prototype.
12211b99
UB		 7779 * tree-if-conv.c (main_tree_if_conversion): Run if doing if conversion,
7780 forcing vectorization of the loop, or if flag_tree_vectorize.
74bf76ed
JJ		 7781 (gate_tree_if_conversion): Similarly.
7782 * tree-inline.c (remap_gimple_stmt): Pass for kind argument to
7783 gimple_build_omp_for.
7784 (copy_cfg_body): set has_force_vect_loops and has_simduid_loops.
7785 * tree-parloops (create_parallel_loop): Pass kind argument to
7786 gimple_build_omp_for.
7787 * tree-pretty-print.c (dump_omp_clause): Add cases for
7788 OMP_CLAUSE_UNIFORM, OMP_CLAUSE_LINEAR, OMP_CLAUSE_SAFELEN,
7789 OMP_CLAUSE__SIMDUID_.
7790 (dump_generic_node): Handle OMP_SIMD.
7791 * tree-ssa-ccp.c (likely_value): Handle IFN_GOMP_SIMD*.
7792 * tree-ssa-loop-ivcanon.c (tree_unroll_loops_completely_1): Do not
7793 unroll OMP_SIMD loops here.
12211b99 7794 * tree-ssa-loop.c (gate_tree_vectorize): Run if has_force_vect_loops.
74bf76ed 7795 * tree-vect-data-refs.c (vect_analyze_data_ref_dependence): Handle
12211b99 7796 loop->safelen.
74bf76ed
JJ		 7797 (vect_analyze_data_refs): Handle simd loops.
7798 * tree-vect-loop.c (vectorizable_live_operation): Handle
7799 IFN_GOMP_SIMD*.
12211b99 7800 * tree-vect-stmts.c (vectorizable_call): Handle IFN_GOMP_SIMD_LANE.
74bf76ed
JJ		 7801 (vectorizable_store): Handle STMT_VINFO_SIMD_LANE_ACCESS_P.
7802 (vectorizable_load): Same.
12211b99 7803 * tree-vectorizer.c: Include hash-table.h and tree-ssa-propagate.h.
74bf76ed
JJ		 7804 (struct simduid_to_vf): New.
7805 (simduid_to_vf::hash): New.
7806 (simduid_to-vf::equal): New.
7807 (struct simd_array_to_simduid): New.
7808 (simd_array_to_simduid::hash): New.
7809 (simd_array_to_simduid::equal): New.
7810 (adjust_simduid_builtins): New.
7811 (struct note_simd_array_uses_struct): New.
7812 (note_simd_array_uses_cb): New.
7813 (note_simd_array_uses): New.
7814 (vectorize_loops): Handle simd hints and adjust simd builtins
7815 accordingly.
7816 * tree-vectorizer.h (struct _stmt_vec_info): Add
7817 simd_lane_access_p field.
7818 (STMT_VINFO_SIMD_LANE_ACCESS_P): New macro.
7819 * tree.c (omp_clause_num_ops): Add entries for OMP_CLAUSE_LINEAR,
7820 OMP_CLAUSE_SAFELEN, OMP_CLAUSE__SIMDUID_, OMP_CLAUSE_UNIFORM.
7821 (omp_clause_code_name): Same.
7822 (walk_tree_1): Handle OMP_CLAUSE_UNIFORM, OMP_CLAUSE_SAFELEN,
7823 OMP_CLAUSE__SIMDUID_, OMP_CLAUSE_LINEAR.
7824 * tree.def (OMP_SIMD): New entry.
12211b99
UB		 7825 * tree.h (enum omp_clause_code): Add entries for OMP_CLAUSE_LINEAR,
7826 OMP_CLAUSE_UNIFORM, OMP_CLAUSE_SAFELEN, OMP_CLAUSE__SIMDUID_.
74bf76ed
JJ		 7827 (OMP_CLAUSE_DECL): Adjust range for new clauses.
7828 (OMP_CLAUSE_LINEAR_NO_COPYIN): New.
7829 (OMP_CLAUSE_LINEAR_NO_COPYOUT): New.
7830 (OMP_CLAUSE_LINEAR_STEP): New.
7831 (OMP_CLAUSE_SAFELEN_EXPR): New.
7832 (OMP_CLAUSE__SIMDUID__DECL): New.
7833 (find_omp_clause): New prototype.
7834
d0cf4e84
L		 78352013-08-27 H.J. Lu <hongjiu.lu@intel.com>
7836
7837 * config/i386/driver-i386.c (host_detect_local_cpu): Update
7838 Haswell processor detection.
7839
992592ec
CW		 78402013-08-27 Christian Widmer <shadow@umbrox.de>
7841
7842 PR target/57927
7843 * config/i386/driver-i386.c (host_detect_local_cpu): Add detection
7844 of Ivy Bridge and Haswell processors. Assume core-avx2 for unknown
7845 AVX2 capable processors.
7846
9116eb22
TB		 78472013-08-27 Tejas Belagod <tejas.belagod@arm.com>
7848
7849 * config/aarch64/arm_neon.h: Replace all inline asm implementations
7850 of vget_low_* with implementations in terms of other intrinsics.
7851
1c5abb53
MG		 78522013-08-27 Marc Glisse <marc.glisse@inria.fr>
7853
7854 PR middle-end/57219
7855 * doc/extend.texi (__builtin_isinf_sign): Restrict the return
7856 values to -1, 0 and 1.
7857
2e100703
VP		 78582013-08-27 Vidya Praveen <vidyapraveen@arm.com>
7859
7860 * config/aarch64/aarch64.md (unspec): Add UNSPEC_SISD_SSHL,
7861 UNSPEC_SISD_USHL, UNSPEC_USHL_2S, UNSPEC_SSHL_2S, UNSPEC_SISD_NEG.
7862 (<optab><mode>3_insn): Remove.
7863 (aarch64_ashl_sisd_or_int_<mode>3): New Pattern.
7864 (aarch64_lshr_sisd_or_int_<mode>3): Likewise.
7865 (aarch64_ashr_sisd_or_int_<mode>3): Likewise.
7866 (define_split for aarch64_lshr_sisd_or_int_di3): Likewise.
7867 (define_split for aarch64_lshr_sisd_or_int_si3): Likewise.
7868 (define_split for aarch64_ashr_sisd_or_int_di3): Likewise.
7869 (define_split for aarch64_ashr_sisd_or_int_si3): Likewise.
7870 (aarch64_sisd_ushl, aarch64_sisd_sshl): Likewise.
7871 (aarch64_ushl_2s, aarch64_sshl_2s, aarch64_sisd_neg_qi): Likewise.
7872 (ror<mode>3_insn): Likewise.
7873 * config/aarch64/predicates.md (aarch64_simd_register): New.
7874
4ded8276
RB		 78752013-08-27 Richard Biener <rguenther@suse.de>
7876
7877 PR tree-optimization/57521
7878 * tree-if-conv.c (if_convertible_bb_p): Verify that at least
7879 one edge is non-critical.
7880 (find_phi_replacement_condition): Make sure to use a non-critical
7881 edge. Cleanup and remove old bug workarounds.
7882 (bb_postdominates_preds): Remove.
7883 (if_convertible_loop_p_1): Do not compute post-dominators.
7884 (combine_blocks): Do not free post-dominators.
7885 (main_tree_if_conversion): Likewise.
7886 (pass_data_if_conversion): Add TODO_verify_ssa.
7887
5aa11061
DD		 78882013-08-27 DJ Delorie <dj@redhat.com>
7889
7890 * config/i386/djgpp.h (ASM_DECLARE_FUNCTION_NAME): New.
7891
fd91cfe3
YZ		 78922013-08-27 Yufeng Zhang <yufeng.zhang@arm.com>
7893
7894 * function.c (assign_parm_find_data_types): Set passed_mode and
7895 nominal_mode to the TYPE_MODE of nominal_type for the built
7896 pointer type in case of the struct-pass-by-reference.
7897
907555ce
JR		 78982013-08-26 Joern Rennecke <joern.rennecke@embecosm.com>
7899
7900 * config/avr/avr-stdint.h (INT16_TYPE): Change default to "int".
7901 (UINT16_TYPE): Change default to "unsigned int".
7902
67518c93
JR		 7903 * config/avr/avr.opt (mfract-convert-truncate): New option.
7904 * config/avr/avr.c (avr_out_fract): Unless TARGET_FRACT_CONV_TRUNC
7905 is set, round negative fractional integers according to n1169
7906 when converting to integer types.
7907
d6d989fc
JH		 79082013-08-26 Jan Hubicka <jh@suse.cz>
7909
7910 * cgraph.c (cgraph_propagate_frequency): Do not assume that virtual
7911 methods can not be called indirectly when their address is not taken.
7912
2aa3da06
JH		 79132013-08-26 Jan Hubicka <jh@suse.cz>
7914
12211b99
UB		 7915 * gimple-fold.c (gimple_get_virt_method_for_binfo): Use
7916 ctor_for_folding.
2aa3da06 7917
0987ffe7
JH		 79182013-08-26 Jan Hubicka <jh@suse.cz>
7919
7920 * ipa.c (comdat_can_be_unshared_p_1): C++ constructors and destructors
7921 can be unshared.
7922
befe8647
JR		 79232013-08-26 Joern Rennecke <joern.rennecke@embecosm.com>
7924
7925 * reload.c (find_valid_class): Allow classes that do not include
7926 FIRST_PSEUDO_REGISTER - 1.
7927
a21e735e
JH		 79282013-08-26 Jan Hubicka <jh@suse.cz>
7929
7930 * cgraph.c (cgraph_redirect_edge_call_stmt_to_callee): Fix formatting;
7931 fix edge count/frequency when speculation failed; fix type check
7932 for the direct call.
7933
e067bd43
JH		 79342013-08-26 Jan Hubicka <jh@suse.cz>
7935
7936 * ipa-prop.c (ipa_print_node_params): Do not ICE during WPA.
7937
e8aec975
JH		 79382013-08-26 Jan Hubicka <jh@suse.cz>
7939
7940 * ipa-inline-transform.c (inline_transform): Be ready for basic block
7941 to be changed by edge redirection.
7942
d0b66480
JH		 79432013-08-26 Jan Hubicka <jh@suse.cz>
7944
12211b99
UB		 7945 * cgraph.c (cgraph_speculative_call_info): Fix parameter order and
7946 formating; add sanity check.
d0b66480
JH		 7947 (cgraph_resolve_speculation): Add FIXME about scaling profiles.
7948 (cgraph_redirect_edge_call_stmt_to_callee): Fix ICE in debug dump.
7949 * ipa-inline.c (heap_edge_removal_hook): Reset node growth cache.
7950 (resolve_noninline_speculation): Update callee keys, too.
7951
0f9aaac7
JH		 79522013-08-26 Jan Hubicka <jh@suse.cz>
7953
7954 * tree.h (tree_decl_with_vis): Add cxx_constructor, cxx_destructor.
7955 (DECL_CXX_CONSTRUCTOR_P, DECL_CXX_DESTRUCTOR_P): New macros.
7956
4b37444e
JR		 79572013-08-26 Joern Rennecke <joern.rennecke@embecosm.com>
7958
7959 * config/i386/i386.c (x86_64_elf_select_section): Put ATTRIBUTE_UNUSED
7960 into proper place.
7961
5339fc0c
UB		 79622013-08-26 Uros Bizjak <ubizjak@gmail.com>
7963
7964 * config/i386/i386.c (ix86_debug_options): Remove prototype.
7965 (x86_64_elf_select_section): Ditto.
7966 (ix86_handle_tm_regparm_attribute): Remove ATTRIBUTE_UNUSED on used
7967 arguments.
7968 (ix86_pass_by_reference): Ditto.
5339fc0c
UB		 7969 (output_set_got): Ditto.
7970 (ix86_unary_operator_ok): Ditto.
7971 (ix86_expand_builtin): Ditto.
7972
d5c3d3ef
JH		 79732013-08-23 Jan Hubicka <jh@suse.cz>
7974
5339fc0c 7975 * cgraph.c (cgraph_turn_edge_to_speculative): Fix debug output.
d5c3d3ef 7976
5b1e7435
JH		 79772013-08-23 Jan Hubicka <jh@suse.cz>
7978
7979 * tree.h (TYPE_FINAL_P, DECL_FINAL_P): New macros.
7980 (tree_decl_with_vis): Add FINAL field.
7981
ead69dac
JL		 79822013-08-23 Jeff Law <law@redhat.com>
7983
7984 * tree-ssa-pre.c (do_regular_insertion): Include the expression in
7985 the debugging dump when the expression is fully redundant.
7986
025311c4
GDR		 79872013-08-23 Gabriel Dos Reis <gdr@integrable-solutions.net>
7988
7989 * diagnostic.c (diagnostic_set_caret_max_width): Use pp_buffer.
7990 * gimple-pretty-print.c (gimple_dump_bb_buff): Likewise.
7991 * pretty-print.c (pp_formatted_text_data): Likewise.
7992 (pp_write_text_to_stream): Likewise.
7993 (pp_write_text_as_dot_label_to_stream): Likewise.
7994 (pp_append_r): Likewise.
7995 (pp_format): Likewise.
7996 (pp_flush): Likewise.
7997 (pp_clear_output_area): Likewise.
7998 (pp_append_text): Likewise.
7999 (pp_formatted_text): Likewise.
8000 (pp_remaining_character_count_for_line): Likewise.
8001 (pp_newline): Likewise.
8002 (pp_character): Likewise.
8003 (output_buffer::~output_buffer): Define.
8004 (pretty_printer::~pretty_printer): Destruct output buffer.
8005 * pretty-print.h (output_buffer::~output_buffer): Declare.
8006 (pretty_printer::~pretty_printer): Declare virtual.
8007
0cadbfaa
MG		 80082013-08-24 Marc Glisse <marc.glisse@inria.fr>
8009
8010 PR other/57324
8011 * hwint.h (HOST_WIDE_INT_UC, HOST_WIDE_INT_1U, HOST_WIDE_INT_M1,
8012 HOST_WIDE_INT_M1U): New macros.
8013 * fold-const.c (sign_bit_p, build_range_check, fold_unary_loc,
8014 fold_binary_loc, fold_ternary_loc): Use the new macros. Use an
8015 unsigned -1 for lshift.
8016 * cse.c (cse_insn): Likewise.
8017 * double-int.c (rshift_double, lshift_double): Likewise.
8018 * builtins.c (fold_builtin_bitop): Likewise.
8019 * combine.c (force_to_mode): Likewise.
8020 * tree.c (integer_pow2p, tree_log2, tree_floor_log2): Likewise.
8021 * simplify-rtx.c (simplify_const_unary_operation,
8022 simplify_const_binary_operation): Likewise.
8023 * tree-stdarg.c (va_list_counter_bump, va_list_ptr_read,
8024 check_va_list_escapes): Likewise.
8025 * rtlanal.c (nonzero_bits1): Likewise.
8026 * expmed.c (expand_smod_pow2): Likewise.
8027 * tree-ssa-structalias.c (UNKNOWN_OFFSET): Use HOST_WIDE_INT_MIN.
8028
009e5353
JH		 80292013-08-23 Jan Hubicka <jh@suse.cz>
8030
8031 * cgraph.c (cgraph_turn_edge_to_speculative): Mark target node
8032 as having address taken.
8033
c4be6568
JH		 80342013-08-23 Jan Hubicka <jh@suse.cz>
8035
64cbf23d
JH		 8036 * ipa-utils.h (method_class_type): Declare.
8037 * ipa-devirt.c (method_class_type): Export.
8038
c4be6568
JH		 8039 * cgraphunit.c (analyze_functions): Do basic devirtualization;
8040 do not walk base classes of anonymous types.
8041
97aba8e9
KK		 80422013-08-23 Kaz Kojima <kkojima@gcc.gnu.org>
8043
8044 PR rtl-optimization/58220
8045 PR regression/58221
8046 * final.c (reemit_insn_block_notes): Use NEXT_INSN to
8047 handle SEQUENCE insns properly.
8048
0fc80001
GDR		 80492013-08-23 Gabriel Dos Reis <gdr@integrable-solutions.net>
8050
8051 * pretty-print.h (pp_newline_and_flush): Declare. Remove macro
8052 definition.
8053 (pp_newline_and_indent): Likewise.
8054 (pp_separate_with): Likewise.
8055 * pretty-print.c (pp_newline_and_flush): Define.
8056 (pp_newline_and_indent): Likewise.
8057 (pp_separate_with): Likewise.
8058
520a5868
JJ		 80592013-08-23 Jakub Jelinek <jakub@redhat.com>
8060
8061 PR target/58218
8062 * config/i386/x86-64.h (TARGET_SECTION_TYPE_FLAGS): Define.
8063 * config/i386/i386.c (x86_64_elf_section_type_flags): New function.
8064
e500c62a
KY		 80652013-08-23 Kirill Yukhin <kirill.yukhin@intel.com>
8066
f43245d1
UB		 8067 * config/i386/predicates.md (ext_sse_reg_operand): New.
8068 * config/i386/i386.md (*movti_internal): Use
e500c62a
KY		 8069 predicate to determine if EVEX is needed.
8070 (*movsi_internal): Ditto.
8071 (*movdf_internal): Ditto.
8072 (*movsf_internal): Ditto.
f43245d1 8073 * config/i386/mmx.md (*mov<mode>_internal): Ditto.
e500c62a 8074
6b00d7dd
JJ		 80752013-08-23 Jakub Jelinek <jakub@redhat.com>
8076
8077 PR tree-optimization/58209
8078 * tree-tailcall.c (process_assignment): Handle POINTER_PLUS_EXPR.
8079 (find_tail_calls): Give up for pointer result types if m is non-NULL.
8080 (adjust_return_value_with_ops): For PLUS_EXPR and pointer result type
8081 emit POINTER_PLUS_EXPR.
8082 (create_tailcall_accumulator): For pointer result type accumulate in
8083 sizetype type.
8084
4f2a9f90
PC		 80852013-08-22 Paolo Carlini <paolo.carlini@oracle.com>
8086
8087 * configure.ac: Add backslashes missing from the last change.
8088 * configure: Regenerate.
8089
7bf4274e
JH		 80902013-08-22 Jan Hubicka <jh@susue.cz>
8091
4f2a9f90
PC		 8092 * ipa.c (function_and_variable_visibility): First remember function
8093 was global and then make it local.
7bf4274e 8094
1bea243d
JB		 80952013-08-22 Julian Brown <julian@codesourcery.com>
8096
8097 * configure.ac: Add aarch64 to list of arches which use "nop" in
8098 debug_line test.
8099 * configure: Regenerate.
8100
bcba7ecc
AK		 81012013-08-22 Andreas Krebbel <Andreas.Krebbel@de.ibm.com>
8102
8103 * config/s390/linux.h (TARGET_LIBC_HAS_FUNCTION): Define as
8104 gnu_libc_has_function.
8105 * config/s390/tpf.h: Likewise.
8106
4c495b0d
JH		 81072013-08-22 Jan Hubicka <jh@susue.cz>
8108
8109 * timevar.c (validate_phases): Add cast.
8110
5e302bca
JH		 81112013-08-22 Jan Hubicka <jh@susue.cz>
8112
8113 * timevar.c (validate_phases): Use size_t for memory.
8114 * timevar.h (struct timevar_time_def): Use size_t for ggc_mem.
8115
da6ca2b5
GDR		 81162013-08-22 Gabriel Dos Reis <gdr@integrable-solutions.net>
8117
8118 * pretty-print.h (output_buffer::output_buffer): Declare.
8119 (pretty_printer::pretty_printer): Likewise.
8120 (pp_construct): Remove.
8121 * pretty-print.c (output_buffer::output_buffer): Define.
8122 (pretty_printer::pretty_printer): Rename from pp_construct. Simplify.
8123 * gimple-pretty-print.c (print_gimple_stmt): Do not call pp_construct.
8124 (print_gimple_expr): Likewise.
8125 (print_gimple_seq): Likewise.
8126 (gimple_dump_bb): Likewise.
8127 * sched-vis.c (dump_value_slim): Likewise.
8128 (dump_insn_slim): Likewise.
8129 (dump_rtl_slim): Likewise.
8130 (str_pattern_slim): Likewise.
8131 * tree-mudflap.c (mf_varname_tree): Likewise.
8132 * graph.c (print_graph_cfg): Likewise.
8133 (start_graph_dump): Likewise.
8134 * tree-pretty-print.c (maybe_init_pretty_print): Likewise. Use
8135 placement-new.
8136 * diagnostic.c (diagnostic_initialize): Simplify early diagnostic
8137 pretty printer initialization.
8138 * coretypes.h (diagnostic_context): Remove superflous type alias
8139 declaration.
8140 (pretty_printer): Likewise. Declare directly as a class.
8141 (pretty_print_info): Remove declaration as class.
8142 * asan.c (asan_emit_stack_protection): Remove call to pp_construct
8143 and pp_clear_output_area.
8144 (asan_add_global): Likewise.
8145
0e1474e5
JH		 81462013-08-22 Jan Hubicka <jh@suse.cz>
8147
8148 * cgraphunit.c (analyze_functions) Use update_type_inheritance_graph.
8149 * ipa-utils.h (update_type_inheritance_graph): Declare.
8150 (possible_polymorphic_call_target_p): Declare.
8151 (possible_polymorphic_call_target_p): New.
8152 * ipa-devirt.c: Update toplevel comments.
8153 (cached_polymorphic_call_targets): Move up.
8154 (odr_type_d): Move ID down.
8155 (polymorphic_type_binfo_p): Update comment.
8156 (odr_hasher::remove): Likewise;
8157 (get_odr_type): Set anonymous_namespace.
8158 (dump_odr_type): Dump it.
8159 (dump_type_inheritance_graph): Do not ICE when there are no ODR types.
8160 (maybe_record_node): Record node in cached_polymorphic_call_targets.
8161 (record_binfo): Add comment.
5339fc0c
UB		 8162 (free_polymorphic_call_targets_hash): Do not ICE when cache is not
8163 built.
0e1474e5
JH		 8164 (devirt_node_removal_hook): Do not iCE when cache is freed.
8165 (possible_polymorphic_call_target_p): New predicate.
8166 (update_type_inheritance_graph): New function.
8167
3f97cb0b
AI		 81682013-08-22 Alexander Ivchenko <alexander.ivchenko@intel.com>
8169 Maxim Kuznetsov <maxim.kuznetsov@intel.com>
8170 Sergey Lega <sergey.s.lega@intel.com>
8171 Anna Tikhonova <anna.tikhonova@intel.com>
8172 Ilya Tocar <ilya.tocar@intel.com>
8173 Andrey Turetskiy <andrey.turetskiy@intel.com>
8174 Ilya Verbin <ilya.verbin@intel.com>
8175 Kirill Yukhin <kirill.yukhin@intel.com>
8176 Michael Zolotukhin <michael.v.zolotukhin@intel.com>
8177
8178 * common/config/i386/i386-common.c (OPTION_MASK_ISA_AVX512F_SET): New.
8179 (OPTION_MASK_ISA_AVX512CD_SET): Ditto.
8180 (OPTION_MASK_ISA_AVX512PF_SET): Ditto.
8181 (OPTION_MASK_ISA_AVX512ER_SET): Ditto.
8182 (OPTION_MASK_ISA_AVX2_UNSET): Update.
8183 (OPTION_MASK_ISA_AVX512F_UNSET): New.
8184 (OPTION_MASK_ISA_AVX512CD_UNSET): Ditto.
8185 (OPTION_MASK_ISA_AVX512PF_UNSET): Ditto.
8186 (OPTION_MASK_ISA_AVX512ER_UNSET): Ditto.
8187 (ix86_handle_option): Handle OPT_mavx512f, OPT_mavx512cd,
8188 OPT_mavx512pf, OPT_mavx512er cases.
8189 * config/i386/constraints.md (v): New constraint.
8190 (Yi, Yj): Replace SSE_REGS with ALL_SSE_REGS.
8191 * config/i386/cpuid.h (bit_AVX512F, bit_AVX512PF, bit_AVX512ER)
8192 (bit_AVX512CD): New.
8193 * config/i386/driver-i386.c (host_detect_local_cpu): Detect
8194 AVX512F, AVX512ER, AVX512PF, AVX512CD features.
8195 * config/i386/i386-c.c (ix86_target_macros_internal):
8196 Conditionally define __AVX512F__, __AVX512ER__, __AVX512CD__,
8197 __AVX512PF__.
8198 * config/i386/i386-modes.def (VECTOR_MODES (INT, 128))
8199 (VECTOR_MODES (FLOAT, 128), INT_MODE (XI, 64)): New modes.
8200 * config/i386/i386.c (regclass_map, dbx_register_map)
8201 (dbx64_register_map, svr4_dbx_register_map): Add new SSE registers.
8202 (gate_insert_vzeroupper): Disable vzeroupper for TARGET_AVX512F.
8203 (ix86_target_string): Define -mavx512f, -mavx512er, -mavx512cd,
8204 -mavx512pf options.
8205 (ix86_option_override_internal): Define PTA_AVX512F, PTA_AVX512ER,
8206 PTA_AVX512PF, PTA_AVX512CD. Handle -mavx512f, -mavx512er, -mavx512cd,
8207 -mavx512pf options. Fix formatting.
8208 (ix86_conditional_register_usage): Squash EXT_REX_SSE_REGs for 32-bit
8209 targets. Squash EVEX_SSE_REGS if AVX512F is disabled.
8210 (ix86_valid_target_attribute_inner_p): Handle -mavx512f, -mavx512er,
8211 -mavx512cd, -mavx512pf options.
8212 (standard_sse_constant_opcode): Add vpternlogd for 512-bit modes.
8213 (print_reg, ix86_print_operand): Handle 'g' to output 512-bit operands.
8214 (ix86_preferred_output_reload_class): Replace SSE_REGS with
8215 ALL_SSE_REGS.
8216 (ix86_hard_regno_mode_ok): Support 512-bit registers.
8217 (ix86_set_reg_reg_cost): Ditto.
8218 (x86_order_regs_for_local_alloc): Ditto.
8219 (MAX_VECT_LEN): Extend to 64-byte.
8220 (ix86_spill_class): Replace SSE_REGS with ALL_SSE_REGS.
8221 * config/i386/i386.h (TARGET_AVX512F, TARGET_AVX512PF)
8222 (TARGET_AVX512ER, TARGET_AVX512CD): New.
8223 (BIGGEST_ALIGNMENT): Extend to 512-bits.
8224 (FIRST_PSEUDO_REGISTER, FIXED_REGISTERS): Add new registers.
8225 (CALL_USED_REGISTERS, REG_ALLOC_ORDER): Likewise.
8226 (VALID_AVX512F_SCALAR_MODE, VALID_AVX512F_REG_MODE): New.
8227 (SSE_REG_MODE_P): Support new modes.
8228 (FIRST_MMX_REG, FIRST_REX_INT_REG, FIRST_REX_SSE_REG): Add comments.
8229 (FIRST_EXT_REX_SSE_REG, LAST_EXT_REX_SSE_REG): New.
8230 (reg_class, REG_CLASS_NAMES): Add EVEX_SSE_REGS, ALL_SSE_REGS.
8231 (SSE_CLASS_P, MAYBE_SSE_CLASS_P): Replace SSE_REGS with ALL_SSE_REGS.
8232 (REG_CLASS_CONTENTS): Add new registers.
8233 (SSE_REGNO_P, SSE_REGNO, HARD_REGNO_RENAME_OK): Support new registers.
8234 (EXT_REX_SSE_REGNO_P): New.
8235 (HI_REGISTER_NAMES): Add new registers.
8236 * config/i386/i386.md: Define constants for new registers.
8237 (mode): Add new 512-bit modes.
8238 (prefix): Support evex prefix.
8239 (isa): Support avx512f, noavx512f, fma_avx512f.
8240 (ssemodesuffix): Add new 512-bit modes.
8241 (movxi): New.
8242 (*movxi_internal_avx512f): Ditto.
8243 (*movdi_internal): Replace constraint "x" with the new constraint "v".
8244 Support MODE_XI.
8245 (*movsi_internal): Likewise.
8246 (*movdf_internal): Likewise.
8247 (*movsf_internal): Likewise.
8248 (*fop_<mode>_comm_sse): Replace constraint "x" with new constraint "v".
8249 (<code><mode>3): Likewise.
5339fc0c
UB		 8250 * config/i386/i386.opt (mavx512f, mavx512pf, mavx512er, mavx512cd):
8251 New.
3f97cb0b
AI		 8252 * config/i386/mmx.md (*mov<mode>_internal): Replace constraint "x"
8253 with the new constraint "v".
8254 * config/i386/sse.md (*mov<mode>_internal): Support new registers and
8255 modes.
8256 (<sse>_loadu<ssemodesuffix><avxsizesuffix>): Replace constraint "x"
8257 with the new constraint "v".
8258 (<sse2>_loaddqu<avxsizesuffix>): Likewise.
8259 (<sse2>_storedqu<avxsizesuffix>): Likewise.
8260 (*<plusminus_insn><mode>3): Likewise.
8261 (<sse>_vm<plusminus_insn><mode>3): Likewise.
8262 (*mul<mode>3): Likewise.
8263 (<sse>_vmmul<mode>3): Likewise.
8264 (<sse>_div<mode>3): Likewise.
8265 (<sse>_vmdiv<mode>3): Likewise.
8266 (<sse>_sqrt<mode>2): Likewise.
8267 (<sse>_vmsqrt<mode>2): Likewise.
8268 (*<code><mode>3_finite): Likewise.
8269 (*<code><mode>3) <smaxmin>: Likewise.
8270 (<sse>_vm<code><mode>3): Likewise.
8271 (*<code><mode>3) <any_logic>: Likewise.
8272 (*fma_fmadd_<mode>): Likewise.
8273 (*fma_fmsub_<mode>): Likewise.
8274 (*fma_fnmadd_<mode>): Likewise.
8275 (*fma_fnmsub_<mode>): Likewise.
8276 (*fma_fmaddsub_<mode>): Likewise.
8277 (*fma_fmsubadd_<mode>): Likewise.
8278 (*fmai_fmadd_<mode>): Likewise.
8279 (*fmai_fmsub_<mode>): Likewise.
8280 (*fmai_fnmadd_<mode>): Likewise.
8281 (*fmai_fnmsub_<mode>): Likewise.
8282 (sse_cvtsi2ss): Likewise.
8283 (sse_cvtsi2ssq): Likewise.
8284 (sse_cvtss2si): Likewise.
8285 (sse_cvtss2si_2): Likewise.
8286 (sse_cvtss2siq): Likewise.
8287 (sse_cvtss2siq_2): Likewise.
8288 (sse_cvttss2si): Likewise.
8289 (sse_cvtss2siq_2): Likewise.
8290 (float<sseintvecmodelower><mode>2): Likewise.
8291 (sse2_cvtsd2si_2): Likewise.
8292 (sse2_cvtsd2siq_2): Likewise.
8293 (*<plusminus_insn><mode>3): Likewise.
8294 (*<sse2_avx2>_<plusminus_insn><mode>3): Likewise.
8295 (*<sse4_1_avx2>_mul<mode>3): Likewise.
8296 (ashr<mode>3): Likewise.
8297 (<shift_insn><mode>3): Likewise.
8298 (avx2_<code><mode>3): Likewise.
8299 (*avx2_<code><mode>3): Likewise.
8300 (*andnot<mode>3): Likewise.
8301 (*<code><mode>3) <any_logic>: Likewise.
8302 (abs<mode>2): Likewise.
8303 (avx2_permvar<mode>): Likewise.
8304 (avx2_perm<mode>_1): Likewise.
8305 (*avx_vpermilp<mode>): Likewise.
8306 (avx_vpermilvar<mode>3): Likewise.
8307 (avx2_ashrv<mode>): Likewise.
8308 (avx2_<shift_insn>v<mode>): Likewise.
8309 * doc/invoke.texi: Document -mavx512f, -mavx512pf, -mavx512er,
8310 -mavx512cd.
8311 * doc/rtl.texi: Document XImode.
8312
dd5e8423
JL		 83132013-08-21 Jeff Law <law@redhat.com>
8314
b9ebee5d
JL		 8315 * tree-flow.h (register_jump_thread): Pass vector of edges
8316 instead of each important edge.
8317 * tree-ssa-threadedge.c (thread_across_edge): Build the jump
8318 thread path into a vector and pass that to register_jump_thread.
8319 * tree-ssa-threadupdate.c (register_jump_thread): Conver the
8320 passed in edge vector to the current 3-edge form.
8321
dd5e8423
JL		 8322 Revert:
8323 2013-08-20 Alexey Makhalov <makhaloff@gmail.com>
8324
9bb6628e 8325 * dce.c (fini_dce): Call df_analyze again just in case
dd5e8423
JL		 8326 delete_unmarked_insns removed anything.
8327
445dc8df
JR		 83282013-08-21 Joern Rennecke <joern.rennecke@embecosm.com>
8329
8330 * reload.h (struct reg_equivs): Rename to ..
8331 (struct reg_equivs_s): .. this.
8332
e0df53dd
ML		 83332013-08-20 Martin Liska <marxin.liska@gmail.com>
8334
8335 * ipa.c (ipa_profile_read_summary): Fix buffer overflow.
8336
e7d1d3eb
RO		 83372013-08-21 Rainer Orth <ro@CeBiTec.Uni-Bielefeld.DE>
8338
5339fc0c 8339 * config/sol2-10.h (TARGET_LIBC_HAS_FUNCTION): Don't nest comment.
e7d1d3eb 8340
5562e26e
JL		 83412013-08-21 Jeff Law <law@redhat.com>
8342
8343 * tree-vrp.c (simplify_stmt_for_jump_threading): Try to
8344 simplify assignments too. If the RHS collapses to a singleton
8345 range, then return the value for the range.
8346
c7ecdec6
KY		 83472013-08-21 Kirill Yukhin <kirill.yukhin@intel.com>
8348
7a5911d3
UB		 8349 * config/i386/sse.md (V16): Rename to...
8350 (VMOVE): this.
8351 (mov<mode>): Update iterator name.
8352 (*mov<mode>_internal): Ditto.
8353 (push<mode>1): Ditto.
8354 (movmisalign<mode>): Ditto.
c7ecdec6 8355
bfa3b50a
JH		 83562013-08-20 Jan Hubicka <jh@suse.cz>
8357
8358 PR bootstrap/58186
8359 * cgraph.c (cgraph_add_edge_to_call_site_hash): Overwrite hash
8360 entry for direct edges.
8361 (cgraph_turn_edge_to_speculative): Fix setting of can_throw_external.
8362
6cd4d135
DM		 83632013-08-20 David Malcolm <dmalcolm@redhat.com>
8364
8365 Revert my last two changes, r201865 and r201864:
8366
8367 Revert r201865:
8368 2013-08-20 David Malcolm <dmalcolm@redhat.com>
8369
8370 Make opt_pass and gcc::pass_manager be GC-managed, so that pass
8371 instances can own GC refs.
8372
8373 * Makefile.in (GTFILES): Add pass_manager.h and tree-pass.h.
8374 * context.c (gcc::context::gt_ggc_mx): Traverse passes_.
8375 (gcc::context::gt_pch_nx): Likewise.
8376 (gcc::context::gt_pch_nx): Likewise.
8377 * ggc.h (gt_ggc_mx <T>): New.
8378 (gt_pch_nx_with_op <T>): New.
8379 (gt_pch_nx <T>): New.
8380 * passes.c (opt_pass::gt_ggc_mx): New.
8381 (opt_pass::gt_pch_nx): New.
8382 (opt_pass::gt_pch_nx_with_op): New.
8383 (pass_manager::gt_ggc_mx): New.
8384 (pass_manager::gt_pch_nx): New.
8385 (pass_manager::gt_pch_nx_with_op): New.
8386 (pass_manager::operator new): Use
8387 ggc_internal_cleared_alloc_stat rather than xcalloc.
8388 * pass_manager.h (class pass_manager): Add GTY((user)) marking.
8389 (pass_manager::gt_ggc_mx): New.
8390 (pass_manager::gt_pch_nx): New.
8391 (pass_manager::gt_pch_nx_with_op): New.
8392 * tree-pass.h (class opt_pass): Add GTY((user)) marking.
8393 (opt_pass::operator new): New.
8394 (opt_pass::gt_ggc_mx): New.
8395 (opt_pass::gt_pch_nx): New.
8396 (opt_pass::gt_pch_nx_with_op): New.
8397
8398 Revert r201864:
8399 2013-08-20 David Malcolm <dmalcolm@redhat.com>
8400
8401 * Makefile.in (GTFILES): Add context.h.
8402 * context.c (gcc::context::operator new): New.
8403 (gcc::context::gt_ggc_mx): New.
8404 (gcc::context::gt_pch_nx): New.
8405 (gcc::context::gt_pch_nx): New.
8406 * context.h (gcc::context): Add GTY((user)) marking.
8407 (gcc::context::operator new): New.
8408 (gcc::context::gt_ggc_mx): New.
8409 (gcc::context::gt_pch_nx): New.
8410 (gcc::context::gt_pch_nx): New.
8411 (g): Add GTY marking.
8412 (gt_ggc_mx (gcc::context *)): New.
8413 (gt_pch_nx (gcc::context *)): New.
8414 (gt_pch_nx (gcc::context *ctxt, gt_pointer_operator op,
8415 void *cookie)): New.
8416 * gengtype.c (open_base_files) <ifiles>: Add context.h.
8417
a046136a
AM		 84182013-08-20 Alexey Makhalov <makhaloff@gmail.com>
8419
8420 * dce.c (fini_dce): Call df_analyze again just in case
8421 delete_unmarked_insns removed anything.
8422
67598720
TJ		 84232013-08-20 Teresa Johnson <tejohnson@google.com>
8424
8425 PR rtl-optimizations/57451
8426 * final.c (reemit_insn_block_notes): Prevent lexical blocks
8427 from crossing split section boundaries.
8428
f8693faf
MGD		 84292013-08-20 Matthew Gretton-Dann <matthew.gretton-dann@linaro.org>
8430
8431 * config/arm/linux-elf.h (MULTILIB_DEFAULTS): Remove definition.
8432 * config/arm/t-linux-eabi (MULTILIB_OPTIONS): Document association
8433 with MULTLIB_DEFAULTS.
8434
7d0b9a9c
NC		 84352013-08-20 Nick Clifton <nickc@redhat.com>
8436
8437 * target.def (narrow_volatile_bitfield): Note that the default
8438 value is false, not !TARGET_STRICT_ALIGN.
8439 * doc/tm.texi: Regenerate.
8440
8edb8dc8
PC		 84412013-08-20 Pavel Chupin <pavel.v.chupin@intel.com>
8442
8443 Fix LIB_SPEC for systems without libpthread.
8444
8445 * config/gnu-user.h: Introduce GNU_USER_TARGET_NO_PTHREADS_LIB_SPEC.
8446 * config/arm/linux-eabi.h: Use GNU_USER_TARGET_NO_PTHREADS_LIB_SPEC
8447 for Android.
8448 * config/i386/linux-common.h: Likewise.
8449 * config/mips/linux-common.h: Likewise.
8450
12211b99 84512013-08-20 Zhouyi Zhou <yizhouzhou@ict.ac.cn>
d71ff3fa
ZZ		 8452
8453 * tree-ssa-ccp.c (get_default_value): Remove redundant condition
8454 checks.
8455
e42c64cb
DM		 84562013-08-20 David Malcolm <dmalcolm@redhat.com>
8457
8458 Make opt_pass and gcc::pass_manager be GC-managed, so that pass
8459 instances can own GC refs.
8460
8461 * Makefile.in (GTFILES): Add pass_manager.h and tree-pass.h.
8462 * context.c (gcc::context::gt_ggc_mx): Traverse passes_.
8463 (gcc::context::gt_pch_nx): Likewise.
8464 (gcc::context::gt_pch_nx): Likewise.
8465 * ggc.h (gt_ggc_mx <T>): New.
8466 (gt_pch_nx_with_op <T>): New.
8467 (gt_pch_nx <T>): New.
8468 * passes.c (opt_pass::gt_ggc_mx): New.
8469 (opt_pass::gt_pch_nx): New.
8470 (opt_pass::gt_pch_nx_with_op): New.
8471 (pass_manager::gt_ggc_mx): New.
8472 (pass_manager::gt_pch_nx): New.
8473 (pass_manager::gt_pch_nx_with_op): New.
8474 (pass_manager::operator new): Use
8475 ggc_internal_cleared_alloc_stat rather than xcalloc.
8476 * pass_manager.h (class pass_manager): Add GTY((user)) marking.
8477 (pass_manager::gt_ggc_mx): New.
8478 (pass_manager::gt_pch_nx): New.
8479 (pass_manager::gt_pch_nx_with_op): New.
8480 * tree-pass.h (class opt_pass): Add GTY((user)) marking.
8481 (opt_pass::operator new): New.
8482 (opt_pass::gt_ggc_mx): New.
8483 (opt_pass::gt_pch_nx): New.
8484 (opt_pass::gt_pch_nx_with_op): New.
8485
910c02a0
DM		 84862013-08-20 David Malcolm <dmalcolm@redhat.com>
8487
8488 * Makefile.in (GTFILES): Add context.h.
8489 * context.c (gcc::context::operator new): New.
8490 (gcc::context::gt_ggc_mx): New.
8491 (gcc::context::gt_pch_nx): New.
8492 (gcc::context::gt_pch_nx): New.
8493 * context.h (gcc::context): Add GTY((user)) marking.
8494 (gcc::context::operator new): New.
8495 (gcc::context::gt_ggc_mx): New.
8496 (gcc::context::gt_pch_nx): New.
8497 (gcc::context::gt_pch_nx): New.
8498 (g): Add GTY marking.
8499 (gt_ggc_mx (gcc::context *)): New.
8500 (gt_pch_nx (gcc::context *)): New.
8501 (gt_pch_nx (gcc::context *ctxt, gt_pointer_operator op,
8502 void *cookie)): New.
8503 * gengtype.c (open_base_files) <ifiles>: Add context.h.
8504
af4b4236
AM		 85052013-08-20 Alan Modra <amodra@gmail.com>
8506
8507 PR target/57865
8508 * config/rs6000/rs6000.c (rs6000_emit_prologue): Correct ool_adjust.
8509 (rs6000_emit_epilogue): Likewise.
8510
12211b99 85112013-08-19 Dehao Chen <dehao@google.com>
0d633222
DC		 8512
8513 * value-prof.c (gimple_ic): Fix the bug of adding EH edge.
8514
e2323f5b
PB		 85152013-08-19 Peter Bergner <bergner@vnet.ibm.com>
8516 Jakub Jelinek <jakub@redhat.com>
8517
8518 * builtins.def (BUILT_IN_FABSD32): New DFP ABS builtin.
8519 (BUILT_IN_FABSD64): Likewise.
8520 (BUILT_IN_FABSD128): Likewise.
8521 * builtins.c (expand_builtin): Add support for new DFP ABS builtins.
8522 (fold_builtin_1): Likewise.
5339fc0c
UB		 8523 * config/rs6000/dfp.md (*negtd2_fpr): Handle non-overlapping
8524 destination and source operands.
e2323f5b
PB		 8525 (*abstd2_fpr): Likewise.
8526 (*nabstd2_fpr): Likewise.
8527
cfbf3ee8
RS		 85282013-08-19 Richard Sandiford <rdsandiford@googlemail.com>
8529
8530 * config/mips/mips.c (mips_adjust_insn_length): Add checks for
8531 JUMP_P and INSN_P.
8532
d3136aeb
AH		 85332013-08-19 Aldy Hernandez <aldyh@redhat.com>
8534
8535 * doc/invoke.texi (-fcilkplus): Clarify that implementation is
8536 incomplete.
8537
d33d9e47
AI		 85382013-08-19 Alexander Ivchenko <alexander.ivchenko@intel.com>
8539
8540 * target.def (TARGET_LIBC_HAS_FUNCTION): New target hook.
8541 * builtins.c (default_libc_has_function): New.
8542 (gnu_libc_has_function): Ditto.
8543 (no_c99_libc_has_function): Ditto.
8544 (expand_builtin_cexpi): Using new target hook TARGET_LIBC_HAS_FUNCTION
8545 instead of TARGET_HAS_SINCOS and TARGET_C99_FUNCTIONS.
8546 (fold_builtin_sincos): Likewise.
8547 (fold_builtin_cexp): Likewise.
8548 * builtins.def (DEF_C94_BUILTIN): Likewise.
8549 (DEF_C99_BUILTIN): Likewise.
8550 (DEF_C99_C90RES_BUILTIN): Likewise.
8551 (DEF_C99_COMPL_BUILTIN): New define. Change all complex c99 builtin
8552 definitions to using this define.
8553 * config/darwin-protos.h (darwin_libc_has_function): New.
5339fc0c 8554 * config/darwin.c (darwin_libc_has_function): Ditto.
d33d9e47
AI		 8555 * config/alpha/linux.h: Remove TARGET_C99_FUNCTIONS and
8556 TARGET_HAS_SINCOS. Redefine TARGET_LIBC_HAS_FUNCTION.
8557 * config/darwin.h: Ditto.
8558 * config/elfos.h: Ditto.
8559 * config/freebsd.h: Ditto.
8560 * config/i386/cygming.h: Ditto.
8561 * config/i386/djgpp.h: Ditto.
8562 * config/i386/i386-interix.h: Ditto.
8563 * config/microblaze/microblaze.h: Ditto.
8564 * config/mmix/mmix.h: Ditto.
8565 * config/gnu-user.h: Ditto.
8566 * config/ia64/hpux.h: Ditto.
8567 * config/pa/pa-hpux.h: Ditto.
8568 * config/pdp11/pdp11.h: Ditto.
8569 * config/picochip/picochip.h: Ditto.
8570 * config/linux.h: Ditto.
8571 * config/netbsd.h: Ditto.
8572 * config/openbsd.h: Ditto.
8573 * config/rs6000/aix43.h: Ditto.
8574 * config/rs6000/aix51.h: Ditto.
8575 * config/rs6000/aix52.h: Ditto.
8576 * config/rs6000/aix53.h: Ditto.
8577 * config/rs6000/aix61.h: Ditto.
8578 * config/rs6000/darwin.h: Ditto.
8579 * config/rs6000/linux.h: Ditto.
8580 * config/rs6000/linux64.h: Ditto.
8581 * config/s390/tpf.h: Ditto.
8582 * config/sol2-10.h: Ditto.
8583 * config/sol2.h: Ditto.
8584 * config/vms/vms.h: Ditto.
8585 * config/vxworks.h: Ditto.
8586 * config/linux-android.c (linux_android_libc_has_function):
8587 New linux-specific implementation of TARGET_LIBC_HAS_FUNCTION.
8588 * config/linux-protos.h (linux_android_libc_has_function):
8589 New declaration.
8590 * config/i386/i386.c (ix86_libc_has_function): New.
8591 * config/i386/i386-protos.h
8592 (ix86_libc_has_function): New declaration.
8593 * config/i386/i386.md
8594 ("isinfxf2"): Change condition for TARGET_LIBC_HAS_FUNCTION.
8595 ("isinf<mode>2): Likewise.
8596 * convert.c (convert_to_integer): Using new target hook
8597 TARGET_LIBC_HAS_FUNCTION istead of TARGET_HAS_SINCOS and
8598 TARGET_C99_FUNCTIONS.
8599 * fortran/f95-lang.c (gfc_init_builtin_functions): Ditto.
8600 * tree-ssa-math-opts.c (execute_cse_sincos): Ditto.
8601 * coretypes.h (function_class): New enum for different
8602 classes of functions.
8603 * defaults.h: Remove TARGET_C99_FUNCTIONS and TARGET_HAS_SINCOS.
8604 * doc/tm.texi.in (TARGET_C99_FUNCTIONS): Remove documentation.
8605 (TARGET_HAS_SINCOS): Likewise.
8606 (TARGET_LIBC_HAS_FUNCTION): New.
8607 * doc/tm.texi: Regenerated.
8608 * targhooks.h (default_libc_has_function): New declaration.
8609 (no_c99_libc_has_function): Ditto.
8610 (gnu_libc_has_function): Ditto.
8611 * system.h: Add the poisoning of TARGET_C99_FUNCTIONS
8612 and TARGET_HAS_SINCOS.
8613
eefe9a99
JH		 86142013-08-18 Jan Hubicka <jh@suse.cz>
8615
8616 * Makeifle-in (ipa-devirt.o): New.
8617 (GTFILES): Add ipa-utils.h and ipa-devirt.c
8618 * cgraphunit.c (decide_is_symbol_needed): Do not care about virtuals.
8619 (analyze_functions): Look into possible targets of polymorphic call.
8620 * dumpfile.c (dump_files): Add type-inheritance dump.
8621 * dumpfile.h (TDI_inheritance): New.
8622 * ipa-devirt.c: New file.
8623 * ipa-utils.h (odr_type_d): Forward declare.
8624 (odr_type): New type.
8625 (build_type_inheritance_graph): Declare.
8626 (possible_polymorphic_call_targets): Declare and introduce inline
8627 variant when only edge is pased.
8628 (dump_possible_polymorphic_call_targets): Likewise.
8629 * timevar.def (TV_IPA_INHERITANCE, TV_IPA_VIRTUAL_CALL): New.
8630 * tree.c (type_in_anonymous_namespace_p): Break out from ...
8631 (types_same_for_odr): ... here.
8632 * tree.h (type_in_anonymous_namespace_p): Declare.
8633
29b89442
JJ		 86342013-08-18 Jakub Jelinek <jakub@redhat.com>
8635
8636 PR tree-optimization/58006
8637 * tree-parloops.c (take_address_of): Don't ICE if get_name
8638 returns NULL.
8639 (eliminate_local_variables_stmt): Remove clobber stmts.
8640
4f219961
EB		 86412013-08-18 Eric Botcazou <ebotcazou@adacore.com>
8642
8643 * cgraphunit.c (handle_alias_pairs): Reset the alias flag after the
8644 error message is issued for an alias to undefined symbol.
8645
1d5755ef
JH		 86462013-08-18 Jan Hubicka <jh@suse.cz>
8647
8648 * cgraph.c (cgraph_create_indirect_edge): Discover
8649 polymorphic calls and record basic info into indirect_info.
8650 * gimple-fold.c (gimple_fold_call): When doing BINFO based
8651 devirtualization, ignore objc function calls.
8652 * ipa-cp.c (initialize_node_lattices): Be ready for polymorphic
8653 call with no parm index info.
8654 * ipa-prop.c (ipa_analyze_call_uses): Likewise.
8655 * tree.c (virtual_method_call_p): New function.
8656 * tree.h (virtual_method_call_p): Declare.
8657
5945bebf
JH		 86582013-08-16 Jan Hubicka <jh@suse.cz>
8659
8660 PR middle-end/58179
8661 * tree.c (obj_type_ref_class): Do not ICE on non-method calls.
8662
47fa5d23
DE		 86632013-08-16 David Edelsohn <dje.gcc@gmail.com>
8664
8665 * config/rs6000/rs6000.md (rs6000_get_timebase_ppc32): Add length
8666 attribute.
8667
c74f54a0
DM		 86682013-08-16 David Malcolm <dmalcolm@redhat.com>
8669
8670 * gengtype.c (type_for_name): Add special-case support for
8671 locating types within the "gcc::" namespace.
8672 (open_base_files): Emit a "using namespace gcc" directive.
8673
f758f299
MM		 86742013-08-16 Michael Meissner <meissner@linux.vnet.ibm.com>
8675
8676 PR target/58160
8677 * config/rs6000/predicates.md (fusion_gpr_mem_load): Allow the
8678 memory rtx to contain ZERO_EXTEND and SIGN_EXTEND.
8679
8680 * config/rs6000/rs6000-protos.h (fusion_gpr_load_p): Pass operands
8681 array instead of each individual operand as a separate argument.
8682 (emit_fusion_gpr_load): Likewise.
8683 (expand_fusion_gpr_load): Add new function declaration.
8684
8685 * config/rs6000/rs6000.c (fusion_gpr_load_p): Change the calling
8686 signature to have the operands passed as an array, instead of as
8687 separate arguments. Allow ZERO_EXTEND to be in the memory
8688 address, and also SIGN_EXTEND if -mpower8-fusion-sign. Do not
8689 depend on the register live/dead flags when peepholes are run.
8690 (expand_fusion_gpr_load): New function to be called from the
8691 peephole2 pass, to change the register that addis sets to be the
8692 target register.
8693 (emit_fusion_gpr_load): Change the calling signature to have the
8694 operands passed as an array, instead of as separate arguments.
8695 Allow ZERO_EXTEND to be in the memory address, and also
8696 SIGN_EXTEND if -mpower8-fusion-sign.
8697
8698 * config/rs6000/rs6000.md (UNSPEC_FUSION_GPR): Delete unused
8699 unspec enumeration.
8700 (power8 fusion peephole/peephole2): Rework the fusion peepholes to
8701 adjust the register addis loads up in the peephole2 pass. Do not
5339fc0c 8702 depend on the register live/dead state when the peephole pass is done.
f758f299 8703
158f4e4f
DM		 87042013-08-16 David Malcolm <dmalcolm@redhat.com>
8705
8706 * gengtype.c (create_user_defined_type): Ensure that the kind
8707 is set to TYPE_USER_STRUCT, fixing a bug seen when an incomplete
8708 declaration is seen before the GTY((user)) marking.
8709
22f8cea5
BE		 87102013-08-16 Bernd Edlinger <bernd.edlinger@hotmail.de>
8711
8712 PR target/58105
8713 * config/i386/i386.c (make_resolver_func): Set DECL_UNINLINABLE.
8714
c49bdb2e
JH		 87152013-08-16 Jan Hubicka <jh@suse.cz>
8716
8717 * gimple-fold.c (gimple_extract_devirt_binfo_from_cst): Add new
8718 arugment expected_type.
8719 (gimple_fold_call): Use it.
8720 * gimple.h (gimple_extract_devirt_binfo_from_cst): Update prototype.
8721 * ipa-cp.c (ipa_get_indirect_edge_target_1): Update.
22f8cea5 8722 * ipa-prop.c (ipa_analyze_virtual_call_uses): Use obj_type_ref_class.
c49bdb2e
JH		 8723 (try_make_edge_direct_virtual_call): Likewise.
8724 * tree.c (obj_type_ref_class): New.
8725 * tree.h (obj_type_ref_class): Use it.
8726
4042dca9
GDR		 87272013-08-16 Gabriel Dos Reis <gdr@integrable-solutions.net>
8728
8729 * sched-vis.c (rtl_slim_pp_initialized): Remove.
8730 (rtl_slim_pp): Likewise.
8731 (init_rtl_slim_pretty_print): Likewise.
8732 (dump_value_slim): Don't call it. Use local pretty printer.
8733 (dump_insn_slim): Likewise.
8734 (dump_rtl_slim): Likewise.
8735 (str_pattern_slim): Likewise.
8736 * tree-mudflap.c (mf_varname_tree): Use local pretty printer.
8737 Simplify.
8738
7a460594
JJ		 87392013-08-16 Jakub Jelinek <jakub@redhat.com>
8740
639dc669
JJ		 8741 PR tree-optimization/58164
8742 * gimple.c (walk_stmt_load_store_addr_ops): For visit_addr
8743 walk gimple_goto_dest of GIMPLE_GOTO.
8744
7a460594
JJ		 8745 PR tree-optimization/58165
8746 * tree-call-cdce.c (shrink_wrap_one_built_in_call): If
8747 bi_call must be the last stmt in a bb, don't split_block, instead
8748 use fallthru edge from it and give up if there is none.
8749 Release conds vector when returning early.
8750
e3c02cc7
XDL		 87512013-08-14 Xinliang David Li <davidxl@google.com>
8752
8753 * config/i386/i386.c (ix86_option_override_internal):
8754 Remove unused variable and field.
8755
6469da2c
BS		 87562013-08-14 Bill Schmidt <wschmidt@linux.vnet.ibm.com>
8757
8758 PR target/57949
22f8cea5 8759 * doc/invoke.texi: Add documentation of mcompat-align-parm option.
6469da2c
BS		 8760 * config/rs6000/rs6000.opt: Add mcompat-align-parm option.
8761 * config/rs6000/rs6000.c (rs6000_function_arg_boundary): For AIX
8762 and Linux, correct BLKmode alignment when 128-bit alignment is
8763 required and compatibility flag is not set.
22f8cea5
BE		 8764 (rs6000_gimplify_va_arg): For AIX and Linux, honor specified alignment
8765 for zero-size arguments when compatibility flag is not set.
6469da2c 8766
4ca890e2
JJ		 87672013-08-14 Jakub Jelinek <jakub@redhat.com>
8768
8769 PR tree-optimization/58145
8770 * tree-sra.c (build_ref_for_offset): If prev_base has
8771 TREE_THIS_VOLATILE or TREE_SIDE_EFFECTS, propagate it to MEM_REF.
8772
daec1759 87732013-08-14 Xinliang David Li <davidxl@google.com>
4ca890e2 8774
daec1759 8775 * config/i386/i386.c (ix86_option_override_internal):
e3c02cc7 8776 Fix uninitialized variable error.
daec1759 8777
3ad20bd4
XDL		 87782013-08-14 Xinliang David Li <davidxl@google.com>
8779
8780 * config/i386/i386.opt: Define two new options.
8781 * config/i386/x86-tune.def: Add arch selector field in macros.
8782 * config/i386/i386.h: Adjust macro definition.
8783 * config/i386/i386.c (ix86_option_override_internal):
8784 Refactor the code.
8785 (parse_mtune_ctrl_str): New function.
8786 (set_ix86_tune_features): New function.
8787 (ix86_function_specific_restore): Call the new helper function.
8788
e839e2a9
AB		 87892013-08-14 Andrey Belevantsev <abel@ispras.ru>
8790
8791 PR rtl-optimization/57662
8792 * sel-sched.c (code_motion_process_successors): When the current insn
8793 is removed after the recursive traversal, break from the loop.
8794 Add comments and debug printouts.
8795
27ed665f
JJ		 87962013-08-14 Jakub Jelinek <jakub@redhat.com>
8797 Alexandre Oliva <aoliva@redhat.com>
8798
8799 PR target/58067
8800 * config/i386/i386.c (ix86_delegitimize_address): For CM_MEDIUM_PIC
8801 and CM_LARGE_PIC ix86_cmodel fall thru into the -m32 code, handle
8802 there also UNSPEC_PLTOFF.
8803
61c1a609
MP		 88042013-08-14 Marek Polacek <polacek@redhat.com>
8805
8806 * ipa-inline-analysis.c (add_clause): Avoid shifting integer
8807 NUM_CONDITIONS bit positions.
8808
0e901c67
CC		 88092013-08-13 Cary Coutant <ccoutant@google.com>
8810
8811 * dwarf2out.c (CHECKSUM_BLOCK): New macro.
8812 (attr_checksum): Hash vector contents instead of pointer.
8813 (attr_checksum_ordered): Likewise.
8814
061eff6d
UB		 88152013-08-13 Uros Bizjak <ubizjak@gmail.com>
8816
8817 * config/i386/sse.md (*sse2_maskmovdqu): Emit addr32 prefix
8818 when Pmode != word_mode. Add length_address attribute.
8819 (sse3_monitor_<mode>): Merge from sse3_monitor and
8820 sse3_monitor64_<mode> insn patterns. Emit addr32 prefix when
8821 Pmode != word_mode. Update insn length attribute.
8822 * config/i386/i386.c (ix86_option_override_internal): Update
8823 ix86_gen_monitor selection for merged sse3_monitor insn.
8824
8ec9249b
JB		 88252013-08-13 Julian Brown <julian@codesourcery.com>
8826
8827 * config/rs6000/rs6000.c (rs6000_legitimize_reload_address): Don't
8828 perform invalid legitimization on greater-than-word-size modes for
8829 TARGET_E500_DOUBLE.
8830
761a8eb7
VM		 88312013-08-13 Vladimir Makarov <vmakarov@redhat.com>
8832
8833 * ira.c (setup_class_translate_array): Use aclass instead of cl
8834 for classes not fully covered by allocno classes.
8835
99c2bd54
JJ		 88362013-08-13 Jakub Jelinek <jakub@redhat.com>
8837
4029a5e0
JJ		 8838 PR tree-optimization/57661
8839 * tree-inline.h (struct copy_body_data): Add blocks_to_copy field.
8840 * tree-inline.c (tree_function_versioning): Initialize it.
8841 (remap_gimple_stmt): Return GIMPLE_NOP for MEM_REF lhs clobber stmts
8842 if id->blocks_to_copy and MEM_REF's SSA_NAME is defined in a block
8843 that is not being copied.
8844
99c2bd54
JJ		 8845 PR sanitizer/56417
8846 * asan.c (instrument_strlen_call): Fix typo in comment.
8847 Use char * type even for the lhs of POINTER_PLUS_EXPR.
8848
b5f8f063
SE		 88492013-08-13 Steve Ellcey <sellcey@mips.com>
8850
8851 * config/mips/mips.md (prefetch): Use lw instead of ld on
8852 loongson in 32bit mode.
8853
4de80584
NC		 88542013-08-13 Nick Clifton <nickc@redhat.com>
8855
8856 * config.gcc: (avr-linux): Allow for tmake_file not being empty.
8857
09ce3660
JH		 88582013-08-13 Jan Hubicka <jh@suse.cz>
8859
8860 * cgraph.c (cgraph_turn_edge_to_speculative): Return newly
8861 introduced edge; fix typo in sanity check.
8862 (cgraph_resolve_speculation): Export; improve diagnostic.
8863 (cgraph_redirect_edge_call_stmt_to_callee): Better diagnostic; cancel
8864 speculation at type mismatch.
8865 * cgraph.h (cgraph_turn_edge_to_speculative): Update.
8866 (cgraph_resolve_speculation): Declare.
8867 (symtab_can_be_discarded): New function.
8868 * value-prof.c (gimple_ic_transform): Remove actual transform code.
8869 * ipa-inline-transform.c (speculation_removed): New global var.
8870 (clone_inlined_nodes): See if speculation can be removed.
8871 (inline_call): If speculations was removed, we growths may not match.
8872 * ipa-inline.c (can_inline_edge_p): Add DISREGARD_LIMITS parameter.
8873 (speculation_useful_p): New function.
8874 (resolve_noninline_speculation): New function.
8875 (inline_small_functions): Resolve useless speculations.
8876 * ipa-inline.h (speculation_useful_p): Declare
8877 * ipa.c (can_replace_by_local_alias): Simplify.
8878 (ipa_profile): Produce speculative calls in non-lto, too;
8879 add simple cost model; produce local aliases.
8880
f971dc24
DM		 88812013-08-13 David Malcolm <dmalcolm@redhat.com>
8882
8883 * config/i386/t-i386 (i386.o): Rename stray PIPELINE_H to
061eff6d 8884 PASS_MANAGER_H.
f971dc24 8885
2e507b9a
PC		 88862013-08-12 Paolo Carlini <paolo.carlini@oracle.com>
8887
8888 * config/i386/i386.c (ix86_function_versions): Use error + inform.
8889
c97fd4b8
UB		 88902013-08-12 Uros Bizjak <ubizjak@gmail.com>
8891
8892 * config/i386/i386.md (floatunssi<mode>2 expand): Use MODEF mode
8893 iterator instead of X87MODEF.
8894
88952013-08-12 Perez Read <netfirewall@gmail.com>
8896
8897 PR target/58132
8898 * config/i386/i386.md (*movabs<mode>_1): Add <ptrsize> PTR before
8899 operand 0 for intel asm alternative.
8900 (*movabs<mode>_2): Ditto for operand 1.
8901
f4f4204c
JG		 89022013-08-12 James Greenhalgh <james.greenhalgh@arm.com>
8903
8904 * config/aarch64/arm_none.h
8905 (vdup<bhsd>_lane_<su><8,16,32,64>): Fix macro call.
8906
133b1a8e
NC		 89072013-08-12 Nick Clifton <nickc@redhat.com>
8908
8909 * config.gcc (m32r-linux): Allow for tmake_file not being empty.
8910
15dd8b3a
YR		 89112013-08-12 Yuri Rumyantsev <ysrumyan@gmail.com>
8912
8913 * config/i386/i386.md (floatunssi<mode>2 expand): Add new
8914 expand for QI/HImode operand to produce more effictive code for
8915 unsigned char(short) --> float(double) conversion.
8916
94109a6a
AM		 89172013-08-12 Alexander Monakov <amonakov@ispras.ru>
8918
8919 * doc/invoke.texi: Mention that -ftls-model does not force the final
8920 model.
8921
4e115102
MP		 89222013-08-12 Marek Polacek <polacek@redhat.com>
8923 Marc Glisse <marc.glisse@inria.fr>
8924
8925 PR tree-optimization/57980
8926 * tree-tailcall.c (process_assignment): Call build_minus_one_cst
8927 when creating -1 constant.
8928
748d71f3
JH		 89292013-08-10 Jan Hubicka <jh@suse.cz>
8930
1f97e193 8931 Workaround binutils PR14342.
748d71f3
JH		 8932 * tree-profile.c (init_ic_make_global_vars): Add LTO path.
8933 (gimple_init_edge_profiler): Likewise.
8934 (gimple_gen_ic_func_profiler): Likewise.
8935
5979aa54
JH		 89362013-08-09 Jan Hubicka <jh@suse.cz>
8937
8938 * cgraph.c (cgraph_create_edge_1): Clear speculative flag.
8939
ad83025e
XDL		 89402013-08-09 Xinliang David Li <davidxl@google.com>
8941
8942 * config/i386/stringop.def: New file.
8943 * config/i386/stringop.opt: New file.
8944 * config/i386/i386-opts.h: Include stringopt.def.
8945 * config/i386/i386.opt: Include stringopt.opt.
8946 * config/i386/i386.c (ix86_option_override_internal):
c97fd4b8 8947 Override default size based stringop inline strategies with options.
ad83025e
XDL		 8948 * config/i386/i386.c (ix86_parse_stringop_strategy_string):
8949 New function.
8950
e59df5fd
JH		 89512013-08-09 Jan Hubicka <jh@suse.cz>
8952
c97fd4b8 8953 * ipa-ref.c (ipa_clear_stmts_in_references): Clear lto_stmt_uid, too.
e59df5fd 8954
634ab819
JH		 89552013-08-09 Jan Hubicka <jh@suse.cz>
8956
8957 * cgraph.c (cgraph_resolve_speculation): Cut frequency to
8958 CGRAPH_FREQ_MAX.
8959 (dump_cgraph_node): Dump profile-id.
8960 * cgraph.h (cgraph_indirect_call_info): Add common_target_id
8961 and common_target_probability.
8962 * lto-cgraph.c (lto_output_edge): Stream common targets.
8963 (lto_output_node): Stream profile ids.
8964 (input_node): Stream profile ids.
8965 (input_edge): Stream common targets.
8966 * lto-streamer-in.c (fixup_call_stmt_edges_1): Fix formatting.
8967 * ipa.c: Include value-prof.h
8968 (ipa_profile_generate_summary): Turn indirect call statement histograms
8969 into common targets.
8970 (ipa_profile): Turn common targets into speculative edges.
8971
2fa3d31b
JH		 89722013-08-09 Jan Hubicka <jh@suse.cz>
8973
8974 * cgraph.h (cgraph_node): Add profile_id.
8975 * value-prof.c (cgraph_node_map): Turn into pointer_map.
c97fd4b8 8976 (init_node_map): Rewrite to handle hashes increas of incremental IDs.
2fa3d31b
JH		 8977 (del_node_map): Update.
8978 (find_func_by_funcdef_no): Replace by ...
8979 (find_func_by_profile_id): ... this one.
8980 (gimple_ic_transform): Do not remove useful histograms when
8981 speculation is not done; dump info when indirect call removal
8982 can happen at LTO.
8983 * value-prof.h (find_func_by_profile_id, gimple_ic): Declare.
8984 * gcov-io.h (__gcov_indirect_call_profiler): Replace by ...
8985 (__gcov_indirect_call_profiler_v2): .. this one.
8986 * profile.h (init_node_map): Update.
8987 * coverage.c (coverage_compute_profile_id): New function.
8988 * coverage.h (coverage_compute_profile_id): Declare.
8989 * tree-profile.c (init_ic_make_global_vars): Make
8990 __gcov_indirect_call_callee and __gcov_indirect_call_counters global.
8991 (gimple_init_edge_profiler): Update prototype of
8992 __gcov_indirect_call_profiler.
8993 (gimple_gen_ic_func_profiler): Simplify.
8994 (tree_profiling): Use init_node_map
8995
042ae7d2
JH		 89962013-08-09 Jan Hubicka <jh@suse.cz>
8997
c97fd4b8
UB		 8998 * cgraphbuild.c (cgraph_rebuild_references): Rebuild only
8999 non-speculative refs.
042ae7d2
JH		 9000 * cgraph.c (cgraph_update_edge_in_call_site_hash): New function.
9001 (cgraph_add_edge_to_call_site_hash): Deal with speculative calls.
9002 (cgraph_set_call_stmt): Likewise.
9003 (cgraph_create_edge_1): Fix release checking compilatoin;
9004 clear lto_stmt_uid.
9005 (cgraph_free_edge): Free indirect info.
9006 (cgraph_turn_edge_to_speculative): New function.
9007 (cgraph_speculative_call_info): New function.
9008 (cgraph_make_edge_direct): Return direct edge; handle speculation.
c97fd4b8 9009 (cgraph_redirect_edge_call_stmt_to_callee): Expand speculative edges.
042ae7d2
JH		 9010 (dump_cgraph_node): Dump speculation.
9011 (verify_edge_count_and_frequency): Accept speculative edges.
9012 (verify_edge_corresponds_to_fndecl): Handle partitioned cgraph.
9013 (verify_cgraph_node): Handle speculation.
9014 * cgraph.h (cgraph_edge): Add SPECULATIVE flag.
9015 (cgraph_set_call_stmt): Update prototype.
9016 (cgraph_make_edge_direct): Update prototype.
9017 (cgraph_speculative_call_info): Declare.
9018 * ipa-cp.c (ipcp_discover_new_direct_edges): Be ready for edge
9019 to change; update call of ipa_find_references.
9020 * ipa-ref.c (ipa_record_reference): Fix return value; clear
9021 lto_stmt_uid and speculative flags.
9022 (ipa_dump_references): Dump speculation.
9023 (ipa_clone_references): Clone speculative flag.
9024 (ipa_clone_referring): Likewise.
9025 (ipa_clone_ref): New function.
9026 (ipa_find_reference): Look into lto_stmt_uids
9027 (ipa_clear_stmts_in_references): Do not clear speculative calls.
9028 * ipa-ref.h (ipa_ref): Add lto_stmt_uid and speculative flags.
9029 (ipa_find_reference): Update declaration.
9030 (ipa_clone_ref): Declare.
9031 * lto-cgraph.c (lto_output_edge): Make lto_stmt_uids start from 0;
9032 stream speculative flag.
9033 (lto_output_ref): Stream statements uids and speculation.
9034 (input_ref): Likewise.
9035 (input_edge): Stream speuclation.
9036 * cgraphclones.c (cgraph_clone_edge): Clone speculation.
9037 (cgraph_set_call_stmt_including_clones): Handle speculation.
9038 * ipa-inline.c (heap_edge_removal_hook): New function.
9039 (inline_small_functions): Register it.
9040 * lto-streamer-in.c (fixup_call_stmt_edges_1): Bounds checking;
9041 also initialize refs.
9042 * ipa-prop.c (ipa_make_edge_direct_to_target): Be ready for
9043 edge to change.
9044 (try_make_edge_direct_simple_call): Likewise.
9045 (try_make_edge_direct_simple_call): Likewise.
9046 (update_indirect_edges_after_inlining): Likewise.
9047 (remove_described_reference): Look proper lto_stmt_uid.
9048 (propagate_controlled_uses): Likewise.
9049 (propagate_controlled_uses): Liekwise.
9050 * tree-inline.c (copy_bb): Copy speculative edges.
9051 (redirect_all_calls): New function.
c97fd4b8
UB		 9052 (copy_cfg_body): Do redirection after loop info is updated.
9053 (delete_unreachable_blocks_update_callgraph): Updadte speculation.
042ae7d2 9054
aa1e10cc
JH		 90552013-08-09 Jan Hubicka <jh@suse.cz>
9056
9057 * lto-streamer-out.c (output_function): Renumber PHIs.
9058 * lto-streamer-in.c (input_function): Likewise.
9059
66adb8eb
JG		 90602013-08-09 James Greenhalgh <james.greenhalgh@arm.com>
9061
9062 * config/aarch64/aarch64-simd-builtins.def (get_lane_signed): Remove.
9063 (get_lane_unsigned): Likewise.
9064 (dup_lane_scalar): Likewise.
9065 (get_lane): enable for VALL.
9066 * config/aarch64/aarch64-simd.md
9067 (aarch64_dup_lane_scalar<mode>): Remove.
9068 (aarch64_get_lane_signed<mode>): Likewise.
9069 (aarch64_get_lane_unsigned<mode>): Likewise.
9070 (aarch64_get_lane_extend<GPI:mode><VDQQH:mode>): New.
9071 (aarch64_get_lane_zero_extendsi<mode>): Likewise.
9072 (aarch64_get_lane<mode>): Enable for all vector modes.
9073 (aarch64_get_lanedi): Remove misleading constraints.
9074 * config/aarch64/arm_neon.h
9075 (__aarch64_vget_lane_any): Define.
9076 (__aarch64_vget<q>_lane_<fpsu><8,16,32,64>): Likewise.
9077 (vget<q>_lane_<fpsu><8,16,32,64>): Use __aarch64_vget_lane macros.
9078 (vdup<bhsd>_lane_<su><8,16,32,64>): Likewise.
9079 * config/aarch64/iterators.md (VDQQH): New.
9080 (VDQQHS): Likewise.
9081 (vwcore): Likewise.
9082
1f65ae7a
EB		 90832013-08-09 Eric Botcazou <ebotcazou@adacore.com>
9084
9085 * configure.ac: Add GAS check for LEON instructions on SPARC.
9086 * configure: Regenerate.
9087 * config.in: Likewise.
9088 * config.gcc (with_cpu): Remove sparc-leon*-* and deal with LEON in the
9089 sparc*-*-* block.
9090 * config/sparc/sparc.opt (LEON, LEON3): New masks.
9091 * config/sparc/sparc.h (ASM_CPU32_DEFAULT_SPEC): Set to AS_LEON_FLAG
9092 for LEON or LEON3.
9093 (ASM_CPU_SPEC): Pass AS_LEON_FLAG if -mcpu=leon or -mcpu=leon3.
9094 (AS_LEON_FLAG): New macro.
9095 * config/sparc/sparc.c (sparc_option_override): Set MASK_LEON for leon
9096 and MASK_LEON3 for leon3 and unset them if HAVE_AS_LEON is not defined.
9097 Deal with LEON and LEON3 for the memory model.
23a6cb78 9098 * config/sparc/sync.md (atomic_compare_and_swap<mode>): Enable if LEON3
1f65ae7a
EB		 9099 (atomic_compare_and_swap<mode>_1): Likewise.
9100 (*atomic_compare_and_swap<mode>_1): Likewise.
9101
ff522f7f
ZC		 91022013-08-09 Zhenqiang Chen <zhenqiang.chen@linaro.org>
9103
9104 * config/arm/neon.md (vcond): Fix floating-point vector
9105 comparisons against 0.
9106
cb1cca12
VM		 91072013-08-08 Vladimir Makarov <vmakarov@redhat.com>
9108
9109 * lra-constraints.c (emit_spill_move): Remove assert.
9110 (process_alt_operands): Add more debugging
9111 output. Increase reject for spilling into memory. Decrease
9112 reject for reloading scratch.
9113 (split_reg): Use HARD_REGNO_CALLER_SAVE_MODE.
9114
a3719e31
SE		 91152013-08-08 Steve Ellcey <sellcey@mips.com>
9116
9117 * config/mips/mti-linux.h (SYSROOT_SUFFIX_SPEC): Add nan2008.
9118 * config/mips/t-mti-elf (MULTILIB_OPTIONS): Make mips16 and
9119 micromips incompatible. Add nan2008.
9120 (MULTILIB_DIRNAMES): Add nan2008.
9121 (MULTILIB_EXCEPTIONS): Remove mips16/micromips entry.
9122 * config/mips/t-mti-linux (MULTILIB_OPTIONS): Make mips16
9123 and micromips incompatible. Add nan2008.
9124 (MULTILIB_DIRNAMES): Add nan2008.
9125 (MULTILIB_EXCEPTIONS): Remove mips16/micromips entry.
9126
f40423e2
RS		 91272013-08-08 Richard Sandiford <rdsandiford@googlemail.com>
9128
9129 PR rtl-optimization/58079
9130 * combine.c (combine_simplify_rtx): Avoid using SUBST if
9131 simplify_comparison has widened a comparison with an integer.
9132
869b9125
KT		 91332013-08-08 Kyrylo Tkachov <kyrylo.tkachov@arm.com>
9134
9135 * config/arm/neon.md (movmisalign<mode>): Disable when we
9136 don't allow unaligned accesses.
9137 (*movmisalign<mode>_neon_store): Likewise.
9138 (*movmisalign<mode>_neon_load): Likewise.
9139 (*movmisalign<mode>_neon_store): Likewise.
9140 (*movmisalign<mode>_neon_load): Likewise.
9141
71cafea9
JH		 91422013-08-08 Jan Hubicka <jh@suse.cz>
9143
9144 * cgraphbuild.c (build_cgraph_edges): Do not walk into debugs.
9145 (make_pass_rebuild_cgraph_edges): Also clear references.
9146 * cgraph.c (verify_cgraph_node): Add basic ipa-ref verifier.
9147 * ipa-inline-transform.c (inline_transform): Remove all references
9148 after inlining.
c97fd4b8
UB		 9149 * cgraphunit.c (expand_function): Remove all references after
9150 expansion.
71cafea9
JH		 9151 * ipa-ref.c (ipa_ref_has_aliases_p): Fix formatting.
9152 (ipa_find_reference): Rewrite to iterator.
9153 (remove_stmt_references): Likewise.
9154 (ipa_clear_stmts_in_references): New function.
9155 * ipa-ref.h (ipa_clear_stmts_in_references): Declare.
03b0ee0a
UB		 9156 * cgraphclones.c (cgraph_materialize_all_clones): Remove or
9157 clear references.
71cafea9
JH		 9158 * ipa-split.c (split_function): Remove references in split function.
9159
27d2e612
RE		 91602013-08-08 Richard Earnshaw <rearnsha@arm.com>
9161
9162 PR target/57431
1f65ae7a 9163 * config/arm/arm/neon.md (neon_vld1_dupdi): New expand pattern.
27d2e612
RE		 9164 (neon_vld1_dup<mode> VD iterator): Iterate over VD not VDX.
9165
5efc661c
RE		 91662013-08-08 Richard Earnshaw <rearnsha@arm.com>
9167
9168 PR target/56979
1f65ae7a 9169 * config/arm/arm.c (aapcs_vfp_allocate): Decompose the argument if the
5efc661c
RE		 9170 suggested mode for the assignment isn't compatible with the
9171 registers required.
9172
f276d31d
BE		 91732013-08-08 Bernd Edlinger <bernd.edlinger@hotmail.de>
9174
9175 PR target/58065
9176 * config/arm/arm.h (MALLOC_ABI_ALIGNMENT): Define.
9177
4b8bc035
XDL		 91782013-08-07 Xinliang David Li <davidxl@google.com>
9179
9180 * config/i386/i386.opt: New option -mtune-ctrl=.
9181 * config/i386/x86-tune.def: New file.
9182 * config/i386/i386.h: include x86-tune.def.
9183 * config/i386/i386.c (ix86_option_override_internal):
9184 Parsing -mtune-ctrl= option and set tune features.
9185
12211b99 91862013-08-07 Oleg Endo <olegendo@gcc.gnu.org>
a5b38740
OE		 9187
9188 PR other/12081
03b0ee0a 9189 * config/rs6000/rs6000.c (gen_2arg_fn_t): Remove typedef.
a5b38740
OE		 9190 (rs6000_emit_swdiv, rs6000_emit_swrsqrt): Don't cast result of GEN_FCN
9191 to gen_2arg_fn_t.
9192
5cf6635b
EB		 91932013-08-07 Eric Botcazou <ebotcazou@adacore.com>
9194
9195 * rtl.h (update_alignments): Declare.
9196 * final.c (grow_label_align): New function extracted from...
9197 (shorten_branches): ...here. Call it.
9198 (update_alignments): New function.
9199 * reorg.c (sibling_labels): New variable.
9200 (get_label_before): Add SIBLING parameter. If it is non-zero, push
9201 the new label along with it onto the sibling_labels vector.
9202 (fill_simple_delay_slots): Adjust call to get_label_before.
9203 (fill_slots_from_thread): Likewise.
9204 (relax_delay_slots): Likewise.
9205 (make_return_insns): Likewise.
9206 (dbr_schedule): Invoke update_alignment on the sibling_labels vector.
9207
3c8ca1ab
EB		 92082013-08-07 Eric Botcazou <ebotcazou@adacore.com>
9209
9210 * diagnostic.c (diagnostic_classify_diagnostic): Accept zero index and
9211 document its semantics.
9212 (diagnostic_report_diagnostic): Adjust accordingly.
9213
5ee5b32c
DM		 92142013-08-07 David Malcolm <dmalcolm@redhat.com>
9215
03b0ee0a 9216 * config/sparc/sparc.c (insert_pass_work_around_errata): Move into...
5ee5b32c
DM		 9217 (sparc_option_override): ...and port to new C++ pass API.
9218 * config/sparc/t-sparc (sparc.o): Add dep on CONTEXT_H
9219
abd566fa
PB		 92202013-08-07 Peter Bergner <bergner@vnet.ibm.com>
9221
9222 * config/rs6000/rs6000.c (htm_expand_builtin) <case 0>: Remove.
9223
2077db1b
CT		 92242013-08-06 Caroline Tice <cmtice@google.com>
9225
9226 * gcc.c (VTABLE_VERIFICATION_SPEC): New definition.
9227 (LINK_COMMAND_SPEC): Add VTABLE_VERIFICATION_SPEC.
9228 * tree-pass.h: Add pass_vtable_verify.
9229 * varasm.c (assemble_variable): Add code to properly set the comdat
9230 section and name for the .vtable_map_vars section.
9231 (assemble_vtyv_preinit_initializer): New function.
9232 (default_sectin_type_flags): Make sure .vtable_map_vars section has
9233 LINK_ONCE flag.
9234 * output.h: Add function decl for assemble_vtv_preinit_initializer.
9235 * vtable-verify.c: New file.
9236 * vtable-verify.h: New file.
9237 * flag-types.h (enum vtv_priority): Defintions for flag_vtable_verify
9238 initialiation levels.
9239 * timevar.def (TV_VTABLE_VERIFICATION): New definition.
9240 * passes.def: Insert pass_vtable_verify.
9241 * aclocal.m4: Reorder includes.
9242 * doc/invoke.texi: Document the -fvtable-verify=, -fvtv-debug, and
9243 -fvtv-counts options.
9244 * config/gnu-user.h (GNU_USER_TARGET_STARTFILE_SPEC): Add vtv_start*.o,
9245 as appropriate, if -fvtable-verify=... is used.
9246 (GNU_USER_TARGET_ENDFILE_SPEC): Add vtv_end*.o as appropriate, if
9247 -fvtable-verify=... is used.
9248 * Makefile.in (OBJS): Add vtable-verify.o to list.
9249 (vtable-verify.o): Add new build rule.
9250 (GTFILES): Add vtable-verify.c to list.
9251 * common.opt (fvtable-verify=): New flag.
9252 (vtv_priority): Values for fvtable-verify= flag.
9253 (fvtv-counts): New flag.
9254 (fvtv-debug): New flag.
9255 * tree.h (save_vtable_map_decl): New extern function decl.
9256
03085d1c
DM		 92572013-08-07 David Malcolm <dmalcolm@redhat.com>
9258
9259 * config/rl78/rl78.c (rl78_devirt_pass): Convert from a struct to...
9260 (pass_rl78_devirt): ...new subclass of rtl_opt_pass along with...
9261 (pass_data_rl78_devirt): ...new pass_data instance and...
9262 (make_pass_rl78_devirt): ...new function.
9263 (rl78_asm_file_start): Port pass registration to new C++ API.
9264
05555c4a
DM		 92652013-08-07 David Malcolm <dmalcolm@redhat.com>
9266
9267 * coretypes.h (rtl_opt_pass): Add.
9268 (gcc::context): Add.
9269 * config/epiphany/epiphany.c (pass_mode_switch_use): New.
9270 (epiphany_init): Port to new C++ pass API.
9271 (epiphany_optimize_mode_switching): Likewise.
9272 * pass_manager.h (pass_manager::get_pass_split_all_insns): New.
9273 (pass_manager::get_pass_mode_switching): New.
9274 (pass_manager::get_pass_peephole2): New.
9275 * mode-switching.c (pass_mode_switching): Add clone method.
9276 * recog.c (pass_peephole2): Add clone method.
9277 (pass_split_all_insns): Add clone method.
9278
c5a12e26
DM		 92792013-08-06 David Malcolm <dmalcolm@redhat.com>
9280
03b0ee0a
UB		 9281 * config/mips/mips.c (insert_pass_mips_machine_reorg2): Move into...
9282 (mips_option_override): ...here, porting to new C++ API for passes.
c5a12e26 9283
a2e2a668
JH		 92842013-08-06 Jan Hubicka <jh@suse.cz>
9285
9286 * cgraph.c (cgraph_get_body): New function based on lto.c
9287 implementation.
9288 * cgraph.h (cgraph_get_body): Declare.
03b0ee0a
UB		 9289 * cgraphclones.c (cgraph_create_virtual_clone): Commonize WPA and
9290 LTO paths.
a2e2a668
JH		 9291 * cgraphunit.c (expand_function): Get body prior expanding.
9292 * ipa.c (function_and_variable_visibility): Use gimple_has_body_p test.
03b0ee0a
UB		 9293 * lto-cgraph.c (lto_output_node): Do not stream bodies we don't
9294 really need.
a2e2a668
JH		 9295 * passes.c (do_per_function_toporder): Get body.
9296 * tree-inline.c (expand_call_inline): Get body prior inlining it.
9297 * tree-ssa-structalias.c (ipa_pta_execute): Get body; skip clones.
9298
15682f24
MJ		 92992013-08-06 Martin Jambor <mjambor@suse.cz>
9300
9301 PR fortran/57987
9302 * cgraphunit.c (cgraph_finalize_function): Assert that nested function
9303 is not re-finalized. Rename second parameter to no_collect.
9304
78f6dd68
MJ		 93052013-08-06 Martin Jambor <mjambor@suse.cz>
9306
9307 PR middle-end/58041
9308 * gimple-ssa-strength-reduction.c (replace_ref): Make sure built
9309 MEM_REF has proper alignment information.
9310
12211b99 93112013-08-05 Oleg Endo <olegendo@gcc.gnu.org>
eb923e54
OE		 9312
9313 PR other/12081
9314 * recog.h (rtx (*insn_gen_fn) (rtx, ...)): Replace typedef with new
9315 class insn_gen_fn.
9316 * expr.c (move_by_pieces_1, store_by_pieces_2): Replace argument
9317 rtx (*) (rtx, ...) with insn_gen_fn.
9318 * genoutput.c (output_insn_data): Cast gen_? function pointers to
9319 insn_gen_fn::stored_funcptr. Add initializer braces.
9320
8ac69a6c
DM		 93212013-08-05 David Malcolm <dmalcolm@redhat.com>
9322
9323 Rewrite how instances of passes are cloned to remove assumptions
9324 about their sizes (thus allowing pass subclasses to have
9325 additional data fields, albeit non-GC-managed ones at this point).
9326
9327 * passes.c (make_pass_instance): Now that passes have clone
9328 methods, rewrite this function to eliminate XNEW and memcpy
9329 calls that used hardcoded sizes. Since this function no longer
9330 creates pass instances, rename it to...
03b0ee0a
UB		 9331 (add_pass_instance): ...this. Document the old way that passes were
9332 numbered and flagged, and rework this function to continue using it.
8ac69a6c
DM		 9333 (next_pass_1): Add an initial_pass argument for use by
9334 add_pass_instance.
9335 (position_pass): When adding multiple instances of a pass, use
9336 the pass's clone method, rather than relying on the XNEW/memcpy
9337 within the former make_pass_instance (now add_pass_instance).
03b0ee0a
UB		 9338 (pass_manager::pass_manager): When invoking next_pass_1, also supply
9339 the initial instance of the current pass within the pass manager.
8ac69a6c 9340
27a4cd48
DM		 93412013-08-05 David Malcolm <dmalcolm@redhat.com>
9342
9343 This is the automated part of the conversion of passes from C
9344 structs to C++ classes.
9345
9346 Patch autogenerated by refactor_passes.py from
9347 https://github.com/davidmalcolm/gcc-refactoring-scripts
9348 revision 03fe39476a4c4ea450b49e087cfa817b5f92021e
9349
9350 * asan.c (pass_asan): Convert from a global struct to a subclass of
9351 gimple_opt_pass along with...
9352 (pass_data_asan): ...new pass_data instance and...
9353 (make_pass_asan): ...new function.
9354 (pass_asan_O0): Convert from a global struct to a subclass of
9355 gimple_opt_pass along with...
9356 (pass_data_asan_O0): ...new pass_data instance and...
9357 (make_pass_asan_O0): ...new function.
9358 * auto-inc-dec.c (pass_inc_dec): Convert from a global struct to a
9359 subclass of rtl_opt_pass along with...
9360 (pass_data_inc_dec): ...new pass_data instance and...
9361 (make_pass_inc_dec): ...new function.
9362 * bb-reorder.c (pass_reorder_blocks): Convert from a global struct to
9363 a subclass of rtl_opt_pass along with...
9364 (pass_data_reorder_blocks): ...new pass_data instance and...
9365 (make_pass_reorder_blocks): ...new function.
9366 (pass_duplicate_computed_gotos): Convert from a global struct to a
9367 subclass of rtl_opt_pass along with...
9368 (pass_data_duplicate_computed_gotos): ...new pass_data instance and...
9369 (make_pass_duplicate_computed_gotos): ...new function.
9370 (pass_partition_blocks): Convert from a global struct to a subclass of
9371 rtl_opt_pass along with...
9372 (pass_data_partition_blocks): ...new pass_data instance and...
9373 (make_pass_partition_blocks): ...new function.
9374 * bt-load.c (pass_branch_target_load_optimize1): Convert from a global
9375 struct to a subclass of rtl_opt_pass along with...
9376 (pass_data_branch_target_load_optimize1): ...new pass_data instance
9377 and...
9378 (make_pass_branch_target_load_optimize1): ...new function.
9379 (pass_branch_target_load_optimize2): Convert from a global struct to a
9380 subclass of rtl_opt_pass along with...
9381 (pass_data_branch_target_load_optimize2): ...new pass_data instance
9382 and...
9383 (make_pass_branch_target_load_optimize2): ...new function.
9384 * cfgcleanup.c (pass_jump): Convert from a global struct to a subclass
9385 of rtl_opt_pass along with...
9386 (pass_data_jump): ...new pass_data instance and...
9387 (make_pass_jump): ...new function.
9388 (pass_jump2): Convert from a global struct to a subclass of
9389 rtl_opt_pass along with...
9390 (pass_data_jump2): ...new pass_data instance and...
9391 (make_pass_jump2): ...new function.
9392 * cfgexpand.c (pass_expand): Convert from a global struct to a
9393 subclass of rtl_opt_pass along with...
9394 (pass_data_expand): ...new pass_data instance and...
9395 (make_pass_expand): ...new function.
9396 * cfgrtl.c (pass_free_cfg): Convert from a global struct to a subclass
9397 of rtl_opt_pass along with...
9398 (pass_data_free_cfg): ...new pass_data instance and...
9399 (make_pass_free_cfg): ...new function.
9400 (pass_into_cfg_layout_mode): Convert from a global struct to a
9401 subclass of rtl_opt_pass along with...
9402 (pass_data_into_cfg_layout_mode): ...new pass_data instance and...
9403 (make_pass_into_cfg_layout_mode): ...new function.
9404 (pass_outof_cfg_layout_mode): Convert from a global struct to a
9405 subclass of rtl_opt_pass along with...
9406 (pass_data_outof_cfg_layout_mode): ...new pass_data instance and...
9407 (make_pass_outof_cfg_layout_mode): ...new function.
9408 * cgraphbuild.c (pass_build_cgraph_edges): Convert from a global
9409 struct to a subclass of gimple_opt_pass along with...
9410 (pass_data_build_cgraph_edges): ...new pass_data instance and...
9411 (make_pass_build_cgraph_edges): ...new function.
9412 (pass_rebuild_cgraph_edges): Convert from a global struct to a
9413 subclass of gimple_opt_pass along with...
9414 (pass_data_rebuild_cgraph_edges): ...new pass_data instance and...
9415 (make_pass_rebuild_cgraph_edges): ...new function.
9416 (pass_remove_cgraph_callee_edges): Convert from a global struct to a
9417 subclass of gimple_opt_pass along with...
9418 (pass_data_remove_cgraph_callee_edges): ...new pass_data instance
9419 and...
9420 (make_pass_remove_cgraph_callee_edges): ...new function.
9421 * combine-stack-adj.c (pass_stack_adjustments): Convert from a global
9422 struct to a subclass of rtl_opt_pass along with...
9423 (pass_data_stack_adjustments): ...new pass_data instance and...
9424 (make_pass_stack_adjustments): ...new function.
9425 * combine.c (pass_combine): Convert from a global struct to a subclass
9426 of rtl_opt_pass along with...
9427 (pass_data_combine): ...new pass_data instance and...
9428 (make_pass_combine): ...new function.
9429 * compare-elim.c (pass_compare_elim_after_reload): Convert from a
9430 global struct to a subclass of rtl_opt_pass along with...
9431 (pass_data_compare_elim_after_reload): ...new pass_data instance
9432 and...
9433 (make_pass_compare_elim_after_reload): ...new function.
9434 * cprop.c (pass_rtl_cprop): Convert from a global struct to a subclass
9435 of rtl_opt_pass along with...
9436 (pass_data_rtl_cprop): ...new pass_data instance and...
9437 (make_pass_rtl_cprop): ...new function.
9438 * cse.c (pass_cse): Convert from a global struct to a subclass of
9439 rtl_opt_pass along with...
9440 (pass_data_cse): ...new pass_data instance and...
9441 (make_pass_cse): ...new function.
9442 (pass_cse2): Convert from a global struct to a subclass of
9443 rtl_opt_pass along with...
9444 (pass_data_cse2): ...new pass_data instance and...
9445 (make_pass_cse2): ...new function.
9446 (pass_cse_after_global_opts): Convert from a global struct to a
9447 subclass of rtl_opt_pass along with...
9448 (pass_data_cse_after_global_opts): ...new pass_data instance and...
9449 (make_pass_cse_after_global_opts): ...new function.
9450 * dce.c (pass_ud_rtl_dce): Convert from a global struct to a subclass
9451 of rtl_opt_pass along with...
9452 (pass_data_ud_rtl_dce): ...new pass_data instance and...
9453 (make_pass_ud_rtl_dce): ...new function.
9454 (pass_fast_rtl_dce): Convert from a global struct to a subclass of
9455 rtl_opt_pass along with...
9456 (pass_data_fast_rtl_dce): ...new pass_data instance and...
9457 (make_pass_fast_rtl_dce): ...new function.
9458 * df-core.c (pass_df_initialize_opt): Convert from a global struct to
9459 a subclass of rtl_opt_pass along with...
9460 (pass_data_df_initialize_opt): ...new pass_data instance and...
9461 (make_pass_df_initialize_opt): ...new function.
9462 (pass_df_initialize_no_opt): Convert from a global struct to a
9463 subclass of rtl_opt_pass along with...
9464 (pass_data_df_initialize_no_opt): ...new pass_data instance and...
9465 (make_pass_df_initialize_no_opt): ...new function.
9466 (pass_df_finish): Convert from a global struct to a subclass of
9467 rtl_opt_pass along with...
9468 (pass_data_df_finish): ...new pass_data instance and...
9469 (make_pass_df_finish): ...new function.
9470 * dse.c (pass_rtl_dse1): Convert from a global struct to a subclass of
9471 rtl_opt_pass along with...
9472 (pass_data_rtl_dse1): ...new pass_data instance and...
9473 (make_pass_rtl_dse1): ...new function.
9474 (pass_rtl_dse2): Convert from a global struct to a subclass of
9475 rtl_opt_pass along with...
9476 (pass_data_rtl_dse2): ...new pass_data instance and...
9477 (make_pass_rtl_dse2): ...new function.
9478 * dwarf2cfi.c (pass_dwarf2_frame): Convert from a global struct to a
9479 subclass of rtl_opt_pass along with...
9480 (pass_data_dwarf2_frame): ...new pass_data instance and...
9481 (make_pass_dwarf2_frame): ...new function.
9482 * except.c (pass_set_nothrow_function_flags): Convert from a global
9483 struct to a subclass of rtl_opt_pass along with...
9484 (pass_data_set_nothrow_function_flags): ...new pass_data instance
9485 and...
9486 (make_pass_set_nothrow_function_flags): ...new function.
9487 (pass_convert_to_eh_region_ranges): Convert from a global struct to a
9488 subclass of rtl_opt_pass along with...
9489 (pass_data_convert_to_eh_region_ranges): ...new pass_data instance
9490 and...
9491 (make_pass_convert_to_eh_region_ranges): ...new function.
9492 * final.c (pass_compute_alignments): Convert from a global struct to a
9493 subclass of rtl_opt_pass along with...
9494 (pass_data_compute_alignments): ...new pass_data instance and...
9495 (make_pass_compute_alignments): ...new function.
9496 (pass_final): Convert from a global struct to a subclass of
9497 rtl_opt_pass along with...
9498 (pass_data_final): ...new pass_data instance and...
9499 (make_pass_final): ...new function.
9500 (pass_shorten_branches): Convert from a global struct to a subclass of
9501 rtl_opt_pass along with...
9502 (pass_data_shorten_branches): ...new pass_data instance and...
9503 (make_pass_shorten_branches): ...new function.
9504 (pass_clean_state): Convert from a global struct to a subclass of
9505 rtl_opt_pass along with...
9506 (pass_data_clean_state): ...new pass_data instance and...
9507 (make_pass_clean_state): ...new function.
9508 * function.c (pass_instantiate_virtual_regs): Convert from a global
9509 struct to a subclass of rtl_opt_pass along with...
9510 (pass_data_instantiate_virtual_regs): ...new pass_data instance and...
9511 (make_pass_instantiate_virtual_regs): ...new function.
9512 (pass_leaf_regs): Convert from a global struct to a subclass of
9513 rtl_opt_pass along with...
9514 (pass_data_leaf_regs): ...new pass_data instance and...
9515 (make_pass_leaf_regs): ...new function.
9516 (pass_thread_prologue_and_epilogue): Convert from a global struct to a
9517 subclass of rtl_opt_pass along with...
9518 (pass_data_thread_prologue_and_epilogue): ...new pass_data instance
9519 and...
9520 (make_pass_thread_prologue_and_epilogue): ...new function.
9521 (pass_match_asm_constraints): Convert from a global struct to a
9522 subclass of rtl_opt_pass along with...
9523 (pass_data_match_asm_constraints): ...new pass_data instance and...
9524 (make_pass_match_asm_constraints): ...new function.
9525 * fwprop.c (pass_rtl_fwprop): Convert from a global struct to a
9526 subclass of rtl_opt_pass along with...
9527 (pass_data_rtl_fwprop): ...new pass_data instance and...
9528 (make_pass_rtl_fwprop): ...new function.
9529 (pass_rtl_fwprop_addr): Convert from a global struct to a subclass of
9530 rtl_opt_pass along with...
9531 (pass_data_rtl_fwprop_addr): ...new pass_data instance and...
9532 (make_pass_rtl_fwprop_addr): ...new function.
9533 * gcse.c (pass_rtl_pre): Convert from a global struct to a subclass of
9534 rtl_opt_pass along with...
9535 (pass_data_rtl_pre): ...new pass_data instance and...
9536 (make_pass_rtl_pre): ...new function.
9537 (pass_rtl_hoist): Convert from a global struct to a subclass of
9538 rtl_opt_pass along with...
9539 (pass_data_rtl_hoist): ...new pass_data instance and...
9540 (make_pass_rtl_hoist): ...new function.
9541 * gimple-low.c (pass_lower_cf): Convert from a global struct to a
9542 subclass of gimple_opt_pass along with...
9543 (pass_data_lower_cf): ...new pass_data instance and...
9544 (make_pass_lower_cf): ...new function.
9545 * gimple-ssa-strength-reduction.c (pass_strength_reduction): Convert
9546 from a global struct to a subclass of gimple_opt_pass along with...
9547 (pass_data_strength_reduction): ...new pass_data instance and...
9548 (make_pass_strength_reduction): ...new function.
9549 * ifcvt.c (pass_rtl_ifcvt): Convert from a global struct to a subclass
9550 of rtl_opt_pass along with...
9551 (pass_data_rtl_ifcvt): ...new pass_data instance and...
9552 (make_pass_rtl_ifcvt): ...new function.
9553 (pass_if_after_combine): Convert from a global struct to a subclass of
9554 rtl_opt_pass along with...
9555 (pass_data_if_after_combine): ...new pass_data instance and...
9556 (make_pass_if_after_combine): ...new function.
9557 (pass_if_after_reload): Convert from a global struct to a subclass of
9558 rtl_opt_pass along with...
9559 (pass_data_if_after_reload): ...new pass_data instance and...
9560 (make_pass_if_after_reload): ...new function.
9561 * init-regs.c (pass_initialize_regs): Convert from a global struct to
9562 a subclass of rtl_opt_pass along with...
9563 (pass_data_initialize_regs): ...new pass_data instance and...
9564 (make_pass_initialize_regs): ...new function.
9565 * ipa-cp.c (pass_ipa_cp): Convert from a global struct to a subclass
9566 of ipa_opt_pass_d along with...
9567 (pass_data_ipa_cp): ...new pass_data instance and...
9568 (make_pass_ipa_cp): ...new function.
9569 * ipa-inline-analysis.c (pass_inline_parameters): Convert from a
9570 global struct to a subclass of gimple_opt_pass along with...
9571 (pass_data_inline_parameters): ...new pass_data instance and...
9572 (make_pass_inline_parameters): ...new function.
9573 * ipa-inline.c (pass_early_inline): Convert from a global struct to a
9574 subclass of gimple_opt_pass along with...
9575 (pass_data_early_inline): ...new pass_data instance and...
9576 (make_pass_early_inline): ...new function.
9577 (pass_ipa_inline): Convert from a global struct to a subclass of
9578 ipa_opt_pass_d along with...
9579 (pass_data_ipa_inline): ...new pass_data instance and...
9580 (make_pass_ipa_inline): ...new function.
9581 * ipa-pure-const.c (pass_local_pure_const): Convert from a global
9582 struct to a subclass of gimple_opt_pass along with...
9583 (pass_data_local_pure_const): ...new pass_data instance and...
9584 (make_pass_local_pure_const): ...new function.
9585 (pass_ipa_pure_const): Convert from a global struct to a subclass of
9586 ipa_opt_pass_d along with...
9587 (pass_data_ipa_pure_const): ...new pass_data instance and...
9588 (make_pass_ipa_pure_const): ...new function.
9589 * ipa-reference.c (pass_ipa_reference): Convert from a global struct
9590 to a subclass of ipa_opt_pass_d along with...
9591 (pass_data_ipa_reference): ...new pass_data instance and...
9592 (make_pass_ipa_reference): ...new function.
9593 * ipa-split.c (pass_split_functions): Convert from a global struct to
9594 a subclass of gimple_opt_pass along with...
9595 (pass_data_split_functions): ...new pass_data instance and...
9596 (make_pass_split_functions): ...new function.
9597 (pass_feedback_split_functions): Convert from a global struct to a
9598 subclass of gimple_opt_pass along with...
9599 (pass_data_feedback_split_functions): ...new pass_data instance and...
9600 (make_pass_feedback_split_functions): ...new function.
9601 * ipa.c (pass_ipa_function_and_variable_visibility): Convert from a
9602 global struct to a subclass of simple_ipa_opt_pass along with...
9603 (pass_data_ipa_function_and_variable_visibility): ...new pass_data
9604 instance and...
9605 (make_pass_ipa_function_and_variable_visibility): ...new function.
9606 (pass_ipa_free_inline_summary): Convert from a global struct to a
9607 subclass of simple_ipa_opt_pass along with...
9608 (pass_data_ipa_free_inline_summary): ...new pass_data instance and...
9609 (make_pass_ipa_free_inline_summary): ...new function.
9610 (pass_ipa_whole_program_visibility): Convert from a global struct to a
9611 subclass of ipa_opt_pass_d along with...
9612 (pass_data_ipa_whole_program_visibility): ...new pass_data instance
9613 and...
9614 (make_pass_ipa_whole_program_visibility): ...new function.
9615 (pass_ipa_profile): Convert from a global struct to a subclass of
9616 ipa_opt_pass_d along with...
9617 (pass_data_ipa_profile): ...new pass_data instance and...
9618 (make_pass_ipa_profile): ...new function.
9619 (pass_ipa_cdtor_merge): Convert from a global struct to a subclass of
9620 ipa_opt_pass_d along with...
9621 (pass_data_ipa_cdtor_merge): ...new pass_data instance and...
9622 (make_pass_ipa_cdtor_merge): ...new function.
9623 * ira.c (pass_ira): Convert from a global struct to a subclass of
9624 rtl_opt_pass along with...
9625 (pass_data_ira): ...new pass_data instance and...
9626 (make_pass_ira): ...new function.
9627 (pass_reload): Convert from a global struct to a subclass of
9628 rtl_opt_pass along with...
9629 (pass_data_reload): ...new pass_data instance and...
9630 (make_pass_reload): ...new function.
9631 * jump.c (pass_cleanup_barriers): Convert from a global struct to a
9632 subclass of rtl_opt_pass along with...
9633 (pass_data_cleanup_barriers): ...new pass_data instance and...
9634 (make_pass_cleanup_barriers): ...new function.
9635 * loop-init.c (pass_loop2): Convert from a global struct to a subclass
9636 of rtl_opt_pass along with...
9637 (pass_data_loop2): ...new pass_data instance and...
9638 (make_pass_loop2): ...new function.
9639 (pass_rtl_loop_init): Convert from a global struct to a subclass of
9640 rtl_opt_pass along with...
9641 (pass_data_rtl_loop_init): ...new pass_data instance and...
9642 (make_pass_rtl_loop_init): ...new function.
9643 (pass_rtl_loop_done): Convert from a global struct to a subclass of
9644 rtl_opt_pass along with...
9645 (pass_data_rtl_loop_done): ...new pass_data instance and...
9646 (make_pass_rtl_loop_done): ...new function.
9647 (pass_rtl_move_loop_invariants): Convert from a global struct to a
9648 subclass of rtl_opt_pass along with...
9649 (pass_data_rtl_move_loop_invariants): ...new pass_data instance and...
9650 (make_pass_rtl_move_loop_invariants): ...new function.
9651 (pass_rtl_unswitch): Convert from a global struct to a subclass of
9652 rtl_opt_pass along with...
9653 (pass_data_rtl_unswitch): ...new pass_data instance and...
9654 (make_pass_rtl_unswitch): ...new function.
9655 (pass_rtl_unroll_and_peel_loops): Convert from a global struct to a
9656 subclass of rtl_opt_pass along with...
9657 (pass_data_rtl_unroll_and_peel_loops): ...new pass_data instance
9658 and...
9659 (make_pass_rtl_unroll_and_peel_loops): ...new function.
9660 (pass_rtl_doloop): Convert from a global struct to a subclass of
9661 rtl_opt_pass along with...
9662 (pass_data_rtl_doloop): ...new pass_data instance and...
9663 (make_pass_rtl_doloop): ...new function.
9664 * lower-subreg.c (pass_lower_subreg): Convert from a global struct to
9665 a subclass of rtl_opt_pass along with...
9666 (pass_data_lower_subreg): ...new pass_data instance and...
9667 (make_pass_lower_subreg): ...new function.
9668 (pass_lower_subreg2): Convert from a global struct to a subclass of
9669 rtl_opt_pass along with...
9670 (pass_data_lower_subreg2): ...new pass_data instance and...
9671 (make_pass_lower_subreg2): ...new function.
9672 * lto-streamer-out.c (pass_ipa_lto_gimple_out): Convert from a global
9673 struct to a subclass of ipa_opt_pass_d along with...
9674 (pass_data_ipa_lto_gimple_out): ...new pass_data instance and...
9675 (make_pass_ipa_lto_gimple_out): ...new function.
9676 (pass_ipa_lto_finish_out): Convert from a global struct to a subclass
9677 of ipa_opt_pass_d along with...
9678 (pass_data_ipa_lto_finish_out): ...new pass_data instance and...
9679 (make_pass_ipa_lto_finish_out): ...new function.
9680 * mode-switching.c (pass_mode_switching): Convert from a global struct
9681 to a subclass of rtl_opt_pass along with...
9682 (pass_data_mode_switching): ...new pass_data instance and...
9683 (make_pass_mode_switching): ...new function.
9684 * modulo-sched.c (pass_sms): Convert from a global struct to a
9685 subclass of rtl_opt_pass along with...
9686 (pass_data_sms): ...new pass_data instance and...
9687 (make_pass_sms): ...new function.
9688 * omp-low.c (pass_expand_omp): Convert from a global struct to a
9689 subclass of gimple_opt_pass along with...
9690 (pass_data_expand_omp): ...new pass_data instance and...
9691 (make_pass_expand_omp): ...new function.
9692 (pass_lower_omp): Convert from a global struct to a subclass of
9693 gimple_opt_pass along with...
9694 (pass_data_lower_omp): ...new pass_data instance and...
9695 (make_pass_lower_omp): ...new function.
9696 (pass_diagnose_omp_blocks): Convert from a global struct to a subclass
9697 of gimple_opt_pass along with...
9698 (pass_data_diagnose_omp_blocks): ...new pass_data instance and...
9699 (make_pass_diagnose_omp_blocks): ...new function.
9700 * passes.c (pass_early_local_passes): Convert from a global struct to
9701 a subclass of simple_ipa_opt_pass along with...
9702 (pass_data_early_local_passes): ...new pass_data instance and...
9703 (make_pass_early_local_passes): ...new function.
9704 (pass_all_early_optimizations): Convert from a global struct to a
9705 subclass of gimple_opt_pass along with...
9706 (pass_data_all_early_optimizations): ...new pass_data instance and...
9707 (make_pass_all_early_optimizations): ...new function.
9708 (pass_all_optimizations): Convert from a global struct to a subclass
9709 of gimple_opt_pass along with...
9710 (pass_data_all_optimizations): ...new pass_data instance and...
9711 (make_pass_all_optimizations): ...new function.
9712 (pass_all_optimizations_g): Convert from a global struct to a subclass
9713 of gimple_opt_pass along with...
9714 (pass_data_all_optimizations_g): ...new pass_data instance and...
9715 (make_pass_all_optimizations_g): ...new function.
9716 (pass_rest_of_compilation): Convert from a global struct to a subclass
9717 of rtl_opt_pass along with...
9718 (pass_data_rest_of_compilation): ...new pass_data instance and...
9719 (make_pass_rest_of_compilation): ...new function.
9720 (pass_postreload): Convert from a global struct to a subclass of
9721 rtl_opt_pass along with...
9722 (pass_data_postreload): ...new pass_data instance and...
9723 (make_pass_postreload): ...new function.
9724 * postreload-gcse.c (pass_gcse2): Convert from a global struct to a
9725 subclass of rtl_opt_pass along with...
9726 (pass_data_gcse2): ...new pass_data instance and...
9727 (make_pass_gcse2): ...new function.
9728 * postreload.c (pass_postreload_cse): Convert from a global struct to
9729 a subclass of rtl_opt_pass along with...
9730 (pass_data_postreload_cse): ...new pass_data instance and...
9731 (make_pass_postreload_cse): ...new function.
9732 * predict.c (pass_profile): Convert from a global struct to a subclass
9733 of gimple_opt_pass along with...
9734 (pass_data_profile): ...new pass_data instance and...
9735 (make_pass_profile): ...new function.
9736 (pass_strip_predict_hints): Convert from a global struct to a subclass
9737 of gimple_opt_pass along with...
9738 (pass_data_strip_predict_hints): ...new pass_data instance and...
9739 (make_pass_strip_predict_hints): ...new function.
9740 * recog.c (pass_peephole2): Convert from a global struct to a subclass
9741 of rtl_opt_pass along with...
9742 (pass_data_peephole2): ...new pass_data instance and...
9743 (make_pass_peephole2): ...new function.
9744 (pass_split_all_insns): Convert from a global struct to a subclass of
9745 rtl_opt_pass along with...
9746 (pass_data_split_all_insns): ...new pass_data instance and...
9747 (make_pass_split_all_insns): ...new function.
9748 (pass_split_after_reload): Convert from a global struct to a subclass
9749 of rtl_opt_pass along with...
9750 (pass_data_split_after_reload): ...new pass_data instance and...
9751 (make_pass_split_after_reload): ...new function.
9752 (pass_split_before_regstack): Convert from a global struct to a
9753 subclass of rtl_opt_pass along with...
9754 (pass_data_split_before_regstack): ...new pass_data instance and...
9755 (make_pass_split_before_regstack): ...new function.
9756 (pass_split_before_sched2): Convert from a global struct to a subclass
9757 of rtl_opt_pass along with...
9758 (pass_data_split_before_sched2): ...new pass_data instance and...
9759 (make_pass_split_before_sched2): ...new function.
9760 (pass_split_for_shorten_branches): Convert from a global struct to a
9761 subclass of rtl_opt_pass along with...
9762 (pass_data_split_for_shorten_branches): ...new pass_data instance
9763 and...
9764 (make_pass_split_for_shorten_branches): ...new function.
9765 * ree.c (pass_ree): Convert from a global struct to a subclass of
9766 rtl_opt_pass along with...
9767 (pass_data_ree): ...new pass_data instance and...
9768 (make_pass_ree): ...new function.
9769 * reg-stack.c (pass_stack_regs): Convert from a global struct to a
9770 subclass of rtl_opt_pass along with...
9771 (pass_data_stack_regs): ...new pass_data instance and...
9772 (make_pass_stack_regs): ...new function.
9773 (pass_stack_regs_run): Convert from a global struct to a subclass of
9774 rtl_opt_pass along with...
9775 (pass_data_stack_regs_run): ...new pass_data instance and...
9776 (make_pass_stack_regs_run): ...new function.
9777 * regcprop.c (pass_cprop_hardreg): Convert from a global struct to a
9778 subclass of rtl_opt_pass along with...
9779 (pass_data_cprop_hardreg): ...new pass_data instance and...
9780 (make_pass_cprop_hardreg): ...new function.
9781 * reginfo.c (pass_reginfo_init): Convert from a global struct to a
9782 subclass of rtl_opt_pass along with...
9783 (pass_data_reginfo_init): ...new pass_data instance and...
9784 (make_pass_reginfo_init): ...new function.
9785 * regmove.c (pass_regmove): Convert from a global struct to a subclass
9786 of rtl_opt_pass along with...
9787 (pass_data_regmove): ...new pass_data instance and...
9788 (make_pass_regmove): ...new function.
9789 * regrename.c (pass_regrename): Convert from a global struct to a
9790 subclass of rtl_opt_pass along with...
9791 (pass_data_regrename): ...new pass_data instance and...
9792 (make_pass_regrename): ...new function.
9793 * reorg.c (pass_delay_slots): Convert from a global struct to a
9794 subclass of rtl_opt_pass along with...
9795 (pass_data_delay_slots): ...new pass_data instance and...
9796 (make_pass_delay_slots): ...new function.
9797 (pass_machine_reorg): Convert from a global struct to a subclass of
9798 rtl_opt_pass along with...
9799 (pass_data_machine_reorg): ...new pass_data instance and...
9800 (make_pass_machine_reorg): ...new function.
9801 * sched-rgn.c (pass_sched): Convert from a global struct to a subclass
9802 of rtl_opt_pass along with...
9803 (pass_data_sched): ...new pass_data instance and...
9804 (make_pass_sched): ...new function.
9805 (pass_sched2): Convert from a global struct to a subclass of
9806 rtl_opt_pass along with...
9807 (pass_data_sched2): ...new pass_data instance and...
9808 (make_pass_sched2): ...new function.
9809 * stack-ptr-mod.c (pass_stack_ptr_mod): Convert from a global struct
9810 to a subclass of rtl_opt_pass along with...
9811 (pass_data_stack_ptr_mod): ...new pass_data instance and...
9812 (make_pass_stack_ptr_mod): ...new function.
9813 * store-motion.c (pass_rtl_store_motion): Convert from a global struct
9814 to a subclass of rtl_opt_pass along with...
9815 (pass_data_rtl_store_motion): ...new pass_data instance and...
9816 (make_pass_rtl_store_motion): ...new function.
9817 * tracer.c (pass_tracer): Convert from a global struct to a subclass
9818 of gimple_opt_pass along with...
9819 (pass_data_tracer): ...new pass_data instance and...
9820 (make_pass_tracer): ...new function.
9821 * trans-mem.c (pass_diagnose_tm_blocks): Convert from a global struct
9822 to a subclass of gimple_opt_pass along with...
9823 (pass_data_diagnose_tm_blocks): ...new pass_data instance and...
9824 (make_pass_diagnose_tm_blocks): ...new function.
9825 (pass_lower_tm): Convert from a global struct to a subclass of
9826 gimple_opt_pass along with...
9827 (pass_data_lower_tm): ...new pass_data instance and...
9828 (make_pass_lower_tm): ...new function.
9829 (pass_tm_init): Convert from a global struct to a subclass of
9830 gimple_opt_pass along with...
9831 (pass_data_tm_init): ...new pass_data instance and...
9832 (make_pass_tm_init): ...new function.
9833 (pass_tm_mark): Convert from a global struct to a subclass of
9834 gimple_opt_pass along with...
9835 (pass_data_tm_mark): ...new pass_data instance and...
9836 (make_pass_tm_mark): ...new function.
9837 (pass_tm_edges): Convert from a global struct to a subclass of
9838 gimple_opt_pass along with...
9839 (pass_data_tm_edges): ...new pass_data instance and...
9840 (make_pass_tm_edges): ...new function.
9841 (pass_tm_memopt): Convert from a global struct to a subclass of
9842 gimple_opt_pass along with...
9843 (pass_data_tm_memopt): ...new pass_data instance and...
9844 (make_pass_tm_memopt): ...new function.
9845 (pass_ipa_tm): Convert from a global struct to a subclass of
9846 simple_ipa_opt_pass along with...
9847 (pass_data_ipa_tm): ...new pass_data instance and...
9848 (make_pass_ipa_tm): ...new function.
9849 * tree-call-cdce.c (pass_call_cdce): Convert from a global struct to a
9850 subclass of gimple_opt_pass along with...
9851 (pass_data_call_cdce): ...new pass_data instance and...
9852 (make_pass_call_cdce): ...new function.
9853 * tree-cfg.c (pass_build_cfg): Convert from a global struct to a
9854 subclass of gimple_opt_pass along with...
9855 (pass_data_build_cfg): ...new pass_data instance and...
9856 (make_pass_build_cfg): ...new function.
9857 (pass_split_crit_edges): Convert from a global struct to a subclass of
9858 gimple_opt_pass along with...
9859 (pass_data_split_crit_edges): ...new pass_data instance and...
9860 (make_pass_split_crit_edges): ...new function.
9861 (pass_warn_function_return): Convert from a global struct to a
9862 subclass of gimple_opt_pass along with...
9863 (pass_data_warn_function_return): ...new pass_data instance and...
9864 (make_pass_warn_function_return): ...new function.
9865 (pass_warn_function_noreturn): Convert from a global struct to a
9866 subclass of gimple_opt_pass along with...
9867 (pass_data_warn_function_noreturn): ...new pass_data instance and...
9868 (make_pass_warn_function_noreturn): ...new function.
9869 (pass_warn_unused_result): Convert from a global struct to a subclass
9870 of gimple_opt_pass along with...
9871 (pass_data_warn_unused_result): ...new pass_data instance and...
9872 (make_pass_warn_unused_result): ...new function.
9873 * tree-cfgcleanup.c (pass_merge_phi): Convert from a global struct to
9874 a subclass of gimple_opt_pass along with...
9875 (pass_data_merge_phi): ...new pass_data instance and...
9876 (make_pass_merge_phi): ...new function.
9877 * tree-complex.c (pass_lower_complex): Convert from a global struct to
9878 a subclass of gimple_opt_pass along with...
9879 (pass_data_lower_complex): ...new pass_data instance and...
9880 (make_pass_lower_complex): ...new function.
9881 (pass_lower_complex_O0): Convert from a global struct to a subclass of
9882 gimple_opt_pass along with...
9883 (pass_data_lower_complex_O0): ...new pass_data instance and...
9884 (make_pass_lower_complex_O0): ...new function.
9885 * tree-eh.c (pass_lower_eh): Convert from a global struct to a
9886 subclass of gimple_opt_pass along with...
9887 (pass_data_lower_eh): ...new pass_data instance and...
9888 (make_pass_lower_eh): ...new function.
9889 (pass_refactor_eh): Convert from a global struct to a subclass of
9890 gimple_opt_pass along with...
9891 (pass_data_refactor_eh): ...new pass_data instance and...
9892 (make_pass_refactor_eh): ...new function.
9893 (pass_lower_resx): Convert from a global struct to a subclass of
9894 gimple_opt_pass along with...
9895 (pass_data_lower_resx): ...new pass_data instance and...
9896 (make_pass_lower_resx): ...new function.
9897 (pass_lower_eh_dispatch): Convert from a global struct to a subclass
9898 of gimple_opt_pass along with...
9899 (pass_data_lower_eh_dispatch): ...new pass_data instance and...
9900 (make_pass_lower_eh_dispatch): ...new function.
9901 (pass_cleanup_eh): Convert from a global struct to a subclass of
9902 gimple_opt_pass along with...
9903 (pass_data_cleanup_eh): ...new pass_data instance and...
9904 (make_pass_cleanup_eh): ...new function.
9905 * tree-emutls.c (pass_ipa_lower_emutls): Convert from a global struct
9906 to a subclass of simple_ipa_opt_pass along with...
9907 (pass_data_ipa_lower_emutls): ...new pass_data instance and...
9908 (make_pass_ipa_lower_emutls): ...new function.
9909 * tree-if-conv.c (pass_if_conversion): Convert from a global struct to
9910 a subclass of gimple_opt_pass along with...
9911 (pass_data_if_conversion): ...new pass_data instance and...
9912 (make_pass_if_conversion): ...new function.
9913 * tree-into-ssa.c (pass_build_ssa): Convert from a global struct to a
9914 subclass of gimple_opt_pass along with...
9915 (pass_data_build_ssa): ...new pass_data instance and...
9916 (make_pass_build_ssa): ...new function.
9917 * tree-loop-distribution.c (pass_loop_distribution): Convert from a
9918 global struct to a subclass of gimple_opt_pass along with...
9919 (pass_data_loop_distribution): ...new pass_data instance and...
9920 (make_pass_loop_distribution): ...new function.
9921 * tree-mudflap.c (pass_mudflap_1): Convert from a global struct to a
9922 subclass of gimple_opt_pass along with...
9923 (pass_data_mudflap_1): ...new pass_data instance and...
9924 (make_pass_mudflap_1): ...new function.
9925 (pass_mudflap_2): Convert from a global struct to a subclass of
9926 gimple_opt_pass along with...
9927 (pass_data_mudflap_2): ...new pass_data instance and...
9928 (make_pass_mudflap_2): ...new function.
9929 * tree-nomudflap.c (pass_mudflap_1): Convert from a global struct to a
9930 subclass of gimple_opt_pass along with...
9931 (pass_data_mudflap_1): ...new pass_data instance and...
9932 (make_pass_mudflap_1): ...new function.
9933 (pass_mudflap_2): Convert from a global struct to a subclass of
9934 gimple_opt_pass along with...
9935 (pass_data_mudflap_2): ...new pass_data instance and...
9936 (make_pass_mudflap_2): ...new function.
9937 * tree-nrv.c (pass_nrv): Convert from a global struct to a subclass of
9938 gimple_opt_pass along with...
9939 (pass_data_nrv): ...new pass_data instance and...
9940 (make_pass_nrv): ...new function.
9941 (pass_return_slot): Convert from a global struct to a subclass of
9942 gimple_opt_pass along with...
9943 (pass_data_return_slot): ...new pass_data instance and...
9944 (make_pass_return_slot): ...new function.
9945 * tree-object-size.c (pass_object_sizes): Convert from a global struct
9946 to a subclass of gimple_opt_pass along with...
9947 (pass_data_object_sizes): ...new pass_data instance and...
9948 (make_pass_object_sizes): ...new function.
9949 * tree-optimize.c (pass_cleanup_cfg_post_optimizing): Convert from a
9950 global struct to a subclass of gimple_opt_pass along with...
9951 (pass_data_cleanup_cfg_post_optimizing): ...new pass_data instance
9952 and...
9953 (make_pass_cleanup_cfg_post_optimizing): ...new function.
9954 (pass_fixup_cfg): Convert from a global struct to a subclass of
9955 gimple_opt_pass along with...
9956 (pass_data_fixup_cfg): ...new pass_data instance and...
9957 (make_pass_fixup_cfg): ...new function.
9958 * tree-pass.h (pass_mudflap_1): Replace declaration with that of...
9959 (make_pass_mudflap_1): ...new function.
9960 (pass_mudflap_2): Replace declaration with that of...
9961 (make_pass_mudflap_2): ...new function.
9962 (pass_asan): Replace declaration with that of...
9963 (make_pass_asan): ...new function.
9964 (pass_asan_O0): Replace declaration with that of...
9965 (make_pass_asan_O0): ...new function.
9966 (pass_tsan): Replace declaration with that of...
9967 (make_pass_tsan): ...new function.
9968 (pass_tsan_O0): Replace declaration with that of...
9969 (make_pass_tsan_O0): ...new function.
9970 (pass_lower_cf): Replace declaration with that of...
9971 (make_pass_lower_cf): ...new function.
9972 (pass_refactor_eh): Replace declaration with that of...
9973 (make_pass_refactor_eh): ...new function.
9974 (pass_lower_eh): Replace declaration with that of...
9975 (make_pass_lower_eh): ...new function.
9976 (pass_lower_eh_dispatch): Replace declaration with that of...
9977 (make_pass_lower_eh_dispatch): ...new function.
9978 (pass_lower_resx): Replace declaration with that of...
9979 (make_pass_lower_resx): ...new function.
9980 (pass_build_cfg): Replace declaration with that of...
9981 (make_pass_build_cfg): ...new function.
9982 (pass_early_tree_profile): Replace declaration with that of...
9983 (make_pass_early_tree_profile): ...new function.
9984 (pass_cleanup_eh): Replace declaration with that of...
9985 (make_pass_cleanup_eh): ...new function.
9986 (pass_sra): Replace declaration with that of...
9987 (make_pass_sra): ...new function.
9988 (pass_sra_early): Replace declaration with that of...
9989 (make_pass_sra_early): ...new function.
9990 (pass_early_ipa_sra): Replace declaration with that of...
9991 (make_pass_early_ipa_sra): ...new function.
9992 (pass_tail_recursion): Replace declaration with that of...
9993 (make_pass_tail_recursion): ...new function.
9994 (pass_tail_calls): Replace declaration with that of...
9995 (make_pass_tail_calls): ...new function.
9996 (pass_tree_loop): Replace declaration with that of...
9997 (make_pass_tree_loop): ...new function.
9998 (pass_tree_loop_init): Replace declaration with that of...
9999 (make_pass_tree_loop_init): ...new function.
10000 (pass_lim): Replace declaration with that of...
10001 (make_pass_lim): ...new function.
10002 (pass_tree_unswitch): Replace declaration with that of...
10003 (make_pass_tree_unswitch): ...new function.
10004 (pass_predcom): Replace declaration with that of...
10005 (make_pass_predcom): ...new function.
10006 (pass_iv_canon): Replace declaration with that of...
10007 (make_pass_iv_canon): ...new function.
10008 (pass_scev_cprop): Replace declaration with that of...
10009 (make_pass_scev_cprop): ...new function.
10010 (pass_empty_loop): Replace declaration with that of...
10011 (make_pass_empty_loop): ...new function.
10012 (pass_record_bounds): Replace declaration with that of...
10013 (make_pass_record_bounds): ...new function.
10014 (pass_graphite): Replace declaration with that of...
10015 (make_pass_graphite): ...new function.
10016 (pass_graphite_transforms): Replace declaration with that of...
10017 (make_pass_graphite_transforms): ...new function.
10018 (pass_if_conversion): Replace declaration with that of...
10019 (make_pass_if_conversion): ...new function.
10020 (pass_loop_distribution): Replace declaration with that of...
10021 (make_pass_loop_distribution): ...new function.
10022 (pass_vectorize): Replace declaration with that of...
10023 (make_pass_vectorize): ...new function.
10024 (pass_slp_vectorize): Replace declaration with that of...
10025 (make_pass_slp_vectorize): ...new function.
10026 (pass_complete_unroll): Replace declaration with that of...
10027 (make_pass_complete_unroll): ...new function.
10028 (pass_complete_unrolli): Replace declaration with that of...
10029 (make_pass_complete_unrolli): ...new function.
10030 (pass_parallelize_loops): Replace declaration with that of...
10031 (make_pass_parallelize_loops): ...new function.
10032 (pass_loop_prefetch): Replace declaration with that of...
10033 (make_pass_loop_prefetch): ...new function.
10034 (pass_iv_optimize): Replace declaration with that of...
10035 (make_pass_iv_optimize): ...new function.
10036 (pass_tree_loop_done): Replace declaration with that of...
10037 (make_pass_tree_loop_done): ...new function.
10038 (pass_ch): Replace declaration with that of...
10039 (make_pass_ch): ...new function.
10040 (pass_ccp): Replace declaration with that of...
10041 (make_pass_ccp): ...new function.
10042 (pass_phi_only_cprop): Replace declaration with that of...
10043 (make_pass_phi_only_cprop): ...new function.
10044 (pass_build_ssa): Replace declaration with that of...
10045 (make_pass_build_ssa): ...new function.
10046 (pass_build_alias): Replace declaration with that of...
10047 (make_pass_build_alias): ...new function.
10048 (pass_build_ealias): Replace declaration with that of...
10049 (make_pass_build_ealias): ...new function.
10050 (pass_dominator): Replace declaration with that of...
10051 (make_pass_dominator): ...new function.
10052 (pass_dce): Replace declaration with that of...
10053 (make_pass_dce): ...new function.
10054 (pass_dce_loop): Replace declaration with that of...
10055 (make_pass_dce_loop): ...new function.
10056 (pass_cd_dce): Replace declaration with that of...
10057 (make_pass_cd_dce): ...new function.
10058 (pass_call_cdce): Replace declaration with that of...
10059 (make_pass_call_cdce): ...new function.
10060 (pass_merge_phi): Replace declaration with that of...
10061 (make_pass_merge_phi): ...new function.
10062 (pass_split_crit_edges): Replace declaration with that of...
10063 (make_pass_split_crit_edges): ...new function.
10064 (pass_pre): Replace declaration with that of...
10065 (make_pass_pre): ...new function.
10066 (pass_profile): Replace declaration with that of...
10067 (make_pass_profile): ...new function.
10068 (pass_strip_predict_hints): Replace declaration with that of...
10069 (make_pass_strip_predict_hints): ...new function.
10070 (pass_lower_complex_O0): Replace declaration with that of...
10071 (make_pass_lower_complex_O0): ...new function.
10072 (pass_lower_complex): Replace declaration with that of...
10073 (make_pass_lower_complex): ...new function.
10074 (pass_lower_vector): Replace declaration with that of...
10075 (make_pass_lower_vector): ...new function.
10076 (pass_lower_vector_ssa): Replace declaration with that of...
10077 (make_pass_lower_vector_ssa): ...new function.
10078 (pass_lower_omp): Replace declaration with that of...
10079 (make_pass_lower_omp): ...new function.
10080 (pass_diagnose_omp_blocks): Replace declaration with that of...
10081 (make_pass_diagnose_omp_blocks): ...new function.
10082 (pass_expand_omp): Replace declaration with that of...
10083 (make_pass_expand_omp): ...new function.
10084 (pass_expand_omp_ssa): Replace declaration with that of...
10085 (make_pass_expand_omp_ssa): ...new function.
10086 (pass_object_sizes): Replace declaration with that of...
10087 (make_pass_object_sizes): ...new function.
10088 (pass_strlen): Replace declaration with that of...
10089 (make_pass_strlen): ...new function.
10090 (pass_fold_builtins): Replace declaration with that of...
10091 (make_pass_fold_builtins): ...new function.
10092 (pass_stdarg): Replace declaration with that of...
10093 (make_pass_stdarg): ...new function.
10094 (pass_early_warn_uninitialized): Replace declaration with that of...
10095 (make_pass_early_warn_uninitialized): ...new function.
10096 (pass_late_warn_uninitialized): Replace declaration with that of...
10097 (make_pass_late_warn_uninitialized): ...new function.
10098 (pass_cse_reciprocals): Replace declaration with that of...
10099 (make_pass_cse_reciprocals): ...new function.
10100 (pass_cse_sincos): Replace declaration with that of...
10101 (make_pass_cse_sincos): ...new function.
10102 (pass_optimize_bswap): Replace declaration with that of...
10103 (make_pass_optimize_bswap): ...new function.
10104 (pass_optimize_widening_mul): Replace declaration with that of...
10105 (make_pass_optimize_widening_mul): ...new function.
10106 (pass_warn_function_return): Replace declaration with that of...
10107 (make_pass_warn_function_return): ...new function.
10108 (pass_warn_function_noreturn): Replace declaration with that of...
10109 (make_pass_warn_function_noreturn): ...new function.
10110 (pass_cselim): Replace declaration with that of...
10111 (make_pass_cselim): ...new function.
10112 (pass_phiopt): Replace declaration with that of...
10113 (make_pass_phiopt): ...new function.
10114 (pass_forwprop): Replace declaration with that of...
10115 (make_pass_forwprop): ...new function.
10116 (pass_phiprop): Replace declaration with that of...
10117 (make_pass_phiprop): ...new function.
10118 (pass_tree_ifcombine): Replace declaration with that of...
10119 (make_pass_tree_ifcombine): ...new function.
10120 (pass_dse): Replace declaration with that of...
10121 (make_pass_dse): ...new function.
10122 (pass_nrv): Replace declaration with that of...
10123 (make_pass_nrv): ...new function.
10124 (pass_rename_ssa_copies): Replace declaration with that of...
10125 (make_pass_rename_ssa_copies): ...new function.
10126 (pass_sink_code): Replace declaration with that of...
10127 (make_pass_sink_code): ...new function.
10128 (pass_fre): Replace declaration with that of...
10129 (make_pass_fre): ...new function.
10130 (pass_check_data_deps): Replace declaration with that of...
10131 (make_pass_check_data_deps): ...new function.
10132 (pass_copy_prop): Replace declaration with that of...
10133 (make_pass_copy_prop): ...new function.
10134 (pass_vrp): Replace declaration with that of...
10135 (make_pass_vrp): ...new function.
10136 (pass_uncprop): Replace declaration with that of...
10137 (make_pass_uncprop): ...new function.
10138 (pass_return_slot): Replace declaration with that of...
10139 (make_pass_return_slot): ...new function.
10140 (pass_reassoc): Replace declaration with that of...
10141 (make_pass_reassoc): ...new function.
10142 (pass_rebuild_cgraph_edges): Replace declaration with that of...
10143 (make_pass_rebuild_cgraph_edges): ...new function.
10144 (pass_remove_cgraph_callee_edges): Replace declaration with that of...
10145 (make_pass_remove_cgraph_callee_edges): ...new function.
10146 (pass_build_cgraph_edges): Replace declaration with that of...
10147 (make_pass_build_cgraph_edges): ...new function.
10148 (pass_local_pure_const): Replace declaration with that of...
10149 (make_pass_local_pure_const): ...new function.
10150 (pass_tracer): Replace declaration with that of...
10151 (make_pass_tracer): ...new function.
10152 (pass_warn_unused_result): Replace declaration with that of...
10153 (make_pass_warn_unused_result): ...new function.
10154 (pass_diagnose_tm_blocks): Replace declaration with that of...
10155 (make_pass_diagnose_tm_blocks): ...new function.
10156 (pass_lower_tm): Replace declaration with that of...
10157 (make_pass_lower_tm): ...new function.
10158 (pass_tm_init): Replace declaration with that of...
10159 (make_pass_tm_init): ...new function.
10160 (pass_tm_mark): Replace declaration with that of...
10161 (make_pass_tm_mark): ...new function.
10162 (pass_tm_memopt): Replace declaration with that of...
10163 (make_pass_tm_memopt): ...new function.
10164 (pass_tm_edges): Replace declaration with that of...
10165 (make_pass_tm_edges): ...new function.
10166 (pass_split_functions): Replace declaration with that of...
10167 (make_pass_split_functions): ...new function.
10168 (pass_feedback_split_functions): Replace declaration with that of...
10169 (make_pass_feedback_split_functions): ...new function.
10170 (pass_strength_reduction): Replace declaration with that of...
10171 (make_pass_strength_reduction): ...new function.
10172 (pass_ipa_lower_emutls): Replace declaration with that of...
10173 (make_pass_ipa_lower_emutls): ...new function.
10174 (pass_ipa_function_and_variable_visibility): Replace declaration with
10175 that of...
10176 (make_pass_ipa_function_and_variable_visibility): ...new function.
10177 (pass_ipa_tree_profile): Replace declaration with that of...
10178 (make_pass_ipa_tree_profile): ...new function.
10179 (pass_early_local_passes): Replace declaration with that of...
10180 (make_pass_early_local_passes): ...new function.
10181 (pass_ipa_whole_program_visibility): Replace declaration with that
10182 of...
10183 (make_pass_ipa_whole_program_visibility): ...new function.
10184 (pass_ipa_lto_gimple_out): Replace declaration with that of...
10185 (make_pass_ipa_lto_gimple_out): ...new function.
10186 (pass_ipa_increase_alignment): Replace declaration with that of...
10187 (make_pass_ipa_increase_alignment): ...new function.
10188 (pass_ipa_inline): Replace declaration with that of...
10189 (make_pass_ipa_inline): ...new function.
10190 (pass_ipa_free_lang_data): Replace declaration with that of...
10191 (make_pass_ipa_free_lang_data): ...new function.
10192 (pass_ipa_free_inline_summary): Replace declaration with that of...
10193 (make_pass_ipa_free_inline_summary): ...new function.
10194 (pass_ipa_cp): Replace declaration with that of...
10195 (make_pass_ipa_cp): ...new function.
10196 (pass_ipa_reference): Replace declaration with that of...
10197 (make_pass_ipa_reference): ...new function.
10198 (pass_ipa_pure_const): Replace declaration with that of...
10199 (make_pass_ipa_pure_const): ...new function.
10200 (pass_ipa_pta): Replace declaration with that of...
10201 (make_pass_ipa_pta): ...new function.
10202 (pass_ipa_lto_finish_out): Replace declaration with that of...
10203 (make_pass_ipa_lto_finish_out): ...new function.
10204 (pass_ipa_tm): Replace declaration with that of...
10205 (make_pass_ipa_tm): ...new function.
10206 (pass_ipa_profile): Replace declaration with that of...
10207 (make_pass_ipa_profile): ...new function.
10208 (pass_ipa_cdtor_merge): Replace declaration with that of...
10209 (make_pass_ipa_cdtor_merge): ...new function.
10210 (pass_cleanup_cfg_post_optimizing): Replace declaration with that
10211 of...
10212 (make_pass_cleanup_cfg_post_optimizing): ...new function.
10213 (pass_init_datastructures): Replace declaration with that of...
10214 (make_pass_init_datastructures): ...new function.
10215 (pass_fixup_cfg): Replace declaration with that of...
10216 (make_pass_fixup_cfg): ...new function.
10217 (pass_expand): Replace declaration with that of...
10218 (make_pass_expand): ...new function.
10219 (pass_instantiate_virtual_regs): Replace declaration with that of...
10220 (make_pass_instantiate_virtual_regs): ...new function.
10221 (pass_rtl_fwprop): Replace declaration with that of...
10222 (make_pass_rtl_fwprop): ...new function.
10223 (pass_rtl_fwprop_addr): Replace declaration with that of...
10224 (make_pass_rtl_fwprop_addr): ...new function.
10225 (pass_jump): Replace declaration with that of...
10226 (make_pass_jump): ...new function.
10227 (pass_jump2): Replace declaration with that of...
10228 (make_pass_jump2): ...new function.
10229 (pass_lower_subreg): Replace declaration with that of...
10230 (make_pass_lower_subreg): ...new function.
10231 (pass_cse): Replace declaration with that of...
10232 (make_pass_cse): ...new function.
10233 (pass_fast_rtl_dce): Replace declaration with that of...
10234 (make_pass_fast_rtl_dce): ...new function.
10235 (pass_ud_rtl_dce): Replace declaration with that of...
10236 (make_pass_ud_rtl_dce): ...new function.
10237 (pass_rtl_dce): Replace declaration with that of...
10238 (make_pass_rtl_dce): ...new function.
10239 (pass_rtl_dse1): Replace declaration with that of...
10240 (make_pass_rtl_dse1): ...new function.
10241 (pass_rtl_dse2): Replace declaration with that of...
10242 (make_pass_rtl_dse2): ...new function.
10243 (pass_rtl_dse3): Replace declaration with that of...
10244 (make_pass_rtl_dse3): ...new function.
10245 (pass_rtl_cprop): Replace declaration with that of...
10246 (make_pass_rtl_cprop): ...new function.
10247 (pass_rtl_pre): Replace declaration with that of...
10248 (make_pass_rtl_pre): ...new function.
10249 (pass_rtl_hoist): Replace declaration with that of...
10250 (make_pass_rtl_hoist): ...new function.
10251 (pass_rtl_store_motion): Replace declaration with that of...
10252 (make_pass_rtl_store_motion): ...new function.
10253 (pass_cse_after_global_opts): Replace declaration with that of...
10254 (make_pass_cse_after_global_opts): ...new function.
10255 (pass_rtl_ifcvt): Replace declaration with that of...
10256 (make_pass_rtl_ifcvt): ...new function.
10257 (pass_into_cfg_layout_mode): Replace declaration with that of...
10258 (make_pass_into_cfg_layout_mode): ...new function.
10259 (pass_outof_cfg_layout_mode): Replace declaration with that of...
10260 (make_pass_outof_cfg_layout_mode): ...new function.
10261 (pass_loop2): Replace declaration with that of...
10262 (make_pass_loop2): ...new function.
10263 (pass_rtl_loop_init): Replace declaration with that of...
10264 (make_pass_rtl_loop_init): ...new function.
10265 (pass_rtl_move_loop_invariants): Replace declaration with that of...
10266 (make_pass_rtl_move_loop_invariants): ...new function.
10267 (pass_rtl_unswitch): Replace declaration with that of...
10268 (make_pass_rtl_unswitch): ...new function.
10269 (pass_rtl_unroll_and_peel_loops): Replace declaration with that of...
10270 (make_pass_rtl_unroll_and_peel_loops): ...new function.
10271 (pass_rtl_doloop): Replace declaration with that of...
10272 (make_pass_rtl_doloop): ...new function.
10273 (pass_rtl_loop_done): Replace declaration with that of...
10274 (make_pass_rtl_loop_done): ...new function.
10275 (pass_web): Replace declaration with that of...
10276 (make_pass_web): ...new function.
10277 (pass_cse2): Replace declaration with that of...
10278 (make_pass_cse2): ...new function.
10279 (pass_df_initialize_opt): Replace declaration with that of...
10280 (make_pass_df_initialize_opt): ...new function.
10281 (pass_df_initialize_no_opt): Replace declaration with that of...
10282 (make_pass_df_initialize_no_opt): ...new function.
10283 (pass_reginfo_init): Replace declaration with that of...
10284 (make_pass_reginfo_init): ...new function.
10285 (pass_inc_dec): Replace declaration with that of...
10286 (make_pass_inc_dec): ...new function.
10287 (pass_stack_ptr_mod): Replace declaration with that of...
10288 (make_pass_stack_ptr_mod): ...new function.
10289 (pass_initialize_regs): Replace declaration with that of...
10290 (make_pass_initialize_regs): ...new function.
10291 (pass_combine): Replace declaration with that of...
10292 (make_pass_combine): ...new function.
10293 (pass_if_after_combine): Replace declaration with that of...
10294 (make_pass_if_after_combine): ...new function.
10295 (pass_ree): Replace declaration with that of...
10296 (make_pass_ree): ...new function.
10297 (pass_partition_blocks): Replace declaration with that of...
10298 (make_pass_partition_blocks): ...new function.
10299 (pass_match_asm_constraints): Replace declaration with that of...
10300 (make_pass_match_asm_constraints): ...new function.
10301 (pass_regmove): Replace declaration with that of...
10302 (make_pass_regmove): ...new function.
10303 (pass_split_all_insns): Replace declaration with that of...
10304 (make_pass_split_all_insns): ...new function.
10305 (pass_fast_rtl_byte_dce): Replace declaration with that of...
10306 (make_pass_fast_rtl_byte_dce): ...new function.
10307 (pass_lower_subreg2): Replace declaration with that of...
10308 (make_pass_lower_subreg2): ...new function.
10309 (pass_mode_switching): Replace declaration with that of...
10310 (make_pass_mode_switching): ...new function.
10311 (pass_sms): Replace declaration with that of...
10312 (make_pass_sms): ...new function.
10313 (pass_sched): Replace declaration with that of...
10314 (make_pass_sched): ...new function.
10315 (pass_ira): Replace declaration with that of...
10316 (make_pass_ira): ...new function.
10317 (pass_reload): Replace declaration with that of...
10318 (make_pass_reload): ...new function.
10319 (pass_clean_state): Replace declaration with that of...
10320 (make_pass_clean_state): ...new function.
10321 (pass_branch_prob): Replace declaration with that of...
10322 (make_pass_branch_prob): ...new function.
10323 (pass_value_profile_transformations): Replace declaration with that
10324 of...
10325 (make_pass_value_profile_transformations): ...new function.
10326 (pass_postreload_cse): Replace declaration with that of...
10327 (make_pass_postreload_cse): ...new function.
10328 (pass_gcse2): Replace declaration with that of...
10329 (make_pass_gcse2): ...new function.
10330 (pass_split_after_reload): Replace declaration with that of...
10331 (make_pass_split_after_reload): ...new function.
10332 (pass_branch_target_load_optimize1): Replace declaration with that
10333 of...
10334 (make_pass_branch_target_load_optimize1): ...new function.
10335 (pass_thread_prologue_and_epilogue): Replace declaration with that
10336 of...
10337 (make_pass_thread_prologue_and_epilogue): ...new function.
10338 (pass_stack_adjustments): Replace declaration with that of...
10339 (make_pass_stack_adjustments): ...new function.
10340 (pass_peephole2): Replace declaration with that of...
10341 (make_pass_peephole2): ...new function.
10342 (pass_if_after_reload): Replace declaration with that of...
10343 (make_pass_if_after_reload): ...new function.
10344 (pass_regrename): Replace declaration with that of...
10345 (make_pass_regrename): ...new function.
10346 (pass_cprop_hardreg): Replace declaration with that of...
10347 (make_pass_cprop_hardreg): ...new function.
10348 (pass_reorder_blocks): Replace declaration with that of...
10349 (make_pass_reorder_blocks): ...new function.
10350 (pass_branch_target_load_optimize2): Replace declaration with that
10351 of...
10352 (make_pass_branch_target_load_optimize2): ...new function.
10353 (pass_leaf_regs): Replace declaration with that of...
10354 (make_pass_leaf_regs): ...new function.
10355 (pass_split_before_sched2): Replace declaration with that of...
10356 (make_pass_split_before_sched2): ...new function.
10357 (pass_compare_elim_after_reload): Replace declaration with that of...
10358 (make_pass_compare_elim_after_reload): ...new function.
10359 (pass_sched2): Replace declaration with that of...
10360 (make_pass_sched2): ...new function.
10361 (pass_stack_regs): Replace declaration with that of...
10362 (make_pass_stack_regs): ...new function.
10363 (pass_stack_regs_run): Replace declaration with that of...
10364 (make_pass_stack_regs_run): ...new function.
10365 (pass_df_finish): Replace declaration with that of...
10366 (make_pass_df_finish): ...new function.
10367 (pass_compute_alignments): Replace declaration with that of...
10368 (make_pass_compute_alignments): ...new function.
10369 (pass_duplicate_computed_gotos): Replace declaration with that of...
10370 (make_pass_duplicate_computed_gotos): ...new function.
10371 (pass_variable_tracking): Replace declaration with that of...
10372 (make_pass_variable_tracking): ...new function.
10373 (pass_free_cfg): Replace declaration with that of...
10374 (make_pass_free_cfg): ...new function.
10375 (pass_machine_reorg): Replace declaration with that of...
10376 (make_pass_machine_reorg): ...new function.
10377 (pass_cleanup_barriers): Replace declaration with that of...
10378 (make_pass_cleanup_barriers): ...new function.
10379 (pass_delay_slots): Replace declaration with that of...
10380 (make_pass_delay_slots): ...new function.
10381 (pass_split_for_shorten_branches): Replace declaration with that of...
10382 (make_pass_split_for_shorten_branches): ...new function.
10383 (pass_split_before_regstack): Replace declaration with that of...
10384 (make_pass_split_before_regstack): ...new function.
10385 (pass_convert_to_eh_region_ranges): Replace declaration with that
10386 of...
10387 (make_pass_convert_to_eh_region_ranges): ...new function.
10388 (pass_shorten_branches): Replace declaration with that of...
10389 (make_pass_shorten_branches): ...new function.
10390 (pass_set_nothrow_function_flags): Replace declaration with that of...
10391 (make_pass_set_nothrow_function_flags): ...new function.
10392 (pass_dwarf2_frame): Replace declaration with that of...
10393 (make_pass_dwarf2_frame): ...new function.
10394 (pass_final): Replace declaration with that of...
10395 (make_pass_final): ...new function.
10396 (pass_rtl_seqabstr): Replace declaration with that of...
10397 (make_pass_rtl_seqabstr): ...new function.
10398 (pass_release_ssa_names): Replace declaration with that of...
10399 (make_pass_release_ssa_names): ...new function.
10400 (pass_early_inline): Replace declaration with that of...
10401 (make_pass_early_inline): ...new function.
10402 (pass_inline_parameters): Replace declaration with that of...
10403 (make_pass_inline_parameters): ...new function.
10404 (pass_update_address_taken): Replace declaration with that of...
10405 (make_pass_update_address_taken): ...new function.
10406 (pass_convert_switch): Replace declaration with that of...
10407 (make_pass_convert_switch): ...new function.
10408 * tree-profile.c (pass_ipa_tree_profile): Convert from a global struct
10409 to a subclass of simple_ipa_opt_pass along with...
10410 (pass_data_ipa_tree_profile): ...new pass_data instance and...
10411 (make_pass_ipa_tree_profile): ...new function.
10412 * tree-sra.c (pass_sra_early): Convert from a global struct to a
10413 subclass of gimple_opt_pass along with...
10414 (pass_data_sra_early): ...new pass_data instance and...
10415 (make_pass_sra_early): ...new function.
10416 (pass_sra): Convert from a global struct to a subclass of
10417 gimple_opt_pass along with...
10418 (pass_data_sra): ...new pass_data instance and...
10419 (make_pass_sra): ...new function.
10420 (pass_early_ipa_sra): Convert from a global struct to a subclass of
10421 gimple_opt_pass along with...
10422 (pass_data_early_ipa_sra): ...new pass_data instance and...
10423 (make_pass_early_ipa_sra): ...new function.
10424 * tree-ssa-ccp.c (pass_ccp): Convert from a global struct to a
10425 subclass of gimple_opt_pass along with...
10426 (pass_data_ccp): ...new pass_data instance and...
10427 (make_pass_ccp): ...new function.
10428 (pass_fold_builtins): Convert from a global struct to a subclass of
10429 gimple_opt_pass along with...
10430 (pass_data_fold_builtins): ...new pass_data instance and...
10431 (make_pass_fold_builtins): ...new function.
10432 * tree-ssa-copy.c (pass_copy_prop): Convert from a global struct to a
10433 subclass of gimple_opt_pass along with...
10434 (pass_data_copy_prop): ...new pass_data instance and...
10435 (make_pass_copy_prop): ...new function.
10436 * tree-ssa-copyrename.c (pass_rename_ssa_copies): Convert from a
10437 global struct to a subclass of gimple_opt_pass along with...
10438 (pass_data_rename_ssa_copies): ...new pass_data instance and...
10439 (make_pass_rename_ssa_copies): ...new function.
10440 * tree-ssa-dce.c (pass_dce): Convert from a global struct to a
10441 subclass of gimple_opt_pass along with...
10442 (pass_data_dce): ...new pass_data instance and...
10443 (make_pass_dce): ...new function.
10444 (pass_dce_loop): Convert from a global struct to a subclass of
10445 gimple_opt_pass along with...
10446 (pass_data_dce_loop): ...new pass_data instance and...
10447 (make_pass_dce_loop): ...new function.
10448 (pass_cd_dce): Convert from a global struct to a subclass of
10449 gimple_opt_pass along with...
10450 (pass_data_cd_dce): ...new pass_data instance and...
10451 (make_pass_cd_dce): ...new function.
10452 * tree-ssa-dom.c (pass_dominator): Convert from a global struct to a
10453 subclass of gimple_opt_pass along with...
10454 (pass_data_dominator): ...new pass_data instance and...
10455 (make_pass_dominator): ...new function.
10456 (pass_phi_only_cprop): Convert from a global struct to a subclass of
10457 gimple_opt_pass along with...
10458 (pass_data_phi_only_cprop): ...new pass_data instance and...
10459 (make_pass_phi_only_cprop): ...new function.
10460 * tree-ssa-dse.c (pass_dse): Convert from a global struct to a
10461 subclass of gimple_opt_pass along with...
10462 (pass_data_dse): ...new pass_data instance and...
10463 (make_pass_dse): ...new function.
10464 * tree-ssa-forwprop.c (pass_forwprop): Convert from a global struct to
10465 a subclass of gimple_opt_pass along with...
10466 (pass_data_forwprop): ...new pass_data instance and...
10467 (make_pass_forwprop): ...new function.
10468 * tree-ssa-ifcombine.c (pass_tree_ifcombine): Convert from a global
10469 struct to a subclass of gimple_opt_pass along with...
10470 (pass_data_tree_ifcombine): ...new pass_data instance and...
10471 (make_pass_tree_ifcombine): ...new function.
10472 * tree-ssa-loop-ch.c (pass_ch): Convert from a global struct to a
10473 subclass of gimple_opt_pass along with...
10474 (pass_data_ch): ...new pass_data instance and...
10475 (make_pass_ch): ...new function.
10476 * tree-ssa-loop.c (pass_tree_loop): Convert from a global struct to a
10477 subclass of gimple_opt_pass along with...
10478 (pass_data_tree_loop): ...new pass_data instance and...
10479 (make_pass_tree_loop): ...new function.
10480 (pass_tree_loop_init): Convert from a global struct to a subclass of
10481 gimple_opt_pass along with...
10482 (pass_data_tree_loop_init): ...new pass_data instance and...
10483 (make_pass_tree_loop_init): ...new function.
10484 (pass_lim): Convert from a global struct to a subclass of
10485 gimple_opt_pass along with...
10486 (pass_data_lim): ...new pass_data instance and...
10487 (make_pass_lim): ...new function.
10488 (pass_tree_unswitch): Convert from a global struct to a subclass of
10489 gimple_opt_pass along with...
10490 (pass_data_tree_unswitch): ...new pass_data instance and...
10491 (make_pass_tree_unswitch): ...new function.
10492 (pass_predcom): Convert from a global struct to a subclass of
10493 gimple_opt_pass along with...
10494 (pass_data_predcom): ...new pass_data instance and...
10495 (make_pass_predcom): ...new function.
10496 (pass_vectorize): Convert from a global struct to a subclass of
10497 gimple_opt_pass along with...
10498 (pass_data_vectorize): ...new pass_data instance and...
10499 (make_pass_vectorize): ...new function.
10500 (pass_graphite): Convert from a global struct to a subclass of
10501 gimple_opt_pass along with...
10502 (pass_data_graphite): ...new pass_data instance and...
10503 (make_pass_graphite): ...new function.
10504 (pass_graphite_transforms): Convert from a global struct to a subclass
10505 of gimple_opt_pass along with...
10506 (pass_data_graphite_transforms): ...new pass_data instance and...
10507 (make_pass_graphite_transforms): ...new function.
10508 (pass_check_data_deps): Convert from a global struct to a subclass of
10509 gimple_opt_pass along with...
10510 (pass_data_check_data_deps): ...new pass_data instance and...
10511 (make_pass_check_data_deps): ...new function.
10512 (pass_iv_canon): Convert from a global struct to a subclass of
10513 gimple_opt_pass along with...
10514 (pass_data_iv_canon): ...new pass_data instance and...
10515 (make_pass_iv_canon): ...new function.
10516 (pass_scev_cprop): Convert from a global struct to a subclass of
10517 gimple_opt_pass along with...
10518 (pass_data_scev_cprop): ...new pass_data instance and...
10519 (make_pass_scev_cprop): ...new function.
10520 (pass_record_bounds): Convert from a global struct to a subclass of
10521 gimple_opt_pass along with...
10522 (pass_data_record_bounds): ...new pass_data instance and...
10523 (make_pass_record_bounds): ...new function.
10524 (pass_complete_unroll): Convert from a global struct to a subclass of
10525 gimple_opt_pass along with...
10526 (pass_data_complete_unroll): ...new pass_data instance and...
10527 (make_pass_complete_unroll): ...new function.
10528 (pass_complete_unrolli): Convert from a global struct to a subclass of
10529 gimple_opt_pass along with...
10530 (pass_data_complete_unrolli): ...new pass_data instance and...
10531 (make_pass_complete_unrolli): ...new function.
10532 (pass_parallelize_loops): Convert from a global struct to a subclass
10533 of gimple_opt_pass along with...
10534 (pass_data_parallelize_loops): ...new pass_data instance and...
10535 (make_pass_parallelize_loops): ...new function.
10536 (pass_loop_prefetch): Convert from a global struct to a subclass of
10537 gimple_opt_pass along with...
10538 (pass_data_loop_prefetch): ...new pass_data instance and...
10539 (make_pass_loop_prefetch): ...new function.
10540 (pass_iv_optimize): Convert from a global struct to a subclass of
10541 gimple_opt_pass along with...
10542 (pass_data_iv_optimize): ...new pass_data instance and...
10543 (make_pass_iv_optimize): ...new function.
10544 (pass_tree_loop_done): Convert from a global struct to a subclass of
10545 gimple_opt_pass along with...
10546 (pass_data_tree_loop_done): ...new pass_data instance and...
10547 (make_pass_tree_loop_done): ...new function.
10548 * tree-ssa-math-opts.c (pass_cse_reciprocals): Convert from a global
10549 struct to a subclass of gimple_opt_pass along with...
10550 (pass_data_cse_reciprocals): ...new pass_data instance and...
10551 (make_pass_cse_reciprocals): ...new function.
10552 (pass_cse_sincos): Convert from a global struct to a subclass of
10553 gimple_opt_pass along with...
10554 (pass_data_cse_sincos): ...new pass_data instance and...
10555 (make_pass_cse_sincos): ...new function.
10556 (pass_optimize_bswap): Convert from a global struct to a subclass of
10557 gimple_opt_pass along with...
10558 (pass_data_optimize_bswap): ...new pass_data instance and...
10559 (make_pass_optimize_bswap): ...new function.
10560 (pass_optimize_widening_mul): Convert from a global struct to a
10561 subclass of gimple_opt_pass along with...
10562 (pass_data_optimize_widening_mul): ...new pass_data instance and...
10563 (make_pass_optimize_widening_mul): ...new function.
10564 * tree-ssa-phiopt.c (pass_phiopt): Convert from a global struct to a
10565 subclass of gimple_opt_pass along with...
10566 (pass_data_phiopt): ...new pass_data instance and...
10567 (make_pass_phiopt): ...new function.
10568 (pass_cselim): Convert from a global struct to a subclass of
10569 gimple_opt_pass along with...
10570 (pass_data_cselim): ...new pass_data instance and...
10571 (make_pass_cselim): ...new function.
10572 * tree-ssa-phiprop.c (pass_phiprop): Convert from a global struct to a
10573 subclass of gimple_opt_pass along with...
10574 (pass_data_phiprop): ...new pass_data instance and...
10575 (make_pass_phiprop): ...new function.
10576 * tree-ssa-pre.c (pass_pre): Convert from a global struct to a
10577 subclass of gimple_opt_pass along with...
10578 (pass_data_pre): ...new pass_data instance and...
10579 (make_pass_pre): ...new function.
10580 (pass_fre): Convert from a global struct to a subclass of
10581 gimple_opt_pass along with...
10582 (pass_data_fre): ...new pass_data instance and...
10583 (make_pass_fre): ...new function.
10584 * tree-ssa-reassoc.c (pass_reassoc): Convert from a global struct to a
10585 subclass of gimple_opt_pass along with...
10586 (pass_data_reassoc): ...new pass_data instance and...
10587 (make_pass_reassoc): ...new function.
10588 * tree-ssa-sink.c (pass_sink_code): Convert from a global struct to a
10589 subclass of gimple_opt_pass along with...
10590 (pass_data_sink_code): ...new pass_data instance and...
10591 (make_pass_sink_code): ...new function.
10592 * tree-ssa-strlen.c (pass_strlen): Convert from a global struct to a
10593 subclass of gimple_opt_pass along with...
10594 (pass_data_strlen): ...new pass_data instance and...
10595 (make_pass_strlen): ...new function.
10596 * tree-ssa-structalias.c (pass_build_alias): Convert from a global
10597 struct to a subclass of gimple_opt_pass along with...
10598 (pass_data_build_alias): ...new pass_data instance and...
10599 (make_pass_build_alias): ...new function.
10600 (pass_build_ealias): Convert from a global struct to a subclass of
10601 gimple_opt_pass along with...
10602 (pass_data_build_ealias): ...new pass_data instance and...
10603 (make_pass_build_ealias): ...new function.
10604 (pass_ipa_pta): Convert from a global struct to a subclass of
10605 simple_ipa_opt_pass along with...
10606 (pass_data_ipa_pta): ...new pass_data instance and...
10607 (make_pass_ipa_pta): ...new function.
10608 * tree-ssa-uncprop.c (pass_uncprop): Convert from a global struct to a
10609 subclass of gimple_opt_pass along with...
10610 (pass_data_uncprop): ...new pass_data instance and...
10611 (make_pass_uncprop): ...new function.
10612 * tree-ssa-uninit.c (pass_late_warn_uninitialized): Convert from a
10613 global struct to a subclass of gimple_opt_pass along with...
10614 (pass_data_late_warn_uninitialized): ...new pass_data instance and...
10615 (make_pass_late_warn_uninitialized): ...new function.
10616 * tree-ssa.c (pass_init_datastructures): Convert from a global struct
10617 to a subclass of gimple_opt_pass along with...
10618 (pass_data_init_datastructures): ...new pass_data instance and...
10619 (make_pass_init_datastructures): ...new function.
10620 (pass_early_warn_uninitialized): Convert from a global struct to a
10621 subclass of gimple_opt_pass along with...
10622 (pass_data_early_warn_uninitialized): ...new pass_data instance and...
10623 (make_pass_early_warn_uninitialized): ...new function.
10624 (pass_update_address_taken): Convert from a global struct to a
10625 subclass of gimple_opt_pass along with...
10626 (pass_data_update_address_taken): ...new pass_data instance and...
10627 (make_pass_update_address_taken): ...new function.
10628 * tree-ssanames.c (pass_release_ssa_names): Convert from a global
10629 struct to a subclass of gimple_opt_pass along with...
10630 (pass_data_release_ssa_names): ...new pass_data instance and...
10631 (make_pass_release_ssa_names): ...new function.
10632 * tree-stdarg.c (pass_stdarg): Convert from a global struct to a
10633 subclass of gimple_opt_pass along with...
10634 (pass_data_stdarg): ...new pass_data instance and...
10635 (make_pass_stdarg): ...new function.
10636 * tree-switch-conversion.c (pass_convert_switch): Convert from a
10637 global struct to a subclass of gimple_opt_pass along with...
10638 (pass_data_convert_switch): ...new pass_data instance and...
10639 (make_pass_convert_switch): ...new function.
10640 * tree-tailcall.c (pass_tail_recursion): Convert from a global struct
10641 to a subclass of gimple_opt_pass along with...
10642 (pass_data_tail_recursion): ...new pass_data instance and...
10643 (make_pass_tail_recursion): ...new function.
10644 (pass_tail_calls): Convert from a global struct to a subclass of
10645 gimple_opt_pass along with...
10646 (pass_data_tail_calls): ...new pass_data instance and...
10647 (make_pass_tail_calls): ...new function.
10648 * tree-vect-generic.c (pass_lower_vector): Convert from a global
10649 struct to a subclass of gimple_opt_pass along with...
10650 (pass_data_lower_vector): ...new pass_data instance and...
10651 (make_pass_lower_vector): ...new function.
10652 (pass_lower_vector_ssa): Convert from a global struct to a subclass of
10653 gimple_opt_pass along with...
10654 (pass_data_lower_vector_ssa): ...new pass_data instance and...
10655 (make_pass_lower_vector_ssa): ...new function.
10656 * tree-vectorizer.c (pass_slp_vectorize): Convert from a global struct
10657 to a subclass of gimple_opt_pass along with...
10658 (pass_data_slp_vectorize): ...new pass_data instance and...
10659 (make_pass_slp_vectorize): ...new function.
10660 (pass_ipa_increase_alignment): Convert from a global struct to a
10661 subclass of simple_ipa_opt_pass along with...
10662 (pass_data_ipa_increase_alignment): ...new pass_data instance and...
10663 (make_pass_ipa_increase_alignment): ...new function.
10664 * tree-vrp.c (pass_vrp): Convert from a global struct to a subclass of
10665 gimple_opt_pass along with...
10666 (pass_data_vrp): ...new pass_data instance and...
10667 (make_pass_vrp): ...new function.
10668 * tree.c (pass_ipa_free_lang_data): Convert from a global struct to a
10669 subclass of simple_ipa_opt_pass along with...
10670 (pass_data_ipa_free_lang_data): ...new pass_data instance and...
10671 (make_pass_ipa_free_lang_data): ...new function.
10672 * tsan.c (pass_tsan): Convert from a global struct to a subclass of
10673 gimple_opt_pass along with...
10674 (pass_data_tsan): ...new pass_data instance and...
10675 (make_pass_tsan): ...new function.
10676 (pass_tsan_O0): Convert from a global struct to a subclass of
10677 gimple_opt_pass along with...
10678 (pass_data_tsan_O0): ...new pass_data instance and...
10679 (make_pass_tsan_O0): ...new function.
10680 * var-tracking.c (pass_variable_tracking): Convert from a global
10681 struct to a subclass of rtl_opt_pass along with...
10682 (pass_data_variable_tracking): ...new pass_data instance and...
10683 (make_pass_variable_tracking): ...new function.
10684 * web.c (pass_web): Convert from a global struct to a subclass of
10685 rtl_opt_pass along with...
10686 (pass_data_web): ...new pass_data instance and...
10687 (make_pass_web): ...new function.
10688 * config/epiphany/epiphany.h (pass_mode_switch_use): Replace
10689 declaration with that of...
10690 (make_pass_mode_switch_use): ...new function.
10691 (pass_resolve_sw_modes): Replace declaration with that of...
10692 (make_pass_resolve_sw_modes): ...new function.
10693 * config/epiphany/mode-switch-use.c (pass_mode_switch_use): Convert
10694 from a global struct to a subclass of rtl_opt_pass along with...
10695 (pass_data_mode_switch_use): ...new pass_data instance and...
10696 (make_pass_mode_switch_use): ...new function.
10697 * config/epiphany/resolve-sw-modes.c (pass_resolve_sw_modes): Convert
10698 from a global struct to a subclass of rtl_opt_pass along with...
10699 (pass_data_resolve_sw_modes): ...new pass_data instance and...
10700 (make_pass_resolve_sw_modes): ...new function.
10701 * config/i386/i386.c (pass_insert_vzeroupper): Convert from a global
10702 struct to a subclass of rtl_opt_pass along with...
10703 (pass_data_insert_vzeroupper): ...new pass_data instance and...
10704 (make_pass_insert_vzeroupper): ...new function.
10705 * config/sparc/sparc.c (pass_work_around_errata): Convert from a
10706 global struct to a subclass of rtl_opt_pass along with...
10707 (pass_data_work_around_errata): ...new pass_data instance and...
10708 (make_pass_work_around_errata): ...new function.
10709 * config/mips/mips.c (pass_mips_machine_reorg2): Convert from a global
10710 struct to a subclass of rtl_opt_pass along with...
10711 (pass_data_mips_machine_reorg2): ...new pass_data instance and...
10712 (make_pass_mips_machine_reorg2): ...new function.
10713
6a389ed5
DM		 107142013-08-05 David Malcolm <dmalcolm@redhat.com>
10715
10716 * passes.c (pass_manager::operator new): New.
10717
f7695dbf
DM		 107182013-08-05 David Malcolm <dmalcolm@redhat.com>
10719
10720 Handwritten part of conversion of passes to C++ classes.
10721
10722 * Makefile.in (PASS_MANAGER_H): Add dep on pass-instances.def.
10723 (toplev.o): Add dep on PASS_MANAGER_H.
10724 * cgraphunit.c (cgraph_process_new_functions): Rework invocation
10725 of early local pases to reflect this moving from a global to a
10726 member of gcc::pass_manager.
10727 (cgraph_add_new_function): Likewise.
10728 * lto-cgraph.c (lto_output_node): Update for conversion of
10729 struct ipa_opt_pass_d to a C++ subclass of opt_pass.
10730 * passes.c (opt_pass::clone): New.
10731 (opt_pass::gate): New.
10732 (opt_pass::execute): New.
10733 (opt_pass::opt_pass): New.
10734 (pass_manager::execute_early_local_passes): New.
10735 (pass_manager::execute_pass_mode_switching): new.
10736 (finish_optimization_passes): Convert to...
10737 (pass_manager::finish_optimization_passes): ...this.
10738 (finish_optimization_passes): Update for conversion of passes to
10739 C++ classes.
10740 (register_dump_files_1): Use has_gate since we cannot portably
10741 check a vtable entry against NULL.
10742 (dump_one_pass): Likewise.
10743 (ipa_write_summaries_2): Likewise.
10744 (ipa_write_optimization_summaries_1): Likewise.
10745 (ipa_read_summaries_1): Likewise.
10746 (ipa_read_optimization_summaries_1): Likewise.
10747 (execute_ipa_stmt_fixups): Likewise.
10748 (pass_manager::pass_manager): Rewrite pass-creation, invoking
10749 pass-creation functions rather than wiring up globals, and
10750 storing the results in fields of pass_manager generated using
10751 pass-instances.def.
10752 (pass_manager::dump_profile_report): Update for conversion of
10753 passes to C++ classes.
10754 (pass_manager::execute_ipa_summary_passes): Likewise.
10755 (execute_one_ipa_transform_pass): Likewise.
10756 (execute_one_pass): Use has_gate and has_execute since we cannot
10757 portably check a vtable entry against NULL.
10758 * pass_manager.h (pass_manager::finish_optimization_passes): New.
10759 (pass_manager): Use pass-instances.def to add fields for the
10760 various pass instances.
10761 * toplev.c (finalize): Update for move of
10762 finish_optimization_passes to a method of gcc::pass_manager.
10763 * toplev.h (finish_optimization_passes): Move to method of class
10764 pass_manager.
10765 * tree-pass.h (struct pass_data): New.
03b0ee0a 10766 (opt_pass): Convert to C++ class, make it a subclass of pass_data.
f7695dbf
DM		 10767 (opt_pass::gate): Convert to virtual function.
10768 (opt_pass::~opt_pass): New.
10769 (opt_pass::clone): New.
10770 (opt_pass::execute): Convert to virtual function.
10771 (opt_pass::opt_pass): New.
10772 (opt_pass::ctxt_): new.
10773 (gimple_opt_pass): Convert to subclass of opt_pass.
10774 (gimple_opt_pass::gimple_opt_pass): New.
10775 (rtl_opt_pass): Convert to subclass of opt_pass.
10776 (rtl_opt_pass::rtl_opt_pass): New.
10777 (ipa_opt_pass_d): Convert to subclass of opt_pass.
10778 (ipa_opt_pass_d::ipa_opt_pass_d): New.
10779 (simple_ipa_opt_pass): Convert to subclass of opt_pass.
10780 (simple_ipa_opt_pass::simple_ipa_opt_pass): New.
10781 * config/i386/i386.c (rest_of_handle_insert_vzeroupper): Rework
10782 invocation of pass_mode_switching to reflect this moving from a
10783 global to a member of gcc::pass_manager.
10784 (ix86_option_override): Rework how pass_insert_vzeroupper is
10785 added to the pass_manager to reflect autogenerated changes.
03b0ee0a 10786 * config/i386/t-i386 (i386.o) Add deps on CONTEXT_H and PASS_MANAGER_H.
f7695dbf 10787
d0b2f831
RE		 107882013-08-05 Richard Earnshaw <rearnsha@arm.com>
10789
10790 PR rtl-optimization/57708
10791 * recog.c (peep2_find_free_register): Validate all regs in a
10792 multi-reg mode.
10793
51a5c0c2
JH		 107942013-08-05 Jan Hubicka <jh@suse.cz>
10795
10796 PR lto/57602
03b0ee0a
UB		 10797 * cgraph.c (verify_cgraph_node): Accept local flags from other
10798 partitions.
51a5c0c2
JH		 10799 * ipa.c (symtab_remove_unreachable_nodes): Do not clear local flag.
10800 (function_and_variable_visibility): Likewise.
10801 * trans-mem.c (ipa_tm_create_version): TM versions are not local.
10802
65f0a120
GDR		 108032013-08-05 Gabriel Dos Reis <gdr@integrable-solutions.net>
10804
10805 * graph.c (init_graph_slim_pretty_print): Remove.
10806 (print_graph_cfg): Do not call it. Use local pretty printer.
10807 (start_graph_dump): Likewise.
10808
b3f80694
GDR		 108092013-08-05 Gabriel Dos Reis <gdr@integrable-solutions.net>
10810
10811 * gimple-pretty-print.c (buffer): Remove.
10812 (initialized): Likewise.
10813 (maybe_init_pretty_print): Likewise.
10814 (print_gimple_stmt): Do not call it. Use non-static local
10815 pretty_printer variable.
10816 (print_gimple_expr): Likewise.
10817 (print_gimple_seq): Likewise.
10818 (gimple_dump_bb): Likewise.
10819
11a877b3
GDR		 108202013-08-05 Gabriel Dos Reis <gdr@integrable-solutions.net>
10821
10822 * asan.c (asan_pp): Remove.
10823 (asan_pp_initialized): Likewise.
10824 (asan_pp_initialize): Likewise.
10825 (asan_pp_string): Take a pretty_printer parameter. Adjust callers.
10826 (asan_emit_stack_protection): Tidy. Use local pretty printer.
10827 (asan_add_global): Likewise.
10828
b066401f
GDR		 108292013-08-04 Gabriel Dos Reis <gdr@integrable-solutions.net>
10830
10831 * pretty-print.h (pp_base): Remove. Adjust dependent macros.
10832 * diagnostic.h (diagnostic_flush_buffer): Adjust.
10833 * pretty-print.c (pp_formatted_text_data): Likewise.
10834 (pp_indent): Rename from pp_base_indent.
10835 (pp_format): Rename from pp_base_format.
10836 (pp_output_formatted_text): Rename from pp_base_output_formatted_text.
10837 (pp_format_verbatim): Rename from pp_base_format_verbatim.
10838 (pp_flush): Rename from pp_base_flush.
10839 (pp_set_line_maximum_length): Rename from
10840 pp_base_set_line_maximum_length.
10841 (pp_clear_output_area): Rename from pp_base_clear_output_area.
10842 (pp_set_prefix): Rename from pp_base_set_prefix.
10843 (pp_destroy_prefix): Rename from pp_base_destroy_prefix.
10844 (pp_emit_prefix): Rename from pp_base_emit_prefix.
10845 (pp_append_text): Rename from pp_base_append_text.
10846 (pp_formatted_text): Rename from pp_base_formatted_text.
10847 (pp_last_position_in_text): Rename from pp_base_last_position_in_text.
10848 (pp_remaining_character_count_for_line): Rename from
10849 pp_base_remaining_character_count_for_line.
10850 (pp_newline): Rename from pp_base_newline.
10851 (pp_character): Rename from pp_base_character.
10852 (pp_string): Rename from pp_base_string.
10853 (pp_maybe_space): Rename from pp_base_maybe_space.
10854 * asan.c (asan_pp_string): Adjust.
10855 (asan_emit_stack_protection): Likewise.
10856 (asan_add_global): Likewise.
10857 * sched-vis.c (str_pattern_slim): Adjust pretty printer function call.
10858 * tree-mudflap.c (mf_varname_tree): Likewise.
10859 * tree-pretty-print.c (pp_tree_identifier): Rename from
10860 pp_base_tree_identifier.
10861 * tree-pretty-print.h (pp_tree_identifier): Remove macro definition.
10862 Declare as function.
10863
137a1a27
GDR		 108642013-08-03 Gabriel Dos Reis <gdr@integrable-solutions.net>
10865
10866 * pretty-print.h (pp_bar_bar): New.
10867 (pp_ampersand_ampersand): Likewise.
10868 (pp_less_equal): Likewise.
10869 (pp_greater_equal): Likewise.
10870 * gimple-pretty-print.c (dump_ternary_rhs): Use specialized pretty
10871 printer functions instead of pp_string or operators and punctuators.
10872 (dump_gimple_call): Likewise.
10873 (dump_gimple_omp_for): Likewise.
10874 (dump_gimple_transaction): Likewise.
10875 (dump_gimple_phi): Likewise.
10876 (pp_gimple_stmt_1): Likewise.
10877 * sched-vis.c (print_insn): Likewise.
10878 * tree-mudflap.c (mf_varname_tree): Likewise.
10879 * tree-pretty-print.c (dump_block_node): Likewise.
10880 (dump_generic_node): Likewise.
10881
815effe1
JH		 108822013-08-02 Jan Hubicka <jh@suse.cz>
10883
10884 * lto-cgraph.c (compute_ltrans_boundary): Add abstract origins into
10885 boundaries.
10886 * lto-streamer-out.c (tree_is_indexable): Results decls and
10887 parm decls are not indexable.
10888 (DFS_write_tree_body): Do not follow args and results.
10889 (hash_tree): Likewise.
10890 (output_functions): Rearrange so struct function is needed
10891 only when real body is output; be able to also ouptut abstract
10892 functions; output DECL_ARGUMENTS and DECL_RESULT.
10893 (lto_output): When not in WPA, ale store abstract functions.
10894 (write_symbol): Do not care about RESULT_DECL.
10895 (output_symbol_p): Handle correctly sbtract decls.
10896 * lto-streamer-in.c (input_function): Rearrange so struct
10897 function can be NULL at entry; allow streaming of
10898 functions w/o body; store DECL_ARGUMENTS and DECL_RESULT.
10899 * ipa.c (symtab_remove_unreachable_nodes): Silence confused
10900 sanity check during LTO.
10901 * tree-streamer-out.c (write_ts_decl_non_common_tree_pointers): Skip
10902 RESULT_DECl and DECL_ARGUMENTS.
10903 * tree-streamer-in.c (lto_input_ts_decl_non_common_tree_pointers):
10904 Likewise.
10905
07838b13
GDR		 109062013-08-03 Gabriel Dos Reis <gdr@integrable-solutions.net>
10907
10908 * pretty-print.h (pp_underscore): New.
10909 (pp_comma): Tidy.
10910 * gimple-pretty-print.c (dump_unary_rhs): Use specialized pretty
10911 printer functions instead of pp_character.
10912 (dump_binary_rhs): Likewise.
10913 (dump_ternary_rhs): Likewise.
10914 (dump_gimple_call_args): Likewise.
10915 (pp_points_to_solution): Likewise.
10916 (dump_gimple_call): Likewise.
10917 (dump_gimple_switch): Likewise.
10918 (dump_gimple_cond): Likewise.
10919 (dump_gimple_bind): Likewise.
10920 (dump_gimple_try): Likewise.
10921 (dump_gimple_omp_for): Likewise.
10922 (dump_gimple_omp_continue): Likewise.
10923 (dump_gimple_omp_single): Likewise.
10924 (dump_gimple_omp_sections): Likewise.
10925 (dump_gimple_omp_block): Likewise.
10926 (dump_gimple_omp_critical): Likewise.
10927 (dump_gimple_transaction): Likewise.
10928 (dump_gimple_asm): Likewise.
10929 (dump_gimple_phi): Likewise.
10930 (dump_gimple_omp_parallel): Likewise.
10931 (dump_gimple_omp_task): Likewise.
10932 (dump_gimple_omp_atomic_load): Likewise.
10933 (dump_gimple_omp_atomic_store): Likewise.
10934 (dump_gimple_mem_ops): Likewise.
10935 (pp_gimple_stmt_1): Likewise.
10936 (pp_cfg_jump): Likewise.
10937 (dump_implicit_edges): Likewise.
10938 (gimple_dump_bb_for_graph): Likewise.
10939 * graph.c (draw_cfg_node): Likewise.
10940 * langhooks.c (lhd_print_error_function): Likewise.
10941 * sched-vis.c (print_exp): Likewise.
10942 (print_value): Likewise.
10943 (print_pattern): Likewise.
10944 (print_insn): Likewise.
10945 (rtl_dump_bb_for_graph): Likewise.
10946 * tree-pretty-print.c (dump_function_declaration): Likewise.
10947 (dump_array_domain): Likewise.
10948 (dump_omp_clause): Likewise.
10949 (dump_location): Likewise.
10950 (dump_generic_node): Likewise.
10951 (print_struct_decl): Likewise.
10952 * diagnostic.c (diagnostic_show_locus): Use pp_space.
10953
bb0d2039
BS		 109542013-08-03 Bill Schmidt <wschmidt@vnet.linux.ibm.com>
10955
10956 * gimple-ssa-strength-reduction.c (replace_mult_candidate): Update
10957 candidate table when replacing a candidate statement.
10958 (replace_rhs_if_not_dup): Likewise.
10959 (replace_one_candidate): Likewise.
10960
bc0ec027 109612013-08-02 Jan Hubicka <jh@suse.cz>
cdaa808a 10962 Martin Liska <marxin.liska@gmail.com>
bc0ec027
JH		 10963
10964 * cgraphunit.c (add_new_function): Fix logic when adding from
10965 late IPA pass.
10966 (assemble_thunk): Rename to ...
03b0ee0a
UB		 10967 (expand_thunk); .. this one; export; get it working with
10968 general functions; make produced gimple valid.
bc0ec027
JH		 10969 * cgraph.h (expand_thunk): Declare.
10970
0e8853ee
JH		 109712013-08-02 Jan Hubicka <jh@suse.cz>
10972
03b0ee0a
UB		 10973 * ipa-cp.c (gather_context_independent_values): Use
10974 ipa_get_param_move_cost.
10975 (get_replacement_map): Remove PARAM; move parameter folding
10976 into tree-inline.c
0e8853ee
JH		 10977 (create_specialized_node): Update.
10978 * ipa-prop.c (ipa_populate_param_decls): Do not look for origins;
10979 assert that we have gimple body; update move_cost.
10980 (count_formal_params): Assert that we have gimple body.
10981 (ipa_dump_param): New function.
10982 (ipa_alloc_node_params): Break out from ...
10983 (ipa_initialize_node_params): ... here.
10984 (ipa_get_vector_of_formal_parms): ICE when used in WPA.
10985 (ipa_write_node_info): Stream move costs.
10986 (ipa_read_node_info): Read move costs.
10987 (ipa_update_after_lto_read): Do not recompute node params.
10988 * ipa-prop.h (ipa_param_descriptor): Add move_cost.
10989 (ipa_get_param): Check we are not in WPA.
10990 (ipa_get_param_move_cost): New.
10991 * tree-inline.c (tree_function_versioning): Fold replacement as needed.
10992 * ipa-inline-analysis.c (inline_node_duplication_hook): Expect only
10993 parm numbers to be present.
10994
ecee672b
VM		 109952013-08-02 Vladimir Makarov <vmakarov@redhat.com>
10996
10997 PR rtl-optimization/58048
10998 * lra-constraints.c (process_alt_operands): Don't check asm
10999 operand on register.
11000
13ec0527
EB		 110012013-08-02 Eric Botcazou <ebotcazou@adacore.com>
11002
11003 * config/sparc/sparc.c (sparc_emit_membar_for_model) <SMM_TSO>: Add
11004 the implied StoreLoad barrier for atomic operations if before.
11005
af15184a
JH		 110062013-08-02 Jan Hubicka <jh@suse.cz>
11007 Martin Liska <marxin.liska@gmail.com>
11008
03b0ee0a
UB		 11009 * cgraph.c (cgraph_function_body_availability): Do not check
11010 cgraph flags.
af15184a
JH		 11011 * cgraph.h (symtab_for_node_and_aliases, symtab_nonoverwritable_alias,
11012 symtab_node_availability): Declare.
11013 * ipa.c (can_replace_by_local_alias): New.
11014 (function_and_variable_visibility): Use it.
03b0ee0a
UB		 11015 * symtab.c (symtab_for_node_and_aliases,
11016 symtab_nonoverwritable_alias_1, symtab_nonoverwritable_alias): New.
af15184a 11017
01e54ef8
VM		 110182013-08-02 Vladimir Makarov <vmakarov@redhat.com>
11019
11020 PR rtl-optimization/57963
03b0ee0a 11021 * lra-constraints.c (reverse_equiv_p, contains_reloaded_insn_p): New.
01e54ef8
VM		 11022 (lra_constraints): Use them.
11023
e2fd7ca7
SN		 110242013-08-02 Sofiane Naci <sofiane.naci@arm.com>
11025
03b0ee0a
UB		 11026 * config/arm/types.md (define_attr "type"): Add "load_acq"
11027 and "store_rel".
e2fd7ca7
SN		 11028 * config/arm/cortex-a53.md (cortex_a53_load1): Update for attribute
11029 changes.
11030 (cortex_a53_store1): Likewise.
11031
0dcc26c3
JH		 110322013-08-01 Jan Hubicka <jh@suse.cz>
11033
03b0ee0a
UB		 11034 * ipa.c (symtab_remove_unreachable_nodes): Nodes in other
11035 partitions are not needed.
0dcc26c3 11036
0bd72901
UB		 110372013-08-01 Uros Bizjak <ubizjak@gmail.com>
11038
11039 * config/i386/i386.h (MAYBE_NON_Q_CLASS_P): New.
11040 * config/i386/i386.c (ix86_secondary_reload): Use INTEGER_CLASS_P and
11041 MAYBE_NON_Q_CLASS_P where appropriate.
11042
12123452
JH		 110432013-08-01 Jan Hubicka <jh@suse.cz>
11044
11045 * cgraph.h (release_function_body): Declare.
11046 * tree.c (free_lang_data_in_decl): Free, parameters and return values
11047 of unused delcarations.
11048
ce852f9c
KT		 110492013-08-01 Kyrylo Tkachov <kyrylo.tkachov@arm.com>
11050
03b0ee0a
UB		 11051 * config/arm/arm.md (minmax_arithsi_non_canon): Emit canonical
11052 RTL form when subtracting a constant.
ce852f9c 11053
c743b246
KT		 110542013-08-01 Kyrylo Tkachov <kyrylo.tkachov@arm.com>
11055
11056 * config/arm/arm.md (peepholes for eq (reg1) (reg2/imm)):
11057 Generate canonical plus rtx with negated immediate instead of minus
11058 where appropriate.
11059 * config/arm/arm.c (thumb2_reorg): Handle ADCS <Rd>, <Rn> case.
11060
c0c123ef
JH		 110612013-08-01 Jan Hubicka <jh@suse.cz>
11062
11063 * cgraph.c (cgraph_release_function_body): Use used_as_abstract_origin.
11064 (cgraph_release_function_body): Likewise.
11065 (cgraph_can_remove_if_no_direct_calls_p): Likewise.
11066 * cgraph.h (cgrpah_node): Rename abstract_and_needed
11067 to used_as_abstract_origin.
11068 * tree-inline-transfrom.c (can_remove_node_now_p_1): Do not remove
11069 symbols used as abstract origins.
11070 * cgraphunit.c (analyze_functions): Update.
0bd72901
UB		 11071 * ipa.c (symtab_remove_unreachable_nodes): Recompute
11072 used_as_abstract_origin.
c0c123ef 11073 * tree-inline.c (tree_function_versioning): Update
0bd72901
UB		 11074 used_as_abstract_origin; be ready for DECL_RESULT and
11075 DECL_ARGUMENTS to be NULL.
c0c123ef 11076
0bd72901
UB		 11077 * lto-symtab.c (lto_symtab_merge_symbols): Merge duplicated nodes
11078 for abstract functions.
11079 * cgraph.h (symtab_real_symbol_p): Abstract declarations are not
11080 real symbols.
c0c123ef 11081
1f029433
JH		 110822013-08-01 Jan Hubicka <jh@suse.cz>
11083
11084 * profile.c (compute_value_histograms): Fix thinko.
11085
c451f4d6
SN		 110862013-08-01 Sofiane Naci <sofiane.naci@arm.com>
11087
11088 * config.gcc (aarch64*-*-*): Add aarch-common.o to extra_objs. Add
11089 aarch-common-protos.h to extra_headers.
11090 (aarch64*-*-*): Add arm/aarch-common-protos.h to tm_p_file.
11091 * config/aarch64/aarch64.md: Include "../arm/cortex-a53.md".
11092 * config/aarch64/t-aarch64 (aarch-common.o): Define.
11093
34b512f5
SN		 110942013-08-01 Sofiane Naci <sofiane.naci@arm.com>
11095
11096 * config/aarch64/aarch64.md (define_attr "type"): Delete.
11097 Include "../arm/types.md". Define "type" attribute for all patterns.
11098 * config/aarch64/aarch64-simd.md (move_lo_quad_<mode>): Update for
11099 attribute changes.
11100
d86e633a
MM		 111012013-07-31 Michael Meissner <meissner@linux.vnet.ibm.com>
11102
11103 * config/rs6000/predicates.md (fusion_gpr_addis): New predicates
11104 to support power8 load fusion.
11105 (fusion_gpr_mem_load): Likewise.
11106
11107 * config/rs6000/rs6000-modes.def (PTImode): Update a comment.
11108
11109 * config/rs6000/rs6000-protos.h (fusion_gpr_load_p): New
11110 declarations for power8 load fusion.
11111 (emit_fusion_gpr_load): Likewise.
11112
11113 * config/rs6000/rs6000.c (rs6000_option_override_internal): If
11114 tuning for power8, turn on fusion mode by default. Turn on sign
11115 extending fusion mode if normal fusion mode is on, and we are at
11116 -O2 or -O3.
11117 (fusion_gpr_load_p): New function, return true if we can fuse an
11118 addis instruction with a dependent load to a GPR.
11119 (emit_fusion_gpr_load): Emit the instructions for power8 load
11120 fusion to GPRs.
11121
0bd72901 11122 * config/rs6000/vsx.md (VSX_M2): New iterator for fusion peepholes.
d86e633a
MM		 11123 (VSX load fusion peepholes): New peepholes to fuse together an
11124 addi instruction with a VSX load instruction.
11125
11126 * config/rs6000/rs6000.md (GPR load fusion peepholes): New
11127 peepholes to fuse an addis instruction with a load to a GPR base
11128 register. If we are supporting sign extending fusions, convert
11129 sign extending loads to zero extending loads and add an explicit
11130 sign extension.
11131
c3f35647
SN		 111322013-07-31 Sofiane Naci <sofiane.naci@arm.com>
11133
11134 * config.gcc (arm*-*-*): Add aarch-common.o to extra_objs. Add
11135 aarch-common-protos.h to extra_headers.
11136 (arm*-*-*): Add arm/aarch-common-protos.h to tm_p_file.
11137 * config/arm/arm.c (arm_early_load_addr_dep): Move from here to ...
11138 (arm_early_store_addr_dep): Likewise.
0bd72901
UB		 11139 (arm_no_early_alu_shift_dep): Likewise.
11140 (arm_no_early_alu_shift_value_dep): Likewise.
11141 (arm_no_early_mul_dep): Likewise.
11142 (arm_no_early_store_addr_dep): Likewise.
11143 (arm_mac_accumulator_is_mul_result): Likewise.
11144 (arm_mac_accumulator_is_result): Likewise.
c3f35647 11145 * config/arm/aarch-common.c: ... here. New file.
0bd72901
UB		 11146 * config/arm/arm-protos.h (arm_early_load_addr_dep): Move from
11147 here to ...
c3f35647 11148 (arm_early_store_addr_dep): Likewise.
0bd72901
UB		 11149 (arm_no_early_alu_shift_dep): Likewise.
11150 (arm_no_early_alu_shift_value_dep): Likewise.
11151 (arm_no_early_mul_dep): Likewise.
11152 (arm_no_early_store_addr_dep): Likewise.
11153 (arm_mac_accumulator_is_mul_result): Likewise.
11154 (arm_mac_accumulator_is_result): Likewise.
c3f35647
SN		 11155 * config/arm/aarch-common-protos.h: ... here. New file.
11156 * config/arm/t-arm (aarch-common.o): Define.
11157
111582013-07-31 Sofiane Naci <sofiane.naci@arm.com>
11159
11160 * config/arm/arm.md: Include new file "types.md".
11161 (define_attr "type"): Move from here to ...
11162 (define_attr "mul32"): Likewise.
11163 (define_attr "mul64"): Likewise.
11164 * config/arm/types.md: ... here. New file.
11165
228c1313
SH		 111662013-07-31 Sebastian Huber <sebastian.huber@embedded-brains.de>
11167
11168 * config.gcc (*-*-rtems*): Use __cxa_atexit by default.
11169 * config/rs6000/rtems.h (TARGET_LIBGCC_SDATA_SECTION): Define.
11170
d6e9046d 111712013-07-31 Jan-Benedict Glaw <jbglaw@lug-owl.de>
8ac499bb
JBG		 11172
11173 * gen-pass-instances.awk: Fix offset of substr().
11174
a167b052
DM		 111752013-07-31 David Malcolm <dmalcolm@redhat.com>
11176
11177 * Makefile.in (pass-instances.def): New.
11178 (passes.o): Replace dependency on passes.def with one on
11179 pass-instances.def
11180
11181 * gen-pass-instances.awk: New.
11182
11183 * passes.c (pass_manager::pass_manager): Use pass-instances.def
11184 rather than passes.def, updating local definition of NEXT_PASS
11185 macro to add an extra NUM parameter (currently unused).
11186
315f8c0e
DM		 111872013-07-30 David Malcolm <dmalcolm@redhat.com>
11188
11189 * Makefile.in (PASS_MANAGER_H): New.
11190 (lto-cgraph.o): Depend on CONTEXT_H and PASS_MANAGER_H.
11191 (passes.o): Likewise.
11192 (statistics.o): Likewise.
11193 (cgraphunit.o): Likewise.
11194 (context.o): Depend on PASS_MANAGER_H.
11195
11196 * pass_manager.h: New.
11197
11198 * cgraphunit.c (cgraph_add_new_function): Update for moves
11199 of globals to fields of pass_manager.
11200 (analyze_function): Likewise.
11201 (expand_function): Likewise.
11202 (ipa_passes): Likewise.
11203 (compile): Likewise.
11204
11205 * context.c (context::context): New.
11206 * context.h (context::context): New.
11207 (context::get_passes): New.
11208 (context::passes_): New.
11209
11210 * lto-cgraph.c (input_node): Update for moves of globals to
11211 fields of pass_manager.
11212
11213 * passes.c (all_passes): Remove, in favor of a field of the
11214 same name within the new class pass_manager.
11215 (all_small_ipa_passes): Likewise.
11216 (all_lowering_passes): Likewise.
11217 (all_regular_ipa_passes): Likewise.
11218 (all_late_ipa_passes): Likewise.
11219 (all_lto_gen_passes): Likewise.
11220 (passes_by_id): Likewise.
11221 (passes_by_id_size): Likewise.
11222 (gcc_pass_lists): Remove, in favor of "pass_lists" field within
11223 the new class pass_manager.
11224 (set_pass_for_id): Convert to...
11225 (pass_manager::set_pass_for_id): ...method.
11226 (get_pass_for_id): Convert to...
11227 (pass_manager::get_pass_for_id): ...method.
11228 (register_one_dump_file): Move body of implementation into...
11229 (pass_manager::register_one_dump_file): ...here.
11230 (register_dump_files_1): Convert to...
11231 (pass_manager::register_dump_files_1): ...method.
11232 (register_dump_files): Convert to...
11233 (pass_manager::register_dump_files): ...method.
11234 (create_pass_tab): Update for moves of globals to fields of
11235 pass_manager.
11236 (dump_passes): Move body of implementation into...
11237 (pass_manager::dump_passes): ...here.
11238 (register_pass): Move body of implementation into...
11239 (pass_manager::register_pass): ...here.
11240 (init_optimization_passes): Convert into...
11241 (pass_manager::pass_manager): ...constructor for new
11242 pass_manager class, and initialize the pass_lists array.
11243 (check_profile_consistency): Update for moves of globals to
11244 fields of pass_manager.
11245 (dump_profile_report): Move body of implementation into...
11246 (pass_manager::dump_profile_report): ...here.
11247 (ipa_write_summaries_1): Update for moves of pass lists from
11248 being globals to fields of pass_manager.
11249 (ipa_write_optimization_summaries): Likewise.
11250 (ipa_read_summaries): Likewise.
11251 (ipa_read_optimization_summaries): Likewise.
11252 (execute_all_ipa_stmt_fixups): Likewise.
11253
11254 * statistics.c (statistics_fini): Update for moves of globals to
11255 fields of pass_manager.
11256
11257 * toplev.c (general_init): Replace call to
11258 init_optimization_passes with construction of the pass_manager
11259 instance.
11260
11261 * tree-pass.h (all_passes): Remove, in favor of a field of the
11262 same name within the new class pass_manager.
11263 (all_small_ipa_passes): Likewise.
11264 (all_lowering_passes): Likewise.
11265 (all_regular_ipa_passes): Likewise.
11266 (all_lto_gen_passes): Likewise.
11267 (all_late_ipa_passes): Likewise.
11268 (passes_by_id): Likewise.
11269 (passes_by_id_size): Likewise.
11270 (gcc_pass_lists): Remove, in favor of "pass_lists" field within
11271 the new class pass_manager.
11272 (get_pass_for_id): Remove.
11273
02cab1c1
RE		 112742013-07-30 Richard Earnshaw <rearnsha@arm.com>
11275
11276 * config.gcc (arm): Require 64-bit host-wide-int for all ARM target
11277 configs.
11278
b2d3f886
RE		 112792013-07-30 Richard Earnshaw <rearnsha@arm.com>
11280
11281 * arm.md (mulhi3): New expand pattern.
11282
605e86fa
JH		 112832013-07-30 Jan Hubicka <jh@suse.cz>
11284 Martin Liska <marxin.liska@gmail.com>
11285
11286 * profile.c (compute_value_histograms): Do not ICE when
11287 there is mismatch only on some counters.
11288
14a87636
ZC		 112892013-07-30 Zhenqiang Chen <zhenqiang.chen@linaro.org>
11290
11291 PR rtl-optimization/57637
11292 * function.c (move_insn_for_shrink_wrap): Also check the
11293 GEN set of the LIVE problem for the liveness analysis
11294 if it exists, otherwise give up.
11295
0100cd3f
BS		 112962013-07-29 Bill Schmidt <wschmidt@vnet.linux.ibm.com>
11297
11298 PR tree-optimization/57993
11299 * gimple-ssa-strength-reduction.c (replace_mult_candidate): Record
11300 replaced statement in the candidate table.
11301 (phi_add_costs): Return infinite cost when the hidden basis does
11302 not dominate all phis on which the candidate is dependent.
11303 (replace_one_candidate): Record replaced statement in the
11304 candidate table.
11305
f0d811f7
JR		 113062013-07-29 Joern Rennecke <joern.rennecke@embecosm.com>
11307
11308 * config/epiphany/epiphany.md (*isub_i+2): New peephole.
11309 (ashlv2si3): New expander.
11310 (*ashlv2si3_i): New define_insn_and_split.
11311 * predicates.md (float_operation): Allow patterns with three
11312 basic sub-patterns.
11313
bba33211
JR		 11314 PR rtl-optimization/58021
11315 * mode-switching.c (create_pre_exit): Always split off preceding
11316 insns if we are not at the basic block head.
11317
ff3f3951
MR		 113182013-07-29 Maciej W. Rozycki <macro@codesourcery.com>
11319
11320 * config/mips/linux.h (GLIBC_DYNAMIC_LINKER): Handle `-mnan=2008'.
11321 (UCLIBC_DYNAMIC_LINKER): New macro.
11322 * config/mips/linux64.h (GLIBC_DYNAMIC_LINKER32): Handle
11323 `-mnan=2008'.
11324 (GLIBC_DYNAMIC_LINKER64, GLIBC_DYNAMIC_LINKERN32): Likewise.
11325 (UCLIBC_DYNAMIC_LINKER32): Undefine macro first. Handle
11326 `-mnan=2008'.
11327 (UCLIBC_DYNAMIC_LINKER64): Redefine macro.
11328 (UCLIBC_DYNAMIC_LINKERN32): Likewise.
11329 * config/mips/mips-modes.def: Remove RESET_FLOAT_FORMAT calls
11330 for SF and DF modes. Use ieee_quad_format for TF mode.
11331 * config/mips/mips-opts.h (mips_ieee_754_setting): New enum.
11332 * config/mips/mips.c (mips_file_start): Output a `.nan' directive.
11333 (mips_option_override): Handle `-mnan=legacy'.
11334 * config/mips/mips.h (TARGET_CPU_CPP_BUILTINS): Handle
11335 `-mabs=2008' and `-mnan=2008'.
11336 (OPTION_DEFAULT_SPECS): Add "nan" default.
11337 (ASM_SPEC): Handle `-mnan='.
11338 [!HAVE_AS_NAN] (HAVE_AS_NAN): New macro.
11339 * config/mips/mips.md (abs<mode>2): Handle `-mabs=2008', update
11340 comment accordingly.
11341 (neg<mode>2): Likewise.
11342 * config/mips/mips.opt (mabs, mnan): New options.
11343 * doc/install.texi (Configuration): Document `--with-nan=' option.
11344 * doc/invoke.texi (Option Summary): List MIPS `-mabs=' and
11345 `-mnan=' options.
11346 (MIPS Options): Document them.
11347 * config.gcc <mips*-*-*>: Handle `--with-nan='.
11348 * configure.ac <mips*-*-*>: Check for GAS `-mnan=2008' support.
11349 * configure: Regenerate.
11350 * config.in: Regenerate.
11351
576e2f37
UB		 113522013-07-29 Uros Bizjak <ubizjak@gmail.com>
11353
11354 * config/i386/i386.md (float post-reload splitters): Do not check
11355 for subregs of SSE registers.
11356
113572013-07-29 Uros Bizjak <ubizjak@gmail.com>
11358 H.J. Lu <hongjiu.lu@intel.com>
11359
11360 PR target/57954
11361 PR target/57988
11362 * config/i386/i386.md (post-reload splitter
11363 to avoid partial SSE reg dependency stalls): New pattern.
11364
a71f0749
DV		 113652013-07-29 Dominik Vogt <vogt@linux.vnet.ibm.com>
11366
11367 * config/s390/s390.md ("movcc"): Swap load and store instructions.
11368
4c97f1cc
JR		 113692013-07-27 Joern Rennecke <joern.rennecke@embecosm.com>
11370
11371 * config/epiphany/epiphany.c (epiphany_compute_frame_size):
11372 Also reserve space for saving UNKNOWN_REGNUM for leaf functions.
11373
b0c733d8
CC		 113742013-07-26 Cary Coutant <ccoutant@google.com>
11375
11376 * dwarf2out.c (die_checksum_ordered): Don't include template
11377 instantiations in signature.
11378 (is_template_parameter): New function.
11379 (is_template_instantiation): New function.
11380 (generate_skeleton_bottom_up): Don't include template instantiations
11381 in type unit DIE.
11382 (generate_skeleton): Likewise.
11383 (break_out_comdat_types): Move recursive call to break out nested
11384 types earlier.
11385 (prune_unused_types_mark_generic_parms_dies): Call
11386 is_template_parameter.
11387
040d8a1c
IB		 113882013-07-26 Ian Bolton <ian.bolton@arm.com>
11389
11390 * config/aarch64/aarch64.md (neg<mode>2): Offer alternative that
11391 uses vector registers.
89fdc743 11392 * config/aarch64/iterators.md: Add attributes rtn and vas.
040d8a1c 11393
85bd4ac6 113942013-07-26 Kyrylo Tkachov <kyrylo.tkachov@arm.com>
576e2f37 11395 Richard Earnshaw <richard.earnshaw@arm.com>
85bd4ac6
KT		 11396
11397 * combine.c (simplify_comparison): Re-canonicalize operands
576e2f37 11398 where appropriate.
85bd4ac6
KT		 11399 * config/arm/arm.md (movcond_addsi): New splitter.
11400
b6051207
SA		 114012013-07-25 Sterling Augustine <saugustine@google.com>
11402
11403 * dwarf2out.c (size_of_pubnames): Move code to...
11404 (include_pubname_in_output): ...here. New.
11405 (want_pubnames): Rearrange.
11406 (output_pubnames): Call include_pubname_in_output. Move assertion.
11407
41a7c215
CM		 114082013-07-25 Cameron McInally <cameron.mcinally@nyu.edu>
11409
11410 * doc/extend.texi: Fix return types for __builtin_ia32_cmp*s builtins.
11411
114122013-07-25 Cameron McInally <cameron.mcinally@nyu.edu>
11413
11414 PR target/38836
11415 * doc/extend.texi: Remove obsolete builtins. Fix
11416 typo for __builtin_ia32_loadss and __builtin_ia32_cmpnltss.
11417
49bde175
JH		 114182013-07-25 Jan Hubicka <jh@suse.cz>
11419
11420 * cgraph.c (release_function_body): Break out from ...
11421 (cgraph_release_function_body): ... this one; also release DECL_RESULT
11422 and DECL_ARGUMENTS.
11423 * ipa-cp.c (get_replacement_map): Add parm_num argument; do not set
11424 old_tree in the map.
11425 (create_specialized_node): Update.
11426 * lto-cgraph.c (output_node_opt_summary): Do not translate old_tree
11427 into index.
41a7c215
CM		 11428 * cgraphclones.c (cgraph_create_virtual_clone): Do not copy
11429 DECL_ARGUMENTS, DECL_INITIAL and DECL_RESULT.
49bde175
JH		 11430 * ipa-prop.c (ipa_populate_param_decls): Look for origin of clones.
11431 * tree-inline.c (initialize_cfun): Initialize DECL_ARGUMENTS and
11432 DECL_RESULT.
11433
4870352d
KT		 114342013-07-25 Kyrylo Tkachov <kyrylo.tkachov@arm.com>
11435
11436 * config/arm/arm.md (arm_addsi3, addsi3_carryin_<optab>,
11437 addsi3_carryin_alt2_<optab>): Correct output template.
11438
1d6aee1c
KT		 114392013-07-25 Kyrylo Tkachov <kyrylo.tkachov@arm.com>
11440
11441 * config/arm/arm-fixed.md (ssmulsa3, usmulusa3):
11442 Adjust for arm_restrict_it.
11443 Remove trailing whitespace.
11444
c7cafd75
MK		 114452013-07-25  Mark Kettenis  <kettenis@openbsd.org>
11446
017d38f5
MK		 11447 * config/pa/pa.c (pa_trampoline_init): Emit __enable_execute_stack
11448 libcall if HAVE_ENABLE_EXECUTE_STACK is defined.
11449
c7cafd75
MK		 11450 * config.gcc (hppa-*-openbsd*): Don't set tmake_file.
11451
00b7527b
VM		 114522013-07-25 Vladimir Makarov <vmakarov@redhat.com>
11453
11454 PR rtl-optimization/57960
11455 * lra-constraints.c (process_alt_operands): Use the right mode
11456 when checking strict_low.
11457
4f63dfc6
JH		 114582013-07-25 Jan Hubicka <jh@suse.cz>
11459
11460 * lto-symtab.c (lto_cgraph_replace_node): Release function body.
11461 * cgraph.c (cgraph_remove_node): Do not release function body
11462 when in cgraph streaming.
11463 * ipa.c (process_references, symtab_remove_unreachable_nodes): Objects
41a7c215
CM		 11464 in other partitions are not considered reachable; fix handling of
11465 clones.
4f63dfc6 11466
e6dcfa18
RR		 114672013-07-25 Ramana Radhakrishnan <ramana.radhakrishnan@arm.com>
11468
11469 * config/arm/arm.md (*sibcall_insn): Remove unnecessary space.
11470
0b93d3b6
RR		 114712013-07-25 Ramana Radhakrishnan <ramana.radhakrishnan@arm.com>
11472
11473 PR target/19599
11474 PR target/57731
bb80c2eb 11475 PR target/57837
e6dcfa18 11476 * config/arm/arm.md ("*sibcall_insn): Replace use of
0b93d3b6 11477 Ss with US. Adjust output for v5 and v4t.
41a7c215 11478 (*sibcall_value_insn): Likewise and loosen predicate on operand0.
e6dcfa18 11479
0b93d3b6
RR		 11480 * config/arm/constraints.md ("Ss"): Rename to US.
11481
690688b3
TG		 114822013-07-25 Terry Guo <terry.guo@arm.com>
11483
11484 * config/arm/arm.c (thumb1_size_rtx_costs): Assign proper cost for
11485 shift_add/shift_sub0/shift_sub1 RTXs.
11486
a4ad093b 114872013-07-24 Bill Schmidt <wschmidt@linux.ibm.com>
12211b99 11488 Anton Blanchard <anton@au1.ibm.com>
a4ad093b
BS		 11489
11490 * config/rs6000/altivec.md (altivec_vpkpx): Handle little endian.
11491 (altivec_vpks<VI_char>ss): Likewise.
11492 (altivec_vpks<VI_char>us): Likewise.
11493 (altivec_vpku<VI_char>us): Likewise.
11494 (altivec_vpku<VI_char>um): Likewise.
11495
8c5005ce
DM		 114962013-07-24 David Malcolm <dmalcolm@redhat.com>
11497
11498 Introduce context class.
11499
11500 * Makefile.in (CONTEXT_H): New.
11501 (OBJS): Add context.o.
11502 (toplev.o): Add CONTEXT_H to dependencies.
11503 (context.o): New.
11504
41a7c215 11505 * toplev.c (general_init): Create the singleton gcc::context instance.
8c5005ce
DM		 11506
11507 * context.c: New.
11508
11509 * context.h: New.
11510
ce4a9422
JR		 115112013-07-24 Joern Rennecke <joern.rennecke@embecosm.com>
11512
11513 PR rtl-optimization/57968
11514 * mode-switching.c (create_pre_exit): Allow instructions that
11515 don't set a return register to need a non-exit mode.
11516
493f4c9e 115172013-07-24 Bill Schmidt <wschmidt@vnet.linux.ibm.com>
12211b99 11518 Anton Blanchard <anton@au1.ibm.com>
493f4c9e 11519
b5860fd3
WS		 11520 * config/rs6000/vector.md (vec_realign_load_<mode>): Reorder input
11521 operands to vperm for little endian.
11522 * config/rs6000/rs6000.c (rs6000_expand_builtin): Use lvsr instead
11523 of lvsl to create the control mask for a vperm for little endian.
493f4c9e 11524
73792b92 115252013-07-23 Bill Schmidt <wschmidt@linux.vnet.ibm.com>
12211b99 11526 Anton Blanchard <anton@au1.ibm.com>
73792b92
BS		 11527
11528 * config/rs6000/rs6000.c (altivec_expand_vec_perm_const): Reverse
11529 two operands for little-endian.
11530
14c21302
SE		 115312013-07-23 Steve Ellcey <sellcey@mips.com>
11532
11533 * config/mips/mips.c (mips_case_values_threshold): New.
11534 (TARGET_CASE_VALUES_THRESHOLD): Define.
11535
a6056198 115362013-07-23 Bill Schmidt <wschmidt@linux.vnet.ibm.com>
12211b99 11537 Anton Blanchard <anton@au1.ibm.com>
a6056198
BS		 11538
11539 * config/rs6000/rs6000.c (altivec_expand_vec_perm_const): Correct
11540 selection of field for vector splat in little endian mode.
11541
dd7a40e1
MM		 115422013-07-23 Michael Meissner <meissner@linux.vnet.ibm.com>
11543
11544 * config/rs6000/vector.md (xor<mode>3): Move 128-bit boolean
11545 expanders to rs6000.md.
11546 (ior<mode>3): Likewise.
11547 (and<mode>3): Likewise.
11548 (one_cmpl<mode>2): Likewise.
11549 (nor<mode>3): Likewise.
11550 (andc<mode>3): Likewise.
11551 (eqv<mode>3): Likewise.
11552 (nand<mode>3): Likewise.
11553 (orc<mode>3): Likewise.
11554
11555 * config/rs6000/rs6000-protos.h (rs6000_split_logical): New
11556 declaration.
11557
11558 * config/rs6000/rs6000.c (rs6000_split_logical_inner): Add support
11559 to split multi-word logical operations.
11560 (rs6000_split_logical_di): Likewise.
11561 (rs6000_split_logical): Likewise.
11562
11563 * config/rs6000/vsx.md (VSX_L2): Delete, no longer used.
11564 (vsx_and<mode>3_32bit): Move 128-bit logical insns to rs6000.md,
11565 and allow TImode operations in 32-bit.
11566 (vsx_and<mode>3_64bit): Likewise.
11567 (vsx_ior<mode>3_32bit): Likewise.
11568 (vsx_ior<mode>3_64bit): Likewise.
11569 (vsx_xor<mode>3_32bit): Likewise.
11570 (vsx_xor<mode>3_64bit): Likewise.
11571 (vsx_one_cmpl<mode>2_32bit): Likewise.
11572 (vsx_one_cmpl<mode>2_64bit): Likewise.
11573 (vsx_nor<mode>3_32bit): Likewise.
11574 (vsx_nor<mode>3_64bit): Likewise.
11575 (vsx_andc<mode>3_32bit): Likewise.
11576 (vsx_andc<mode>3_64bit): Likewise.
11577 (vsx_eqv<mode>3_32bit): Likewise.
11578 (vsx_eqv<mode>3_64bit): Likewise.
11579 (vsx_nand<mode>3_32bit): Likewise.
11580 (vsx_nand<mode>3_64bit): Likewise.
11581 (vsx_orc<mode>3_32bit): Likewise.
11582 (vsx_orc<mode>3_64bit): Likewise.
11583
11584 * config/rs6000/rs6000.h (VLOGICAL_REGNO_P): Always allow vector
11585 logical types in GPRs.
11586
11587 * config/rs6000/altivec.md (altivec_and<mode>3): Move 128-bit
11588 logical insns to rs6000.md, and allow TImode operations in
11589 32-bit.
11590 (altivec_ior<mode>3): Likewise.
11591 (altivec_xor<mode>3): Likewise.
11592 (altivec_one_cmpl<mode>2): Likewise.
11593 (altivec_nor<mode>3): Likewise.
11594 (altivec_andc<mode>3): Likewise.
11595
11596 * config/rs6000/rs6000.md (BOOL_128): New mode iterators and mode
11597 attributes for moving the 128-bit logical operations into
11598 rs6000.md.
11599 (BOOL_REGS_OUTPUT): Likewise.
11600 (BOOL_REGS_OP1): Likewise.
11601 (BOOL_REGS_OP2): Likewise.
11602 (BOOL_REGS_UNARY): Likewise.
11603 (BOOL_REGS_AND_CR0): Likewise.
11604 (one_cmpl<mode>2): Add support for DI logical operations on
11605 32-bit, splitting the operations to 32-bit.
11606 (anddi3): Likewise.
11607 (iordi3): Likewise.
11608 (xordi3): Likewise.
11609 (and<mode>3, 128-bit types): Rewrite 2013-06-06 logical operator
11610 changes to combine the 32/64-bit code, allow logical operations on
11611 TI mode in 32-bit, and to use similar match_operator patterns like
11612 scalar mode uses. Combine the Altivec and VSX code for logical
11613 operations, and move it here.
11614 (ior<mode>3, 128-bit types): Likewise.
11615 (xor<mode>3, 128-bit types): Likewise.
11616 (one_cmpl<mode>3, 128-bit types): Likewise.
11617 (nor<mode>3, 128-bit types): Likewise.
11618 (andc<mode>3, 128-bit types): Likewise.
11619 (eqv<mode>3, 128-bit types): Likewise.
11620 (nand<mode>3, 128-bit types): Likewise.
11621 (orc<mode>3, 128-bit types): Likewise.
11622 (and<mode>3_internal): Likewise.
11623 (bool<mode>3_internal): Likewise.
11624 (boolc<mode>3_internal1): Likewise.
11625 (boolc<mode>3_internal2): Likewise.
11626 (boolcc<mode>3_internal1): Likewise.
11627 (boolcc<mode>3_internal2): Likewise.
11628 (eqv<mode>3_internal1): Likewise.
11629 (eqv<mode>3_internal2): Likewise.
11630 (one_cmpl1<mode>3_internal): Likewise.
11631
12211b99 116322013-07-23 David Holsgrove <david.holsgrove@xilinx.com>
9b69d4b4
DH		 11633
11634 * config/microblaze/microblaze.c (microblaze_expand_prologue):
11635 Rename flag_stack_usage to flag_stack_usage_info.
11636
12211b99 116372013-07-23 David Holsgrove <david.holsgrove@xilinx.com>
450b0ebb 11638
a5f257fd
ME		 11639 * config/microblaze/sync.md: New file.
11640 * config/microblaze/microblaze.md: Include sync.md
11641 * config/microblaze/microblaze.c: Add print_operand 'y'.
11642 * config/microblaze/constraints.md: Add memory_contraint
11643 'Q' which is a single register.
450b0ebb 11644
49fba14f
EB		 116452013-07-23 Eric Botcazou <ebotcazou@adacore.com>
11646
11647 * doc/invoke.texi (SPARC Options): Document new leon3 processor value.
11648
c7b3b99f
PCC		 116492013-07-22 Po-Chun Chang <pchang9@cs.wisc.edu>
11650
11651 * reload.c (find_reloads): Exit loop once we find this operand
11652 cannot be reloaded somehow for this alternative.
11653
11654 * reload.c (find_reloads): Exit loop once we find a hard register.
11655
11656 * rtlanal.c (computed_jump_p): Exit loop once we find label
11657 reference is used.
11658
11659 * i386.c (ix86_pad_returns): Exit loop after setting replace.
11660
11661 * cfgloopmanip.c (remove_path): Exit loop after setting
11662 irred_invalidated.
11663
11664 * gensupport.c (subst_dup): Avoid loop if code is not
11665 MATCH_DUP nor MATCH_OP_DUP.
11666
c8fbf1fa
NBJ		 116672013-07-23 Nicklas Bo Jensen <nbjensen@gmail.com>
11668
11669 * doc/md.texi (Machine-Specific Peephole Optimizers): Fix a typo.
11670
9259db42
YZ		 116712013-07-23 Yufeng Zhang <yufeng.zhang@arm.com>
11672
11673 * config/aarch64/aarch64.c (aarch64_hard_regno_mode_ok): Also return
11674 true for SP_REGNUM if mode == ptr_mode.
11675 * config/aarch64/aarch64.h (ADDITIONAL_REGISTER_NAMES): Add "wsp"
11676 with value R0_REGNUM + 31.
11677
00edcfbe
YZ		 116782013-07-23 Yufeng Zhang <yufeng.zhang@arm.com>
11679
11680 * config/aarch64/aarch64.c (aarch64_pad_arg_upward): In big-endian,
11681 pad pointer-typed argument downward.
11682
43be9a95
YZ		 116832013-07-23 Yufeng Zhang <yufeng.zhang@arm.com>
11684
11685 * config/aarch64/aarch64.h (TARGET_CPU_CPP_BUILTINS): Define _ILP32
11686 and __ILP32__ when the ILP32 model is in use.
11687
28514dda
YZ		 116882013-07-23 Yufeng Zhang <yufeng.zhang@arm.com>
11689
11690 * config/aarch64/aarch64.c (POINTER_BYTES): New define.
11691 (aarch64_load_symref_appropriately): In the case of
11692 SYMBOL_SMALL_ABSOLUTE, use the mode of 'dest' instead of Pmode
11693 to generate new rtx; likewise to the case of SYMBOL_SMALL_GOT.
11694 (aarch64_expand_mov_immediate): In the case of SYMBOL_FORCE_TO_MEM,
11695 change to pass 'ptr_mode' to force_const_mem and zero-extend 'mem'
11696 if 'mode' doesn't equal to 'ptr_mode'.
11697 (aarch64_output_mi_thunk): Add an assertion on the alignment of
11698 'vcall_offset'; change to call aarch64_emit_move differently depending
11699 on whether 'Pmode' equals to 'ptr_mode' or not; use 'POINTER_BYTES'
11700 to calculate the upper bound of 'vcall_offset'.
11701 (aarch64_cannot_force_const_mem): Change to also return true if
11702 mode != ptr_mode.
11703 (aarch64_legitimize_reload_address): In the case of large
11704 displacements, add new local variable 'xmode' and an assertion
11705 based on it; change to use 'xmode' to generate the new rtx and
11706 reload.
11707 (aarch64_asm_trampoline_template): Change to generate the template
11708 differently depending on TARGET_ILP32 or not; change to use
11709 'POINTER_BYTES' in the argument passed to assemble_aligned_integer.
11710 (aarch64_trampoline_size): Removed.
11711 (aarch64_trampoline_init): Add new local constant 'tramp_code_sz'
11712 and replace immediate literals with it. Change to use 'ptr_mode'
11713 instead of 'DImode' and call convert_memory_address if the mode
11714 of 'fnaddr' doesn't equal to 'ptr_mode'.
11715 (aarch64_elf_asm_constructor): Change to use assemble_aligned_integer
11716 to output symbol.
11717 (aarch64_elf_asm_destructor): Likewise.
11718 * config/aarch64/aarch64.h (TRAMPOLINE_SIZE): Change to be dependent
11719 on TARGET_ILP32 instead of aarch64_trampoline_size.
11720 * config/aarch64/aarch64.md (movsi_aarch64): Add new alternatives
11721 of 'mov' between WSP and W registers as well as 'adr' and 'adrp'.
11722 (loadwb_pair<GPI:mode>_<PTR:mode>): Rename to ...
11723 (loadwb_pair<GPI:mode>_<P:mode>): ... this. Replace PTR with P.
11724 (storewb_pair<GPI:mode>_<PTR:mode>): Likewise; rename to ...
11725 (storewb_pair<GPI:mode>_<P:mode>): ... this.
11726 (add_losym): Change to 'define_expand' and call gen_add_losym_<mode>
11727 depending on the value of 'mode'.
11728 (add_losym_<mode>): New.
11729 (ldr_got_small_<mode>): New, based on ldr_got_small.
11730 (ldr_got_small): Remove.
11731 (ldr_got_small_sidi): New.
11732 * config/aarch64/iterators.md (P): New.
11733 (PTR): Change to 'ptr_mode' in the condition.
11734
17a819cb
YZ		 117352013-07-23 Yufeng Zhang <yufeng.zhang@arm.com>
11736
11737 * config.gcc (aarch64*-*-*): Support --with-abi.
11738 (aarch64*-*-elf): Support --with-multilib-list.
11739 (aarch64*-*-linux*): Likewise.
11740 (supported_defaults): Add abi to aarch64*-*-*.
11741 * configure.ac: Mention AArch64 for --with-multilib-list.
11742 * configure: Re-generated.
11743 * config/aarch64/biarchilp32.h: New file.
11744 * config/aarch64/biarchlp64.h: New file.
11745 * config/aarch64/aarch64-elf.h (ENDIAN_SPEC): New define.
11746 (ABI_SPEC): Ditto.
11747 (MULTILIB_DEFAULTS): Ditto.
11748 (DRIVER_SELF_SPECS): Ditto.
11749 (ASM_SPEC): Update to also substitute -mabi.
11750 * config/aarch64/aarch64-elf-raw.h (LINK_SPEC): Add linker script
11751 file whose name depends on -mabi= and -mbig-endian.
11752 * config/aarch64/aarch64.h (LONG_TYPE_SIZE): Change to depend on
11753 TARGET_ILP32.
11754 (POINTER_SIZE): New define.
11755 (POINTERS_EXTEND_UNSIGNED): Ditto.
11756 (enum aarch64_abi_type): New enumeration tag.
11757 (AARCH64_ABI_LP64, AARCH64_ABI_ILP32): New enumerators.
11758 (AARCH64_ABI_DEFAULT): Define to AARCH64_ABI_LP64 if undefined.
11759 (TARGET_ILP32): New define.
11760 * config/aarch64/aarch64.opt (mabi): New.
11761 (aarch64_abi): New.
11762 (ilp32, lp64): New values for -mabi.
11763 * config/aarch64/t-aarch64 (comma): New define.
11764 (MULTILIB_OPTIONS): Ditto.
11765 (MULTILIB_DIRNAMES): Ditto.
11766 * config/aarch64/t-aarch64-linux (MULTIARCH_DIRNAME): New define.
11767 * doc/invoke.texi: Document -mabi for AArch64.
11768
9f9cbdce
GJL		 117692013-07-23 Georg-Johann Lay <avr@gjlay.de>
11770
11771 * config/avr/avr.md: Explain asm print modifier 'r' for REG.
11772
7061977b 117732013-07-22 Bill Schmidt <wschmidt@vnet.linux.ibm.com>
12211b99 11774 Anton Blanchard <anton@au1.ibm.com>
7061977b
BS		 11775
11776 * config/rs6000/rs6000.c (rs6000_expand_vector_init): Fix
11777 endianness when selecting field to splat.
11778
75a41878
EC		 117792013-07-22 Eric Christopher <echristo@gmail.com>
11780
11781 * dwarf2out.c (die_odr_checksum): New function to use
11782 CHECKSUM_ macros and ULEB128 for DIE tag.
11783 (generate_type_signature): Use.
41a7c215 11784
38ae58ca
EB		 117852013-07-22 Eric Botcazou <ebotcazou@adacore.com>
11786
11787 * config.gcc (sparc*-*-*): Accept leon3 processor.
11788 (sparc-leon*-*): Merge with sparc*-*-* and add leon3 support.
11789 * doc/invoke.texi (SPARC Options): Adjust -mfix-ut699 entry.
11790 * config/sparc/sparc-opts.h (enum processor_type): Add PROCESSOR_LEON3.
11791 * config/sparc/sparc.opt (enum processor_type): Add leon3.
11792 (mfix-ut699): Adjust comment.
11793 * config/sparc/sparc.h (TARGET_CPU_leon3): New define.
11794 (CPP_CPU32_DEFAULT_SPEC): Add leon3 support.
11795 (CPP_CPU_SPEC): Likewise.
11796 (ASM_CPU_SPEC): Likewise.
11797 * config/sparc/sparc.c (leon3_cost): New constant.
11798 (sparc_option_override): Add leon3 support.
11799 (mem_ref): New function.
11800 (sparc_gate_work_around_errata): Return true if -mfix-ut699 is enabled.
11801 (sparc_do_work_around_errata): Look into the instruction in the delay
11802 slot and adjust accordingly. Add fix for the data cache nullify issues
11803 of the UT699. Change insertion position for the NOP.
11804 * config/sparc/leon.md (leon_fpalu, leon_fpmds, write_buf): Delete.
11805 (leon3_load): New reservation.
11806 (leon_store): Bump latency to 2.
11807 (grfpu): New automaton.
11808 (grfpu_alu): New unit.
11809 (grfpu_ds): Likewise.
11810 (leon_fp_alu): Adjust.
11811 (leon_fp_mult): Delete.
11812 (leon_fp_div): Split into leon_fp_divs and leon_fp_divd.
11813 (leon_fp_sqrt): Split into leon_fp_sqrts and leon_fp_sqrtd.
11814 * config/sparc/sparc.md (cpu): Add leon3.
11815 * config/sparc/sync.md (atomic_exchangesi): Disable if -mfix-ut699.
11816 (swapsi): Likewise.
11817 (atomic_test_and_set): Likewise.
11818 (ldstub): Likewise.
11819
04dfc6df
JU		 118202013-07-22 Jürgen Urban <JuergenUrban@gmx.de>
11821
11822 * config.gcc (mips*-*-*): Add --with-fpu support. Make single the
11823 default for R5900 targets.
11824 * config/mips/mips.h (OPTION_DEFAULT_SPECS): Handle --with-fpu.
11825 (ISA_HAS_LDC1_SDC1): Set to false for TARGET_MIPS5900.
11826 * config/mips/mips.c (mips_option_override): Report an error for
11827 -march=r5900 -mhard-float -mdouble-float. Use spu_single_format
11828 for -march=r5900 -mhard-float.
11829
c7b3b99f 118302013-07-22 Po-Chun Chang <pchang9@cs.wisc.edu>
ae382ebd
PCC		 11831
11832 * df-problems.c (can_move_insns_across): Exit loop once we
11833 find a non-fixed, non-global register.
11834
11835 * ipa-pure-const.c (propagate_nothrow): Exit loop after
11836 setting can_throw.
11837
11838 * omega.c (omega_eliminate_red): Break after setting red_found.
11839 (omega_problem_has_red_equations): Similarly after setting found.
11840 (omega_query_variable): Similarly after setting coupled.
11841
9b6e6981
MP		 118422013-07-22 Marek Polacek <polacek@redhat.com>
11843
11844 * gimplify.c: Don't include gimple.h twice.
11845
5e5f7673
KT		 118462013-07-22 Kyrylo Tkachov <kyrylo.tkachov@arm.com>
11847
11848 * config/arm/constraints.md (Pd): Allow TARGET_THUMB
11849 instead of TARGET_THUMB1.
11850 (Pz): New constraint.
11851 * config/arm/arm.md (arm_addsi3): Add alternatives for 16-bit
11852 encodings.
11853 (compare_negsi_si): Likewise.
11854 (compare_addsi2_op0): Likewise.
11855 (compare_addsi2_op1): Likewise.
11856 (addsi3_carryin_<optab>): Likewise.
11857 (addsi3_carryin_alt2_<optab>): Likewise.
11858 (addsi3_carryin_shift_<optab>): Disable cond_exec variant
11859 for arm_restrict_it.
11860 (subsi3_carryin): Likewise.
11861 (arm_subsi3_insn): Add alternatives for 16-bit encoding.
11862 (minmax_arithsi): Disable for arm_restrict_it.
11863 (minmax_arithsi_non_canon): Adjust for arm_restrict_it.
11864 (satsi_<SAT:code>): Disable cond_exec variant for arm_restrict_it.
11865 (satsi_<SAT:code>_shift): Likewise.
11866 (arm_shiftsi3): Add alternative for 16-bit encoding.
11867 (arm32_movhf): Disable for arm_restrict_it.
11868 (arm_cmpdi_unsigned): Add alternatives for 16-bit encoding.
11869 (arm_movtas_ze): Disable cond_exec variant for arm_restrict_it.
11870
859abddd
SN		 118712013-07-22 Sofiane Naci <sofiane.naci@arm.com>
11872
11873 * config/arm/arm.md (attribute "insn"): Delete.
11874 (attribute "type"): Add "mov_imm", "mov_reg", "mov_shift",
11875 "mov_shift_reg", "mvn_imm", "mvn_reg", "mvn_shift" and "mvn_shift_reg".
11876 (not_shiftsi): Update for attribute change.
11877 (not_shiftsi_compare0): Likewise.
11878 (not_shiftsi_compare0_scratch): Likewise.
11879 (arm_one_cmplsi2): Likewise.
11880 (thumb1_one_cmplsi2): Likewise.
11881 (notsi_compare0): Likewise.
11882 (notsi_compare0_scratch): Likewise.
11883 (thumb1_movdi_insn): Likewise.
11884 (arm_movsi_insn): Likewise.
11885 (movhi_insn_arch4): Likewise.
11886 (movhi_bytes): Likewise.
11887 (arm_movqi_insn): Likewise.
11888 (thumb1_movqi_insn): Likewise.
11889 (arm32_movhf): Likewise.
11890 (thumb1_movhf): Likewise.
11891 (arm_movsf_soft_insn): Likewise.
11892 (thumb1_movsf_insn): Likewise.
11893 (thumb_movdf_insn): Likewise.
11894 (movsicc_insn): Likewise.
11895 (movsfcc_soft_insn): Likewise.
11896 (and_scc): Likewise.
11897 (cond_move): Likewise.
11898 (if_move_not): Likewise.
11899 (if_not_move): Likewise.
11900 (if_shift_move): Likewise.
11901 (if_move_shift): Likewise.
11902 (if_shift_shift): Likewise.
11903 (if_not_arith): Likewise.
11904 (if_arith_not): Likewise.
11905 (cond_move_not): Likewise.
11906 * config/arm/neon.md (neon_mov<mode>): Update for attribute change.
11907 (neon_mov<mode>): Likewise.
11908 * config/arm/vfp.md (arm_movsi_vfp): Update for attribute change.
11909 (thumb2_movsi_vfp): Likewise.
11910 (movsf_vfp): Likewise.
11911 (thumb2_movsf_vfp): Likewise.
371e77e3
UB		 11912 * config/arm/arm.c (xscale_sched_adjust_cost): Update for attribute
11913 change.
859abddd
SN		 11914 (cortexa7_older_only): Likewise.
11915 (cortexa7_younger): Likewise.
11916 * config/arm/arm1020e.md (1020alu_op): Update for attribute change.
11917 (1020alu_shift_op): Likewise.
11918 (1020alu_shift_reg_op): Likewise.
11919 * config/arm/arm1026ejs.md (alu_op): Update for attribute change.
11920 (alu_shift_op): Likewise.
11921 (alu_shift_reg_op): Likewise.
11922 * config/arm/arm1136jfs.md (11_alu_op): Update for attribute change.
11923 (11_alu_shift_op): Likewise.
11924 (11_alu_shift_reg_op): Likewise.
11925 * config/arm/arm926ejs.md (9_alu_op): Update for attribute change.
11926 (9_alu_shift_reg_op): Likewise.
371e77e3
UB		 11927 * config/arm/cortex-a15.md (cortex_a15_alu): Update for attribute
11928 change.
859abddd
SN		 11929 (cortex_a15_alu_shift): Likewise.
11930 (cortex_a15_alu_shift_reg): Likewise.
11931 * config/arm/cortex-a5.md (cortex_a5_alu): Update for attribute change.
11932 (cortex_a5_alu_shift): Likewise.
371e77e3
UB		 11933 * config/arm/cortex-a53.md (cortex_a53_alu): Update for attribute
11934 change.
859abddd 11935 (cortex_a53_alu_shift): Likewise.
371e77e3
UB		 11936 * config/arm/cortex-a7.md (cortex_a7_alu_imm): Update for attribute
11937 change.
859abddd
SN		 11938 (cortex_a7_alu_reg): Likewise.
11939 (cortex_a7_alu_shift): Likewise.
11940 * config/arm/cortex-a8.md (cortex_a8_alu): Update for attribute change.
11941 (cortex_a8_alu_shift): Likewise.
11942 (cortex_a8_alu_shift_reg): Likewise.
11943 (cortex_a8_mov): Likewise.
11944 * config/arm/cortex-a9.md (cortex_a9_dp): Update for attribute change.
11945 (cortex_a9_dp_shift): Likewise.
11946 * config/arm/cortex-m4.md (cortex_m4_alu): Update for attribute change.
11947 * config/arm/cortex-r4.md (cortex_r4_alu): Update for attribute change.
11948 (cortex_r4_mov): Likewise.
11949 (cortex_r4_alu_shift): Likewise.
11950 (cortex_r4_alu_shift_reg): Likewise.
11951 * config/arm/fa526.md (526_alu_op): Update for attribute change.
11952 (526_alu_shift_op): Likewise.
11953 * config/arm/fa606te.md (606te_alu_op): Update for attribute change.
11954 * config/arm/fa626te.md (626te_alu_op): Update for attribute change.
11955 (626te_alu_shift_op): Likewise.
11956 * config/arm/fa726te.md (726te_shift_op): Update for attribute change.
11957 (726te_alu_op): Likewise.
11958 (726te_alu_shift_op): Likewise.
11959 (726te_alu_shift_reg_op): Likewise.
11960 * config/arm/fmp626.md (mp626_alu_op): Update for attribute change.
11961 (mp626_alu_shift_op): Likewise.
11962 * config/arm/marvell-pj4.md (pj4_alu_e1): Update for attribute change.
11963 (pj4_alu_e1_conds): Likewise.
11964 (pj4_alu): Likewise.
11965 (pj4_alu_conds): Likewise.
11966 (pj4_shift): Likewise.
11967 (pj4_shift_conds): Likewise.
11968 (pj4_alu_shift): Likewise.
11969 (pj4_alu_shift_conds): Likewise.
11970
95b97fac
KT		 119712013-07-22 Kyrylo Tkachov <kyrylo.tkachov@arm.com>
11972
11973 * config/arm/predicates.md (shiftable_operator_strict_it):
11974 New predicate.
11975 * config/arm/thumb2.md (thumb_andsi_not_shiftsi_si):
11976 Disable cond_exec version for arm_restrict_it.
11977 (thumb2_smaxsi3): Convert to generate cond_exec.
11978 (thumb2_sminsi3): Likewise.
11979 (thumb32_umaxsi3): Likewise.
11980 (thumb2_uminsi3): Likewise.
11981 (thumb2_abssi2): Adjust constraints for arm_restrict_it.
11982 (thumb2_neg_abssi2): Likewise.
11983 (thumb2_mov_scc): Add alternative for 16-bit encoding.
11984 (thumb2_movsicc_insn): Adjust alternatives.
11985 (thumb2_mov_negscc): Disable for arm_restrict_it.
11986 (thumb2_mov_negscc_strict_it): New pattern.
11987 (thumb2_mov_notscc_strict_it): New pattern.
11988 (thumb2_mov_notscc): Disable for arm_restrict_it.
11989 (thumb2_ior_scc): Likewise.
11990 (thumb2_ior_scc_strict_it): New pattern.
11991 (thumb2_cond_move): Adjust for arm_restrict_it.
11992 (thumb2_cond_arith): Disable for arm_restrict_it.
11993 (thumb2_cond_arith_strict_it): New pattern.
11994 (thumb2_cond_sub): Adjust for arm_restrict_it.
11995 (thumb2_movcond): Likewise.
11996 (thumb2_extendqisi_v6): Disable cond_exec variant for arm_restrict_it.
11997 (thumb2_zero_extendhisi2_v6): Likewise.
11998 (thumb2_zero_extendqisi2_v6): Likewise.
11999 (orsi_notsi_si): Likewise.
12000 (orsi_not_shiftsi_si): Likewise.
12001
d5a6ef82
GJL		 120022013-07-22 Georg-Johann Lay <avr@gjlay.de>
12003
12004 * config/avr/avr.c (avr_out_xload): No SBIS around LPM so that
12005 instruction sequence is 1 byte shorter.
12006
d9e71154
UB		 120072013-07-22 Uros Bizjak <ubizjak@gmail.com>
12008
12009 * config/i386/i386.md (nonlocal_goto_receiver): Delete insn if
12010 it is not needed after split.
12011
8c6381ec
IS		 120122013-07-20 Iain Sandoe <iain@codesourcery.com>
12013
12014 PR target/51784
12015 * config/i386/i386.c (output_set_got) [TARGET_MACHO]: Adjust to emit a
12016 second label for nonlocal goto receivers. Don't output pic base labels
12017 unless we're producing PIC; mark that action unreachable().
12018 (ix86_save_reg): If the function contains a nonlocal label, save the
12019 PIC base reg.
12020 * config/darwin-protos.h (machopic_should_output_picbase_label): New.
f43245d1 12021 * config/darwin.c (emitted_pic_label_num): New GTY.
8c6381ec
IS		 12022 (update_pic_label_number_if_needed): New.
12023 (machopic_output_function_base_name): Adjust for nonlocal receiver
12024 case.
12025 (machopic_should_output_picbase_label): New.
12026 * config/i386/i386.md (enum unspecv): UNSPECV_NLGR: New.
12027 (nonlocal_goto_receiver): New insn and split.
12028
d05d0709
JG		 120292013-07-20 James Greenhalgh <james.greenhalgh@arm.com>
12030
12031 * config/aarch64/aarch64-builtins.c
12032 (aarch64_fold_builtin): Fold abs in all modes.
12033 * config/aarch64/aarch64-simd-builtins.def
12034 (abs): Enable for all modes.
12035 * config/aarch64/arm_neon.h
12036 (vabs<q>_s<8,16,32,64): Rewrite using builtins.
12037 (vabs_f64): Add missing intrinsic.
12038
ba590e63
IB		 120392013-07-19 Ian Bolton <ian.bolton@arm.com>
12040
12041 * config/aarch64/arm_neon.h (vabs_s64): New function
12042
02371798
GJL		 120432013-07-19 Georg-Johann Lay <avr@gjlay.de>
12044
12045 PR target/57516
12046 * config/avr/avr-fixed.md (round<mode>3_const): Turn expander to insn.
12047 * config/avr/avr.md (adjust_len): Add `round'.
12048 * config/avr/avr-protos.h (avr_out_round): New prototype.
12049 (avr_out_plus): Add `out_label' argument.
12050 * config/avr/avr.c (avr_out_plus_1): Add `out_label' argument.
12051 (avr_out_plus): Pass down `out_label' to avr_out_plus_1.
12052 Handle the case where `insn' is just a pattern.
12053 (avr_out_bitop): Handle the case where `insn' is just a pattern.
12054 (avr_out_round): New function.
12055 (avr_adjust_insn_length): Handle ADJUST_LEN_ROUND.
12056
12211b99 120572013-07-18 David Holsgrove <david.holsgrove@xilinx.com>
b950ac25
DH		 12058
12059 * config/microblaze/microblaze.c (microblaze_expand_prologue):
12060 Add check for flag_stack_usage to handle -fstack-usage support
12061
12211b99 120622013-07-18 Pat Haugen <pthaugen@us.ibm.com>
ed1fd999
PH		 12063
12064 * config/rs6000/rs6000.c (rs6000_option_override_internal): Adjust flag
12065 interaction for new Power8 flags and VSX.
12066
df9dda2d
ST		 120672013-07-18 Sriraman Tallam <tmsriram@google.com>
12068
12069 PR middle-end/57698
12070 * tree-inline.c (expand_call_inline): Emit errors during
12071 early_inlining only if optimization is not turned on.
12072
0f9cdef4
DM		 120732013-07-18 David Malcolm <dmalcolm@redhat.com>
12074
12075 * passes.def: New.
12076
12077 * passes.c (init_optimization_passes): Move the construction of
12078 the pass hierarchy into a new passes.def file.
12079
12080 * Makefile.in (passes.o): Add dependency on passes.def.
12081
2efa4087
DM		 120822013-07-18 David Malcolm <dmalcolm@redhat.com>
12083
12084 * passes.c (init_optimization_passes): Introduce macros for
12085 constructing the tree of passes (INSERT_PASSES_AFTER,
12086 PUSH_INSERT_PASSES_WITHIN, POP_INSERT_PASSES,
12087 TERMINATE_PASS_LIST).
12088
bc404e1b
VM		 120892013-07-18 Vladimir Makarov <vmakarov@redhat.com>
12090 Wei Mi <wmi@google.com>
12091
12092 PR rtl-optimization/57878
12093 * lra-assigns.c (assign_by_spills): Move non_reload_pseudos to the
12094 top.
12095 (reload_pseudo_compare_func): Check nregs first for reload
12096 pseudos.
12097
d675843f
DM		 120982013-07-18 David Malcolm <dmalcolm@redhat.com>
12099
12100 * tree-pass.h (pass_ipa_lto_wpa_fixup): Remove redundant decl.
12101
c7b3b99f 121022013-07-18 Po-Chun Chang <pchang9@cs.wisc.edu>
bcb97fbd
PCC		 12103
12104 * read-rtl.c (validate_const_int): Once an invalid character is
12105 seen, quit the loop.
12106
12107 * gengtype.c (write_roots): Similarly once we find the "deletable"
12108 or "if_marked" option.
12109
52fceb44
SN		 121102013-07-18 Sofiane Naci <sofiane.naci@arm.com>
12111
12112 * config/arm/arm.md (attribute "insn"): Delete values "mrs", "msr",
12113 "xtab" and "sat". Move value "clz" from here to ...
12114 (attriubte "type"): ... here.
12115 (satsi_<SAT:code>): Delete "insn" attribute.
12116 (satsi_<SAT:code>_shift): Likewise.
12117 (arm_zero_extendqisi2addsi): Likewise.
12118 (arm_extendqisi2addsi): Likewise.
12119 (clzsi2): Update for attribute changes.
12120 (rbitsi2): Likewise.
371e77e3
UB		 12121 * config/arm/arm-fixed.md (arm_ssatsihi_shift): Delete "insn"
12122 attribute.
52fceb44
SN		 12123 (arm_usatsihi): Likewise.
12124 * config/arm/cortex-a8.md (cortex_a8_alu): Update for attribute change.
12125
006bd006
SN		 121262013-07-18 Sofiane Naci <sofiane.naci@arm.com>
12127
12128 * config/arm/arm.md (attribute "type"): Rename "simple_alu_imm" to
371e77e3
UB		 12129 "arlo_imm". Rename "alu_reg" to "arlo_reg". Rename "simple_alu_shift"
12130 to "extend". Split "alu_shift" into "shift" and "arlo_shift". Split
006bd006
SN		 12131 "alu_shift_reg" into "shift_reg" and "arlo_shift_reg". List types
12132 in alphabetical order.
12133 (attribute "core_cycles"): Update for attribute changes.
12134 (arm_addsi3): Likewise.
12135 (addsi3_compare0): Likewise.
12136 (addsi3_compare0_scratch): Likewise.
12137 (addsi3_compare_op1): Likewise.
12138 (addsi3_compare_op2): Likewise.
12139 (compare_addsi2_op0): Likewise.
12140 (compare_addsi2_op1): Likewise.
12141 (addsi3_carryin_shift_<optab>): Likewise.
12142 (subsi3_carryin_shift): Likewise.
12143 (rsbsi3_carryin_shift): Likewise.
12144 (arm_subsi3_insn): Likewise.
12145 (subsi3_compare0): Likewise.
12146 (subsi3_compare): Likewise.
12147 (arm_andsi3_insn): Likewise.
12148 (thumb1_andsi3_insn): Likewise.
12149 (andsi3_compare0): Likewise.
12150 (andsi3_compare0_scratch): Likewise.
12151 (zeroextractsi_compare0_scratch
12152 (andsi_not_shiftsi_si): Likewise.
12153 (iorsi3_insn): Likewise.
12154 (iorsi3_compare0): Likewise.
12155 (iorsi3_compare0_scratch): Likewise.
12156 (arm_xorsi3): Likewise.
12157 (thumb1_xorsi3_insn): Likewise.
12158 (xorsi3_compare0): Likewise.
12159 (xorsi3_compare0_scratch): Likewise.
12160 (satsi_<SAT:code>_shift): Likewise.
12161 (rrx): Likewise.
12162 (arm_shiftsi3): Likewise.
12163 (shiftsi3_compare0): Likewise.
12164 (not_shiftsi): Likewise.
12165 (not_shiftsi_compare0): Likewise.
12166 (not_shiftsi_compare0_scratch): Likewise.
12167 (arm_one_cmplsi2): Likewise.
12168 (thumb_one_complsi2): Likewise.
12169 (notsi_compare0): Likewise.
12170 (notsi_compare0_scratch): Likewise.
12171 (thumb1_zero_extendhisi2): Likewise.
12172 (arm_zero_extendhisi2): Likewise.
12173 (arm_zero_extendhisi2_v6): Likewise.
12174 (arm_zero_extendhisi2addsi): Likewise.
12175 (thumb1_zero_extendqisi2): Likewise.
12176 (thumb1_zero_extendqisi2_v6): Likewise.
12177 (arm_zero_extendqisi2): Likewise.
12178 (arm_zero_extendqisi2_v6): Likewise.
12179 (arm_zero_extendqisi2addsi): Likewise.
12180 (thumb1_extendhisi2): Likewise.
12181 (arm_extendhisi2): Likewise.
12182 (arm_extendhisi2_v6): Likewise.
12183 (arm_extendqisi): Likewise.
12184 (arm_extendqisi_v6): Likewise.
12185 (arm_extendqisi2addsi): Likewise.
12186 (thumb1_extendqisi2): Likewise.
12187 (thumb1_movdi_insn): Likewise.
12188 (arm_movsi_insn): Likewise.
12189 (movsi_compare0): Likewise.
12190 (movhi_insn_arch4): Likewise.
12191 (movhi_bytes): Likewise.
12192 (arm_movqi_insn): Likewise.
12193 (thumb1_movqi_insn): Likewise.
12194 (arm32_movhf): Likewise.
12195 (thumb1_movhf): Likewise.
12196 (arm_movsf_soft_insn): Likewise.
12197 (thumb1_movsf_insn): Likewise.
12198 (movdf_soft_insn): Likewise.
12199 (thumb_movdf_insn): Likewise.
12200 (arm_cmpsi_insn): Likewise.
12201 (cmpsi_shiftsi): Likewise.
12202 (cmpsi_shiftsi_swp): Likewise.
12203 (arm_cmpsi_negshiftsi_si): Likewise.
12204 (movsicc_insn): Likewise.
12205 (movsfcc_soft_insn): Likewise.
12206 (arith_shiftsi): Likewise.
12207 (arith_shiftsi_compare0
12208 (arith_shiftsi_compare0_scratch
12209 (sub_shiftsi): Likewise.
12210 (sub_shiftsi_compare0
12211 (sub_shiftsi_compare0_scratch
12212 (and_scc): Likewise.
12213 (cond_move): Likewise.
12214 (if_plus_move): Likewise.
12215 (if_move_plus): Likewise.
12216 (if_move_not): Likewise.
12217 (if_not_move): Likewise.
12218 (if_shift_move): Likewise.
12219 (if_move_shift): Likewise.
12220 (if_shift_shift): Likewise.
12221 (if_not_arith): Likewise.
12222 (if_arith_not): Likewise.
12223 (cond_move_not): Likewise.
12224 (thumb1_ashlsi3): Set type attribute.
12225 (thumb1_ashrsi3): Likewise.
12226 (thumb1_lshrsi3): Likewise.
12227 (thumb1_rotrsi3): Likewise.
12228 (shiftsi3_compare0_scratch): Likewise.
12229 * config/arm/neon.md (neon_mov<mode>): Update for attribute changes.
12230 (neon_mov<mode>): Likewise.
371e77e3
UB		 12231 * config/arm/thumb2.md (thumb_andsi_not_shiftsi_si): Update for
12232 attribute changes.
006bd006
SN		 12233 (thumb2_movsi_insn): Likewise.
12234 (thumb2_cmpsi_neg_shiftsi): Likewise.
12235 (thumb2_extendqisi_v6): Likewise.
12236 (thumb2_zero_extendhisi2_v6): Likewise.
12237 (thumb2_zero_extendqisi2_v6): Likewise.
12238 (thumb2_shiftsi3_short): Likewise.
12239 (thumb2_addsi3_compare0_scratch): Likewise.
12240 (orsi_not_shiftsi_si): Likewise.
12241 * config/arm/vfp.md (arm_movsi_vfp): Update for attribute changes.
12242 * config/arm/arm-fixed.md (arm_ssatsihi_shift): Update for attribute
12243 changes.
12244 * config/arm/arm1020e.md (1020alu_op): Update for attribute changes.
12245 (1020alu_shift_op): Likewise.
12246 (1020alu_shift_reg_op): Likewise.
12247 * config/arm/arm1026ejs.md (alu_op): Update for attribute changes.
12248 (alu_shift_op): Likewise.
12249 (alu_shift_reg_op): Likewise.
12250 * config/arm/arm1136jfs.md (11_alu_op): Update for attribute changes.
12251 (11_alu_shift_op): Likewise.
12252 (11_alu_shift_reg_op): Likewise.
12253 * config/arm/arm926ejs.md (9_alu_op): Update for attribute changes.
12254 (9_alu_shift_reg_op): Likewise.
371e77e3
UB		 12255 * config/arm/cortex-a15.md (cortex_a15_alu): Update for attribute
12256 changes.
006bd006
SN		 12257 (cortex_a15_alu_shift): Likewise.
12258 (cortex_a15_alu_shift_reg): Likewise.
371e77e3
UB		 12259 * config/arm/cortex-a5.md (cortex_a5_alu): Update for attribute
12260 changes.
006bd006
SN		 12261 (cortex_a5_alu_shift): Likewise.
12262 * config/arm/cortex-a53.md (cortex_a53_alu) : Update for attribute
12263 changes.
12264 (cortex_a53_alu_shift): Likewise.
12265 * config/arm/cortex-a7.md (cortex_a7_alu_imm): Update for attribute
12266 changes.
12267 (cortex_a7_alu_reg): Likewise.
12268 (cortex_a7_alu_shift): Likewise.
371e77e3
UB		 12269 * config/arm/cortex-a8.md (cortex_a8_alu): Update for attribute
12270 changes.
006bd006
SN		 12271 (cortex_a8_alu_shift): Likewise.
12272 (cortex_a8_alu_shift_reg): Likewise.
12273 (cortex_a8_mov): Likewise.
12274 * config/arm/cortex-a9.md (cortex_a9_dp): Update for attribute changes.
12275 (cortex_a9_dp_shift): Likewise.
371e77e3
UB		 12276 * config/arm/cortex-m4.md (cortex_m4_alu): Update for attribute
12277 changes.
12278 * config/arm/cortex-r4.md (cortex_r4_alu): Update for attribute
12279 changes.
006bd006
SN		 12280 (cortex_r4_mov): Likewise.
12281 (cortex_r4_alu_shift): Likewise.
12282 (cortex_r4_alu_shift_reg): Likewise.
12283 * config/arm/fa526.md (526_alu_op): Update for attribute changes.
12284 (526_alu_shift_op): Likewise.
12285 * config/arm/fa606te.md (606te_alu_op): Update for attribute changes.
12286 * config/arm/fa626te.md (626te_alu_op): Update for attribute changes.
12287 (626te_alu_shift_op): Likewise.
12288 * config/arm/fa726te.md (726te_shift_op): Update for attribute changes.
12289 (726te_alu_op): Likewise.
12290 (726te_alu_shift_op): Likewise.
12291 (726te_alu_shift_reg_op): Likewise.
12292 * config/arm/fmp626.md (mp626_alu_op): Update for attribute changes.
12293 (mp626_alu_shift_op): Likewise.
12294 * config/arm/marvell-pj4.md (pj4_alu_e1): Update for attribute changes.
12295 (pj4_alu_e1_conds): Likewise.
12296 (pj4_alu): Likewise.
12297 (pj4_alu_conds): Likewise.
12298 (pj4_shift): Likewise.
12299 (pj4_shift_conds): Likewise.
12300 (pj4_alu_shift): Likewise.
12301 (pj4_alu_shift_conds): Likewise.
371e77e3
UB		 12302 * config/arm/arm.c (xscale_sched_adjust_cost): Update for attribute
12303 changes.
006bd006
SN		 12304 (cortexa7_older_only): Likewise.
12305 (cortexa7_younger): Likewise.
12306
651df1b2
DM		 123072013-07-18 David Malcolm <dmalcolm@redhat.com>
12308
12309 * ipa-pure-const.c (generate_summary): Rename to...
12310 (pure_const_generate_summary): ... this.
12311
ea717bd7
IS		 123122013-07-17 Iain Sandoe <iain@codesourcery.com>
12313
12314 * config/rs6000/darwin.h (REGISTER_NAMES): Add HTM registers.
12315
a1259a13
YR		 123162013-07-17 Yvan Roux <yvan.roux@linaro.org>
12317
12318 PR target/57909
12319 * config/arm/arm.c (gen_movmem_ldrd_strd): Fix unaligned load/store
12320 usage in HI mode.
12321
167f68ed
AK		 123222013-07-17 Andreas Krebbel <Andreas.Krebbel@de.ibm.com>
12323
12324 * config/s390/s390.c: (s390_expand_builtin): Allow -mhtm to be
12325 enabled without -march=zEC12.
12326 * config/s390/s390.h (TARGET_HTM): Do not require EC12 machine
12327 flags to be set.
12328
f900a982
MR		 123292013-07-16 Maciej W. Rozycki <macro@codesourcery.com>
12330
12331 * config/mips/mips.h (ISA_HAS_FP4): Correct formatting.
12332 (ISA_HAS_FP_MADD4_MSUB4): Also enable for ISA_MIPS32R2.
12333 (ISA_HAS_NMADD4_NMSUB4): Remove the MODE argument; rewrite in
12334 terms of ISA_HAS_FP4, and also enable for ISA_MIPS32R2.
12335 (ISA_HAS_NMADD3_NMSUB3): Remove the MODE argument.
12336 * config/mips/mips.c (mips_rtx_costs) <PLUS>: Check for
12337 ISA_HAS_FP_MADD4_MSUB4 || ISA_HAS_FP_MADD3_MSUB3 rather than
12338 ISA_HAS_FP4.
12339 <MINUS, NEG>: Update according to changes to ISA_HAS_NMADD4_NMSUB4
12340 and ISA_HAS_NMADD3_NMSUB3.
12341 * config/mips/mips.md (nmadd4<mode>, nmadd3<mode>): Likewise.
12342 (nmadd4<mode>_fastmath, nmadd3<mode>_fastmath): Likewise.
12343 (nmsub4<mode>, nmsub3<mode>): Likewise.
12344 (nmsub4<mode>_fastmath, nmsub3<mode>_fastmath): Likewise.
12345
d4ed27eb
MR		 123462013-07-16 Maciej W. Rozycki <macro@codesourcery.com>
12347
12348 * config/mips/mips.h (ISA_HAS_NMADD4_NMSUB4): Remove
12349 TARGET_MIPS5400 checking.
12350
7a5add18
PB		 123512013-07-16 Jakub Jelinek <jakub@redhat.com>
12352 Peter Bergner <bergner@vnet.ibm.com>
12353
12354 * config/rs6000/rs6000.h (FIRST_PSEUDO_REGISTERS): Mention HTM
12355 registers in the comment.
12356 (DWARF_FRAME_REGISTERS): Subtract also the 3 HTM registers.
12357 (DWARF_REG_TO_UNWIND_COLUMN): Use DWARF_FRAME_REGISTERS
12358 rather than FIRST_PSEUDO_REGISTERS.
12359
123602013-07-16 Peter Bergner <bergner@vnet.ibm.com>
c694dc01
PB		 12361
12362 * config/rs6000/rs6000.c (rs6000_option_override_internal): Do not
12363 enable extra ISA flags with TARGET_HTM.
12364
d2ab0929
MR		 123652013-07-16 Maciej W. Rozycki <macro@codesourcery.com>
12366
12367 * config/mips/mips.h (ISA_HAS_MULS, ISA_HAS_MSAC, ISA_HAS_MACC):
12368 Fix comment typos.
12369
839c74bc
CH		 123702013-07-15 Cong Hou <congh@google.com>
12371
12372 * tree-vect-data-refs.c (dr_group_sort_cmp): Do not use hash function
12373 in compare function for sorting.
12374
7a5add18 123752013-07-15 Peter Bergner <bergner@vnet.ibm.com>
0258b6e4
PB		 12376
12377 * config.gcc (powerpc*-*-*): Install htmintrin.h and htmxlintrin.h.
12378 * config/rs6000/t-rs6000 (MD_INCLUDES): Add htm.md.
12379 * config/rs6000/rs6000.opt: Add -mhtm option.
12380 * config/rs6000/rs6000-cpus.def (POWERPC_MASKS): Add OPTION_MASK_HTM.
12381 (ISA_2_7_MASKS_SERVER): Add OPTION_MASK_HTM.
12382 * config/rs6000/rs6000-c.c (rs6000_target_modify_macros): Define
12383 __HTM__ if the HTM instructions are available.
12384 * config/rs6000/predicates.md (u3bit_cint_operand, u10bit_cint_operand,
12385 htm_spr_reg_operand): New define_predicates.
12386 * config/rs6000/rs6000.md (define_attr "type"): Add htm.
12387 (TFHAR_REGNO, TFIAR_REGNO, TEXASR_REGNO): New define_constants.
12388 Include htm.md.
12389 * config/rs6000/rs6000-builtin.def (BU_HTM_0, BU_HTM_1, BU_HTM_2,
12390 BU_HTM_3, BU_HTM_SPR0, BU_HTM_SPR1): Add support macros for defining
12391 HTM builtin functions.
12392 * config/rs6000/rs6000.c (RS6000_BUILTIN_H): New macro.
12393 (rs6000_reg_names, alt_reg_names): Add HTM SPR register names.
12394 (rs6000_init_hard_regno_mode_ok): Add support for HTM instructions.
12395 (rs6000_builtin_mask_calculate): Likewise.
12396 (rs6000_option_override_internal): Likewise.
12397 (bdesc_htm): Add new HTM builtin support.
12398 (htm_spr_num): New function.
12399 (htm_spr_regno): Likewise.
12400 (rs6000_htm_spr_icode): Likewise.
12401 (htm_expand_builtin): Likewise.
12402 (htm_init_builtins): Likewise.
12403 (rs6000_expand_builtin): Add support for HTM builtin functions.
12404 (rs6000_init_builtins): Likewise.
371e77e3
UB		 12405 (rs6000_invalid_builtin, rs6000_opt_mask): Add support for -mhtm
12406 option.
0258b6e4
PB		 12407 * config/rs6000/rs6000.h (ASM_CPU_SPEC): Add support for -mhtm.
12408 (TARGET_HTM, MASK_HTM): Define macros.
12409 (FIRST_PSEUDO_REGISTER): Adjust for new HTM SPR registers.
12410 (FIXED_REGISTERS): Likewise.
12411 (CALL_USED_REGISTERS): Likewise.
12412 (CALL_REALLY_USED_REGISTERS): Likewise.
12413 (REG_ALLOC_ORDER): Likewise.
12414 (enum reg_class): Likewise.
12415 (REG_CLASS_NAMES): Likewise.
12416 (REG_CLASS_CONTENTS): Likewise.
12417 (REGISTER_NAMES): Likewise.
12418 (ADDITIONAL_REGISTER_NAMES): Likewise.
12419 (RS6000_BTC_SPR, RS6000_BTC_VOID, RS6000_BTC_32BIT, RS6000_BTC_64BIT,
12420 RS6000_BTC_MISC_MASK, RS6000_BTM_HTM): New macros.
12421 (RS6000_BTM_COMMON): Add RS6000_BTM_HTM.
12422 * config/rs6000/htm.md: New file.
12423 * config/rs6000/htmintrin.h: New file.
12424 * config/rs6000/htmxlintrin.h: New file.
12425
87dd8ab0
MS		 124262013-07-15 Marcus Shawcroft <marcus.shawcroft@arm.com>
12427
12428 * config/aarch64/aarch64-protos.h (aarch64_symbol_type):
12429 Define SYMBOL_TINY_GOT, update comment.
12430 * config/aarch64/aarch64.c
12431 (aarch64_load_symref_appropriately): Handle SYMBOL_TINY_GOT.
12432 (aarch64_expand_mov_immediate): Likewise.
12433 (aarch64_print_operand): Likewise.
12434 (aarch64_classify_symbol): Likewise.
12435 * config/aarch64/aarch64.md (UNSPEC_GOTTINYPIC): Define.
12436 (ldr_got_tiny): Define.
12437
52d676b6
TG		 124382013-07-13 Tobias Grosser <tobias@grosser.es>
12439
12440 PR tree-optimization/54094
12441 * graphite-clast-to-gimple.c (translate_clast_for_loop): Derive the
12442 scheduling dimension for the parallelism check from the polyhedral
12443 information in the AST.
12444 * graphite-dependences.c (carries_deps): Do not assume the schedule is
12445 in 2D + 1 form.
12446
286e8fc1
JM		 124472013-07-13 Jason Merrill <jason@redhat.com>
12448
12449 * print-tree.c (debug_vec_tree): Use debug_raw.
12450 (debug_raw (vec<tree, va_gc> &)): New.
12451 (debug_raw (vec<tree, va_gc> *)): New.
12452 * tree.h: Declare them.
12453
e43257e8
BC		 124542013-07-13 Bin Cheng <bin.cheng@arm.com>
12455
12456 * ifcvt.c (ifcvt_after_combine): New static variable.
12457 (cheap_bb_rtx_cost_p): Set scale to REG_BR_PROB_BASE when optimizing
12458 for size.
12459 (if_convert): New parameter after_combine. Set ifcvt_after_combine.
12460 (rest_of_handle_if_conversion, rest_of_handle_if_after_combine,
12461 rest_of_handle_if_after_reload): Pass new argument for if_convert.
12462
da582d46
MR		 124632013-07-12 Maciej W. Rozycki <macro@codesourcery.com>
12464
12465 * config/mips/mips.c (mips_expand_call): Remove empty statement.
12466
18b0ea8f
MM		 124672013-07-12 Michael Matz <matz@suse.de>
12468
12469 PR middle-end/55771
12470 * convert.c (convert_to_real): Reject non-float inner types.
12471
e4f0f84d
TB		 124722013-07-12 Tejas Belagod <tejas.belagod@arm.com>
12473
12474 * config/aarch64/aarch64-protos.h
12475 (aarch64_simd_immediate_valid_for_move): Remove.
12476 * config/aarch64/aarch64.c (simd_immediate_info): New member.
12477 (aarch64_simd_valid_immediate): Recognize idioms for shifting ones
12478 cases.
12479 (aarch64_output_simd_mov_immediate): Print the correct shift specifier.
12480
76500022
SE		 124812013-07-11 Steve Ellcey <sellcey@mips.com>
12482
12483 * config/mips/mips.c (mips_conditional_register_usage): Do not
12484 use t[0-7] registers in MIPS16 mode when optimizing for size.
12485
78039734
ST		 124862013-07-11 Sriraman Tallam <tmsriram@google.com>
12487
12488 * config/i386/i386.c (dispatch_function_versions): Fix array
12489 indexing of function_version_info to match actual_versions.
12490
7b24b675
TJ		 124912013-07-11 Teresa Johnson <tejohnson@google.com>
12492
12493 * vec.h (struct va_gc): Move release out-of-line.
12494 (va_gc::release): Call ggc_free on released vec.
12495
85d4cbb8
UW		 124962013-07-11 Ulrich Weigand <Ulrich.Weigand@de.ibm.com>
12497
12498 * config/rs6000/rs6000.md (""*tls_gd_low<TLSmode:tls_abi_suffix>"):
12499 Require GOT register as additional operand in UNSPEC.
12500 ("*tls_ld_low<TLSmode:tls_abi_suffix>"): Likewise.
12501 ("*tls_got_dtprel_low<TLSmode:tls_abi_suffix>"): Likewise.
12502 ("*tls_got_tprel_low<TLSmode:tls_abi_suffix>"): Likewise.
12503 ("*tls_gd<TLSmode:tls_abi_suffix>"): Update splitter.
12504 ("*tls_ld<TLSmode:tls_abi_suffix>"): Likewise.
12505 ("tls_got_dtprel_<TLSmode:tls_abi_suffix>"): Likewise.
12506 ("tls_got_tprel_<TLSmode:tls_abi_suffix>"): Likewise.
12507
52b3f9e2
GJL		 125082013-07-11 Georg-Johann Lay <avr@gjlay.de>
12509
12510 PR target/57631
12511 * config/avr/avr.c (avr_set_current_function): Sanity-check signal
12512 name seen by assembler/linker rather if available.
12513
3f3e266e
AS		 125142013-07-11 Andreas Schwab <schwab@suse.de>
12515
12516 * config/aarch64/aarch64-linux.h (CPP_SPEC): Define.
12517
21231ca6
VM		 125182013-07-10 Vladimir Makarov <vmakarov@redhat.com>
12519
371e77e3 12520 * lra-constraints.c (curr_insn_transform): Switch off optional reloads.
21231ca6 12521
d331d374
JM		 125222013-07-10 Joseph Myers <joseph@codesourcery.com>
12523
673c2f63
JM		 12524 * doc/tm.texi.in: Move hook documentation to ....
12525 * target.def: ... here.
12526
d331d374
JM		 12527 * doc/tm.texi.in (TARGET_CANONICALIZE_COMPARISON): Remove stray
12528 text on @hook line.
12529 * doc/tm.texi: Regenerate.
12530
dfeadaa0
PC		 125312013-07-10 Paolo Carlini <paolo.carlini@oracle.com>
12532
12533 PR c++/57869
12534 * doc/invoke.texi: Document -Wconditionally-supported.
12535
bb6f2bac
GJL		 125362013-07-10 Georg-Johann Lay <avr@gjlay.de>
12537
12538 PR target/57844
12539 * config/avr/avr.c (avr_prologue_setup_frame): Trunk -size to mode
12540 of my_fp.
12541
5219b232
GJL		 125422013-07-10 Georg-Johann Lay <avr@gjlay.de>
12543
12544 PR target/57506
12545 * config/avr/avr-mcus.def (atmega16hva, atmega16hva2, atmega16hvb)
12546 (atmega16m1, atmega16u4, atmega32a, atmega32c1, atmega32hvb)
12547 (atmega32m1, atmega32u4, atmega32u6, atmega64c1, atmega64m1):
12548 Remove duplicate devices.
12549 * config/avr/gen-avr-mmcu-texi.c (print_mcus): Fail on duplicate MCUs.
12550 * config/avr/t-multilib: Regenerate.
12551 * config/avr/avr-tables.opt: Regenerate.
12552 * doc/avr-mmcu.texi: Regenerate.
12553
283045c0
GJL		 125542013-07-10 Georg-Johann Lay <avr@gjlay.de>
12555
12556 PR target/56987
12557 * config/avr/avr.opt (Waddr-space-convert): Fix typo.
12558
ceb2d59f 125592013-07-10 Graham Stott <graham.stott@btinternet.com>
6a2fa4b2 12560
ceb2d59f
GS		 12561 * config/mips/mips.c (mips_rtx_costs): Very slightly increase
12562 the cost of MULT when optimizing for size.
12563
055e0a99
JBG		 125642013-07-10 Jan-Benedict Glaw <jbglaw@lug-owl.de>
12565
12566 * config/cr16/cr16-protos.h: Don't include target.h.
12567
ef43e868
JM		 125682013-07-09 Joseph Myers <joseph@codesourcery.com>
12569
12570 * config/rs6000/rs6000.c (rs6000_init_hard_regno_mode_ok): Only
12571 adjust register size for TDmode and TFmode for VSX registers.
12572
4a283fdf
KT		 125732013-07-08 Kai Tietz <ktietz@redhat.com>
12574
12575 PR target/56892
12576 * config/i386/i386.c (TARGET_FUNCTION_ATTRIBUTE_INLINABLE_P): Define as
12577 hook_bool_const_tree_true.
12578
2cf4c39e
AK		 125792013-07-08 Andreas Krebbel <Andreas.Krebbel@de.ibm.com>
12580
12581 * config/s390/s390.c: Replace F*_REGNUM with FPR*_REGNUM.
12582 * config/s390/s390.h: Remove F*_REGNUM macro definitions.
12583 * config/s390/s390.md: Define FPR*_REGNUM constants.
12584 Fix FPR2_REGNUM constant (18 -> 17).
12585 ("*trunc<BFP:mode><DFP_ALL:mode>2")
12586 ("*trunc<DFP_ALL:mode><BFP:mode>2")
12587 ("trunc<BFP:mode><DFP_ALL:mode>2")
12588 ("trunc<DFP_ALL:mode><BFP:mode>2")
12589 ("*extend<BFP:mode><DFP_ALL:mode>2")
12590 ("*extend<DFP_ALL:mode><BFP:mode>2")
12591 ("extend<BFP:mode><DFP_ALL:mode>2")
12592 ("extend<DFP_ALL:mode><BFP:mode>2"): Replace FPR2_REGNUM with
12593 FPR4_REGNUM.
12594
e6ac0270 125952013-07-08 Graham Stott <graham.stott@btinternet.com>
e6ac0270 12596
6a2fa4b2 12597 * Makefile.in: (c-family-warn): Define to $(STRICT_WARN)
64a92610 12598
b89b22fc
AK		 125992013-07-08 Andreas Krebbel <Andreas.Krebbel@de.ibm.com>
12600
12601 * config/s390/s390.c: Rename cfun_set_fpr_bit to cfun_set_fpr_save
12602 and cfun_fpr_bit_p to cfun_fpr_save_p.
12603 (s390_frame_area, s390_register_info, s390_frame_info)
12604 (s390_emit_prologue, s390_emit_epilogue)
12605 (s390_conditional_register_usage): Use the *_REGNUM macros for FPR
12606 register numbers.
12607 * config/s390/s390.h: Define *_REGNUM macros for floating point
12608 register numbers.
12609
07f398aa
EB		 126102013-07-08 Eric Botcazou <ebotcazou@adacore.com>
12611
12612 * Makefile.in (tree-ssa-reassoc.o): Add dependency on $(PARAMS_H).
12613
75fefa91
PCC		 126142013-07-08 Po-Chun Chang <pchang9@cs.wisc.edu>
12615
12616 PR rtl-optimization/57786
12617 * combine.c (distribute_notes) <case REG_DEAD>: Change all_used to bool
12618 and break out of the loop when it is set to false.
12619
046f1eee
JJ		 126202013-07-08 Jakub Jelinek <jakub@redhat.com>
12621
8140c065
JJ		 12622 PR target/57819
12623 * simplify-rtx.c (simplify_unary_operation_1) <case ZERO_EXTEND>:
12624 Simplify (zero_extend:SI (subreg:QI (and:SI (reg:SI)
12625 (const_int 63)) 0)).
12626 * combine.c (make_extraction): Create ZERO_EXTEND or SIGN_EXTEND
12627 using simplify_gen_unary instead of gen_rtx_*_EXTEND.
12628 * config/i386/i386.md (*jcc_bt<mode>_1): New define_insn_and_split.
12629
046f1eee
JJ		 12630 PR rtl-optimization/57829
12631 * simplify-rtx.c (simplify_binary_operation_1) <case IOR>: Ensure that
12632 mask bits outside of mode are just sign-extension from mode to HWI.
12633
930b700b
MZ		 126342013-07-08 Michael Zolotukhin <michael.v.zolotukhin@gmail.com>
12635
12636 * config/i386/i386-opts.h (enum stringop_alg): Add vector_loop.
12637 * config/i386/i386.c (expand_set_or_movmem_via_loop): Use
12638 adjust_address instead of change_address to keep info about alignment.
12639 (emit_strmov): Remove.
12640 (emit_memmov): New function.
12641 (expand_movmem_epilogue): Refactor to properly handle bigger sizes.
12642 (expand_movmem_epilogue): Likewise and return updated rtx for
12643 destination.
12644 (expand_constant_movmem_prologue): Likewise and return updated rtx for
12645 destination and source.
12646 (decide_alignment): Refactor, handle vector_loop.
12647 (ix86_expand_movmem): Likewise.
12648 (ix86_expand_setmem): Likewise.
12649 * config/i386/i386.opt (Enum): Add vector_loop to option stringop_alg.
12650
7b9d1bd8
UB		 126512013-07-07 Uros Bizjak <ubizjak@gmail.com>
12652
12653 * config/i386/driver-i386.c (host_detect_local_cpu): Do not check
12654 signature_TM2_ebx, it interferes with signature_INTEL_ebx.
12655
eabb5f48
UB		 126562013-07-06 Uros Bizjak <ubizjak@gmail.com>
12657
12658 * config/i386/sse.md (sse_movlhps): Change alternative 3
12659 of operand 2 to "m".
12660
126612013-07-06 Uros Bizjak <ubizjak@gmail.com>
12662
12663 PR target/57807
12664 * config/i386/sse.md (iptr): New mode attribute.
12665 (sse2_movq128): Add pointer size overrides for Intel asm dialect.
12666 (<sse>_vm<plusminus_insn><mode>3): Ditto.
12667 (<sse>_vmmul<mode>3): Ditto.
12668 (<sse>_vmdiv<mode>3): Ditto.
12669 (sse_vmrcpv4sf2): Ditto.
12670 (<sse>_vmsqrt<mode>2): Ditto.
12671 (sse_vmrsqrtv4sf2): Ditto.
12672 (<sse>_vm<code><mode>3): Ditto.
12673 (avx_vmcmp<mode>3): Ditto.
12674 (<sse>_vmmaskcmp<mode>3): Ditto.
12675 (<sse>_comi): Ditto.
12676 (<sse>_ucomi): Ditto.
12677 (*xop_vmfrcz_<mode>): Ditto.
12678 (*fmai_fmadd_<mode>): Ditto.
12679 (*fmai_fmsub_<mode>): Ditto.
12680 (*fmai_fnmadd_<mode>): Ditto.
12681 (*fmai_fnmsub_<mode>): Ditto.
12682 (*fma4i_vmfmadd_<mode>): Ditto.
12683 (*fma4i_vmfmsub_<mode>): Ditto.
12684 (*fma4i_vmfnmadd_<mode>): Ditto.
12685 (*fma4i_vmfnmsub_<mode>): Ditto.
12686 (*xop_vmfrcz_<mode>): Ditto.
12687 (sse_cvtps2pi): Ditto.
12688 (sse_cvttps2pi): Ditto.
12689 (sse_cvtss2si): Ditto.
12690 (sse_cvtss2si_2): Ditto.
12691 (sse_cvtss2siq_2): Ditto.
12692 (sse_cvttss2si): Ditto.
12693 (sse_cvttss2siq): Ditto.
12694 (sse_cvtsd2si): Ditto.
12695 (sse_cvtsd2si_2): Ditto.
12696 (sse_cvtsd2siq_2): Ditto.
12697 (sse_cvttsd2si): Ditto.
12698 (sse_cvttsd2siq): Ditto.
12699 (sse_cvtsd2ss): Ditto.
12700 (sse_cvtss2sd): Ditto.
12701 (avx2_pbroadcast<mode>): Ditto.
12702 (avx2_pbroadcast<mode>_1): Ditto.
12703 (*avx_vperm_broadcast_v4sf): Ditto.
12704
12705 (sse_movhlps): Ditto for movlp[sd]/movhp[sd] alternatives.
12706 (sse_movlhps): Ditto.
12707 (sse_storehps): Ditto.
12708 (sse_loadhps): Ditto.
12709 (sse_storelps): Ditto.
12710 (sse_loadlps): Ditto.
12711 (*vec_concatv4sf): Ditto.
12712 (*vec_interleave_highv2df): Ditto.
12713 (*vec_interleave_lowv2df): Ditto.
12714 (*vec_extractv2df_1_sse): Ditto.
12715 (*vec_extractv2df_0_sse): Ditto.
12716 (sse2_storelpd): Ditto.
12717 (sse2_loadlpd): Ditto.
12718 (sse2_movsd): Ditto.
12719 (*vec_concatv4si): Ditto.
12720 (vec_concatv2di): Ditto.
12721
12722 * config/i386/mmx.md (mmx_punpcklbw): Add pointer size overrides
12723 for Intel asm dialect.
12724 (mmx_punpcklwd): Ditto.
12725 (mmx_punpckldq): Ditto.
12726
12727 * config/i386/i386.c (ix86_print_operand) ['H']: Output 'qword ptr'
12728 for intel assembler dialect.
12729
1f6eac90
JJ		 127302013-07-06 Jakub Jelinek <jakub@redhat.com>
12731
12732 PR target/29776
12733 * fold-const.c (tree_call_nonnegative_warnv_p): Return true
12734 for BUILT_IN_C{LZ,LRSB}*.
12735 * tree.h (CASE_INT_FN): Add FN##IMAX case.
12736 * tree-vrp.c (extract_range_basic): Handle
12737 BUILT_IN_{FFS,PARITY,POPCOUNT,C{LZ,TZ,LRSB}}*. For
12738 BUILT_IN_CONSTANT_P if argument isn't (D) of PARM_DECL,
12739 fall thru to code calling set_value*.
12740 * builtins.c (expand_builtin): Remove *IMAX cases.
12741 (fold_builtin_bitop): For BUILT_IN_CLRSB* return NULL_TREE
12742 if width is bigger than 2*HWI.
12743
2b778c9d
VM		 127442013-07-05 Vladimir Makarov <vmakarov@redhat.com>
12745
12746 PR rtl-optimization/55342
12747 * lra-int.h (lra_subreg_reload_pseudos): New.
12748 * lra.c: Add undoing optional reloads to the block diagram.
12749 (lra_subreg_reload_pseudos): New.
12750 (lra_optional_reload_pseudos): Change comments.
12751 (lra): Init and clear lra_subreg_reload_pseudos. Clear
12752 lra_optional_reload_pseudos after undo transformations.
12753 * lra-assigns.c (pseudo_prefix_title): New.
12754 (lra_setup_reg_renumber): Use it.
12755 (spill_for): Ditto. Check subreg reload pseudos too.
12756 (assign_by_spills): Consider subreg reload pseudos too.
12757 * lra-constraints.c (simplify_operand_subreg): Use
12758 lra_subreg_reload_pseudos instead of lra_optional_reload_pseudos.
12759 (curr_insn_transform): Recognize and do optional reloads.
12760 (undo_optional_reloads): New.
12761 (lra_undo_inheritance): Call undo_optional_reloads.
12762
207156e4
TQ		 127632013-07-05 Thomas Quinot <quinot@adacore.com>
12764
12765 * tree-complex.c (expand_complex_operations_1): Fix typo.
12766
8990e73a
TB		 127672013-07-04 Tejas Belagod <tejas.belagod@arm.com>
12768
12769 * config/aarch64/aarch64-protos.h (cpu_vector_cost): New.
12770 (tune_params): New member 'const vec_costs'.
12771 * config/aarch64/aarch64.c (generic_vector_cost): New.
12772 (generic_tunings): New member 'generic_vector_cost'.
12773 (aarch64_builtin_vectorization_cost): New.
12774 (aarch64_add_stmt_cost): New.
12775 (TARGET_VECTORIZE_ADD_STMT_COST): New.
12776 (TARGET_VECTORIZE_BUILTIN_VECTORIZATION_COST): New.
12777
f7bc421d
JJ		 127782013-07-03 Jakub Jelinek <jakub@redhat.com>
12779
12780 PR target/57777
12781 * config/i386/predicates.md (vsib_address_operand): Disallow
12782 SYMBOL_REF or LABEL_REF in parts.disp if TARGET_64BIT && flag_pic.
12783
e90d1568
HPN		 127842013-07-03 Hans-Peter Nilsson <hp@bitrange.com>
12785
12786 PR middle-end/55030
12787 * stmt.c (expand_nl_goto_receiver): Remove almost-copy of
12788 expand_builtin_setjmp_receiver.
12789 (expand_label): Adjust, call expand_builtin_setjmp_receiver
12790 with NULL for the label parameter.
12791 * builtins.c (expand_builtin_setjmp_receiver): Don't clobber
12792 the frame-pointer. Adjust comments.
12793 [HAVE_builtin_setjmp_receiver]: Emit builtin_setjmp_receiver
12794 only if LABEL is non-NULL.
12795
cbd7413b
YZ		 127962013-07-03 Yufeng Zhang <yufeng.zhang@arm.com>
12797
12798 * config/aarch64/aarch64.h (enum arm_abi_type): Remove.
12799 (ARM_ABI_AAPCS64): Ditto.
12800 (arm_abi): Ditto.
12801 (ARM_DEFAULT_ABI): Ditto.
12802
dec11868
JG		 128032013-07-03 James Greenhalgh <james.greenhalgh@arm.com>
12804
12805 * config/aarch64/aarch64-builtins.c
12806 (aarch64_simd_expand_builtin): Handle AARCH64_SIMD_STORE1.
12807 * config/aarch64/aarch64-simd-builtins.def (ld1): New.
12808 (st1): Likewise.
12809 * config/aarch64/aarch64-simd.md
12810 (aarch64_ld1<VALL:mode>): New.
12811 (aarch64_st1<VALL:mode>): Likewise.
12812 * config/aarch64/arm_neon.h
12813 (vld1<q>_<fpsu><8, 16, 32, 64>): Convert to RTL builtins.
12814
786fb9b6
ST		 128152013-07-02 Sriraman Tallam <tmsriram@google.com>
12816
0ba6aff3
UB		 12817 * config/i386/i386.c (gate_insert_vzeroupper): Check if
12818 target ISA is AVX.
786fb9b6
ST		 12819 (ix86_option_override_internal):Turn on all -mavx target flags by
12820 default as they are dependent on AVX anyway.
12821
4b043553
CC		 128222013-07-02 Cary Coutant <ccoutant@google.com>
12823
12824 * dwarf2out.c (loc_checksum): Call hash_loc_operands for a
12825 deterministic hash.
12826 (loc_checksum_ordered): Likewise.
12827 (hash_loc_operands): Remove inline keyword.
12828
c4e87a13
JJ		 128292013-07-02 Jakub Jelinek <jakub@redhat.com>
12830
12831 PR tree-optimization/57741
12832 * tree-vect-loop.c (vect_is_simple_iv_evolution): Disallow
12833 non-INTEGRAL_TYPE_P non-SCALAR_FLOAT_TYPE_P SSA_NAME step_exprs,
12834 or SCALAR_FLOAT_TYPE_P SSA_NAMEs if !flag_associative_math.
12835 Allow REAL_CST step_exprs if flag_associative_math.
12836 (get_initial_def_for_induction): Handle SCALAR_FLOAT_TYPE_P step_expr.
12837
26366d28
IB		 128382013-07-02 Ian Bolton <ian.bolton@arm.com>
12839
0ba6aff3 12840 * config/aarch64/aarch64-simd.md (absdi2): Support abs for DI mode.
26366d28 12841
2879bb2b
IB		 128422013-07-02 Ian Bolton <ian.bolton@arm.com>
12843
12844 * config/aarch64/aarch64.md (*extr_insv_reg<mode>): New pattern.
12845
afc5e8a6
KT		 128462013-07-02 Kyrylo Tkachov <kyrylo.tkachov@arm.com>
12847
12848 * config/arm/arm.md (arm_andsi3_insn): Add alternatives for 16-bit
12849 encoding.
12850 (iorsi3_insn): Likewise.
12851 (arm_xorsi3): Likewise.
12852
31f8442b
SN		 128532013-07-01 Sofiane Naci <sofiane.naci@arm.com>
12854
12855 * arm.md (attribute "wtype"): Delete. Move attribute values from here
12856 to ...
12857 (attribute "type"): ... here, and prefix with "wmmx_".
12858 (attribute "core_cycles"): Update for attribute changes.
12859 * iwmmxt.md (tbcstv8qi): Update for attribute changes.
12860 (tbcstv4hi): Likewise.
12861 (tbcstv2si): Likewise.
12862 (iwmmxt_iordi3): Likewise.
12863 (iwmmxt_xordi3): Likewise.
12864 (iwmmxt_anddi3): Likewise.
12865 (iwmmxt_nanddi3): Likewise.
12866 (iwmmxt_arm_movdi): Likewise.
12867 (iwmmxt_movsi_insn): Likewise.
12868 (mov<mode>_internal): Likewise.
12869 (and<mode>3_iwmmxt): Likewise.
12870 (ior<mode>3_iwmmxt): Likewise.
12871 (xor<mode>3_iwmmxt): Likewise.
12872 (add<mode>3_iwmmxt): Likewise.
12873 (ssaddv8qi3): Likewise.
12874 (ssaddv4hi3): Likewise.
12875 (ssaddv2si3): Likewise.
12876 (usaddv8qi3): Likewise.
12877 (usaddv4hi3): Likewise.
12878 (usaddv2si3): Likewise.
12879 (sub<mode>3_iwmmxt): Likewise.
12880 (sssubv8qi3): Likewise.
12881 (sssubv4hi3): Likewise.
12882 (sssubv2si3): Likewise.
12883 (ussubv8qi3): Likewise.
12884 (ussubv4hi3): Likewise.
12885 (ussubv2si3): Likewise.
12886 (mulv4hi3_iwmmxt): Likewise.
12887 (smulv4hi3_highpart): Likewise.
12888 (umulv4hi3_highpart): Likewise.
12889 (iwmmxt_wmacs): Likewise.
12890 (iwmmxt_wmacsz): Likewise.
12891 (iwmmxt_wmacu): Likewise.
12892 (iwmmxt_wmacuz): Likewise.
12893 (iwmmxt_clrdi): Likewise.
12894 (iwmmxt_clrv8qi): Likewise.
12895 (iwmmxt_clr4hi): Likewise.
12896 (iwmmxt_clr2si): Likewise.
12897 (iwmmxt_uavgrndv8qi3): Likewise.
12898 (iwmmxt_uavgrndv4hi3): Likewise.
12899 (iwmmxt_uavgv8qi3): Likewise.
12900 (iwmmxt_uavgv4hi3): Likewise.
12901 (iwmmxt_tinsrb): Likewise.
12902 (iwmmxt_tinsrh): Likewise.
12903 (iwmmxt_tinsrw): Likewise.
12904 (iwmmxt_textrmub): Likewise.
12905 (iwmmxt_textrmsb): Likewise.
12906 (iwmmxt_textrmuh): Likewise.
12907 (iwmmxt_textrmsh): Likewise.
12908 (iwmmxt_textrmw): Likewise.
12909 (iwmxxt_wshufh): Likewise.
12910 (eqv8qi3): Likewise.
12911 (eqv4hi3): Likewise.
12912 (eqv2si3): Likewise.
12913 (gtuv8qi3): Likewise.
12914 (gtuv4hi3): Likewise.
12915 (gtuv2si3): Likewise.
12916 (gtv8qi3): Likewise.
12917 (gtv4hi3): Likewise.
12918 (gtv2si3): Likewise.
12919 (smax<mode>3_iwmmxt): Likewise.
12920 (umax<mode>3_iwmmxt): Likewise.
12921 (smin<mode>3_iwmmxt): Likewise.
12922 (umin<mode>3_iwmmxt): Likewise.
12923 (iwmmxt_wpackhss): Likewise.
12924 (iwmmxt_wpackwss): Likewise.
12925 (iwmmxt_wpackdss): Likewise.
12926 (iwmmxt_wpackhus): Likewise.
12927 (iwmmxt_wpackwus): Likewise.
12928 (iwmmxt_wpackdus): Likewise.
12929 (iwmmxt_wunpckihb): Likewise.
12930 (iwmmxt_wunpckihh): Likewise.
12931 (iwmmxt_wunpckihw): Likewise.
12932 (iwmmxt_wunpckilb): Likewise.
12933 (iwmmxt_wunpckilh): Likewise.
12934 (iwmmxt_wunpckilw): Likewise.
12935 (iwmmxt_wunpckehub): Likewise.
12936 (iwmmxt_wunpckehuh): Likewise.
12937 (iwmmxt_wunpckehuw): Likewise.
12938 (iwmmxt_wunpckehsb): Likewise.
12939 (iwmmxt_wunpckehsh): Likewise.
12940 (iwmmxt_wunpckehsw): Likewise.
12941 (iwmmxt_wunpckelub): Likewise.
12942 (iwmmxt_wunpckeluh): Likewise.
12943 (iwmmxt_wunpckeluw): Likewise.
12944 (iwmmxt_wunpckelsb): Likewise.
12945 (iwmmxt_wunpckelsh): Likewise.
12946 (iwmmxt_wunpckelsw): Likewise.
12947 (ror<mode>3): Likewise.
12948 (ashr<mode>3_iwmmxt): Likewise.
12949 (lshr<mode>3_iwmmxt): Likewise.
12950 (ashl<mode>3_iwmmxt): Likewise.
12951 (ror<mode>3_di): Likewise.
12952 (ashr<mode>3_di): Likewise.
12953 (lshr<mode>3_di): Likewise.
12954 (ashl<mode>3_di): Likewise.
12955 (iwmmxt_wmadds): Likewise.
12956 (iwmmxt_wmaddu): Likewise.
12957 (iwmmxt_tmia): Likewise.
12958 (iwmmxt_tmiaph): Likewise.
12959 (iwmmxt_tmiabb): Likewise.
12960 (iwmmxt_tmiatb): Likewise.
12961 (iwmmxt_tmiabt): Likewise.
12962 (iwmmxt_tmiatt): Likewise.
12963 (iwmmxt_tmovmskb): Likewise.
12964 (iwmmxt_tmovmskh): Likewise.
12965 (iwmmxt_tmovmskw): Likewise.
12966 (iwmmxt_waccb): Likewise.
12967 (iwmmxt_wacch): Likewise.
12968 (iwmmxt_waccw): Likewise.
12969 (iwmmxt_waligni): Likewise.
12970 (iwmmxt_walignr): Likewise.
12971 (iwmmxt_walignr0): Likewise.
12972 (iwmmxt_walignr1): Likewise.
12973 (iwmmxt_walignr2): Likewise.
12974 (iwmmxt_walignr3): Likewise.
12975 (iwmmxt_wsadb): Likewise.
12976 (iwmmxt_wsadh): Likewise.
12977 (iwmmxt_wsadbz): Likewise.
12978 (iwmmxt_wsadhz): Likewise.
12979 * iwmmxt2.md (iwmmxt_wabs<mode>3): Update for attribute changes.
12980 (iwmmxt_wabsdiffb): Likewise.
12981 (iwmmxt_wabsdiffh): Likewise.
12982 (iwmmxt_wabsdiffw): Likewise.
12983 (iwmmxt_waddsubhx): Likewise
12984 (iwmmxt_wsubaddhx): Likewise.
12985 (addc<mode>3): Likewise.
12986 (iwmmxt_avg4): Likewise.
12987 (iwmmxt_avg4r): Likewise.
12988 (iwmmxt_wmaddsx): Likewise.
12989 (iwmmxt_wmaddux): Likewise.
12990 (iwmmxt_wmaddsn): Likewise.
12991 (iwmmxt_wmaddun): Likewise.
12992 (iwmmxt_wmulwsm): Likewise.
12993 (iwmmxt_wmulwum): Likewise.
12994 (iwmmxt_wmulsmr): Likewise.
12995 (iwmmxt_wmulumr): Likewise.
12996 (iwmmxt_wmulwsmr): Likewise.
12997 (iwmmxt_wmulwumr): Likewise.
12998 (iwmmxt_wmulwl): Likewise.
12999 (iwmmxt_wqmulm): Likewise.
13000 (iwmmxt_wqmulwm): Likewise.
13001 (iwmmxt_wqmulmr): Likewise.
13002 (iwmmxt_wqmulwmr): Likewise.
13003 (iwmmxt_waddbhusm): Likewise.
13004 (iwmmxt_waddbhusl): Likewise.
13005 (iwmmxt_wqmiabb): Likewise.
13006 (iwmmxt_wqmiabt): Likewise.
13007 (iwmmxt_wqmiatb): Likewise.
13008 (iwmmxt_wqmiatt): Likewise.
13009 (iwmmxt_wqmiabbn): Likewise.
13010 (iwmmxt_wqmiabtn): Likewise.
13011 (iwmmxt_wqmiatbn): Likewise.
13012 (iwmmxt_wqmiattn): Likewise.
13013 (iwmmxt_wmiabb): Likewise.
13014 (iwmmxt_wmiabt): Likewise.
13015 (iwmmxt_wmiatb): Likewise.
13016 (iwmmxt_wmiatt): Likewise.
13017 (iwmmxt_wmiabbn): Likewise.
13018 (iwmmxt_wmiabtn): Likewise.
13019 (iwmmxt_wmiatbn): Likewise.
13020 (iwmmxt_wmiattn): Likewise.
13021 (iwmmxt_wmiawbb): Likewise.
13022 (iwmmxt_wmiawbt): Likewise.
13023 (iwmmxt_wmiawtb): Likewise.
13024 (iwmmxt_wmiawtt): Likewise.
13025 (iwmmxt_wmiawbbn): Likewise.
13026 (iwmmxt_wmiawbtn): Likewise.
13027 (iwmmxt_wmiawtbn): Likewise.
13028 (iwmmxt_wmiawttn): Likewise.
13029 (iwmmxt_wmerge): Likewise.
13030 (iwmmxt_tandc<mode>3): Likewise.
13031 (iwmmxt_torc<mode>3): Likewise.
13032 (iwmmxt_torvsc<mode>3): Likewise.
13033 (iwmmxt_textrc<mode>3): Likewise.
13034 * marvell-f-iwmmxt.md (wmmxt_shift): Update for attribute changes.
13035 (wmmxt_pack): Likewise.
13036 (wmmxt_mult_c1): Likewise.
13037 (wmmxt_mult_c2): Likewise.
13038 (wmmxt_alu_c1): Likewise.
13039 (wmmxt_alu_c2): Likewise.
13040 (wmmxt_alu_c3): Likewise.
13041 (wmmxt_transfer_c1): Likewise.
13042 (wmmxt_transfer_c2): Likewise.
13043 (wmmxt_transfer_c3): Likewise.
13044 (marvell_f_iwmmxt_wstr): Likewise.
13045 (marvell_f_iwmmxt_wldr): Likewise.
13046
bdb7bf8a
YZ		 130472013-06-29 Yufeng Zhang <yufeng.zhang@arm.com>
13048
0ba6aff3 13049 * config/aarch64/aarch64.c: Remove junk from the beginning of the file.
bdb7bf8a 13050
e6631d38
VM		 130512013-06-28 Vladimir Makarov <vmakarov@redhat.com>
13052
13053 Revert:
13054 2013-06-28 Vladimir Makarov <vmakarov@redhat.com>
13055 * lra-constraints.c (need_for_split_p): Check call used hard regs
13056 living through calls.
13057
13058 * lra-constraints.c (inherit_in_ebb): Reset live_hard_regs for
13059 call used regs for call insn.
13060
53c1275b
JJ		 130612013-06-28 Jakub Jelinek <jakub@redhat.com>
13062
13063 PR target/57736
0ba6aff3
UB		 13064 * config/i386/i386.c (ix86_expand_builtin): If target == NULL and
13065 mode is VOIDmode, don't create a VOIDmode pseudo to copy result into.
53c1275b 13066
1141ed3f
BI		 130672013-06-28 Balaji V. Iyer <balaji.v.iyer@intel.com>
13068
13069 * builtins.def: Fixed the function type of CILKPLUS_BUILTIN.
e6631d38 13070
e32e4c4a
VM		 130712013-06-28 Vladimir Makarov <vmakarov@redhat.com>
13072
13073 * lra-constraints.c (need_for_split_p): Check call used hard regs
13074 living through calls.
13075
f161bfd3
MM		 130762013-06-28 Michael Meissner <meissner@linux.vnet.ibm.com>
13077
13078 PR target/57744
13079 * config/rs6000/rs6000.h (MODES_TIEABLE_P): Do not allow PTImode
13080 to tie with any other modes. Eliminate Altivec vector mode tests,
13081 since these are a subset of ALTIVEC or VSX vector modes. Simplify
13082 code, to return 0 if testing MODE2 for a condition, if we've
13083 already tested MODE1 for the same condition.
13084
7eda14e1
MS		 130852013-06-28 Marcus Shawcroft <marcus.shawcroft@arm.com>
13086
13087 * config/aarch64/aarch64.c (aarch64_cannot_force_const_mem): Adjust
13088 layout.
13089
c822f852
MS		 130902013-06-28 Marcus Shawcroft <marcus.shawcroft@arm.com>
13091
13092 * config/aarch64/aarch64-protos.h (aarch64_symbol_type):
13093 Update comment w.r.t SYMBOL_TINY_ABSOLUTE.
13094
da4f13a4
MS		 130952013-06-28 Marcus Shawcroft <marcus.shawcroft@arm.com>
13096
0ba6aff3
UB		 13097 * config/aarch64/aarch64-protos.h (aarch64_classify_symbol_expression):
13098 Define.
da4f13a4
MS		 13099 (aarch64_symbolic_constant_p): Remove.
13100 * config/aarch64/aarch64.c (aarch64_classify_symbol_expression): Remove
13101 static. Fix line length and white space.
13102 (aarch64_symbolic_constant_p): Remove.
13103 * config/aarch64/predicates.md (aarch64_valid_symref):
13104 Use aarch64_classify_symbol_expression.
13105
956a95a5
KT		 131062013-06-28 Kyrylo Tkachov <kyrylo.tkachov@arm.com>
13107
13108 * config/arm/constraints.md (Ts): New constraint.
13109 * config/arm/arm.md (arm_movqi_insn): Add alternatives for
13110 16-bit encodings.
13111 (compare_scc): Use "Ts" constraint for operand 0.
13112 (ior_scc_scc): Likewise.
13113 (and_scc_scc): Likewise.
13114 (and_scc_scc_nodom): Likewise.
13115 (ior_scc_scc_cmp): Likewise for operand 7.
13116 (and_scc_scc_cmp): Likewise.
13117 * config/arm/thumb2.md (thumb2_movsi_insn):
13118 Add alternatives for 16-bit encodings.
13119 (thumb2_movhi_insn): Likewise.
13120 (thumb2_movsicc_insn): Likewise.
13121 (thumb2_and_scc): Take 'and' outside cond_exec. Use "Ts" constraint.
13122 (thumb2_negscc): Use "Ts" constraint.
13123 Move mvn instruction outside cond_exec block.
13124 * config/arm/vfp.md (thumb2_movsi_vfp): Add alternatives
13125 for 16-bit encodings.
13126
31dfce10
KT		 131272013-06-28 Kyrylo Tkachov <kyrylo.tkachov@arm.com>
13128
13129 * config/arm/arm.md (arm_mulsi3_v6): Add alternative for 16-bit
13130 encoding.
13131 (mulsi3addsi_v6): Disable predicable variant for arm_restrict_it.
13132 (mulsi3subsi): Likewise.
13133 (mulsidi3adddi): Likewise.
13134 (mulsidi3_v6): Likewise.
13135 (umulsidi3_v6): Likewise.
13136 (umulsidi3adddi_v6): Likewise.
13137 (smulsi3_highpart_v6): Likewise.
13138 (umulsi3_highpart_v6): Likewise.
13139 (mulhisi3tb): Likewise.
13140 (mulhisi3bt): Likewise.
13141 (mulhisi3tt): Likewise.
13142 (maddhisi4): Likewise.
13143 (maddhisi4tb): Likewise.
13144 (maddhisi4tt): Likewise.
13145 (maddhidi4): Likewise.
13146 (maddhidi4tb): Likewise.
13147 (maddhidi4tt): Likewise.
13148 (zeroextractsi_compare0_scratch): Likewise.
13149 (insv_zero): Likewise.
13150 (insv_t2): Likewise.
13151 (anddi_notzesidi_di): Likewise.
13152 (anddi_notsesidi_di): Likewise.
13153 (andsi_notsi_si): Likewise.
13154 (iordi_zesidi_di): Likewise.
13155 (xordi_zesidi_di): Likewise.
13156 (andsi_iorsi3_notsi): Likewise.
13157 (smax_0): Likewise.
13158 (smax_m1): Likewise.
13159 (smin_0): Likewise.
13160 (not_shiftsi): Likewise.
13161 (unaligned_loadsi): Likewise.
13162 (unaligned_loadhis): Likewise.
13163 (unaligned_loadhiu): Likewise.
13164 (unaligned_storesi): Likewise.
13165 (unaligned_storehi): Likewise.
13166 (extv_reg): Likewise.
13167 (extzv_t2): Likewise.
13168 (divsi3): Likewise.
13169 (udivsi3): Likewise.
13170 (arm_zero_extendhisi2addsi): Likewise.
13171 (arm_zero_extendqisi2addsi): Likewise.
13172 (compareqi_eq0): Likewise.
13173 (arm_extendhisi2_v6): Likewise.
13174 (arm_extendqisi2addsi): Likewise.
13175 (arm_movt): Likewise.
13176 (thumb2_ldrd): Likewise.
13177 (thumb2_ldrd_base): Likewise.
13178 (thumb2_ldrd_base_neg): Likewise.
13179 (thumb2_strd): Likewise.
13180 (thumb2_strd_base): Likewise.
13181 (thumb2_strd_base_neg): Likewise.
13182 (arm_negsi2): Add alternative for 16-bit encoding.
13183 (arm_one_cmplsi2): Likewise.
13184
1572e697
KT		 131852013-06-28 Kyrylo Tkachov <kyrylo.tkachov@arm.com>
13186
13187 * config/arm/predicates.md (arm_cond_move_operator): New predicate.
13188 * config/arm/arm.md (movsfcc): Use arm_cond_move_operator predicate.
13189 (movdfcc): Likewise.
13190 * config/arm/vfp.md (*thumb2_movsf_vfp):
13191 Disable predication for arm_restrict_it.
13192 (*thumb2_movsfcc_vfp): Disable for arm_restrict_it.
13193 (*thumb2_movdfcc_vfp): Likewise.
13194 (*abssf2_vfp, *absdf2_vfp, *negsf2_vfp, *negdf2_vfp,*addsf3_vfp,
13195 *adddf3_vfp, *subsf3_vfp, *subdf3_vfpc, *divsf3_vfp,*divdf3_vfp,
13196 *mulsf3_vfp, *muldf3_vfp, *mulsf3negsf_vfp, *muldf3negdf_vfp,
13197 *mulsf3addsf_vfp, *muldf3adddf_vfp, *mulsf3subsf_vfp,
13198 *muldf3subdf_vfp, *mulsf3negsfaddsf_vfp, *fmuldf3negdfadddf_vfp,
13199 *mulsf3negsfsubsf_vfp, *muldf3negdfsubdf_vfp, *fma<SDF:mode>4,
13200 *fmsub<SDF:mode>4, *fnmsub<SDF:mode>4, *fnmadd<SDF:mode>4,
13201 *extendsfdf2_vfp, *truncdfsf2_vfp, *extendhfsf2, *truncsfhf2,
13202 *truncsisf2_vfp, *truncsidf2_vfp, fixuns_truncsfsi2, fixuns_truncdfsi2,
13203 *floatsisf2_vfp, *floatsidf2_vfp, floatunssisf2, floatunssidf2,
13204 *sqrtsf2_vfp, *sqrtdf2_vfp, *cmpsf_vfp, *cmpsf_trap_vfp, *cmpdf_vfp,
13205 *cmpdf_trap_vfp, <vrint_pattern><SDF:mode>2):
13206 Disable predication for arm_restrict_it.
13207
ebf8f0ea
KY		 132082013-06-28 Kirill Yukhin <kirill.yukhin@intel.com>
13209
13210 * config/i386/bmiintrin.h (_bextr_u32): New.
13211 (_bextr_u64): Ditto.
13212
a290fcda
RS		 132132013-06-27 Richard Sandiford <rdsandiford@googlemail.com>
13214
13215 * config.gcc (mips*-mti-elf*, mips*-sde-elf*, mips64r5900-*-elf*)
13216 (mips64r5900el-*-elf*): Include mips/n32-elf.h.
13217 * config/mips/sde.h (LOCAL_LABEL_PREFIX, NO_DOLLAR_IN_LABEL)
13218 (LONG_DOUBLE_TYPE_SIZE, LIBGCC2_LONG_DOUBLE_TYPE_SIZE): Move to...
13219 * config/mips/n32-elf.h: ...this new file.
13220
ea169996
MG		 132212013-06-27 Marc Glisse <marc.glisse@inria.fr>
13222
13223 PR target/57224
13224 * config/i386/i386.c (enum ix86_builtins, bdesc_args): Remove
13225 IX86_BUILTIN_CMPNGTSS and IX86_BUILTIN_CMPNGESS.
13226
12211b99 132272013-06-27 Catherine Moore <clm@codesourcery.com>
8e932114
CM		 13228
13229 * config/mips/mips-tables.opt: Regenerate.
13230 * config/mips/mips-cpus.def: Add m14ke and m14kec.
0ba6aff3 13231 * config/mips/mips.h (BASE_DRIVER_SELF_SPECS): m14ke* implies -mdspr2.
8e932114 13232 * doc/invoke.texi: Add -m14kc.
ebf8f0ea 13233
2444244e
JJ		 132342013-06-27 Jakub Jelinek <jakub@redhat.com>
13235
cd06d2a2
JJ		 13236 PR target/57623
13237 * config/i386/i386.md (bmi_bextr_<mode>): Swap predicates and
13238 constraints of operand 1 and 2.
13239
2444244e
JJ		 13240 PR target/57623
13241 * config/i386/i386.md (bmi2_bzhi_<mode>3): Swap AND arguments
13242 to match RTL canonicalization. Swap predicates and
13243 constraints of operand 1 and 2.
13244
2f259720
VM		 132452013-06-27 Vladimir Makarov <vmakarov@redhat.com>
13246
0ba6aff3
UB		 13247 * lra-constraints.c (inherit_in_ebb): Process static hard regs too.
13248 Process OP_INOUT regs for splitting too.
2f259720 13249
fef4d2b3
JJ		 132502013-06-27 Jakub Jelinek <jakub@redhat.com>
13251
13252 * tree-vect-stmts.c (vectorizable_store): Move ptr_incr var
13253 decl before the loop, initialize to NULL.
13254 (vectorizable_load): Initialize ptr_incr to NULL.
13255
79ee9826
MJ		 132562013-06-27 Martin Jambor <mjambor@suse.cz>
13257
13258 PR lto/57208
13259 * ipa-ref.h (ipa_maybe_record_reference): Declare.
13260 * ipa-ref.c (ipa_maybe_record_reference): New function.
13261 * cgraphclones.c (cgraph_create_virtual_clone): Use it.
13262 * ipa-cp.c (create_specialized_node): Record potential references from
13263 aggvals.
13264 * Makefile.in (ipa-ref.o): Add IPA_REF_H to dependencies.
13265
e18b4a81
YZ		 132662013-06-27 Yufeng Zhang <yufeng.zhang@arm.com>
13267
13268 * config/aarch64/aarch64.c (aarch64_force_temporary): Add an extra
13269 parameter 'mode' of type 'enum machine_mode mode'; change to pass
13270 'mode' to force_reg.
13271 (aarch64_add_offset): Update calls to aarch64_force_temporary.
13272 (aarch64_expand_mov_immediate): Likewise.
13273
9c023bf0
YZ		 132742013-06-27 Yufeng Zhang <yufeng.zhang@arm.com>
13275
13276 * config/aarch64/aarch64.c (aarch64_add_offset): Change to pass
13277 'mode' to aarch64_plus_immediate and gen_rtx_PLUS.
13278
5a3fe9b6
AK		 132792013-06-27 Andreas Krebbel <Andreas.Krebbel@de.ibm.com>
13280
13281 * config/s390/s390.c: Rename UNSPEC_CCU_TO_INT to
13282 UNSPEC_STRCMPCC_TO_INT and UNSPEC_CCZ_TO_INT to UNSPEC_CC_TO_INT.
13283 (struct machine_function): Add tbegin_p.
13284 (s390_canonicalize_comparison): Fold CC mode compares to
13285 conditional jump if possible.
13286 (s390_emit_jump): Return the emitted jump.
13287 (s390_branch_condition_mask, s390_branch_condition_mnemonic):
13288 Handle CCRAWmode compares.
13289 (s390_option_override): Default to -mhtm if available.
13290 (s390_reg_clobbered_rtx): Handle floating point regs as well.
13291 (s390_regs_ever_clobbered): Use s390_regs_ever_clobbered also for
13292 FPRs instead of df_regs_ever_live_p.
13293 (s390_optimize_nonescaping_tx): New function.
13294 (s390_init_frame_layout): Extend clobbered_regs array to cover
13295 FPRs as well.
13296 (s390_emit_prologue): Call s390_optimize_nonescaping_tx.
13297 (s390_expand_tbegin): New function.
13298 (enum s390_builtin): New enum definition.
13299 (code_for_builtin): New array definition.
13300 (s390_init_builtins): New function.
13301 (s390_expand_builtin): New function.
13302 (TARGET_INIT_BUILTINS): Define.
13303 (TARGET_EXPAND_BUILTIN): Define.
0ba6aff3 13304 * common/config/s390/s390-common.c (processor_flags_table): Add PF_TX.
5a3fe9b6
AK		 13305 * config/s390/predicates.md (s390_comparison): Handle CCRAWmode.
13306 (s390_alc_comparison): Likewise.
13307 * config/s390/s390-modes.def: Add CCRAWmode.
13308 * config/s390/s390.h (processor_flags): Add PF_TX.
13309 (TARGET_CPU_HTM): Define macro.
13310 (TARGET_HTM): Define macro.
13311 (TARGET_CPU_CPP_BUILTINS): Define __HTM__ for htm.
13312 * config/s390/s390.md: Rename UNSPEC_CCU_TO_INT to
13313 UNSPEC_STRCMPCC_TO_INT and UNSPEC_CCZ_TO_INT to UNSPEC_CC_TO_INT.
13314 (UNSPECV_TBEGIN, UNSPECV_TBEGINC, UNSPECV_TEND, UNSPECV_TABORT)
0ba6aff3 13315 (UNSPECV_ETND, UNSPECV_NTSTG, UNSPECV_PPA): New unspecv enum values.
5a3fe9b6
AK		 13316 (TBEGIN_MASK, TBEGINC_MASK): New constants.
13317 ("*cc_to_int"): Move up.
13318 ("*mov<mode>cc", "*cjump_64", "*cjump_31"): Accept integer
13319 constants other than 0.
13320 ("*ccraw_to_int"): New insn and splitter definition.
13321 ("tbegin", "tbegin_nofloat", "tbegin_retry")
13322 ("tbegin_retry_nofloat", "tbeginc", "tend", "tabort")
13323 ("tx_assist"): New expander.
13324 ("tbegin_1", "tbegin_nofloat_1", "*tbeginc_1", "*tend_1")
13325 ("*tabort_1", "etnd", "ntstg", "*ppa"): New insn definition.
13326 * config/s390/s390.opt: Add -mhtm option.
13327 * config/s390/s390-protos.h (s390_emit_jump): Add return type.
13328 * config/s390/htmxlintrin.h: New file.
13329 * config/s390/htmintrin.h: New file.
13330 * config/s390/s390intrin.h: New file.
13331 * doc/extend.texi: Document htm builtins.
13332 * config.gcc: Add the new header files to extra_headers.
13333
9e216629
TS		 133342013-06-26 Thomas Schwinge <thomas@codesourcery.com>
13335
13336 * config/i386/gnu.h [TARGET_LIBC_PROVIDES_SSP]
0ba6aff3 13337 (TARGET_CAN_SPLIT_STACK, TARGET_THREAD_SPLIT_STACK_OFFSET): Undefine.
9e216629 13338
fd8c65e7 133392013-06-26 Michael Meissner <meissner@linux.vnet.ibm.com>
12211b99
UB		 13340 Pat Haugen <pthaugen@us.ibm.com>
13341 Peter Bergner <bergner@vnet.ibm.com>
fd8c65e7
MM		 13342
13343 * config/rs6000/power8.md: New.
13344 * config/rs6000/rs6000-cpus.def (RS6000_CPU table): Adjust processor
13345 setting for power8 entry.
13346 * config/rs6000/t-rs6000 (MD_INCLUDES): Add power8.md.
13347 * config/rs6000/rs6000.c (is_microcoded_insn, is_cracked_insn): Adjust
13348 test for Power4/Power5 only.
13349 (insn_must_be_first_in_group, insn_must_be_last_in_group): Add Power8
13350 support.
13351 (force_new_group): Adjust comment.
13352 * config/rs6000/rs6000.md: Include power8.md.
13353
b24a2ce5
GY		 133542013-06-26 Greta Yorsh <Greta.Yorsh@arm.com>
13355
13356 * config/arm/arm.h (MAX_CONDITIONAL_EXECUTE): Define macro.
13357 * config/arm/arm-protos.h (arm_max_conditional_execute): New
13358 declaration.
13359 (tune_params): Update comment.
13360 * config/arm/arm.c (arm_cortex_a15_tune): Set max_cond_insns to 2.
13361 (arm_max_conditional_execute): New function.
13362 (thumb2_final_prescan_insn): Use max_insn_skipped and
13363 MAX_INSN_PER_IT_BLOCK to compute maximum instructions in a block.
13364
f6fdeb62
JJ		 133652013-06-25 Jakub Jelinek <jakub@redhat.com>
13366
13367 PR tree-optimization/57705
13368 * tree-vect-loop.c (vect_is_simple_iv_evolution): Allow
13369 SSA_NAME step, provided that it is not defined inside the loop.
0ba6aff3 13370 (vect_analyze_scalar_cycles_1): Disallow SSA_NAME step in nested loop.
f6fdeb62
JJ		 13371 (get_initial_def_for_induction): Handle SSA_NAME IV step.
13372
c13bc3d9
MJ		 133732013-06-25 Martin Jambor <mjambor@suse.cz>
13374
13375 PR middle-end/57670
13376 * cgraph.h (cgraph_indirect_call_info): New flag member_ptr.
13377 * ipa-prop.c (ipa_print_node_jump_functions): Mark member pointer
13378 calls in the dump.
13379 (ipa_note_param_call): Initialize member_ptr flag.
13380 (ipa_analyze_indirect_call_uses): Set member_ptr flag.
13381 (ipa_make_edge_direct_to_target): Bail out if member_ptr is set.
13382 (ipa_write_indirect_edge_info): Stream member_ptr flag.
13383 (ipa_read_indirect_edge_info): Likewise.
13384
8a845901
RB		 133852013-06-25 Richard Biener <rguenther@suse.de>
13386
13387 PR middle-end/56977
13388 * passes.c (init_optimization_passes): Move pass_fold_builtins
13389 and pass_dce earlier with -Og.
13390
33be0bec
EB		 133912013-06-25 Eric Botcazou <ebotcazou@adacore.com>
13392
13393 * expr.c (expand_expr_real_1) <ARRAY_REF>: Fix formatting glitches.
13394 <BIT_FIELD_REF>: Remove trailing TAB.
13395 * varasm.c (output_constructor_bitfield): Fix formatting glitch and
13396 remove blank line.
13397
5fe8e757
MJ		 133982013-06-24 Martin Jambor <mjambor@suse.cz>
13399
13400 PR tree-optimization/57358
13401 * ipa-prop.c (ipa_func_spec_opts_forbid_analysis_p): New function.
13402 (ipa_compute_jump_functions_for_edge): Bail out if it returns true.
13403 (ipa_analyze_params_uses): Generate pessimistic info when true.
13404
44a60244
MJ		 134052013-06-24 Martin Jambor <mjambor@suse.cz>
13406
13407 PR tree-optimization/57539
13408 * cgraphclones.c (cgraph_clone_node): Add parameter new_inlined_to, set
13409 global.inlined_to of the new node to it. All callers changed.
13410 * ipa-inline-transform.c (clone_inlined_nodes): New variable
13411 inlining_into, pass it to cgraph_clone_node.
13412 * ipa-prop.c (ipa_propagate_indirect_call_infos): Do not call
13413 ipa_free_edge_args_substructures.
13414 (ipa_edge_duplication_hook): Only add edges from inlined nodes to
13415 rdesc linked list. Do not assert rdesc edges have inlined caller.
13416 Assert we have found an rdesc in the rdesc list.
13417
7c5848b8
RB		 134182013-06-24 Richard Biener <rguenther@suse.de>
13419
0ba6aff3 13420 * pointer-set.h (struct pointer_set_t): Move here from pointer-set.c.
7c5848b8
RB		 13421 (pointer_set_lookup): Declare.
13422 (class pointer_map): New template class implementing a
13423 generic pointer to T map.
13424 (pointer_map<T>::pointer_map, pointer_map<T>::~pointer_map,
13425 pointer_map<T>::contains, pointer_map<T>::insert,
13426 pointer_map<T>::traverse): New functions.
13427 * pointer-set.c (struct pointer_set_t): Moved to pointer-set.h.
13428 (pointer_set_lookup): New function.
13429 (pointer_set_contains): Use pointer_set_lookup.
13430 (pointer_set_insert): Likewise.
13431 (insert_aux): Remove.
13432 (struct pointer_map_t): Embed a pointer_set_t.
13433 (pointer_map_create): Adjust.
13434 (pointer_map_destroy): Likewise.
13435 (pointer_map_contains): Likewise.
13436 (pointer_map_insert): Likewise.
13437 (pointer_map_traverse): Likewise.
13438 * tree-streamer.h (struct streamer_tree_cache_d): Use a
13439 pointer_map<unsigned> instead of a pointer_map_t.
13440 * tree-streamer.c (streamer_tree_cache_insert_1): Adjust.
13441 (streamer_tree_cache_lookup): Likewise.
13442 (streamer_tree_cache_create): Likewise.
13443 (streamer_tree_cache_delete): Likewise.
13444 * lto-streamer.h (struct lto_tree_ref_encoder): Use a
13445 pointer_map<unsigned> instead of a pointer_map_t.
13446 (lto_init_tree_ref_encoder): Adjust.
13447 (lto_destroy_tree_ref_encoder): Likewise.
13448 * lto-section-out.c (lto_output_decl_index): Likewise.
13449 (lto_record_function_out_decl_state): Likewise.
13450 * dominance.c (iterate_fix_dominators): Use pointer_map<int>.
13451
3dbc97a9
RB		 134522013-06-24 Richard Biener <rguenther@suse.de>
13453
13454 PR tree-optimization/57488
13455 * tree-ssa-pre.c (insert): Clear NEW sets before each iteration.
13456
8345965a
AM		 134572013-06-24 Alan Modra <amodra@gmail.com>
13458
13459 * config/rs6000/rs6000.c (vspltis_constant): Correct for little-endian.
13460 (gen_easy_altivec_constant): Likewise.
13461 * config/rs6000/predicates.md (easy_vector_constant_add_self,
13462 easy_vector_constant_msb): Likewise.
13463
b7df379f
JJ		 134642013-06-23 Jakub Jelinek <jakub@redhat.com>
13465
13466 PR target/57688
13467 * common/config/i386/i386-common.c (ix86_handle_option): For OPT_mlzcnt
13468 add missing return true.
13469
12211b99 134702013-06-23 Oleg Endo <olegendo@gcc.gnu.org>
0bcf9a09
OE		 13471
13472 PR target/52483
13473 * config/sh/predicates.md (general_extend_operand): Invoke
13474 general_movsrc_operand for memory operands.
13475 (general_movsrc_operand): Allow reg+reg addressing, do not use
13476 general_operand for memory operands.
13477
97db2bf7
ST		 134782013-06-23 Sriraman Tallam <tmsriram@google.com>
13479
13480 * config/i386/i386.c (ix86_pragma_target_parse): Restore target
13481 when current target options does not apply.
13482 * config/i386/i386-protos.h (ix86_reset_previous_fndecl): New function.
13483 * config/i386/i386.c (ix86_reset_previous_fndecl): Ditto.
0ba6aff3
UB		 13484 * config/i386/bmiintrin.h: Pass appropriate target
13485 attributes to header.
97db2bf7
ST		 13486 * config/i386/mmintrin.h: Ditto.
13487 * config/i386/nmmintrin.h: Ditto.
13488 * config/i386/avx2intrin.h: Ditto.
13489 * config/i386/fxsrintrin.h: Ditto.
13490 * config/i386/tbmintrin.h: Ditto.
13491 * config/i386/xsaveintrin.h: Ditto.
13492 * config/i386/f16cintrin.h: Ditto.
13493 * config/i386/xtestintrin.h: Ditto.
13494 * config/i386/xsaveoptintrin.h: Ditto.
13495 * config/i386/bmi2intrin.h: Ditto.
13496 * config/i386/lzcntintrin.h: Ditto.
13497 * config/i386/smmintrin.h: Ditto.
13498 * config/i386/wmmintrin.h: Ditto.
13499 * config/i386/x86intrin.h: Remove all header include guards.
13500 * config/i386/prfchwintrin.h: Ditto.
13501 * config/i386/pmmintrin.h: Ditto.
13502 * config/i386/tmmintrin.h: Ditto.
13503 * config/i386/xmmintrin.h: Ditto.
13504 * config/i386/popcntintrin.h: Ditto.
13505 * config/i386/rdseedintrin.h: Ditto.
13506 * config/i386/ammintrin.h: Ditto.
13507 * config/i386/emmintrin.h: Ditto.
13508 * config/i386/immintrin.h: Remove all header include guards.
13509 * config/i386/fma4intrin.h: Ditto.
13510 * config/i386/lwpintrin.h: Ditto.
13511 * config/i386/xopintrin.h: Ditto.
13512 * config/i386/ia32intrin.h: Ditto.
13513 * config/i386/avxintrin.h: Ditto.
13514 * config/i386/rtmintrin.h: Ditto.
13515 * config/i386/fmaintrin.h: Ditto.
13516 * config/i386/mm3dnow.h: Ditto.
13517
495e6879
ST		 135182013-06-22 Sriraman Tallam <tmsriram@google.com>
13519
13520 * common/config/i386/i386-common.c: Handle LZCNT.
13521
11e8f687
AK		 135222013-06-22 Andi Kleen <ak@linux.intel.com>
13523
13524 * doc/extend.texi: Use __atomic_store_n instead of
13525 __atomic_store in HLE example.
13526
12211b99 135272013-06-22 Oleg Endo <olegendo@gcc.gnu.org>
664dd721
OE		 13528
13529 * config/sh/sh.c: Remove <cstdlib> workaround.
13530
927f908b
AK		 135312013-06-21 Andi Kleen <ak@linux.intel.com>
13532
0ba6aff3 13533 * doc/extend.texi: Dont use __atomic_clear in HLE example. Fix typo.
927f908b 13534
cca410b6
AK		 135352013-06-21 Andi Kleen <ak@linux.intel.com>
13536
13537 * doc/extend.texi: Document that __atomic_clear and
13538 __atomic_test_and_set should only be used with bool.
13539
d84db1ed
JH		 135402013-06-20 Jan Hubicka <jh@suse.cz>
13541
13542 * gimple-fold.c (gimple_extract_devirt_binfo_from_cst): Use
13543 types_same_for_odr.
13544 * tree.c (decls_same_for_odr): New function.
13545 (same_for_edr): New function.
13546 (types_same_for_odr): New function.
13547 (get_binfo_at_offset): Use it.
13548 * tree.h (types_same_for_odr): Declare.
13549
12211b99 135502013-06-20 Oleg Endo <olegendo@gcc.gnu.org>
469a31a9
OE		 13551 Jason Merrill <jason@redhat.com>
13552
13553 * system.h: Include <cstdlib> as well as <stdlib.h>.
13554
49b4ff63
UB		 135552013-06-20 Uros Bizjak <ubizjak@gmail.com>
13556
13557 PR target/57655
13558 * config/i386/i386.c (construct_container): Report error if
13559 long double is used with disabled x87 float returns.
13560
7a89ad00
JH		 135612013-06-20 Jan Hubicka <jh@suse.cz>
13562
13563 * lto-cgraph.c (input_symtab): Do not set cgraph state.
13564
12211b99 135652013-06-20 Joern Rennecke <joern.rennecke@embecosm.com>
bd280792
JR		 13566
13567 PR rtl-optimization/57425
13568 PR rtl-optimization/57569
13569 * alias.c (write_dependence_p): Remove parameters mem_mode and
13570 canon_mem_addr. Add parameters x_mode, x_addr and x_canonicalized.
13571 Changed all callers.
13572 (canon_anti_dependence): Get comments and semantics in sync.
13573 Add parameter mem_canonicalized. Changed all callers.
13574 * rtl.h (canon_anti_dependence): Update prototype.
13575
d16e9a99
RB		 135762013-06-20 Richard Biener <rguenther@suse.de>
13577
13578 * data-streamer-in.c (streamer_read_uhwi): Optimize single
13579 byte case, inline streamer_read_uchar and defer section
13580 overrun check.
13581
c3a9b91b
RB		 135822013-06-20 Richard Biener <rguenther@suse.de>
13583
13584 PR tree-optimization/57584
13585 * tree-ssa-loop-niter.c (expand_simple_operations): Avoid including
13586 SSA names into the expanded expression that take part in
13587 abnormal coalescing.
13588
371e77e3 135892013-06-19 Sharad Singhai <singhai@google.com>
fc8a650e
SS		 13590
13591 * gcov.c (print_usage): Handle new option.
13592 (process_args): Ditto.
13593 (get_gcov_intermediate_filename): New function.
13594 (output_intermediate_file): New function.
13595 (output_gcov_file): New function
13596 (generate_results): Handle new option.
13597 (release_function): Relase demangled name.
13598 (read_graph_file): Handle demangled name.
13599 (output_lines): Ditto.
13600 * doc/gcov.texi: Document gcov intermediate format.
13601
11f5b71a
VM		 136022013-06-19 Vladimir Makarov <vmakarov@redhat.com>
13603
13604 PR bootstrap/57604
13605 * lra.c (emit_add3_insn, emit_add2_insn): New functions.
13606 (lra_emit_add): Use the functions. Add comment about Y as an
13607 address segment.
13608
e1f9340b
DE		 136092013-06-19 David Edelsohn <dje.gcc@gmail.com>
13610
13611 PR driver/57652
13612 * collect2.c (collect_atexit): New.
13613 (collect_exit): Delete.
13614 (main): Register collect_atexit with atexit.
13615 (collect_wait): Change collect_exit to exit.
13616 (do_wait): Same.
13617 * collect2.h (collect_exit): Delete.
13618 * tlink.c (do_tlink): Rename exit to ret. Change collect_exit to exit.
13619
e3f9e0ac
WM		 136202013-06-19 Wei Mi <wmi@google.com>
13621
13622 PR rtl-optimization/57518
13623 * ira.c (set_paradoxical_subreg): Set pdx_subregs[regno]
13624 if regno is used in paradoxical subreg.
13625 (update_equiv_regs): Check pdx_subregs[regno] before
13626 set a reg to be equivalent with a mem.
13627
a09f1a76
MK		 136282013-06-19 Matthias Klose <doko@ubuntu.com>
13629
13630 PR driver/57651
13631 * file-find.h (find_a_file): Add a mode parameter.
13632 * file-find.c (find_a_file): Likewise.
13633 * gcc-ar.c (main): Call find_a_file with R_OK for the plugin,
13634 with X_OK for the executables.
13635 * collect2.c (main): Call find_a_file with X_OK.
13636
7d18b0ad
SE		 136372013-06-19 Steve Ellcey <sellcey@mips.com>
13638
13639 PR target/56942
13640 * config/mips/mips.md (casesi_internal_mips16_<mode>):
13641 Use NEXT_INSN instead of next_real_insn.
13642
6a6dac52
JH		 136432013-06-19 Jan Hubicka <jh@suse.cz>
13644
13645 * cgraph.h (const_value_known_p): Replace by ...
13646 (ctor_for_folding): .. this one.
13647 * cgraphunit.c (process_function_and_variable_attributes): Use it.
13648 * lto-cgraph.c (compute_ltrans_boundary): Use ctor_for_folding.
13649 * expr.c (expand_expr_real_1): Likewise.
13650 (string_constant): Likewise.
13651 * tree-ssa-loop-ivcanon.c (constant_after_peeling): Likewise.
13652 * ipa.c (process_references): Likewise.
13653 (symtab_remove_unreachable_nodes): Likewise.
13654 * ipa-inline-analysis.c (param_change_prob): Likewise.
13655 * gimple-fold.c (canonicalize_constructor_val): Likewise.
13656 (get_base_constructor): Likwise.
13657 * varpool.c (varpool_remove_node): Likewise.
13658 (varpool_remove_initializer): LIkewise.
13659 (dump_varpool_node): LIkwise.
13660 (const_value_known_p): Rewrite to ...
13661 (ctor_for_folding): ... this one.
13662
216c12ab
JJ		 136632013-06-19 Jakub Jelinek <jakub@redhat.com>
13664
13665 PR driver/57651
13666 * gcc-ar.c (main): If not CROSS_DIRECTORY_STRUCTURE, look for
13667 PERSONALITY in $PATH derived prefixes.
13668
d53e2f99
JL		 136692013-06-19 Jeff Law <law@redhat.com>
13670
ecdbf306
JL		 13671 * tree-ssa-forwprop.c (simplify_bitwise_binary_boolean): Fix typo
13672 in comment.
13673
d53e2f99
JL		 13674 * tree-ssa-forwprop.c (simplify_bitwise_binary_boolean): New function.
13675 (simplify_bitwise_binary): Use it to simpify certain binary ops on
13676 booleans.
13677
89e7fc04
SN		 136782013-06-19 Sofiane Naci <sofiane.naci@arm.com>
13679
13680 * config/arm/vfp.md: Move VFP instruction classification documentation
13681 to ...
13682 * config/arm/arm.md: ... here. Update instruction classification
13683 documentation.
13684
5e7f6aaa
RE		 136852013-06-19 Richard Earnshaw <rearnsha@arm.com>
13686
13687 arm.md (split for eq(reg, 0)): Add variants for ARMv5 and Thumb2.
13688 (peepholes for eq(reg, not-0)): Ensure condition register is dead after
13689 pattern. Use more efficient sequences on ARMv5 and Thumb2.
13690
5fdc1e5d
SB		 136912013-06-19 Steven Bosscher <steven@gcc.gnu.org>
13692
13693 PR target/57609
13694 * config/s390/s390.c (s390_chunkify_start): Replace next_real_insn
13695 with NEXT_INSN. Use tablejump_p to check for jump table data
13696 insns.
13697
e952afba
PC		 136982013-06-19 Paolo Carlini <paolo.carlini@oracle.com>
13699
13700 PR c++/56544
13701 * doc/cpp.texi [Standard Predefined Macros, __cplusplus]: Document
13702 that now in C++ the value is correct per the C++ standards.
13703
3fd0303a
RB		 137042013-06-19 Richard Biener <rguenther@suse.de>
13705
13706 * expr.c (expand_expr_real_1): Use SCOPE_FILE_SCOPE_P to check
13707 for global context.
13708
b04e4b08
AK		 137092013-06-19 Andreas Krebbel <Andreas.Krebbel@de.ibm.com>
13710
13711 Revert:
13712 2013-06-18 Andreas Krebbel <Andreas.Krebbel@de.ibm.com>
13713
13714 PR target/57609
13715 * config/s390/s390.c (s390_chunkify_start): Replace next_real_insn
13716 with next_active_insn.
13717
bef8491a
ST		 137182013-06-18 Sriraman Tallam <tmsriram@google.com>
13719
13720 * ipa-inline.c (inline_always_inline_functions): Pretend always_inline
13721 functions are inlined during failures to flag an error.
13722 * tree-inline.c (expand_call_inline): Allow the error to be flagged
13723 in early inline pass.
22fc7d3d 13724
c21b257d
L		 137252013-06-18 H.J. Lu <hongjiu.lu@intel.com>
13726
13727 * config/i386/i386.c (initial_ix86_tune_features): Fix a typo
13728 in comments.
13729
33255ae3
JB		 137302013-06-18 Julian Brown <julian@codesourcery.com>
13731
13732 * config/arm/arm.c (neon_vector_mem_operand): Add strict argument.
13733 Permit virtual register pre-reload if !strict.
13734 (coproc_secondary_reload_class): Adjust for neon_vector_mem_operand
13735 change.
13736 * config/arm/arm-protos.h (neon_vector_mem_operand): Adjust
13737 prototype.
13738 * config/arm/neon.md (movmisalign<mode>): Use
13739 neon_perm_struct_or_reg_operand instead of
13740 neon_struct_or_register_operand.
13741 (*movmisalign<mode>_neon_load, *movmisalign<mode>_neon_store): Use
13742 neon_permissive_struct_operand instead of neon_struct_operand.
13743 * config/arm/constraints.md (Un, Um, Us): Adjust calls to
13744 neon_vector_mem_operand.
13745 * config/arm/predicates.md (neon_struct_operand): Adjust call to
13746 neon_vector_mem_operand.
13747 (neon_permissive_struct_operand): New.
13748 (neon_struct_or_register_operand): Rename to...
13749 (neon_perm_struct_or_reg_operand): This. Adjust call to
13750 neon_vector_mem_operand.
13751
d579fcda
RB		 137522013-06-18 Richard Biener <rguenther@suse.de>
13753
13754 * Makefile.in (LTO_STREAMER_H): Add pointer-set.h dependency.
13755 * lto-streamer.h: Include pointer-set.h.
13756 (struct lto_decl_slot): Remove.
13757 (struct lto_tree_ref_encoder): Make tree_hash_table a pointer-map.
13758 Remove next_index entry.
13759 (lto_hash_decl_slot_node, lto_eq_decl_slot_node,
13760 lto_hash_type_slot_node, lto_eq_type_slot_node): Remove.
13761 (lto_init_tree_ref_encoder): Adjust.
13762 (lto_destroy_tree_ref_encoder): Likewise.
13763 * lto-section-out.c (lto_hash_decl_slot_node, lto_eq_decl_slot_node,
13764 lto_hash_type_slot_node, lto_eq_type_slot_node): Remove.
13765 (lto_output_decl_index): Adjust.
13766 (lto_new_out_decl_state): Likewise.
13767 (lto_record_function_out_decl_state): Likewise.
13768 * lto-streamer-out.c (copy_function): Likewise.
13769
31ee20ba
RB		 137702013-06-18 Richard Biener <rguenther@suse.de>
13771
13772 * Makefile.in (cgraphunit.o): Add $(CFGLOOP_H) dependency.
13773 * cgraphunit.c: Include cfgloop.h.
13774 (init_lowered_empty_function): Initialize the loop tree.
13775 (assemble_thunk): Insert new BBs into loops.
13776
ac0511f2
RB		 137772013-06-18 Richard Biener <rguenther@suse.de>
13778
13779 * tree-streamer.h (streamer_tree_cache_create): Adjust prototype.
13780 * tree-streamer.c (streamer_tree_cache_create): Make maintaining
13781 the map from cache entry to cache index optional.
13782 (streamer_tree_cache_replace_tree): Adjust accordingly.
13783 (streamer_tree_cache_append): Likewise.
13784 (streamer_tree_cache_delete): Likewise.
13785 * lto-streamer-in.c (lto_data_in_create): Do not maintain the
13786 streamer cache map from cache entry to cache index.
13787 * lto-streamer-out.c (create_output_block): Adjust.
13788
09485a08
SN		 137892013-06-18 Sofiane Naci <sofiane.naci@arm.com>
13790
22fc7d3d
UB		 13791 * config/arm/arm.md (attribute "insn"): Move multiplication and
13792 division attributes to...
09485a08
SN		 13793 (attribute "type"): ... here. Remove mult.
13794 (attribute "mul32"): New attribute.
13795 (attribute "mul64"): Add umaal.
13796 (*arm_mulsi3): Update attributes.
13797 (*arm_mulsi3_v6): Likewise.
13798 (*thumb_mulsi3): Likewise.
13799 (*thumb_mulsi3_v6): Likewise.
13800 (*mulsi3_compare0): Likewise.
13801 (*mulsi3_compare0_v6): Likewise.
13802 (*mulsi_compare0_scratch): Likewise.
13803 (*mulsi_compare0_scratch_v6): Likewise.
13804 (*mulsi3addsi): Likewise.
13805 (*mulsi3addsi_v6): Likewise.
13806 (*mulsi3addsi_compare0): Likewise.
13807 (*mulsi3addsi_compare0_v6): Likewise.
13808 (*mulsi3addsi_compare0_scratch): Likewise.
13809 (*mulsi3addsi_compare0_scratch_v6): Likewise.
13810 (*mulsi3subsi): Likewise.
13811 (*mulsidi3adddi): Likewise.
13812 (*mulsi3addsi_v6): Likewise.
13813 (*mulsidi3adddi_v6): Likewise.
13814 (*mulsidi3_nov6): Likewise.
13815 (*mulsidi3_v6): Likewise.
13816 (*umulsidi3_nov6): Likewise.
13817 (*umulsidi3_v6): Likewise.
13818 (*umulsidi3adddi): Likewise.
13819 (*umulsidi3adddi_v6): Likewise.
13820 (*smulsi3_highpart_nov6): Likewise.
13821 (*smulsi3_highpart_v6): Likewise.
13822 (*umulsi3_highpart_nov6): Likewise.
13823 (*umulsi3_highpart_v6): Likewise.
13824 (mulhisi3): Likewise.
13825 (*mulhisi3tb): Likewise.
13826 (*mulhisi3bt): Likewise.
13827 (*mulhisi3tt): Likewise.
13828 (maddhisi4): Likewise.
13829 (*maddhisi4tb): Likewise.
13830 (*maddhisi4tt): Likewise.
13831 (maddhidi4): Likewise.
13832 (*maddhidi4tb): Likewise.
13833 (*maddhidi4tt): Likewise.
13834 (divsi3): Likewise.
13835 (udivsi3): Likewise.
13836 * config/arm/thumb2.md (thumb2_mulsi_short): Update attributes.
13837 (thumb2_mulsi_short_compare0): Likewise.
13838 (thumb2_mulsi_short_compare0_scratch): Likewise.
13839 * config/arm/arm1020e.md (1020mult1): Update attribute change.
13840 (1020mult2): Likewise.
13841 (1020mult3): Likewise.
13842 (1020mult4): Likewise.
13843 (1020mult5): Likewise.
13844 (1020mult6): Likewise.
22fc7d3d
UB		 13845 * config/arm/cortex-a15.md (cortex_a15_mult32): Update attribute
13846 change.
09485a08
SN		 13847 (cortex_a15_mult64): Likewise.
13848 (cortex_a15_sdiv): Likewise.
13849 (cortex_a15_udiv): Likewise.
13850 * config/arm/arm1026ejs.md (mult1): Update attribute change.
13851 (mult2): Likewise.
13852 (mult3): Likewise.
13853 (mult4): Likewise.
13854 (mult5): Likewise.
13855 (mult6): Likewise.
13856 * config/arm/marvell-pj4.md (pj4_ir_mul): Update attribute change.
13857 (pj4_ir_div): Likewise.
13858 * config/arm/arm1136jfs.md (11_mult1): Update attribute change.
13859 (11_mult2): Likewise.
13860 (11_mult3): Likewise.
13861 (11_mult4): Likewise.
13862 (11_mult5): Likewise.
13863 (11_mult6): Likewise.
13864 (11_mult7): Likewise.
13865 * config/arm/cortex-a8.md (cortex_a8_mul): Update attribute change.
13866 (cortex_a8_mla): Likewise.
13867 (cortex_a8_mull): Likewise.
13868 (cortex_a8_smulwy): Likewise.
13869 (cortex_a8_smlald): Likewise.
13870 * config/arm/cortex-m4.md (cortex_m4_alu): Update attribute change.
13871 * config/arm/cortex-r4.md (cortex_r4_mul_4): Update attribute change.
13872 (cortex_r4_mul_3): Likewise.
13873 (cortex_r4_mla_4): Likewise.
13874 (cortex_r4_mla_3): Likewise.
13875 (cortex_r4_smlald): Likewise.
13876 (cortex_r4_mull): Likewise.
13877 (cortex_r4_sdiv): Likewise.
13878 (cortex_r4_udiv): Likewise.
13879 * config/arm/cortex-a7.md (cortex_a7_mul): Update attribute change.
13880 (cortex_a7_idiv): Likewise.
13881 * config/arm/arm926ejs.md (9_mult1): Update attribute change.
13882 (9_mult2): Likewise.
13883 (9_mult3): Likewise.
13884 (9_mult4): Likewise.
13885 (9_mult5): Likewise.
13886 (9_mult6): Likewise.
13887 * config/arm/cortex-a53.md (cortex_a53_mul): Update attribute change.
13888 (cortex_a53_sdiv): Likewise.
13889 (cortex_a53_udiv): Likewise.
13890 * config/arm/fa726te.md (726te_mult_op): Update attribute change.
13891 * config/arm/fmp626.md (mp626_mult1): Update attribute change.
13892 (mp626_mult2): Likewise.
13893 (mp626_mult3): Likewise.
13894 (mp626_mult4): Likewise.
13895 * config/arm/fa526.md (526_mult1): Update attribute change.
13896 (526_mult2): Likewise.
13897 * config/arm/arm-generic.md (mult): Update attribute change.
13898 (mult_ldsched_strongarm): Likewise.
13899 (mult_ldsched): Likewise.
13900 (multi_cycle): Likewise.
13901 * config/arm/cortex-a5.md (cortex_a5_mul): Update attribute change.
13902 * config/arm/fa606te.md (606te_mult1): Update attribute change.
13903 (606te_mult2): Likewise.
13904 (606te_mult3): Likewise.
13905 (606te_mult4): Likewise.
13906 * config/arm/cortex-a9.md (cortex_a9_mult16): Update attribute change.
13907 (cortex_a9_mac16): Likewise.
13908 (cortex_a9_multiply): Likewise.
13909 (cortex_a9_mac): Likewise.
13910 (cortex_a9_multiply_long): Likewise.
13911 * config/arm/fa626te.md (626te_mult1): Update attribute change.
13912 (626te_mult2): Likewise.
13913 (626te_mult3): Likewise.
13914 (626te_mult4): Likewise.
13915
8951f345
RB		 139162013-06-18 Richard Biener <rguenther@suse.de>
13917
13918 PR lto/57334
13919 * lto-symtab.c (lto_symtab_merge_decls): Process nodes properly.
13920
d8101d05
AK		 139212013-06-18 Andreas Krebbel <Andreas.Krebbel@de.ibm.com>
13922
13923 PR target/57609
13924 * config/s390/s390.c (s390_chunkify_start): Replace next_real_insn
13925 with next_active_insn.
13926
69eff9da
AM		 139272013-06-18 Alan Modra <amodra@gmail.com>
13928
13929 * config/rs6000/rs6000.h (enum data_align): New.
13930 (LOCAL_ALIGNMENT, DATA_ALIGNMENT): Use rs6000_data_alignment.
13931 (DATA_ABI_ALIGNMENT): Define.
13932 (CONSTANT_ALIGNMENT): Correct comment.
13933 * config/rs6000/rs6000-protos.h (rs6000_data_alignment): Declare.
13934 * config/rs6000/rs6000.c (rs6000_data_alignment): New function.
13935
8ed00d76
DM		 139362013-06-17 David Malcolm <dmalcolm@redhat.com>
13937
13938 * ggc-page.c (ggc_pch_write_object) <d>: Remove erroneous
13939 ATTRIBUTE_UNUSED marking.
13940
3d33d06b
SN		 139412013-06-17 Sofiane Naci <sofiane.naci@arm.com>
13942
13943 * config/aarch64/aarch64-simd.md (aarch64_dup_lane<mode>): Add r<-w
13944 alternative and update.
13945 (aarch64_dup_lanedi): Delete.
13946 * config/aarch64/arm_neon.h (vdup<bhsd>_lane_*): Update.
13947 * config/aarch64/aarch64-simd-builtins.def: Update.
13948
ee03e71d
RB		 139492013-06-17 Richard Biener <rguenther@suse.de>
13950
13951 * lto-streamer.h (enum LTO_tags): Add LTO_tree_scc.
13952 (lto_input_scc): Declare.
13953 (lto_input_tree_1): Likewise.
13954 (struct lto_stats_d): Add num_tree_bodies_output and
13955 num_pickle_refs_output.
13956 * lto-streamer-in.c (lto_read_body): Use streamer_tree_cache_get_tree.
13957 (lto_read_tree_1): Split out from ...
13958 (lto_read_tree): ... this.
13959 (lto_input_scc): New function.
13960 (lto_input_tree_1): Split out from ...
13961 (lto_input_tree): ... this. Handle LTO_tree_scc.
13962 (lto_data_in_create): Create the streamer cache without hashes.
13963 * lto-streamer-out.c (create_output_block): Create the streamer
13964 cache with hashes when not doing WPA.
13965 (lto_write_tree_1): Split out from ...
13966 (lto_write_tree): ... this.
13967 (get_symbol_initial_value): New function.
13968 (lto_output_tree_1): Split out from ...
13969 (lto_output_tree): ... this. Write trees as series of SCCs
13970 using a DFS walk via DFS_write_tree.
13971 (struct sccs, struct scc_entry): New types.
13972 (next_dfs_num, sccstack, sccstate, sccstate_obstack): New globals.
13973 (DFS_write_tree_body): New function.
13974 (DFS_write_tree): Likewise.
13975 (hash_tree): Likewise.
13976 (scc_entry_compare): Likewise.
13977 (hash_scc): Likewise.
13978 (tree_is_indexable): DEBUG_EXPR_DECLs are local entities.
13979 * tree-streamer-in.c (lto_input_ts_list_tree_pointers): Stream
13980 TREE_CHAIN as regular reference.
13981 (streamer_read_integer_cst): Remove.
13982 (streamer_get_pickled_tree): Adjust.
13983 * tree-streamer-out.c (streamer_write_chain): Disable streaming
13984 of DECL_EXTERNALs in BLOCK_VARS for now.
13985 (write_ts_list_tree_pointers): Stream TREE_CHAIN as regular
13986 reference.
13987 * tree-streamer.c (streamer_tree_cache_add_to_node_array):
13988 Add hash value argument and record that if hashes are recorded
13989 in the cache.
13990 (streamer_tree_cache_insert_1): Adjust.
13991 (streamer_tree_cache_insert): Likewise.
13992 (streamer_tree_cache_insert_at): Rename to ...
13993 (streamer_tree_cache_replace_tree): ... this and adjust.
13994 (streamer_tree_cache_append): Adjust.
13995 (record_common_node): Likewise.
13996 (streamer_tree_cache_create): Add argument whether to
13997 record hash values together with trees.
13998 (streamer_tree_cache_delete): Adjust.
13999 * tree-streamer.h (struct streamer_tree_cache_d): Add
14000 vector of hashes.
14001 (streamer_read_integer_cst): Remove.
14002 (streamer_tree_cache_insert): Adjust.
14003 (streamer_tree_cache_append): Likewise.
14004 (streamer_tree_cache_insert_at): Rename to ...
14005 (streamer_tree_cache_replace_tree): ... this and adjust.
14006 (streamer_tree_cache_create): Add argument whether to record hashes.
14007 (streamer_tree_cache_get): Rename to ...
14008 (streamer_tree_cache_get_tree): ... this.
14009 (streamer_tree_cache_get_hash): New function.
14010 * tree.c (cache_integer_cst): New function.
14011 * tree.h (cache_integer_cst): Declare.
14012 (ANON_AGGRNAME_FORMAT, ANON_AGGRNAME_P): Move here from cp/cp-tree.h.
14013 * lto-symtab.c (lto_varpool_replace_node): Only release
14014 DECL_INITIAL of non-prevailing decls.
14015 * varpool.c (varpool_remove_initializer): Do not release
14016 DECL_INITIAL when we are still in CGRAPH_LTO_STREAMING.
14017
0de86a92
JU		 140182013-06-16 Jürgen Urban <JuergenUrban@gmx.de>
14019
14020 * config/mips/mips.h (ISA_HAS_MUL3): Include TARGET_MIPS5900.
14021 (ISA_HAS_MULT, ISA_HAS_DMULT, ISA_HAS_DIV, ISA_HAS_DDIV): New macros.
14022 * config/mips/mips.md (mul<mode>3, mul<mode>3_internal)
14023 (mul<mode>3_r4000): Require ISA_HAS_<D>MULT.
14024 (mul<mode>3_mul3): Handle TARGET_MIPS5900.
14025 (mulsidi3_64bit_dmul): Remove redundant TARGET_64BIT test.
14026 (<su>muldi3_highpart, <su>muldi3_highpart_internal, <u>mulditi3)
14027 (<u>mulditi3_internal, <u>mulditi3_r4000): Require ISA_HAS_DMULT
14028 instead of TARGET_64BIT.
14029 (divmod<mode>4, udivmod<mode>4, <u>divmod<GPR:mode>4_hilo_<HILO:mode>):
14030 Require ISA_HAS_<D>DIV.
14031
74240413
RS		 140322013-06-16 Richard Sandiford <rdsandiford@googlemail.com>
14033
14034 * config.gcc (mips*-mti-linux*, mips64*-*-linux*, mipsisa64*-*-linux*)
14035 (mips*-*-linux*): Move default with_llsc setting to where other
14036 defaults are set.
14037 (mips*-*-vxworks*): Move with_arch default from with_cpu block to
14038 with_arch block.
14039 (mips64r5900-*-*, mips64r5900el-*-*, mipsr5900-*-*, mipsr5900el-*-*):
14040 Likewise. Remove default with_tune setting. Move default float
14041 setting to its own block. Handle with_llsc in the same block as above.
14042
12211b99 140432013-06-16 Joern Rennecke <joern.rennecke@embecosm.com>
393f9fed
JR		 14044
14045 PR rtl-optimization/57425
14046 PR rtl-optimization/57569
299a5f6a 14047 * alias.c (write_dependence_p): Add new parameters mem_mode,
393f9fed
JR		 14048 canon_mem_addr and mem_canonicalized. Change type of writep to bool.
14049 Changed all callers.
14050 (canon_anti_dependence): New function.
14051 * cse.c (check_dependence): Use canon_anti_dependence.
14052 * cselib.c (cselib_invalidate_mem): Likewise.
14053 * rtl.h (canon_anti_dependence): Declare.
14054
b259d352
JU		 140552013-06-16 Jürgen Urban <JuergenUrban@gmx.de>
14056
14057 * config/mips/mips.h (ISA_HAS_LL_SC): Exclude TARGET_MIPS5900.
14058 * config/mips/mips.c (mips_start_ll_sc_sync_block): Output
14059 ".set mips3" for 64-bit targets.
14060
4de09b85
DC		 140612013-06-15 Dehao Chen <dehao@google.com>
14062
14063 * tree-flow.h (gimple_check_call_matching_types): Add new argument.
14064 * gimple-low.c (gimple_check_call_matching_types): Likewise.
14065 (gimple_check_call_args): Likewise.
14066 * value-prof.c (check_ic_target): Likewise.
14067 * ipa-inline.c (early_inliner): Likewise.
14068 * ipa-prop.c (update_indirect_edges_after_inlining): Likewise.
14069 * cgraph.c (cgraph_create_edge_1): Likewise.
14070 (cgraph_make_edge_direct): Likewise.
14071
d1aa4a76
MM		 140722013-06-14 Michael Meissner <meissner@linux.vnet.ibm.com>
14073
14074 PR target/57615
14075 * config/rs6000/rs6000.md (mov<mode>_ppc64): Call
14076 rs6000_output_move_128bit to handle emitting quad memory
14077 operations. Set attribute length to 8 bytes.
14078
12211b99 140792013-06-14 Vidya Praveen <vidyapraveen@arm.com>
24244e4d
VP		 14080
14081 * config/aarch64/aarch64-simd.md (aarch64_<su>mlal_lo<mode>):
14082 New pattern.
14083 (aarch64_<su>mlal_hi<mode>, aarch64_<su>mlsl_lo<mode>): Likewise.
14084 (aarch64_<su>mlsl_hi<mode>, aarch64_<su>mlal<mode>): Likewise.
14085 (aarch64_<su>mlsl<mode>): Likewise.
14086
f5273574
MS		 140872013-06-14 Mike Stump <mikestump@comcast.net>
14088
14089 * Makefile.in (TARGET_H): Add insn-codes.h.
14090
ef7e7ab2
AM		 140912013-06-14 Alan Modra <amodra@gmail.com>
14092
14093 PR middle-end/57134
14094 PR middle-end/57586
14095 * expr.c (expand_expr_real_1 <normal_inner_ref>): Pass
14096 EXPAND_MEMORY and EXPAND_WRITE to recursive call. Don't use
14097 bitfield expansion when EXPAND_MEMORY.
14098 (expand_expr_real_1 <VIEW_CONVERT_EXPR>): Pass modifier likewise.
14099
135faab6
MM		 141002013-06-13 Michael Meissner <meissner@linux.vnet.ibm.com>
14101
14102 * config/rs6000/rs6000.c (rs6000_option_override_internal): Move
14103 test for clearing quad memory on 32-bit later.
14104
948a1fd9
MG		 141052013-06-13 Marc Glisse <marc.glisse@inria.fr>
14106
14107 * fold-const.c (negate_expr_p): Handle VECTOR_CST.
14108 (fold_negate_expr): Likewise.
14109 (fold_real_zero_addition_p): Handle vectors.
14110 (fold_binary_loc) <PLUS_EXPR, MINUS_EXPR>: Likewise.
14111
0597fb9c
AM		 141122013-06-14 Alan Modra <amodra@gmail.com>
14113
14114 * varasm.c (force_const_mem): Revert 2013-06-07 change.
14115
e5b962d0
JH		 141162013-06-13 Jan Hubicka <jh@suse.cz>
14117
14118 * ipa.c (cgraph_externally_visible_p, varpool_externally_visible_p):
14119 Local comdats are not externally visible.
14120 * symtab.c (dump_symtab_base): Dump externally visible.
14121 (verify_symtab_base): Verify back links in the symtab hash.
14122
a8355e51
BC		 141232013-06-13 Bin Cheng <bin.cheng@arm.com>
14124
14125 * fold-const.c (operand_equal_p): Consider NOP_EXPR and
14126 CONVERT_EXPR as equal nodes.
14127
8f7e6e33
BC		 141282013-06-13 Bin Cheng <bin.cheng@arm.com>
14129
22fc7d3d 14130 * rtlanal.c (noop_move_p): Check the code to be executed for COND_EXEC.
8f7e6e33 14131
5d418483
MG		 141322013-06-13 Marc Glisse <marc.glisse@inria.fr>
14133
14134 * tree-ssa-forwprop.c (simplify_bitwise_binary, associate_plusminus):
14135 Generalize to complex and vector.
14136 * tree.c (build_all_ones_cst): New function.
14137 * tree.h (build_all_ones_cst): Declare it.
14138
50751417
AM		 141392013-06-13 Alan Modra <amodra@gmail.com>
14140
14141 * config/rs6000/rs6000.h (LONG_DOUBLE_LARGE_FIRST): Define.
14142 * config/rs6000/rs6000.md (signbittf2): New insn.
14143 (extenddftf2_internal): Use LONG_DOUBLE_LARGE_FIRST.
14144 (abstf2_internal, cmptf_internal2): Likewise.
14145 * config/rs6000/spe.md (spe_abstf2_cmp, spe_abstf2_tst): Likewise.
14146
4b02c962 141472013-06-12 Michael Meissner <meissner@linux.vnet.ibm.com>
12211b99
UB		 14148 Pat Haugen <pthaugen@us.ibm.com>
14149 Peter Bergner <bergner@vnet.ibm.com>
4b02c962
MM		 14150
14151 * config/rs6000/rs6000.c (emit_load_locked): Add support for
14152 power8 byte, half-word, and quad-word atomic instructions.
14153 (emit_store_conditional): Likewise.
14154 (rs6000_expand_atomic_compare_and_swap): Likewise.
14155 (rs6000_expand_atomic_op): Likewise.
14156
14157 * config/rs6000/sync.md (larx): Add new modes for power8.
14158 (stcx): Likewise.
14159 (AINT): New mode iterator to include TImode as well as normal
14160 integer modes on power8.
14161 (fetchop_pred): Use int_reg_operand instead of gpc_reg_operand so
14162 that VSX registers are not considered. Use AINT mode iterator
14163 instead of INT1 to allow inclusion of quad word atomic operations
14164 on power8.
14165 (load_locked<mode>): Likewise.
14166 (store_conditional<mode>): Likewise.
14167 (atomic_compare_and_swap<mode>): Likewise.
14168 (atomic_exchange<mode>): Likewise.
14169 (atomic_nand<mode>): Likewise.
14170 (atomic_fetch_<fetchop_name><mode>): Likewise.
14171 (atomic_nand_fetch<mode>): Likewise.
14172 (mem_thread_fence): Use gen_loadsync_<mode> instead of enumerating
14173 each type.
14174 (ATOMIC): On power8, add QImode, HImode modes.
14175 (load_locked<QHI:mode>_si): Varients of load_locked for QI/HI
14176 modes that promote to SImode.
14177 (load_lockedti): Convert TImode arguments to PTImode, so that we
14178 get a guaranteed even/odd register pair.
14179 (load_lockedpti): Likewise.
14180 (store_conditionalti): Likewise.
14181 (store_conditionalpti): Likewise.
14182
14183 * config/rs6000/rs6000.md (QHI): New mode iterator for power8
14184 atomic load/store instructions.
14185 (HSI): Likewise.
14186
710b6cc1
RS		 141872013-06-12 Richard Sandiford <rdsandiford@googlemail.com>
14188
14189 * config/mips/mips.md (extended_mips16): Include GOT and constant-pool
14190 loads.
14191 (insn_count): New attribute, with most cases extracted from...
14192 (length): ...here. Redefine most cases in terms of insn_count.
14193 (single_insn): Delete.
14194 (can_delay): Use insn_count to check for single instructions.
14195 (*mul<mode>3_r4300, mul<mode>3_r4000, *mul_acc_si, *mul_acc_si_r3900)
14196 (*msac_using_macc, *mul_sub_si, <u>mulsidi3_32bit_r4000)
14197 (<u>mulsidi3_64bit_r4000, <su>muldi3_highpart_internal)
14198 (<su>mulsi3_highpart_split, <su>muldi3_highpart_internal)
14199 (<u>mulditi3_r4000, *div<mode>3, *recip<mode>3, divmod<mode>4)
14200 (udivmod<mode>4, sqrt<mode>2, *rsqrt<mode>a, *rsqrt<mode>b)
14201 (fix_truncdfsi2_macro, fix_truncsfsi2_macro, *lea_high64)
14202 (*lea64, cprestore_<mode>, clear_hazard_<mode>, <unnamed insn>)
14203 (casesi_internal_mips16_<mode>, *tls_get_tp_<mode>_split)
14204 (tls_get_tp_mips16, *tls_get_tp_mips16_call_<mode>): Use "insn_count"
14205 rather than "length".
14206 (tls_get_tp_<mode>): Likewise. Remove redundant "no_delay" attribute.
14207 * config/mips/mips-ps-3d.md (mips_c_cond_4s, mips_cabs_cond_4s):
14208 Use "insn_count" rather than "length".
14209 * config/mips/mips-dsp.md
14210 (mips_l<SHORT:size><u>x_ext<GPR:mode>_<P:mode>)
14211 (mips_l<GPR:size>x_<P:mode>, *mips_lw<u>x_<P:mode>_ext): Remove
14212 length attributes.
14213
1951f101
MG		 142142013-06-12 Marc Glisse <marc.glisse@inria.fr>
14215
14216 PR tree-optimization/57361
14217 * tree-ssa-dse.c (dse_possible_dead_store_p): Handle self-assignment.
14218
8b033a8a
SN		 142192013-06-12 Sofiane Naci <sofiane.naci@arm.com>
14220
22fc7d3d
UB		 14221 * config/aarch64/aarch64-simd.md (aarch64_combine<mode>): Convert
14222 to split.
8b033a8a
SN		 14223 (aarch64_simd_combine<mode>): New instruction expansion.
14224 * config/aarch64/aarch64-protos.h (aarch64_split_simd_combine): New
14225 function prototype.
14226 * config/aarch64/aarch64.c (aarch64_split_combine): New function.
14227 * config/aarch64/iterators.md (Vdbl): Add entry for DF.
14228
ca0f62a8
JH		 142292013-06-12 Jan Hubicka <jh@suse.cz>
14230
14231 * cgraph.c (verify_edge_corresponds_to_fndecl): Be lax about
14232 decl has when in streaming stage.
14233 * lto-symtab.c (lto_symtab_merge_symbols): Likewise.
14234 * cgraph.h (cgraph_state): Add CGRAPH_LTO_STREAMING.
14235
12211b99 142362013-06-12 Roland Stigge <stigge@antcom.de>
cddd69eb
RS		 14237
14238 PR target/57578
14239 * config/rs6000/t-linux (MULTIARCH_DIRNAME): Fix SPE version detection.
14240
bfdeda2c
JJ		 142412013-06-12 Jakub Jelinek <jakub@redhat.com>
14242
14243 PR tree-optimization/57537
14244 * tree-vect-patterns.c (vect_recog_widen_mult_pattern): If
14245 vect_handle_widen_op_by_const, convert oprnd1 to half_type1.
14246
a4fa02d1
RB		 142472013-06-12 Richard Biener <rguenther@suse.de>
14248
14249 * data-streamer.h (streamer_write_char_stream): CSE
14250 obs->current_pointer.
14251 * data-streamer-out.c (streamer_write_uhwi_stream): Inline
14252 streamer_write_char_stream manually and optimize the resulting loop.
14253 (streamer_write_hwi_stream): Likewise.
14254
bbf9ad07
JH		 142552013-06-12 Jan Hubicka <jh@suse.cz>
14256
14257 * lto-symtab.c (lto_symtab_merge_symbols): Populate symtab hashtable.
14258 * cgraph.h (varpool_create_empty_node): Declare.
14259 * lto-cgraph.c (input_node, input_varpool_node): Forcingly create
14260 duplicated nodes.
14261 * symtab.c (symtab_unregister_node): Be lax about missin entries
14262 in node hash.
14263 (symtab_get_node): Update comment.
14264 * varpool.c (varpool_create_empty_node): Break out from ...
14265 (varpool_node_for_decl): ... here.
14266 * lto-streamer.h (lto_file_decl_data): Add RESOLUTION_MAP.
14267
a53f90ad
EB		 142682013-06-12 Eric Botcazou <ebotcazou@adacore.com>
14269
14270 * expr.c (expand_expr_real_1) <TARGET_MEM_REF>: Use straight-line flow.
14271 <MEM_REF>: Use 'type' instead of TREE_TYPE (exp) and tidy up the first
14272 part. Use straight-line flow at the end.
14273 <COMPONENT_REF>: Remove superfluous else.
14274 <VIEW_CONVERT_EXPR>: Use 'type' instead of TREE_TYPE (exp).
14275
946f9306
JJ		 142762013-06-12 Jakub Jelinek <jakub@redhat.com>
14277
14278 PR target/56564
14279 * varasm.c (decl_binds_to_current_def_p): Call binds_local_p
14280 target hook even for !TREE_PUBLIC decls. If no resolution info
14281 is available, return false for common and external decls.
14282
a7fca6f0
KP		 142832013-06-12 Kaushik Phatak <kaushik.phatak@kpitcummins.com>
14284
14285 * config/rl78/constraints.md (U): New constraint.
14286 * config/rl78/rl78.md (*mulqi3_rl78,*mulhi3_rl78,*mulhi3_g13): Add
14287 valloc attribute.
14288
c7d8f446
MM		 142892013-06-11 Michael Meissner <meissner@linux.vnet.ibm.com>
14290
14291 PR target/57589
14292 * config/rs6000/driver-rs6000.c (elf_platform): Make buffer static
14293 to allow returning address to AT_PLATFORM name.
14294
08346abd
JH		 142952013-06-11 Jan Hubicka <jh@suse.cz>
14296
14297 * cgraph.c (cgraph_create_function_alias): Set weakref flag.
14298 * cgraph.h (symtab_node_base): Add weakref flag.
14299 * cgraphunit.c (cgraph_reset_node): Clear weakref flag.
14300 (handle_alias_pairs): Set weakref flag, do not set DECL_EXTERNAL.
14301 (output_weakrefs): Use weakref flag.
14302 * fold-const.c (simple_operand_p): Handle WEAK.
14303 * gimple-fold.c (can_refer_decl_in_current_unit_p): Drop weakref.
14304 * ipa.c (varpool_externally_visible_p): Drop weakref.
14305 (function_and_variable_visibility): Update comment; fix weakref
14306 sanity checks; do not clear DECL_WEAK on them.
14307 * lto-cgraph.c (lto_output_node): update.
14308 (lto_output_varpool_node): Update.
14309 (input_overwrite_node): Update.
14310 (input_node): Update.
14311 (input_varpool_node): Update.
14312 * lto-symtab.c (lto_symtab_symbol_p): Do not special case weakrefs.
14313 (lto_symtab_merge_symbols): Add sanity check.
14314 (lto_symtab_prevailing_decl): Do not special case weakrefs.
14315 * passes.c (rest_of_decl_compilation): Set static flag, too.
14316 * symtab.c (dump_symtab_base): Dump weakref.
14317 (verify_symtab_base): Sanity check weakrefs.
14318 (symtab_make_decl_local): Remove duplicated code.
14319 (symtab_alias_ultimate_target): Simplify.
14320 * varpool.c (varpool_create_variable_alias): Set weakref flag.
c7d8f446 14321
2e6a2f88
TV		 143222013-06-11 Tom de Vries <tom@codesourcery.com>
14323
14324 * genautomata.c (gen_regexp_sequence): Handle els_num == -1. Handle
14325 sequence_vect == NULL.
14326
d9af3652
DD		 143272013-06-11 DJ Delorie <dj@redhat.com>
14328
14329 * config/rl78/rl78.c (TARGET_UNWIND_WORD_MODE): Define.
14330 (rl78_unwind_word_mode): New.
14331
ffa66012
DM		 143322013-06-11 David Malcolm <dmalcolm@redhat.com>
14333
14334 * final.c (debug_prefix_maps): Make static.
14335
783f2d35
DM		 143362013-06-11 David Malcolm <dmalcolm@redhat.com>
14337
14338 * function.c (initial_trampoline): Remove stray copy.
14339
7af26def
SN		 143402013-06-11 Sofiane Naci <sofiane.naci@arm.com>
14341
14342 * config/aarch64/aarch64-simd.md (move_lo_quad_<mode>): Update.
14343
97756c0e
MJ		 143442013-06-11 Martin Jambor <mjambor@suse.cz>
14345
14346 * ipa-cp.c (ipa_get_indirect_edge_target_1): Check that param_index is
14347 within bounds at the beginning of the function.
14348
2c7b8bf6
AM		 143492013-06-11 Alan Modra <amodra@gmail.com>
14350
14351 * varasm.c (get_section): Don't die on !DECL_P decl. Tidy error
14352 reporting.
14353 (get_named_section): Don't NULL !DECL_P decl.
14354
77cea46e
IZ		 143552013-06-11 Igor Zamyatin <igor.zamyatin@intel.com>
14356
14357 * doc/invoke.texi (core-avx2): Document.
14358 (slm): Likewise.
14359 (atom): Updated with MOVBE.
14360
888f0920
RB		 143612013-06-11 Richard Biener <rguenther@suse.de>
14362
22fc7d3d 14363 * collect2.c (main): Do not redirect ld stdout/stderr when debugging.
888f0920 14364
6b438d58
AB		 143652013-06-11 Anton Blanchard <anton@samba.org>
14366
14367 * config/rs6000/rs6000.c (rs6000_adjust_atomic_subword): Calculate
14368 correct shift value in little-endian mode.
14369
3bc66938
JJ		 143702013-06-11 Jakub Jelinek <jakub@redhat.com>
14371
14372 PR target/56564
14373 * varasm.c (get_variable_align): Move #endif to the right place.
14374
a5dc7f88
CC		 143752013-06-10 Cary Coutant <ccoutant@google.com>
14376
14377 * dwarf2out.c (hash_external_ref): Use die_symbol or signature
14378 for hash so that hash table traversal order is deterministic.
14379
19be72ab 143802013-06-10 Michael Meissner <meissner@linux.vnet.ibm.com>
12211b99
UB		 14381 Pat Haugen <pthaugen@us.ibm.com>
14382 Peter Bergner <bergner@vnet.ibm.com>
19be72ab
MM		 14383
14384 * config/rs6000/vector.md (GPR move splitter): Do not split moves
14385 of vectors in GPRS if they are direct moves or quad word load or
14386 store moves.
14387
14388 * config/rs6000/rs6000-protos.h (rs6000_output_move_128bit): Add
14389 declaration.
14390 (direct_move_p): Likewise.
14391 (quad_load_store_p): Likewise.
14392
14393 * config/rs6000/rs6000.c (enum rs6000_reg_type): Simplify register
14394 classes into bins based on the physical register type.
14395 (reg_class_to_reg_type): Likewise.
14396 (IS_STD_REG_TYPE): Likewise.
14397 (IS_FP_VECT_REG_TYPE): Likewise.
14398 (reload_fpr_gpr): Arrays to determine what insn to use if we can
14399 use direct move instructions.
14400 (reload_gpr_vsx): Likewise.
14401 (reload_vsx_gpr): Likewise.
14402 (rs6000_init_hard_regno_mode_ok): Precalculate the register type
14403 information that is a simplification of register classes. Also
14404 precalculate direct move reload helpers.
14405 (direct_move_p): New function to return true if the operation can
14406 be done as a direct move instruciton.
14407 (quad_load_store_p): New function to return true if the operation
14408 is a quad memory operation.
14409 (rs6000_legitimize_address): If quad memory, only allow register
14410 indirect for TImode addresses.
14411 (rs6000_legitimate_address_p): Likewise.
14412 (enum reload_reg_type): Delete, replace with rs6000_reg_type.
14413 (rs6000_reload_register_type): Likewise.
14414 (register_to_reg_type): Return register type.
14415 (rs6000_secondary_reload_simple_move): New helper function for
14416 secondary reload and secondary memory needed to identify anything
14417 that is a simple move, and does not need reloading.
14418 (rs6000_secondary_reload_direct_move): New helper function for
14419 secondary reload to identify cases that can be done with several
14420 instructions via the direct move instructions.
14421 (rs6000_secondary_reload_move): New helper function for secondary
14422 reload to identify moves between register types that can be done.
14423 (rs6000_secondary_reload): Add support for quad memory operations
14424 and for direct move.
14425 (rs6000_secondary_memory_needed): Likewise.
14426 (rs6000_debug_secondary_memory_needed): Change argument names.
14427 (rs6000_output_move_128bit): New function to return the move to
14428 use for 128-bit moves, including knowing about the various
14429 limitations of quad memory operations.
14430
14431 * config/rs6000/vsx.md (vsx_mov<mode>): Add support for quad
14432 memory operations. call rs6000_output_move_128bit for the actual
14433 instruciton(s) to generate.
14434 (vsx_movti_64bit): Likewise.
14435
14436 * config/rs6000/rs6000.md (UNSPEC_P8V_FMRGOW): New unspec values.
14437 (UNSPEC_P8V_MTVSRWZ): Likewise.
14438 (UNSPEC_P8V_RELOAD_FROM_GPR): Likewise.
14439 (UNSPEC_P8V_MTVSRD): Likewise.
14440 (UNSPEC_P8V_XXPERMDI): Likewise.
14441 (UNSPEC_P8V_RELOAD_FROM_VSX): Likewise.
14442 (UNSPEC_FUSION_GPR): Likewise.
14443 (FMOVE128_GPR): New iterator for direct move.
22fc7d3d 14444 (f32_lv): New mode attribute for load/store of SFmode/SDmode values.
19be72ab
MM		 14445 (f32_sv): Likewise.
14446 (f32_dm): Likewise.
14447 (zero_extend<mode>di2_internal1): Add support for power8 32-bit
14448 loads and direct move instructions.
14449 (zero_extendsidi2_lfiwzx): Likewise.
14450 (extendsidi2_lfiwax): Likewise.
14451 (extendsidi2_nocell): Likewise.
14452 (floatsi<mode>2_lfiwax): Likewise.
14453 (lfiwax): Likewise.
14454 (floatunssi<mode>2_lfiwzx): Likewise.
14455 (lfiwzx): Likewise.
14456 (fix_trunc<mode>_stfiwx): Likewise.
14457 (fixuns_trunc<mode>_stfiwx): Likewise.
14458 (mov<mode>_hardfloat, 32-bit floating point): Likewise.
14459 (mov<move>_hardfloat64, 64-bit floating point): Likewise.
14460 (parity<mode>2_cmpb): Set length/type attr.
14461 (unnamed shift right patterns, mov<mode>_internal2): Change type attr
14462 for 'mr.' to fast_compare.
14463 (bpermd_<mode>): Change type attr to popcnt.
14464 (p8_fmrgow_<mode>): New insns for power8 direct move support.
14465 (p8_mtvsrwz_1): Likewise.
14466 (p8_mtvsrwz_2): Likewise.
14467 (reload_fpr_from_gpr<mode>): Likewise.
14468 (p8_mtvsrd_1): Likewise.
14469 (p8_mtvsrd_2): Likewise.
14470 (p8_xxpermdi_<mode>): Likewise.
14471 (reload_vsx_from_gpr<mode>): Likewise.
14472 (reload_vsx_from_gprsf): Likewise.
14473 (p8_mfvsrd_3_<mode>): LIkewise.
14474 (reload_gpr_from_vsx<mode>): Likewise.
14475 (reload_gpr_from_vsxsf): Likewise.
14476 (p8_mfvsrd_4_disf): Likewise.
14477 (multi-word GPR splits): Do not split direct moves or quad memory
14478 operations.
14479
16876bdc
DM		 144802013-06-10 David Malcolm <dmalcolm@redhat.com>
14481
14482 * tree-into-ssa.c (interesting_blocks): Make static.
14483
df8a1d28
JJ		 144842013-06-10 Jakub Jelinek <jakub@redhat.com>
14485
14486 PR target/56564
14487 * varasm.c (align_variable): Don't use DATA_ALIGNMENT or
14488 CONSTANT_ALIGNMENT if !decl_binds_to_current_def_p (decl).
14489 Use DATA_ABI_ALIGNMENT for that case instead if defined.
14490 (get_variable_align): New function.
14491 (get_variable_section, emit_bss, emit_common,
14492 assemble_variable_contents, place_block_symbol): Use
14493 get_variable_align instead of DECL_ALIGN.
14494 (assemble_noswitch_variable): Add align argument, use it
14495 instead of DECL_ALIGN.
14496 (assemble_variable): Adjust caller. Use get_variable_align
14497 instead of DECL_ALIGN.
14498 * config/i386/i386.h (DATA_ALIGNMENT): Adjust x86_data_alignment
14499 caller.
14500 (DATA_ABI_ALIGNMENT): Define.
14501 * config/i386/i386-protos.h (x86_data_alignment): Adjust prototype.
14502 * config/i386/i386.c (x86_data_alignment): Add opt argument. If
14503 opt is false, only return the psABI mandated alignment increase.
14504 * config/c6x/c6x.h (DATA_ALIGNMENT): Renamed to...
14505 (DATA_ABI_ALIGNMENT): ... this.
14506 * config/mmix/mmix.h (DATA_ALIGNMENT): Renamed to...
14507 (DATA_ABI_ALIGNMENT): ... this.
14508 * config/mmix/mmix.c (mmix_data_alignment): Adjust function comment.
14509 * config/s390/s390.h (DATA_ALIGNMENT): Renamed to...
14510 (DATA_ABI_ALIGNMENT): ... this.
14511 * doc/tm.texi.in (DATA_ABI_ALIGNMENT): Document.
14512 * doc/tm.texi: Regenerated.
14513
57b29ca6
UB		 145142013-06-10 Uros Bizjak <ubizjak@gmail.com>
14515
14516 * config/alpha/alpha.c (alpha_emit_xfloating_compare): Also use
14517 cmp_code to construct REG_EQUAL note.
14518
3e56ed50
JJ		 145192013-06-09 Jakub Jelinek <jakub@redhat.com>
14520
14521 PR target/57568
14522 * config/i386/i386.md (TARGET_READ_MODIFY_WRITE peepholes): Ensure
14523 that operands[2] doesn't overlap with operands[0].
14524
74fe2a1d
DE		 145252013-06-09 David Edelsohn <dje.gcc@gmail.com>
14526 Jan Hubicka <jh@suse.cz>
14527
14528 * config/rs6000/rs6000.c (print_operand, 'z'): Remove historical
14529 hack to mark symbols as used.
14530
ed52a84e
VM		 145312013-06-08 Vladimir Makarov <vmakarov@redhat.com>
14532
14533 PR rtl-optimization/57559
14534 * lra-constraints.c (process_alt_operands): Don't discourage
14535 memory with known offset for offsetable memory constraint.
14536 * lra.c (lra_emit_add): Exchange y and z for 2-op add insn.
14537
299316ed
EB		 145382013-06-08 Eric Botcazou <ebotcazou@adacore.com>
14539
14540 * varasm.c (struct oc_local_state): Reorder fields.
14541 (output_constructor_bitfield): Replace OUTER parameter with BIT_OFFSET
14542 and adjust accordingly.
14543 (output_constructor): Reorder initialization code and adjust call to
14544 output_constructor_bitfield.
14545
88ac60d3
JH		 145462013-06-07 Jan Hubicka <jh@suse.cz>
14547
14548 * symtab.c (symtab_resolve_alias): Do not remove alias attribute.
14549
4cdce1a8
DM		 145502013-06-07 David Malcolm <dmalcolm@redhat.com>
14551
14552 * tree-object-size.c (unknown): Make const.
14553
b5e0425c
AK		 145542013-06-07 Andreas Krebbel <Andreas.Krebbel@de.ibm.com>
14555
14556 * config/s390/s390.md (cpu_facility): Add cpu_zarch.
14557 ("*movmem_short", "*clrmem_short", "*cmpmem_short): Use cpu_zarch
14558 for last alternative in the cpu_facility attribute.
14559
79678d04
KT		 145602013-06-07 Kyrylo Tkachov <kyrylo.tkachov@arm.com>
14561
14562 PR target/56315
14563 * config/arm/arm.md (*xordi3_insn): Change to insn_and_split.
14564 (xordi3): Change operand 2 constraint to arm_xordi_operand.
14565 * config/arm/arm.c (const_ok_for_dimode_op): Handle XOR.
14566 * config/arm/constraints.md (Dg): New constraint.
14567 * config/arm/neon.md (xordi3_neon): Remove.
14568 (neon_veor<mode>): Generate xordi3 instead of xordi3_neon.
14569 * config/arm/predicates.md (arm_xordi_operand): New predicate.
14570
b31ddbdb
KT		 145712013-06-07 Kyrylo Tkachov <kyrylo.tkachov@arm.com>
14572
14573 * config/arm/arm.md (anddi3_insn): Remove duplicate alternatives.
14574 Clean up alternatives.
14575
8f90b7d4
AM		 145762013-06-07 Alan Modra <amodra@gmail.com>
14577
14578 * config/rs6000/rs6000.c (setup_incoming_varargs): Round up
14579 va_list_gpr_size.
14580
961ce119
AM		 145812013-06-07 Alan Modra <amodra@gmail.com>
14582
14583 * varasm.c (force_const_mem): Assert mode is not VOID or BLK.
14584
33e49835
KT		 145852013-06-07 Kyrylo Tkachov <kyrylo.tkachov@arm.com>
14586
14587 * config/arm/constraints.md (Df): New constraint.
14588 * config/arm/arm.md (iordi3_insn): Use Df constraint instead of De.
14589 Correct length attribute for last two alternatives.
14590
acd17ae6
AM		 145912013-06-07 Alan Modra <amodra@gmail.com>
14592
14593 * config/rs6000/rs6000.c (rs6000_option_override_internal): Don't
14594 override user -mfp-in-toc.
14595 (offsettable_ok_by_alignment): Consider just the current access
14596 rather than the whole object, unless BLKmode. Handle
14597 CONSTANT_POOL_ADDRESS_P constants that lack a decl too.
14598 (use_toc_relative_ref): Allow CONSTANT_POOL_ADDRESS_P constants
14599 for -mcmodel=medium.
14600 * config/rs6000/linux64.h (SUBSUBTARGET_OVERRIDE_OPTIONS): Don't
14601 override user -mfp-in-toc or -msum-in-toc. Default to
14602 -mno-fp-in-toc for -mcmodel=medium.
14603
73310b0e
DD		 146042013-06-06 DJ Delorie <dj@redhat.com>
14605
14606 * config/rl78/rl78.c (rl78_valid_pointer_mode): New, implements
14607 TARGET_VALID_POINTER_MODE.
14608
0bd62dca 146092013-06-06 Michael Meissner <meissner@linux.vnet.ibm.com>
12211b99
UB		 14610 Pat Haugen <pthaugen@us.ibm.com>
14611 Peter Bergner <bergner@vnet.ibm.com>
0bd62dca
MM		 14612
14613 * doc/extend.texi (PowerPC AltiVec/VSX Built-in Functions):
14614 Document new power8 builtins.
14615
14616 * config/rs6000/vector.md (and<mode>3): Add a clobber/scratch of a
14617 condition code register, to allow 128-bit logical operations to be
14618 done in the VSX or GPR registers.
14619 (nor<mode>3): Use the canonical form for nor.
14620 (eqv<mode>3): Add expanders for power8 xxleqv, xxlnand, xxlorc,
14621 vclz*, and vpopcnt* vector instructions.
14622 (nand<mode>3): Likewise.
14623 (orc<mode>3): Likewise.
14624 (clz<mode>2): LIkewise.
14625 (popcount<mode>2): Likewise.
14626
14627 * config/rs6000/predicates.md (int_reg_operand): Rework tests so
14628 that only the GPRs are recognized.
14629
14630 * config/rs6000/rs6000-c.c (altivec_overloaded_builtins): Add
14631 support for new power8 builtins.
14632
14633 * config/rs6000/rs6000-builtin.def (xscvspdpn): Add new power8
14634 builtin functions.
14635 (xscvdpspn): Likewise.
14636 (vclz): Likewise.
14637 (vclzb): Likewise.
14638 (vclzh): Likewise.
14639 (vclzw): Likewise.
14640 (vclzd): Likewise.
14641 (vpopcnt): Likewise.
14642 (vpopcntb): Likewise.
14643 (vpopcnth): Likewise.
14644 (vpopcntw): Likewise.
14645 (vpopcntd): Likewise.
14646 (vgbbd): Likewise.
14647 (vmrgew): Likewise.
14648 (vmrgow): Likewise.
14649 (eqv): Likewise.
14650 (eqv_v16qi3): Likewise.
14651 (eqv_v8hi3): Likewise.
14652 (eqv_v4si3): Likewise.
14653 (eqv_v2di3): Likewise.
14654 (eqv_v4sf3): Likewise.
14655 (eqv_v2df3): Likewise.
14656 (nand): Likewise.
14657 (nand_v16qi3): Likewise.
14658 (nand_v8hi3): Likewise.
14659 (nand_v4si3): Likewise.
14660 (nand_v2di3): Likewise.
14661 (nand_v4sf3): Likewise.
14662 (nand_v2df3): Likewise.
14663 (orc): Likewise.
14664 (orc_v16qi3): Likewise.
14665 (orc_v8hi3): Likewise.
14666 (orc_v4si3): Likewise.
14667 (orc_v2di3): Likewise.
14668 (orc_v4sf3): Likewise.
14669 (orc_v2df3): Likewise.
14670
14671 * config/rs6000/rs6000.c (rs6000_option_override_internal): Only
14672 allow power8 quad mode in 64-bit.
14673 (rs6000_builtin_vectorized_function): Add support to vectorize
14674 ISA 2.07 count leading zeros, population count builtins.
14675 (rs6000_expand_vector_init): On ISA 2.07 use xscvdpspn to form
14676 V4SF vectors instead of xscvdpsp to avoid IEEE related traps.
14677 (builtin_function_type): Add vgbbd builtin function which takes an
14678 unsigned argument.
14679 (altivec_expand_vec_perm_const): Add support for new power8 merge
14680 instructions.
14681
14682 * config/rs6000/vsx.md (VSX_L2): New iterator for 128-bit types,
14683 that does not include TImdoe for use with 32-bit.
14684 (UNSPEC_VSX_CVSPDPN): Support for power8 xscvdpspn and xscvspdpn
14685 instructions.
14686 (UNSPEC_VSX_CVDPSPN): Likewise.
14687 (vsx_xscvdpspn): Likewise.
14688 (vsx_xscvspdpn): Likewise.
14689 (vsx_xscvdpspn_scalar): Likewise.
14690 (vsx_xscvspdpn_directmove): Likewise.
14691 (vsx_and<mode>3): Split logical operations into 32-bit and
14692 64-bit. Add support to do logical operations on TImode as well as
14693 VSX vector types. Allow logical operations to be done in either
14694 VSX registers or in general purpose registers in 64-bit mode. Add
14695 splitters if GPRs were used. For AND, add clobber of CCmode to
14696 allow use of ANDI on GPRs. Rewrite nor to use the canonical RTL
14697 encoding.
14698 (vsx_and<mode>3_32bit): Likewise.
14699 (vsx_and<mode>3_64bit): Likewise.
14700 (vsx_ior<mode>3): Likewise.
14701 (vsx_ior<mode>3_32bit): Likewise.
14702 (vsx_ior<mode>3_64bit): Likewise.
14703 (vsx_xor<mode>3): Likewise.
14704 (vsx_xor<mode>3_32bit): Likewise.
14705 (vsx_xor<mode>3_64bit): Likewise.
14706 (vsx_one_cmpl<mode>2): Likewise.
14707 (vsx_one_cmpl<mode>2_32bit): Likewise.
14708 (vsx_one_cmpl<mode>2_64bit): Likewise.
14709 (vsx_nor<mode>3): Likewise.
14710 (vsx_nor<mode>3_32bit): Likewise.
14711 (vsx_nor<mode>3_64bit): Likewise.
14712 (vsx_andc<mode>3): Likewise.
14713 (vsx_andc<mode>3_32bit): Likewise.
14714 (vsx_andc<mode>3_64bit): Likewise.
14715 (vsx_eqv<mode>3_32bit): Add support for power8 xxleqv, xxlnand,
14716 and xxlorc instructions.
14717 (vsx_eqv<mode>3_64bit): Likewise.
14718 (vsx_nand<mode>3_32bit): Likewise.
14719 (vsx_nand<mode>3_64bit): Likewise.
14720 (vsx_orc<mode>3_32bit): Likewise.
14721 (vsx_orc<mode>3_64bit): Likewise.
14722
14723 * config/rs6000/rs6000.h (VLOGICAL_REGNO_P): Update comment.
14724
14725 * config/rs6000/altivec.md (UNSPEC_VGBBD): Add power8 vgbbd
14726 instruction.
14727 (p8_vmrgew): Add power8 vmrgew and vmrgow instructions.
14728 (p8_vmrgow): Likewise.
14729 (altivec_and<mode>3): Add clobber of CCmode to allow AND using
14730 GPRs to be split under VSX.
14731 (p8v_clz<mode>2): Add power8 count leading zero support.
14732 (p8v_popcount<mode>2): Add power8 population count support.
14733 (p8v_vgbbd): Add power8 gather bits by bytes by doubleword
14734 support.
14735
14736 * config/rs6000/rs6000.md (eqv<mode>3): Add support for powerp eqv
14737 instruction.
14738
14739 * config/rs6000/altivec.h (vec_eqv): Add defines to export power8
14740 builtin functions.
14741 (vec_nand): Likewise.
14742 (vec_vclz): Likewise.
14743 (vec_vclzb): Likewise.
14744 (vec_vclzd): Likewise.
14745 (vec_vclzh): Likewise.
14746 (vec_vclzw): Likewise.
14747 (vec_vgbbd): Likewise.
14748 (vec_vmrgew): Likewise.
14749 (vec_vmrgow): Likewise.
14750 (vec_vpopcnt): Likewise.
14751 (vec_vpopcntb): Likewise.
14752 (vec_vpopcntd): Likewise.
14753 (vec_vpopcnth): Likewise.
14754 (vec_vpopcntw): Likewise.
14755
37684c46
VM		 147562013-06-06 Vladimir Makarov <vmakarov@redhat.com>
14757
14758 PR rtl-optimization/57468
14759 * config/i386/i386.c (inline_secondary_memory_needed): Ignore
14760 spilled pseudos.
14761
d89ae27c
VM		 147622013-06-06 Vladimir Makarov <vmakarov@redhat.com>
14763
14764 PR rtl-optimization/57459
14765 * lra-constraints.c (update_ebb_live_info): Fix typo for operand
14766 type when setting live regs.
14767
3597e113
VM		 147682013-06-06 Vladimir Makarov <vmakarov@redhat.com>
14769
14770 * config/s390/s390.opt (mlra): New option.
14771 * config/s390/s390.c (s390_decompose_address): Check displacement
14772 for all registers for LRA.
14773 (s390_secondary_reload): Don't used secondary reloads for LRA.
14774 (s390_lra_p): New function.
14775 (TARGET_LRA_P): Define.
14776 * config/s390/s390.md (*movmem_short, *clrmem_short): Change value
14777 of attribute cpu_facility to zarch for the last alternative.
14778 (*cmpmem_short): Ditto.
14779
01037aeb
EB		 147802013-06-06 Eric Botcazou <ebotcazou@adacore.com>
14781
14782 * config/arm/arm.c (arm_r3_live_at_start_p): New predicate.
14783 (arm_compute_static_chain_stack_bytes): Use it. Tidy up.
14784 (arm_expand_prologue): Likewise.
14785
3371a64f
TJ		 147862013-06-06 Teresa Johnson <tejohnson@google.com>
14787
14788 PR c++/53743
14789 * ifcvt.c (find_if_case_1): Replace BB_COPY_PARTITION with assert
14790 as this is now done by redirect_edge_and_branch_force.
14791 * function.c (thread_prologue_and_epilogue_insns): Insert new bb after
14792 barriers, and fix interaction with splitting.
14793 * emit-rtl.c (try_split): Copy REG_CROSSING_JUMP notes.
14794 * cfgcleanup.c (try_forward_edges): Fix early return value to properly
14795 reflect changes made in the routine.
14796 * bb-reorder.c (emit_barrier_after_bb): Move to cfgrtl.c.
14797 (fix_up_fall_thru_edges): Remove incorrect check for bb layout order
14798 since this is called in cfglayout mode, and replace partition fixup
14799 with assert as that is now done by force_nonfallthru_and_redirect.
14800 (add_reg_crossing_jump_notes): Handle the fact that some jumps may
14801 already be marked with region crossing note.
14802 (insert_section_boundary_note): Make non-static, gate on flag
14803 has_bb_partition, rewrite to also check for multiple partitions.
14804 (rest_of_handle_reorder_blocks): Remove call to
14805 insert_section_boundary_note, now done later during free_cfg.
14806 (duplicate_computed_gotos): Don't duplicate partition crossing edge.
14807 * bb-reorder.h (insert_section_boundary_note): Declare.
14808 * Makefile.in (cfgrtl.o): Depend on bb-reorder.h
14809 * cfgrtl.c (rest_of_pass_free_cfg): If partitions exist
14810 invoke insert_section_boundary_note.
14811 (try_redirect_by_replacing_jump): Remove unnecessary
14812 check for region crossing note.
14813 (fixup_partition_crossing): New function.
14814 (rtl_redirect_edge_and_branch): Fixup partition boundaries.
14815 (emit_barrier_after_bb): Move here from bb-reorder.c, handle insertion
14816 in non-cfglayout mode.
14817 (force_nonfallthru_and_redirect): Fixup partition boundaries,
14818 remove old code that tried to do this. Emit barrier correctly
14819 when we are in cfglayout mode.
14820 (last_bb_in_partition): New function.
14821 (rtl_split_edge): Correctly fixup partition boundaries.
14822 (commit_one_edge_insertion): Remove old code that tried to
14823 fixup region crossing edge since this is now handled in
14824 split_block, and set up insertion point correctly since
14825 block may now end in a jump.
14826 (verify_hot_cold_block_grouping): Guard against checking when not in
14827 linearized RTL mode.
14828 (rtl_verify_edges): Add checks for incorrect/missing REG_CROSSING_JUMP
14829 notes.
14830 (rtl_verify_flow_info_1): Move verify_hot_cold_block_grouping to
14831 rtl_verify_flow_info, so not called in cfglayout mode.
14832 (rtl_verify_flow_info): Move verify_hot_cold_block_grouping here.
14833 (fixup_reorder_chain): Remove old code that attempted to fixup region
14834 crossing note as this is now handled in force_nonfallthru_and_redirect.
14835 (duplicate_insn_chain): Don't duplicate switch section notes.
14836 (rtl_can_remove_branch_p): Remove unnecessary check for region crossing
14837 note.
14838 * basic-block.h (emit_barrier_after_bb): Declare.
14839
66071e10
KT		 148402013-06-06 Kyrylo Tkachov <kyrylo.tkachov@arm.com>
14841
14842 * config/arm/arm-fixed.md (add<mode>3,usadd<mode>3,ssadd<mode>3,
14843 sub<mode>3, ussub<mode>3, sssub<mode>3, arm_ssatsihi_shift,
14844 arm_usatsihi): Adjust alternatives for arm_restrict_it.
14845
0e26bf3d
KT		 148462013-06-06 Kyrylo Tkachov <kyrylo.tkachov@arm.com>
14847
14848 * config/arm/arm-ldmstm.ml: Set "predicable_short_it" to "no"
14849 where appropriate.
14850 * config/arm/ldmstm.md: Regenerate.
14851
12b4e7ef
KT		 148522013-06-06 Kyrylo Tkachov <kyrylo.tkachov@arm.com>
14853
14854 * config/arm/sync.md (atomic_loaddi_1):
14855 Disable predication for arm_restrict_it.
14856 (arm_load_exclusive<mode>): Likewise.
14857 (arm_load_exclusivesi): Likewise.
14858 (arm_load_exclusivedi): Likewise.
14859 (arm_load_acquire_exclusive<mode>): Likewise.
14860 (arm_load_acquire_exclusivesi): Likewise.
14861 (arm_load_acquire_exclusivedi): Likewise.
14862 (arm_store_exclusive<mode>): Likewise.
14863 (arm_store_exclusive<mode>): Likewise.
14864 (arm_store_release_exclusivedi): Likewise.
14865 (arm_store_release_exclusive<mode>): Likewise.
14866
15d16c8a
RB		 148672013-06-06 Richard Biener <rguenther@suse.de>
14868
14869 * lto-streamer.h (enum LTO_tags): Move LTO_tree_pickle_reference
14870 after LTO_null.
14871 (lto_tag_is_tree_code_p): Adjust.
14872 (lto_tag_is_gimple_code_p): Likewise.
14873 (lto_gimple_code_to_tag): Likewise.
14874 (lto_tag_to_gimple_code): Likewise.
14875 (lto_tree_code_to_tag): Likewise.
14876 (lto_tag_to_tree_code): Likewise.
14877 * data-streamer.h (streamer_write_hwi_in_range): Use
14878 uhwi streaming to stream the normalized range.
14879 (streamer_read_hwi_in_range): Likewise.
14880
17a7fc37
KT		 148812013-06-05 Kyrylo Tkachov <kyrylo.tkachov@arm.com>
14882
14883 * config/arm/arm.md (enabled_for_depr_it): New attribute.
14884 (predicable_short_it): Likewise.
14885 (predicated): Likewise.
14886 (enabled): Handle above.
14887 (define_cond_exec): Set predicated attribute to yes.
14888
b57ca59b
MS		 148892013-06-05 Mike Stump <mikestump@comcast.net>
14890
14891 * gdbinit.in (__FUNCTION__): Add.
14892
c1e183a9
UB		 148932013-06-05 Uros Bizjak <ubizjak@gmail.com>
14894
14895 * config/alpha/alpha.c (alpha_emit_conditional_move): Swap all
14896 GE, GT, GEU and GTU compares, modulo DImode compares with zero.
14897
0378bddb
JH		 148982013-06-05 Jan Hubicka <jh@suse.cz>
14899
14900 * varasm.c (mark_decl_referenced): Revert the removal until targets
14901 are fixed.
14902
8f8a2057
DE		 149032013-06-05 David Edelsohn <dje.gcc@gmail.com>
14904
14905 * config/rs6000/rs6000.c (print_operand, 'z'): Use DECL_PRESERVE_P
14906 instead of mark_decl_referenced.
14907
edb983b2
JH		 149082013-06-05 Jan Hubicka <jh@suse.cz>
14909
14910 * cgraph.c (cgraph_remove_node): Clear forced_by_abi.
14911 (cgraph_node_cannot_be_local_p_1): Honnor symbol.forced_by_abi
14912 and symtab_used_from_object_file_p.
14913 (cgraph_make_node_local_1): Clear forced_by_abi.
14914 (cgraph_can_remove_if_no_direct_calls_and): Use forced_by_abi
14915 * cgraph.h (symtab_node_base): Add forced_by_abi.
14916 (decide_is_variable_needed): Remove.
14917 (varpool_can_remove_if_no_refs): Honnor symbol.forced_by_abi.
14918 * cgraphunit.c (cgraph_decide_is_function_needed): Rename to ..
14919 (decide_is_symbol_needed): ... this one; handle symbols in general;
14920 always analyze virtuals; honnor forced_by_abi.
14921 (cgraph_finalize_function): Update.
14922 (varpool_finalize_decl): Update.
14923 (symbol_defined_and_needed): Remove.
14924 (analyze_functions): Update.
14925 * lto-cgraph.c (lto_output_node, lto_output_varpool_node,
14926 output_refs, input_overwrite_node): Handle forced_by_abi.
14927 * ipa.c (cgraph_address_taken_from_non_vtable_p): Rename to ...
14928 (address_taken_from_non_vtable_p): ... this one.
14929 (comdat_can_be_unshared_p_1): New function.
14930 (cgraph_comdat_can_be_unshared_p): Rename to ...
14931 (comdat_can_be_unshared_p): ... this one; handle symbols in general.
14932 (varpool_externally_visible_p): Use comdat_can_be_unshared_p.
14933 (function_and_variable_visibility): Clear forced_by_abi as needed.
14934 * trans-mem.c (ipa_tm_mark_forced_by_abi_node): New functoin.
14935 (ipa_tm_create_version_alias, ipa_tm_create_version): Update.
edb983b2
JH		 14936 * symtab.c (dump_symtab_base): Dump forced_by_abi.
14937 * varpool.c (decide_is_variable_needed): Remove.
14938
9912dbe5
KT		 149392013-06-05 Kyrylo Tkachov <kyrylo.tkachov@arm.com>
14940
14941 * config/arm/arm.c (MAX_INSN_PER_IT_BLOCK): New macro.
14942 (arm_option_override): Override arm_restrict_it where appropriate.
14943 (thumb2_final_prescan_insn): Use MAX_INSN_PER_IT_BLOCK.
14944 * config/arm/arm.opt (mrestrict-it): New command-line option.
14945 * doc/invoke.texi: Document -mrestrict-it.
14946
34c136b6
DM		 149472013-06-05 David Malcolm <dmalcolm@redhat.com>
14948
14949 * tsan.c (tsan_atomic_table): Make const.
14950
a367df53
RB		 149512013-06-05 Richard Biener <rguenther@suse.de>
14952
14953 * tree-streamer.c (streamer_tree_cache_insert_1): Update the
14954 index associated with the tree we are supposed to replace.
14955 * tree-streamer-out.c (pack_ts_base_value_fields): Output
14956 TREE_ASM_WRITTEN as zero for everything but SSA names.
14957
70d51a19
DM		 149582013-06-05 David Malcolm <dmalcolm@redhat.com>
14959
14960 * tree-ssa-structalias.c (call_stmt_vars): Make static.
14961
b9bd2075
JH		 149622013-06-04 Jan Hubicka <jh@suse.cz>
14963
14964 * lto-cgraph.c (get_alias_symbol): Remove weakref sanity check.
14965 (input_node, input_varpool_node): Handle correctly external same
14966 body aliases.
14967 * ipa.c (symtab_remove_unreachable_nodes): Do not remove external
14968 nodes at ltrans stage.
14969
aaae719d
JH		 149702013-06-04 Jan Hubicka <jh@suse.cz>
14971
14972 * ipa-inline.c (update_caller_keys): Fix availability test.
14973 (update_callee_keys): Likewise.
14974 * symtab.c (symtab_alias_ultimate_target): Make availaiblity logic
14975 to follow ELF standard.
14976
107eea2c
JU		 149772013-06-04 Jürgen Urban <JuergenUrban@gmx.de>
14978
14979 * config.gcc (mipsr5900-*-elf*, mipsr5900el-*-elf*, mips64r5900-*-elf*)
14980 (mips64r5900el-*-elf*): New configurations.
14981 * config/mips/mips-cpus.def (r5900): New processor.
14982 * config/mips/mips-tables.opt: Regenerate.
14983 * config/mips/mips.c (mips_rtx_cost_data): Add an R5900 entry.
14984 (mips_issue_rate): Handle PROCESSOR_R5900.
14985 (mips_reorg_process_insns): Force reorder mode for the R5900.
14986 * config/mips/mips.h (TARGET_MIPS5900): Define.
14987 (ISA_HAS_CONDMOVE, ISA_HAS_PREFETCH, ISA_HAS_HILO_INTERLOCKS): Include
14988 TARGET_MIPS5900.
14989 (ISA_HAS_LOAD_DELAY, ISA_HAS_XFER_DELAY, ISA_HAS_FCMP_DELAY): Exclude
14990 TARGET_MIPS5900.
14991 * config/mips/mips.md (processor): Add r5900.
14992 (MOVECC): Disallow CCmode conditions for TARGET_MIPS5900.
14993
b7342d25
IB		 149942013-06-04 Ian Bolton <ian.bolton@arm.com>
14995
14996 * config/aarch64/aarch64.md (*mov<mode>_aarch64): Call
14997 into function to generate MOVI instruction.
c1e183a9 14998 * config/aarch64/aarch64.c (aarch64_simd_container_mode): New function.
b7342d25
IB		 14999 (aarch64_preferred_simd_mode): Turn into wrapper.
15000 (aarch64_output_scalar_simd_mov_immediate): New function.
15001 * config/aarch64/aarch64-protos.h: Add prototype for above.
15002
81c2dfb9
IB		 150032013-06-04 Ian Bolton <ian.bolton@arm.com>
15004
15005 * config/aarch64/aarch64.c (simd_immediate_info): Remove
15006 element_char member.
15007 (sizetochar): Return signed char.
15008 (aarch64_simd_valid_immediate): Remove elchar and other
15009 unnecessary variables.
15010 (aarch64_output_simd_mov_immediate): Take rtx instead of &rtx.
15011 Calculate element_char as required.
15012 * config/aarch64/aarch64-protos.h: Update and move prototype
15013 for aarch64_output_simd_mov_immediate.
15014 * config/aarch64/aarch64-simd.md (*aarch64_simd_mov<mode>):
15015 Update arguments.
15016
48063b9d
IB		 150172013-06-04 Ian Bolton <ian.bolton@arm.com>
15018
15019 * config/aarch64/aarch64.c (simd_immediate_info): Struct to hold
15020 information completed by aarch64_simd_valid_immediate.
15021 (aarch64_legitimate_constant_p): Update arguments.
15022 (aarch64_simd_valid_immediate): Work with struct rather than many
15023 pointers.
15024 (aarch64_simd_scalar_immediate_valid_for_move): Update arguments.
15025 (aarch64_simd_make_constant): Update arguments.
15026 (aarch64_output_simd_mov_immediate): Work with struct rather than
15027 many pointers. Output immediate directly rather than as operand.
15028 * config/aarch64/aarch64-protos.h (aarch64_simd_valid_immediate):
15029 Update prototype.
15030 * config/aarch64/constraints.md (Dn): Update arguments.
15031
3ea63f60
IB		 150322013-06-04 Ian Bolton <ian.bolton@arm.com>
15033
15034 * config/aarch64/aarch64.c (aarch64_simd_valid_immediate): No
15035 longer static.
15036 (aarch64_simd_immediate_valid_for_move): Remove.
15037 (aarch64_simd_scalar_immediate_valid_for_move): Update call.
15038 (aarch64_simd_make_constant): Update call.
15039 (aarch64_output_simd_mov_immediate): Update call.
15040 * config/aarch64/aarch64-protos.h (aarch64_simd_valid_immediate):
15041 Add prototype.
15042 * config/aarch64/constraints.md (Dn): Update call.
15043
d8edd899
IB		 150442013-06-04 Ian Bolton <ian.bolton@arm.com>
15045
15046 * config/aarch64/aarch64.c (aarch64_simd_valid_immediate): Change
15047 return type to bool for prototype.
15048 (aarch64_legitimate_constant_p): Check for true instead of not -1.
15049 (aarch64_simd_valid_immediate): Fix up each return to return a bool.
15050 (aarch64_simd_immediate_valid_for_move): Update retval for bool.
15051
44b20bb8
CM		 150522013-06-04 Catherine Moore <clm@codesourcery.com>
15053
15054 * config/mips/mips.opt (meva): New.
c1e183a9 15055 * config/mips/mips.h (TARGET_CPU_CPP_BUILTINS): Define __mips_eva.
44b20bb8
CM		 15056 (ASM_SPEC): Handle -meva.
15057 * doc/invoke.texi (meva): Document.
15058
52befbd8
AM		 150592013-06-04 Alan Modra <amodra@gmail.com>
15060
15061 * config/rs6000/rs6000.c (output_toc): Correct little-endian float
15062 constant output.
15063
aadaf24e
KT		 150642013-06-04 Kyrylo Tkachov <kyrylo.tkachov@arm.com>
15065
15066 * rtl.def: Add extra fourth optional field to define_cond_exec.
15067 * gensupport.c (process_one_cond_exec): Process attributes from
15068 define_cond_exec.
15069 * doc/md.texi: Document fourth field in define_cond_exec.
15070
0cd9e9ee
EB		 150712013-06-04 Eric Botcazou <ebotcazou@adacore.com>
15072
15073 * expmed.c (extract_bit_field_1): In the larger-than-a-word case, factor
15074 out the processing order as in store_bit_field_1.
15075
cb0f665d
JH		 150762013-06-04 Jan Hubicka <jh@suse.cz>
15077
15078 PR middle-end/57500
15079 * cgraphunit.c (cgraph_process_same_body_aliases): Create
15080 non-VAR_DECL node if it does not exist yet.
15081
53984b9b
RS		 150822013-06-03 Richard Sandiford <rdsandiford@googlemail.com>
15083
15084 * config.gcc (mipsisa64sr71k-*-elf*, mipsisa64sb1-*-elf*)
15085 (mipsisa64sb1el-*-elf*, mips64-*-elf*, mips64el-*-elf*)
15086 (mips64orion-*-elf*, mips64orionel-*-elf*): Remove
15087 target_cpu_default setting.
15088
5d318fd4
TJ		 150892013-06-03 Teresa Johnson <tejohnson@google.com>
15090
15091 * dumpfile.c (opt_info_switch_p): Change -fopt-info
15092 default to -fopt-info=optimized instead of all.
15093 * doc/invoke.texi: Ditto.
15094 * tree-vectorizer.c (vectorize_loops): Emit loop vectorization
15095 success under MSG_OPTIMIZED_LOCATIONS, and use dump_printf_loc.
15096 (execute_vect_slp): Emit BB vectorization success under
15097 MSG_OPTIMIZED_LOCATIONS.
15098 * tree-vect-slp.c (vect_slp_transform_bb): Change
15099 MSG_OPTIMIZED_LOCATIONS to MSG_NOTE.
15100 * tree-vect-loop.c (vect_transform_loop): Ditto.
15101
ec3af349
JM		 151022013-06-03 Jason Merrill <jason@redhat.com>
15103
15104 PR c++/57415
15105 * config/rs6000/rs6000-c.c (altivec_resolve_overloaded_builtin):
15106 Use TARGET_EXPR for C++.
15107
1baf9159
JJ		 151082013-06-03 Jakub Jelinek <jakub@redhat.com>
15109
15110 PR rtl-optimization/57268
0cd9e9ee 15111 * sched-deps.c (sched_analyze_2): Don't flush_pending_lists
1baf9159
JJ		 15112 if DEBUG_INSN_P (insn).
15113
15114 Reapply
15115 2013-05-31 Dinar Temirbulatov <dinar@kugelworks.com>
15116
15117 PR rtl-optimization/57268
15118 * sched-deps.c (sched_analyze_2): Flush dependence lists if
15119 the sum of the read and write lists exceeds MAX_PENDING_LIST_LENGTH.
15120
0f1d3965
YR		 151212013-06-03 Yuri Rumyantsev <yuri.s.rumyantsev@intel.com>
15122
15123 * config/i386/i386.c (ix86_lea_outperforms): Fix formatting.
15124 (ix86_avoid_lea_for_addr): Likewise.
15125 (exact_dependency_1): Likewise.
15126 (ix86_adjust_cost): Likewise.
15127 (swap_top_of_ready_list): Fix formatting and !reload_completed check
15128 removed.
15129 (do_reorder_for_imul): Fix typo, formatting and
15130 !reload_completed check removed.
15131 (ix86_sched_reorder): Fix typo and formatting.
15132 (fold_builtin_cpu): Move M_INTEL_SLM at the end of processor types
15133 list.
15134
3bc30db4
SN		 151352013-06-03 Sofiane Naci <sofiane.naci@arm.com>
15136
15137 * config/aarch64/aarch64.md (*movdi_aarch64): Define "simd" attribute.
15138
19cc6fac
EB		 151392013-06-03 Eric Botcazou <ebotcazou@adacore.com>
15140
15141 * varasm.c (output_constant) <CONSTRUCTOR>: Minor formatting tweak.
15142 <STRING_CST>: Likewise.
15143 <VECTOR_CST>: Likewise.
15144
c68a6e08
JW		 151452013-06-01 Janus Weil <janus@gcc.gnu.org>
15146 Mikael Morin <mikael@gcc.gnu.org>
15147
15148 * configure.ac: Add AC_HEADER_TIOCGWINSZ macro.
15149 * config.in: Regenerated.
15150 * configure: Regenerated.
15151
38e55e5c
JH		 151522013-06-01 Jan Hubicka <jh@suse.cz>
15153
15154 PR middle-end/57366
15155 * cgraphunit.c (compile): When weakref is not supported,
15156 set up transparent aliases before final output pass.
15157 * varasm.c (assemble_alias): Do not try to do it here.
15158
eb51d2ff
JH		 151592013-06-01 Jan Hubicka <jh@suse.cz>
15160
15161 PR middle-end/57467
15162 * passes.c (for_per_function): Skip unanalyzed functions.
15163
40a7fe1e
JH		 151642013-06-01 Jan Hubicka <jh@suse.cz>
15165
15166 * lto-symtab.c (lto_symtab_merge_cgraph_nodes_1): Rename to ...
15167 (lto_symtab_merge_symbols_1): ... this one.
15168 (lto_symtab_merge_cgraph_nodes): Rename to ...
15169 (lto_symtab_merge_symbols): ... this one; simplify.
15170 * cgraph.c (same_body_aliases_done): Rename to ...
15171 (cpp_implicit_aliases_done): ... this one.
15172 (cgraph_create_function_alias): Update.
15173 (cgraph_same_body_alias): Update.
c1e183a9 15174 (dump_cgraph_node): Remove alias dumping; simplify thunk dumping.
40a7fe1e 15175 (verify_edge_corresponds_to_fndecl): Simplify.
c1e183a9 15176 * cgraph.h (symtab_node_base): Add cpp_implicit_alias, alias_target.
40a7fe1e
JH		 15177 (cgraph_node): Remove same_body_alias.
15178 (varpool_node): Remove alias_of and extra_name_alias.
15179 (same_body_aliases_done): Rename to ..
15180 (cpp_implicit_aliases_done): ... this one.
15181 (symtab_alias_ultimate_target): Add default parameter.
15182 (symtab_resolve_alias): New function.
15183 (fixup_same_cpp_alias_visibility): Declare.
15184 (cgraph_function_node): Add default parameter.
15185 (cgraph_node_asm_name): Likewise.
15186 (cgraph_function_or_thunk_node): Add default parameter; do
15187 not ICE when it is NULL.
15188 (varpool_variable_node): Likewise.
15189 * tree-emutls.c (create_emultls_var): Update.
15190 (ipa_lower_emutls): Update.
15191 * cgraphunit.c (cgraph_decide_is_function_needed): Update.
15192 (cgraph_reset_node): Reset alias info.
15193 (cgraph_finalize_function): Update.
15194 (fixup_same_cpp_alias_visibility): Move to symtab.c.
15195 (analyze_function): Simplify.
15196 (cgraph_process_same_body_aliases): Simplify.
15197 (analyze_functions): Fixup same body aliases.
15198 (handle_alias_pairs): Simplify.
15199 (assemble_thunk): Update.
15200 (assemble_thunks_and_aliases): Update.
15201 (output_weakrefs): Rewrite.
15202 * lto-cgraph.c (lto_output_node): Rewrite alias handling.
15203 (lto_output_varpool_node): Likewise.
15204 (compute_ltrans_boundary): Remve assert.
15205 (get_alias_symbol): New functoin.
15206 (input_node): Rewrite alias handling.
15207 (input_varpool_node): Likewise.
15208 * ipa-pure-const.c (propagate_pure_const): Fix formating.
15209 * ipa.c (process_references): Handle weakrefs correctly.
15210 (symtab_remove_unreachable_nodes): Likewise.
15211 * trans-mem.c (get_cg_data): Update.
15212 (ipa_tm_create_version_alias): Update.
15213 (ipa_tm_execute): Update.
15214 * symtab.c (dump_symtab_base): Dump aliases.
15215 (verify_symtab_base): Verify aliases.
15216 (symtab_node_availability): New function.
15217 (symtab_alias_ultimate_target): Simplify.
15218 (fixup_same_cpp_alias_visibility): Move here from cgraphunit.c;
15219 handle all the fixup cases.
15220 (symtab_resolve_alias): New function.
15221 * passes.c (ipa_write_summaries): Handle weakrefs.
15222 * varpool.c (varpool_analyze_node): Simplify.
15223 (assemble_aliases): Update.
15224 (varpool_create_variable_alias): Simplify.
15225 (varpool_extra_name_alias): Simplify.
15226 * lto-streamer.h (lto_symtab_merge_cgraph_nodes): Rename to...
15227 (lto_symtab_merge_symbols): ... this one.
15228
21759881
DT		 152292013-06-01 Dinar Temirbulatov <dinar@kugelworks.com>
15230
15231 Revert
15232 PR rtl-optimization/57268
15233 * sched-deps.c (sched_analyze_2): Flush dependence lists if
15234 the sum of the read and write lists exceeds MAX_PENDING_LIST_LENGTH.
15235
daf4e940
TB		 152362013-06-01 Tobias Burnus <burnus@net-b.de>
15237
15238 Partially reverted:
15239 2013-05-31 Tobias Burnus <burnus@net-b.de>
15240
15241 PR middle-end/57073
15242 * tree-ssa-math-opts.c (execute_cse_sincos): Move check
15243 further up.
15244
c359d8d0
DT		 152452013-05-31 Dinar Temirbulatov <dinar@kugelworks.com>
15246
15247 PR rtl-optimization/57268
15248 * sched-deps.c (sched_analyze_2): Flush dependence lists if
15249 the sum of the read and write lists exceeds MAX_PENDING_LIST_LENGTH.
15250
4172215d
EB		 152512013-05-31 Eric Botcazou <ebotcazou@adacore.com>
15252
15253 * config/rs6000/predicates.md (rs6000_cbranch_operator): Accept some
15254 unordered comparison operators when -fno-trapping-math is in effect
15255 on the e500.
15256 * config/rs6000/rs6000.c (rs6000_generate_compare): Remove dead code
15257 and implement unordered comparison operators properly on the e500.
15258
a8c50132
EB		 152592013-05-31 Eric Botcazou <ebotcazou@adacore.com>
15260
15261 * simplify-rtx.c (simplify_byte_swapping_operation): Use proper macro
15262 for constant scalar integers.
15263 (simplify_relational_operation_1): Likewise.
15264
44626634
SB		 152652013-05-31 Segher Boessenkool <segher@kernel.crashing.org>
15266
15267 * config/rs6000/rs6000-opts.h (enum processor_type): Reorder.
15268 * config/rs6000/rs6000.md (cpu): Reorder. Split long line.
15269 Fix comment.
15270
55805e54
YR		 152712013-05-31 Yuri Rumyantsev <yuri.s.rumyantsev@intel.com>
15272 Igor Zamyatin <igor.zamyatin@intel.com>
15273
15274 Silvermont (SLM) architecture performance tuning.
15275 * config/i386/i386.h (enum ix86_tune_indices): Add
15276 X86_TUNE_SPLIT_MEM_OPND_FOR_FP_CONVERTS.
15277 (TARGET_SPLIT_MEM_OPND_FOR_FP_CONVERTS): New define.
15278
15279 * config/i386/i386.c (initial_ix86_tune_features)
15280 <X86_TUNE_SPLIT_MEM_OPND_FOR_FP_CONVERTS>: Initialize.
15281 (ix86_lea_outperforms): Handle Silvermont tuning.
15282 (ix86_avoid_lea_for_add): Add new argument to ix86_lea_outperforms
15283 call.
15284 (ix86_use_lea_for_mov): Likewise.
15285 (ix86_avoid_lea_for_addr): Likewise.
15286 (ix86_lea_for_add_ok): Likewise.
15287 (exact_dependency_1): New function.
15288 (exact_store_load_dependency): Likewise.
15289 (ix86_adjust_cost): Handle Silvermont tuning.
15290 (do_reoder_for_imul): Likewise.
15291 (swap_top_of_ready_list): New function.
15292 (ix86_sched_reorder): Changed to handle Silvermont tuning.
15293
15294 * config/i386/i386.md (peepholes that split memory operand in fp
15295 converts): New.
15296
e19c9de2
MS		 152972013-05-31 Marcus Shawcroft <marcus.shawcroft@arm.com>
15298
15299 * config/aarch64/aarch64.c (aarch64_load_symref_appropriately):
15300 Remove un-necessary braces.
15301
38e6c9a6
MS		 153022013-05-31 Marcus Shawcroft <marcus.shawcroft@arm.com>
15303
15304 * config/aarch64/aarch64.c (aarch64_classify_symbol):
15305 Use SYMBOL_TINY_ABSOLUTE for AARCH64_CMODEL_TINY_PIC.
15306
e3530904
TB		 153072013-05-31 Tobias Burnus <burnus@net-b.de>
15308
15309 PR middle-end/57073
c1e183a9 15310 * tree-ssa-math-opts.c (execute_cse_sincos): Move check further up.
e3530904 15311
b6af05a9
KT		 153122013-05-31 Kyrylo Tkachov <kyrylo.tkachov@arm.com>
15313
15314 PR target/56315
15315 * config/arm/arm.c (const_ok_for_dimode_op): Handle IOR.
15316 * config/arm/arm.md (*iordi3_insn): Change to insn_and_split.
15317 * config/arm/neon.md (iordi3_neon): Remove.
15318 (neon_vorr<mode>): Generate iordi3 instead of iordi3_neon.
15319 * config/arm/predicates.md (imm_for_neon_logic_operand):
15320 Move to earlier in the file.
15321 (neon_logic_op2): Likewise.
15322 (arm_iordi_operand_neon): New predicate.
15323
f800c166
RB		 153242013-05-31 Richard Biener <rguenther@suse.de>
15325
15326 PR tree-optimization/57478
15327 PR tree-optimization/57453
15328 * tree-vect-slp.c (vect_bb_slp_scalar_cost): Uses in PHI nodes
15329 are life as well.
15330
e571fa59
KP		 153312013-05-31 Kaushik Phatak <kaushik.phatak@kpitcummins.com>
15332
15333 * config/rl78/rl78.md (mulqi3,mulhi3): New define_expands.
15334 (*mulqi3_rl78,*mulhi3_rl78,*mulhi3_g13): New define_insns.
15335
0fa6e0ef
TB		 153362013-05-30 Tobias Burnus <burnus@net-b.de>
15337 Thomas Koenig <tkoenig@gcc.gnu.org>
15338
15339 PR middle-end/57073
15340 * tree-ssa-math-opts.c (execute_cse_sincos): Optimize
15341 powi (-1.0, k) to (k & 1) ? -1.0 : 1.0.
15342
4688ddf5
SB		 153432013-05-30 Steven Bosscher <steven@gcc.gnu.org>
15344
0fa6e0ef 15345 * rtlanal.c (tablejump_p): Expect table and label to be adjacent.
4688ddf5 15346
3b9ceb4b
VM		 153472013-05-30 Vladimir Makarov <vmakarov@redhat.com>
15348
15349 * target.def (register_usage_leveling_p): New hook.
15350 * targhooks.c (default_register_usage_leveling_p): New.
15351 * targhooks.h (default_register_usage_leveling_p): New prototype.
15352 * lra-assigns.c (register_usage_leveling_p): Use the hook.
15353 * doc/tm.texi.in (TARGET_REGISTER_USAGE_LEVELING_P): New hook.
15354 * doc/tm.texi: Update.
15355 * config/i386/i386.c (TARGET_REGISTER_USAGE_LEVELING_P): Define.
15356
65f9e789
IB		 153572013-05-30 Ian Bolton <ian.bolton@arm.com>
15358
15359 * config/aarch64/aarch64.md (insv<mode>): New define_expand.
15360 (*insv_reg<mode>): New define_insn.
15361
12211b99 153622013-05-30 Joern Rennecke <joern.rennecke@embecosm.com>
ca035367
JR		 15363
15364 PR rtl-optimization/57439
15365 * postreload.c (move2add_valid_value_p): Check that we have
15366 a zero subreg_regno_offset when accessing the register in
15367 the requested mode.
15368
0b871ccf
YR		 153692013-05-30 Yuri Rumyantsev <yuri.s.rumyantsev@intel.com>
15370 Igor Zamyatin <igor.zamyatin@intel.com>
15371
15372 Silvermont (SLM) architecture pipeline model, tuning and
15373 insn selection.
15374 * config.gcc: Add slm config options and target.
15375
15376 * config/i386/slm.md: New.
15377
15378 * config/i386/driver-i386.c (host_detect_local_cpu): Check movbe.
15379
f43245d1 15380 * config/i386/i386-c.c (ix86_target_macros_internal): New case
0b871ccf
YR		 15381 PROCESSOR_SLM.
15382 (ix86_target_macros_internal): Likewise.
15383
f43245d1 15384 * config/i386/i386.c (slm_cost): New cost.
0b871ccf
YR		 15385 (m_SLM): New macro flag.
15386 (initial_ix86_tune_features): Set m_SLM.
15387 (x86_accumulate_outgoing_args): Likewise.
15388 (x86_arch_always_fancy_math_387): Likewise.
15389 (processor_target_table): Add slm cost.
15390 (cpu_names): Add slm cpu name.
15391 (x86_option_override_internal): Set SLM ISA.
15392 (ix86_issue_rate): New case PROCESSOR_SLM.
15393 (ia32_multipass_dfa_lookahead): Likewise.
15394 (fold_builtin_cpu): Add slm.
15395
15396 * config/i386/i386.h (TARGET_SLM): New target macro.
15397 (target_cpu_default): Add TARGET_CPU_DEFAULT_slm.
15398 (processor_type): Add PROCESSOR_SLM.
15399
15400 * config/i386/i386.md (cpu): Add new value "slm".
15401 (slm.md): Include slm.md.
15402
24d5b097
XG		 154032013-05-30 Bernd Schmidt <bernds@codesourcery.com>
15404 Zhenqiang Chen <zhenqiang.chen@linaro.org>
15405
15406 * config/arm/arm-protos.h: Add and update function protos.
15407 * config/arm/arm.c (use_simple_return_p): New added.
15408 (thumb2_expand_return): Check simple_return flag.
15409 * config/arm/arm.md: Add simple_return and conditional simple_return.
15410 * config/arm/iterators.md: Add iterator for return and simple_return.
15411
c1cccc15
ZC		 154122013-05-30 Zhenqiang Chen <zhenqiang.chen@linaro.org>
15413
15414 * config/arm/arm.c (arm_add_cfa_adjust_cfa_note): New added.
15415 (arm_emit_multi_reg_pop): Add REG_CFA_ADJUST_CFA notes.
15416 (arm_emit_vfp_multi_reg_pop): Likewise.
15417 (thumb2_emit_ldrd_pop): Likewise.
15418 (arm_expand_epilogue): Add misc REG_CFA notes.
15419 (arm_unwind_emit): Skip REG_CFA_ADJUST_CFA and REG_CFA_RESTORE.
15420
3a4f280b
LC		 154212013-05-29 Lawrence Crowl <crowl@google.com>
15422
15423 * config/arm/t-arm: Update for below.
15424
15425 * config/arm/arm.c (arm_libcall_uses_aapcs_base::libcall_htab):
15426 Change type to hash_table. Update dependent calls and types.
15427
15428 * config/i386/t-cygming: Update for below.
15429
15430 * config/i386/t-interix: Update for below.
15431
15432 * config/i386/winnt.c (i386_pe_section_type_flags::htab):
15433 Change type to hash_table. Update dependent calls and types.
15434 (i386_find_on_wrapper_list::wrappers): Likewise.
15435
15436 * config/ia64/t-ia64: Update for below.
15437
15438 * config/ia64/ia64.c (bundle_state_table):
15439 Change type to hash_table. Update dependent calls and types.
15440
15441 * config/mips/mips.c (mips_reorg_process_insns::htab):
15442 Change type to hash_table. Update dependent calls and types.
15443
15444 * config/sol2.c (solaris_comdat_htab):
15445 Change type to hash_table. Update dependent calls and types.
15446
15447 * config/t-sol2: Update for above.
15448
1388a0e3
TJ		 154492013-05-29 Teresa Johnson <tejohnson@google.com>
15450
15451 * passes.c (dump_passes): Use FOR_EACH_FUNCTION since
15452 functions are not yet marked as defined.
15453
a5965b52 154542013-05-29 Michael Meissner <meissner@linux.vnet.ibm.com>
12211b99
UB		 15455 Pat Haugen <pthaugen@us.ibm.com>
15456 Peter Bergner <bergner@vnet.ibm.com>
a5965b52
MM		 15457
15458 * config/rs6000/vector.md (VEC_I): Add support for new power8 V2DI
15459 instructions.
15460 (VEC_A): Likewise.
15461 (VEC_C): Likewise.
15462 (vrotl<mode>3): Likewise.
15463 (vashl<mode>3): Likewise.
15464 (vlshr<mode>3): Likewise.
15465 (vashr<mode>3): Likewise.
15466
15467 * config/rs6000/rs6000-c.c (altivec_overloaded_builtins): Add
15468 support for power8 V2DI builtins.
15469
15470 * config/rs6000/rs6000-builtin.def (abs_v2di): Add support for
15471 power8 V2DI builtins.
15472 (vupkhsw): Likewise.
15473 (vupklsw): Likewise.
15474 (vaddudm): Likewise.
15475 (vminsd): Likewise.
15476 (vmaxsd): Likewise.
15477 (vminud): Likewise.
15478 (vmaxud): Likewise.
15479 (vpkudum): Likewise.
15480 (vpksdss): Likewise.
15481 (vpkudus): Likewise.
15482 (vpksdus): Likewise.
15483 (vrld): Likewise.
15484 (vsld): Likewise.
15485 (vsrd): Likewise.
15486 (vsrad): Likewise.
15487 (vsubudm): Likewise.
15488 (vcmpequd): Likewise.
15489 (vcmpgtsd): Likewise.
15490 (vcmpgtud): Likewise.
15491 (vcmpequd_p): Likewise.
15492 (vcmpgtsd_p): Likewise.
15493 (vcmpgtud_p): Likewise.
15494 (vupkhsw): Likewise.
15495 (vupklsw): Likewise.
15496 (vaddudm): Likewise.
15497 (vmaxsd): Likewise.
15498 (vmaxud): Likewise.
15499 (vminsd): Likewise.
15500 (vminud): Likewise.
15501 (vpksdss): Likewise.
15502 (vpksdus): Likewise.
15503 (vpkudum): Likewise.
15504 (vpkudus): Likewise.
15505 (vrld): Likewise.
15506 (vsld): Likewise.
15507 (vsrad): Likewise.
15508 (vsrd): Likewise.
15509 (vsubudm): Likewise.
15510
15511 * config/rs6000/rs6000.c (rs6000_init_hard_regno_mode_ok): Add
15512 support for power8 V2DI instructions.
15513
15514 * config/rs6000/altivec.md (UNSPEC_VPKUHUM): Add support for
15515 power8 V2DI instructions. Combine pack and unpack insns to use an
15516 iterator for each mode. Check whether a particular mode supports
15517 Altivec instructions instead of just checking TARGET_ALTIVEC.
15518 (UNSPEC_VPKUWUM): Likewise.
15519 (UNSPEC_VPKSHSS): Likewise.
15520 (UNSPEC_VPKSWSS): Likewise.
15521 (UNSPEC_VPKUHUS): Likewise.
15522 (UNSPEC_VPKSHUS): Likewise.
15523 (UNSPEC_VPKUWUS): Likewise.
15524 (UNSPEC_VPKSWUS): Likewise.
15525 (UNSPEC_VPACK_SIGN_SIGN_SAT): Likewise.
15526 (UNSPEC_VPACK_SIGN_UNS_SAT): Likewise.
15527 (UNSPEC_VPACK_UNS_UNS_SAT): Likewise.
15528 (UNSPEC_VPACK_UNS_UNS_MOD): Likewise.
15529 (UNSPEC_VUPKHSB): Likewise.
15530 (UNSPEC_VUNPACK_HI_SIGN): Likewise.
15531 (UNSPEC_VUNPACK_LO_SIGN): Likewise.
15532 (UNSPEC_VUPKHSH): Likewise.
15533 (UNSPEC_VUPKLSB): Likewise.
15534 (UNSPEC_VUPKLSH): Likewise.
15535 (VI2): Likewise.
15536 (VI_char): Likewise.
15537 (VI_scalar): Likewise.
15538 (VI_unit): Likewise.
15539 (VP): Likewise.
15540 (VP_small): Likewise.
15541 (VP_small_lc): Likewise.
15542 (VU_char): Likewise.
15543 (add<mode>3): Likewise.
15544 (altivec_vaddcuw): Likewise.
15545 (altivec_vaddu<VI_char>s): Likewise.
15546 (altivec_vadds<VI_char>s): Likewise.
15547 (sub<mode>3): Likewise.
15548 (altivec_vsubcuw): Likewise.
15549 (altivec_vsubu<VI_char>s): Likewise.
15550 (altivec_vsubs<VI_char>s): Likewise.
15551 (altivec_vavgs<VI_char>): Likewise.
15552 (altivec_vcmpbfp): Likewise.
15553 (altivec_eq<mode>): Likewise.
15554 (altivec_gt<mode>): Likewise.
15555 (altivec_gtu<mode>): Likewise.
15556 (umax<mode>3): Likewise.
15557 (smax<mode>3): Likewise.
15558 (umin<mode>3): Likewise.
15559 (smin<mode>3): Likewise.
15560 (altivec_vpkuhum): Likewise.
15561 (altivec_vpkuwum): Likewise.
15562 (altivec_vpkshss): Likewise.
15563 (altivec_vpkswss): Likewise.
15564 (altivec_vpkuhus): Likewise.
15565 (altivec_vpkshus): Likewise.
15566 (altivec_vpkuwus): Likewise.
15567 (altivec_vpkswus): Likewise.
15568 (altivec_vpks<VI_char>ss): Likewise.
15569 (altivec_vpks<VI_char>us): Likewise.
15570 (altivec_vpku<VI_char>us): Likewise.
15571 (altivec_vpku<VI_char>um): Likewise.
15572 (altivec_vrl<VI_char>): Likewise.
15573 (altivec_vsl<VI_char>): Likewise.
15574 (altivec_vsr<VI_char>): Likewise.
15575 (altivec_vsra<VI_char>): Likewise.
15576 (altivec_vsldoi_<mode>): Likewise.
15577 (altivec_vupkhsb): Likewise.
15578 (altivec_vupkhs<VU_char>): Likewise.
15579 (altivec_vupkls<VU_char>): Likewise.
15580 (altivec_vupkhsh): Likewise.
15581 (altivec_vupklsb): Likewise.
15582 (altivec_vupklsh): Likewise.
15583 (altivec_vcmpequ<VI_char>_p): Likewise.
15584 (altivec_vcmpgts<VI_char>_p): Likewise.
15585 (altivec_vcmpgtu<VI_char>_p): Likewise.
15586 (abs<mode>2): Likewise.
15587 (vec_unpacks_hi_v16qi): Likewise.
15588 (vec_unpacks_hi_v8hi): Likewise.
15589 (vec_unpacks_lo_v16qi): Likewise.
15590 (vec_unpacks_hi_<VP_small_lc>): Likewise.
15591 (vec_unpacks_lo_v8hi): Likewise.
15592 (vec_unpacks_lo_<VP_small_lc>): Likewise.
15593 (vec_pack_trunc_v8h): Likewise.
15594 (vec_pack_trunc_v4si): Likewise.
15595 (vec_pack_trunc_<mode>): Likewise.
15596
15597 * config/rs6000/altivec.h (vec_vaddudm): Add defines for power8
15598 V2DI builtins.
15599 (vec_vmaxsd): Likewise.
15600 (vec_vmaxud): Likewise.
15601 (vec_vminsd): Likewise.
15602 (vec_vminud): Likewise.
15603 (vec_vpksdss): Likewise.
15604 (vec_vpksdus): Likewise.
15605 (vec_vpkudum): Likewise.
15606 (vec_vpkudus): Likewise.
15607 (vec_vrld): Likewise.
15608 (vec_vsld): Likewise.
15609 (vec_vsrad): Likewise.
15610 (vec_vsrd): Likewise.
15611 (vec_vsubudm): Likewise.
15612 (vec_vupkhsw): Likewise.
15613 (vec_vupklsw): Likewise.
15614
e70670cf
JH		 156152013-05-29 Jan Hubicka <jh@suse.cz>
15616
15617 * cgraph.h (symtab_node_base): Add definition, alias and analyzed
15618 flags; reorder rest of fields in more consistent way.
15619 (varpool_node): Remove analyzed, finalized and alias.
15620 (cgraph_ndoe): Likewise.
15621 (symtab_alias_ultimate_target): New function.
15622 (cgraph_function_node): Move offline.
15623 (cgraph_reset_node): Declare.
15624 (cgraph_comdat_can_be_unshared_p): Remove.
15625 (varpool_remove_initializer): Declare.
15626 (varpool_first_defined_variable, varpool_next_defined_variable
15627 cgraph_first_defined_function, cgraph_next_defined_function): Update.
15628 (cgraph_function_with_gimple_body_p): Update.
15629 (varpool_all_refs_explicit_p): Update.
15630 (symtab_alias_target): New function.
15631 (cgraph_alias_aliased_node, varpool_alias_aliased_node): Rename to ...
15632 (cgraph_alias_target, varpool_alias_target): .. this one; simplify.
c1e183a9
UB		 15633 (cgraph_function_or_thunk_node): Simplify using
15634 symtab_alias_ultimate_target.
e70670cf
JH		 15635 (varpool_variable_node): Likewise.
15636 * cgraph.c (cgraph_create_function_alias): Update.
15637 (cgraph_add_thunk): Update.
15638 (cgraph_remove_node): Update.
15639 (dump_cgraph_node): Do not dump removed flags.
15640 (cgraph_function_body_availability): Update.
15641 (cgraph_propagate_frequency): Update.
15642 (verify_cgraph_node): Check sanity of local flag.
15643 (cgraph_function_node): Move here from cgraph.h; revamp for
15644 cgraph_function_or_thunk_node.
15645 * lto-symtab.c (lto_varpool_replace_node): Update.
15646 (lto_symtab_resolve_can_prevail_p): Update.
15647 (lto_symtab_merge_cgraph_nodes): Update.
15648 * ipa-cp.c (determine_versionability, initialize_node_lattices,
15649 propagate_constants_accross_call, devirtualization_time_bonus,
15650 ipcp_propagate_stage): Update.
15651 * tree-emutls.c (create_emultls_var, ipa_lower_emutls): Update.
c1e183a9
UB		 15652 * ipa-inline-transform.c (clone_inlined_nodes,
15653 preserve_function_body_p): Update.
e70670cf
JH		 15654 * ipa-reference.c (propagate): Update.
15655 (write_node_summary_p): Update.
15656 * toplev.c (wrapup_global_declaration_2): Update.
15657 * cgraphunit.c (cgraph_analyze_function): Rename to ...
15658 (analyze_function) ... this one.
15659 (cgraph_process_new_functions): Update.
15660 (cgraph_reset_node): Export.
15661 (cgraph_finalize_function): Update.
15662 (cgraph_add_new_function): Update.
15663 (process_function_and_variable_attributes): Update.
15664 (varpool_finalize_decl): Update.
15665 (symbol_finalized): Remove.
15666 (symbol_finalized_and_needed): Rename to ...
15667 (symbol_defined_and_needed): ... update.
15668 (cgraph_analyze_functions): Update.
15669 (handle_alias_pairs): Update.
15670 (mark_functions_to_output): Update.
15671 (assemble_thunk): Update.
15672 (output_in_order): Update.
15673 (output_weakrefs): Update.
15674 (finalize_compilation_unit): Update.
15675 * lto-cgraph.c (reachable_from_other_partition_p, lto_output_node,
15676 lto_output_varpool_node, compute_ltrans_boundary, input_overwrite_node,
15677 input_node, input_varpool_node): Update.
15678 * dbxout.c (dbxout_expand_expr): Update.
15679 * cgraphclones.c (cgraph_clone_node): Update.
15680 (cgraph_copy_node_for_versioning): Update.
15681 (cgraph_materialize_clone): Update.
15682 (cgraph_materialize_all_clones): Update.
15683 * ipa-pure-const.c (analyze_function, pure_const_write_summary,
15684 propagate_pure_const, propagate_nothrow): Update.
15685 * lto-streamer-out.c (lto_output, write_symbol): Update.
15686 * ipa-utils.c (ipa_reverse_postorder): Update.
15687 * ipa-inline.c (can_inline_edge_p): Update.
15688 (update_caller_keys, ipa_inline): Update.
15689 * dwarf2out.c (reference_to_unused,
15690 premark_types_used_by_global_vars_helper): Update.
15691 * tree-eh.c (tree_could_trap_p): Update.
15692 * ipa-split.c (consider_split, execute_split_functions): Update.
15693 * ipa.c (cgraph_non_local_node_p_1, cgraph_local_node_p,
c1e183a9
UB		 15694 has_addr_references_p): Update; move ahead in file for better
15695 readability.
e70670cf
JH		 15696 (process_references): Simplify.
15697 (symtab_remove_unreachable_nodes): Update; cleanup way function/var
15698 bodies are removed.
15699 (cgraph_comdat_can_be_unshared_p): Make static.
15700 (cgraph_externally_visible_p): Update.
15701 (varpool_externally_visible_p): Update.
15702 (function_and_variable_visibility): Update.
15703 * trans-mem.c (get_cg_data, ipa_tm_mayenterirr_function,
15704 ipa_tm_mark_force_output_node): Update.
15705 * ipa-inline-analysis.c (dump_inline_summary, initialize_inline_failed,
15706 estimate_edge_devirt_benefit, inline_generate_summary,
15707 inline_write_summary): Update.
15708 * gimple-fold.c (can_refer_decl_in_current_unit_p): Update.
15709 * ipa-prop.c (ipa_compute_jump_functions): Update.
c1e183a9
UB		 15710 (ipa_print_node_params, ipa_prop_read_section,
15711 ipa_update_after_lto_read, read_replacements_section): Update.
e70670cf
JH		 15712 * varasm.c (mark_decl_referenced): Update.
15713 (assemble_alias, dump_tm_clone_pairs): Update.
15714 * tree-inline.c (copy_bb): Update.
15715 (estimate_num_insns, optimize_inline_calls, tree_function_versioning):
15716 Update.
15717 * symtab.c (dump_symtab_base): Print new flags.
15718 (verify_symtab_base): Verify new flags.
15719 (symtab_alias_ultimate_target): New function.
15720 * tree-ssa-structalias.c (get_constraint_for_ssa_var,
15721 create_variable_info_for, associate_varinfo_to_alias, ipa_pta_execute):
15722 Update.
c1e183a9
UB		 15723 * passes.c (ipa_write_summaries, ipa_write_optimization_summaries):
15724 Update.
e70670cf
JH		 15725 * i386.c (ix86_get_function_versions_dispatcher,
15726 ix86_generate_version_dispatcher_body): Update.
15727 (fold_builtin_cpu): Use varpool_add_new_variable.
15728 * varpool.c (varpool_remove_initializer): Break out from ...
15729 (varpool_remove_node): ... this one.
15730 (dump_varpool_node, varpool_node_for_asm,
15731 cgraph_variable_initializer_availability, varpool_analyze_node,
15732 varpool_assemble_decl, varpool_remove_unreferenced_decls,
c1e183a9
UB		 15733 varpool_finalize_named_section_flags, varpool_create_variable_alias):
15734 Update.
e70670cf 15735
182802ad
JH		 157362013-05-29 Jan Hubicka <jh@suse.cz>
15737
15738 * passes.c (init_optimization_passes): Move OMP expansion into lowering.
15739
8190b609
ER		 157402013-05-29 Easwaran Raman <eraman@google.com>
15741
15742 PR tree-optimization/57442
15743 * tree-ssa-reassoc.c (appears_later_in_bb): Return correct value
15744 when control exits the main loop.
15745
69f5aa9b
SKS		 157462013-05-29 Sandeep Kumar Singh <Sandeep.Singh2@kpitcummins.com>
15747
15748 * rx/rx.h (TARGET_CPU_CPP_BUILTINS): Add macros for RX100, RX200,
15749 and RX600.
c1e183a9 15750 * rx/rx.opt: Add macro for rx100 with string rx100 and value RX100.
69f5aa9b
SKS		 15751 * rx/rx-opts.h (rx_cpu_types): Add new cpu type rx100.
15752 * rx/t-rx: Add rx100 under multi library matches option for nofpu
15753 option.
15754
4b847da9
BS		 157552013-05-29 Bill Schmidt <wschmidt@linux.vnet.ibm.com>
15756
15757 PR tree-optimization/57441
15758 * gimple-ssa-strength-reduction.c (analyze_candidates_and_replace):
15759 Don't limit size of incr_vec to number of candidates.
15760
4fc43c39
SE		 157612013-05-29 Steve Ellcey <sellcey@imgtec.com>
15762
15763 * config/mips/mti-linux.h (SYSROOT_SUFFIX_SPEC): Add micromips
15764 and mips16 directories.
c1e183a9 15765 * config/mips/t-mti-linux (MULTILIB_OPTIONS): Add micromips and mips16.
4fc43c39
SE		 15766 (MULTILIB_DIRNAMES): Ditto.
15767 (MULTILIB_EXCEPTIONS): Add new exceptions.
15768 * config/mips/t-mti-elf (MULTILIB_OPTIONS): Add micromips.
15769 (MULTILIB_DIRNAMES): Ditto.
15770 (MULTILIB_EXCEPTIONS): Add new exceptions.
15771
12211b99 157722012-05-29 Chris Schlumberger-Socha <chris.schlumberger-socha@arm.com>
a5350ddc
CSS		 15773 Marcus Shawcroft <marcus.shawcroft@arm.com>
15774
15775 * config/aarch64/aarch64-protos.h (aarch64_symbol_type): Define
15776 SYMBOL_TINY_ABSOLUTE.
15777 * config/aarch64/aarch64.c (aarch64_load_symref_appropriately): Handle
15778 SYMBOL_TINY_ABSOLUTE.
15779 (aarch64_expand_mov_immediate): Likewise.
15780 (aarch64_classify_symbol): Likewise.
15781 (aarch64_mov_operand_p): Remove ATTRIBUTE_UNUSED.
15782 Permit SYMBOL_TINY_ABSOLUTE.
15783 * config/aarch64/predicates.md (aarch64_mov_operand): Permit CONST.
15784
12211b99 157852013-05-29 Chris Schlumberger-Socha <chris.schlumberger-socha@arm.com>
17f4d4bf
CSS		 15786 Marcus Shawcroft <marcus.shawcroft@arm.com>
15787
15788 * config/aarch64/aarch64.c (aarch64_classify_symbol): Remove comment.
15789 Refactor if/switch. Replace gcc_assert with if.
15790
c0186656
GG		 157912013-05-29 Ganesh Gopalasubramanian <Ganesh.Gopalasubramanian@amd.com>
15792
15793 * config/i386/i386.c (initial_ix86_tune_features): Enable
15794 FP Reassociation for AMD bdver1 and bdver2.
15795
d20188f3
MJ		 157962013-05-29 Martin Jambor <mjambor@suse.cz>
15797
15798 * tree-cfg.c (verify_expr): Verify that BIT_FIELD_REF, REALPART_EXPR
15799 and IMAGPART_EXPR do not occur within other handled_components.
15800
292cba13
RB		 158012013-05-29 Richard Biener <rguenther@suse.de>
15802
15803 * tree-vect-slp.c (vect_bb_slp_scalar_cost): Guard vinfo
15804 access on whether the use is in the BB we currently try to
15805 vectorize.
15806 (vect_bb_vectorization_profitable_p): Pass the BB we currently
15807 vectorize to vect_bb_slp_scalar_cost.
15808
6eddf228
RB		 158092013-05-29 Richard Biener <rguenther@suse.de>
15810
15811 * tree-vect-slp.c (vect_bb_slp_scalar_cost): New function
15812 computing scalar cost offsetted by stmts that are kept live
15813 by scalar uses.
15814 (vect_bb_vectorization_profitable_p): Use vect_bb_slp_scalar_cost
15815 for computation of scalar cost.
15816
7df36117
SE		 158172013-05-28 Steve Ellcey <sellcey@mips.com>
15818
15819 * config/mips/mips-cpus.def (mips32r2): Change processor type.
15820
36536d79
BI		 158212013-05-28 Balaji V. Iyer <balaji.v.iyer@intel.com>
15822
15823 * doc/extend.texi (C Extensions): Added documentation about Cilk Plus
15824 array notation built-in reduction functions.
15825 * doc/passes.texi (Passes): Added documentation about changes done
15826 for Cilk Plus.
15827 * doc/invoke.texi (C Dialect Options): Added documentation about
15828 the -fcilkplus flag.
15829 * Makefile.in (C_COMMON_OBJS): Added c-family/array-notation-common.o.
15830 (BUILTINS_DEF): Depend on cilkplus.def.
15831 * builtins.def: Include cilkplus.def. Define DEF_CILKPLUS_BUILTIN.
15832 * builtin-types.def: Define BT_FN_INT_PTR_PTR_PTR.
15833 * cilkplus.def: New file.
15834
12211b99 158352013-05-28 Joern Rennecke <joern.rennecke@embecosm.com>
b49eefa5
JR		 15836
15837 PR rtl-optimization/57439
15838 * postreload.c (move2add_use_add2_insn): Use gen_lowpart_common.
15839
9daf6dbb
ER		 158402013-05-28 Easwaran Raman <eraman@google.com>
15841
15842 PR tree-optimization/57337
15843 * tree-ssa-reassoc.c (appears_later_in_bb): New function.
15844 (find_insert_point): Correctly identify the insertion point
15845 when two statements with the same UID is compared.
15846
fbd7e877
RB		 158472013-05-28 Richard Biener <rguenther@suse.de>
15848
15849 PR tree-optimization/56787
15850 * tree-vect-data-refs.c (vect_analyze_data_refs): Drop clobbers
15851 from the list of data references.
15852 * tree-vect-loop.c (vect_determine_vectorization_factor): Skip
15853 clobbers.
15854 (vect_analyze_loop_operations): Likewise.
15855 (vect_transform_loop): Remove clobbers.
15856
bbba1117
MJ		 158572013-05-28 Martin Jambor <mjambor@suse.cz>
15858
15859 * tree-cfg.c (verify_expr): Verify that BIT_FIELD_REFs, IMAGPART_EXPRs
15860 and REALPART_EXPRs have scalar type.
15861
bd388c2a
RB		 158622013-05-28 Richard Biener <rguenther@suse.de>
15863
15864 PR tree-optimization/57411
15865 * tree-ssa-copy.c (may_propagate_copy): Cannot propagate
15866 virtual operands.
15867 * tree-ssa-dom.c (eliminate_const_or_copy): Special-case
15868 virtual operand propagation.
15869
2f56a311
EB		 158702013-05-28 Eric Botcazou <ebotcazou@adacore.com>
15871
15872 * config/sparc/sparc.c (sparc_expand_vec_perm_bmask): Use %g0 as
15873 destination register for bmasksi_vis.
15874 (vector_init_bshuffle): Likewise.
15875 * config/sparc/sparc.md (vec_perm_constv8qi): Likewise.
15876
5c3eacbb
EB		 158772013-05-28 Eric Botcazou <ebotcazou@adacore.com>
15878
15879 * doc/invoke.texi (SPARC Options): Document -mfix-ut699.
15880 * builtins.c (expand_builtin_mathfn) <BUILT_IN_SQRT>: Try to widen the
15881 mode if the instruction isn't available in the original mode.
15882 * config/sparc/sparc.opt (mfix-ut699): New option.
15883 * config/sparc/sparc.md (muldf3_extend): Disable if -mfix-ut699.
15884 (divdf3): Turn into expander.
15885 (divdf3_nofix): New insn.
15886 (divdf3_fix): Likewise.
15887 (divsf3): Disable if -mfix-ut699.
15888 (sqrtdf2): Turn into expander.
15889 (sqrtdf2_nofix): New insn.
15890 (sqrtdf2_fix): Likewise.
15891 (sqrtsf2): Disable if -mfix-ut699.
15892
a1756c0a
RB		 158932013-05-27 Richard Biener <rguenther@suse.de>
15894
15895 PR middle-end/57412
15896 * omp-low.c (expand_omp_atomic_pipeline): Use the correct latch
15897 block for the new loop.
15898
5a892248
RB		 158992013-05-27 Richard Biener <rguenther@suse.de>
15900
15901 PR tree-optimization/57343
15902 * tree-ssa-loop-niter.c (number_of_iterations_ne_max): Do not
15903 use multiple_of_p if not TYPE_OVERFLOW_UNDEFINED.
15904 (number_of_iterations_cond): Do not build the folded tree.
15905
d1de852b
RB		 159062013-05-27 Richard Biener <rguenther@suse.de>
15907
15908 Revert
15909 PR middle-end/57381
15910 * fold-const.c (operand_equal_p): Compare FIELD_DECLs with
15911 OEP_CONSTANT_ADDRESS_OF retained.
15912
15913 PR tree-optimization/57417
15914 * tree-ssa-sccvn.c (vn_reference_fold_indirect): Fix test
15915 for unchanged base.
15916 (set_ssa_val_to): Compare addresses using
15917 get_addr_base_and_unit_offset.
15918
12211b99 159192013-05-27 Joern Rennecke <joern.rennecke@embecosm.com>
7894bc6b
JR		 15920
15921 PR rtl-optimization/56833
15922 * postreload.c (move2add_record_mode): New function.
15923 (move2add_record_sym_value, move2add_valid_value_p): Likewise.
15924 (move2add_use_add2_insn): Use move2add_record_sym_value.
15925 (move2add_use_add3_insn): Likewise.
15926 (reload_cse_move2add): Use move2add_valid_value_p and
15927 move2add_record_mode. Invalidate call-clobbered and REG_INC
15928 affected regs by setting reg_mode to VOIDmode.
15929 (move2add_note_store): Don't pretend the inside of a SUBREG is
15930 the actual destination. Invalidate single/leading registers by
15931 setting reg_mode to VOIDmode.
15932 Use move2add_record_sym_value, move2add_valid_value_p and
15933 move2add_record_mode.
15934
b03be25f
RB		 159352013-05-27 Richard Biener <rguenther@suse.de>
15936
15937 PR tree-optimization/57396
15938 * tree-affine.c (double_int_constant_multiple_p): Properly
15939 return false for val == 0 and div != 0.
15940
44e88db2
RS		 159412013-05-25 Richard Sandiford <rdsandiford@googlemail.com>
15942
15943 * config/mips/mips.h: Use #elif in preprocessor conditions.
15944
3b859704
RS		 159452013-05-25 Richard Sandiford <rdsandiford@googlemail.com>
15946
15947 PR target/53916
15948 * config/mips/constraints.md (kl): New constraint.
15949 * config/mips/mips.md (divmod<mode>4, udivmod<mode>4): Delete.
15950 (divmod<mode>4_internal): Rename to divmod<mode>4. Use "kl" as the
15951 constraint for operand 0. Split after CSE for MIPS16. Emit a move
15952 from LO for MIPS16.
15953 (udivmod<mode>4_internal): Likewise udivmod<mode>4.
15954
c979d5f5
RS		 159552013-05-25 Richard Sandiford <rdsandiford@googlemail.com>
15956
15957 PR target/55777
15958 * config/mips/mips.c (mips_can_inline_p): New function.
15959 (TARGET_CAN_INLINE_P): Define.
15960
8e90de43
SB		 159612013-05-25 Steven Bosscher <steven@gcc.gnu.org>
15962
15963 * sched-int.h (ds_t, dw_t): Make unsigned int.
15964 Fix documentation that describes how all the ds_t bits are used.
15965 Reserve the last bit for delayed-branch scheduling.
15966 (BITS_PER_DEP_STATUS): Move to ds_t typedef.
15967 (BITS_PER_DEP_WEAK): Fix definition and documentation.
15968 (gen_dep_weak_1): Remove prototype.
15969 * sched-deps.c (get_dep_weak_1): Make static.
15970 * target.def (speculate_insn, needs_block_p, gen_spec_check,
15971 get_insn_spec_ds, get_insn_checked_ds): Adjust hook prototypes.
15972 * doc/tm.texi: Regenerate.
15973 * config/ia64/ia64.c (ia64_needs_block_p): Update prototype.
15974
cb5cb194
SB		 159752013-05-24 Steven Bosscher <steven@gcc.gnu.org>
15976
15977 PR debug/56950
15978 * haifa-sched.c (sched_extend_bb): Ignore DEBUG_INSNs.
15979
5bd93ff6
NS		 159802013-05-24 Nathan Sidwell <nathan@codesourcery.com>
15981 Sandra Loosemore <sandra@codesourcery.com>
15982
15983 * config.gcc (powerpc-*): Allow native for with-cpu.
15984
2343af65
JL		 159852013-05-24 Jeff Law <law@redhat.com>
15986
15987 PR tree-optimization/57124
15988 * tree-vrp.c (simplify_cond_using_ranges): Only simplify a
15989 conversion feeding a condition if the range has an overflow
15990 if -fstrict-overflow. Add warnings for when we do make the
15991 transformation.
15992
3ad695b9
DC		 159932013-05-24 Dehao Chen <dehao@google.com>
15994
04960246 15995 * tree-cfg.c (locus_discrim_map): Fix the typo.
3ad695b9
DC		 15996 (locus_discrim_hasher): Likewise.
15997 (locus_discrim_hasher::hash): Likewise.
15998 (locus_discrim_hasher::equal): Likewise.
15999
82338059
MJ		 160002013-05-24 Martin Jambor <mjambor@suse.cz>
16001
16002 PR tree-optimization/57294
16003 * cgraph.h (ipa_record_stmt_references): Declare.
16004 * cgraphbuild.c (ipa_record_stmt_references): New function.
16005 (build_cgraph_edges): Use ipa_record_stmt_references.
16006 (rebuild_cgraph_edges): Likewise.
16007 (cgraph_rebuild_references): Likewise.
16008 * ipa-prop.c (ipa_modify_call_arguments): Discard references
16009 associated with the old statement and build references from the
16010 newly built statements.
16011 * ipa-ref.c (ipa_remove_stmt_references): New function.
16012 * ipa-ref.h (ipa_remove_stmt_references): Declare.
16013
1ccd4874
VM		 160142013-05-24 Vladimir Makarov <vmakarov@redhat.com>
16015
55805e54 16016 * lra-constraints.c (emit_spill_move): Use smaller mode for
1ccd4874 16017 mem-mem moves.
55805e54 16018 (check_and_process_move): Consider mem-reg moves for secondary
1ccd4874 16019 too.
55805e54 16020 (curr_insn_transform): Don't lose insns emitted before for
1ccd4874
VM		 16021 secondary memory moves.
16022 (inherit_in_ebb): Mark defined reg. Add usage only if it is not a
16023 reg set up in the current insn.
16024
25e25c73
DC		 160252013-05-24 Dehao Chen <dehao@google.com>
16026
04960246 16027 * tree-cfg.c (locus_descrim_hasher::hash): Change discriminator
25e25c73
DC		 16028 hash function.
16029 (locus_descrim_hasher::equal): Likewise.
04960246 16030 (build_gimple_cfg): New discriminator assignment algorithm.
25e25c73
DC		 16031 (make_edges): Likewise.
16032 (next_discriminator_for_locus): Likewise.
16033 (same_line_p): Likewise.
16034 (assign_discriminators): Likewise.
16035 (make_cond_expr_edges): Likewise.
16036 (make_gimple_switch_edges): Likewise.
16037 (make_goto_expr_edges): Likewise.
16038 (make_gimple_asm_edges): Likewise.
16039
50d38551
IB		 160402013-05-24 Ian Bolton <ian.bolton@arm.com>
16041
16042 * config/aarch64/aarch64.c (aarch64_print_operand): Change the
16043 X format specifier to only display bottom 16 bits.
16044 * config/aarch64/aarch64.md (insv_imm<mode>): Allow any size of
16045 immediate to match for operand 2, since it will be masked.
16046
aea0101d
RB		 160472013-05-24 Richard Biener <rguenther@suse.de>
16048
16049 PR tree-optimization/57287
16050 * tree-ssa-uninit.c (compute_uninit_opnds_pos): Disregard
16051 all SSA names that occur in abnormal PHIs.
16052
634e03d3
AI		 160532013-05-24 Alexander Ivchenko <alexander.ivchenko@intel.com>
16054
16055 PR tree-ssa/57385
16056 * tree-ssa-sccvn.c (fully_constant_vn_reference_p): Check
16057 that index is not negative.
16058
b17c024f
EB		 160592013-05-24 Eric Botcazou <ebotcazou@adacore.com>
16060
16061 PR rtl-optimization/55177
16062 * simplify-rtx.c (simplify_unary_operation_1) <NOT>: Deal with BSWAP.
16063 (simplify_byte_swapping_operation): New.
16064 (simplify_binary_operation_1): Call it for AND, IOR and XOR.
16065 (simplify_relational_operation_1): Deal with BSWAP.
16066
46aeac1b
RH		 160672013-05-23 Richard Henderson <rth@redhat.com>
16068
16069 PR target/56742
16070 * config/i386/i386.c (ix86_seh_fixup_eh_fallthru): New.
16071 (ix86_reorg): Call it.
16072
70cc1536
UB		 160732013-05-23 Uros Bizjak <ubizjak@gmail.com>
16074
16075 PR target/57379
16076 * config/alpha/alpha.md (unspec): Add UNSPEC_XFLT_COMPARE.
16077 * config/alpha/alpha.c (alpha_emit_xfloating_compare): Construct
16078 REG_EQUAL note as UNSPEC_XFLT_COMPARE unspec.
16079
b3851501
CB		 160802013-05-23 Christian Bruel <christian.bruel@st.com>
16081
16082 PR debug/57351
16083 * config/arm/arm.c (arm_dwarf_register_span): Do not use dbx number.
16084
12211b99 160852013-05-23 Chris Schlumberger-Socha <chris.schlumberger-socha@arm.com>
4e194912
CSS		 16086 Marcus Shawcroft <marcus.shawcroft@arm.com>
16087
16088 * config/aarch64/aarch64.md (*movdi_aarch64): Replace Usa with S.
16089 * config/aarch64/constraints.md (Usa): Remove.
16090 * doc/md.texi (AArch64 Usa): Remove.
16091
12211b99 160922013-05-23 Chris Schlumberger-Socha <chris.schlumberger-socha@arm.com>
83f8c414
CSS		 16093 Marcus Shawcroft <marcus.shawcroft@arm.com>
16094
16095 * config/aarch64/aarch64-protos.h (aarch64_mov_operand_p): Define.
16096 * config/aarch64/aarch64.c (aarch64_mov_operand_p): Define.
16097 * config/aarch64/predicates.md (aarch64_const_address): Remove.
16098 (aarch64_mov_operand): Use aarch64_mov_operand_p.
16099
12211b99 161002013-05-23 Vidya Praveen <vidyapraveen@arm.com>
b5574232
VP		 16101
16102 * config/aarch64/aarch64-simd.md (clzv4si2): Support for CLZ
16103 instruction (AdvSIMD).
16104 * config/aarch64/aarch64-builtins.c
16105 (aarch64_builtin_vectorized_function): Handler for BUILT_IN_CLZ.
16106 * config/aarch64/aarch-simd-builtins.def: Entry for CLZ.
16107
76c36cb1 161082013-05-23 Martin Jambor <mjambor@suse.cz>
7d2fb524
MJ		 16109
16110 PR middle-end/57347
16111 * tree.h (contains_bitfld_component_ref_p): Declare.
16112 * tree-sra.c (contains_bitfld_comp_ref_p): Move...
70cc1536
UB		 16113 * tree.c (contains_bitfld_component_ref_p): ...here. Adjust its
16114 caller.
7d2fb524
MJ		 16115 * ipa-prop.c (determine_known_aggregate_parts): Check that LHS does
16116 not access a bit-field. Assert all final offsets are byte-aligned.
16117
ce521ae6
RB		 161182013-05-23 Richard Biener <rguenther@suse.de>
16119
16120 PR tree-optimization/57380
16121 * tree-ssa-phiprop.c (propagate_with_phi): Do not require at
16122 least one invariant or re-used load.
16123 * passes.c (init_optimization_passes): Move pass_phiprop before
16124 pass_forwprop.
16125
75c7257f
JG		 161262013-05-23 James Greenhalgh <james.greenhalgh@arm.com>
16127
16128 * config/aarch64/aarch64-simd.md
16129 (aarch64_cm<optab>di): Add clobber of CC_REGNUM to unsplit pattern.
16130
af6d44b5
RB		 161312013-05-23 Richard Biener <rguenther@suse.de>
16132
16133 PR middle-end/57381
16134 * fold-const.c (operand_equal_p): Compare FIELD_DECLs with
16135 OEP_CONSTANT_ADDRESS_OF retained.
16136
bd3647bf
JJ		 161372013-05-23 Jakub Jelinek <jakub@redhat.com>
16138
16139 PR middle-end/57344
70cc1536
UB		 16140 * expmed.c (store_split_bit_field): If op0 is a REG or SUBREG of a REG,
16141 don't lower unit. Handle unit not being always BITS_PER_WORD.
bd3647bf 16142
a55757ea
RB		 161432013-05-23 Richard Biener <rguenther@suse.de>
16144
16145 PR rtl-optimization/57341
16146 * ira.c (validate_equiv_mem_from_store): Use anti_dependence
16147 instead of true_dependence.
16148
cfb00b41
DM		 161492013-05-22 David Malcolm <dmalcolm@redhat.com>
16150
dd1c676f
DM		 16151 * bb-reorder.c (branch_threshold): Make const.
16152 (exec_threshold): Ditto.
cfb00b41 16153
f62511da 161542013-05-22 Michael Meissner <meissner@linux.vnet.ibm.com>
12211b99
UB		 16155 Pat Haugen <pthaugen@us.ibm.com>
16156 Peter Bergner <bergner@vnet.ibm.com>
f62511da
MM		 16157
16158 * doc/extend.texi (PowerPC AltiVec/VSX Built-in Functions): Add
16159 documentation for the power8 crypto builtins.
16160
16161 * config/rs6000/t-rs6000 (MD_INCLUDES): Add crypto.md.
16162
16163 * config/rs6000/rs6000-builtin.def (BU_P8V_AV_1): Add support
16164 macros for defining power8 builtin functions.
16165 (BU_P8V_AV_2): Likewise.
16166 (BU_P8V_AV_P): Likewise.
16167 (BU_P8V_VSX_1): Likewise.
16168 (BU_P8V_OVERLOAD_1): Likewise.
16169 (BU_P8V_OVERLOAD_2): Likewise.
16170 (BU_CRYPTO_1): Likewise.
16171 (BU_CRYPTO_2): Likewise.
16172 (BU_CRYPTO_3): Likewise.
16173 (BU_CRYPTO_OVERLOAD_1): Likewise.
16174 (BU_CRYPTO_OVERLOAD_2): Likewise.
16175 (XSCVSPDP): Fix typo, point to the correct instruction.
16176 (VCIPHER): Add power8 crypto builtins.
16177 (VCIPHERLAST): Likewise.
16178 (VNCIPHER): Likewise.
16179 (VNCIPHERLAST): Likewise.
16180 (VPMSUMB): Likewise.
16181 (VPMSUMH): Likewise.
16182 (VPMSUMW): Likewise.
16183 (VPERMXOR_V2DI): Likewise.
16184 (VPERMXOR_V4SI: Likewise.
16185 (VPERMXOR_V8HI: Likewise.
16186 (VPERMXOR_V16QI: Likewise.
16187 (VSHASIGMAW): Likewise.
16188 (VSHASIGMAD): Likewise.
16189 (VPMSUM): Likewise.
16190 (VPERMXOR): Likewise.
16191 (VSHASIGMA): Likewise.
16192
16193 * config/rs6000/rs6000-c.c (rs6000_target_modify_macros): Define
16194 __CRYPTO__ if the crypto instructions are available.
16195 (altivec_overloaded_builtins): Add support for overloaded power8
16196 builtins.
16197
16198 * config/rs6000/rs6000.c (rs6000_expand_ternop_builtin): Add
16199 support for power8 crypto builtins.
16200 (builtin_function_type): Likewise.
16201 (altivec_init_builtins): Add support for builtins that take vector
16202 long long (V2DI) arguments.
16203
16204 * config/rs6000/crypto.md: New file, define power8 crypto
16205 instructions.
16206
162072013-05-22 Michael Meissner <meissner@linux.vnet.ibm.com>
12211b99
UB		 16208 Pat Haugen <pthaugen@us.ibm.com>
16209 Peter Bergner <bergner@vnet.ibm.com>
f62511da
MM		 16210
16211 * doc/invoke.texi (Option Summary): Add power8 options.
16212 (RS/6000 and PowerPC Options): Likewise.
16213
16214 * doc/md.texi (PowerPC and IBM RS6000 constraints): Update to use
16215 constraints.md instead of rs6000.h. Reorder w* constraints. Add
16216 wm, wn, wr documentation.
16217
f43245d1 16218 * config/rs6000/constraints.md (wm): New constraint for VSX
f62511da
MM		 16219 registers if direct move instructions are enabled.
16220 (wn): New constraint for no registers.
16221 (wq): New constraint for quad word even GPR registers.
16222 (wr): New constraint if 64-bit instructions are enabled.
16223 (wv): New constraint if power8 vector instructions are enabled.
16224 (wQ): New constraint for quad word memory locations.
16225
f43245d1 16226 * config/rs6000/predicates.md (const_0_to_15_operand): New
f62511da
MM		 16227 constraint for 0..15 for crypto instructions.
16228 (gpc_reg_operand): If VSX allow registers in VSX registers as well
16229 as GPR and floating point registers.
16230 (int_reg_operand): New predicate to match only GPR registers.
16231 (base_reg_operand): New predicate to match base registers.
16232 (quad_int_reg_operand): New predicate to match even GPR registers
16233 for quad memory operations.
16234 (vsx_reg_or_cint_operand): New predicate to allow vector logical
16235 operations in both GPR and VSX registers.
16236 (quad_memory_operand): New predicate for quad memory operations.
16237 (reg_or_indexed_operand): New predicate for direct move support.
16238
f43245d1 16239 * config/rs6000/rs6000-cpus.def (ISA_2_5_MASKS_EMBEDDED):
f62511da
MM		 16240 Inherit from ISA_2_4_MASKS, not ISA_2_2_MASKS.
16241 (ISA_2_7_MASKS_SERVER): New mask for ISA 2.07 (i.e. power8).
16242 (POWERPC_MASKS): Add power8 options.
16243 (power8 cpu): Use ISA_2_7_MASKS_SERVER instead of specifying the
16244 various options.
16245
f43245d1 16246 * config/rs6000/rs6000-c.c (rs6000_target_modify_macros):
f62511da
MM		 16247 Define _ARCH_PWR8 and __POWER8_VECTOR__ for power8.
16248
f43245d1 16249 * config/rs6000/rs6000.opt (-mvsx-timode): Add documentation.
f62511da
MM		 16250 (-mpower8-fusion): New power8 options.
16251 (-mpower8-fusion-sign): Likewise.
16252 (-mpower8-vector): Likewise.
16253 (-mcrypto): Likewise.
16254 (-mdirect-move): Likewise.
16255 (-mquad-memory): Likewise.
16256
f43245d1 16257 * config/rs6000/rs6000.c (power8_cost): Initial definition for power8.
f62511da
MM		 16258 (rs6000_hard_regno_mode_ok): Make PTImode only match even GPR
16259 registers.
70cc1536 16260 (rs6000_debug_reg_print): Print the base register class if -mdebug=reg.
f62511da
MM		 16261 (rs6000_debug_vector_unit): Add p8_vector.
16262 (rs6000_debug_reg_global): If -mdebug=reg, print power8 constraint
16263 definitions. Also print fusion state.
16264 (rs6000_init_hard_regno_mode_ok): Set up power8 constraints.
16265 (rs6000_builtin_mask_calculate): Add power8 builtin support.
16266 (rs6000_option_override_internal): Add support for power8.
16267 (rs6000_common_init_builtins): Add debugging for skipped builtins
16268 if -mdebug=builtin.
16269 (rs6000_adjust_cost): Add power8 support.
16270 (rs6000_issue_rate): Likewise.
16271 (insn_must_be_first_in_group): Likewise.
16272 (insn_must_be_last_in_group): Likewise.
16273 (force_new_group): Likewise.
16274 (rs6000_register_move_cost): Likewise.
16275 (rs6000_opt_masks): Likewise.
16276
16277 * config/rs6000/rs6000.h (ASM_CPU_POWER8_SPEC): If we don't have a
16278 power8 capable assembler, default to power7 options.
16279 (TARGET_DIRECT_MOVE): Likewise.
16280 (TARGET_CRYPTO): Likewise.
16281 (TARGET_P8_VECTOR): Likewise.
16282 (VECTOR_UNIT_P8_VECTOR_P): Define power8 vector support.
16283 (VECTOR_UNIT_VSX_OR_P8_VECTOR_P): Likewise.
16284 (VECTOR_MEM_P8_VECTOR_P): Likewise.
16285 (VECTOR_MEM_VSX_OR_P8_VECTOR_P): Likewise.
16286 (VECTOR_MEM_ALTIVEC_OR_VSX_P): Likewise.
16287 (TARGET_XSCVDPSPN): Likewise.
16288 (TARGET_XSCVSPDPN): Likewsie.
16289 (TARGET_SYNC_HI_QI): Likewise.
16290 (TARGET_SYNC_TI): Likewise.
16291 (MASK_CRYPTO): Likewise.
16292 (MASK_DIRECT_MOVE): Likewise.
16293 (MASK_P8_FUSION): Likewise.
16294 (MASK_P8_VECTOR): Likewise.
70cc1536
UB		 16295 (REG_ALLOC_ORDER): Move fr13 to be lower in priority so that the TFmode
16296 temporary used by some of the direct move instructions to get two FP
16297 temporary registers does not force creation of a stack frame.
f62511da
MM		 16298 (VLOGICAL_REGNO_P): Allow vector logical operations in GPRs.
16299 (MODES_TIEABLE_P): Move the VSX tests above the Altivec tests so
16300 that any VSX registers are tieable, even if they are also an
16301 Altivec vector mode.
16302 (r6000_reg_class_enum): Add wm, wr, wv constraints.
16303 (RS6000_BTM_P8_VECTOR): Power8 builtin support.
16304 (RS6000_BTM_CRYPTO): Likewise.
16305 (RS6000_BTM_COMMON): Likewise.
16306
16307 * config/rs6000/rs6000.md (cpu attribute): Add power8.
16308 * config/rs6000/rs6000-opts.h (PROCESSOR_POWER8): Likewise.
16309 (enum rs6000_vector): Add power8 vector support.
16310
73a1a707
RR		 163112013-05-22 Ramana Radhakrishnan <ramana.radhakrishnan@arm.com>
16312
16313 PR target/19599
16314 PR target/57340
16315 * config/arm/arm.c (any_sibcall_uses_r3): Rename to ..
16316 (any_sibcall_could_use_r3): this and handle indirect calls.
16317 (arm_get_frame_offsets): Rename use of any_sibcall_uses_r3.
16318
d28073d4
BS		 163192013-05-22 Bill Schmidt <wschmidt@linux.vnet.ibm.com>
16320
16321 * config/rs6000/rs6000.h (MALLOC_ABI_ALIGNMENT): New #define.
16322
bee0b10c
RB		 163232013-05-22 Richard Biener <rguenther@suse.de>
16324
16325 PR middle-end/57349
16326 * profile.c (branch_prob): Do not split blocks that are
16327 abnormally receiving from ECF_RETURNS_TWICE functions.
16328
98409b51
RS		 163292013-05-22 Richard Sandiford <rsandifo@linux.vnet.ibm.com>
16330
16331 * recog.c (offsettable_address_addr_space_p): Fix calculation of
16332 address mode. Move pointer mode initialization to the same place.
16333
c0602ab8
MZ		 163342013-05-22 Michael Zolotukhin <michael.v.zolotukhin@intel.com>
16335
16336 * read-rtl.c (copy_rtx_for_iterators): Continue applying iterators
16337 while it has any effect.
16338
4f7a634e
ER		 163392013-05-21 Easwaran Raman <eraman@google.com>
16340
16341 PR tree-optimization/57322
9daf6dbb
ER		 16342 * tree-ssa-reassoc.c (build_and_add_sum): If a BB is empty, set the
16343 UID of the statement added to the BB to be 1.
4f7a634e 16344
a4ce1258
JJ		 163452013-05-21 Jakub Jelinek <jakub@redhat.com>
16346
16347 PR tree-optimization/57331
70cc1536
UB		 16348 * tree-vrp.c (simplify_cond_using_ranges): Don't optimize comparison
16349 of conversion from pointer type to integral type with integer.
a4ce1258 16350
1b14621a
MJ		 163512013-05-21 Martin Jambor <mjambor@suse.cz>
16352
16353 PR lto/57289
16354 * ipa-prop.c (ipa_read_node_info): Process param_used and
16355 controlled_uses in the same order as when writing.
16356
e60661f0
MG		 163572013-05-21 Magnus Granberg <baldrick@free.fr>
16358
16359 PR plugins/56754
ee49aa34 16360 * Makefile.in (PLUGIN_HEADERS): Add $(TARGET_H).
e60661f0 16361
4a61be9a
RB		 163622013-05-21 Richard Biener <rguenther@suse.de>
16363
16364 PR tree-optimization/57318
16365 * tree-ssa-loop-ivcanon.c (tree_estimate_loop_size): Do not
16366 estimate stmts with side-effects as likely eliminated.
16367
c52da5f7
RB		 163682013-05-21 Richard Biener <rguenther@suse.de>
16369
16370 PR tree-optimization/57330
16371 * cgraph.c (cgraph_redirect_edge_call_stmt_to_callee): Properly
16372 preserve the call stmts fntype.
16373
7ec67e2a
RB		 163742013-05-21 Richard Biener <rguenther@suse.de>
16375
16376 PR tree-optimization/57303
16377 * tree-ssa-sink.c (statement_sink_location): Improve killing
16378 stmt detection and properly handle self-assignments.
16379
b112d513
CB		 163802013-05-21 Christian Bruel <christian.bruel@st.com>
16381
55805e54
YR		 16382 * dwarf2out.c (multiple_reg_loc_descriptor): Use dbx_reg_number for
16383 spanning registers. LEAF_REG_REMAP is supported only for contiguous
b112d513
CB		 16384 registers. Set register size out of the PARALLEL loop.
16385
14c2ec26
OE		 163862013-05-20 Oleg Endo <olegendo@gcc.gnu.org>
16387
16388 PR target/56547
16389 * config/sh/sh.md (fmasf4): Remove empty constraints strings.
16390 (*fmasf4, *fmasf4_media): New insns.
16391
da734fa1
RS		 163922013-05-19 Richard Sandiford <rdsandiford@googlemail.com>
16393
16394 * config/mips/mips.h (BASE_INSN_LENGTH, NOP_INSN_LENGTH): New macros.
16395 * config/mips/mips.c (mips_symbol_insns, mips_address_insns)
16396 (mips_const_insns, mips_split_const_insns, mips_load_store_insns)
16397 (mips_idiv_insns): Update the comments to say that the returned
16398 instruction counts are in units of BASE_INSN_LENGTH.
16399 (mips_adjust_insn_length): Multiply the mips_load_label_num_insns
16400 by BASE_INSN_LENGTH rather than 4. Add the jump separately,
16401 using 2 rather than 4 as the length of indirect MIPS16 and
16402 microMIPS jumps. Use NOP_INSN_LENGTH rather than 4 as the
16403 length of a NOP. Don't divide MIPS16 lengths by 2.
16404 (mips16_split_long_branches): Assume a branch is long if the
16405 length is greater than 4 rather than 8.
16406 * config/mips/mips.md (length): Give MIPS16 lengths directly,
16407 rather than multiplying them by 2. Multiply instruction counts
16408 by BASE_INSN_LENGTH rather than 4.
16409 (*jump_mips16, tls_get_tp_mips16_<mode>)
16410 (*tls_get_tp_mips16_call_<mode>): Divide lengths by 2.
16411
13719e8b
RS		 164122013-05-19 Richard Sandiford <rdsandiford@googlemail.com>
16413
16414 * config/mips/mips.md (extended_mips16): Remove branch case.
16415 (length): Remove duplicated extended_mips16 test.
16416
c3850d14
RS		 164172013-05-19 Richard Sandiford <rdsandiford@googlemail.com>
16418
16419 * config/mips/t-sde: Don't build 64-bit microMIPS multilibs.
16420
8da2e059
RS		 164212013-05-18 Richard Sandiford <rdsandiford@googlemail.com>
16422
16423 * recog.h (Recog_data): Rename to...
16424 (recog_data_d): ...this.
16425 (recog_data): Update accordingly.
16426 * recog.c (recog_data): Likewise.
16427 * reload.c (save_recog_data): Likewise.
16428 * config/picochip/picochip.c (picochip_saved_recog_data): Likewise.
16429 (picochip_save_recog_data, picochip_restore_recog_data): Likewise.
16430
2df013f3
JB		 164312013-05-17 Julian Brown <julian@codesourcery.com>
16432
16433 * gcse.c (compute_ld_motion_mems): If a non-simple MEM is
16434 found in a REG_EQUAL note, invalidate it.
16435
371e77e3 164362013-05-17 Easwaran Raman <eraman@google.com>
933f507d
ER		 16437
16438 * tree-ssa-reassoc.c (find_insert_point): New function.
16439 (insert_stmt_after): Likewise.
16440 (get_def_stmt): Likewise.
16441 (ensure_ops_are_available): Likewise.
16442 (not_dominated_by): Likewise.
16443 (rewrite_expr_tree): Do not move statements beyond what is
16444 necessary. Remove call to swap_ops_for_binary_stmt...
16445 (reassociate_bb): ... and move it here.
16446 (build_and_add_sum): Assign UIDs for new statements.
16447 (linearize_expr): Likewise.
16448 (do_reassoc): Renumber gimple statement UIDs.
16449
e01c7cca
JH		 164502013-05-17 Jan Hubicka <jh@suse.cz>
16451
16452 * lto-symtab.c (lto_symtab_merge_cgraph_nodes): Resolve cross module
16453 weakrefs.
16454 * cgraph.c (dump_cgraph_node): Do not ice on unresolved alias.
e2420687
UB		 16455 * cgraphunit.c (handle_alias_pairs): Store target of unresolved
16456 weakrefs.
e01c7cca
JH		 16457 (output_weakrefs): Update.
16458
c3272a92
PCC		 164592013-05-17 Po-Chun Chang <pchang9@cs.wisc.edu>
16460 Martin Jambor <mjambor@suse.cz>
16461
16462 PR middle-end/57276
16463 * ipa-cp.c (cgraph_edge_brings_all_agg_vals_for_node): Break when a
16464 value that corresponds to the given aggval is found in values vector.
16465
11c2aa39
UB		 164662013-05-17 Uros Bizjak <ubizjak@gmail.com>
16467
16468 * config/i386/driver-i386.c (host_detect_local_cpu): Pass mmx, 3dnow,
16469 sse, sse2, sse3, ssse3 and sse4a flags to options.
16470
abecc8c6
DM		 164712013-05-17 David Malcolm <dmalcolm@redhat.com>
16472
16473 * gengtype-state.c: (s_expr_writer): New class, to handle
16474 prettifying of output layout of s-expressions.
16475 (state_writer): New class, to write out gtype.state.
16476 (state_written_type_count): Move this variable into member data of
16477 state_writer.
16478 (s_expr_writer::s_expr_writer): New code: constructor for new class
16479 (state_writer::state_writer(): ditto
16480 (s_expr_writer::write_new_line): New function
16481 (s_expr_writer::write_any_indent): ditto
16482 (s_expr_writer::begin_s_expr): ditto
16483 (s_expr_writer::end_s_expr): ditto
16484 (write_state_fileloc): convert to method of state_writer...
16485 (state_writer:: write_state_fileloc): ...and use methods of
16486 s_expr_writer to write indentation into the gtype.state output file
16487 to visually represent the hierarchical structure of the list
16488 structures
16489 (write_state_fields): ditto, renaming to...
16490 (state_writer::write_state_fields)
16491 (write_state_a_string): ditto, renaming to...
16492 (state_writer::write_state_a_string)
16493 (write_state_string_option): ditto, renaming to...
16494 (state_writer::write_state_string_option)
16495 (write_state_type_option): ditto, renaming to...
16496 (state_writer::write_state_type_option)
16497 (write_state_nested_option): ditto, renaming to...
16498 (state_writer::write_state_nested_option)
16499 (write_state_option): ditto, renaming to...
16500 (state_writer::write_state_option)
16501 (write_state_options): ditto, renaming to...
16502 (state_writer::write_state_options)
16503 (write_state_lang_bitmap): ditto, renaming to...
16504 (state_writer::write_state_lang_bitmap)
16505 (write_state_version): ditto, renaming to...
16506 (state_writer::write_state_version)
16507 (write_state_scalar_type): ditto, renaming to...
16508 (state_writer::write_state_scalar_type)
16509 (write_state_string_type): ditto, renaming to...
16510 (state_writer::write_state_string_type)
16511 (write_state_undefined_type): ditto, renaming to...
16512 (state_writer::write_state_undefined_type)
16513 (write_state_struct_union_type): ditto, renaming to...
16514 (state_writer::write_state_struct_union_type)
16515 (write_state_struct_type): ditto, renaming to...
16516 (state_writer::write_state_struct_type)
16517 (write_state_user_struct_type): ditto, renaming to...
16518 (state_writer::write_state_user_struct_type)
16519 (write_state_lang_struct_type): ditto, renaming to...
16520 (state_writer::write_state_lang_struct_type)
16521 (write_state_param_struct_type): ditto, renaming to...
16522 (state_writer::write_state_param_struct_type)
16523 (write_state_pointer_type): ditto, renaming to...
16524 (state_writer::write_state_pointer_type)
16525 (write_state_array_type): ditto, renaming to...
16526 (state_writer::write_state_array_type)
16527 (write_state_gc_used): ditto, renaming to...
16528 (state_writer::write_state_gc_used)
16529 (write_state_common_type_content): ditto, renaming to...
16530 (state_writer::write_state_common_type_content)
16531 (write_state_type): ditto, renaming to...
16532 (state_writer::write_state_type)
16533 (write_state_pair_list): ditto, renaming to...
16534 (state_writer::write_state_pair_list)
16535 (write_state_pair): ditto, renaming to...
16536 (state_writer::write_state_pair)
16537 (write_state_typedefs): ditto, renaming to...
16538 (state_writer::write_state_typedefs)
16539 (write_state_structures): ditto, renaming to...
16540 (state_writer::write_state_structures)
16541 (write_state_param_structs): ditto, renaming to...
16542 (state_writer::write_state_param_structs)
16543 (write_state_variables): ditto, renaming to...
16544 (state_writer::write_state_variables)
16545 (write_state_srcdir): ditto, renaming to...
16546 (state_writer::write_state_srcdir)
16547 (write_state_files_list): ditto, renaming to...
16548 (state_writer::write_state_files_list)
16549 (write_state_languages): ditto, renaming to...
16550 (state_writer::write_state_languages)
16551 (write_state): create a state_writer instance and use it when
16552 writing out the state file
16553
d6545f29
MS		 165542013-05-17 Mike Stump <mikestump@comcast.net>
16555
816a3f73 16556 PR rtl-optimization/57304
d6545f29
MS		 16557 * web.c (union_match_dups): Ensure that DF_REF_LOC exists before
16558 accessing DF_REF_REAL_LOC.
16559
38c821cf
JJ		 165602013-05-17 Jakub Jelinek <jakub@redhat.com>
16561
16562 PR rtl-optimization/57281
16563 PR rtl-optimization/57300
16564 * config/i386/i386.md (extendsidi2_1 dead reg splitter): Remove.
16565 (extendsidi2_1 peephole2s): Add instead 2 new peephole2s, that undo
16566 what the other splitter did if the registers are dead.
16567
2a293391
RB		 165682013-05-17 Richard Biener <rguenther@suse.de>
16569
16570 * tree-ssa-alias.c (stmt_kills_ref_p_1): Properly compare
16571 MEM_REF offsets.
16572
17042d2b
JJ		 165732013-05-17 Jakub Jelinek <jakub@redhat.com>
16574
16575 * gcc.c (SANITIZER_SPEC): Reject -fsanitize=address -fsanitize=thread
16576 linking.
16577
5b115c1f
MP		 165782013-05-17 Marek Polacek <polacek@redhat.com>
16579
11c2aa39
UB		 16580 * tree-ssa-strlen.c (handle_char_store): Don't invalidate cached
16581 length when doing non-zero store of storing '\0' to '\0'.
5b115c1f 16582
52d84413
JJ		 165832013-05-17 Jakub Jelinek <jakub@redhat.com>
16584
68119618
JJ		 16585 * tree-vect-patterns.c (vect_recog_rotate_pattern): For
16586 vect_external_def oprnd1 with loop_vinfo, try to emit
16587 optional cast, negation and and stmts on the loop preheader
16588 edge instead of into the pattern def seq.
16589
52d84413
JJ		 16590 PR tree-optimization/57051
16591 * fold-const.c (const_binop) <case VEC_LSHIFT_EXPR,
16592 case VEC_RSHIFT_EXPR>: Fix BYTES_BIG_ENDIAN handling.
16593
17b962bd
NC		 165942013-05-16 Nick Clifton <nickc@redhat.com>
16595
16596 * config/rl78/rl78.c (rl78_attribute_table): Add naked.
16597 (rl78_is_naked_func): New function.
56aefbf7
UB		 16598 (rl78_expand_prologue): Skip prologue generation for naked functions.
16599 (rl78_expand_epilogue): Skip epilogue generation for naked functions.
17b962bd
NC		 16600 * doc/extend.texi (naked): Add RL78 to the list of processors
16601 that supports this attribute.
16602
b1a0f84e
JL		 166032013-05-16 Jeff Law <law@redhat.com>
16604
16605 * Makefile.in (tree-switch-conversion.o): Depend on $(OPTABS_H).
16606
af0e415b
UB		 166072013-05-16 Uros Bizjak <ubizjak@gmail.com>
16608
16609 * config/i386/driver-i386.c (host_detect_local_cpu): Determine
16610 cache parameters using detect_caches_amd also for CYRIX,
16611 NSC and TM2 signatures.
16612
19db293a
UB		 166132013-05-16 Uros Bizjak <ubizjak@gmail.com>
16614 Dzianis Kahanovich <mahatma@eu.by>
16615
16616 PR target/45359
16617 PR target/46396
16618 * config/i386/driver-i386.c (host_detect_local_cpu): Detect
16619 VIA/Centaur processors and determine their cache parameters
16620 using detect_caches_amd.
16621
251a41b9
TJ		 166222013-05-16 Teresa Johnson <tejohnson@google.com>
16623
16624 * cfgrtl.c (verify_hot_cold_block_grouping): Return err.
16625 (rtl_verify_edges): New function.
16626 (rtl_verify_bb_insns): Ditto.
16627 (rtl_verify_bb_pointers): Ditto.
16628 (rtl_verify_bb_insn_chain): Ditto.
16629 (rtl_verify_fallthru): Ditto.
16630 (rtl_verify_bb_layout): Ditto.
16631 (rtl_verify_flow_info_1): Outline checks into new functions.
16632 (rtl_verify_flow_info): Ditto.
16633
f14540b6
SE		 166342013-05-16 Steve Ellcey <sellcey@imgtec.com>
16635
16636 * cfghooks.c (copy_bbs): Add update_dominance argument.
16637 * cfghooks.h (copy_bbs): Update prototype.
16638 * tree-cfg.c (gimple_duplicate_sese_region):
16639 Add update_dominance argument.
16640 * tree-flow.h (gimple_duplicate_sese_region): Update prototype.
16641 * tree-ssa-loop-ch.c (copy_loop_headers): Update
16642 gimple_duplicate_sese_region call.
16643 * tree-vect-loop-manip.c (slpeel_tree_duplicate_loop_to_edge_cfg):
16644 Update copy_bbs call.
16645 * cfgloopmanip.c (duplicate_loop_to_header_edge): Ditto.
16646 * trans-mem.c (ipa_uninstrument_transaction): Ditto.
16647
7e9a3abb
JJ		 166482013-05-16 Jakub Jelinek <jakub@redhat.com>
16649
16650 * tree-vectorizer.h (NUM_PATTERNS): Increment.
16651 * tree-vect-patterns.c (vect_vect_recog_func_ptrs): Add
16652 vect_recog_rotate_pattern.
16653 (vect_recog_rotate_pattern): New function.
16654
427b248d
JM		 166552013-05-16 Jason Merrill <jason@redhat.com>
16656
16657 * Makefile.in (LLINKER): New variable.
16658 (mostlyclean): Remove link mutex.
16659 * configure.ac: Handle --enable-link-mutex.
16660 * lock-and-run.sh: New script.
16661
b871e3d2
RR		 166622013-05-16 Ramana Radhakrishnan <ramana.radhakrishnan@arm.com>
16663
16664 PR target/19599
16665 * config/arm/arm.c (arm_function_ok_for_sibcall): Add check
16666 for NULL decl.
16667
ce858126
RO		 166682013-05-16 Rainer Orth <ro@CeBiTec.Uni-Bielefeld.DE>
16669
16670 * reorg.c (link_cc0_insns): Wrap in #ifdef HAVE_cc0.
16671
798d3d04
GY		 166722013-05-16 Greta Yorsh <Greta.Yorsh@arm.com>
16673
16674 * config/arm/arm-protos.h (gen_movmem_ldrd_strd): New declaration.
16675 * config/arm/arm.c (next_consecutive_mem): New function.
16676 (gen_movmem_ldrd_strd): Likewise.
16677 * config/arm/arm.md (movmemqi): Update condition and code.
16678 (unaligned_loaddi, unaligned_storedi): New patterns.
16679
0baddc45
RO		 166802013-05-16 Rainer Orth <ro@CeBiTec.Uni-Bielefeld.DE>
16681
16682 * config.gcc: Obsolete *-*-solaris2.9*.
16683 * doc/install.texi (Specific, *-*-solaris2*): Document it.
16684
aa9d5bcf
RB		 166852013-05-16 Richard Biener <rguenther@suse.de>
16686
16687 * passes.c (init_optimization_passes): Move pass_parallelize_loops
16688 earlier, after GRAPHITE transforms and IV canonicalization.
16689
5a0f4dd3
JJ		 166902013-05-16 Jakub Jelinek <jakub@redhat.com>
16691
16692 * omp-low.c (extract_omp_for_data): For collapsed loops,
16693 if at least one of the loops is known at compile time to
16694 iterate zero times, set count to 0.
16695 (expand_omp_regimplify_p): New function.
16696 (expand_omp_for_generic): For collapsed loops, if at least
16697 one of the loops isn't known to iterate at least once,
16698 add runtime check with setting count to 0.
16699 (expand_omp_for_static_nochunk, expand_omp_for_static_chunk):
16700 For unsigned types if it isn't known at compile time that
16701 the loop will iterate at least once, add runtime check to bypass
16702 the whole loop if initial condition isn't true.
16703
e3753785
NS		 167042013-05-16 Nathan Sidwell <nathan@codesourcery.com>
16705
16706 * varasm.c (default_use_anchors_for_symbol_p): Use decl_replaceable_p.
16707
43bb4dd1
MG		 167082013-05-16 Marc Glisse <marc.glisse@inria.fr>
16709
16710 PR middle-end/57286
16711 * fold-const.c (fold_ternary_loc) <VEC_COND_EXPR>: Disable some
16712 transformations to avoid an infinite loop.
16713
3571dde6
MP		 167142013-05-16 Marek Polacek <polacek@redhat.com>
16715
16716 * tree-scalar-evolution.c (scev_const_prop): Add more dumps.
16717
b8b3f0ca
LE		 167182013-05-15 Leif Ekblad <leif@rdos.net>
16719
16720 * config/i386/i386.c (ix86_decompose_address): Use
16721 DEFAULT_TLS_SEG_REG to access TLS segment register.
16722 * config/i386/i386.h (DEFAULT_TLS_SEG_REG): New define.
16723 * config/i386/rdos.h (DEFAULT_TLS_SEG_REG): Ditto.
16724 (TARGET_TLS_DIRECT_SEG_REFS_DEFAULT): Ditto.
16725
e299a383
RS		 167262013-05-15 Richard Sandiford <rdsandiford@googlemail.com>
16727
16728 PR target/57260
16729 * config/mips/mips.c (mips_function_ok_for_sibcall): Don't allow
16730 sibling calls to functions that would normally be lazily bound,
16731 unless $gp is call-clobbered.
16732
e7208ea3
UB		 167332013-05-15 Uros Bizjak <ubizjak@gmail.com>
16734
19db293a 16735 * config/i386/i386.c (ix86_option_override_internal): Update
e7208ea3 16736 processor_alias_table for missing PTA_PRFCHW and PTA_FXSR flags. Add
5d280e2b
UB		 16737 PTA_POPCNT to corei7 entry. Do not enable SSE prefetch on
16738 non-SSE 3dNow! targets. Enable TARGET_PRFCHW for TARGET_3DNOW targets.
e7208ea3
UB		 16739 * config/i386/i386.md (prefetch): Enable for TARGET_PRFCHW instead
16740 of TARGET_3DNOW.
16741 (*prefetch_3dnow): Enable for TARGET_PRFCHW only.
16742
e7413f3d
AS		 167432013-05-15 Andreas Schwab <schwab@suse.de>
16744
16745 * config/m68k/m68k.md (*rotlhi3_lowpart, *rotlqi3_lowpart): Name
16746 for rotlhi3+1 and rotlqi3+1, resp. Fix reference to non-existing
16747 third operand.
16748
ad4db775
TJ		 167492013-05-15 Teresa Johnson <tejohnson@google.com>
16750
16751 * loop-unroll.c (report_unroll_peel): Check decision before
16752 emitting unroll/peel message.
16753
af205f67
TJ		 167542013-05-15 Teresa Johnson <tejohnson@google.com>
16755
16756 * function.h (has_bb_partition): New rtl_data flag.
16757 (bb_reorder_complete): Ditto.
16758 * cfgcleanup.c (try_crossjump_to_edge): Check for has_bb_partition
16759 instead of flag_reorder_blocks_and_partition.
16760 * cfgrtl.c (verify_hot_cold_block_grouping): Moved from bb-reorder.c,
16761 with some enhancements.
16762 (rtl_verify_flow_info_1): Call verify_hot_cold_block_grouping.
16763 * bb-reorder.c (connect_traces): Check for has_bb_partition
16764 instead of flag_reorder_blocks_and_partition.
16765 (verify_hot_cold_block_grouping): Moved to cfgrtl.c.
16766 (reorder_basic_blocks): Set bb_reorder_complete flag, remove call to
16767 verify_hot_cold_block_grouping.
16768 (partition_hot_cold_basic_blocks): Set has_bb_partition.
16769
9adcfa3c
RR		 167702013-05-15 Ramana Radhakrishnan <ramana.radhakrishnan@arm.com>
16771
16772 PR target/19599
16773 * config/arm/predicates.md (call_insn_operand): New predicate.
16774 * config/arm/constraints.md ("Cs", "Ss"): New constraints.
16775 * config/arm/arm.md (*call_insn, *call_value_insn): Match only
16776 if insn is not a tail call.
16777 (*sibcall_insn, *sibcall_value_insn): Adjust for tailcalling through
16778 registers.
16779 * config/arm/arm.h (enum reg_class): New caller save register class.
16780 (REG_CLASS_NAMES): Likewise.
16781 (REG_CLASS_CONTENTS): Likewise.
16782 * config/arm/arm.c (arm_function_ok_for_sibcall): Allow tailcalling
16783 without decls.
16784
ccb3ad87
RB		 167852013-05-15 Richard Biener <rguenther@suse.de>
16786
16787 * tree-vect-loop.c (vect_transform_loop): Use MSG_NOTE instead
16788 of MSG_OPTIMIZED_LOCATIONS.
16789 * tree-vect-slp.c (vect_make_slp_decision): Likewise.
16790 (vect_slp_transform_bb): Indicate location in MSG_OPTIMIZED_LOCATIONS
16791 message.
16792 * tree-vectorizer.c (vectorize_loops): Use MSG_NOTE instead
16793 of MSG_OPTIMIZED_LOCATIONS.
16794 (execute_vect_slp): Likewise.
16795 * tree-vect-loop-manip.c (vect_do_peeling_for_loop_bound): Likewise.
16796 (vect_create_cond_for_alias_checks): Likewise.
16797 * tree-vect-patterns.c (vect_recog_dot_prod_pattern): Likewise.
16798 (vect_recog_widen_mult_pattern): Likewise.
16799 (vect_recog_widen_sum_pattern): Likewise.
16800 (vect_recog_over_widening_pattern): Likewise.
16801 (vect_recog_widen_shift_pattern): Likewise.
16802 (vect_recog_vector_vector_shift_pattern): Likewise.
16803 (vect_recog_divmod_pattern): Likewise.
16804 (vect_recog_mixed_size_cond_pattern): Likewise.
16805 (vect_recog_bool_pattern): Likewise.
16806 (vect_pattern_recog_1): Likewise.
16807
48b1474e
MJ		 168082013-05-15 Martin Jambor <mjambor@suse.cz>
16809
16810 * ipa-prop.c (ipa_make_edge_direct_to_target): Redirect calls to
16811 non-functions to builtin_unreachable.
16812 * ipa-inline-transform.c (inline_call): Do not assert estimates were
16813 correct when new direct edges were discovered.
16814
9de04252
MJ		 168152013-05-15 Martin Jambor <mjambor@suse.cz>
16816
16817 * ipa-prop.c (ipa_print_node_jump_functions): Print symbol order in
16818 header, print symbol order instead of node uid, print more information
16819 about indirect edge targets.
16820 (ipa_make_edge_direct_to_target): Print symbol order instead of node
16821 uids.
16822 (ipa_make_edge_direct_to_target): Likewise.
16823 (remove_described_reference): Likewise.
16824 (propagate_controlled_uses): Likewise.
16825 (ipa_print_node_params): Also print symbol order.
16826 (ipcp_transform_function): Print symbol order instead of node uids.
16827 * cgraph.c (cgraph_redirect_edge_call_stmt_to_callee): Likewise.
16828 (cgraph_get_create_real_symbol_node): Likewise.
16829 * ipa-cp.c (print_lattice): Likewise.
16830 (print_all_lattices): Likewise.
16831 (determine_versionability): Likewise.
16832 (initialize_node_lattices): Likewise.
16833 (estimate_local_effects): Likewise.
16834 (update_profiling_info): Likewise.
16835 (create_specialized_node): Likewise.
16836 (perhaps_add_new_callers): Likewise.
16837 (decide_about_value): Likewise.
16838 (decide_whether_version_node): Likewise.
16839 (identify_dead_nodes): Likewise.
16840 * ipa-inline-analysis.c (dump_inline_edge_summary): Likewise.
16841 (dump_inline_summary): Likewise.
16842 (estimate_node_size_and_time): Likewise.
16843 (inline_analyze_function): Likewise.
16844 * ipa-inline.c (report_inline_failed_reason): Likewise.
16845 (want_early_inline_function_p): Likewise.
16846 (edge_badness): Likewise.
16847 (update_edge_key): Likewise.
16848 (inline_small_functions): Likewise. Add dumping of order to two other
16849 dumps.
16850 * ipa-pure-const.c (pure_const_read_summary): Print symbol order
16851 instead of node uids.
16852 (propagate_pure_const): Likewise.
16853 (propagate_pure_const): Likewise.
16854 * ipa-utils.c (dump_cgraph_node_set): Likewise.
16855 * lto-cgraph.c (input_node): Explicitly specify we dump uid.
16856 * lto-symtab.c (lto_cgraph_replace_node): Print symbol order instead
16857 of node uids.
16858 * tree-pretty-print.c (dump_function_header): Likewise.
16859 * tree-sra.c (convert_callers_for_node): Dump in traditional format.
16860 Print symbol order instead of node uids.
16861
1dd03b91
AK		 168622013-05-15 Andreas Krebbel <Andreas.Krebbel@de.ibm.com>
16863
16864 * config/s390/s390.c (s390_register_move_cost): Don't impose the
16865 FPR<->GPR move cost penalty if ldgr/lgdr can be used.
16866
fdf6a7b9
RB		 168672013-05-15 Richard Biener <rguenther@suse.de>
16868
16869 PR tree-optimization/57275
16870 * tree-vect-data-refs.c (vect_analyze_data_ref_dependence): Fix
16871 return value for fail to do runtime alias checks for gather loads.
16872
2d6e4603
JH		 168732013-05-15 Jan Hubicka <jh@suse.cz>
16874
16875 PR lto/57038
16876 PR lto/47375
e7208ea3
UB		 16877 * lto-symtab.c (lto_symtab_symbol_p): Add external symbol;
16878 weakrefs are not external.
16879 (lto_symtab_merge_decls): Fix thinko when dealing with
16880 non-lto_symtab decls.
2d6e4603
JH		 16881 (lto_symtab_merge_cgraph_nodes): Use lto_symtab_symbol_p.
16882 (lto_symtab_prevailing_decl): Get int sync with lto_symtab_symbol_p.
16883 * varpool.c (dump_varpool_node): Dump more flags.
16884
83f44b39
GG		 168852013-05-15 Ganesh Gopalasubramanian <Ganesh.Gopalasubramanian@amd.com>
16886
16887 * config/i386/i386.c (processor_alias_table): Add instruction
16888 FSGSBASE for AMD bdver3 architecture.
16889
2e55d062
JJ		 168902013-05-14 Jakub Jelinek <jakub@redhat.com>
16891
16892 * tree.c (warn_deprecated_use): Print file:line using locus color.
16893 * diagnostic.c (diagnostic_report_current_module): Print file:line
16894 and file:line:column using locus color.
16895
0bfdb81e
MS		 168962013-05-14 Mike Stump <mikestump@comcast.net>
16897
16898 * gdbinit.in: Add __null.
16899
a508ef22
MS		 169002013-05-14 Mike Stump <mikestump@comcast.net>
16901
16902 * recog.h: Rename struct recog_data to Recog_data.
16903 * recog.c: Likewise.
16904 * reload.c (can_reload_into): Likewise.
16905 * config/picochip/picochip.c: Likewise.
16906
e7180acb
MS		 169072013-05-14 Mike Stump <mikestump@comcast.net>
16908
16909 * web.c (union_match_dups): Also check DF_REF_REAL_LOC.
16910
61216c26
SB		 169112013-05-14 Steven Bosscher <steven@gcc.gnu.org>
16912
fd6beed4
SB		 16913 * resource.h (struct resources): Remove unch_memory member.
16914 (CLEAR_RESOURCE): Don't clear unch_memory.
16915 * resource.c (mark_referenced_resources): Don't set it.
16916 (mark_set_resources): Likewise.
16917 (mark_target_live_regs): Don't clear it.
16918 (init_resource_info): Likewise.
16919 * reorg.c (resource_conflicts_p): Don't compare it.
16920 (redundant_insn): Don't set it.
16921
61216c26
SB		 16922 * rtl.h (next_label, skip_consecutive_labels, link_cc0_insns):
16923 Remove prototypes.
16924 * emit-rtl.c (next_label): Remove unused function.
16925 (skip_consecutive_labels, link_cc0_insns): Move to ...
16926 * reorg.c (skip_consecutive_labels, link_cc0_insns): ... here, the
16927 only place where these functions are used, and make them static.
16928
418d1b87
MG		 169292013-05-14 Marc Glisse <marc.glisse@inria.fr>
16930
16931 * fold-const.c (fold_negate_expr): Handle vectors.
16932 (fold_truth_not_expr): Make it static.
16933 (fold_invert_truthvalue): New static function.
16934 (invert_truthvalue_loc): Handle vectors. Do not call
16935 fold_truth_not_expr directly.
16936 (fold_unary_loc) <BIT_NOT_EXPR>: Handle comparisons.
16937 <TRUTH_NOT_EXPR>: Do not cast to boolean.
16938 (fold_comparison): Handle vector constants.
16939 (fold_binary_loc) <TRUTH_XOR_EXPR>: Remove redundant code.
16940 (fold_ternary_loc) <VEC_COND_EXPR>: Adapt more COND_EXPR optimizations.
16941 * tree.h (fold_truth_not_expr): Remove declaration.
16942
fc21784d
JG		 169432013-05-14 James Greenhalgh <james.greenhalgh@arm.com>
16944
16945 * config/aarch64/aarch64-simd.md
16946 (aarch64_vcond_internal<mode>): Rename to...
16947 (aarch64_vcond_internal<mode><mode>): ...This, for integer modes.
16948 (aarch64_vcond_internal<VDQF_COND:mode><VDQF:mode>): ...This for
16949 float modes. Clarify all iterator modes.
16950 (vcond<mode><mode>): Use new name for vcond expanders.
16951 (vcond<v_cmp_result><mode>): Likewise.
16952 (vcondu<mode><mode>: Likewise.
16953 * config/aarch64/iterators.md (VDQF_COND): New.
16954
d4c52634
MG		 169552013-05-14 Marc Glisse <marc.glisse@inria.fr>
16956
16957 PR bootstrap/57266
16958 * fold-const.c (fold_binary_loc) <shift>: Use an unsigned
16959 variable for the shift amount. Check that we shift by non-negative
16960 amounts.
16961
2b261262
CLT		 169622013-05-14 Chung-Lin Tang <cltang@codesourcery.com>
16963
16964 PR target/42017
16965 * config/arm/arm.h (EPILOGUE_USES): Only return true
16966 for LR_REGNUM after epilogue_completed.
16967
12211b99 169682013-05-14 Joern Rennecke <joern.rennecke@embecosm.com>
6e022d7b
JR		 16969
16970 * config/avr/avr.c (avr_encode_section_info): Bail out if the type
e7208ea3 16971 is error_mark_node.
6e022d7b 16972
56cf7859
RO		 169732013-05-14 Rainer Orth <ro@CeBiTec.Uni-Bielefeld.DE>
16974
16975 PR target/57261
16976 * configure.ac (gcc_cv_ld_as_needed): Disable before Solaris 11
16977 and Solaris 11+/x86 with gld.
16978 * configure: Regenerate.
16979
75776c6d
JJ		 169802013-05-14 Jakub Jelinek <jakub@redhat.com>
16981
16982 * expmed.c (expand_shift_1): Canonicalize rotates by
16983 constant bitsize / 2 to bitsize - 1.
7f998021 16984 * simplify-rtx.c (simplify_binary_operation_1) <case ROTATE,
75776c6d
JJ		 16985 case ROTATERT>: Likewise.
16986
16987 Revert:
16988 2013-05-10 Jakub Jelinek <jakub@redhat.com>
16989
16990 * config/i386/i386.md (rotateinv): New code attr.
16991 (*<rotate_insn><mode>3_1, *<rotate_insn>si3_1_zext,
16992 *<rotate_insn>qi3_1_slp): Emit rorl %eax instead of
16993 roll $31, %eax, etc.
16994
df35498a
RB		 169952013-05-14 Richard Biener <rguenther@suse.de>
16996
16997 PR middle-end/57235
16998 * tree-eh.c (sink_clobbers): Give up for successors with
16999 multiple predecessors and no virtual uses.
17000
cc6e7ece
EB		 170012013-05-14 Eric Botcazou <ebotcazou@adacore.com>
17002
17003 * config/sparc/sp64-elf.h (CPP_SUBTARGET_SPEC): Delete.
17004 * config/sparc/openbsd64.h (CPP_SUBTARGET_SPEC): Likewise.
17005
f778dd4d
JJ		 170062013-05-14 Jakub Jelinek <jakub@redhat.com>
17007
17008 PR middle-end/57251
17009 * expr.c (expand_expr_real_2) <case WIDEN_MULT_EXPR>: Handle
17010 the case when both op0 and op1 have VOIDmode.
17011
bad4df9b
KP		 170122013-05-14 Kaushik Phatak <kaushik.phatak@kpitcummins.com>
17013
17014 * config/rl78/rl78.md(mulsi3_g13): Add additional 'nop' required
17015 in multiply-accumulate mode.
17016
56f3e9ac
GW		 170172013-05-13 Guozhi Wei <carrot@google.com>
17018
17019 * dwarf2asm.c (dw2_output_indirect_constant_1): Mark new decl STATIC.
17020
a3d7ab92
KT		 170212013-05-13 Kai Tietz <ktietz@redhat.com>
17022
17023 PR target/56975
17024 * config/i386/cygming.h (TARGET_PECOFF): Define as true.
e7208ea3 17025 * config/i386/i386.h (TARGET_PECOFF): Define by default as false.
a3d7ab92 17026 (PIC_OFFSET_TABLE_REGNUM): Use TARGET_PECOFF.
e7208ea3 17027 * config/i386/i386.c (ix86_option_override_internal): Likewise.
a3d7ab92
KT		 17028 (ix86_expand_prologue): Likewise.
17029 (ix86_expand_split_stack_prologue): Likewise.
17030 (legitimate_pic_address_disp_p): Likewise.
17031 (legitimize_pic_address): Likewise.
17032 (legitimize_tls_address): Likewise.
17033 (legitimize_pe_coff_symbol): Likewise.
17034 (output_pic_addr_const): Likewise.
17035 (construct_plt_address): Likewise.
17036 (ix86_expand_call): Likewise.
17037 (x86_output_mi_thunk): Likewise.
17038 (x86_function_profiler): Likewise.
17039
c59b7e28
SN		 170402013-05-13 Sofiane Naci <sofiane.naci@arm.com>
17041
17042 * config/aarch64/aarch64-simd.md (aarch64_simd_mov<mode>): Group
17043 similar switch cases.
17044 (aarch64_simd_mov): Rename to aarch64_split_simd_mov. Update.
17045 (aarch64_simd_mov_to_<mode>low): Delete.
17046 (aarch64_simd_mov_to_<mode>high): Delete.
17047 (move_lo_quad_<mode>): Add w<-r alternative.
17048 (aarch64_simd_move_hi_quad_<mode>): Likewise.
17049 (aarch64_simd_mov_from_*): Update type attribute.
17050 * config/aarch64/aarch64.c (aarch64_split_simd_move): Refacror switch
17051 statement.
17052
5f28524a
JH		 170532013-05-13 Jan Hubicka <jh@suse.cz>
17054
17055 * mode-switching.c (optimize_mode_switching): Set correct RTL profile.
17056 * config/i386/i386.c (ix86_compute_frame_layout,
e7208ea3
UB		 17057 ix86_expand_epilogue, emit_i387_cw_initialization,
17058 ix86_expand_vector_move_misalign, ix86_fp_comparison_strategy,
17059 ix86_local_alignment): Fix use of size/speed predicates.
5f28524a 17060
ae6fa899
JJ		 170612013-05-13 Jakub Jelinek <jakub@redhat.com>
17062
17063 PR tree-optimization/45216
17064 PR tree-optimization/57157
17065 * tree-ssa-forwprop.c (simplify_rotate): Only recognize
17066 the (-Y) & (B - 1) variant if OP is |.
17067 * expmed.c (expand_shift_1): For rotations by const0_rtx just
17068 return shifted. Use (-op1) & (prec - 1) as other_amount
17069 instead of prec - op1.
17070
4502fe8d
MJ		 170712013-05-13 Martin Jambor <mjambor@suse.cz>
17072
17073 PR middle-end/42371
17074 * ipa-prop.h (IPA_UNDESCRIBED_USE): New macro.
17075 (ipa_constant_data): New type.
17076 (ipa_jump_func): Use ipa_constant_data to hold information about
17077 constant jump functions.
17078 (ipa_get_jf_constant): Adjust to jump function type changes.
17079 (ipa_get_jf_constant_rdesc): New function.
17080 (ipa_param_descriptor): New field controlled_uses.
17081 (ipa_get_controlled_uses): New function.
17082 (ipa_set_controlled_uses): Likewise.
17083 * ipa-ref.h (ipa_find_reference): Declare.
17084 * ipa-prop.c (ipa_cst_ref_desc): New type.
17085 (ipa_print_node_jump_functions_for_edge): Adjust for jump function type
17086 changes.
17087 (ipa_set_jf_constant): Likewise. Also create reference descriptions.
17088 New parameter cs. Adjust all callers.
17089 (ipa_analyze_params_uses): Detect uncontrolled and controlled uses.
17090 (remove_described_reference): New function.
17091 (jfunc_rdesc_usable): Likewise.
17092 (try_make_edge_direct_simple_call): Decrement controlled use count,
17093 attempt to remove reference if it hits zero.
17094 (combine_controlled_uses_counters): New function.
17095 (propagate_controlled_uses): Likewise.
17096 (ipa_propagate_indirect_call_infos): Call propagate_controlled_uses.
17097 (ipa_edge_duplication_hook): Duplicate reference descriptions.
17098 (ipa_print_node_params): Print described use counter.
17099 (ipa_write_jump_function): Adjust to jump function type changes.
17100 (ipa_read_jump_function): New parameter CS, pass it to
17101 ipa_set_jf_constant. Adjust caller.
17102 (ipa_write_node_info): Stream controlled use count
17103 (ipa_read_node_info): Likewise.
17104 * cgraph.c (cgraph_mark_address_taken_node): Bail out instead of
17105 asserting.
17106 * ipa-cp.c (ipcp_discover_new_direct_edges): Decrement controlled use
17107 count. Remove cloning-added reference if it reaches zero.
17108 * ipa-ref.c (ipa_find_reference): New function.
17109
0864bfc2
GG		 171102013-05-13 Ganesh Gopalasubramanian <Ganesh.Gopalasubramanian@amd.com>
17111
e7208ea3 17112 * config/i386/i386.c (processor_target_table): Modified default
0864bfc2
GG		 17113 alignment values for AMD BD and BT architectures.
17114
640bfeb2
MG		 171152013-05-13 Marc Glisse <marc.glisse@inria.fr>
17116
17117 * tree-vect-generic.c (uniform_vector_p): Move ...
17118 * tree.c (uniform_vector_p): ... here.
17119 * tree.h (uniform_vector_p): Declare it.
17120 * fold-const.c (fold_binary_loc) <shift>: Turn the second argument
17121 into a scalar.
17122
3a60f32b
JJ		 171232013-05-13 Jakub Jelinek <jakub@redhat.com>
17124
198fe1bf
JJ		 17125 PR tree-optimization/57230
17126 * tree-ssa-strlen.c (handle_char_store): Record length for
17127 array store from STRING_CST.
17128
3a60f32b
JJ		 17129 PR tree-optimization/57230
17130 * tree-ssa-strlen.c (handle_char_store): Add missing integer_zerop
17131 check.
17132
566be57c
JR		 171332013-05-12 Joern Rennecke <joern.rennecke@embecosm.com>
17134
17135 * config/epiphany/epiphany.c (epiphany_init): Check size of
17136 NUM_MODES_FOR_MODE_SWITCHING.
17137 (epiphany_expand_prologue):
17138 Remove CONFIG_REGNUM initial value handling code.
17139 (epiphany_optimize_mode_switching): Handle EPIPHANY_MSW_ENTITY_CONFIG.
17140 (epiphany_mode_needed, epiphany_mode_entry_exit): Likewise.
2710a27a 17141 (emit_set_fp_mode, epiphany_mode_after): Likewise.
566be57c
JR		 17142 (epiphany_mode_needed) <Handle EPIPHANY_MSW_ENTITY_AND>:
17143 Don't return 1 for FP_MODE_NONE.
17144 * config/epiphany/epiphany.h (NUM_MODES_FOR_MODE_SWITCHING):
17145 Add value for EPIPHANY_MSW_ENTITY_CONFIG.
17146 (EPIPHANY_MSW_ENTITY_CONFIG, EPIPHANY_MSW_ENTITY_NUM): Define.
17147 * config/epiphany/epiphany.md (save_config): New pattern.
17148
0f2c2331
UB		 171492013-05-12 Uros Bizjak <ubizjak@gmail.com>
17150
17151 * config/i386/i386.md (*zero_extendsidi2): Add *x->?r alternative.
17152
5b3f93c7
UB		 171532013-05-10 Uros Bizjak <ubizjak@gmail.com>
17154
17155 * config/i386/i386.md (memory): Handle sseishft1.
17156 * config/i386/sse.md (*vec_extractv4si): Remove memory attribute.
17157 (*vec_extractv2di_1): Ditto.
17158
1f873f0e
VM		 171592013-05-10 Vladimir Makarov <vmakarov@redhat.com>
17160
17161 * lra-assigns.c (find_hard_regno_for): Add 1 to the cost of call
17162 saved registers.
17163
341427fa 171642013-05-10 Sebastian Huber <sebastian.huber@embedded-brains.de>
d7de6455
SH		 17165
17166 * config/arm/t-rtems-eabi: Remove mthumb/march=armv7 multilib.
17167 Add mthumb/march=armv7-a multilib.
17168 Add mthumb/march=armv7-r multilib.
17169 Add mthumb/march=armv7-a/mfpu=neon/mfloat-abi=hard multilib.
17170
9e69bdde
RC		 171712013-05-10 Ralf Corsépius <ralf.corsepius@rtems.org>
17172
17173 * config/v850/t-rtems: Add more multilibs.
17174
9ff09a22
RB		 171752013-05-10 Richard Biener <rguenther@suse.de>
17176
17177 PR tree-optimization/57214
17178 * tree-ssa-loop-ivcanon.c (propagate_constants_for_unrolling): Do
17179 not propagate from SSA names that occur in abnormal PHI nodes.
17180
a5e0cd1d
MG		 171812013-05-10 Marc Glisse <marc.glisse@inria.fr>
17182
17183 * stor-layout.c (element_precision): New function.
17184 * machmode.h (element_precision): Declare it.
17185 * tree.c (build_minus_one_cst): New function.
17186 (element_precision): Likewise.
17187 * tree.h (build_minus_one_cst): Declare new function.
17188 (element_precision): Likewise.
17189 * fold-const.c (operand_equal_p): Use element_precision.
17190 (fold_binary_loc): Handle vector types.
17191 * convert.c (convert_to_integer): Use element_precision.
17192 * gimple.c (iterative_hash_canonical_type): Handle complex and vectors
17193 separately.
17194
cb2558bc
RS		 171952013-05-10 Richard Sandiford <rdsandiford@googlemail.com>
17196
17197 * config/mips/mips-protos.h (m16_uimm3_b, m16_simm4_1, m16_nsimm4_1)
17198 (m16_simm5_1, m16_nsimm5_1, m16_uimm5_4, m16_nuimm5_4, m16_simm8_1)
17199 (m16_nsimm8_1, m16_uimm8_1, m16_nuimm8_1, m16_uimm8_m1_1, m16_uimm8_4)
17200 (m16_nuimm8_4, m16_simm8_8, m16_nsimm8_8): Delete.
17201 * config/mips/mips.c (m16_check_op, m16_uimm3_b, m16_simm4_1)
17202 (m16_nsimm4_1, m16_simm5_1, m16_nsimm5_1, m16_uimm5_4, m16_nuimm5_4)
17203 (m16_simm8_1, m16_nsimm8_1, m16_uimm8_1, m16_nuimm8_1, m16_uimm8_m1_1)
17204 (m16_uimm8_4, m16_nuimm8_4, m16_simm8_8, m16_nsimm8_8): Delete.
17205 * config/mips/constraints.md (Udb8, Usb5, Usb8, Usd8, Uub8, Uuw5)
17206 (Uuw8): New constraints.
17207 (Usb4): Move into alphabetical order.
17208 * config/mips/predicates.md (db8_operand, sb5_operand, sb8_operand)
17209 (sd8_operand, ub8_operand, uw8_operand): New predicates.
17210 * config/mips/mips.md (*xor<mode>3, *xor<mode>3_mips16): Name
17211 previously unnamed patterns.
17212 (*add<mode>3_mips16, *xor<mode>3_mips16, *<optab>si3_mips16)
17213 (*ashldi3_mips16, *ashrdi3_mips16, *lshrdi3_mips16)
17214 (*slt<u>_<GPR:mode><GPR2:mode>_mips16)
17215 (*sle<u>_<GPR:mode><GPR2:mode>_mips16): Use constraints instead
17216 of set_attr_alternative/if_then_else. Use extended_mips16 instead
17217 of specific lengths.
17218
cb3b8d33
JJ		 172192013-05-10 Jakub Jelinek <jakub@redhat.com>
17220
6f93c008
JJ		 17221 * config/i386/i386.md (rotateinv): New code attr.
17222 (*<rotate_insn><mode>3_1, *<rotate_insn>si3_1_zext,
17223 *<rotate_insn>qi3_1_slp): Emit rorl %eax instead of
17224 roll $31, %eax, etc.
17225
cb3b8d33
JJ		 17226 PR tree-optimization/45216
17227 PR tree-optimization/57157
17228 * tree-ssa-forwprop.c (simplify_rotate): New function.
17229 (ssa_forward_propagate_and_combine): Call it.
17230
afb119be
RB		 172312013-05-10 Richard Biener <rguenther@suse.de>
17232
17233 * tree-vect-data-refs.c (vect_enhance_data_refs_alignment): Do not
17234 disable peeling when we version for aliasing.
17235 (vector_alignment_reachable_p): Honor explicit user alignment.
17236 (vect_supportable_dr_alignment): Likewise.
17237 * tree-vect-loop-manip.c (vect_can_advance_ivs_p): Use
17238 STMT_VINFO_LOOP_PHI_EVOLUTION_PART instead of recomputing it.
17239 * tree-vect-loop.c (vect_transform_loop): First apply versioning,
17240 then peeling to arrange for the cost-model check to come first.
17241
01ae4861
AM		 172422013-05-10 Alan Modra <amodra@gmail.com>
17243
17244 * configure.ac (HAVE_AS_TLS): Swap powerpc64 and powerpc cases.
17245 (HAVE_LD_LARGE_TOC): Don't mention AIX in help text.
17246 * configure: Regenerate.
17247
ebc9a431
AM		 172482013-05-10 Alan Modra <amodra@gmail.com>
17249
17250 PR target/55033
17251 * varasm.c (default_elf_select_section): Move !DECL_P check..
17252 (get_named_section): ..to here before calling get_section_name.
17253 Adjust assertion.
17254 (default_section_type_flags): Add DECL_P check.
17255 * config/i386/winnt.c (i386_pe_section_type_flags): Likewise.
17256 * config/rs6000/rs6000.c (rs6000_xcoff_section_type_flags): Likewise.
17257
d4bca93c
JR		 172582013-05-09 Joern Rennecke <joern.rennecke@embecosm.com>
17259
17260 * config/epiphany/epiphany.c (epiphany_expand_prologue):
17261 When using gen_stack_adjust_str with a register offset, add a
17262 REG_FRAME_RELATED_EXPR note.
17263
60ca9a65
UB		 172642013-05-09 Uros Bizjak <ubizjak@gmail.com>
17265
17266 * config/i386/sse.md (*vec_extractv4si_0_zext): New pattern.
17267 (*vec_extractv4si_zext_mem): Ditto.
17268 (*vec_extractv2di): Add 0->x and x->x alternatives.
17269 * config/i386/mmx.md (*vec_extractv2si_zext_mem): New pattern.
17270 * config/i386/i386.md (*zero_extendsidi2): Add *Yj->?r alternative.
17271
a3409c02
JM		 172722013-05-09 Jason Merrill <jason@redhat.com>
17273
0138d6b2
JM		 17274 N3639 C++1y VLA support
17275 * gimplify.c (gimplify_vla_decl): Don't touch an existing
17276 DECL_VALUE_EXPR.
17277
a3409c02
JM		 17278 * tree.c (build_constructor_va): New.
17279 * tree.h: Declare it.
17280
66e6b990
MJ		 172812013-05-09 Martin Jambor <mjambor@suse.cz>
17282
17283 PR lto/57084
17284 * gimple-fold.c (canonicalize_constructor_val): Call
17285 cgraph_get_create_real_symbol_node instead of cgraph_get_create_node.
17286
64cfa6c0
JH		 172872013-05-09 Jan Hubicka <jh@suse.cz>
17288 Richard Biener <rguenther@suse.de>
17289
17290 PR lto/54095
17291 * symtab.c (symtab_make_decl_local): Do not add private names.
17292
c3167b00
JH		 172932013-05-09 Jan Hubicka <jh@suse.cz>
17294
17295 PR lto/54095
17296 * symtab.c (insert_to_assembler_name_hash): Handle clones.
17297 (unlink_from_assembler_name_hash): Likewise.
17298 (symtab_prevail_in_asm_name_hash, symtab_register_node,
17299 symtab_unregister_node, symtab_initialize_asm_name_hash,
17300 change_decl_assembler_name): Update.
17301
12dc6974
SN		 173022013-05-09 Sofiane Naci <sofiane.naci@arm.com>
17303
17304 * config/aarch64/aarch64.md: New movtf split.
17305 (*movtf_aarch64): Update.
17306 (aarch64_movdi_tilow): Handle TF modes and rename to
17307 aarch64_movdi_<mode>low.
17308 (aarch64_movdi_tihigh): Handle TF modes and rename to
17309 aarch64_movdi_<mode>high
17310 (aarch64_movtihigh_di): Handle TF modes and rename to
17311 aarch64_mov<mode>high_di
17312 (aarch64_movtilow_di): Handle TF modes and rename to
17313 aarch64_mov<mode>low_di
17314 (aarch64_movtilow_tilow): Remove spurious whitespace.
17315 * config/aarch64/aarch64.c (aarch64_split_128bit_move): Handle TFmode
17316 splits.
17317 (aarch64_print_operand): Update.
17318
227eb343
AM		 173192013-05-09 Alan Modra <amodra@gmail.com>
17320
17321 * configure.ac (HAVE_AS_TLS): Enable tests for powerpcle and
17322 powerpc64le.
17323 * configure: Regenerate.
17324
0b013847
UB		 173252013-05-08 Uros Bizjak <ubizjak@gmail.com>
17326
17327 * config/i386/mmx.md (*vec_extract* splitters): Simplify post-reload
17328 splitter preparation statements.
17329 * config/i386/sse.md (*vec_extract* splitters): Ditto.
17330 (*avx_vperm_broadcast_<mode>): Use adjust_address instead of
17331 adjust_address_nv.
17332
1dc3d6e9
BS		 173332013-05-08 Bill Schmidt <wschmidt@linux.vnet.ibm.com>
17334
17335 * gimple-ssa-strength-reduction.c (count_candidates): Change
17336 return value to int.
17337 (analyze_candidates_and_replace): Change type of length to int.
17338
e61e7d28
UB		 173392013-05-08 Uros Bizjak <ubizjak@gmail.com>
17340
17341 * config/i386/sse.md (PEXTR_MODE, PEXTR_MODEx): Remove.
17342 (*vec_extract<mode>): Use VI12_128 mode iterator.
17343 (*vec_extract<mode>_mem): Ditto.
17344 (*vec_extract*_mem splitters): Merge splitters using VI_128 mode
17345 attribute.
17346
4fbfcf44
DN		 173472013-05-08 Diego Novillo <dnovillo@google.com>
17348
17349 PR bootstrap/54659
17350
17351 Revert:
4fbfcf44
DN		 17352 2012-08-17 Diego Novillo <dnovillo@google.com>
17353
e61e7d28
UB		 17354 PR bootstrap/54281
17355 * configure.ac: Add libintl.h to AC_CHECK_HEADERS list.
17356 * config.in: Regenerate.
17357 * configure: Regenerate.
17358 * intl.h: Always include libintl.h if HAVE_LIBINTL_H is set.
4fbfcf44 17359
702d8703
JH		 173602013-05-08 Jan Hubicka <jh@suse.cz>
17361
17362 PR lto/54095
17363 * cgraph.c (cgraph_make_node_local_1): Se unique_name.
17364 * cgraph.h (symtab_node_base): Add unique_name.
17365 * lto-cgraph.c (lto_output_node, lto_output_varpool_node,
17366 input_overwrite_node, input_varpool_node): Stream unique_name.
17367 * cgraphclones.c (cgraph_create_virtual_clone,
17368 cgraph_function_versioning): Set unique_name.
17369 * ipa.c (function_and_variable_visibility): Set unique_name.
17370
8b28cf47
BS		 173712013-05-08 Bill Schmidt <wschmidt@linux.vnet.ibm.com>
17372
17373 * gimple-ssa-strength-reduction.c (find_phi_def): Revert former "fix."
17374 (alloc_cand_and_find_basis): Restrict conditional candidate
17375 processing to CAND_MULTs.
17376
e86074fd
JH		 173772013-05-08 Jan Hubicka <jh@suse.cz>
17378
17379 PR lto/54095
17380 lto-symtab.c (lto_symtab_symbol_p): New function.
17381 (lto_symtab_resolve_can_prevail_p, lto_symtab_resolve_symbols,
17382 lto_symtab_resolve_symbols, lto_symtab_merge_decls_2,
17383 lto_symtab_merge_decls_1, lto_symtab_merge_cgraph_nodes_1):
17384 Skip static symbols.
17385
44398cbe
PC		 173862013-05-08 Paolo Carlini <paolo.carlini@oracle.com>
17387
17388 PR tree-optimization/57200
17389 * tree-ssa-loop-niter.c (do_warn_aggressive_loop_optimizations):
17390 Only call inform if the preceding warning_at returns true.
17391
f6bc1c4a
HS		 173922013-05-07 Han Shen <shenhan@google.com>
17393
17394 * cfgexpand.c (record_or_union_type_has_array_p): New function.
17395 (expand_used_vars): Add logic handling '-fstack-protector-strong'.
17396 * common.opt (fstack-protector-strong): New option.
17397 * doc/cpp.texi (__SSP_STRONG__): New builtin "__SSP_STRONG__".
17398 * doc/invoke.texi (Optimization Options): Document
17399 "-fstack-protector-strong".
17400 * gcc.c (LINK_SSP_SPEC): Add 'fstack-protector-strong'.
17401
4ffecb1f
SB		 174022013-05-06 Steven Bosscher <steven@gcc.gnu.org>
17403
17404 * config/mips/mips.c (mips_machine_reorg2): Return 0.
17405
5a107a0f
VM		 174062013-05-07 Vladimir Makarov <vmakarov@redhat.com>
17407
17408 * ira.c (update_equiv_regs): Add insn having equiv memory even if
17409 it is not lhs of the insn.
17410 (setup_reg_equiv): Remove insn having equiv memory which it is not
17411 lhs of the insn.
17412 * lra-constraints.c (process_address): Try to improve generation
17413 code for address base + disp.
17414 (lra_constraints): Make correct the code for checking insn setting
17415 up backward equivalence. Remove insn only if it is in the init
17416 insn list.
17417 * lra-eliminations.c (update_reg_eliminate): Change return value.
17418 (lra_eliminate): Use the result.
17419
3f5783ea
UB		 174202013-05-07 Uros Bizjak <ubizjak@gmail.com>
17421
17422 * config/i386/sse.md (ssescalarnummask): New mode attribute.
17423 (PEXTR_MODE, PEXTR_MODEx): New mode iterators.
17424 (*vec_extract<mode>): Merge from *sse4_1_pextrb_memory and
17425 *sse4_1_pextrw_memory using PEXTR_MODE mode iterator. Handle
17426 register target operands.
17427 (*vec_extractv8hi_sse2): New pattern.
17428 (*vec_extractv16qi_zext): Rename from *sse4_1_pextrb_<mode>.
17429 (*vec_extractv8hi_zext): Rename from *sse2_pextrw_<mode>.
17430 (*vec_extract<mode>_mem): New insn and split pattern.
17431
8a5800b8
CL		 174322013-05-07 Christophe Lyon <christophe.lyon@linaro.org>
17433
17434 * config/arm/arm.c (arm_asan_shadow_offset): New function.
17435 (TARGET_ASAN_SHADOW_OFFSET): Define.
17436 * config/arm/linux-eabi.h (ASAN_CC1_SPEC): Define.
17437 (LINUX_OR_ANDROID_CC): Add ASAN_CC1_SPEC.
17438
7bf55a70
BS		 174392013-05-07 Bill Schmidt <wschmidt@linux.vnet.ibm.com>
17440
17441 * gimple-ssa-strength-reduction.c (MAX_INCR_VEC_LEN): New constant.
17442 (incr_vec_index): Return -1 if increment not found.
17443 (create_add_on_incoming_edge): Assert if increment not found.
17444 (record_increment): Limit number of increments recorded.
17445 (all_phi_incrs_profitable): Return false if an increment not found.
17446 (replace_profitable_candidates): Don't process increments that were
17447 not recorded.
17448 (analyze_candidates_and_replace): Limit size of incr_vec.
17449
3f8825c0
RB		 174502013-05-07 Richard Biener <rguenther@suse.de>
17451
17452 * calls.c (special_function_p): setjmp-like functions are leaf.
17453 * builtins.def (BUILT_IN_SETJMP): setjmp is leaf.
17454 * tree-inline.c (update_ssa_across_abnormal_edges): Remove assert.
17455
fd4842cd
SN		 174562013-05-07 Sofiane Naci <sofiane.naci@arm.com>
17457
17458 * config/aarch64/aarch64-simd.md (*aarch64_simd_mov<mode>): call splitter.
17459 (aarch64_simd_mov<mode>): New expander.
17460 (aarch64_simd_mov_to_<mode>low): New instruction pattern.
17461 (aarch64_simd_mov_to_<mode>high): Likewise.
17462 (aarch64_simd_mov_from_<mode>low): Likewise.
17463 (aarch64_simd_mov_from_<mode>high): Likewise.
17464 (aarch64_dup_lane<mode>): Update.
17465 (aarch64_dup_lanedi): New instruction pattern.
17466 * config/aarch64/aarch64-protos.h (aarch64_split_simd_move): New prototype.
17467 * config/aarch64/aarch64.c (aarch64_split_simd_move): New function.
17468
a7a7d10e
BS		 174692013-05-07 Bill Schmidt <wschmidt@linux.vnet.ibm.com>
17470
17471 * gimple-ssa-strength-reduction.c (lazy_create_slsr_reg): Remove.
17472 (replace_mult_candidate): Remove unnecessary argument; remove
17473 unnecessary parameter from call to introduce_cast_before_cand.
17474 (replace_unconditional_candidate): Remove unnecessary parameter
17475 from call to replace_mult_candidate.
17476 (replace_conditional_candidate): Likewise.
17477 (insert_initializers): Use make_temp_ssa_name.
17478 (introduce_cast_before_cand): Remove unnecessary argument; use
17479 make_temp_ssa_name.
17480 (replace_one_candidate): Remove unnecessary argument; remove
17481 unnecessary parameter from calls to introduce_cast_before_cand.
17482 (replace_profitable_candidates): Remove unnecessary parameters
17483 from calls to replace_one_candidate.
17484
29105868
BS		 174852013-05-07 Bill Schmidt <wschmidt@linux.vnet.ibm.com>
17486
17487 * gimple-ssa-strength-reduction.c (find_phi_def): Don't record a
17488 phi def as possibly hiding a basis for a CAND_ADD whose operands
17489 have been commuted in the analysis.
17490 (alloc_cand_and_find_basis): Add parms to call to find_phi_def.
17491
4095f9fa
N		 174922013-05-07 Naveen H.S <Naveen.Hurugalawadi@caviumnetworks.com>
17493
17494 * config/aarch64/aarch64.md
17495 (cmp_swp_<optab><ALLX:mode>_shft_<GPI:mode>): Restrict the
17496 shift value between 0-4.
17497
39e843e8
RB		 174982013-05-07 Richard Biener <rguenther@suse.de>
17499
17500 * double-int.h (rshift): New overload.
17501 * double-int.c (rshift): New function.
17502 * tree-ssa-sccvn.c (copy_reference_ops_from_ref): Optimize.
17503 (create_reference_ops_from_ref): Remove.
17504 (vn_reference_insert): Use shared ops for constructing the
17505 reference and copy it.
17506
0a1a83cb
RB		 175072013-05-07 Richard Biener <rguenther@suse.de>
17508
17509 PR middle-end/57190
17510 * tree-eh.c (sink_clobbers): Properly propagate
17511 SSA_NAME_OCCURS_IN_ABNORMAL_PHI.
17512
43303d6f
JJ		 175132013-05-07 Jakub Jelinek <jakub@redhat.com>
17514
ba7e83f8
JJ		 17515 PR tree-optimization/57149
17516 * tree-ssa-uninit.c (uninit_undefined_value_p): New inline.
17517 (can_skip_redundant_opnd, compute_uninit_opnds_pos,
17518 collect_phi_def_edges, execute_late_warn_uninitialized): Use
17519 uninit_undefined_value_p instead of ssa_undefined_value_p.
17520
43303d6f
JJ		 17521 PR debug/57184
17522 * expr.c (expand_expr_addr_expr_1): Handle COMPOUND_LITERAL_EXPR
17523 for modifier == EXPAND_INITIALIZER.
17524
14523c25
AB		 175252013-05-07 Anton Blanchard <anton@samba.org>
17526
17527 * configure.ac (HAVE_LD_LARGE_TOC): Use correct linker emulation
17528 for powerpc64 little endian.
17529 * configure: Regenerate.
17530
cb7c8be9
GS		 175312013-05-06 Graham Stott <grahams@btinternet.com>
17532
17533 * expmed.c (init_expmed_rtl): Remove unused fields reg_fld, plus_fld,
17534 mult_fld, sdiv_fld1, udiv_fld1, sdiv_32_fld1, smod_32_fld1,
17535 wide_mult_fld1, wide_lshr_fld1, shift_fld1, shift_mult_fld1,
17536 shift_add_fld1, shift_sub0_fld1, shift_sub1_fld1.
17537
175382013-05-06 Graham Stott <grahams@btinternet.com>
a8fd4225
GS		 17539
17540 * gensupport.c (add_predicate_code): Also exclude SCRATCH from rtx
17541 codes which allow non-lvalues.
17542
9a0ee7b0
MG		 175432013-05-06 Marc Glisse <marc.glisse@inria.fr>
17544
17545 * tree.c (integer_all_onesp) <COMPLEX_CST>: Test that both
17546 components are all 1s.
17547 (integer_minus_onep): New function.
17548 * tree.h (integer_minus_onep): Declare it.
17549 * fold-const.c (fold_binary_loc) <MULT_EXPR>: Test
17550 integer_minus_onep instead of integer_all_onesp.
17551
f2c17ea9
OE		 175522013-05-06 Oleg Endo <olegendo@gcc.gnu.org>
17553
17554 PR target/52933
17555 * config/sh/sh.md (*cmp_div0s_0, *cmp_div0s_1, *movsicc_div0s): Add
17556 variations of these patterns.
17557
f75e6a51
UB		 175582013-05-06 Uros Bizjak <ubizjak@gmail.com>
17559
17560 * config/i386/i386.md (isa): Add x64_sse4 member.
17561 (enabled): Handle x64_sse4.
17562 (*movdi_internal): Add *x->?r alternative to emit pextrq $0,%xmm,%reg
17563 instruction for 64bit SSE4_1 targets. Update insn attributes.
17564 (*movsi_internal): Add *x->?r alternative to emit pextrd $0,%xmm,%reg
17565 instruction for SSE4_1 targets. Update insn attributes.
17566 * config/i386/sse.md (*vec_extract<ssevecmodelower>_0): Merge
17567 with *sse4_1_pextrd and *sse4_1_pextrq having const_0 selector.
17568 (*vec_extractv2di_1): Merge with *sse4_1_pextrq having
17569 const_1 selector.
17570 (*vec_extractv4si): Rename from *sse4_1_pextrd.
17571 (*vec_extractv4si_zext): Rename from *sse4_1_pextrd_zext.
17572 (*vec_extract<ssevecmodelower>_0 splitters): Merge splitters together.
17573
a986d468
OE		 175742013-05-06 Oleg Endo <olegendo@gcc.gnu.org>
17575
17576 PR target/57108
17577 * config/sh/sh.md (tstsi_t_zero_extract_eq): Use QIHISIDI mode iterator.
17578
382522cb
MK		 175792013-05-06 Maxim Kuznetsov <maks.kuznetsov@gmail.com>
17580
17581 * final.c (do_assembler_dialects): Don't handle curly braces and
17582 vertical bar escaped by % as dialect delimiters.
17583 (output_asm_insn): Print curly braces and vertical bar if escaped
17584 by % and ASSEMBLER_DIALECT defined.
17585 * doc/tm.texi.in (ASSEMBLER_DIALECT): Document new standard escapes.
17586 * doc/tm.texi: Regenerated.
17587
fb0d5c60
SB		 175882013-05-06 Steven Bosscher <steven@gcc.gnu.org>
17589
fb0d5c60
SB		 17590 * config/mips/mips.c: Include tree-pass.h.
17591 (mips_reorg): Split in pre- and post-dbr_schedule parts.
17592 (mips_machine_reorg2): Move mips_reorg post-dbr_schedule parts here.
17593 (pass_mips_machine_reorg2): New machine specific pass.
17594 (insert_pass_mips_machine_reorg2): New pass plugin definition.
17595 (mips_option_override): Register the new pass.
17596 * rtl.h (cleanup_barriers): Remove prototype.
17597 (dbr_schedule): Likewise.
17598 * jump.c (cleanup_barriers): Make static.
17599 * reorg.c (dbr_schedule): Likewise.
17600
aa06a978
RB		 176012013-05-06 Richard Biener <rguenther@suse.de>
17602
17603 PR tree-optimization/57185
17604 * tree-parloops.c (add_field_for_reduction): Handle anonymous
17605 SSA names properly.
17606
0b953bec
UB		 176072013-05-06 Uros Bizjak <ubizjak@gmail.com>
17608
17609 PR target/57106
17610 * config/i386/i386.c (add_parameter_dependencies): Add dependence
17611 between "first_arg" and "insn", not "last" and "insn".
17612
28708525
WS		 176132013-05-06 Bill Schmidt <wschmidt@linux.vnet.ibm.com>
17614
17615 * gimple-ssa-strength-reduction.c (slsr_process_phi): Re-enable.
17616 (find_candidates_in_block): Re-enable slsr_process_phi.
17617 (create_phi_basis): Fix double counting of candidate adjustment.
17618
0107dca2
RB		 176192013-05-06 Richard Biener <rguenther@suse.de>
17620
17621 PR middle-end/57147
17622 * tree-cfg.c (gimple_purge_dead_abnormal_call_edges): If
17623 the edge is also fallthru, preserve it and just clear the
17624 abnormal flag.
17625 * tree-cfgcleanup.c (remove_fallthru_edge): If the edge is
17626 also complex, preserve that and just clear the fallthru flag.
17627 * tree-inline.c (update_ssa_across_abnormal_edges): Also
17628 update virtual operands.
17629
470d4d13
AM		 176302013-05-06 Alan Modra <amodra@gmail.com>
17631
17632 * config/rs6000/linux.h (DEFAULT_ASM_ENDIAN): Define.
17633 (LINK_OS_LINUX_EMUL): Use ENDIAN_SELECT.
17634 * config/rs6000/linux64.h (DEFAULT_ASM_ENDIAN): Define.
17635 * config/rs6000/sysv4le.h (DEFAULT_ASM_ENDIAN): Define.
17636 (LINK_TARGET_SPEC): Use ENDIAN_SELECT.
17637 * config/rs6000/sysv4.h (DEFAULT_ASM_ENDIAN): Define as -mbig.
17638
8f1dbf8d
AM		 176392013-05-06 Alan Modra <amodra@gmail.com>
17640
17641 * config/rs6000/sysv4.h (ENDIAN_SELECT): Define, extracted from
17642 (ASM_SPEC): ..here. Emit DEFAULT_ASM_ENDIAN too.
17643 (DEFAULT_ASM_ENDIAN): Define.
17644 (CC1_SPEC, LINK_TARGET_SPEC): Use ENDIAN_SELECT.
17645 * config/rs6000/linux64.h (ASM_SPEC32): Remove endian options.
17646 Update -K PIC clause from sysv4.h.
17647 (ASM_SPEC_COMMON): Use ENDIAN_SELECT.
17648 (LINK_OS_LINUX_EMUL32, LINK_OS_LINUX_EMUL64): Likewise.
17649
54890767
AM		 176502013-05-06 Alan Modra <amodra@gmail.com>
17651
17652 * config/rs6000/rs6000.md (bswapdi 2nd splitter): Don't swap words
17653 twice for little-endian.
17654 (ashrdi3_no_power, ashrdi3): Support little-endian.
17655
2353515d
OE		 176562013-05-06 Oleg Endo <olegendo@gcc.gnu.org>
17657
17658 PR target/55303
17659 * config/sh/sh.c (sh_rtx_costs): Handle SMIN and SMAX cases.
17660 * config/sh/sh.md (*clips, uminsi3, *clipu, clipu_one): New insns and
17661 related expanders.
17662 * config/sh/iterators.md (SMIN_SMAX): New code iterator.
17663 * config/sh/predicates.md (arith_reg_or_0_or_1_operand,
17664 clips_min_const_int, clips_max_const_int, clipu_max_const_int):
17665 New predicates.
17666
33e67557
SB		 176672013-05-05 Steven Bosscher <steven@gcc.gnu.org>
17668 John David Anglin <dave.anglin@nrc-cnrc.gc.ca>
17669
17670 * config.gcc (hppa*-*-*): Remove MASK_BIG_SWITCH from CPU default.
17671 * config/pa/pa.opt: Make mbig-switch a no-op.
17672 * config/pa/pa.h (TARGET_DEFAULT): Remove MASK_BIG_SWITCH.
17673 (CASE_VECTOR_MODE): Always return SImode.
17674 (ASM_OUTPUT_ADDR_VEC_ELT, ASM_OUTPUT_ADDR_DIFF_ELT): Remove code
17675 for the !TARGET_BIG_SWITCH case.
17676 * config/pa/pa-linux.h: Likewise.
17677 * config/pa/pa-openbsd.h: Likewise.
17678 * config/pa/pa-hpux.h: Define TARGET_DEFAULT to 0.
17679 * config/pa/pa.md (short_jump): Remove define_insn.
17680 (casesi): Remove code for the !TARGET_BIG_SWITCH case.
17681 (casesi0): Remove define_insn.
17682 (type): Remove btable_branch.
17683 (pa_combine_type): Likewise.
17684 (in_nullified_branch_delay): Likewise.
17685 (in_call_delay): Likewise.
17686 (define_delay): Likewise.
17687 (define_insn_reservation "Z3"): Likewise.
17688 (define_insn_reservation "Z4"): Likewise.
17689 * config/pa/pa.c (pa_reorg): Remove code for !TARGET_BIG_SWITCH.
17690 (pa_adjust_insn_length): Remove adjustment for btable branches.
17691 * doc/invoke.texi (HPPA Options): Delete documentation for mbig-switch
17692 and mno-big-switch
17693
3095685e
UB		 176942013-05-05 Uros Bizjak <ubizjak@gmail.com>
17695
17696 * config/i386/sse.md (*vec_extract<ssevecmodelower>_0): Merge
17697 from sse2_stored and *sse2_storeq_rex64 using SWI48 mode iterator.
17698 Add m->r,x alternatives.
17699 (*vec_extract<ssevecmodelower>_0 splitters): Merge V2DI and V4SI
17700 splitters using SWI48x mode iterator.
17701 (*vec_extract_v2di_0_sse): Rename from *sse2_storeq. Disable for
17702 TARGET_64BIT. Add m->x alternative.
17703 (*vec_extractv4si_mem): Rename from *vec_ext_v4si_mem.
17704 Add o->x alternative. Enable for TARGET_SSE.
17705 (sse_storeq): Remove expander.
17706 (*vec_extractv2di_1): Enable for TARGET_SSE. Split alternatives
17707 with memory input operand.
17708 (*vec_extractv2di_1 splitter): New.
17709 (*vec_extractv4sf_mem): Rename from *vec_extract_v4sf_mem.
17710 * config/i386/i386.md (ssevecmodelower): New mode attribute.
17711
4b36ae28
SB		 177122013-05-04 Segher Boessenkool <segher@kernel.crashing.org>
17713
17714 * config/rs6000/rs6000.c (INT_P): Reformat. Delete obsolete comment.
17715 (INT_LOWPART): Delete.
17716 (extract_MB): Adjust.
17717 (extract_ME): Adjust.
17718 (print_operand): Adjust.
17719
da226db2
SB		 177202013-05-04 Segher Boessenkool <segher@kernel.crashing.org>
17721
17722 * config/rs6000/predicates.md (reg_or_add_cint_operand,
17723 reg_or_sub_cint_operand): Delete "HOST_BITS_PER_WIDE_INT == 32" case.
17724 (reg_or_logical_cint_operand, easy_fp_constant,
17725 logical_const_operand): Delete "CONST_DOUBLE" case.
17726 * config/rs6000/rs6000.c (num_insns_constant_wide): Delete
17727 "HOST_BITS_PER_WIDE_INT == 64" test.
3095685e 17728 (num_insns_constant): Ditto. Delete CONST_DOUBLE DImode/VOIDmode case.
da226db2
SB		 17729 (build_mask64_2_operands): Delete "HOST_BITS_PER_WIDE_INT >= 64" test.
17730 (rs6000_emit_set_const): Delete CONST_DOUBLE case.
17731 (rs6000_emit_set_long_const): Delete "HOST_BITS_PER_WIDE_INT >= 64"
17732 test.
17733 (includes_rldic_lshift_p, includes_rldicr_lshift_p): Delete
17734 CONST_DOUBLE DImode/VOIDmode case.
17735 (INT_P, INT_LOWPART): Delete CONST_DOUBLE case.
17736 (print_operand): Delete "HOST_BITS_PER_WIDE_INT == 32" case. Delete
17737 CONST_DOUBLE VOIDmode case.
17738 (output_toc): Delete "HOST_BITS_PER_WIDE_INT == 32" case.
17739 (rs6000_rtx_costs): Delete CONST_DOUBLE DImode/VOIDmode case.
17740 * config/rs6000/rs6000.md (iordi3, xordi3, splitter for these):
17741 Delete CONST_DOUBLE case.
17742 (splitters for mov FMOVE64 const_double): Delete
17743 "HOST_BITS_PER_WIDE_INT == 32" case. Delete
17744 "HOST_BITS_PER_WIDE_INT >= 64" test.
17745 (splitter for mov DI const_int): Delete "HOST_BITS_PER_WIDE_INT == 32"
17746 case.
17747 (mov DI const_double): Delete.
17748
40de22d6
JJ		 177492013-05-04 Jakub Jelinek <jakub@redhat.com>
17750
17751 * combine.c (combine_simplify_rtx) <case SUBREG>: If nonzero_bits
17752 on op shows all bits zero in mode of a lowpart subreg, return zero.
17753
5ec6aff2
MM		 177542013-05-03 Michael Meissner <meissner@linux.vnet.ibm.com>
17755
17756 PR target/57150
17757 * config/rs6000/rs6000.h (HARD_REGNO_CALLER_SAVE_MODE): Use DFmode
17758 to save TFmode registers and DImode to save TImode registers for
17759 caller save operations.
17760 (HARD_REGNO_CALL_PART_CLOBBERED): TFmode and TDmode do not need to
17761 mark being partially clobbered since they only use the first
17762 double word.
17763
17764 * config/rs6000/rs6000.c (rs6000_init_hard_regno_mode_ok): TFmode
17765 and TDmode only use the upper 64-bits of each VSX register.
17766
2cefad90
BS		 177672013-05-03 Bill Schmidt <wschmidt@linux.vnet.ibm.com>
17768
17769 * gimple-ssa-strength-reduction.c (slsr_process_phi): Disable.
17770 (find_candidates_in_block): Disable slsr_process_phi.
17771
d6d7eee1
GW		 177722013-05-03 Guozhi Wei <carrot@google.com>
17773
17774 * coverage.c (coverage_obj_init): Move the construction of gcov
3095685e
UB		 17775 constructor to ...
17776 (build_init_ctor): ... here.
d6d7eee1 17777
9b92d12b
BS		 177782013-05-03 Bill Schmidt <wschmidt@linux.vnet.ibm.com>
17779
17780 * gimple-ssa-strength-reduction.c (cand_kind): Add CAND_PHI.
17781 (slsr_cand_d): Redefine def_phi.
17782 (stride_status, phi_adjust_status, count_phis_status): New enums.
17783 (find_phi_def): New.
17784 (find_basis_for_base_expr): New.
17785 (find_basis_for_candidate): Handle hidden bases.
17786 (alloc_cand_and_find_basis): Handle phi candidates.
17787 (slsr_process_phi): New.
17788 (create_mul_ssa_cand): Exclude phi base candidates; use integer_onep.
17789 (create_mul_imm_cand): Likewise.
17790 (create_add_ssa_cand): Exclude phi base candidates.
17791 (create_add_imm_cand): Likewise.
17792 (slsr_process_cast): Likewise.
17793 (slsr_process_copy): Likewise.
17794 (find_candidates_in_block): Handle phi candidates.
17795 (dump_candidate): Likewise.
17796 (unconditional_cands): Delete.
17797 (unconditional_cands_with_known_stride_p): Delete.
17798 (phi_dependent_cand_p): New.
17799 (cand_increment): Handle phi-dependent candidates.
17800 (replace_dependent): Delete.
17801 (replace_mult_candidate): New.
17802 (replace_unconditional_candidate): New.
17803 (incr_vec_index): Move to avoid forward reference.
17804 (create_add_on_incoming_edge): New.
17805 (create_phi_basis): New.
17806 (replace_dependents): Delete.
17807 (replace_conditional_candidate): New.
17808 (phi_add_costs): New.
17809 (replace_uncond_cands_and_profitable_phis): New.
17810 (record_increment): Handle phi adjustments.
17811 (record_phi_increments): New.
17812 (record_increments): Handle phi adjustments.
17813 (phi_incr_cost): New.
17814 (lowest_cost_path): Handle phis.
17815 (total_savings): Likewise.
17816 (analyze_increments): Likewise.
17817 (ncd_with_phi): New.
17818 (ncd_of_cand_and_phis): New.
17819 (nearest_common_dominator_for_cands): Handle phi increments.
17820 (all_phi_incrs_profitable): New.
17821 (replace_profitable_candidates): Handle phi-dependent candidates.
17822 (analyze_candidates_and_replace): Likewise.
17823
68f073d4
TJ		 178242013-05-03 Teresa Johnson <tejohnson@google.com>
17825
17826 PR bootstrap/57154
17827 * sched-rgn.c (compute_dom_prob_ps): Ensure accumulated probabilities
17828 do not exceed REG_BR_PROB_BASE.
17829
a4ee7cb9
JL		 178302013-05-03 Jeff Law <law@redhat.com>
17831
ade67f70 17832 PR tree-optimization/57144
a4ee7cb9
JL		 17833 * tree-vrp.c (simplify_cond_using_ranges): Verify the constant
17834 operand of the condition will bit into the new type when eliminating
17835 a cast feeding a condition.
17836
47954c4e
JJ		 178372013-05-03 Jakub Jelinek <jakub@redhat.com>
17838
17839 PR rtl-optimization/57130
3095685e
UB		 17840 * combine.c (make_compound_operation) <case SUBREG>: Pass SET instead
17841 of COMPARE as in_code to the recursive call if needed.
47954c4e 17842
3c21604f
UB		 178432013-05-03 Uros Bizjak <ubizjak@gmail.com>
17844
17845 * config/i386/i386.md (isa): Add x64_sse4_noavx and x64_avx members.
17846 (enabled): Handle new members.
17847 * config/i386/sse.md (*vec_concatv2si): Merge from
17848 *vec_concatv2si_sse2 and vec_concatv2si_sse.
17849 (vec_concatv2di): Merge with *vec_concatv2di_rex64.
17850
12211b99 178512013-05-03 Joern Rennecke <joern.rennecke@embecosm.com>
95c03b36
JR		 17852
17853 PR tree-optimization/57027
17854 * tree-ssa-math-opts.c (convert_mult_to_fma): When checking
17855 for fnms opportunity, check we got the prerequisite kind
17856 of tree / gimple before using accessor functions.
17857
07bfc9ec
RB		 178582013-05-03 Richard Biener <rguenther@suse.de>
17859
17860 * double-int.h (lshift): New overload without precision
17861 and arith argument.
17862 (operator *=, operator +=, operator -=): Move ...
17863 * double-int.c (operator *=, operator +=, operator -=): ... here
17864 and implement more efficiently.
17865 (mul_double_with_sign): Remove.
17866 (lshift_double): Adjust to take unsinged shift argument, push
17867 dispatching code to callers.
17868 (mul_double_wide_with_sign): Add early out for callers that
17869 are not interested in high parts or overflow.
17870 (lshift): New function.
17871 (lshift, rshift, alshift, arshift, llshift, lrshift): Add
17872 dispatch code here.
17873 (lrotate, rrotate): Use logical shifts.
17874 * expr.c (get_inner_reference): Use lshift.
17875 * fixed-value.c (do_fixed_divide): Likewise.
17876 * tree-dfa.c (get_ref_base_and_extent): Likewise.
17877 * tree-ssa-alias.c (indirect_ref_may_alias_decl_p): Likewise.
17878 (indirect_refs_may_alias_p): Likewise.
17879 (stmt_kills_ref_p_1): Likewise.
17880
7769bb64
VP		 178812013-05-03 Vidya Praveen <vidyapraveen@arm.com>
17882
17883 * config/aarch64/aarch64-simd.md (simd_fabd): Correct the description.
17884
1d0c8e5c
VP		 178852013-05-03 Vidya Praveen <vidyapraveen@arm.com>
17886
17887 * config/aarch64/aarch64-simd.md (*fabd_scalar<mode>3): Support
17888 scalar form of FABD instruction.
17889
f15feaf9
VM		 178902013-05-02 Vladimir Makarov <vmakarov@redhat.com>
17891
17892 * lra-constraints.c (process_alt_operands): Add checking alt
17893 number to choose the best alternative.
17894
d90e76d4
RB		 178952013-05-02 Richard Biener <rguenther@suse.de>
17896
17897 * tree-eh.c (cleanup_empty_eh_merge_phis): Remove rename_virts
17898 bitmap and its handling.
17899 (pass_cleanup_eh): Set todo_flags_finish to TODO_verify_ssa.
17900
9f8e7a96
RB		 179012013-05-02 Richard Biener <rguenther@suse.de>
17902
17903 PR middle-end/57140
17904 * tree-inline.c (copy_loops): Properly handle removed loops.
17905 (copy_cfg_body): Mark destination loops for fixup if source
17906 loops needed fixup.
17907
f3a81b39
GY		 179082013-05-02 Greta Yorsh <Greta.Yorsh@arm.com>
17909
17910 PR target/56732
17911 * config/arm/arm.c (arm_expand_epilogue): Check really_return before
17912 generating simple_return for naked functions.
17913
7b920a9a
MJ		 179142013-05-02 Martin Jambor <mjambor@suse.cz>
17915
17916 PR middle-end/56988
17917 * ipa-prop.h (ipa_agg_replacement_value): New flag by_ref.
17918 * ipa-cp.c (ipa_get_indirect_edge_target_1): Also check that by_ref
17919 flags match.
17920 (find_aggregate_values_for_callers_subset): Fill in the by_ref flag of
17921 ipa_agg_replacement_value structures.
17922 (known_aggs_to_agg_replacement_list): Likewise.
17923 * ipa-prop.c (write_agg_replacement_chain): Stream by_ref flag.
17924 (read_agg_replacement_chain): Likewise.
17925 (ipcp_transform_function): Also check that by_ref flags match.
17926
2c41c19d
RB		 179272013-05-02 Richard Biener <rguenther@suse.de>
17928
17929 * graphds.h (struct graph): Add obstack member.
17930 * graphds.c (new_graph): Initialize obstack and allocate
17931 vertices from it.
17932 (add_edge): Allocate edge from the obstack.
3c21604f 17933 (free_graph): Free the obstack instead of all edges and vertices.
2c41c19d 17934
8b47039c
TJ		 179352013-05-02 Teresa Johnson <tejohnson@google.com>
17936
17937 * loop-unswitch.c (unswitch_loop): Use helper routines with rounding
17938 divides.
17939 * cfg.c (update_bb_profile_for_threading): Ditto.
17940 * tree-inline.c (copy_bb): Ditto.
17941 (copy_edges_for_bb): Ditto.
17942 (initialize_cfun): Ditto.
17943 (copy_cfg_body): Ditto.
17944 (expand_call_inline): Ditto.
17945 * ipa-inline-analysis.c (estimate_edge_size_and_time): Ditto.
17946 (estimate_node_size_and_time): Ditto.
17947 (inline_merge_summary): Ditto.
17948 * cgraphclones.c (cgraph_clone_edge): Ditto.
17949 (cgraph_clone_node): Ditto.
17950 * sched-rgn.c (compute_dom_prob_ps): Ditto.
17951 (compute_trg_info): Ditto.
17952
da65928c
IB		 179532013-05-02 Ian Bolton <ian.bolton@arm.com>
17954
17955 * config/aarch64/aarch64.md (movsi_aarch64): Only allow to/from
17956 S reg when fp attribute set.
17957 (movdi_aarch64): Only allow to/from D reg when fp attribute set.
17958
add3c965
IB		 179592013-05-02 Ian Bolton <ian.bolton@arm.com>
17960
17961 * config/aarch64/aarch64.md (*and_one_cmpl<mode>3_compare0):
17962 New pattern.
17963 (*and_one_cmplsi3_compare0_uxtw): Likewise.
17964 (*and_one_cmpl_<SHIFT:optab><mode>3_compare0): Likewise.
17965 (*and_one_cmpl_<SHIFT:optab>si3_compare0_uxtw): Likewise.
17966
fdd43ac4
RB		 179672013-05-02 Richard Biener <rguenther@suse.de>
17968
17969 * tree-scalar-evolution.c (scev_info_hasher): Remove.
17970 (struct instantiate_cache_entry): New type.
17971 (struct instantiate_cache_entry_hasher): New hashtable descriptor.
17972 (struct instantiate_cache_type): New type.
17973 (set_instantiated_value, get_instantiated_value): Remove.
17974 (get_instantiated_value_entry): New function.
17975 (instantiate_scev_name): Use the new cache and adjust.
17976 (instantiate_scev_poly): Adjust.
17977 (instantiate_scev_binary): Likewise.
17978 (instantiate_array_ref): Likewise.
17979 (instantiate_scev_convert): Likewise.
17980 (instantiate_scev_not): Likewise.
17981 (instantiate_scev_3): Likewise.
17982 (instantiate_scev_2): Likewise.
17983 (instantiate_scev_r): Likewise.
17984 (instantiate_scev): Likewise.
17985 (resolve_mixers): Likewise.
17986
36ff9dfb
VM		 179872013-05-01 Vladimir Makarov <vmakarov@redhat.com>
17988
17989 PR target/57091
17990 * lra-constraints.c (best_small_class_operands_num): Remove.
17991 (process_alt_operands): Remove small_class_operands_num. Take
17992 small classes operands into losers and only if the operand is not
17993 matched. Modify debugging output.
17994 (curr_insn_transform): Remove best_small_class_operands_num.
17995 Print insn name.
17996
36054fab
JG		 179972013-05-01 James Greenhalgh <james.greenhalgh@arm.com>
17998
17999 * config/aarch64/aarch64-builtins.c
18000 (aarch64_gimple_fold_builtin.c): Fold more modes for reduc_splus_.
18001 * config/aarch64/aarch64-simd-builtins.def
18002 (reduc_splus_): Add new modes.
18003 (reduc_uplus_): New.
18004 * config/aarch64/aarch64-simd.md (aarch64_addvv4sf): Remove.
18005 (reduc_uplus_v4sf): Likewise.
18006 (reduc_splus_v4sf): Likewise.
18007 (aarch64_addv<mode>): Likewise.
18008 (reduc_uplus_<mode>): Likewise.
18009 (reduc_splus_<mode>): Likewise.
18010 (aarch64_addvv2di): Likewise.
18011 (reduc_uplus_v2di): Likewise.
18012 (reduc_splus_v2di): Likewise.
18013 (aarch64_addvv2si): Likewise.
18014 (reduc_uplus_v2si): Likewise.
18015 (reduc_splus_v2si): Likewise.
18016 (reduc_<sur>plus_<mode>): New.
18017 (reduc_<sur>plus_v2di): Likewise.
18018 (reduc_<sur>plus_v2si): Likewise.
18019 (reduc_<sur>plus_v4sf): Likewise.
18020 (aarch64_addpv4sf): Likewise.
18021 * config/aarch64/arm_neon.h
18022 (vaddv<q>_<s,u,f><8, 16, 32, 64): Rewrite using builtins.
18023 * config/aarch64/iterators.md (unspec): Remove UNSPEC_ADDV,
18024 add UNSPEC_SADDV, UNSPEC_UADDV.
18025 (SUADDV): New.
18026 (sur): Add UNSPEC_SADDV, UNSPEC_UADDV.
18027
6dce23a8
JG		 180282013-05-01 James Greenhalgh <james.greenhalgh@arm.com>
18029
18030 * config/aarch64/arm_neon.h
18031 (v<max,min><nm><q><v>_<sfu><8, 16, 32, 64>): Rewrite using builtins.
18032
1598945b
JG		 180332013-05-01 James Greenhalgh <james.greenhalgh@arm.com>
18034
18035 * config/aarch64/aarch64-builtins
18036 (aarch64_gimple_fold_builtin): Fold reduc_<su><maxmin>_ builtins.
18037
998eaf97
JG		 180382013-05-01 James Greenhalgh <james.greenhalgh@arm.com>
18039
18040 * config/aarch64/aarch64-simd-builtins.def
18041 (reduc_smax_): New.
18042 (reduc_smin_): Likewise.
18043 (reduc_umax_): Likewise.
18044 (reduc_umin_): Likewise.
18045 (reduc_smax_nan_): Likewise.
18046 (reduc_smin_nan_): Likewise.
18047 (fmax): Remove.
18048 (fmin): Likewise.
18049 (smax): Update for V2SF, V4SF and V2DF modes.
18050 (smin): Likewise.
18051 (smax_nan): New.
18052 (smin_nan): Likewise.
18053 * config/aarch64/aarch64-simd.md (<maxmin><mode>3): Rename to...
18054 (<su><maxmin><mode>3): ...This, refactor.
18055 (s<maxmin><mode>3): New.
18056 (<maxmin_uns><mode>3): Likewise.
18057 (reduc_<maxmin_uns>_<mode>): Refactor.
18058 (reduc_<maxmin_uns>_v4sf): Likewise.
18059 (reduc_<maxmin_uns>_v2si): Likewise.
18060 (aarch64_<fmaxmin><mode>: Remove.
18061 * config/aarch64/arm_neon.h (vmax<q>_f<32,64>): Rewrite to use
18062 new builtin names.
18063 (vmin<q>_f<32,64>): Likewise.
18064 * config/iterators.md (unspec): Add UNSPEC_FMAXNMV, UNSPEC_FMINNMV.
18065 (FMAXMIN): New.
18066 (su): Add mappings for smax, smin, umax, umin.
18067 (maxmin): New.
18068 (FMAXMINV): Add UNSPEC_FMAXNMV, UNSPEC_FMINNMV.
18069 (FMAXMIN): Rename as...
18070 (FMAXMIN_UNS): ...This.
18071 (maxminv): Remove.
18072 (fmaxminv): Likewise.
18073 (fmaxmin): Likewise.
18074 (maxmin_uns): New.
18075 (maxmin_uns_op): Likewise.
18076
bd11644e
JG		 180772013-05-01 James Greenhalgh <james.greenhalgh@arm.com>
18078
18079 * config/aarch64/arm_neon.h
18080 (vac<ge, gt><sd>_f<32, 64>): Rename to...
18081 (vca<ge, gt><sd>_f<32, 64>): ...this, reimpliment in C.
18082 (vca<ge, gt, lt, le><q>_f<32, 64>): Reimpliment in C.
18083
75dd5ace
JG		 180842013-05-01 James Greenhalgh <james.greenhalgh@arm.com>
18085
18086 * config/aarch64/aarch64-simd.md (*aarch64_fac<optab><mode>): New.
18087 * config/aarch64/iterators.md (FAC_COMPARISONS): New.
18088
7c19979f
JG		 180892013-05-01 James Greenhalgh <james.greenhalgh@arm.com>
18090
18091 * config/aarch64/aarch64-simd.md
18092 (vcond<mode>_internal): Handle special cases for constant masks.
18093 (vcond<mode><mode>): Allow nonmemory_operands for outcome vectors.
18094 (vcondu<mode><mode>): Likewise.
18095 (vcond<v_cmp_result><mode>): New.
18096
bb60efd9
JG		 180972013-05-01 James Greenhalgh <james.greenhalgh@arm.com>
18098
18099 * config/aarch64/aarch64-builtins.c (BUILTIN_VALLDI): Define.
18100 (aarch64_fold_builtin): Add folding for cm<eq,ge,gt,tst>.
18101 * config/aarch64/aarch64-simd-builtins.def
18102 (cmeq): Update to BUILTIN_VALLDI.
18103 (cmgt): Likewise.
18104 (cmge): Likewise.
18105 (cmle): Likewise.
18106 (cmlt): Likewise.
18107 * config/aarch64/arm_neon.h
18108 (vc<eq, lt, le, gt, ge, tst><z><qsd>_<fpsu><8,16,32,64>): Remap
18109 to builtins or C as appropriate.
18110
889b9412
JG		 181112013-05-01 James Greenhalgh <james.greenhalgh@arm.com>
18112
18113 * config/aarch64/aarch64-simd-builtins.def (cmhs): Rename to...
18114 (cmgeu): ...This.
18115 (cmhi): Rename to...
18116 (cmgtu): ...This.
18117 * config/aarch64/aarch64-simd.md
18118 (simd_mode): Add SF.
18119 (aarch64_vcond_internal): Use new names for unsigned comparison insns.
18120 (aarch64_cm<optab><mode>): Rewrite to not use UNSPECs.
18121 * config/aarch64/aarch64.md (*cstore<mode>_neg): Rename to...
18122 (cstore<mode>_neg): ...This.
18123 * config/aarch64/iterators.md
18124 (VALLF): new.
18125 (unspec): Remove UNSPEC_CM<EQ, LE, LT, GE, GT, HS, HI, TST>.
18126 (COMPARISONS): New.
18127 (UCOMPARISONS): Likewise.
18128 (optab): Add missing comparisons.
18129 (n_optab): New.
18130 (cmp_1): Likewise.
18131 (cmp_2): Likewise.
18132 (CMP): Likewise.
18133 (cmp): Remove.
18134 (VCMP_S): Likewise.
18135 (VCMP_U): Likewise.
18136 (V_cmp_result): Add DF, SF modes.
18137 (v_cmp_result): Likewise.
18138 (v): Likewise.
18139 (vmtype): Likewise.
18140 * config/aarch64/predicates.md (aarch64_reg_or_fp_zero): New.
18141
0a7dbb76
GY		 181422013-05-01 Greta Yorsh <Greta.Yorsh@arm.com>
18143
18144 * config/arm/thumb2.md (thumb2_smaxsi3,thumb2_sminsi3): Convert
18145 define_insn to define_insn_and_split.
18146 (thumb32_umaxsi3,thumb2_uminsi3): Likewise.
18147 (thumb2_negdi2,thumb2_abssi2,thumb2_neg_abssi2): Likewise.
18148 (thumb2_mov_scc,thumb2_mov_negscc,thumb2_mov_notscc): Likewise.
18149 (thumb2_movsicc_insn,thumb2_and_scc,thumb2_ior_scc): Likewise.
18150 (thumb2_negscc): Likewise.
18151
fb614ca6
GY		 181522013-04-30 Greta Yorsh <Greta.Yorsh@arm.com>
18153
18154 * config/arm/thumb2.md (thumb2_incscc, thumb2_decscc): Delete.
18155
9e64a0bf
GY		 181562013-04-30 Greta Yorsh <Greta.Yorsh@arm.com>
18157
18158 * config/arm/thumb2.md: Remove trailing whitespaces.
18159
d6b28156
RS		 181602013-04-30 Richard Sandiford <rsandifo@linux.vnet.ibm.com>
18161
18162 * explow.c (plus_constant): Pass "mode" to immed_double_int_const.
18163 Use gen_int_mode rather than GEN_INT.
18164
f91674c3
L		 181652013-04-30 H.J. Lu <hongjiu.lu@intel.com>
18166
b0dec607 18167 * value-prof.c (stream_in_histogram_value): Remove the strayed
f91674c3
L		 18168 debug_gimple_stmt.
18169
3551257c
RB		 181702013-04-30 Richard Biener <rguenther@suse.de>
18171
18172 PR middle-end/57122
3c21604f 18173 * cfghooks.c (split_edge): Properly check for the loop latch edge.
3551257c 18174
4c1aff1c
RB		 181752013-04-30 Richard Biener <rguenther@suse.de>
18176
18177 PR middle-end/57107
18178 * tree-eh.c (sink_clobbers): Preserve virtual SSA form.
18179
54b8379a
AB		 181802013-04-30 Andrey Belevantsev <abel@ispras.ru>
18181
18182 PR rtl-optimization/56957
18183 PR rtl-optimization/57105
54b8379a
AB		 18184 * sel-sched.c (move_op_orig_expr_found): Remove insn_emitted
18185 variable. Use just INSN_UID for determining whether an insn
18186 should be only disconnected from the insn stream.
18187 * sel-sched-ir.h (EXPR_WAS_CHANGED): Remove.
18188
abc27962
JJ		 181892013-04-30 Jakub Jelinek <jakub@redhat.com>
18190
18191 PR tree-optimization/57104
18192 * tsan.c (instrument_expr): Don't instrument accesses to
18193 DECL_HARD_REGISTER VAR_DECLs.
18194
0fc822d0
RB		 181952013-04-30 Richard Biener <rguenther@suse.de>
18196
18197 * function.h (loops_for_fn): New inline function.
18198 (set_loops_for_fn): Likewise.
18199 * cfgloop.h (place_new_loop): Add struct function parameter.
18200 (get_loop): Likewise.
18201 (get_loops): Likewise.
18202 (number_of_loops): Likewise.
18203 (fel_next): Adjust.
18204 (fel_init): Likewise.
18205 * cfg.c (get_loop_copy): Adjust.
18206 * cfgloop.c (flow_loops_dump): Likewise.
18207 (record_loop_exits): Likewise.
18208 (verify_loop_structure): Likewise.
18209 * cfgloopanal.c (mark_irreducible_loops): Likewise.
18210 (estimate_reg_pressure_cost): Likewise.
18211 (mark_loop_exit_edges): Likewise.
18212 * cfgloopmanip.c (place_new_loop): Likewise.
18213 (add_loop): Likewise.
18214 (duplicate_loop): Likewise.
18215 * graph.c (draw_cfg_nodes): Likewise.
18216 * graphite-clast-to-gimple.c (translate_clast_user): Likewise.
18217 * graphite-sese-to-poly.c (build_scop_scattering): Likewise.
18218 (extract_affine_chrec): Likewise.
18219 (build_scop_iteration_domain): Likewise.
18220 * graphite.c (graphite_initialize): Likewise.
18221 * ira-build.c (create_loop_tree_nodes): Likewise.
18222 (more_one_region_p): Likewise.
18223 (rebuild_regno_allocno_maps): Likewise.
18224 (mark_loops_for_removal): Likewise.
18225 (mark_all_loops_for_removal): Likewise.
18226 (remove_unnecessary_regions): Likewise.
18227 (ira_build): Likewise.
18228 * ira-emit.c (setup_entered_from_non_parent_p): Likewise.
18229 * loop-init.c (fix_loop_structure): Likewise.
18230 (gate_rtl_move_loop_invariants): Likewise.
18231 (gate_rtl_unswitch): Likewise.
18232 (gate_rtl_unroll_and_peel_loops): Likewise.
18233 (rtl_doloop): Likewise.
18234 * lto-streamer-in.c (input_cfg): Likewise.
18235 * lto-streamer-out.c (output_cfg): Likewise.
18236 * modulo-sched.c (sms_schedule): Likewise.
18237 * predict.c (tree_estimate_probability): Likewise.
18238 (tree_estimate_probability_driver): Likewise.
18239 (estimate_loops): Likewise.
18240 * tree-cfg.c (fixup_loop_arrays_after_move): Likewise.
18241 (move_sese_region_to_fn): Likewise.
18242 (debug_loop_num): Likewise.
18243 * tree-chrec.c (chrec_evaluate): Likewise.
18244 (hide_evolution_in_other_loops_than_loop): Likewise.
18245 (chrec_component_in_loop_num): Likewise.
18246 (reset_evolution_in_loop): Likewise.
18247 (evolution_function_is_invariant_rec_p): Likewise.
18248 * tree-if-conv.c (main_tree_if_conversion): Likewise.
18249 * tree-inline.c (copy_loops): Likewise.
18250 (copy_cfg_body): Likewise.
18251 (tree_function_versioning): Likewise.
18252 * tree-loop-distribution.c (rdg_flag_loop_exits): Likewise.
18253 * tree-scalar-evolution.c (chrec_contains_symbols_defined_in_loop):
18254 Likewise.
18255 (add_to_evolution_1): Likewise.
18256 (scev_const_prop): Likewise.
18257 * tree-scalar-evolution.h (get_chrec_loop): Likewise.
18258 * tree-ssa-loop-ch.c (copy_loop_headers): Likewise.
18259 * tree-ssa-loop-im.c (analyze_memory_references): Likewise.
18260 (tree_ssa_lim_initialize): Likewise.
18261 * tree-ssa-loop-manip.c (rewrite_into_loop_closed_ssa): Likewise.
18262 (verify_loop_closed_ssa): Likewise.
18263 * tree-ssa-loop.c (tree_ssa_loop_init): Likewise.
18264 (tree_ssa_loop_im): Likewise.
18265 (tree_ssa_loop_unswitch): Likewise.
18266 (tree_vectorize): Likewise.
18267 (check_data_deps): Likewise.
18268 (tree_ssa_loop_ivcanon): Likewise.
18269 (tree_ssa_loop_bounds): Likewise.
18270 (tree_complete_unroll): Likewise.
18271 (tree_complete_unroll_inner): Likewise.
18272 (tree_parallelize_loops): Likewise.
18273 (tree_ssa_loop_prefetch): Likewise.
18274 (tree_ssa_loop_ivopts): Likewise.
18275 * tree-ssa.c (execute_update_addresses_taken): Liekwise.
18276 * tree-vectorizer.c (vectorize_loops): Likewise.
18277
37953bd3
MF		 182782013-04-29 Mike Frysinger <vapier@gentoo.org>
18279
18280 * config/arm/bpabi.h (EABI_LINK_SPEC): Define.
18281 (BPABI_LINK_SPEC): Use new EABI_LINK_SPEC.
18282 * config/arm/linux-eabi.h (LINK_SPEC): Replace BE8_LINK_SPEC
18283 with EABI_LINK_SPEC.
18284
f9ed28db
UB		 182852013-04-29 Uros Bizjak <ubizjak@gmail.com>
18286
18287 PR target/44578
18288 * config/i386/i386.md (*zero_extendsidi2): Add "!" to m->?*y
18289 alternative.
18290
deca73f5
VM		 182912013-04-29 Vladimir Makarov <vmakarov@redhat.com>
18292
18293 PR target/57097
37953bd3 18294 * lra-constraints.c (process_alt_operands): Discourage a bit more
deca73f5
VM		 18295 using memory for pseudos. Print cost dump for alternatives.
18296 Modify cost values for conflicts with early clobbers.
18297 (curr_insn_transform): Spill pseudos reassigned to NO_REGS.
18298
fdca7d03
UB		 182992013-04-29 Uros Bizjak <ubizjak@gmail.com>
18300
18301 PR target/57098
18302 * config/i386/i386.c (ix86_expand_vec_perm): Validize constant memory.
18303
0b064172
IB		 183042013-04-29 Ian Bolton <ian.bolton@arm.com>
18305
18306 * config/aarch64/aarch64.md (movsi_aarch64): Support LDR/STR
18307 from/to S register.
18308 (movdi_aarch64): Support LDR/STR from/to D register.
18309
473cec55
IB		 183102013-04-29 Ian Bolton <ian.bolton@arm.com>
18311
18312 * common/config/aarch64/aarch64-common.c: Enable REE pass at O2
18313 or higher by default.
18314
a6f30e66
RB		 183152013-04-29 Richard Biener <rguenther@suse.de>
18316
18317 PR middle-end/57075
18318 * tree-inline.c (copy_edges_for_bb): Still split the bbs,
18319 even if not adding abnormal edges for calls that can make
18320 abnormal gotos.
18321
0c2b2040
RB		 183222013-04-29 Richard Biener <rguenther@suse.de>
18323
18324 PR middle-end/57103
18325 * tree-cfg.c (move_stmt_op): Fix condition under which to update
18326 TREE_BLOCK.
18327 (move_stmt_r): Remove redundant checking.
18328
f41f80f9
TJ		 183292013-04-29 Teresa Johnson <tejohnson@google.com>
18330
18331 PR bootstrap/57077
18332 * basic-block.h (apply_scale): New function.
18333 (apply_probability): Use apply_scale.
18334 * gimple-streamer-in.c (input_bb): Ditto.
18335 * lto-streamer-in.c (input_cfg): Ditto.
18336 * lto-cgraph.c (merge_profile_summaries): Ditto.
18337 * tree-optimize.c (execute_fixup_cfg): Ditto.
fdca7d03 18338 * tree-inline.c (copy_bb): Update comment to use apply_scale.
f41f80f9
TJ		 18339 (copy_edges_for_bb): Ditto.
18340 (copy_cfg_body): Ditto.
18341
315bbd2e
TV		 183422013-04-29 Tom de Vries <tom@codesourcery.com>
18343
18344 * tree-ssa-tail-merge.c (find_same_succ_bb): Skip loop latch bbs.
18345 (replace_block_by): Don't set LOOPS_NEED_FIXUP.
18346 (tail_merge_optimize): Handle current_loops == NULL.
18347
ebbd90d8
JL		 183482013-04-26 Jeff Law <law@redhat.com>
18349
18350 * tree-vrp.c (range_fits_type_p): Move to earlier point in file.
18351 (simplify_cond_using_ranges): Generalize code to simplify
18352 COND_EXPRs where one argument is a constant and the other
18353 is an SSA_NAME created by an integral type conversion.
18354
8b9b57eb
KT		 183552013-04-29 Kyrylo Tkachov <kyrylo.tkachov@arm.com>
18356
18357 * config/arm/arm.md (store_minmaxsi): Use only when
18358 optimize_insn_for_size_p.
18359
9498e5dc
CB		 183602013-04-29 Christian Bruel <christian.bruel@st.com>
18361
18362 PR target/57108
18363 * sh.md (tstsi_t_zero_extract_eq): Set mode for operand 0.
18364
6093bc06
RB		 183652013-04-29 Richard Biener <rguenther@suse.de>
18366
18367 PR middle-end/57089
fdca7d03
UB		 18368 * omp-low.c (expand_omp_taskreg): If the parent function had a broken
18369 loop tree make sure to schedule a fixup for the child as well.
6093bc06
RB		 18370 (expand_omp_for_generic): Properly add loops.
18371 (expand_omp_for_static_nochunk): Likewise.
18372 (expand_omp_for_static_chunk): Likewise.
18373 (expand_omp_for): For the degenerate case fixup loops.
18374 (expand_omp_sections): Fix default bb placement in loops.
18375 (expand_omp_atomic_pipeline): Properly add loops.
18376
84aacbfd
KT		 183772013-04-29 Kyrylo Tkachov <kyrylo.tkachov@arm.com>
18378
18379 * predict.c: Fix typo in comment above #define PROB_VERY_UNLIKELY.
18380
8dee4479
TV		 183812013-04-29 Tom de Vries <tom@codesourcery.com>
18382
18383 * tree-ssa-tail-merge.c: Update header comment.
18384
47934dc4
JG		 183852013-04-29 James Greenhalgh <james.greenhalgh@arm.com>
18386
18387 * config/aarch64/arm_neon.h
18388 (vcvt<sd>_f<32,64>_s<32,64>): Rewrite in C.
18389 (vcvt<q>_f<32,64>_s<32,64>): Rewrite using builtins.
18390 (vcvt_<high_>_f<32,64>_f<32,64>): Likewise.
18391 (vcvt<qsd>_<su><32,64>_f<32,64>): Likewise.
18392 (vcvta<qsd>_<su><32,64>_f<32,64>): Likewise.
18393 (vcvtm<qsd>_<su><32,64>_f<32,64>): Likewise.
18394 (vcvtn<qsd>_<su><32,64>_f<32,64>): Likewise.
18395 (vcvtp<qsd>_<su><32,64>_f<32,64>): Likewise.
18396
384be29f
JG		 183972013-04-29 James Greenhalgh <james.greenhalgh@arm.com>
18398
18399 * config/aarch64/aarch64-simd.md
18400 (<optab><VDQF:mode><fcvt_target>2): New, maps to fix, fixuns.
18401 (<fix_trunc_optab><VDQF:mode><fcvt_target>2): New, maps to
18402 fix_trunc, fixuns_trunc.
18403 (ftrunc<VDQF:mode>2): New.
18404 * config/aarch64/iterators.md (optab): Add fix, fixuns.
18405 (fix_trunc_optab): New.
18406
0386b123
JG		 184072013-04-29 James Greenhalgh <james.greenhalgh@arm.com>
18408
18409 * config/aarch64/aarch64-builtins.c
18410 (aarch64_builtin_vectorized_function): Vectorize over ifloorf,
18411 iceilf, lround, iroundf.
18412
00fcb892
UB		 184132013-04-29 Uros Bizjak <ubizjak@gmail.com>
18414
18415 PR target/54349
18416 * config/i386/i386.h (enum ix86_tune_indices)
18417 <X86_TUNE_INTER_UNIT_MOVES_TO_VEC, X86_TUNE_INTER_UNIT_MOVES_FROM_VEC>:
18418 New, split from X86_TUNE_INTER_UNIT_MOVES.
18419 <X86_TUNE_INTER_UNIT_MOVES>: Remove.
18420 (TARGET_INTER_UNIT_MOVES_TO_VEC): New define.
18421 (TARGET_INTER_UNIT_MOVES_FROM_VEC): Ditto.
18422 (TARGET_INTER_UNIT_MOVES): Remove.
18423 * config/i386/i386.c (initial_ix86_tune_features): Update.
18424 Disable X86_TUNE_INTER_UNIT_MOVES_FROM_VEC for m_ATHLON_K8 only.
18425 (ix86_expand_convert_uns_didf_sse): Use
18426 TARGET_INTER_UNIT_MOVES_TO_VEC instead of TARGET_INTER_UNIT_MOVES.
18427 (ix86_expand_vector_init_one_nonzero): Ditto.
18428 (ix86_expand_vector_init_interleave): Ditto.
18429 (inline_secondary_memory_needed): Return true for moves from SSE class
18430 registers for !TARGET_INTER_UNIT_MOVES_FROM_VEC targets and for moves
18431 to SSE class registers for !TARGET_INTER_UNIT_MOVES_TO_VEC targets.
18432 * config/i386/constraints.md (Yi, Ym): Depend on
18433 TARGET_INTER_UNIT_MOVES_TO_VEC.
18434 (Yj, Yn): New constraints.
18435 * config/i386/i386.md (*movdi_internal): Change constraints of
18436 operand 1 from Yi to Yj and from Ym to Yn.
18437 (*movsi_internal): Ditto.
18438 (*movdf_internal): Ditto.
18439 (*movsf_internal): Ditto.
18440 (*float<SWI48x:mode><X87MODEF:mode>2_1): Use
18441 TARGET_INTER_UNIT_MOVES_TO_VEC instead of TARGET_INTER_UNIT_MOVES.
18442 (*float<SWI48x:mode><X87MODEF:mode>2_1 splitters): Ditto.
18443 (floatdi<X87MODEF:mode>2_i387_with_xmm): Ditto.
18444 (floatdi<X87MODEF:mode>2_i387_with_xmm splitters): Ditto.
18445 * config/i386/sse.md (movdi_to_sse): Ditto.
18446 (sse2_stored): Change constraint of operand 1 from Yi to Yj.
18447 Use TARGET_INTER_UNIT_MOVES_FROM_VEC instead of
18448 TARGET_INTER_UNIT_MOVES.
18449 (sse_storeq_rex64): Change constraint of operand 1 from Yi to Yj.
18450 (sse_storeq_rex64 splitter): Use TARGET_INTER_UNIT_MOVES_FROM_VEC
18451 instead of TARGET_INTER_UNIT_MOVES.
18452 * config/i386/mmx.md (*mov<mode>_internal): Change constraint of
18453 operand 1 from Yi to Yj and from Ym to Yn.
18454
4c871069
JG		 184552013-04-29 James Greenhalgh <james.greenhalgh@arm.com>
18456
18457 * config/aarch64/aarch64-simd-builtins.def (vec_unpacks_hi_): New.
18458 (float_truncate_hi_): Likewise.
18459 (float_extend_lo_): Likewise.
18460 (float_truncate_lo_): Likewise.
18461 * config/aarch64/aarch64-simd.md (vec_unpacks_lo_v4sf): New.
18462 (aarch64_float_extend_lo_v2df): Likewise.
18463 (vec_unpacks_hi_v4sf): Likewise.
18464 (aarch64_float_truncate_lo_v2sf): Likewise.
18465 (aarch64_float_truncate_hi_v4sf): Likewise.
18466 (vec_pack_trunc_v2df): Likewise.
18467 (vec_pack_trunc_df): Likewise.
18468
1709ff9b
JG		 184692013-04-29 James Greenhalgh <james.greenhalgh@arm.com>
18470
18471 * config/aarch64/aarch64-builtins.c
18472 (aarch64_fold_builtin): Fold float conversions.
18473 * config/aarch64/aarch64-simd-builtins.def
18474 (floatv2si, floatv4si, floatv2di): New.
18475 (floatunsv2si, floatunsv4si, floatunsv2di): Likewise.
18476 * config/aarch64/aarch64-simd.md
18477 (<optab><fcvt_target><VDQF:mode>2): New, expands to float and floatuns.
18478 * config/aarch64/iterators.md (FLOATUORS): New.
18479 (optab): Add float, floatuns.
18480 (su_optab): Likewise.
18481
ce966824
JG		 184822013-04-29 James Greenhalgh <james.greenhalgh@arm.com>
18483
18484 * config/aarch64/aarch64-builtins.c
18485 (aarch64_builtin_vectorized_function): Use new names for
18486 fcvt builtins.
18487 * config/aarch64/aarch64-simd-builtins.def (fcvtzs): Split as...
18488 (lbtruncv2sf, lbtruncv4sf, lbtruncv2df): ...This.
18489 (fcvtzu): Split as...
18490 (lbtruncuv2sf, lbtruncuv4sf, lbtruncuv2df): ...This.
18491 (fcvtas): Split as...
18492 (lroundv2sf, lroundv4sf, lroundv2df, lroundsf, lrounddf): ...This.
18493 (fcvtau): Split as...
18494 (lrounduv2sf, lrounduv4sf, lrounduv2df, lroundusf, lroundudf): ...This.
18495 (fcvtps): Split as...
18496 (lceilv2sf, lceilv4sf, lceilv2df): ...This.
18497 (fcvtpu): Split as...
18498 (lceiluv2sf, lceiluv4sf, lceiluv2df, lceilusf, lceiludf): ...This.
18499 (fcvtms): Split as...
18500 (lfloorv2sf, lfloorv4sf, lfloorv2df): ...This.
18501 (fcvtmu): Split as...
18502 (lflooruv2sf, lflooruv4sf, lflooruv2df, lfloorusf, lfloorudf): ...This.
18503 (lfrintnv2sf, lfrintnv4sf, lfrintnv2df, lfrintnsf, lfrintndf): New.
18504 (lfrintnuv2sf, lfrintnuv4sf, lfrintnuv2df): Likewise.
18505 (lfrintnusf, lfrintnudf): Likewise.
18506 * config/aarch64/aarch64-simd.md
18507 (l<fcvt_pattern><su_optab><fcvt_target><VDQF:mode>2): Convert to
18508 define_insn.
18509 (aarch64_fcvt<frint_suffix><su><mode>): Remove.
18510 * config/aarch64/iterators.md (FCVT): Include UNSPEC_FRINTN.
18511 (fcvt_pattern): Likewise.
18512
b9de24fe
JG		 185132013-04-29 James Greenhalgh <james.greenhalgh@arm.com>
18514
18515 * config/aarch64/aarch64-simd.md
18516 (l<fcvt_pattern><su_optab><fcvt_target><VDQF:mode>2): Rename to...
18517 (l<fcvt_pattern><su_optab><VDQF:mode><fcvt_target>2): ... This.
18518
77a205be
JG		 185192013-04-29 James Greenhalgh <james.greenhalgh@arm.com>
18520
18521 * config/aarch64/arm_neon.h (vrndq<a,m,n,p>_f<32, 64>): Rename to...
18522 (vrnd<a,m,n,p>q_f<32, 64>): ...This, implement using builtin.
18523 (vrnd<a,m,n,p>_f32): Implement using builtins.
18524 (vrnd<i,x><q>_f<32, 64>): New.
18525
0659ce6f
JG		 185262013-04-29 James Greenhalgh <james.greenhalgh@arm.com>
18527
18528 * config/aarch64/aarch64-builtins.c
18529 (aarch64_builtin_vectorized_function): Fold to standard pattern names.
18530 * config/aarch64/aarch64-simd-builtins.def (frintn): New.
18531 (frintz): Rename to...
18532 (btrunc): ...this.
18533 (frintp): Rename to...
18534 (ceil): ...this.
18535 (frintm): Rename to...
18536 (floor): ...this.
18537 (frinti): Rename to...
18538 (nearbyint): ...this.
18539 (frintx): Rename to...
18540 (rint): ...this.
18541 (frinta): Rename to...
18542 (round): ...this.
18543 * config/aarch64/aarch64-simd.md
18544 (aarch64_frint<frint_suffix><mode>): Delete.
18545 (<frint_pattern><mode>2): Convert to insn.
18546 * config/aarch64/aarch64.md (unspec): Add UNSPEC_FRINTN.
18547 * config/aarch64/iterators.md (FRINT): Add UNSPEC_FRINTN.
18548 (frint_pattern): Likewise.
18549 (frint_suffix): Likewise.
18550
ea78906a
RB		 185512013-04-29 Richard Biener <rguenther@suse.de>
18552
18553 PR tree-optimization/57081
18554 * loop-init.c: Include tree-flow.h.
18555 (loop_optimizer_finalize): Free number of iteration estimates.
18556 * Makefile.in (loop-init.o): Add $(TREE_FLOW_H) dependency.
18557
baee1763
JJ		 185582013-04-29 Jakub Jelinek <jakub@redhat.com>
18559
94dc5332
JJ		 18560 PR tree-optimization/57083
18561 * tree-vrp.c (extract_range_from_binary_expr_1): For LSHIFT_EXPR with
18562 non-singleton shift count range, zero extend low_bound for uns case.
18563
baee1763
JJ		 18564 * config/i386/predicates.md (general_vector_operand): New predicate.
18565 * config/i386/i386.c (const_vector_equal_evenodd_p): New function.
18566 (ix86_expand_mul_widen_evenodd): Force op1 resp. op2 into register
18567 if they aren't nonimmediate operands. If their original values
18568 satisfy const_vector_equal_evenodd_p, don't shift them.
18569 * config/i386/sse.md (mul<mode>3): Use general_vector_operand
18570 predicates. For the SSE4.1 case force operands[{1,2}] into registers
18571 if not nonimmediate_operand.
18572 (vec_widen_smult_even_v4si): Use nonimmediate_operand predicates
18573 instead of register_operand.
18574 (vec_widen_<s>mult_odd_<mode>): Use general_vector_operand predicates.
18575
a9073727 185762013-04-28 Eric Botcazou <ebotcazou@adacore.com>
22ad5e0a
EB		 18577
18578 * stor-layout.c (finalize_size_functions): Allocate a structure and
18579 reset cfun before dumping the functions.
18580
ba8011e6
JJ		 185812013-04-27 Jakub Jelinek <jakub@redhat.com>
18582
d6fde69e
JJ		 18583 * config/i386/i386.c (ix86_expand_call): Make cregs_size unsigned.
18584
ba8011e6
JJ		 18585 PR target/56866
18586 * config/i386/i386.c (ix86_expand_mul_widen_evenodd): Don't
18587 use xop_pmacsdqh if uns_p.
18588 * config/i386/sse.md (xop_rotr<mode>3): Fix up computation of
18589 the immediate rotate count.
18590
2c62cbaa
VM		 185912013-04-26 Vladimir Makarov <vmakarov@redhat.com>
18592
18593 * rtl.h (struct rtx_def): Add comment for field jump.
18594 (LRA_SUBREG_P): New macro.
18595 * recog.c (register_operand): Check LRA_SUBREG_P.
18596 * lra.c (lra): Add note at the end of RTL code. Align non-empty
18597 stack frame.
18598 * lra-spills.c (lra_spill): Align stack after spilling pseudos.
18599 (lra_final_code_change): Skip subreg change for operators.
18600 * lra-eliminations.c (eliminate_regs_in_insn): Make return earlier
18601 if there are no operand changes.
18602 * lra-constraints.c (curr_insn_set): New.
18603 (match_reload): Set LRA_SUBREG_P.
18604 (emit_spill_move): Ditto.
18605 (check_and_process_move): Use curr_insn_set. Process only single
18606 set insns. Don't initialize sec_mem_p and change_p.
18607 (simplify_operand_subreg): Use LRA_SUBREG_P.
18608 (reg_in_class_p): New function.
18609 (process_alt_operands): Use it. Use #if HAVE_ATTR_enabled instead
18610 of #ifdef. Add code to remove cycling.
18611 (process_address): Check EXTRA_CONSTRAINT_STR. Process even if
18612 non-null disp. Reload inner instead of disp when base and index
18613 are null. Try to put lo_sum into register.
18614 (EBB_PROBABILITY_CUTOFF): Redefine probability in percents.
18615 (check_and_process_move): Move code for move cost check to
18616 simple_move_p. Remove equiv_substitution.
18617 (simple_move_p): New function.
18618 (curr_insn_transform): Initialize sec_mem_p and change_p. Set up
18619 curr_insn_set. Call check_and_process_move only for single set
18620 insns. Use the new function. Move call of check_and_process_move
18621 after operand equiv substitution and address process.
18622
e7d764f3
JJ		 186232013-04-26 Jakub Jelinek <jakub@redhat.com>
18624
18625 PR go/57045
18626 * tree-ssa-uninit.c (compute_uninit_opnds_pos): In functions
18627 with nonlocal goto receivers or returns twice calls, ignore
18628 unininitialized values from abnormal edges to nl goto receiver
18629 or returns twice call.
18630
41e10689
JJ		 186312013-04-26 Jakub Jelinek <jakub@redhat.com>
18632
18633 PR tree-optimization/57051
18634 * fold-const.c (const_binop): Handle VEC_LSHIFT_EXPR
18635 and VEC_RSHIFT_EXPR if shift count is a multiple of element
18636 bitsize.
18637
d7ed20db
RB		 186382013-04-26 Richard Biener <rguenther@suse.de>
18639
18640 * omp-low.c (finalize_task_copyfn): Do not drop PROP_loops.
18641 (expand_omp_taskreg): Likewise. Mark loops for fixup.
18642 * tree-cfg.c (move_block_to_fn): Remap loop fathers.
18643 (fixup_loop_arrays_after_move): New function.
18644 (move_sese_region_to_fn): Properly outline the loop tree parts
18645 of the SESE region.
18646
df93505e
UB		 186472013-04-26 Uros Bizjak <ubizjak@gmail.com>
18648
18649 * config/i386/i386.md (type, unit): Fix long lines.
18650
dd366ec3
RB		 186512013-04-26 Richard Biener <rguenther@suse.de>
18652
18653 * Makefile.in (lto-streamer-in.o): Add $(CFGLOOP_H) dependency.
18654 (lto-streamer-out.o): Likewise.
18655 * cfgloop.c (init_loops_structure): Export, add struct function
18656 argument and adjust.
18657 (flow_loops_find): Adjust.
18658 * cfgloop.h (enum loop_estimation): Add EST_LAST.
18659 (init_loops_structure): Declare.
18660 * lto-streamer-in.c: Include cfgloop.h.
18661 (input_cfg): Input the loop tree.
18662 * lto-streamer-out.c: Include cfgloop.h.
18663 (output_cfg): Output the loop tree.
18664 (output_struct_function_base): Do not drop PROP_loops.
18665
a9e0d843
RB		 186662013-03-26 Richard Biener <rguenther@suse.de>
18667
18668 * tree-cfg.c (execute_build_cfg): Build the loop tree.
18669 (pass_build_cfg): Provide PROP_loops.
18670 (move_sese_region_to_fn): Remove loops that are outlined into fn
18671 for now.
18672 * tree-inline.c: Include cfgloop.h.
18673 (initialize_cfun): Do not drop PROP_loops.
18674 (copy_loops): New function.
18675 (copy_cfg_body): Copy loop structure.
18676 (tree_function_versioning): Initialize destination loop tree.
18677 * tree-ssa-loop.c (pass_tree_loop_init): Do not provide PROP_loops.
18678 (pass_parallelize_loops): Do IL verification.
18679 * loop-init.c (loop_optimizer_init): Fixup loops if required.
18680 * tree-optimize.c (execute_fixup_cfg): If we need to cleanup
18681 the CFG make sure we fixup loops as well.
18682 * tree-ssa-tail-merge.c: Include cfgloop.h.
18683 (replace_block_by): When merging loop latches mark loops for fixup.
18684 * lto-streamer-out.c (output_struct_function_base): Drop
18685 PROP_loops for now.
18686 * tree-ssa-phiopt.c: Include tree-scalar-evolution.h.
18687 (tree_ssa_cs_elim): Initialize the loop optimizer and SCEV.
18688 * ipa-split.c: Include cfgloop.h.
18689 (split_function): Add the new return block to the loop tree root.
18690 * tree-cfgcleanup.c (remove_forwarder_block_with_phi): Return
18691 whether we have removed the forwarder block.
18692 (merge_phi_nodes): If we removed a forwarder mark loops for fixup.
18693 * cfgloop.h (place_new_loop): Declare.
18694 * cfgloopmanip.c (place_new_loop): Export.
18695 * Makefile.in (asan.o): Add $(CFGLOOP_H) dependency.
18696 (tree-switch-conversion.o): Likewise.
18697 (tree-complex.o): Likewise.
18698 (tree-inline.o): Likewise.
18699 (tree-ssa-tailmerge.o): Likewise.
18700 (ipa-split.o): Likewise.
18701 (tree-ssa-phiopt.o): Add $(SCEV_H) dependency.
18702 (tree-ssa-copy.o): Likewise.
18703 * tree-switch-conversion.c: Include cfgloop.h
18704 (process_switch): If we emit a bit-test cascade, schedule loops
18705 for fixup.
18706 * tree-complex.c: Include cfgloop.h.
18707 (expand_complex_div_wide): Properly add new basic-blocks to loops.
18708 * asan.c: Include cfgloop.h.
18709 (create_cond_insert_point): Properly add new basic-blocks to
18710 loops, schedule loop fixup.
18711 * cfgloop.c (verify_loop_structure): Check that looks are not
18712 marked for fixup.
18713 * omp-low.c (expand_parallel_call): Properly add new basic-blocks
18714 to loops.
18715 (expand_omp_for_generic): Likewise.
18716 (expand_omp_sections): Likewise.
18717 (expand_omp_atomic_pipeline): Schedule loops for fixup.
18718 * tree-ssa-copy.c: Include tree-scalar-evolution.h.
18719 (fini_copy_prop): Disable DCE in substitute_and_fold if SCEV
18720 is initialized, not when loops are present.
18721 * tree-parloops.c (parallelize_loops): Remove checking here.
18722 * passes.c (init_optimization_passes): Schedule a copy-propagation
18723 pass before complete unrolling of inner loops.
18724
e78e8a0b
JJ		 187252013-04-26 Jakub Jelinek <jakub@redhat.com>
18726
a2e836b2
JJ		 18727 * Makefile.in (toplev.o): Depend on diagnostic-color.h.
18728 * diagnostic-color.c (should_colorize): Remove _WIN32 version.
18729 (colorize_init): Add argument to _WIN32 version.
18730 * toplev.c: Include diagnostic-color.h.
18731 (process_options): Default to -fdiagnostics-color=auto if
18732 GCC_COLORS env var is in the environment.
18733 * common.opt (fdiagnostics-color=): Add Var and Init.
df93505e
UB		 18734 * doc/invoke.texi (-fdiagnostics-color=): Document that if GCC_COLORS
18735 env var is in the environment, the default is auto rather than never.
a2e836b2 18736
e78e8a0b
JJ		 18737 * diagnostic.h (file_name_as_prefix): Add context argument.
18738 * diagnostic.c (file_name_as_prefix): Likewise. Colorize
18739 the string as locus.
18740 * langhooks.c (lhd_print_error_function): Adjust caller.
18741
013e5ef9
LC		 187422013-04-25 Lawrence Crowl <crowl@google.com>
18743
18744 * var-tracking.c (shared_hash_def::htab):
18745 Change type to hash_table. Update dependent calls and types.
18746
4a8fb1a1
LC		 187472013-04-25 Lawrence Crowl <crowl@google.com>
18748
18749 * Makefile.in: Update as needed below.
18750
18751 * alloc-pool.c (static hash_table <alloc_pool_hasher> alloc_pool_hash):
18752 Move declaration to after the type's method definitons.
18753
18754 * attribs.c (htab_t scoped_attributes::attribute_hash):
18755 Change type to hash_table. Update dependent calls and types.
18756
18757 * bitmap.c (htab_t bitmap_desc_hash):
18758 Change type to hash_table. Update dependent calls and types.
18759
18760 * cselib.c (htab_t cselib_hash_table):
18761 Change type to hash_table. Update dependent calls and types.
18762
18763 * data-streamer.h (struct string_slot): Move to lto-streamer.h.
18764 (hash_string_slot_node): Move implementation into lto-streamer.h
18765 struct string_slot_hasher.
18766 (eq_string_slot_node): Likewise.
18767
18768 * data-streamer-out.c: Update output_block::string_hash_table
18769 dependent calls and types.
18770
18771 * dwarf2cfi.c (htab_t trace_index):
18772 Change type to hash_table. Update dependent calls and types.
18773
18774 * dwarf2out.c (htab_t break_out_includes::cu_hash_table):
18775 Change type to hash_table. Update dependent calls and types.
18776 (htab_t copy_decls_for_unworthy_types::decl_table): Likewise.
18777 (htab_t optimize_external_refs::map): Likewise.
18778 (htab_t output_comp_unit::extern_map): Likewise.
18779 (htab_t output_comdat_type_unit::extern_map): Likewise.
18780 (htab_t output_macinfo::macinfo_htab): Likewise.
18781 (htab_t optimize_location_lists::htab): Likewise.
18782 (htab_t dwarf2out_finish::comdat_type_table): Likewise.
18783
18784 * except.c (htab_t ehspec_hash_type):
18785 Change type to hash_table. Update dependent calls and types.
18786 (assign_filter_values::ttypes): Likewise.
18787 (assign_filter_values::ehspec): Likewise.
18788 (sjlj_assign_call_site_values::ar_hash): Likewise.
18789 (convert_to_eh_region_ranges::ar_hash): Likewise.
18790
18791 * gcse.c (htab_t pre_ldst_table):
18792 Change type to hash_table. Update dependent calls and types.
18793
18794 * ggc-common.c (htab_t saving_htab):
18795 Change type to hash_table. Update dependent calls and types.
18796 (htab_t loc_hash): Likewise.
18797 (htab_t ptr_hash): Likewise.
18798 (call_count): Rename ggc_call_count.
18799 (call_alloc): Rename ggc_call_alloc.
18800 (loc_descriptor): Rename make_loc_descriptor.
18801 (add_statistics): Rename ggc_add_statistics.
18802
18803 * ggc-common.c (saving_htab):
18804 Change type to hash_table. Update dependent calls and types.
18805
18806 * gimple.h (struct gimplify_ctx): Move to gimplify-ctx.h.
18807 (push_gimplify_context): Likewise.
18808 (pop_gimplify_context): Likewise.
18809 (struct gimple_temp_hash_elt): Added.
18810 (struct gimplify_hasher): Likewise.
18811 (struct gimplify_ctx.temp_htab):
18812 Change type to hash_table. Update dependent calls and types.
18813
18814 * gimple-fold.c: Include gimplify-ctx.h.
18815
18816 * gimple-ssa-strength-reduction.c (htab_t base_cand_map):
18817 Change type to hash_table. Update dependent calls and types.
18818 (base_cand_dump_callback): Rename to ssa_base_cand_dump_callback to
18819 avoid potential global name collision.
18820
18821 * gimplify.c: Include gimplify-ctx.h.
18822 (struct gimple_temp_hash_elt): Move to gimplify-ctx.h.
18823 (htab_t gimplify_ctx::temp_htab):
18824 Update dependent calls and types for new type hash_table.
18825 (gimple_tree_hash): Move into gimplify_hasher in gimplify-ctx.h.
18826 (gimple_tree_eq): Move into gimplify_hasher in gimplify-ctx.h.
18827
18828 * gimplify-ctx.h: New.
18829 (struct gimple_temp_hash_elt): Move from gimplify.c.
18830 (class gimplify_hasher): New.
18831 (struct gimplify_ctx): Move from gimple.h.
18832 (htab_t gimplify_ctx::temp_htab):
18833 Change type to hash_table. Update dependent calls and types.
18834
18835 * graphite-clast-to-gimple.c: Include graphite-htab.h.
18836 (htab_t ivs_params::newivs_index):
18837 Change type to hash_table. Update dependent calls and types.
18838 (htab_t ivs_params::params_index): Likewise.
18839 (htab_t print_generated_program::params_index): Likewise.
18840 (htab_t gloog::newivs_index): Likewise.
18841 (htab_t gloog::params_index): Likewise.
18842
18843 * graphite.c: Include graphite-htab.h.
18844 4htab_t graphite_transform_loops::bb_pbb_mapping):
18845 Change type to hash_table. Update dependent calls and types.
18846
18847 * graphite-clast-to-gimple.h: (extern gloog) Move to graphite-htab.h.
18848 (bb_pbb_map_hash): Fold into bb_pbb_htab_type in graphite-htab.h.
18849 (eq_bb_pbb_map): Fold into bb_pbb_htab_type in graphite-htab.h.
18850
18851 * graphite-dependences.c: Include graphite-htab.h.
18852 (loop_is_parallel_p): Change hash table type of parameter.
18853
18854 * graphite-htab.h: New.
18855 (typedef hash_table <bb_pbb_hasher> bb_pbb_htab_type): New.
18856 (extern find_pbb_via_hash): Move from graphite-poly.h.
18857 (extern loop_is_parallel_p): Move from graphite-poly.h.
18858 (extern get_loop_body_pbbs): Move from graphite-poly.h.
18859
18860 * graphite-poly.h (extern find_pbb_via_hash): Move to graphite-htab.h.
18861 (extern loop_is_parallel_p): Move to graphite-htab.h.
18862 (extern get_loop_body_pbbs): Move to graphite-htab.h.
18863
18864 * haifa-sched.c (htab_t delay_htab):
18865 Change type to hash_table. Update dependent calls and types.
18866 (htab_t delay_htab_i2): Likewise.
18867
18868 * ira-color.c (htab_t allocno_hard_regs_htab):
18869 Change type to hash_table. Update dependent calls and types.
18870
18871 * ira-costs.c (htab_t cost_classes_htab):
18872 Change type to hash_table. Update dependent calls and types.
18873
18874 * loop-invariant.c (htab_t merge_identical_invariants::eq):
18875 Change type to hash_table. Update dependent calls and types.
18876
18877 * loop-iv.c (htab_t bivs):
18878 Change type to hash_table. Update dependent calls and types.
18879
18880 * loop-unroll.c (htab_t opt_info::insns_to_split):
18881 Change type to hash_table. Update dependent calls and types.
18882 (htab_t opt_info::insns_with_var_to_expand): Likewise.
18883
18884 * lto-streamer.h (struct string_slot): Move from data-streamer.h
18885 (struct string_slot_hasher): New.
18886 (htab_t output_block::string_hash_table):
18887 Change type to hash_table. Update dependent calls and types.
18888
18889 * lto-streamer-in.c (freeing_string_slot_hasher): New.
18890 (htab_t file_name_hash_table):
18891 Change type to hash_table. Update dependent calls and types.
18892
18893 * lto-streamer-out.c: Update output_block::string_hash_table dependent
18894 calls and types.
18895
18896 * lto-streamer.c (htab_t tree_htab):
18897 Change type to hash_table. Update dependent calls and types.
18898
18899 * omp-low.c: Include gimplify-ctx.h.
18900
18901 * passes.c (htab_t name_to_pass_map):
18902 Change type to hash_table. Update dependent calls and types.
18903 (pass_traverse): Rename to passes_pass_traverse.
18904
18905 * plugin.c (htab_t event_tab):
18906 Change type to hash_table. Update dependent calls and types.
18907
18908 * postreload-gcse.c (htab_t expr_table):
18909 Change type to hash_table. Update dependent calls and types.
18910 (dump_hash_table_entry): Rename dump_expr_hash_table_entry.
18911
18912 * sese.c (debug_rename_map_1): Make extern.
18913 (htab_t copy_bb_and_scalar_dependences::rename_map):
18914 Change type to hash_table. Update dependent calls and types.
18915
18916 * sese.h (extern debug_rename_map): Move to .c file.
18917
18918 * store-motion.c (htab_t store_motion_mems_table):
18919 Change type to hash_table. Update dependent calls and types.
18920
18921 * trans-mem.c (htab_t tm_new_mem_hash):
18922 Change type to hash_table. Update dependent calls and types.
18923
18924 * tree-browser.c (htab_t TB_up_ht):
18925 Change type to hash_table. Update dependent calls and types.
18926
18927 * tree-cfg.c (htab_t discriminator_per_locus):
18928 Change type to hash_table. Update dependent calls and types.
18929
18930 * tree-complex.c: Include tree-hasher.h
18931 (htab_t complex_variable_components):
18932 Change type to hash_table. Update dependent calls and types.
18933
18934 * tree-eh.c (htab_t finally_tree):
18935 Change type to hash_table. Update dependent calls and types.
18936
18937 * tree-flow.h (extern int_tree_map_hash): Moved into tree-hasher
18938 struct int_tree_hasher.
18939 (extern int_tree_map_eq): Likewise.
18940 (uid_decl_map_hash): Removed.
18941 (extern decl_tree_map_eq): Likewise.
18942
18943 * tree-hasher.h: New.
18944 (struct int_tree_hasher): New.
18945 (typedef int_tree_htab_type): New.
18946
18947 * tree-inline.c: Include gimplify-ctx.h.
18948
18949 * tree-mudflap.c: Include gimplify-ctx.h.
18950
18951 * tree-parloops.c: Include tree-hasher.h.
18952 (htab_t eliminate_local_variables_stmt::decl_address):
18953 Change type to hash_table. Update dependent calls and types.
18954 (htab_t separate_decls_in_region::decl_copies): Likewise.
18955
18956 * tree-scalar-evolution.c (htab_t resolve_mixers::cache):
18957 Change type to hash_table. Update dependent calls and types.
18958
18959 * tree-sra.c (candidates):
18960 Change type to hash_table. Update dependent calls and types.
18961
18962 * tree-ssa.c (int_tree_map_eq): Moved into struct int_tree_hasher
18963 in tree-flow.h.
18964 (int_tree_map_hash): Likewise.
18965
18966 * tree-ssa-dom.c (htab_t avail_exprs):
18967 Change type to hash_table. Update dependent calls and types.
18968
18969 * tree-ssa-live.c (var_map_base_init::tree_to_index):
18970 Change type to hash_table. Update dependent calls and types.
18971
18972 * tree-ssa-loop-ivopts.c (struct ivopts_data.inv_expr_tab):
18973 Change type to hash_table. Update dependent calls and types.
18974
18975 * tree-ssa-phiopt.c (seen_ssa_names):
18976 Change type to hash_table. Update dependent calls and types.
18977
18978 * tree-ssa-strlen.c (decl_to_stridxlist_htab):
18979 Change type to hash_table. Update dependent calls and types.
18980
18981 * tree-ssa-uncprop.c (equiv):
18982 Change type to hash_table. Update dependent calls and types.
18983
c5a44004
JJ		 189842013-04-25 Jakub Jelinek <jakub@redhat.com>
18985
18986 PR rtl-optimization/57003
18987 * regcprop.c (copyprop_hardreg_forward_1): If ksvd.ignore_set_reg,
18988 call note_stores with kill_clobbered_value callback again after
18989 killing regs_invalidated_by_call.
18990
09962a4a
JG		 189912013-04-25 James Greenhalgh <james.greenhalgh@arm.com>
18992
18993 * config/aarch64/aarch64-simd.md
18994 (aarch64_simd_bsl<mode>_internal): Rewrite RTL to not use UNSPEC_BSL.
18995 (aarch64_simd_bsl<mode>): Likewise.
18996 * config/aarch64/iterators.md (unspec): Remove UNSPEC_BSL.
18997
ea28bb0b
MP		 189982013-04-25 Marek Polacek <polacek@redhat.com>
18999
19000 PR tree-optimization/57066
3c21604f 19001 * builtins.c (fold_builtin_logb): Return +Inf for -Inf.
ea28bb0b 19002
96659611
JG		 190032013-04-25 James Greenhalgh <jame.greenhalgh@arm.com>
19004
19005 * config/aarch64/aarch64-simd.md (neg<mode>2): Use VDQ iterator.
19006
9697e620
JG		 190072013-04-25 James Greenhalgh <james.greenhalgh@arm.com>
19008
19009 * config/aarch64/aarch64-builtins.c
19010 (aarch64_fold_builtin): New.
19011 * config/aarch64/aarch64-protos.h (aarch64_fold_builtin): New.
19012 * config/aarch64/aarch64.c (TARGET_FOLD_BUILTIN): Define.
19013 * config/aarch64/aarch64-simd-builtins.def (abs): New.
19014 * config/aarch64/arm_neon.h
19015 (vabs<q>_<f32, 64>): Implement using __builtin_aarch64_fabs.
19016
0ac198d3
JG		 190172013-04-25 James Greenhalgh <james.greenhalgh@arm.com>
19018 Tejas Belagod <tejas.belagod@arm.com>
19019
19020 * config/aarch64/aarch64-builtins.c
19021 (aarch64_gimple_fold_builtin): New.
19022 * config/aarch64/aarch64-protos.h (aarch64_gimple_fold_builtin): New.
19023 * config/aarch64/aarch64-simd-builtins.def (addv): New.
19024 * config/aarch64/aarch64-simd.md (addpv4sf): New.
19025 (addvv4sf): Update.
19026 * config/aarch64/aarch64.c (TARGET_GIMPLE_FOLD_BUILTIN): Define.
19027
58cff58c
N		 190282013-04-25 Naveen H.S <Naveen.Hurugalawadi@caviumnetworks.com>
19029
df93505e 19030 * config/aarch64/aarch64.md
58cff58c
N		 19031 (*cmp_swp_<optab><ALLX:mode>_shft_<GPI:mode>): New pattern.
19032
7e0228bf
N		 190332013-04-25 Naveen H.S <Naveen.Hurugalawadi@caviumnetworks.com>
19034
19035 * config/aarch64/aarch64.md (*ngc<mode>): New pattern.
19036 (*ngcsi_uxtw): New pattern.
19037
5819f96f 190382013-04-25 Kyrylo Tkachov <kyrylo.tkachov@arm.com>
3c21604f 19039 Julian Brown <julian@codesourcery.com>
5819f96f
KT		 19040
19041 * config/arm/arm.c (neon_builtin_type_mode): Add T_V4HF.
19042 (TB_DREG): Add T_V4HF.
19043 (v4hf_UP): New macro.
19044 (neon_itype): Add NEON_FLOAT_WIDEN, NEON_FLOAT_NARROW.
df93505e 19045 (arm_init_neon_builtins): Handle NEON_FLOAT_WIDEN, NEON_FLOAT_NARROW.
5819f96f
KT		 19046 Handle initialisation of V4HF. Adjust initialisation of reinterpret
19047 built-ins.
df93505e 19048 (arm_expand_neon_builtin): Handle NEON_FLOAT_WIDEN, NEON_FLOAT_NARROW.
5819f96f
KT		 19049 (arm_vector_mode_supported_p): Handle V4HF.
19050 (arm_mangle_map): Handle V4HFmode.
19051 * config/arm/arm.h (VALID_NEON_DREG_MODE): Add V4HF.
19052 * config/arm/arm_neon_builtins.def: Add entries for
19053 vcvtv4hfv4sf, vcvtv4sfv4hf.
19054 * config/arm/neon.md (neon_vcvtv4sfv4hf): New pattern.
19055 (neon_vcvtv4hfv4sf): Likewise.
19056 * config/arm/neon-gen.ml: Handle half-precision floating point
19057 features.
19058 * config/arm/neon-testgen.ml: Handle Requires_FP_bit feature.
19059 * config/arm/arm_neon.h: Regenerate.
19060 * config/arm/neon.ml (type elts): Add F16.
19061 (type vectype): Add T_float16x4, T_floatHF.
19062 (type vecmode): Add V4HF.
19063 (type features): Add Requires_FP_bit feature.
19064 (elt_width): Handle F16.
19065 (elt_class): Likewise.
19066 (elt_of_class_width): Likewise.
19067 (mode_of_elt): Refactor.
19068 (type_for_elt): Handle F16, fix error messages.
19069 (vectype_size): Handle T_float16x4.
19070 (vcvt_sh): New function.
19071 (ops): Add entries for vcvt_f16_f32, vcvt_f32_f16.
19072 (string_of_vectype): Handle T_floatHF, T_float16, T_float16x4.
19073 (string_of_mode): Handle V4HF.
19074 * doc/arm-neon-intrinsics.texi: Regenerate.
19075
1ef395e4
JG		 190762013-04-25 James Greenhalgh <james.greenhalgh@arm.com>
19077
19078 * config/aarch64/aarch64.c (aarch64_print_operand): Fix asm_fprintf
19079 format specifier in 'X' case.
19080
41c34e94
AM		 190812013-04-25 Alan Modra <amodra@gmail.com>
19082
19083 PR target/57052
19084 * config/rs6000/rs6000.md (rotlsi3_internal7): Rename to
19085 rotlsi3_internal7le and condition on !BYTES_BIG_ENDIAN.
19086 (rotlsi3_internal8be): New BYTES_BIG_ENDIAN insn.
19087 Repeat for many other rotate/shift and mask patterns using subregs.
19088 Name lshiftrt insns.
19089 (ashrdisi3_noppc64): Rename to ashrdisi3_noppc64be and condition
19090 on WORDS_BIG_ENDIAN.
19091
b9a7eb5d
AM		 190922013-04-25 Alan Modra <amodra@gmail.com>
19093
19094 * config.gcc: Support little-endian powerpc-linux targets.
19095 * config/rs6000/linux.h (LINK_OS_LINUX_EMUL): Define.
19096 (LINK_OS_LINUX_SPEC): Define.
19097 * config/rs6000/linuxspe.h (TARGET_DEFAULT):
19098 Preserve MASK_LITTLE_ENDIAN.
19099 * config/rs6000/default64.h (TARGET_DEFAULT): Likewise.
19100 * config/rs6000/linuxaltivec.h (TARGET_DEFAULT): Likewise.
19101 * config/rs6000/linux64.h (OPTION_LITTLE_ENDIAN): Don't zero.
19102 (LINK_OS_LINUX_EMUL32, LINK_OS_LINUX_EMUL64): Define.
19103 (LINK_OS_LINUX_SPEC32, LINK_OS_LINUX_SPEC64): Use above.
19104 * config/rs6000/rs6000.c (output_toc): Don't use .tc for TARGET_ELF.
19105 Correct fp word order for little-endian. Don't shift toc entries
19106 smaller than a word for little-endian.
19107 * config/rs6000/rs6000.md (bswaphi2, bswapsi2 split): Comment.
19108 (bswapdi2 splits): Correct low-part subreg for little-endian.
19109 Remove wrong BYTES_BIG_ENDIAN tests, and rename vars to remove
19110 low/high where such is correct only for be.
19111 * config/rs6000/sysv4.h (SUBTARGET_OVERRIDE_OPTIONS): Allow
19112 little-endian for -mcall-aixdesc.
19113
87f73374
AM		 191142013-04-25 Alan Modra <amodra@gmail.com>
19115
19116 * config/rs6000/rs6000.c (rs6000_secondary_reload_inner): Use
19117 replace_equiv_address_nv.
19118
cabf91cd
AM		 191192013-04-25 Alan Modra <amodra@gmail.com>
19120
19121 * config/rs6000/rs6000.c (rs6000_emit_set_long_const): Tidy.
19122
0ae24cc8
VM		 191232013-04-24 Vladimir Makarov <vmakarov@redhat.com>
19124
19125 Revert:
19126 2013-04-24 Vladimir Makarov <vmakarov@redhat.com>
19127 * rtl.h (struct rtx_def): ...
cabf91cd 19128
77bce07c
VM		 191292013-04-24 Vladimir Makarov <vmakarov@redhat.com>
19130
19131 PR rtl-optimizations/57046
19132 * lra-constraints (split_reg): Set up lra_risky_transformations_p
19133 for multi-reg splits.
19134
0db63e7f
L		 191352013-04-24 H.J. Lu <hongjiu.lu@intel.com>
19136
19137 * config/i386/x86-64.h (ASM_SPEC): Support -mx32.
19138
3362b6b6
SA		 191392013-04-24 Sterling Augustine <saugustine@google.com>
19140
19141 * dwarf2out.c (skeleton_debug_str_hash, add_skeleton_AT_string)
19142 (comp_dir_string, debug_str_dwo_section): New.
19143 (DEBUG_STR_DWO_SECTION): Rename to ...
19144 (DEBUG_DWO_STR_SECTION): ... this.
19145 (DEBUG_NORM_STR_SECTION): Delete.
19146 (DEBUG_STR_SECTION, DEBUG_STR_SECTION_FLAGS): Edit definitions.
19147 (DEBUG_STR_DWO_SECTION_FLAGS): New.
19148 (find_AT_string): Move most logic to ...
19149 (find_AT_string_in_table): ... here. New.
19150 (add_top_level_skeleton_die_attrs): Call comp_dir_string and
19151 add_skeleton_AT_string. Delete logic.
19152 (output_skeleton_debug_sections): Remove call to
19153 add_top_level_skeleton_die_attrs.
19154 (add_comp_dir_attribute): Move logic to comp_dir_string.
19155 (dwarf2out_init): Initialize debug_str_dwo_section.
19156 (output_indirect_string): Call find_string_form.
19157 (output_indirect_strings): Rewrite.
19158 (prune_unused_types): Empty skeleton_debug_str_hash.
19159 Call get_skeleton_type_unit and add_top_level_skeleton_die_attrs.
19160 (dwarf2out_finish): Call output_indirect_strings.
19161
e93e18e9
PC		 191622013-04-24 Paolo Carlini <paolo.carlini@oracle.com>
19163
19164 * doc/cpp.texi: Remove __GXX_EXPERIMENTAL_CXX1Y__.
19165
f6ce35ac
VM		 191662013-04-24 Vladimir Makarov <vmakarov@redhat.com>
19167
cabf91cd 19168 * rtl.h (struct rtx_def): Add comment for field jump.
f6ce35ac
VM		 19169 (LRA_SUBREG_P): New macro.
19170 * recog.c (register_operand): Check LRA_SUBREG_P.
19171 * lra.c (lra): Add note at the end of RTL code. Align non-empty
19172 stack frame.
19173 * lra-spills.c (lra_spill): Align stack after spilling pseudos.
19174 (lra_final_code_change): Skip subreg change for operators.
19175 * lra-eliminations.c (eliminate_regs_in_insn): Make return earlier
19176 if there are no operand changes.
19177 * lra-constraints.c (curr_insn_set): New.
19178 (match_reload): Set LRA_SUBREG_P.
19179 (emit_spill_move): Ditto.
19180 (check_and_process_move): Use curr_insn_set. Process only single
19181 set insns. Don't initialize sec_mem_p and change_p.
19182 (simplify_operand_subreg): Use LRA_SUBREG_P.
19183 (reg_in_class_p): New function.
19184 (process_alt_operands): Use it. Use #if HAVE_ATTR_enabled instead
19185 of #ifdef. Add code to remove cycling.
19186 (process_address): Check EXTRA_CONSTRAINT_STR. Process even if
19187 non-null disp. Reload inner instead of disp when base and index
19188 are null. Try to put lo_sum into register.
19189 (EBB_PROBABILITY_CUTOFF): Redefine probability in percents.
cabf91cd 19190 (check_and_process_move): Move code for move cost check to
f6ce35ac 19191 simple_move_p. Remove equiv_substitution.
cabf91cd 19192 (simple_move_p): New function.
f6ce35ac
VM		 19193 (curr_insn_transform): Initialize sec_mem_p and change_p. Set up
19194 curr_insn_set. Call check_and_process_move only for single set
19195 insns. Use the new function. Move call of check_and_process_move
19196 after operand equiv substitution and address process.
19197
38047d90
JG		 191982013-04-24 James Greenhalgh <james.greenhalgh@arm.com>
19199
19200 * config/aarch64/arm_neon.h (vld1<q>_lane*): Fix constraints.
19201 (vld1<q>_dup_<sufp><8, 16, 32, 64>): Likewise.
19202 (vld1<q>_<sufp><8, 16, 32, 64>): Likewise.
19203
13f39b2e
PC		 192042013-04-24 Paolo Carlini <paolo.carlini@oracle.com>
19205
19206 * doc/cpp.texi: Document __GXX_EXPERIMENTAL_CXX1Y__.
19207
97c116dc
MP		 192082013-04-24 Marek Polacek <polacek@redhat.com>
19209
19210 * tree-scalar-evolution.h (analyze_scalar_evolution): Remove.
19211 * tree-scalar-evolution.c (get_exit_conditions_rec): Likewise.
19212 (select_loops_exit_conditions): Likewise.
19213 (number_of_iterations_for_all_loops): Likewise.
19214 (analyze_scalar_evolution_for_all_loop_phi_nodes): Likewise.
19215 (scev_analysis): Likewise.
19216
83082391 192172013-04-02 Catherine Moore <clm@codesourcery.com>
12211b99 19218 Chao-ying Fu <fu@mips.com>
83082391 19219
cabf91cd
AM		 19220 * config/mips/micromips.md (jraddiusp): New pattern.
19221 * config/mips/mips.c (mips_expand_epilogue): Use the JRADDIUSP
19222 instruction if possible.
83082391 19223
19e34aa2
AM		 192242013-04-24 Alan Modra <amodra@gmail.com>
19225
19226 * config/rs6000/driver-rs6000.c (elf_dcachebsize): Fix comment pasto.
19227
fdb6603c
JB		 192282013-04-24 Julian Brown <julian@codesourcery.com>
19229 Chung-Lin Tang <cltang@codesourcery.com>
19230
19231 * dwarf2out.c (gen_enumeration_type_die): Fix HOST_BITS_PER_WIDE_INT
df93505e
UB		 19232 dependency behavior in enumeration type DIE generation. Add TODO note
19233 to comments about future DW_FORM_sdata/udata re-work of related code.
fdb6603c 19234
bf190e8d
LC		 192352013-04-23 Lawrence Crowl <crowl@google.com>
19236
19237 * Makefile.in: Update as needed below.
19238
19239 * hash-table.h (class hash_table):
19240 Correct many methods with parameter types compare_type to the correct
19241 value_type. (Correct code was unlikely to notice the change.)
19242 (hash_table::elements_with_deleted) New.
19243 (class hashtable::iterator): New.
19244 (hashtable::begin()): New.
19245 (hashtable::end()): New.
19246 (FOR_EACH_HASH_TABLE_ELEMENT): New.
19247
19248 * statistics.c (statistics_hashes):
19249 Change type to hash_table. Update dependent calls and types.
19250
19251 * tree-into-ssa.c (var_infos):
19252 Change type to hash_table. Update dependent calls and types.
19253
19254 * tree-ssa-coalesce.c (struct coalesce_list_d.list):
19255 Change type to hash_table. Update dependent calls and types.
19256
19257 * tree-ssa-loop-im.c (struct mem_ref.refs):
19258 Change type to hash_table. Update dependent calls and types.
19259
19260 * tree-ssa-reassoc.c (undistribute_ops_list::ctable):
19261 Change type to hash_table. Update dependent calls and types.
19262
19263 * tree-ssa-sccvn.c (vn_tables_s::nary):
19264 Change type to hash_table. Update dependent calls and types.
19265 (vn_tables_s::phis): Likewise.
19266 (vn_tables_s::references): Likewise.
19267
19268 * tree-ssa-sccvn.h (vn_nary_op_eq): Update parameter and return types.
19269 (vn_reference_eq): Update parameter and return types.
19270
19271 * tree-ssa-structalias.c (pointer_equiv_class_table):
19272 Change type to hash_table. Update dependent calls and types.
19273 (location_equiv_class_table): Likewise.
19274
19275 * tree-vect-data-refs.c: Consequential changes for making
19276 peeling a hash_table.
19277
19278 * tree-vect-loop.c (new_loop_vec_info): Dependent hash_table update.
19279 (destroy_loop_vec_info): Dependent hash_table update.
19280
19281 * tree-vectorizer.h (peeling_htab):
19282 Change type to hash_table. Update dependent calls and types.
19283
d70a81dd
SC		 192842013-04-23 Shiva Chen <shiva0217@gmail.com>
19285
cabf91cd
AM		 19286 * lra-assigns.c (find_hard_regno_for): Use lra_reg_val_equal_p
19287 to check the register content is equal or not.
19288 * lra-constraints.c (match_reload): Use lra_assign_reg_val
19289 to assign register content record.
19290 * lra-eliminations.c (update_reg_eliminate): Use
d70a81dd 19291 lra_update_reg_val_offset to update register content offset.
cabf91cd
AM		 19292 * lra-int.h (struct lra_reg): Add offset member.
19293 (lra_reg_val_equal_p): New static inline function.
19294 (lra_update_reg_val_offset): New static inline function.
19295 (lra_assign_reg_val): New static inline function.
19296 * lra.c (lra_create_new_reg): Use lra_assign_reg_val
19297 to assign register content record.
19298 (initialize_lra_reg_info_element): Initial offset to zero.
d70a81dd 19299
b894a1f3
CM		 193002013-04-23 Catherine Moore <clm@codesourcery.com>
19301
19302 * config/mips/mips.md (*movhi_internal, *movqi_internal): New
19303 operands. Record compression.
19304
ecd14de9
XDL		 193052013-04-23 Xinliang David Li <davidxl@google.com>
19306
19307 * cfghhooks.c (dump_bb_for_graph): Support 'slim' graph dump.
19308
92e776e9
RB		 193092013-04-23 Richard Biener <rguenther@suse.de>
19310
19311 PR middle-end/57036
19312 * tree-inline.c (copy_edges_for_bb): Add can_make_abnormal_goto
19313 parameter, only add abnormal goto edges from the copied body
19314 if the call could perform abnormal gotos.
19315 (copy_cfg_body): Adjust.
19316
a15ee567
SN		 193172013-04-23 Sofiane Naci <sofiane.naci@arm.com>
19318
19319 * config/aarch64/aarch64.md (*mov<mode>_aarch64): Add simd attribute.
19320
08c52234
AS		 193212013-04-23 Andreas Schwab <schwab@linux-m68k.org>
19322
19323 * coretypes.h (gimple_stmt_iterator): Add struct to make
19324 compatible with C.
19325
999c1171
RB		 193262013-04-23 Richard Biener <rguenther@suse.de>
19327
19328 PR tree-optimization/57026
19329 * tree-vrp.c (simplify_conversion_using_ranges): Do not propagate
19330 from SSA names occuring in abnormal PHI nodes.
19331
53669259
AK		 193322013-04-22 Andi Kleen <ak@linux.intel.com>
19333
19334 * lto/lto.c (print_lto_report_1): Fix LTO report names.
19335
1a0ad150
AK		 193362013-04-22 Andi Kleen <ak@linux.intel.com>
19337
19338 * lto/lto.c (print_lto_report_1): Declare early.
19339 (read_cgraph_and_symbols): Call print_lto_report_1 early.
19340
057f8f20
AK		 193412013-04-22 Andi Kleen <ak@linux.intel.com>
19342
19343 * common.opt (-flto-report-wpa): Add.
19344 * doc/invoke.texi (-flto-report-wpa): Add.
19345 * lto/lto.c (do_whole_program_analysis): Check for lto-report-wpa.
19346 (lto_main): dito.
19347
473b1e05
XDL		 193482013-04-22 Xinliang David Li <davidxl@google.com>
19349
19350 * graph.c (draw_cfg_node_succ_edges): Add branch probility as label.
19351 * cfghhooks.c (dump_bb_for_graph): Dump profile count and frquency.
19352 * Makefile.in: New dependency
19353
cabf91cd 19354 David Daney <ddaney.cavm@gmail.com>
b1485a33 19355
cabf91cd
AM		 19356 * configure.ac (gcc_cv_as_micromips_support): Use the
19357 --fatal-warnings option.
19358 * configure: Regenerate.
b1485a33 19359
829d0168
MP		 193602013-04-22 Marek Polacek <polacek@redhat.com>
19361
19362 PR sanitizer/56990
19363 * tsan.c (instrument_expr): Don't instrument expression
19364 in case its size is zero.
19365
6d9b7208
UB		 193662013-04-22 Uros Bizjak <ubizjak@gmail.com>
19367
19368 PR target/57032
19369 Revert:
19370 2013-03-17 Uros Bizjak <ubizjak@gmail.com>
19371
19372 * config/alpha/alpha.c (TARGET_LRA_P): New define.
19373
ea679d55
JG		 193742013-04-22 James Greenhalgh <james.greenhalgh@arm.com>
19375
19376 * coretypes.h (gimple_stmt_iterator_d): Forward declare.
19377 (gimple_stmt_iterator): New typedef.
19378 * gimple.h (gimple_stmt_iterator): Rename to...
19379 (gimple_stmt_iterator_d): ... This.
19380 * doc/tm.texi.in (TARGET_FOLD_BUILTIN): Detail restriction that
19381 trees be valid for GIMPLE and GENERIC.
19382 (TARGET_GIMPLE_FOLD_BUILTIN): New.
19383 * gimple-fold.c (gimple_fold_call): Call target hook
19384 gimple_fold_builtin.
19385 * hooks.c (hook_bool_gsiptr_false): New.
19386 * hooks.h (hook_bool_gsiptr_false): New.
19387 * target.def (fold_stmt): New.
19388 * doc/tm.texi: Regenerate.
19389
88a581da
VM		 193902013-04-22 Vladimir Makarov <vmakarov@redhat.com>
19391
19392 PR target/57018
19393 * lra-eliminations.c (mark_not_eliminable): Prevent elimination of
19394 a set sp if no stack realignment.
19395
92be22dc
NC		 193962013-04-22 Nick Clifton <nickc@redhat.com>
19397
19398 * config.gcc (tilegx-linux): Extend extra_objs rather than
19399 overwriting it.
19400 (tilepro-linux): Likewise.
19401
0ddec79f
JG		 194022013-04-22 James Greenhalgh <james.greenhalgh@arm.com>
19403
19404 * config/aarch64/aarch64-builtins.c
19405 (CF): Remove.
19406 (CF0, CF1, CF2, CF3, CF4, CF10): New.
19407 (VAR<1-12>): Add MAP parameter.
19408 (BUILTIN_*): Likewise.
19409 * config/aarch64/aarch64-simd-builtins.def: Set MAP parameter.
19410 * config/aarch64/aarch64-simd.md (aarch64_sshl_n<mode>): Remove.
19411 (aarch64_ushl_n<mode>): Likewise.
19412 (aarch64_sshr_n<mode>): Likewise.
19413 (aarch64_ushr_n<mode>): Likewise.
19414 (aarch64_<maxmin><mode>): Likewise.
19415 (aarch64_sqrt<mode>): Likewise.
19416 * config/aarch64/arm_neon.h (vshl<q>_n_*): Use new builtin names.
19417 (vshr<q>_n_*): Likewise.
19418
0050faf8
JG		 194192013-04-22 James Greenhalgh <james.greenhalgh@arm.com>
19420
19421 * config/aarch64/aarch64-builtins.c
19422 (aarch64_simd_builtin_type_mode): Handle SF types.
19423 (sf_UP): Define.
19424 (BUILTIN_GPF): Define.
19425 (aarch64_init_simd_builtins): Handle SF types.
19426 * config/aarch64/aarch64-simd-builtins.def (frecpe): Add support.
19427 (frecps): Likewise.
19428 (frecpx): Likewise.
19429 * config/aarch64/aarch64-simd.md
19430 (simd_types): Update simd_frcp<esx> to simd_frecp<esx>.
19431 (aarch64_frecpe<mode>): New.
19432 (aarch64_frecps<mode>): Likewise.
19433 * config/aarch64/aarch64.md (unspec): Add UNSPEC_FRECP<ESX>.
19434 (v8type): Add frecp<esx>.
19435 (aarch64_frecp<FRECP:frecp_suffix><mode>): New.
19436 (aarch64_frecps<mode>): Likewise.
19437 * config/aarch64/iterators.md (FRECP): New.
19438 (frecp_suffix): Likewise.
19439 * config/aarch64/arm_neon.h
19440 (vrecp<esx><qsd>_<fd><32, 64>): Convert to using builtins.
19441
0fad3dbc 194422013-04-22 Christian Bruel <christian.bruel@st.com>
1818d01c
CB		 19443
19444 PR target/56995
19445 * config/sh/sh.h (enum reg_class): Remove DF_HI_REGS.
19446 (REG_CLASS_NAMES): Idem.
19447 (REG_CLASS_CONTENTS): Idem.
19448 (REGCLASS_HAS_FP_REG): Idem.
19449 * config/sh/sh.c (sh_cannot_change_mode_class): Idem.
19450 (sh_conditional_register_usage): Idem.
19451
3e8a33f9
JL		 194522013-04-21 Jeff Law <law@redhat.com>
19453
19454 * tree-ssa-forwprop.c (simplify_conversion_from_bitmask): New function.
19455 (ssa_forward_propagate_and_combine): Use it.
19456
f38e1b0a
VM		 194572013-04-19 Vladimir Makarov <vmakarov@redhat.com>
19458
19459 * lra.c: Update the flow chart diagram.
19460
682303da
VM		 194612013-04-19 Vladimir Makarov <vmakarov@redhat.com>
19462
19463 PR rtl-optimization/56847
19464 * lra-constraints.c (process_alt_operands): Discourage alternative
19465 with non-matche doffsettable memory constraint fro memory with
19466 known offset.
19467
f6b64c35
RB		 194682013-04-19 Richard Biener <rguenther@suse.de>
19469
19470 PR tree-optimization/56982
19471 * builtins.def (BUILT_IN_LONGJMP): longjmp is not a leaf
19472 function.
19473 * gimplify.c (gimplify_call_expr): Notice special calls.
19474 (gimplify_modify_expr): Likewise.
19475 * tree-cfg.c (make_abnormal_goto_edges): Handle setjmp-like
19476 abnormal control flow receivers.
19477 (call_can_make_abnormal_goto): Handle cfun->calls_setjmp
19478 in the same way as cfun->has_nonlocal_labels.
19479 (gimple_purge_dead_abnormal_call_edges): Likewise.
19480 (stmt_starts_bb_p): Make setjmp-like abnormal control flow
19481 receivers start a basic-block.
19482
01d8bf07
RB		 194832013-04-19 Richard Biener <rguenther@suse.de>
19484
19485 * tree-vectorizer.h (struct _slp_instance): Move load_permutation
19486 member ...
19487 (struct _slp_tree): ... here. Make it a vector of unsigned ints.
19488 (SLP_INSTANCE_LOAD_PERMUTATION): Remove.
19489 (SLP_TREE_LOAD_PERMUTATION): Add.
19490 (vect_transform_slp_perm_load): Adjust prototype.
19491 * tree-vect-slp.c (vect_free_slp_tree): Adjust.
19492 (vect_free_slp_instance): Likewise.
19493 (vect_create_new_slp_node): Likewise.
19494 (vect_supported_slp_permutation_p): Remove.
19495 (vect_slp_rearrange_stmts): Adjust.
19496 (vect_supported_load_permutation_p): Likewise. Inline
19497 vect_supported_slp_permutation_p here.
19498 (vect_analyze_slp_instance): Compute load permutations per
19499 slp node instead of per instance.
19500 (vect_get_slp_defs): Adjust.
19501 (vect_transform_slp_perm_load): Likewise.
19502 (vect_schedule_slp_instance): Remove redundant code.
19503 (vect_schedule_slp): Remove hack for PR56270, add it ...
19504 * tree-vect-stmts.c (vectorizable_load): ... here, do not
19505 CSE loads for SLP. Adjust.
19506
ede22fc3
GY		 195072013-04-19 Greta Yorsh <Greta.Yorsh@arm.com>
19508
19509 * config/arm/arm.c (load_multiple_sequence, ldm_stm_operation_p): Fix
19510 spelling in two comments.
19511
67bc84fb
GY		 195122013-04-19 Greta Yorsh <Greta.Yorsh@arm.com>
19513
19514 PR target/56797
19515 * config/arm/arm.c (load_multiple_sequence): Require SP
19516 as base register for loads if SP is in the register list.
19517
e248d83f
MJ		 195182013-04-19 Martin Jambor <mjambor@suse.cz>
19519
19520 PR tree-optimization/56718
19521 * ipa-cp.c (ipa_value_from_known_type_jfunc): Moved...
19522 * ipa-prop.c (ipa_binfo_from_known_type_jfunc): ...here, renamed
19523 and made public. Adjusted all callers.
19524 (ipa_intraprocedural_devirtualization): New function.
19525 * ipa-prop.h (ipa_binfo_from_known_type_jfunc): Declare.
19526 (ipa_intraprocedural_devirtualization): Likewise.
19527 * Makefile.in (tree-ssa-pre.o): Add ipa-prop.h to dependencies.
19528
4891e8f8
RB		 195292013-04-19 Richard Biener <rguenther@suse.de>
19530
19531 PR tree-optimization/57000
19532 * tree-ssa-reassoc.c (pass_reassoc): Add TODO_update_ssa_only_virtuals.
19533
dad89f7c
TG		 195342013-04-19 Terry Guo <terry.guo@arm.com>
19535
19536 * config/arm/cortex-m4-fpu.md (cortex_m4_v): Delete cpu unit.
19537 Replace with ...
19538 (cortex_m4_v_a, cortex_m4_v_b): ... new cpu units.
19539 (cortex_m4_v, cortex_m4_exa_va, cortex_m4_exb_vb): New reservations.
19540 (cortex_m4_fmacs): Use new reservations.
19541 (cortex_m4_f_load, cortex_m4_f_store): Likewise.
19542
72ea0d47
VM		 195432013-04-18 Vladimir Makarov <vmakarov@redhat.com>
19544
f1e6512c 19545 PR rtl-optimization/56999
72ea0d47
VM		 19546 * lra-coalesce.c (coalescable_pseudo_p): Remove 2nd parameter and
19547 related code.
19548 (lra_coalesce): Remove split_origin_bitmap and related code.
19549 * lra.c (lra): Coalesce after undoing inheritance. Recreate live
19550 ranges if necessary.
19551
780a5b71
UB		 195522013-04-18 Uros Bizjak <ubizjak@gmail.com>
19553
19554 * config/i386/i386.c (x86_64_ms_sysv_extra_clobbered_registers):
19555 New array.
19556 (ix86_expand_call): Remove clobbered_registers array and use
19557 x86_64_ms_sysv_extra_clobbered_registers instead.
19558 * config/i386/i386.h (x86_64_ms_sysv_extra_clobbered_registers):
19559 Declare here.
19560 * config/i386/predicates.md (call_rex64_ms_sysv_operation): New
19561 predicate.
19562 * config/i386/i386.md (*call_rex64_ms_sysv): Use
19563 call_rex64_ms_sysv_operation predicate. Remove explicit clobbers.
19564 (*call_value_rex64_ms_sysv): Ditto.
19565
6f5a366a
CC		 195662013-04-18 Cary Coutant <ccoutant@google.com>
19567
19568 * dwarf2out.c (output_pubnames): Check die_perennial_p of
19569 parent instead of die_mark.
19570
475b8f37
DN		 195712013-04-18 Diego Novillo <dnovillo@google.com>
19572
19573 * gimple.c (create_gimple_tmp): New.
19574 (get_expr_type): New.
19575 (build_assign): New.
19576 (build_type_cast): New.
19577 * gimple.h (enum ssa_mode): Define.
19578 (gimple_seq_set_location): New.
19579 * asan.c (build_check_stmt): Change some gimple_build_* calls
780a5b71 19580 to use build_assign and build_type_cast.
475b8f37 19581
08940f33
RB		 195822013-04-18 Richard Biener <rguenther@suse.de>
19583
19584 * tree-vect-data-refs.c (vect_analyze_group_access): Properly
19585 handle negative step. Remove redundant checks.
19586 (vect_create_data_ref_ptr): Avoid ICEs with non-constant steps.
19587 * tree-vect-stmts.c (vectorizable_load): Instead of asserting
19588 for negative step and grouped loads fail to vectorize.
19589
0e0f87d4
SB		 195902013-04-18 Steven Bosscher <steven@gcc.gnu.org>
19591
19592 * emit-rtl.c (reset_insn_used_flags): New function.
19593 (reset_all_used_flags): Use it.
19594 (verify_insn_sharing): New function.
19595 (verify_rtl_sharing): Fix verification for SEQUENCEs.
19596
4c445590
JJ		 195972013-04-18 Jakub Jelinek <jakub@redhat.com>
19598
19599 PR tree-optimization/56984
19600 * tree-vrp.c (register_edge_assert_for_2): For (x >> M) < N
19601 and (x >> M) >= N don't register any assertion if N << M is the
19602 minimum value.
19603
6873ecab
SB		 196042013-04-18 Steven Bosscher <steven@gcc.gnu.org>
19605
19606 * lower-subreg.c (resolve_simple_move): If called self-recursive,
19607 do not delete_insn insns that have not yet been emitted, only
19608 unlink them with remove_insn.
19609 * df-scan.c (df_insn_delete): Revert r197492.
19610
3ccb989e
SB		 196112013-04-17 Steven Bosscher <steven@gcc.gnu.org>
19612
19613 * emit-rtl.c (link_insn_into_chain): Handle chaining of SEQUENCEs.
19614 * reorg.c (emit_delay_sequence): Simplify with emit-rtl API.
19615
196162013-04-17 Greta Yorsh <Greta.Yorsh@arm.com>
0480fff0
GY		 19617
19618 * config/arm/arm.md (movsicc_insn): Convert define_insn into
19619 define_insn_and_split.
19620 (and_scc,ior_scc,negscc): Likewise.
19621 (cmpsi2_addneg, subsi3_compare): Convert to named patterns.
19622
3ccb989e 196232013-04-17 Greta Yorsh <Greta.Yorsh@arm.com>
367075fe
GY		 19624
19625 * config/arm/arm.c (use_return_insn): Return 0 for targets that
19626 can benefit from using a sequence of LDRD instructions in epilogue
19627 instead of a single LDM instruction.
19628
6d10a203
MLI		 196292013-04-17 Manuel López-Ibáñez <manu@gcc.gnu.org>
19630
19631 PR 45688
19632 * doc/extend.texi: Fix typo.
19633
6983e6b5
RB		 196342013-04-17 Richard Biener <rguenther@suse.de>
19635
19636 * tree-vect-slp.c (vect_build_slp_tree_1): Split out from ...
19637 (vect_build_slp_tree): ... here.
19638 (vect_build_slp_tree_1): Compute which stmts of the SLP group
19639 match. Remove special-casing of mismatched complex loads.
19640 (vect_build_slp_tree): Based on the result from vect_build_slp_tree_1
19641 re-try the match with swapped commutative operands.
19642 (vect_supported_load_permutation_p): Remove special-casing of
19643 mismatched complex loads.
19644 (vect_analyze_slp_instance): Adjust.
19645
ef23e6a2
RB		 196462013-04-17 Richard Biener <rguenther@suse.de>
19647
19648 PR rtl-optimization/56921
19649 * cfgloop.h (struct loop): Add simple_loop_desc member.
19650 (struct niter_desc): Mark with GTY(()).
19651 (simple_loop_desc): Do not use aux field but simple_loop_desc.
19652 * loop-iv.c (get_simple_loop_desc): Likewise.
19653 (free_simple_loop_desc): Likewise.
19654
19655 Revert
19656 2013-04-16 Richard Biener <rguenther@suse.de>
19657
19658 PR rtl-optimization/56921
19659 * loop-init.c (pass_rtl_move_loop_invariants): Add
19660 TODO_do_not_ggc_collect to todo_flags_finish.
19661 (pass_rtl_unswitch): Same.
19662 (pass_rtl_unroll_and_peel_loops): Same.
19663 (pass_rtl_doloop): Same.
19664
fc6f94f5
EB		 196652013-04-17 Eric Botcazou <ebotcazou@adacore.com>
19666
19667 * tree-ssa-alias.c (nonoverlapping_component_refs_of_decl_p): New.
19668 (decl_refs_may_alias_p): Add REF1 and REF2 parameters.
19669 Use nonoverlapping_component_refs_of_decl_p to disambiguate component
19670 references.
19671 (refs_may_alias_p_1): Adjust call to decl_refs_may_alias_p.
19672 * tree-streamer.c (record_common_node): Adjust reference in comment.
19673
10a88311
TG		 196742013-04-17 Terry Guo <terry.guo@arm.com>
19675
19676 * config/arm/cortex-m4.md: Add a new bypass.
19677
6d9b7208 196782013-04-16 Naveen H.S <Naveen.Hurugalawadi@caviumnetworks.com>
f3d096b0
N		 19679
19680 * config/aarch64/aarch64.md (*adds_<optab><mode>_multp2):
19681 New pattern.
19682 (*subs_<optab><mode>_multp2): New pattern.
19683 (*adds_<optab><ALLX:mode>_<GPI:mode>): New pattern.
19684 (*subs_<optab><ALLX:mode>_<GPI:mode>): New pattern.
19685
6d9b7208 196862013-04-16 Naveen H.S <Naveen.Hurugalawadi@caviumnetworks.com>
860ad33f
N		 19687
19688 * config/aarch64/aarch64.md (*adds_mul_imm_<mode>): New pattern.
19689 (*subs_mul_imm_<mode>): New pattern.
19690
18a6701e
DE		 196912013-04-16 David Edelsohn <dje.gcc@gmail.com>
19692
19693 PR target/56948
19694 * config/rs6000/vsx.md (vsx_mov<mode>): Add j->r alternative.
19695 (vsx_movti_64bit): Change j->wa to O->wa. Add n->r alternative.
19696 (vsx_movti_32bit): Change j->wa to O->wa.
19697
07c37b2f
RB		 196982013-04-16 Richard Biener <rguenther@suse.de>
19699
19700 PR rtl-optimization/56921
19701 * loop-init.c (pass_rtl_move_loop_invariants): Add
19702 TODO_do_not_ggc_collect to todo_flags_finish.
19703 (pass_rtl_unswitch): Same.
19704 (pass_rtl_unroll_and_peel_loops): Same.
19705 (pass_rtl_doloop): Same.
19706
0e0f87d4 197072013-04-16 Greta Yorsh <Greta.Yorsh@arm.com>
974ba22d
GY		 19708
19709 * config/arm/arm.c (emit_multi_reg_push): New declaration
19710 for an existing function.
19711 (arm_emit_strd_push): New function.
19712 (arm_expand_prologue): Used here.
19713 (arm_emit_ldrd_pop): New function.
19714 (arm_expand_epilogue): Used here.
19715 (arm_get_frame_offsets): Update condition.
19716 (arm_emit_multi_reg_pop): Add a special case for load of a single
19717 register with writeback.
19718
5e8e2af4
UB		 197192013-04-16 Uros Bizjak <ubizjak@gmail.com>
19720
19721 * doc/invoke.texi (i386 Option): Reword -mstack-protector-guard
19722 description.
19723
9fd9ccf7
RB		 197242013-04-16 Richard Biener <rguenther@suse.de>
19725
19726 PR tree-optimization/56756
19727 * tree-ssa-loop-im.c (struct first_mem_ref_loc_1): New functor.
19728 (first_mem_ref_loc): New.
19729 (execute_sm): Place the load temporarily before a previous
19730 access instead of in the latch edge to ensure its SSA dependencies
19731 are defined at points dominating the load.
19732
96fba521
SB		 197332013-04-16 Steven Bosscher <steven@gcc.gnu.org>
19734
4c8af858
SB		 19735 * cfgrtl.c (cfg_layout_merge_blocks): Revert r184005, implement
19736 correct fix by moving header and footer insn to the footer of
19737 the merged basic block. Clear BB_END of the merged-away block.
19738
96fba521
SB		 19739 PR middle-end/43631
19740 * emit-rtl.c (make_note_raw): New function.
19741 (link_insn_into_chain): New static inline function.
19742 (add_insn): Use it.
19743 (add_insn_before, add_insn_after): Factor insn chain linking code...
19744 (add_insn_before_nobb, add_insn_after_nobb): ...here, new functions
19745 using link_insn_into_chain.
19746 (note_outside_basic_block_p): New helper function for emit_note_after
19747 and emit_note_before.
19748 (emit_note_after): Use nobb variant of add_insn_after if the note
19749 should not be contained in a basic block.
19750 (emit_note_before): Use nobb variant of add_insn_before if the note
19751 should not be contained in a basic block.
19752 (emit_note_copy): Use make_note_raw.
19753 (emit_note): Likewise.
19754 * bb-reorder.c (insert_section_boundary_note): Remove hack to set
19755 BLOCK_FOR_INSN to NULL manually for NOTE_INSN_SWITCH_TEXT_SECTIONS.
19756 * jump.c (cleanup_barriers): Use reorder_insns_nobb to avoid making
19757 the moved barrier the tail of the basic block it follows.
19758 * var-tracking.c (pass_variable_tracking): Add TODO_verify_flow.
19759
7b8265ba
JJ		 197602013-04-15 Jakub Jelinek <jakub@redhat.com>
19761
19762 PR tree-optimization/56962
19763 * gimple-ssa-strength-reduction.c (record_increment): Only set
19764 initializer if gimple_assign_rhs_code is {,POINTER_}PLUS_EXPR and
19765 either rhs1 or rhs2 is equal to c->base_expr.
19766
5185d248
RB		 197672013-04-15 Richard Biener <rguenther@suse.de>
19768
19769 PR tree-optimization/56933
19770 * tree-vectorizer.h (struct _stmt_vec_info): Remove read_write_dep
19771 member.
19772 (GROUP_READ_WRITE_DEPENDENCE): Remove.
19773 (STMT_VINFO_GROUP_READ_WRITE_DEPENDENCE): Likewise.
19774 * tree-vect-data-refs.c (vect_analyze_group_access): Move
19775 dependence check ...
19776 vect_analyze_data_ref_dependence (vect_analyze_data_ref_dependence):
19777 ... here.
19778 * tree-vect-stmts.c (new_stmt_vec_info): Do not initialize
19779 GROUP_READ_WRITE_DEPENDENCE.
19780
a24243a0
AK		 197812013-04-15 Andreas Krebbel <Andreas.Krebbel@de.ibm.com>
19782
19783 * emit-rtl.c (reset_all_used_flags): New function.
19784 (verify_rtl_sharing): Call reset_all_used_flags before and after
19785 performing the checks.
19786
1c50eada
KT		 197872013-04-15 Kyrylo Tkachov <kyrylo.tkachov@arm.com>
19788
19789 * config/arm/arm.c (const_ok_for_dimode_op): Handle AND case.
19790 * config/arm/arm.md (*anddi3_insn): Change to insn_and_split.
19791 * config/arm/constraints.md (De): New constraint.
19792 * config/arm/neon.md (anddi3_neon): Delete.
19793 (neon_vand<mode>): Expand to standard anddi3 pattern.
19794 * config/arm/predicates.md (imm_for_neon_inv_logic_operand):
19795 Move earlier in the file.
19796 (neon_inv_logic_op2): Likewise.
19797 (arm_anddi_operand_neon): New predicate.
19798
e927b6ad
RO		 197992013-04-15 Rainer Orth <ro@CeBiTec.Uni-Bielefeld.DE>
19800
19801 * configure.ac (gcc_cv_ld_as_needed): Set
19802 gcc_cv_ld_as_needed_option, gcc_cv_no_as_needed_option.
19803 Use -z ignore, -z record on *-*-solaris2*.
19804 (HAVE_LD_AS_NEEDED): Update comment.
19805 (LD_AS_NEEDED_OPTION, LD_NO_AS_NEEDED_OPTION): Define.
19806 * configure: Regenerate.
19807 * config.in: Regenerate.
19808 * gcc.c (init_gcc_specs) [USE_LD_AS_NEEDED]: Use
19809 LD_AS_NEEDED_OPTION, LD_NO_AS_NEEDED_OPTION.
19810 * config/sol2.h [HAVE_LD_AS_NEEDED] (USE_LD_AS_NEEDED): Define.
19811 * doc/tm.texi.in (USE_LD_AS_NEEDED): Allow for --as-needed
19812 equivalents. Fix markup.
19813 * doc/tm.texi: Regenerate.
19814
e0ea8797
AH		 198152013-04-15 Andrew Hsieh <andrewhsieh.google.com>
19816
19817 * config/i386/i386.opt: New option mstack-protector-guard=.
19818 * config/i386/i386-opts.h: Add enum stack_protector_guard.
19819 * config/i386/i386.h: Define TARGET_SSP_GLOBAL_GUARD and
19820 TARGET_SSP_TLS_GUARD.
19821 * config/i386/i386.c (ix86_option_override_internal): Set
19822 ix86_stack_protector_guard.
19823 * config/i386/i386.md (stack_protect_set): Enable for
19824 TARGET_SSP_TLS_GUARD only.
19825 (stack_protect_set_<mode>): Ditto.
19826 (stack_protect_test): Ditto.
19827 (stack_protect_test_<mode>): Ditto.
19828 * doc/invoke.texi (i386 Option): Document.
19829
811b72f9
EB		 198302013-04-15 Eric Botcazou <ebotcazou@adacore.com>
19831
19832 PR target/56890
19833 * config/sparc/sparc.c (enum sparc_mode_class): Add H_MODE value.
19834 (S_MODES): Set H_MODE bit.
19835 (SF_MODES): Set only S_MODE and SF_MODE bits.
19836 (DF_MODES): Set SF_MODES and only D_MODE and DF_MODE bits.
19837 (sparc_init_modes) <MODE_INT>: Set H_MODE bit for sub-word modes.
19838 <MODE_VECTOR_INT>: Do not set SF_MODE for sub-word modes.
19839 <MODE_FLOAT>: Likewise.
19840
5529fdd6
JY		 198412013-04-15 Joey Ye <joey.ye@arm.com>
19842
19843 * config/arm/arm.c (thumb_far_jump_used_p): Fix typo in comments.
19844
517b1da2
JY		 198452013-04-15 Joey Ye <joey.ye@arm.com>
19846
19847 * config/arm/arm.c (thumb1_final_prescan_insn): Assert lr save
19848 for real far jump.
19849 (thumb_far_jump_used_p): Count instruction size and set
19850 far_jump_used.
19851
01007ae0
EB		 198522013-04-14 Eric Botcazou <ebotcazou@adacore.com>
19853
19854 * reorg.c (fill_simple_delay_slots): Reindent block of code.
19855 * resource.c (mark_target_live_regs): Reformat conditional block.
19856
c46f6580
SB		 198572013-04-13 Steven Bosscher <steven@gcc.gnu.org>
19858
19859 * sched-deps.c (deps_analyze_insn): Do not check for EH_REGION insn
19860 notes, they are emitted only just before final.
19861 * sched-int.h: Include insn-attr.h before checking INSN_SCHEDULING.
19862
1f397f45
SB		 198632013-04-13 Steven Bosscher <steven@gcc.gnu.org>
19864
19865 * emit-rtl.c (remove_insn): Do not call df_insn_delete here.
19866 * cfgrtl.c (delete_insn): Call it here instead.
19867 * lra-spills.c (lra_final_code_change): Use delete_insn.
19868 * haifa-sched.c (sched_remove_insn): Likewise.
19869 * sel-sched-ir.c (return_nop_to_pool): Clear INSN_DELETED_P for nops
19870 returning to the nop pool.
19871 (sel_remove_insn): Simplify the only_disconnect case via remove_insn,
19872 use delete_insn for definitive removal. Clear BLOCK_FOR_INSN.
19873
58a51369
SB		 198742013-04-12 Steven Bosscher <steven@gcc.gnu.org>
19875
19876 * doc/tm.texi.in (LOOP_ALIGN): Remove loop note references.
19877 * doc/tm.texi: Regenerated.
19878
33159866
UB		 198792013-04-12 Uros Bizjak <ubizjak@gmail.com>
19880
19881 * config/i386/i386.c (ix86_hard_regno_mode_ok): Use ANY_QI_REGNO_P in
19882 QImode checks.
19883
226e378f
SB		 198842013-04-12 Steven Bosscher <steven@gcc.gnu.org>
19885
19886 * df-core.c (df_find_def): Compare register numbers.
19887 (df_find_use): Likewise.
19888
fafb9b18
VM		 198892013-04-12 Vladimir Makarov <vmakarov@redhat.com>
19890
19891 PR target/56903
19892 * config/i386/i386.c (ix86_hard_regno_mode_ok): Add
19893 lra_in_progress for return.
19894
9a946fd6
GY		 198952013-04-12 Greta Yorsh <Greta.Yorsh@arm.com>
19896
19897 * config/arm/arm.md (mov_scc,mov_negscc,mov_notscc): Convert
19898 define_insn into define_insn_and_split and emit movsicc patterns.
19899
41b83758
GY		 199002013-04-12 Greta Yorsh <Greta.Yorsh@arm.com>
19901
19902 * config/arm/arm.c (gen_operands_ldrd_strd): Initialize "base".
19903
d3afd9aa
RB		 199042013-04-12 Richard Biener <rguenther@suse.de>
19905
19906 * tree-pass.h (TODO_do_not_ggc_collect): New.
19907 * passes.c (execute_one_ipa_transform_pass): Honor
19908 TODO_do_not_ggc_collect.
19909 (execute_one_pass): Likewise.
19910
19911 Revert
19912 2013-04-10 Richard Biener <rguenther@suse.de>
19913
19914 * passes.c (init_optimization_passes): Remove reload pass.
19915 * ira.c (do_reload): Merge into ...
19916 (ira): ... this.
19917 (rest_of_handle_reload): Remove.
19918 (pass_reload): Likewise.
19919 * config/i386/i386.c (ix86_option_override): Refer to ira instead
19920 of reload for vzeroupper pass placement.
19921
06f9b387
JJ		 199222013-04-12 Jakub Jelinek <jakub@redhat.com>
19923
19924 PR tree-optimization/56918
19925 PR tree-optimization/56920
19926 * fold-const.c (int_const_binop_1): Use op1.mul_with_sign (op2, ...)
19927 instead of op1 - op2. Pass 2 * TYPE_PRECISION (type) as second
19928 argument to rshift method. For 2 * HOST_BITS_PER_WIDE_INT precision
19929 use wide_mul_with_sign method.
19930
953094d2
RB		 199312013-04-12 Richard Biener <rguenther@suse.de>
19932
19933 * gimple.c (is_gimple_constant): Vector CONSTRUCTORs should
19934 not be considered a gimple constant.
19935
bb506982
MG		 199362013-04-12 Marc Glisse <marc.glisse@inria.fr>
19937
19938 * fold-const.c (const_binop): Handle vector shifts by a scalar.
19939 (fold_binary_loc): Call const_binop also for mixed vector-scalar
19940 operations.
19941
4b84d650
JJ		 199422013-04-12 Manuel López-Ibáñez <manu@gcc.gnu.org>
19943 Jakub Jelinek <jakub@redhat.com>
19944
19945 * opts.c: Include diagnostic-color.h.
19946 (common_handle_option): Handle OPT_fdiagnostics_color_.
19947 * Makefile.in (OBJS-libcommon): Add diagnostic-color.o.
19948 (diagnostic.o, opts.o, pretty-print.o): Depend on diagnostic-color.h.
19949 (diagnostic-color.o): New.
19950 * common.opt (fdiagnostics-color, fdiagnostics-color=): New options.
19951 (diagnostic_color_rule): New enum.
19952 * dwarf2out.c (gen_producer_string): Don't print -fdiagnostics-color*.
19953 * langhooks.c (lhd_print_error_function): Add %r "locus" and %R around
19954 the location string.
19955 * diagnostic.def: Add 3rd argument to DEFINE_DIAGNOSTIC_KIND macros,
19956 either NULL, or color kind.
19957 * diagnostic-color.c: New file.
19958 * diagnostic-color.h: New file.
19959 * diagnostic-core.h (DEFINE_DIAGNOSTIC_KIND): Adjust macro for 3
19960 arguments.
19961 * doc/invoke.texi (-fdiagnostics-color): Document.
19962 * pretty-print.h (pp_show_color): Define.
19963 (struct pretty_print_info): Add show_color field.
19964 * diagnostic.c: Include diagnostic-color.h.
19965 (diagnostic_build_prefix): Adjust for 3 argument DEFINE_DIAGNOSTIC_KIND
19966 macros. Colorize error:, warning: etc. strings and also the location
19967 string.
19968 (diagnostic_show_locus): Colorize the caret line.
19969 * pretty-print.c: Include diagnostic-color.h.
19970 (pp_base_format): Handle %r and %R format specifiers. Colorize strings
19971 inside of %< %> quotes or quoted through q format modifier.
19972
067a1e71
AK		 199732013-04-12 Andreas Krebbel <Andreas.Krebbel@de.ibm.com>
19974
33159866 19975 * ifcvt.c (end_ifcvt_sequence): Mark a and b for unsharing as well.
067a1e71 19976
33159866 199772013-04-11 Naveen H.S <Naveen.Hurugalawadi@caviumnetworks.com>
b056c910
N		 19978
19979 * config/aarch64/aarch64.c (aarch64_select_cc_mode): Allow NEG
19980 code in CC_NZ mode.
19981 * config/aarch64/aarch64.md (*neg_<shift><mode>3_compare0): New
19982 pattern.
19983
7b55f98f
MP		 199842013-04-11 Marek Polacek <polacek@redhat.com>
19985
19986 PR tree-optimization/48184
33159866 19987 * params.def (PARAM_ALIGN_THRESHOLD): Increase the minimum value to 1.
7b55f98f 19988
966b587e
EB		 199892013-04-11 Eric Botcazou <ebotcazou@adacore.com>
19990
19991 * stor-layout.c (skip_simple_constant_arithmetic): Move to...
19992 * tree.c (skip_simple_constant_arithmetic): ...here and make public.
19993 (skip_simple_arithmetic): Tidy up.
19994 * tree.h (skip_simple_constant_arithmetic): Declare.
19995
33159866 199962013-04-11 Naveen H.S <Naveen.Hurugalawadi@caviumnetworks.com>
f8bf91ab
N		 19997
19998 * config/aarch64/aarch64.h (REVERSIBLE_CC_MODE): Define.
19999
1d42c1ec
RB		 200002013-04-11 Richard Biener <rguenther@suse.de>
20001
20002 * tree-vect-loop.c (get_initial_def_for_induction): Properly
20003 generate vector constants.
20004
4ba5ea11
RB		 200052013-04-11 Richard Biener <rguenther@suse.de>
20006
20007 PR tree-optimization/56878
20008 * tree-flow.h (outermost_invariant_loop_for_expr): Declare.
20009 * tree-ssa-loop-ivopts.c (outermost_invariant_loop_for_expr):
20010 New function.
20011 * tree-vect-data-refs.c (vect_enhance_data_refs_alignment):
20012 Prefer to align the DR with the most invariant base address.
20013
f0defe58
SKS		 200142013-04-11 Senthil Kumar Selvaraj <senthil_kumar.selvaraj@atmel.com>
20015
20016 * opts.c (common_handle_option): Fix formatting and add FALLTHRU
20017 comment.
20018
d07458be
JG		 200192013-04-11 James Greenhalgh <james.greenhalgh@arm.com>
20020
20021 * config/aarch64/aarch64-simd.md (aarch64_vcond_internal): Fix
20022 floating-point vector comparisons against 0.
20023
146b8692
JJ		 200242013-04-11 Jakub Jelinek <jakub@redhat.com>
20025
20026 PR tree-optimization/56899
20027 * fold-const.c (extract_muldiv_1): Apply distributive law
20028 only if TYPE_OVERFLOW_WRAPS (ctype).
20029
b8578ff7
BC		 200302013-04-11 Bin Cheng <bin.cheng@arm.com>
20031
20032 PR target/56124
20033 * ira-costs.c (scan_one_insn): Check whether the source rtx of
20034 loading has side effect.
20035
0ea8a6f9
SB		 200362013-04-10 Steven Bosscher <steven@gcc.gnu.org>
20037
20038 * config/sparc/sparc.c: Include tree-pass.h.
20039 (TARGET_MACHINE_DEPENDENT_REORG): Do not redefine.
20040 (sparc_reorg): Rename to sparc_do_work_around_errata. Move to
20041 head of file. Change return type. Split off gate function.
20042 (sparc_gate_work_around_errata): New function.
20043 (pass_work_around_errata): New pass definition.
20044 (insert_pass_work_around_errata) New pass insert definition to
20045 insert pass_work_around_errata just after delayed-branch scheduling.
20046 (sparc_option_override): Insert the pass.
20047 * config/sparc/t-sparc (sparc.o): Add TREE_PASS_H dependence.
20048
42e37616
DM		 200492013-04-10 David S. Miller <davem@davemloft.net>
20050
89deeb3b
DM		 20051 * config/sparc/sparc.h (ASM_CPU_SPEC): Pass -Av8 if -mcpu=supersparc
20052 or -mcpu=hypersparc.
20053
42e37616
DM		 20054 * target.def (cstore_mode): New hook.
20055 * target.h: Include insn-codes.h
20056 * targhooks.c: Likewise.
20057 (default_cstore_mode): New function.
20058 * targhooks.h: Declare it.
20059 * doc/tm.texi.in: New hook slot for TARGET_CSTORE_MODE.
20060 * doc/tm.texi: Rebuild.
20061 * expmed.c (emit_cstore): Obtain cstore boolean result mode using
20062 target hook, rather than inspecting the insn_data.
20063 * config/sparc/sparc.c (sparc_cstore_mode): New function.
20064 (TARGET_CSTORE_MODE): Redefine.
20065 (emit_scc_insn): When TARGET_ARCH64, emit new 64-bit boolean
20066 result patterns.
20067 * config/sparc/predicates.md (cstore_result_operand): New special
20068 predicate.
20069 * config/sparc/sparc.md (cstoresi4, cstoredi4, cstore<F:mode>4):
20070 Use it for operand 0.
20071 (*seqsi_special): Rewrite using 'P' mode iterator on operand 0.
20072 (*snesi_special): Likewise.
20073 (*snesi_zero): Likewise.
20074 (*seqsi_zero): Likewise.
20075 (*sltu_insn): Likewise.
20076 (*sgeu_insn): Likewise.
20077 (*seqdi_special): Make operand 0 and comparison operation be of
20078 DImode.
20079 (*snedi_special): Likewise.
20080 (*snedi_special_vis3): Likewise.
20081 (*neg_snesi_zero): Rename to *neg_snesisi_zero.
20082 (*neg_snesi_sign_extend): Rename to *neg_snesidi_zero.
20083 (*snesi_zero_extend): Delete, covered by 'P' mode iterator.
20084 (*neg_seqsi_zero): Rename to *neg_seqsisi_zero.
20085 (*neg_seqsi_sign_extend): Rename to *neg_seqsidi_zero.
20086 (*seqsi_zero_extend): Delete, covered by 'P' mode iterator.
20087 (*sltu_extend_sp64): Likewise.
20088 (*neg_sltu_insn): Rename to *neg_sltusi_insn.
20089 (*neg_sltu_extend_sp64): Rename to *neg_sltudi_insn.
20090 (*sgeu_extend_sp64): Delete, covered by 'P' mode iterator.
20091 (*neg_sgeu_insn): Rename to *neg_sgeusi_insn.
20092 (*neg_sgeu_extend_sp64): Rename to *neg_sgeudi_insn.
20093
95ca411e
YZ		 200942013-04-10 Yufeng Zhang <yufeng.zhang@arm.com>
20095
20096 * config/aarch64/aarch64.c (aarch64_print_extension): New function.
20097 (aarch64_start_file): Use the new function.
20098
6782438d 200992013-04-10 Senthil Kumar Selvaraj <senthil_kumar.selvaraj@atmel.com>
33159866 20100 Jason Merrill <jason@redhat.com>
6782438d
SKS		 20101
20102 * common.opt: Add -gdwarf.
20103 * opts.c (common_handle_option): Handle it.
20104 * gcc.c (ASM_DEBUG_SPEC): Don't expect "-2" for DWARF.
20105
bb313b93
RB		 201062013-04-10 Richard Biener <rguenther@suse.de>
20107
20108 * passes.c (execute_todo): Do not call ggc_collect conditional here.
20109 (execute_one_ipa_transform_pass): But unconditionally here.
20110 (execute_one_pass): And here.
20111 (init_optimization_passes): Remove reload pass.
20112 * tree-pass.h (TODO_ggc_collect): Remove.
20113 (pass_reload): Likewise.
20114 * ira.c (do_reload): Merge into ...
20115 (ira): ... this.
20116 (rest_of_handle_reload): Remove.
20117 (pass_reload): Likewise.
20118 * config/i386/i386.c (ix86_option_override): Refer to ira instead
20119 of reload for vzeroupper pass placement.
20120 * <everywhere>: Remove TODO_ggc_collect from todo_flags_start
20121 and todo_flags_finish of all passes.
20122
793d9a16
RB		 201232013-04-10 Richard Biener <rguenther@suse.de>
20124
20125 * tree-vectorizer.h (struct _slp_oprnd_info): Remove
20126 first_const_oprnd field, rename first_def_type to first_op_type.
20127 * tree-vect-slp.c (vect_create_oprnd_info): Adjust.
20128 (vect_get_and_check_slp_defs): Always use the type of the
20129 operand. Allow mixed vect_external_def, vect_constant_def types.
20130 (vect_get_constant_vectors): Handle mixed vect_external_def,
20131 vect_constant_def types.
20132
12211b99 201332013-04-10 Joern Rennecke <joern.rennecke@embecosm.com>
ee8a9b7b
JR		 20134
20135 PR tree-optimization/55524
20136 * tree-ssa-math-opts.c
20137 (convert_mult_to_fma): Don't use an fms construct
20138 when we don't have an fms operation, but fnma, and it looks
20139 likely that we'll be able to use the latter.
20140
12211b99 201412013-04-10 Zhouyi Zhou <yizhouzhou@ict.ac.cn>
6957a6f6
ZZ		 20142
20143 * cif-code.def (OVERWRITABLE): Correct the comment for overwritable
20144 function.
20145 * ipa-inline.c (can_inline_edge_p): Let dump mechanism report the
20146 inline fail caused by overwritable functions.
20147
34ab4a5b
CJW		 201482013-04-10 Chung-Ju Wu <jasonwucj@gmail.com>
20149
20150 * combine.c (simplify_compare_const): Use GET_MODE_MASK to filter out
20151 unnecessary bits in the constant power of two case.
20152
abf9bfbc
RB		 201532013-04-10 Richard Biener <rguenther@suse.de>
20154
20155 * tree-vect-slp.c (vect_get_and_check_slp_defs): Remove
20156 broken code swapping operands.
20157 (vect_build_slp_tree): Do not compute load permutations here.
20158 (vect_analyze_slp_instance): Compute load permutations here,
20159 after building the SLP tree.
20160
f408477e
CB		 201612013-04-09 Christian Bruel <christian.bruel@st.com>
20162
20163 * config/sh/sh.md (barrier_align): Use next/prev_active_insn instead
20164 of next/prev_real_insn.
20165
5ac42672
JH		 201662013-04-09 Jan Hubicka <jh@suse.cz>
20167
abf9bfbc
RB		 20168 * ipa.c (cgraph_externally_visible_p, varpool_externally_visible_p):
20169 Drop aliased parameter.
5ac42672
JH		 20170 (function_and_variable_visibility): Do not handle alias pairs.
20171 * cgraph.c (varpool_externally_visible_p): Update prototype.
20172 * varpool.c (varpool_add_new_variable): Update.
20173
5017f1d2
KT		 201742013-04-09 Kyrylo Tkachov <kyrylo.tkachov@arm.com>
20175
20176 * config/arm/arm.md (minmax_arithsi_non_canon): New pattern.
20177
48eecbee
SB		 201782013-04-09 Steven Bosscher <steven@gcc.gnu.org>
20179
a949cf1c
SB		 20180 * sched-vis.c (print_pattern): Print SEQUENCE of insns as insns.
20181
48eecbee
SB		 20182 * config/sparc/sparc.md: Use define_c_enum for "unspec" and "unspecv".
20183
75ef8e3d
MP		 201842013-04-09 Marek Polacek <polacek@redhat.com>
20185
20186 PR tree-optimization/48762
33159866 20187 * params.def (PARAM_MAX_CSE_INSNS): Increase the minimum value to 1.
75ef8e3d 20188
23847df4
RB		 201892013-04-09 Richard Biener <rguenther@suse.de>
20190
20191 * tree-vect-slp.c (vect_get_and_check_slp_defs): Remove code
20192 dealing with cost.
20193 (vect_build_slp_tree): Likewise.
20194 (vect_analyze_slp_cost_1, vect_analyze_slp_cost): New functions
20195 calculating the cost of a SLP instance.
20196 (vect_analyze_slp_instance): Use it from here, after building
20197 the SLP tree.
20198
ea3a0fde
JJ		 201992013-04-09 Jakub Jelinek <jakub@redhat.com>
20200
20201 PR middle-end/56883
20202 * omp-low.c (expand_omp_for_generic, expand_omp_for_static_nochunk,
20203 expand_omp_for_static_chunk): Use simple_p = true in
20204 force_gimple_operand_gsi calls when assigning to addressable decls.
20205
a32dfe9d
JL		 202062013-04-09 Jeff Law <law@redhat.com>
20207
20208 * tree-vrp.c (simplify_cond_using_ranges): Simplify test of boolean
20209 when the boolean was created by converting a wider object which
20210 had a boolean range.
20211
d755c7ef
RB		 202122013-04-09 Richard Biener <rguenther@suse.de>
20213
20214 * tree-vectorizer.h (slp_void_p): Remove.
20215 (slp_tree): Typedef before _slp_tree declaration.
20216 (struct _slp_tree): Use a vector of slp_tree as children.
20217 (vect_get_place_in_interleaving_chain): Remove.
20218 * tree-vect-data-refs.c (vect_get_place_in_interleaving_chain):
20219 Move ...
20220 * tree-vect-slp.c (vect_get_place_in_interleaving_chain): ... here
20221 and make static.
20222 (vect_free_slp_tree, vect_print_slp_tree, vect_mark_slp_stmts,
20223 vect_mark_slp_stmts_relevant, vect_slp_rearrange_stmts,
20224 vect_detect_hybrid_slp_stmts, vect_slp_analyze_node_operations,
20225 vect_schedule_slp_instance, vect_remove_slp_scalar_calls):
20226 Use slp_node instead of slp_void_p and adjust.
20227
3d741091
RB		 202282013-04-09 Richard Biener <rguenther@suse.de>
20229
20230 * tree-ssa-loop-manip.c (rewrite_into_loop_closed_ssa): Avoid
20231 work that is not necessary.
20232
39307ba7
JJ		 202332013-04-09 Jakub Jelinek <jakub@redhat.com>
20234
20235 PR tree-optimization/56854
20236 * tree-ssa-forwprop.c (forward_propagate_addr_expr_1): Don't
20237 forward into clobber stmts if it would change MEM_REF lhs into
20238 non-MEM_REF.
20239
343881fd
MK		 202402013-04-09 Maxim Kuvyrkov <maxim@kugelworks.com>
20241
20242 * tree.c (type_hash_lookup, type_hash_add): Make static.
20243 * tree.h (type_hash_lookup, type_hash_add): Remove global declarations.
20244
3922658a
RB		 202452013-04-09 Richard Biener <rguenther@suse.de>
20246
20247 * tree.h (unsave_expr_now): Remove.
20248 * tree-inline.c (mark_local_for_remap_r): Remove.
20249 (unsave_expr_1): Likewise.
20250 (unsave_r): Likewise.
20251 (unsave_expr_now): Likewise.
20252 * tree-ssa-copy.c (replace_exp_1): Use unshare_expr.
20253 (propagate_tree_value): Likewise.
20254
9fb6b620
SB		 202552013-04-08 Steven Bosscher <steven@gcc.gnu.org>
20256
20257 * doc/rtl.texi (sequence): Rewrite documentation to match the
20258 current use of SEQUENCE rtl objects.
20259 * rtl.def (SEQUENCE): Likewise.
20260
20261 * doc/rtl.texi (NOTE_INSN_EH_REGION_BEG, NOTE_INSN_EH_REGION_END):
20262 Update documentation.
20263 (NOTE_INSN_LOOP_BEG, NOTE_INSN_LOOP_END, NOTE_INSN_LOOP_CONT,
20264 NOTE_INSN_LOOP_VTOP): Remove documentation for non-existing notes.
20265
20266 * reg-notes.def (REG_EH_CONTEXT): Remove unused note.
20267
8ddb5a29
TJ		 202682013-04-08 Teresa Johnson <tejohnson@google.com>
20269
20270 * basic-block.h (GCOV_COMPUTE_SCALE): Define.
20271 * ipa-inline-analysis.c (param_change_prob): Use helper rounding divide
33159866 20272 methods.
8ddb5a29
TJ		 20273 (estimate_edge_size_and_time): Add comment to suggest using rounding
20274 methods.
20275 (estimate_node_size_and_time): Ditto.
20276 (remap_edge_change_prob): Use helper rounding divide methods.
20277 * value-prof.c (gimple_divmod_fixed_value_transform): Ditto.
20278 (gimple_mod_pow2_value_transform): Ditto.
20279 (gimple_mod_subtract_transform): Ditto.
20280 (gimple_ic_transform): Ditto.
20281 (gimple_stringops_transform): Ditto.
20282 * stmt.c (conditional_probability): Ditto.
20283 (emit_case_dispatch_table): Ditto.
20284 * lto-cgraph.c (merge_profile_summaries): Ditto.
20285 * tree-optimize.c (execute_fixup_cfg): Ditto.
20286 * cfgcleanup.c (try_forward_edges): Ditto.
20287 * cfgloopmanip.c (scale_loop_profile): Ditto.
20288 (loopify): Ditto.
20289 (duplicate_loop_to_header_edge): Ditto.
20290 (lv_adjust_loop_entry_edge): Ditto.
20291 * tree-vect-loop.c (vect_transform_loop): Ditto.
20292 * profile.c (compute_branch_probabilities): Ditto.
20293 * cfgbuild.c (compute_outgoing_frequencies): Ditto.
20294 * lto-streamer-in.c (input_cfg): Ditto.
20295 * gimple-streamer-in.c (input_bb): Ditto.
20296 * ipa-cp.c (update_profiling_info): Ditto.
20297 (update_specialized_profile): Ditto.
20298 * tree-vect-loop-manip.c (slpeel_tree_peel_loop_to_edge): Ditto.
20299 * cfg.c (update_bb_profile_for_threading): Add comment to suggest using
33159866 20300 rounding methods.
8ddb5a29
TJ		 20301 * sched-rgn.c (compute_dom_prob_ps): Ditto.
20302 (compute_trg_info): Ditto.
20303 * cfgrtl.c (force_nonfallthru_and_redirect): Ditto.
20304 (purge_dead_edges): Ditto.
20305 * loop-unswitch.c (unswitch_loop): Ditto.
20306 * cgraphclones.c (cgraph_clone_edge): Ditto.
20307 (cgraph_clone_node): Ditto.
20308 * tree-inline.c (copy_bb): Ditto.
20309 (copy_edges_for_bb): Ditto.
20310 (initialize_cfun): Ditto.
20311 (copy_cfg_body): Ditto.
20312 (expand_call_inline): Ditto.
20313
661e6bd7
KT		 203142013-04-08 Kai Tietz <ktietz@redhat.com>
20315
20316 * config/i386/cygwin.h (EXTRA_OS_CPP_BUILTINS): Replaced
20317 TARGET_CYGWIN64 by TARGET_64BIT.
20318
105766f3
JR		 203192013-04-08 Joern Rennecke <joern.rennecke@embecosm.com>
20320
20321 * config/epiphany/epiphany.md (GPR_1): New constant.
20322 (define_expand "mov<mode>cc): FAIL if gen_compare_reg returned 0.
20323 * config/epiphany/epiphany.c (gen_compare_reg):
20324 For flag_finite_math_only, avoid swapping operands when r0 and/or r1
20325 is already in place.
20326 Use GPR_0 / GPR_1 instead of 0/1 for r0/r1 register numbers.
20327 Don't require being called during rtl expansion; If y operlaps r0,
20328 return 0.
20329 (epiphany_compute_frame_size, epiphany_expand_prologue): Use GPR_1.
20330 (epiphany_expand_epilogue): Likewise.
20331
8afab237
JR		 20332 * config/epiphany/epiphany.c (epiphany_select_cc_mode):
20333 Don't use CC_FPmode for ORDERED / UNORDERED.
33159866 20334 * config/epiphany/epiphany.md (cmpsf_ord): Make pattern unconditional.
8afab237 20335
fa7707d6
JR		 20336 * config/epiphany/constraints.md (CnL): New constraint.
20337 * config/epiphany/epiphany.md (addsi3_i): Add r/r/CnL alternative.
20338 * config/epiphany/predicates.md (add_operand): Allow 1024.
20339
5138e18d
JR		 20340 * config/epiphany/epiphany.md (logical_op): New code iterator.
20341 (op_mnc): New code attribute.
20342 (<op_mnc>_f, mov_f, cstoresi4): New patterns.
20343 (mov_f+1, mov_f+2): New peephole2 patterns.
20344
2ccc703d
JR		 20345 * config/epiphany/epiphany.md (mov_f+2): New peephole2 pattern.
20346 (cstoresi4): Also allow re-use of zero result when doing a NE
20347 comparison to a non-zero operand.
aefb0819 20348 Use (clobber (scratch)) for first insn if the gpr output is not needed.
2ccc703d 20349
093ac0a5
JR		 20350 * config/epiphany/epiphany.md (<insn_opname>v2si3):
20351 Use gen_addsi3_i / gen_subsi3_i.
20352
f223bb13
JJ		 203532013-04-08 Jakub Jelinek <jakub@redhat.com>
20354
20355 PR c++/34949
20356 PR c++/50243
20357 * tree-eh.c (optimize_clobbers): Only remove clobbers if bb doesn't
20358 contain anything but clobbers, at most one __builtin_stack_restore,
20359 optionally debug stmts and final resx, and if it has at least one
20360 incoming EH edge. Don't check for SSA_NAME on LHS of a clobber.
20361 (sink_clobbers): Don't check for SSA_NAME on LHS of a clobber.
20362 Instead of moving clobbers with MEM_REF LHS with SSA_NAME address
20363 which isn't defaut definition, remove them.
20364 (unsplit_eh, cleanup_empty_eh): Use single_{pred,succ}_{p,edge}
20365 instead of EDGE_COUNT comparisons or EDGE_{PRED,SUCC}.
20366 * tree-ssa-ccp.c (execute_fold_all_builtins): Remove clobbers
20367 with MEM_REF LHS with SSA_NAME address.
20368
4481581f
JL		 203692013-04-08 Jeff Law <law@redhat.com>
20370
20371 * gimple.c (canonicalize_cond_expr_cond): Rewrite x ^ y into x != y.
a32dfe9d 20372
451dabda
RB		 203732013-04-08 Richard Biener <rguenther@suse.de>
20374
20375 * gimple-pretty-print.c (debug_gimple_stmt): Do not print
20376 extra newline.
20377 * tree-vect-loop.c (vect_determine_vectorization_factor): Dump
20378 determined vector type.
20379 (vect_analyze_data_refs): Likewise.
20380 (vect_get_new_vect_var): Adjust.
20381 (vect_create_destination_var): Preserve SSA name versions.
20382 * tree-vect-stmts.c (get_vectype_for_scalar_type_and_size): Do
20383 not dump anything here.
20384
3b088b47
JR		 203852013-04-08 Joern Rennecke <joern.rennecke@embecosm.com>
20386
20387 * config/epiphany/epiphany.h (struct GTY (()) machine_function):
20388 Add member lr_slot_known.
20389 * config/epiphany/epiphany.md (reload_insi_ra): Compute lr_slot_offs
20390 if necessary.
20391 * config/epiphany/epiphany.c (epiphany_compute_frame_size):
20392 Remove code that sets lr_slot_offset according to what a previous
20393 version of epiphany_emit_save_restore used to do.
20394 (epiphany_emit_save_restore): When doing an lr save or restore,
20395 set/verify lr_slot_known and lr_slot_offset.
20396
d8484d41
XQ		 203972013-04-08 Xinyu Qi <xyqi@marvell.com>
20398
33159866
UB		 20399 PR target/54338
20400 * config/arm/arm.h (REG_CLASS_CONTENTS): Include IWMMXT_GR_REGS
d8484d41
XQ		 20401 in ALL_REGS.
20402
481be1c4
RB		 204032013-04-08 Richard Biener <rguenther@suse.de>
20404
20405 * alias.c (find_base_term): Fix thinko in previous change.
20406
401f3a81
JJ		 204072013-04-08 Jakub Jelinek <jakub@redhat.com>
20408
20409 * tree-loop-distribution.c (const_with_all_bytes_same): New function.
20410 (generate_memset_builtin): Only handle integer_all_onesp as -1 val if
20411 TYPE_PRECISION is equal to mode bitsize. Use const_with_all_bytes_same
20412 if possible to compute val.
20413 (classify_partition): Verify CONSTRUCTOR doesn't have any elts.
20414 For QImode integers don't require anything about precision. Use
20415 const_with_all_bytes_same to find out if the constant doesn't have
20416 repeated bytes in it.
20417
03ed99a8
AK		 204182013-04-08 Andreas Krebbel <Andreas.Krebbel@de.ibm.com>
20419
20420 * config/s390/s390.c (s390_expand_insv): Only accept insertions
20421 within mode size.
20422
781b2e62
MP		 204232013-04-08 Marek Polacek <polacek@redhat.com>
20424
20425 PR rtl-optimization/48182
20426 * params.def (PARAM_MIN_CROSSJUMP_INSNS): Increase the minimum
20427 value to 1.
20428
27e430a2
JDA		 204292013-04-06 John David Anglin <dave.anglin@nrc-cnrc.gc.ca>
20430
20431 PR target/55487
20432 * config/pa/pa.c (legitimize_pic_address): Before incrementing label
20433 nuses, make sure we have a label.
20434
4902aa64
BS		 204352013-04-05 Bill Schmidt <wschmidt@linux.vnet.ibm.com>
20436
20437 PR target/56843
20438 * config/rs6000/rs6000.c (rs6000_emit_swdiv_high_precision): Remove.
20439 (rs6000_emit_swdiv_low_precision): Remove.
20440 (rs6000_emit_swdiv): Rewrite to handle between one and four
20441 iterations of Newton-Raphson generally; modify required number of
20442 iterations for some cases.
20443 * config/rs6000/rs6000.h (RS6000_RECIP_HIGH_PRECISION_P): Remove.
20444
7bca81dc
SB		 204452013-04-05 Steven Bosscher <steven@gcc.gnu.org>
20446
20447 * bb-reorder.c (fix_crossing_unconditional_branches): Remove a
20448 set-but-unused variable.
20449
20450 * cgraph.c (cgraph_release_function_body): Clear cfun->cfg to make
20451 basic blocks of released function bodies garbage-collectable.
20452
20453 * ree.c (find_and_remove_re): Do not call df_finish_pass here.
20454 (struct rtl_opt_pass): Add TODO_df_finish.
20455
20456 * rtl.def (DEFINE_SUBST, DEFINE_SUBST_ATTR): Add documentation.
20457
4542a38a
GY		 204582013-04-05 Greta Yorsh <Greta.Yorsh@arm.com>
20459
20460 * config/arm/constraints.md (q): New constraint.
20461 * config/arm/ldrdstrd.md: New file.
20462 * config/arm/arm.md (ldrdstrd.md) New include.
20463 (arm_movdi): Use "q" instead of "r" constraint
20464 for double-word memory access.
20465 (movdf_soft_insn): Likewise.
20466 * config/arm/vfp.md (movdi_vfp): Likewise.
20467 * config/arm/t-arm (MD_INCLUDES): Add ldrdstrd.md.
33159866 20468 * config/arm/arm-protos.h (gen_operands_ldrd_strd): New declaration.
4542a38a
GY		 20469 * config/arm/arm.c (gen_operands_ldrd_strd): New function.
20470 (mem_ok_for_ldrd_strd): Likewise.
20471 (output_move_double): Update assertion.
20472
2385b218
GY		 204732013-04-05 Greta Yorsh <Greta.Yorsh@arm.com>
20474
20475 * config/arm/arm.md: Comment on splitting Thumb1 patterns.
20476
75fe1cb5
GY		 204772013-04-05 Greta Yorsh <Greta.Yorsh@arm.com>
20478
20479 * config/arm/arm.md (arm_smax_insn): Convert define_insn into
20480 define_insn_and_split.
20481 (arm_smin_insn,arm_umaxsi3,arm_uminsi3): Likewise.
20482
dd660e8e
GY		 204832013-04-05 Greta Yorsh <Greta.Yorsh@arm.com>
20484
20485 * config/arm/arm.md (arm_ashldi3_1bit): Convert define_insn into
20486 define_insn_and_split.
33159866 20487 (arm_ashrdi3_1bit,arm_lshrdi3_1bit): Likewise.
dd660e8e
GY		 20488 (shiftsi3_compare): New pattern.
20489 (rrx): New pattern.
20490 * config/arm/unspecs.md (UNSPEC_RRX): New.
20491
045e472c
GY		 204922013-04-05 Greta Yorsh <Greta.Yorsh@arm.com>
20493
20494 * config/arm/arm.md (negdi_extendsidi): New pattern.
20495 (negdi_zero_extendsidi): Likewise.
20496
3f3bf1a8
GY		 204972013-04-05 Greta Yorsh <Greta.Yorsh@arm.com>
20498
20499 * config/arm/arm.md (andsi_iorsi3_notsi): Convert define_insn into
20500 define_insn_and_split.
20501 (arm_negdi2,arm_abssi2,arm_neg_abssi2): Likewise.
20502 (arm_cmpdi_insn,arm_cmpdi_unsigned): Likewise.
20503
b0b49556
GY		 205042013-04-05 Greta Yorsh <Greta.Yorsh@arm.com>
20505
20506 * config/arm/arm.md (arm_subdi3): Convert define_insn into
20507 define_insn_and_split.
20508 (subdi_di_zesidi,subdi_di_sesidi): Likewise.
20509 (subdi_zesidi_di,subdi_sesidi_di,subdi_zesidi_zesidi): Likewise.
20510
d633dd84
GY		 205112013-04-05 Greta Yorsh <Greta.Yorsh@arm.com>
20512
20513 * config/arm/arm.md (subsi3_carryin): New pattern.
20514 (subsi3_carryin_const): Likewise.
20515 (subsi3_carryin_compare,subsi3_carryin_compare_const): Likewise.
20516 (subsi3_carryin_shift,rsbsi3_carryin_shift): Likewise.
20517
f4499066
GY		 205182013-04-05 Greta Yorsh <Greta.Yorsh@arm.com>
20519
20520 * config/arm/arm.md (incscc,arm_incscc,decscc,arm_decscc): Delete.
20521
ceef6fd9
GY		 205222013-04-05 Greta Yorsh <Greta.Yorsh@arm.com>
20523
20524 * config/arm/arm.md (addsi3_carryin_<optab>): Set attribute predicable.
d633dd84 20525 (addsi3_carryin_alt2_<optab>,addsi3_carryin_shift_<optab>): Likewise.
ceef6fd9 20526
ddbdd8a7
KT		 205272013-04-05 Kyrylo Tkachov <kyrylo.tkachov@arm.com>
20528
20529 * config/arm/arm.c (arm_expand_builtin): Change fcode
20530 type to unsigned int.
20531
8456d78a
RR		 205322013-04-05 Ramana Radhakrishnan <ramana.radhakrishnan@arm.com>
20533
20534 * doc/invoke.texi (ARM Options): Document cortex-a53 support.
20535
526c230f
ILT		 205362013-04-04 Ian Lance Taylor <iant@google.com>
20537
33159866
UB		 20538 * doc/standards.texi (Standards): The Go frontend supports the Go 1
20539 language standard.
526c230f 20540
3cfbe04d
SB		 205412013-04-04 Steven Bosscher <steven@gcc.gnu.org>
20542
20543 PR middle-end/56729
20544 * df-scan.c (df_insn_delete): Disable failing assert.
20545
dfa3f8d0
KT		 205462013-04-04 Kyrylo Tkachov <kyrylo.tkachov@arm.com>
20547
20548 * config/arm/arm-protos.h (arm_builtin_vectorized_function):
20549 New function prototype.
20550 * config/arm/arm.c (TARGET_VECTORIZE_BUILTINS): Define.
20551 (TARGET_VECTORIZE_BUILTIN_VECTORIZED_FUNCTION): Likewise.
20552 (arm_builtin_vectorized_function): New function.
20553
f7837758
KT		 205542013-04-04 Kyrylo Tkachov <kyrylo.tkachov@arm.com>
20555
20556 * config/arm/arm_neon_builtins.def: New file.
20557 * config/arm/arm.c (neon_builtin_data): Move contents to
20558 arm_neon_builtins.def.
20559 (enum arm_builtins): Include neon builtin definitions.
20560 (ARM_BUILTIN_NEON_BASE): Move from enum to macro.
33159866 20561 * config/arm/t-arm (arm.o): Add dependency on arm_neon_builtins.def.
f7837758 20562
39c1b6db
MP		 205632013-04-04 Marek Polacek <polacek@redhat.com>
20564
20565 PR tree-optimization/48186
20566 * predict.c (maybe_hot_frequency_p): Return false if
20567 HOT_BB_FREQUENCY_FRACTION is 0.
20568 (cgraph_maybe_hot_edge_p): Likewise.
20569
314f64eb
RB		 205702013-04-04 Richard Biener <rguenther@suse.de>
20571
20572 PR tree-optimization/56826
20573 * tree-vect-slp.c (vect_build_slp_tree): Compute ncopies
20574 more accurately.
20575
51a905b2
RB		 205762013-04-04 Richard Biener <rguenther@suse.de>
20577
20578 PR tree-optimization/56213
20579 * tree-vect-data-refs.c (vect_check_strided_load): Remove.
33159866 20580 (vect_analyze_data_refs): Allow all non-nested loads as strided loads.
51a905b2 20581
f1bf4f3a
RB		 205822013-04-04 Richard Biener <rguenther@suse.de>
20583
20584 PR tree-optimization/56837
20585 * tree-loop-distribution.c (classify_partition): For non-zero
20586 values require that the value has the same precision as its
20587 mode to be useful as memset value.
20588
0bca7ded
NC		 205892013-04-03 Nick Clifton <nickc@redhat.com>
20590
33159866 20591 * config/v850/v850e3v5.md (fmasf4): Use fmaf.s on E3V5 architectures.
0bca7ded
NC		 20592 (fmssf4): Use fmsf.s on E3V5 architectures.
20593 (fnmasf4): Use fnmaf.s on E3V5 architectures.
20594 (fnmssf4): Use fnmsf.s on E3V5 architectures.
20595
b4019227
JL		 205962013-04-03 Jeff Law <law@redhat.com>
20597
20598 * Makefile.in (lra-constraints.o): Depend on $(OPTABS_H).
20599 (lra-eliminations.o): Likewise.
20600
f57ddb5b
TJ		 206012013-04-03 Teresa Johnson <tejohnson@google.com>
20602
20603 * gcov-io.c (compute_working_sets): Moved most of body of old
33159866 20604 compute_working_sets here from profile.c.
f57ddb5b 20605 * gcov-io.h (NUM_GCOV_WORKING_SETS): Moved here from profile.c.
33159866
UB		 20606 (gcov_working_set_t): Moved typedef here from basic-block.h
20607 (compute_working_set): Declare.
f57ddb5b
TJ		 20608 * profile.c (NUM_GCOV_WORKING_SETS): Moved to gcov-io.h.
20609 (get_working_sets): Renamed from compute_working_set,
33159866 20610 replace most of body with call to new compute_working_sets.
f57ddb5b 20611 (get_exec_counts): Replace call to compute_working_sets
33159866
UB		 20612 to get_working_sets.
20613 * profile.h (get_working_sets): Renamed from compute_working_set.
f57ddb5b 20614 * lto-cgraph.c (input_symtab): Replace call to compute_working_sets
33159866 20615 to get_working_sets.
f57ddb5b
TJ		 20616 * basic-block.h (gcov_working_set_t): Moved to gcov-io.h.
20617 * gcov-dump.c (dump_working_sets): New function.
20618
12211b99 206192013-04-03 Kenneth Zadeck <zadeck@naturalbridge.com>
5e0919f1
KZ		 20620
20621 * hwint.c (sext_hwi, zext_hwi): New functions.
20622 * hwint.h (HOST_BITS_PER_HALF_WIDE_INT, HOST_HALF_WIDE_INT,
20623 HOST_HALF_WIDE_INT_PRINT, HOST_HALF_WIDE_INT_PRINT_C,
20624 HOST_HALF_WIDE_INT_PRINT_DEC, HOST_HALF_WIDE_INT_PRINT_DEC_C,
20625 HOST_HALF_WIDE_INT_PRINT_UNSIGNED, HOST_HALF_WIDE_INT_PRINT_HEX,
20626 HOST_HALF_WIDE_INT_PRINT_HEX_PURE): New symbols.
20627 (sext_hwi, zext_hwi): New functions.
0bca7ded 20628
be672e08
JL		 206292013-04-03 Jeff Law <law@redhat.com>
20630
20631 PR tree-optimization/56799
33159866
UB		 20632 * tree-ssa-dom.c (record_equivalences_from_incoming_edge): Bring
20633 back test for widening conversion erroneously dropped in prior change.
be672e08 20634
9d821fa5
KT		 206352013-04-03 Kyrylo Tkachov <kyrylo.tkachov@arm.com>
20636
20637 PR target/56809
20638 * config/aarch64/aarch64.c (is_jump_table): Use next_active_insn
20639 instead of next_real_insn.
20640
71c581e7
MP		 206412013-04-03 Marek Polacek <polacek@redhat.com>
20642
20643 PR sanitizer/55702
33159866 20644 * tsan.c (instrument_func_exit): Allow BUILT_IN_RETURN functions.
71c581e7 20645
4a32ef80
KT		 206462013-04-03 Kyrylo Tkachov <kyrylo.tkachov@arm.com>
20647
20648 PR target/56809
20649 * config/arm/arm.c (is_jump_table): Use next_active_insn instead of
20650 next_real_insn.
20651 (thumb1_output_casesi): Likewise.
20652 (thumb2_output_casesi): Likewise.
20653
1b2253d4
RB		 206542013-04-03 Richard Biener <rguenther@suse.de>
20655
20656 PR tree-optimization/56817
20657 * tree-ssa-loop-ivcanon.c (tree_unroll_loops_completely):
20658 Split out ...
20659 (tree_unroll_loops_completely_1): ... new function to manually
20660 walk the loop tree, properly defering outer loops of unrolled
20661 loops to later iterations.
20662
38000232
MG		 206632013-04-03 Marc Glisse <marc.glisse@inria.fr>
20664
20665 * tree-vect-stmts.c (vectorizable_store): Accept BIT_FIELD_REF.
20666 (vectorizable_load): Likewise.
20667 * tree-vect-slp.c (vect_build_slp_tree): Likewise.
20668 * tree-vect-data-refs.c (vect_create_data_ref_ptr): Handle VECTOR_TYPE.
20669
3fa35298
MG		 206702013-04-03 Marc Glisse <marc.glisse@inria.fr>
20671
20672 * tree-flow-inline.h (get_addr_base_and_unit_offset_1): Handle
20673 BIT_FIELD_REF.
20674
b3d45ff0
UW		 206752013-04-03 Ulrich Weigand <Ulrich.Weigand@de.ibm.com>
20676
20677 * config/spu/spu.c (emit_nop_for_insn): Handle JUMP_TABLE_DATA.
20678
ec9202a8
BC		 206792013-04-03 Bin Cheng <bin.cheng@arm.com>
20680
20681 * rtl.h (AUTO_INC_DEC): Fix typo of HAVE_POST_MODIFY_DISP.
20682
6805bd36
MG		 206832013-04-03 Marc Glisse <marc.glisse@inria.fr>
20684
20685 PR tree-optimization/56790
33159866
UB		 20686 * fold-const.c (fold_ternary_loc) <VEC_COND_EXPR>: Add constant
20687 folding.
6805bd36 20688
da694a77
MG		 206892013-04-03 Marc Glisse <marc.glisse@inria.fr>
20690
20691 * simplify-rtx.c (simplify_binary_operation_1) <VEC_SELECT>:
20692 Handle VEC_MERGE.
20693 (simplify_ternary_operation) <VEC_MERGE>: Use unsigned HOST_WIDE_INT
20694 for masks. Test for side effects. Handle nested VEC_MERGE. Handle
20695 equal arguments.
20696
4e7d7b3d
JJ		 206972013-04-03 Jakub Jelinek <jakub@redhat.com>
20698
20699 PR c/19449
20700 * tree.h (force_folding_builtin_constant_p): New decl.
20701 * builtins.c (force_folding_builtin_constant_p): New variable.
20702 (fold_builtin_constant_p): Fold immediately also if
20703 force_folding_builtin_constant_p.
20704
e6c9d234
RB		 207052013-04-03 Richard Biener <rguenther@suse.de>
20706
20707 PR tree-optimization/56812
20708 * tree-vect-data-refs.c (vect_slp_analyze_data_ref_dependence):
20709 DRs of the same interleaving chain are independent.
20710
984d07dd
JM		 207112013-04-02 Jason Merrill <jason@redhat.com>
20712
20713 * gdbinit.in (pbb): Use debug fn.
20714
622849c9
LC		 207152013-04-02 Lawrence Crowl <crowl@google.com>
20716
20717 * sese.h (struct ivtype_map_elt_s): Remove unused.
20718 (extern debug_ivtype_map): Remove unused.
20719 (extern eq_ivtype_map_elts): Remove unused.
20720 * sese.c (debug_ivtype_map): Removed unused.
20721 (debug_ivtype_map_1): Removed unused.
20722 (debug_ivtype_elt): Remove unused.
20723 (eq_ivtype_map_elts): Remove unused.
20724
20725
82c0e1a0
KT		 207262013-04-02 Kai Tietz <ktietz@redhat.com>
20727
20728 PR target/52790
20729 * config/i386/cygming.h (SUB_TARGET_RECORD_STUB): New sub-target macro.
20730 * config/i386/i386-protos.h (i386_pe_record_stub): Add new prototype.
33159866
UB		 20731 * config/i386/i386.c (legitimize_pe_coff_extern_decl): New static
20732 function.
82c0e1a0
KT		 20733 (legitimize_pe_coff_symbol): Likewise.
20734 (is_imported_p): New helper-function.
20735 (ix86_option_override_internal): Make MEDIUM_PIC the default code-model
20736 for Windows x64 targets.
20737 (ix86_expand_prologue): Optimize for pe-coff targets.
20738 (ix86_expand_split_stack_prologue): Adjust for pe-coff targets.
20739 (legitimate_pic_address_disp_p): Adjust for x64 pe-coff to support
20740 medium/large code-model.
20741 (legitimize_pic_address): Likewise.
20742 (legitimize_tls_address): Likewise.
20743 (ix86_expand_call): Likewise.
20744 (x86_output_mi_thunk): Likewise.
20745 (get_dllimport_decl): Add new beimport argument.
20746 (construct_plt_address): Don't assert for x64 pe-coff targets.
20747 * config/i386/i386.h (PIC_OFFSET_TABLE_REGNUM): Adjust for x64 pe-coff
20748 targets.
20749 (SYMBOL_FLAG_STUBVAR): New macro.
20750 (SYMBOL_REF_STUBVAR_P): Likewise.
20751 * config/i386/winnt.c (stub_list): New structure.
20752 (stub_head): New local variable.
20753 (i386_pe_record_stub): New function.
20754 (i386_pe_file_end): Emit refptr-stubs.
20755
5d751b0c
JJ		 207562013-04-02 Jakub Jelinek <jakub@redhat.com>
20757
09bb4c99
JJ		 20758 PR rtl-optimization/56745
20759 * ifcvt.c (cond_exec_find_if_block): Don't try to optimize
20760 if then_bb has no successors and else_bb is EXIT_BLOCK_PTR.
20761
a1d8947a
JJ		 20762 PR c++/34949
20763 * tree-ssa-alias.c (stmt_kills_ref_p_1): If base != ref->base
20764 and both of them are MEM_REFs, just compare first argument for
20765 equality and attempt to deal even with differing offsets.
20766
5d751b0c
JJ		 20767 PR c++/34949
20768 * tree-cfg.c (verify_gimple_assign_single): Allow lhs
20769 of gimple_clobber_p to be MEM_REF.
20770 * gimplify.c (gimplify_modify_expr): Gimplify *to_p of
20771 an assignment from TREE_CLOBBER_P. Allow it to be MEM_REF
20772 after gimplification.
20773 * asan.c (get_mem_ref_of_assignment): Don't instrument
20774 gimple_clobber_p stmts.
20775 * tree-ssa-dse.c (dse_optimize_stmt): Allow DSE of
20776 gimple_clobber_p stmt if they have MEM_REF lhs and
20777 are dead because of another gimple_clobber_p stmt.
20778 * tree-ssa-live.c (clear_unused_block_pointer): Treat
20779 gimple_clobber_p stmts like debug stmts.
20780 (remove_unused_locals): Remove clobbers with MEM_REF lhs
20781 that refer to unused VAR_DECLs or uninitialized values.
20782 * tree-sra.c (sra_ipa_reset_debug_stmts): Also remove
20783 gimple_clobber_p stmts if they refer to removed parameters.
20784 (get_repl_default_def_ssa_name, sra_ipa_modify_expr): Fix up
20785 formatting.
20786
e594716a
UB		 207872013-04-02 Uros Bizjak <ubizjak@gmail.com>
20788
20789 * config/i386/i386.md (*testqi_ext_3): Merge with *testqi_ext_3_rex64
20790 using SWI48 mode attribute.
20791
7a80735b
WM		 207922013-04-02 Wei Mi <wmi@google.com>
20793
20794 * config/i386/i386.c (ix86_rtx_costs): Set proper rtx cost for
20795 ashl<mode>3_mask, *<shift_insn><mode>3_mask and
20796 *<rotate_insn><mode>3_mask in i386.md.
20797
f423a9e4
AI		 207982013-04-02 Alexander Ivchenko <alexander.ivchenko@intel.com>
20799
20800 * config.gcc (arm*-*-linux-*): Remove duplicate t-linux-android.
20801
90eb75f2
RB		 208022013-04-02 Richard Biener <rguenther@suse.de>
20803
20804 PR tree-optimization/56778
20805 * tree-vect-data-refs.c (vect_analyze_data_ref_dependence):
20806 Runtime alias tests are not supported for gather loads.
20807 * tree-vect-loop-manip.c (vect_loop_versioning): Insert
20808 stmts referenced from SSA operands before updating SSA form.
20809
d8c69a92
IC		 208102013-04-02 Ian Caulfield <ian.caulfield@arm.com>
20811 Ramana Radhakrishnan <ramana.radhakrishnan@arm.com>
20812
20813 * config/arm/arm-arches.def (armv8-a): Default to cortex-a53.
e594716a
UB		 20814 * config/arm/t-arm (MD_INCLUDES): Depend on cortex-a53.md.
20815 * config/arm/cortex-a53.md: New file.
20816 * config/arm/bpabi.h (BE8_LINK_SPEC): Handle cortex-a53.
20817 * config/arm/arm.md (generic_sched, generic_vfp): Handle cortex-a53.
20818 * config/arm/arm.c (arm_issue_rate): Likewise.
20819 * config/arm/arm-tune.md: Regenerate
20820 * config/arm/arm-tables.opt: Regenerate.
20821 * config/arm/arm-cores.def: Add cortex-a53.
d8c69a92 20822
239eb04c
ZC		 208232013-04-02 Zhenqiang Chen <zhenqiang.chen@arm.com>
20824
20825 * config/arm/uclinux-elf.h: Add %L to LINK_GCC_C_SEQUENCE_SPEC for
20826 non-static link.
20827
c902d3c8
SN		 208282013-04-02 Sofiane Naci <sofiane.naci@arm.com>
20829
d8c69a92
IC		 20830 * config/aarch64/aarch64.md (*mov<mode>_aarch64): Add variants for
20831 scalar load/store operations using B/H registers.
c902d3c8
SN		 20832 (*zero_extend<SHORT:mode><GPI:mode>2_aarch64): Likewise.
20833
051d0e2f
SN		 208342013-04-02 Sofiane Naci <sofiane.naci@arm.com>
20835
20836 * config/aarch64/aarch64.md (*mov<mode>_aarch64): Add alternatives for
20837 scalar move.
20838 * config/aarch64/aarch64.c
20839 (aarch64_simd_scalar_immediate_valid_for_move): New.
20840 * config/aarch64/aarch64-protos.h
20841 (aarch64_simd_scalar_immediate_valid_for_move): New.
20842 * config/aarch64/constraints.md (Dh, Dq): New.
20843 * config/aarch64/iterators.md (hq): New.
20844
0ee1e3d9
EB		 208452013-04-02 Eric Botcazou <ebotcazou@adacore.com>
20846
20847 * reorg.c (get_branch_condition): Deal with conditional returns.
20848 (fill_simple_delay_slots): Remove dead code dealing with jumps.
20849
136fb3f7
WM		 208502013-04-01 Wei Mi <wmi@google.com>
20851
20852 * config/i386/i386.md (*ashl<mode>3_mask): Rewrite as define_insn.
20853 Truncate operand 2 using %b asm operand modifier.
20854 (*<shift_insn><mode>3_mask): Ditto.
20855 (*<rotate_insn><mode>3_mask): Ditto.
20856
6388c738
SB		 208572013-04-01 Steven Bosscher <steven@gcc.gnu.org>
20858
20859 PR middle-end/56798
20860 * cfgbuild.c (inside_basic_block_p): Restore check broken at r197234.
20861
f7a4d826
KK		 208622013-03-31 Kaz Kojima <kkojima@gcc.gnu.org>
20863
20864 * config/sh/sh.md (casesi_worker_1): Use next_active_insn instead
20865 of next_real_insn.
20866 (casesi_worker_2, casesi_shift_media, casesi_load_media): Likewise.
20867
4b943a49
LC		 208682013-03-30 Lawrence Crowl <crowl@google.com>
20869
20870 * dse.c (clear_alias_sets): Remove never set.
20871 (disqualified_clear_alias_sets): Remove never set.
20872 (clear_alias_mode_pool): Remove never set.
20873 (dse_step0): Remove condition that is never true.
20874 (canon_address): Remove condition that is never true.
20875 (dse_step7): Remove condition that is never true.
20876 (rest_of_handle_dse): Remove condition that is never true.
20877 (rest_of_handle_dse::did_global): Remove never read from above.
20878 (dse_step2_spill): Remove never called from above.
20879 (dse_step5_spill): Remove never called from above.
20880
39718607
SB		 208812013-03-30 Steven Bosscher <steven@gcc.gnu.org>
20882
da5c6bde
SB		 20883 * doc/md.texi (Standard Names) <casesi>: Update documentation for
20884 JUMP_TABLE_DATA changes.
20885 * doc/tm.texi.in (Dispatch Tables) <ASM_OUTPUT_CASE_LABEL>: Likewise.
20886 * doc/rtl.texi (Flags) <INSN_DELETED_P, SCHED_GROUP_P>: Likewise.
20887 (Insns) <jump_table_data>: New entry.
20888 * doc/tm.texi: Regenerate.
20889
39718607
SB		 20890 * cfgrtl.c (fixup_reorder_chain): Do not emit barriers to BB_FOOTER.
20891
20892 * postreload-gcse.c (bb_has_well_behaved_predecessors): Correct test
20893 for table jump at the end of a basic block using tablejump_p.
20894 * targhooks.c (default_invalid_within_doloop): Likewise.
20895 * config/rs6000/rs6000.c (TARGET_INVALID_WITHIN_DOLOOP): Remove
20896 target hook implementation that is identical to the default hook.
20897 (rs6000_invalid_within_doloop): Remove.
20898
20899 * bb-reorder.c (fix_crossing_unconditional_branches): Remove set but
20900 unused variable from tablejump_p call.
20901
20902 * rtl.def (JUMP_TABLE_DATA): New RTX_INSN object.
20903 * rtl.h (RTX_PREV, RTX_NEXT): Adjust for new JUMP_TABLE_DATA.
20904 (INSN_DELETED_P): Likewise.
20905 (emit_jump_table_data): New prototype.
20906 * gengtype.c (adjust_field_rtx_def): Handle JUMP_TABLE_DATA fields
20907 after 4th as unused.
20908 * print-rtl.c (print_rtl): Handle JUMP_TABLE_DATA.
20909 * sched-vis.c (print_insn): Likewise.
20910 * emit-rtl.c (active_insn_p): Consider JUMP_TABLE_DATA an active
20911 insn for compatibility with back ends that use next_active_insn to
20912 identify jump table data.
20913 (set_insn_deleted): Remove no longer useful JUMP_TABLE_DATA_P check.
20914 (remove_insn): Likewise.
20915 (emit_insn): Do not accept JUMP_TABLE_DATA objects in insn chains
20916 to be emitted.
20917 (emit_debug_insn, emit_jump_insn, emit_call_insn, emit_label): Idem.
20918 (emit_jump_table_data): New function.
20919
20920 * cfgbuild.c (inside_basic_block_p): A JUMP_INSN is always inside a
20921 basic block, a JUMP_TABLE_DATA never is.
20922 (control_flow_insn_p): JUMP_TABLE_DATA is not a control flow insn.
20923 * cfgrtl.c (duplicate_insn_chain): Split handling of JUMP_TABLE_DATA
20924 off from code handling real insns.
20925 * final.c (get_attr_length_1): Simplify for JUMP_INSNs.
20926 * function.c (instantiate_virtual_regs): Remove JUMP_TABLE_DATA_P
20927 test, now redundant because JUMP_TABLE_DATA is not an INSN_P insn.
20928 * gcse.c (insert_insn_end_basic_block): Likewise, JUMP_TABLE_DATA_P
20929 is not a NONDEBUG_INSN_P.
20930 * ira-costs.c (scan_one_insn): Likewise.
20931 * jump.c (mark_all_labels): Likewise.
20932 (mark_jump_label_1): Likewise.
20933 * lra-eliminations.c (eliminate_regs_in_insn): Likewise.
20934 * lra.c (get_insn_freq): Expect all insns reaching here to be in
20935 a basic block.
20936 (check_rtl): Remove JUMP_TABLE_DATA_P test, not a NONDEBUG_INSN_P insn.
20937 * predict.c (expensive_function_p): Use FOR_BB_INSNS.
20938 * reload1.c (calculate_needs_all_insns): Call set_label_offsets for
20939 JUMP_TABLE_DATA_P insns.
20940 (calculate_elim_costs_all_insns): Likewise.
20941 (set_label_offsets): Recurse on the PATTERN of JUMP_TABLE_DATA insns.
20942 (elimination_costs_in_insn): Remove redundant JUMP_TABLE_DATA_P test.
20943 (delete_output_reload): Code style fixups.
20944 * reorg.c (dbr_schedule): Move JUMP_TABLE_DATA_P up to avoid setting
20945 insn flags on this non-insn.
20946 * sched-rgn.c (add_branch_dependences): Treat JUMP_TABLE_DATA insns
20947 as scheduling barriers, for pre-change compatibility.
20948 * stmt.c (emit_case_dispatch_table): Emit jump table data not as
20949 JUMP_INSN objects but instead as JUMP_TABLE_DATA objects.
20950
20951 * config/alpha/alpha.c (alpha_does_function_need_gp): Remove
20952 redundant JUMP_TABLE_DATA_P test.
20953 * config/arm/arm.c (thumb_far_jump_used_p): Likewise.
20954 * config/frv/frv.c (frv_function_contains_far_jump): Likewise.
20955 (frv_for_each_packet): Likewise.
20956 * config/i386/i386.c (min_insn_size): Likewise.
20957 (ix86_avoid_jump_mispredicts): Likewise.
20958 * config/m32r/m32r.c (m32r_is_insn): Likewise.
20959 * config/mep/mep.c (mep_reorg_erepeat): Likewise.
20960 * config/mips/mips.c (USEFUL_INSN_P): Likewise.
20961 (mips16_insn_length): Robustify.
20962 (mips_has_long_branch_p): Remove redundant JUMP_TABLE_DATA_P test.
20963 (mips16_split_long_branches): Likewise.
20964 * config/pa/pa.c (pa_combine_instructions): Likewise.
20965 * config/rs6000/rs6000.c (get_next_active_insn): Treat
20966 JUMP_TABLE_DATA objects as active insns, like in active_insn_p.
20967 * config/s390/s390.c (s390_chunkify_start): Treat JUMP_TABLE_DATA
20968 as contributing to pool range lengths.
20969 * config/sh/sh.c (find_barrier): Restore check for ADDR_DIFF_VEC.
20970 Remove redundant JUMP_TABLE_DATA_P test.
20971 (sh_loop_align): Likewise.
20972 (split_branches): Likewise.
20973 (sh_insn_length_adjustment): Likewise.
20974 * config/spu/spu.c (get_branch_target): Likewise.
20975
0208f7da
JH		 209762013-03-29 Jan Hubicka <jh@suse.cz>
20977
20978 * lto-cgraph.c (output_profile_summary, input_profile_summary): Use
20979 gcov streaming; stream hot bb threshold to ltrans.
20980 * predict.c (get_hot_bb_threshold): Break out from ....
20981 (maybe_hot_count_p): ... here.
20982 (set_hot_bb_threshold): New function.
20983 * lto-section-in.c (lto_section_name): Add profile.
20984 * profile.h (get_hot_bb_threshold, set_hot_bb_threshold): Declare.
20985 * ipa.c: Include hash-table.h, tree-inline.h, profile.h, lto-streamer.h
20986 and data-streamer.h
20987 (histogram_entry): New structure.
20988 (histogram, histogram_pool): New global vars.
20989 (histogram_hash): New structure.
20990 (histogram_hash::hash): New method.
20991 (histogram_hash::equal): Likewise.
20992 (account_time_size): New function.
20993 (cmp_counts): New function.
20994 (dump_histogram): New function.
20995 (ipa_profile_generate_summary): New function.
20996 (ipa_profile_write_summary): New function.
20997 (ipa_profile_read_summary): New function.
20998 (ipa_profile): Decide on threshold.
e594716a
UB		 20999 (pass_ipa_profile): Add ipa_profile_write_summary and
21000 ipa_profile_read_summary.
0208f7da
JH		 21001 * Makefile.in (ipa.o): Update dependencies.
21002 * lto-streamer.h (LTO_section_ipa_profile): New section.
21003
5a6ccc94
GDR		 210042013-03-29 Gabriel Dos Reis <gdr@integrable-solutions.net>
21005
21006 * tree.h (VAR_P): New.
21007
39385fa6
PC		 210082013-03-29 Paolo Carlini <paolo.carlini@oracle.com>
21009
21010 PR lto/56777
21011 * doc/invoke.texi ([-fwhole-program]): Fix typo.
21012
34f0d87a
SB		 210132013-03-29 Steven Bosscher <steven@gcc.gnu.org>
21014
21015 * cfgbuild.c (inside_basic_block_p): Use JUMP_TABLE_DATA_P in lieu
21016 of tests for JUMP_P and a ADDR_DIFF_VEC or ADDR_VEC pattern.
21017 (control_flow_insn_p): Likewise.
21018 * cfgrtl.c (duplicate_insn_chain): Likewise.
21019 * final.c (get_attr_length_1): Likewise.
21020 (shorten_branches): Likewise.
21021 (final_scan_insn): Likewise.
21022 * function.c (instantiate_virtual_regs): Likewise.
21023 * gcse.c (insert_insn_end_basic_block): Likewise.
21024 * ira-costs.c (scan_one_insn): Likewise.
21025 * lra-eliminations.c (eliminate_regs_in_insn): Likewise.
21026 * lra.c (check_rtl): Likewise.
21027 * reload1.c (elimination_costs_in_insn): Likewise.
21028 * reorg.c (follow_jumps): Likewise.
21029
21030 * config/arm/arm.c (is_jump_table): Use JUMP_TABLE_DATA_P in lieu
21031 of tests for JUMP_P and a ADDR_DIFF_VEC or ADDR_VEC pattern.
21032 (thumb_far_jump_used_p): Likewise.
21033 * config/bfin/bfin.c (workaround_rts_anomaly): Likewise.
21034 (workaround_speculation): Likewise.
21035 (add_sched_insns_for_speculation): Likewise.
21036 * config/c6x/c6x.c (reorg_emit_nops): Likewise.
21037 * config/frv/frv.c (frv_function_contains_far_jump): Likewise.
21038 (frv_for_each_packet): Likewise.
21039 * config/i386/i386.c (ix86_avoid_jump_mispredicts): Likewise.
21040 * config/ia64/ia64.c (emit_all_insn_group_barriers): Likewise.
21041 (final_emit_insn_group_barriers): Likewise.
21042 * config/m32r/m32r.c (m32r_is_insn): Likewise.
21043 * config/mips/mips.c (USEFUL_INSN_P): Likewise.
21044 (mips16_insn_length): Likewise.
21045 * config/pa/pa.c (pa_reorg): Likewise.
21046 (pa_combine_instructions): Likewise.
21047 * config/rs6000/rs6000.c (rs6000_invalid_within_doloop): Likewise.
21048 * config/sh/sh.c (fixup_addr_diff_vecs): Likewise.
21049 (sh_reorg): Likewise.
21050 (split_branches): Likewise.
21051 * config/spu/spu.c (get_branch_target): Likewise.
21052
21053 * config/s390/s390.c (s390_chunkify_start): Simplify logic using
21054 JUMP_TABLE_DATA_P.
21055
4ac761b0
KY		 210562013-03-29 Kirill Yukhin <kirill.yukhin@intel.com>
21057
39385fa6 21058 * config/i386/avx2intrin.h (_mm256_broadcastsi128_si256):
4ac761b0
KY		 21059 Fix declaration name.
21060
58a49189
LC		 210612013-03-28 Lawrence Crowl <crowl@google.com>
21062
21063 * graphds.h (struct graph.indicies): Remove unused.
21064 * graphite-poly.h (struct graph.original_pddrs): Remove unused.
21065 (SCOP_ORIGINAL_PDDRS): Remove unused.
21066 * sese.h (extern insert_loop_close_phis): Removed unused.
21067 (extern insert_guard_phis): Removed unused.
21068 (extern ivtype_map_elt_info): Removed unused.
21069 (new_ivtype_map_elt): Removed unused.
21070 * sese.c (ivtype_map_elt_info): Removed unused.
21071
7b3b6ae4
LC		 210722013-03-28 Lawrence Crowl <crowl@google.com>
21073
21074 * Makefile.in: Add several missing include dependences.
21075 (DUMPFILE_H): New.
21076 (test-dump.o): New. This object is not added to any executable,
21077 but is present for ad-hoc testing.
21078 * bitmap.c
21079 (debug (const bitmap_head_def &)): New.
21080 (debug (const bitmap_head_def *)): New.
21081 * bitmap.h
21082 (extern debug (const bitmap_head_def &)): New.
21083 (extern debug (const bitmap_head_def *)): New.
21084 * cfg.c
21085 (debug (edge_def &)): New.
21086 (debug (edge_def *)): New.
21087 * cfghooks.c
21088 (debug (basic_block_def &)): New.
21089 (debug (basic_block_def *)): New.
21090 * dumpfile.h
21091 (dump_node (const_tree, int, FILE *)): Correct source file.
21092 * dwarf2out.c
21093 (debug (die_struct &)): New.
21094 (debug (die_struct *)): New.
21095 * dwarf2out.h
21096 (extern debug (die_struct &)): New.
21097 (extern debug (die_struct *)): New.
21098 * gimple-pretty-print.c
21099 (debug (gimple_statement_d &)): New.
21100 (debug (gimple_statement_d *)): New.
21101 * gimple-pretty-print.h
21102 (extern debug (gimple_statement_d &)): New.
21103 (extern debug (gimple_statement_d *)): New.
21104 * ira-build.c
21105 (debug (ira_allocno_copy &)): New.
21106 (debug (ira_allocno_copy *)): New.
21107 (debug (ira_allocno &)): New.
21108 (debug (ira_allocno *)): New.
21109 * ira-int.h
21110 (extern debug (ira_allocno_copy &)): New.
21111 (extern debug (ira_allocno_copy *)): New.
21112 (extern debug (ira_allocno &)): New.
21113 (extern debug (ira_allocno *)): New.
21114 * ira-lives.c
21115 (debug (live_range &)): New.
21116 (debug (live_range *)): New.
21117 * lra-int.h
21118 (debug (lra_live_range &)): New.
21119 (debug (lra_live_range *)): New.
21120 * lra-lives.c
21121 (debug (lra_live_range &)): New.
21122 (debug (lra_live_range *)): New.
21123 * omega.c
21124 (debug (omega_pb_d &)): New.
21125 (debug (omega_pb_d *)): New.
21126 * omega.h
21127 (extern debug (omega_pb_d &)): New.
21128 (extern debug (omega_pb_d *)): New.
21129 * print-rtl.c
21130 (debug (const rtx_def &)): New.
21131 (debug (const rtx_def *)): New.
21132 * print-tree.c
21133 (debug_tree (tree): Move within file.
21134 (debug_raw (const tree_node &)): New.
21135 (debug_raw (const tree_node *)): New.
21136 (dump_tree_via_hooks (const tree_node *, int)): New.
21137 (debug (const tree_node &)): New.
21138 (debug (const tree_node *)): New.
21139 (debug_verbose (const tree_node &)): New.
21140 (debug_verbose (const tree_node *)): New.
21141 (debug_head (const tree_node &)): New.
21142 (debug_head (const tree_node *)): New.
21143 (debug_body (const tree_node &)): New.
21144 (debug_body (const tree_node *)): New.
21145 (debug_vec_tree (tree): Move and reimplement in terms of dump.
21146 (debug (vec<tree, va_gc> &)): New.
21147 (debug (vec<tree, va_gc> *)): New.
21148 * rtl.h
21149 (extern debug (const rtx_def &)): New.
21150 (extern debug (const rtx_def *)): New.
21151 * sbitmap.c
21152 (debug_raw (simple_bitmap_def &)): New.
21153 (debug_raw (simple_bitmap_def *)): New.
21154 (debug (simple_bitmap_def &)): New.
21155 (debug (simple_bitmap_def *)): New.
21156 * sbitmap.h
21157 (extern debug (simple_bitmap_def &)): New.
21158 (extern debug (simple_bitmap_def *)): New.
21159 (extern debug_raw (simple_bitmap_def &)): New.
21160 (extern debug_raw (simple_bitmap_def *)): New.
21161 * sel-sched-dump.c
21162 (debug (vinsn_def &)): New.
21163 (debug (vinsn_def *)): New.
21164 (debug_verbose (vinsn_def &)): New.
21165 (debug_verbose (vinsn_def *)): New.
21166 (debug (expr_def &)): New.
21167 (debug (expr_def *)): New.
21168 (debug_verbose (expr_def &)): New.
21169 (debug_verbose (expr_def *)): New.
21170 (debug (vec<rtx> &)): New.
21171 (debug (vec<rtx> *)): New.
21172 * sel-sched-dump.h
21173 (extern debug (vinsn_def &)): New.
21174 (extern debug (vinsn_def *)): New.
21175 (extern debug_verbose (vinsn_def &)): New.
21176 (extern debug_verbose (vinsn_def *)): New.
21177 (extern debug (expr_def &)): New.
21178 (extern debug (expr_def *)): New.
21179 (extern debug_verbose (expr_def &)): New.
21180 (extern debug_verbose (expr_def *)): New.
21181 (extern debug (vec<rtx> &)): New.
21182 (extern debug (vec<rtx> *)): New.
21183 * sel-sched-ir.h
21184 (_list_iter_cond_expr): Make inline instead of static.
21185 * sreal.c
21186 (debug (sreal &)): New.
21187 (debug (sreal *)): New.
21188 * sreal.h
21189 (extern debug (sreal &)): New.
21190 (extern debug (sreal *)): New.
21191 * tree.h
21192 (extern debug_raw (const tree_node &)): New.
21193 (extern debug_raw (const tree_node *)): New.
21194 (extern debug (const tree_node &)): New.
21195 (extern debug (const tree_node *)): New.
21196 (extern debug_verbose (const tree_node &)): New.
21197 (extern debug_verbose (const tree_node *)): New.
21198 (extern debug_head (const tree_node &)): New.
21199 (extern debug_head (const tree_node *)): New.
21200 (extern debug_body (const tree_node &)): New.
21201 (extern debug_body (const tree_node *)): New.
21202 (extern debug (vec<tree, va_gc> &)): New.
21203 (extern debug (vec<tree, va_gc> *)): New.
21204 * tree-cfg.c
21205 (debug (struct loop &)): New.
21206 (debug (struct loop *)): New.
21207 (debug_verbose (struct loop &)): New.
21208 (debug_verbose (struct loop *)): New.
21209 * tree-dump.c: Add header dependence.
21210 * tree-flow.h
21211 (extern debug (struct loop &)): New.
21212 (extern debug (struct loop *)): New.
21213 (extern debug_verbose (struct loop &)): New.
21214 (extern debug_verbose (struct loop *)): New.
21215 * tree-data-ref.c
21216 (debug (data_reference &)): New.
21217 (debug (data_reference *)): New.
21218 (debug (vec<data_reference_p> &)): New.
21219 (debug (vec<data_reference_p> *)): New.
21220 (debug (vec<ddr_p> &)): New.
21221 (debug (vec<ddr_p> *)): New.
21222 * tree-data-ref.h
21223 (extern debug (data_reference &)): New.
21224 (extern debug (data_reference *)): New.
21225 (extern debug (vec<data_reference_p> &)): New.
21226 (extern debug (vec<data_reference_p> *)): New.
21227 (extern debug (vec<ddr_p> &)): New.
21228 (extern debug (vec<ddr_p> *)): New.
21229 * tree-ssa-alias.c
21230 (debug (pt_solution &)): New.
21231 (debug (pt_solution *)): New.
21232 * tree-ssa-alias.h
21233 (extern debug (pt_solution &)): New.
21234 (extern debug (pt_solution *)): New.
21235 * tree-ssa-alias.c
21236 (debug (_var_map &)): New.
21237 (debug (_var_map *)): New.
21238 (debug (tree_live_info_d &)): New.
21239 (debug (tree_live_info_d *)): New.
21240 * tree-ssa-alias.h
21241 (extern debug (_var_map &)): New.
21242 (extern debug (_var_map *)): New.
21243 (extern debug (tree_live_info_d &)): New.
21244 (extern debug (tree_live_info_d *)): New.
21245
be77e1e5
JH		 212462013-03-28 Jan Hubicka <jh@suse.cz>
21247
21248 * lto-cgraph.c (merge_profile_summaries): Fix overflows.
21249
777e6976
IB		 212502013-03-28 Ian Bolton <ian.bolton@arm.com>
21251
21252 * config/aarch64/aarch64.md (aarch64_can_eliminate): Keep frame
21253 record only when desired or required.
21254
9cd347ae
UB		 212552013-03-28 Uros Bizjak <ubizjak@gmail.com>
21256
21257 * config/i386/i386.md (*vec_extract2vdi_1): Merge with
21258 *vec_extractv2di_1_rex64. Use x64 isa attribute.
21259
33159866 212602013-03-28 Naveen H.S <Naveen.Hurugalawadi@caviumnetworks.com>
2a1d1991
N		 21261
21262 * config/aarch64/aarch64.md (*and<mode>3_compare0): New pattern.
21263 (*andsi3_compare0_uxtw): New pattern.
21264 (*and_<SHIFT:optab><mode>3_compare0): New pattern.
21265 (*and_<SHIFT:optab>si3_compare0_uxtw): New pattern.
21266
89ab31c1
JH		 212672013-03-28 Jan Hubicka <jh@suse.cz>
21268
21269 * data-streamer-in.c (streamer_read_gcov_count): New function.
21270 * gimple-streamer-out.c: Include value-prof.h.
21271 (output_gimple_stmt): Output histogram.
21272 (output_bb): Use streamer_write_gcov_count.
21273 * value-prof.c: Include data-streamer.h
21274 (dump_histogram_value): Add HIST_TYPE_MAX.
21275 (stream_out_histogram_value): New function.
21276 (stream_in_histogram_value): New function.
21277 * value-prof.h (enum hist_type): Add HIST_TYPE_MAX.
21278 (stream_out_histogram_value, stream_in_histogram_value): Declare.
21279 * data-streamer-out.c (streamer_write_gcov_count): New function.
21280 (streamer_write_gcov_count_stream): New function.
21281 * lto-cgraph.c (lto_output_edge): Update counter streaming.
21282 (lto_output_node): Likewise.
21283 (input_node, input_edge): Likewise.
21284 * lto-streamer-out.c (output_cfg): Update streaming.
21285 * lto-streamer-in.c (input_cfg): Likewise.
21286 * data-streamer.h (streamer_write_gcov_count,
21287 streamer_write_gcov_count_stream, streamer_read_gcov_count): Declare.
21288 * gimple-streamer-in.c: Include value-prof.h
21289 (input_gimple_stmt): Input histograms.
21290 (input_bb): Update profile streaming.
21291
e594716a 212922013-03-28 Kenneth Zadeck <zadeck@naturalbridge.com>
b4c73eed 21293
33159866 21294 * genmodes.c (emit_max_int): New function.
8697be17 21295 (emit_insn_modes_h): Added call to emit_max_function.
39385fa6
PC		 21296 * doc/rtl.texi (MAX_BITSIZE_MODE_ANY_INT, MAX_BITSIZE_MODE_ANY_MODE):
21297 Added doc.
8697be17 21298 * machmode.def: Fixed comment.
89ab31c1 21299
e594716a 213002013-03-28 Kenneth Zadeck <zadeck@naturalbridge.com>
8697be17
KZ		 21301
21302 * combine.c (try_combine): Removed useless assert.
b4c73eed
KZ		 21303 * cselib.c (rtx_equal_for_cselib_1): Removed unnecessary parens.
21304
74946978
MP		 213052013-03-28 Marek Polacek <polacek@redhat.com>
21306 Richard Biener <rguenther@suse.de>
21307
21308 PR tree-optimization/56695
21309 * tree-vect-stmts.c (vectorizable_condition): Unconditionally
21310 build signed result of a vector comparison.
21311 * tree-cfg.c (verify_gimple_comparison): Check that a result
21312 of a vector comparison has signed type.
21313
a64b9c26
RB		 213142013-03-28 Richard Biener <rguenther@suse.de>
21315
21316 PR tree-optimization/37021
21317 * tree-vect-slp.c (vect_build_slp_tree): When not unrolling
21318 do not restrict gaps between groups.
21319 * tree-vect-stmts.c (vectorizable_load): Properly account for
21320 a gap between groups.
21321
a9dc2a2f
EB		 213222013-03-28 Eric Botcazou <ebotcazou@adacore.com>
21323
21324 * toplev.c (process_options): Do not disable -fomit-frame-pointer on a
21325 general basis if unwind info is requested and ACCUMULATE_OUTGOING_ARGS
21326 is not enabled.
21327
53cb97f9
GP		 213282013-03-27 Gerald Pfeifer <gerald@pfeifer.com>
21329
21330 * doc/invoke.texi (AVR Options): Tweak link for AVR-LibC user manual.
21331 * doc/extend.texi (Named Address Spaces): Ditto.
21332 (Variable Attributes): Ditto.
21333
b802ae5c
KT		 213342013-03-27 Kai Tietz <ktietz@redhat.com>
21335
eddae10a
KT		 21336 * config.build: Add support for cygwin x64 target.
21337 * config.gcc: Likewise.
21338 * config.host: Likewise.
21339 * configure.ac: Likewise
21340 * configure: Regenerated.
b802ae5c 21341
371e77e3 213422013-03-27 Kai Tietz <ktietz@redhat.com>
8269de54
KT		 21343
21344 * config/i386/cygwin-stdint.h: Add support for cygwin x64 target.
21345 * config/i386/t-cygwin-w64: New file.
21346 * config/i386/cygwin-w64.h: New file.
21347 * config/i386/cygwin.h (EXTRA_OS_CPP_BUILTINS): Extend
21348 and add support for x64-cygwin target.
21349 (CPP_SPEC): Likewise.
21350 (CXX_WRAP_SPEC_LIST): Undefine before define.
21351 (LIBGCJ_SONAME): Use 15 as version.
21352
f49b33cb
RB		 213532013-03-27 Richard Biener <rguenther@suse.de>
21354
21355 PR tree-optimization/56716
21356 * tree-ssa-structalias.c (perform_var_substitution): Adjust
21357 dumping for ref nodes.
21358
b37a6ce5
MJ		 213592013-03-27 Martin Jambor <mjambor@suse.cz>
21360
21361 PR tree-optimization/55334
21362 * ipa-cp.c (initialize_node_lattices): Allow IPA-CP through and to
21363 restricted pointers to arrays.
21364
9469b9b2
GDR		 213652013-03-27 Gabriel Dos Reis <gdr@integrable-solutions.net>
21366
21367 * Makefile.in (.SUFFIXES): Add .cc.
21368 (.c.o): Apply same recipe for implicit rule .cc.o.
21369
7d24f650
RB		 213702013-03-27 Richard Biener <rguenther@suse.de>
21371
21372 PR tree-optimization/37021
21373 * tree-vect-data-refs.c (vect_check_strided_load): Allow
21374 REALPART/IMAGPART_EXPRs around the supported refs.
21375 * tree-ssa-structalias.c (find_func_aliases): Assume that
21376 floating-point values are not used to transfer pointers.
21377
2f251a05
AI		 213782013-03-27 Alexander Ivchenko <alexander.ivchenko@intel.com>
21379
f43245d1
UB		 21380 * target.def (TARGET_HAS_IFUNC_P): New target hook.
21381 * doc/tm.texi.in (TARGET_HAS_IFUNC_P): New.
21382 * doc/tm.texi: Regenerate.
21383 * targhooks.h (default_has_ifunc_p): New.
21384 * targhooks.c (default_has_ifunc_p): Ditto.
21385 * config/linux-protos.h: New file.
21386 * config/linux-android.h (TARGET_HAS_IFUNC_P): Using version of this
21387 hook for linux which disables support of indirect functions in android.
21388 * config/linux-android.c: New file.
21389 * config/t-linux-android.c: Ditto.
21390 * config.gcc: Added new object file linux-android.o.
21391 * config/i386/i386.c (ix86_get_function_versions_dispatcher):
2f251a05 21392 Using TARGET_HAS_IFUNC hook instead of HAVE_GNU_INDIRECT_FUNCTION.
f43245d1 21393 * varasm.c (do_assemble_alias): Likewise.
2f251a05
AI		 21394 * configure.ac: Define HAVE_GNU_INDIRECT_FUNCTION as zero if the target
21395 doesn't support indirect functions.
21396 * configure: Regenerate.
21397
78b4e425
BC		 213982013-03-27 Bin Cheng <bin.cheng@arm.com>
21399
21400 PR target/56102
21401 * config/arm/arm.c (thumb1_rtx_costs, thumb1_size_rtx_costs): Fix
21402 rtx costs for SET/ASHIFT/ASHIFTRT/LSHIFTRT/ROTATERT patterns with
21403 mult-word mode.
21404
a5ba7b92
AK		 214052013-03-27 Andreas Krebbel <Andreas.Krebbel@de.ibm.com>
21406
21407 * config/s390/s390.h (TARGET_FLT_EVAL_METHOD): Define.
21408
be624986
TG		 214092013-03-27 Terry Guo <terry.guo@arm.com>
21410
21411 * config/arm/arm-cores.def: Added core cortex-r7.
21412 * config/arm/arm-tune.md: Regenerated.
21413 * config/arm/arm-tables.opt: Regenerated.
21414 * doc/invoke.texi: Added entry for core cortex-r7.
21415
0a514f47
WL		 214162013-03-27 Walter Lee <walt@tilera.com>
21417
21418 * config/tilegx/tilegx.c (tilegx_expand_prologue): Avoid
21419 double-decrement of next_scratch_regno.
21420
9332b0d2
WL		 214212013-03-27 Walter Lee <walt@tilera.com>
21422
801d9b2a 21423 * config/tilegx/tilegx.md (insn_v1mulu): Fix predicates on
9332b0d2
WL		 21424 input operands.
21425 (insn_v1mulus): Ditto.
21426 (insn_v2muls): Ditto.
21427
f54ea5dd
WL		 214282013-03-27 Walter Lee <walt@tilera.com>
21429
e594716a 21430 * config/tilegx/tilegx.h (ASM_OUTPUT_ADDR_VEC_ELT): Delete extra tab.
f54ea5dd
WL		 21431 (ASM_OUTPUT_ADDR_DIFF_ELT): Ditto.
21432
bf60f4ca
WL		 214332013-03-27 Walter Lee <walt@tilera.com>
21434
e594716a 21435 * config/tilegx/tilegx.md (*sibcall_insn): Fix type atribute for jr.
bf60f4ca
WL		 21436 (*sibcall_value): Ditto.
21437
e3b51eeb
WL		 214382013-03-27 Walter Lee <walt@tilera.com>
21439
21440 * config/tilegx/tilegx.md (insn_mnz_<mode>): Replaced by ...
21441 (insn_mnz_v8qi): ... this ...
21442 (insn_mnz_v4hi): ... and this. Replace (const_int 0) with the
21443 vector equivalent.
21444 (insn_v<n>mnz): Replaced by ...
21445 (insn_v1mnz): ... this ...
21446 (insn_v2mnz): ... and this. Replace (const_int 0) with the vector
21447 equivalent.
21448 (insn_mz_<mode>): Replaced by ...
21449 (insn_mz_v8qi): ... this ...
21450 (insn_mz_v4hi): ... and this. Replace (const_int 0) with the
21451 vector equivalent.
21452 (insn_v<n>mz): Replaced by ...
21453 (insn_v1mz): ... this ...
21454 (insn_v2mz): ... and this. Replace (const_int 0) with the vector
21455 equivalent.
21456
4fc7b145
EB		 214572013-03-26 Eric Botcazou <ebotcazou@adacore.com>
21458
21459 * doc/invoke.texi (SPARC options): Remove -mlittle-endian.
21460
6f33817e
RM		 214612013-03-26 Roland McGrath <mcgrathr@google.com>
21462
21463 * config/arm/arm.c (arm_print_operand: case 'w'): Use fputs rather
21464 than fprintf with a non-constant, non-format string.
21465
93a4145b
UB		 214662013-03-26 Uros Bizjak <ubizjak@gmail.com>
21467
21468 * config/i386/i386.md (*cmpqi_ext_1): Merge with *cmpqi_ext_1_rex64
21469 using nox64 isa attribute. Use nonimmediate_x86nomem_operand as
21470 operand 0 predicate.
21471 (*cmpqi_ext_3): Merge with *cmpqi_ext_3_rex64 using nox64 isa
21472 attribute. Use general_x64nomem_operand as operand 1 predicate.
21473 (*movqi_extv_1): Merge with *movqi_extv_1_rex64 using nox64 isa
21474 attribute. Use nonimmediate_x64nomem_operand as operand 0 predicate.
21475 (*movqi_extzv_2): Merge with *movqi_extzv_2_rex64 using nox64 isa
21476 attribute. Use nonimmediate_x64nomem_operand as operand 0 predicate.
21477 (mov<mode>_insv_1): Remove expander. Merge insn with
21478 movsi_insv_1 using SWI48 mode iterator and nox64 isa attribute.
21479 Use general_x64nomem_operand as operand 1 predicate.
21480 (addqi_ext_1): Merge with *addqi_ext_1_rex64 using nox64 isa attribute.
21481 (*testqi_ext_1): Merge with *testqi_ext_1_rex64 using nox64 isa
21482 attribute. Use nonimmediate_x64nomem_operand as operand 1 predicate.
21483 (*andqi_ext_1): Merge with *andqi_ext_1_rex64 using nox64 isa
21484 attribute. Use nonimmediate_x64nomem_operand as operand 2 predicate.
21485 (*<code>qi_ext_1): Merge with *<code>qi_ext_1_rex64 using nox64 isa
21486 attribute. Use nonimmediate_x64nomem_operand as operand 1 predicate.
21487 (*xorqi_cc_ext_1): Merge with *xorqi_cc_ext_1_rex64 using nox64
21488 isa attribute. Use general_x64nomem_operand as operand 2 predicate.
21489 * config/i386/predicates.md (nonimmediate_x64nomem_operand): New.
21490 (general_x64nomem_operand): Ditto.
21491
c6a9ed5a 214922013-03-26 Sebastian Huber <sebastian.huber@embedded-brains.de>
df599020
SH		 21493
21494 * config/rtems.opt: Add -pthread option.
21495
31b0a960
RB		 214962013-03-26 Richard Biener <rguenther@suse.de>
21497
93a4145b
UB		 21498 * alias.c (find_base_term): Avoid redundant and not used recursion.
21499 (base_alias_check): Get the initial base term from the caller.
31b0a960
RB		 21500 (true_dependence_1): Compute and pass base terms to base_alias_check.
21501 (write_dependence_p): Likewise.
21502 (may_alias_p): Likewise.
21503
79517551
SN		 215042013-03-26 Sofiane Naci <sofiane.naci@arm.com>
21505
21506 * config/aarch64/aarch64.c (aarch64_classify_address): Support
21507 PC-relative load in SI modes and above only.
21508
a76213b9
XQ		 215092013-03-26 Xinyu Qi <xyqi@marvell.com>
21510
21511 * config/arm/arm.h (FIRST_IWMMXT_GR_REGNUM): Add comment.
33159866
UB		 21512 * config/arm/iwmmxt.md (WCGR0): Update.
21513 (WCGR1, WCGR2, WCGR3): Likewise.
a76213b9 21514
37ff9355
UB		 215152013-03-26 Uros Bizjak <ubizjak@gmail.com>
21516
21517 * config/i386/i386.md (*movdfcc_1): Merge with *movdfcc_1_rex64.
21518 Use x64 and nox64 isa attributes.
21519
30b0317c
RB		 215202013-03-26 Richard Biener <rguenther@suse.de>
21521
21522 * emit-rtl.c (set_mem_attributes_minus_bitpos): Remove
21523 alignment computations and rely on get_object_alignment_1
21524 for the !TYPE_P case.
21525 Commonize DECL/COMPONENT_REF handling in the ARRAY_REF path.
21526
cbcdb140
WL		 215272013-03-26 Walter Lee <walt@tilera.com>
21528
21529 * config/tilegx/tilegx.h (PROFILE_BEFORE_PROLOGUE): Define.
21530 * config/tilegx/tilepro.h (PROFILE_BEFORE_PROLOGUE): Define.
21531
dd884525 215322013-03-25 Jeff Law <law@redhat.com>
be672e08 21533
33159866
UB		 21534 * tree-ssa-dom.c (record_equivalences_from_incoming_edge): Add missing
21535 check for INTEGRAL_TYPE_P that was missing due to checking in
dd884525
JL		 21536 wrong version of prior patch.
21537
0051d3ec
WL		 215382013-03-25 Walter Lee <walt@tilera.com>
21539
21540 * config/tilegx/tilegx-builtins.h (enum tilegx_builtin): Add
21541 TILEGX_INSN_SHUFFLEBYTES1.
21542 * config/tilegx/tilegx.c (tilegx_builtin_info): Add entry for
21543 shufflebytes1.
21544 (tilegx_builtins): Ditto.
21545 * config/tilegx/tilegx.md (insn_shufflebytes1): New pattern.
21546
065a3605
WL		 215472013-03-25 Walter Lee <walt@tilera.com>
21548
21549 * config/tilegx/tilegx.md (floatsisf2): New pattern.
0051d3ec
WL		 21550 (floatunssisf2): New pattern.
21551 (floatsidf2): New pattern.
21552 (floatunssidf2): New pattern.
065a3605 21553
5b2a3c82
WL		 215542013-03-25 Walter Lee <walt@tilera.com>
21555
21556 * config/tilegx/tilegx.c (expand_set_cint64_one_inst): Inline
21557 tests for constraint J, K, N, P.
21558
192ea533
WL		 215592013-03-25 Walter Lee <walt@tilera.com>
21560
21561 * config/tilegx/tilegx.c (tilegx_asm_preferred_eh_data_format):
21562 Use indirect/pcrel encoding.
21563 * config/tilepro/tilepro.c (tilepro_asm_preferred_eh_data_format):
21564 Ditto.
21565
855e0d0b
SE		 215662013-03-25 Steve Ellcey <sellcey@mips.com>
21567
21568 * config/mips/mmips-cpus.def (74kc, 74kf2_1, 74kf, 74kf, 74kf1_1,
21569 74kfx, 74kx, 74kf3_2): Add PTF_AVOID_IMADD.
21570 * config/mips/mips.c (mips_option_override): Set IMADD default.
21571 * config/mips/mips.h (PTF_AVOID_IMADD): New.
21572 (ISA_HAS_MADD_MSUB): Remove MIPS16 check.
21573 (GENERATE_MADD_MSUB): Remove TUNE_74K check, add MIPS16 check.
21574 * config/mips/mips.md (mimadd): New flag for integer madd/msub.
21575 * doc/invoke.texi (-mimadd/-mno-imadd): New.
21576
39e45653
JL		 215772013-03-25 Jeff Law <law@redhat.com>
21578
21579 * tree-ssa-dom.c (record_equivalences_from_incoming_edge): Rework
21580 slightly to avoid creating and folding useless trees. Simplify
21581 slightly by restricting to INTEGER_CSTs and using int_fits_type_p.
21582
7482c470
UB		 215832013-03-25 Uros Bizjak <ubizjak@gmail.com>
21584
21585 * config/i386/i386.md (*zero_extendsidi2): Merge with
21586 *zero_extendsidi2_rex64. Use x64 and nox64 isa attributes.
21587 * config/i386/predicates.md (x86_64_zext_operand): Rename from
21588 x86_64_zext_general_operand. Use nonimmediate_operand on 32bit
21589 targets. Clarify comment.
21590
4a53743e
MJ		 215912013-03-25 Martin Jambor <mjambor@suse.cz>
21592
21593 * ipa-prop.c (ipa_write_jump_function): Stream simple and aritmetic
21594 pass-through jump functions differently.
21595 (ipa_read_jump_function): Likewise. Also use setter functions to set
21596 up jump functions.
21597
162712de
MJ		 215982013-03-25 Martin Jambor <mjambor@suse.cz>
21599
21600 * ipa-cp.c (ipa_get_indirect_edge_target): Renamed to
21601 ipa_get_indirect_edge_target_1, added parameter agg_reps and ability to
21602 process it.
21603 (ipa_get_indirect_edge_target): New function.
21604 (devirtualization_time_bonus): New parameter known_aggs, pass it to
21605 ipa_get_indirect_edge_target. Update all callers.
21606 (ipcp_discover_new_direct_edges): New parameter aggvals. Pass it to
21607 ipa_get_indirect_edge_target_1 instead of calling
21608 ipa_get_indirect_edge_target.
21609 (create_specialized_node): Pass aggvlas to
21610 ipcp_discover_new_direct_edges.
21611
a5a4c20a
KT		 216122013-03-25 Kyrylo Tkachov <kyrylo.tkachov@arm.com>
21613
21614 * config/arm/arm.md (f_sels, f_seld): New types.
21615 (*cmov<mode>): New pattern.
21616 * config/arm/predicates.md (arm_vsel_comparison_operator): New
21617 predicate.
21618
f992b9fc
KT		 216192013-03-25 Kai Tietz <ktietz@redhat.com>
21620
21621 * config/i386/xm-mingw32.h (__USE_MINGW_ANSI_STDIO): Enable
21622 POSIX-printf for mingw-hosted builds.
21623
c16fd676
RB		 216242013-03-25 Richard Biener <rguenther@suse.de>
21625
21626 PR middle-end/56694
21627 * tree-eh.c (lower_eh_must_not_throw): Strip BLOCKs from the
21628 must-not-throw stmt location.
21629
5ad29f12
KT		 216302013-03-25 Kyrylo Tkachov <kyrylo.tkachov@arm.com>
21631
21632 * config/arm/arm.c (arm_emit_load_exclusive): Add acq parameter.
21633 Emit load-acquire versions when acq is true.
21634 (arm_emit_store_exclusive): Add rel parameter.
21635 Emit store-release versions when rel is true.
21636 (arm_split_compare_and_swap): Use acquire-release instructions
21637 instead.
21638 of barriers when appropriate.
21639 (arm_split_atomic_op): Likewise.
21640 * config/arm/arm.h (TARGET_HAVE_LDACQ): New macro.
21641 * config/arm/unspecs.md (VUNSPEC_LAX): New unspec.
21642 (VUNSPEC_SLX): Likewise.
21643 (VUNSPEC_LDA): Likewise.
21644 (VUNSPEC_STL): Likewise.
21645 * config/arm/sync.md (atomic_load<mode>): New pattern.
21646 (atomic_store<mode>): Likewise.
21647 (arm_load_acquire_exclusive<mode>): Likewise.
21648 (arm_load_acquire_exclusivesi): Likewise.
21649 (arm_load_acquire_exclusivedi): Likewise.
21650 (arm_store_release_exclusive<mode>): Likewise.
21651
03a7dddb
CM		 216522013-03-25 Catherine Moore <clm@codesourcery.com>
21653
21654 * config/mips/constraints.md (u, Udb7 Uead, Uean, Uesp, Uib3,
21655 Uuw6, Usb4, ZS, ZT, ZU, ZV, ZW): New constraints.
21656 * config/mip/predicates.md (lwsp_swsp_operand,
21657 lw16_sw16_operand, lhu16_sh16_operand, lbu16_operand,
21658 sb16_operand, db4_operand, db7_operand, ib3_operand,
21659 sb4_operand, ub4_operand, uh4_operand, uw4_operand,
21660 uw5_operand, uw6_operand, addiur2_operand, addiusp_operand,
21661 andi16_operand): New predicates.
21662 * config/mips/mips.md (compression): New attribute.
21663 (enabled): New attribute.
21664 (length): Consider compression in computing length.
21665 (shift_compression): New code attribute.
21666 (*add<mode>3): New operands. Record compression.
21667 (sub<mode>3): Likewise.
21668 (one_cmpl<mode>2): Likewise.
21669 (*and<mode>3): Likewise.
21670 (*ior<mode>3): Likewise.
21671 (unnamed pattern for xor): Likewise.
21672 (*zero_extend<SHORT:mode><GPR:mode>2): Likewise.
21673 (*<optab><mode>3): Likewise.
21674 (*mov<mode>_internal: Likewise.
21675 * config/mips/mips-protos.h (mips_signed_immediate_p): New.
21676 (mips_unsigned_immediate_p): New.
21677 (umips_lwsp_swsp_address_p): New.
21678 (m16_based_address_p): New.
21679 * config/mips/mips-protos.h (mips_signed_immediate_p): New prototype.
21680 (mips_unsigned_immediate_p): New prototype.
21681 (lwsp_swsp_address_p): New prototype.
21682 (m16_based_address_p): New prototype.
21683 * config/mips/mips.c (mips_unsigned_immediate_p): New function.
21684 (mips_signed_immediate_p): New function.
21685 (m16_based_address_p): New function.
21686 (lwsp_swsp_address_p): New function.
21687 (mips_print_operand_punctuation): Recognize short delay slot insns
21688 for microMIPS.add<mode>3"
21689
f35c297f 216902013-03-25 Kyrylo Tkachov <kyrylo.tkachov@arm.com>
a5784ce6 21691
f35c297f
KT		 21692 PR target/56720
21693 * config/arm/iterators.md (v_cmp_result): New mode attribute.
21694 * config/arm/neon.md (vcond<mode><mode>): Handle unordered cases.
21695
051b9446
RB		 216962013-03-25 Richard Biener <rguenther@suse.de>
21697
21698 PR tree-optimization/56689
21699 * tree-vrp.c (execute_vrp): Mark loops for fixup if we removed
21700 any edge.
21701
374001cb
RB		 217022013-03-25 Richard Biener <rguenther@suse.de>
21703
21704 * tree-ssa-loop-im.c (struct mem_ref): Use bitmap_head instead
21705 of bitmap.
21706 (memory_references): Likewise.
21707 (outermost_indep_loop, mem_ref_alloc, mark_ref_stored,
21708 gather_mem_refs_stmt, record_dep_loop, ref_indep_loop_p_1,
21709 ref_indep_loop_p_2, find_refs_for_sm): Adjust.
21710 (gather_mem_refs_in_loops): Fold into ...
21711 (analyze_memory_references): ... this. Move initialization
21712 to tree_ssa_lim_initialize.
21713 (fill_always_executed_in): Rename to ...
21714 (fill_always_executed_in_1): ... this.
21715 (fill_always_executed_in): Move contains_call computation to
21716 this new function from ...
21717 (tree_ssa_lim_initialize): ... here.
21718 (tree_ssa_lim): Call fill_always_executed_in.
21719
57895947
EB		 217202013-03-25 Eric Botcazou <ebotcazou@adacore.com>
21721
21722 * postreload.c (reload_combine): Fix code detecting returns.
21723
1e1b18c1
EB		 217242013-03-25 Eric Botcazou <ebotcazou@adacore.com>
21725
21726 * function.c (emit_use_return_register_into_block): On cc0 targets,
21727 do not emit the sequence between cc0 setter and user.
21728
9216baf1
KT		 217292013-03-25 Kai Tietz <ktietz@redhat.com>
21730
1e1b18c1
EB		 21731 * config/i386/predicates.md (local_symbolic_operand): Interpret
21732 dll-imported symbols as none-local.
9216baf1 21733
76421b44
RB		 217342013-03-25 Richard Biener <rguenther@suse.de>
21735
21736 * tree-ssa-loop-im.c (struct depend): Remove.
21737 (struct lim_aux_data): Make depends a vec of gimples.
21738 (free_lim_aux_data): Adjust.
21739 (add_dependency): Likewise.
21740 (set_level): Likewise.
21741
d154bfa2
RB		 217422013-03-25 Richard Biener <rguenther@suse.de>
21743
21744 PR middle-end/56434
21745 * calls.c (expand_call): Use MALLOC_ABI_ALIGNMENT to annotate
21746 the pointer returned by calls with ECF_MALLOC set.
21747
2cd9804e
UB		 217482013-03-24 Uros Bizjak <ubizjak@gmail.com>
21749
acef5fe0
UB		 21750 * config/i386/mmx.md (mov<mode>): Add ?!Ym,r and r,?!Ym alternatives.
21751
217522013-03-24 Uros Bizjak <ubizjak@gmail.com>
21753
21754 * config/i386/mmx.md (mov<mode>): Merge with movv2sf expander
2cd9804e
UB		 21755 using MMXMODE mode iterator.
21756 (*move<mode>_internal): Merge with *movv2sf_internal and
21757 *movv2sf_internal_rex64 using MMXMODE mode iterator.
21758
225ccc68
SB		 217592013-03-23 Steven Bosscher <steven@gcc.gnu.org>
21760
85c0f02d
SB		 21761 * gcse.c (oprs_unchanged_p): Respect flag_gcse_lm.
21762 (record_last_mem_set_info): Likewise.
21763
225ccc68
SB		 21764 * df-core.c (rest_of_handle_df_initialize): Use XCNEWVEC instead
21765 of XNEWVEC followed by memset.
21766 (df_worklist_dataflow): Use XNEWVEC instead of xmalloc with a cast.
21767
b64925dc
SB		 217682013-03-23 Steven Bosscher <steven@gcc.gnu.org>
21769
21770 * config/avr/avr.c, config/bfin/bfin.c, config/c6x/c6x.c,
21771 config/epiphany/epiphany.c, config/frv/frv.c, config/ia64/ia64.c,
21772 config/iq2000/iq2000.c, config/mcore/mcore.c, config/mep/mep.c,
21773 config/mmix/mmix.c, config/pa/pa.c, config/rs6000/rs6000.c,
21774 config/s390/s390.c, config/sparc/sparc.c, config/spu/spu.c,
21775 config/stormy16/stormy16.c, config/v850/v850.c, config/xtensa/xtensa.c,
21776 dwarf2out.c, hw-doloop.c, resource.c, rtl.h : Where applicable, use
21777 the predicates NOTE_P, NONJUMP_INSN_P, JUMP_P, CALL_P, LABEL_P, and
21778 BARRIER_P instead of GET_CODE.
21779
b9aaf52e
EB		 217802013-03-23 Eric Botcazou <ebotcazou@adacore.com>
21781
21782 * config/sparc/sparc.c (sparc_emit_probe_stack_range): Fix small
21783 inaccuracy in the probing code.
21784
21785 * config/sparc/sparc.md (ctrapsi4): Add predicate for operand #3.
21786 (ctrapdi4): Likewise.
21787
66de4d7c
EB		 217882013-03-23 Eric Botcazou <ebotcazou@adacore.com>
21789
21790 * calls.c (expand_call): Add missing guard to code handling return
21791 of non-BLKmode structures in MSB.
21792 * function.c (expand_function_end): Likewise.
21793
4f25c8fb
EB		 217942013-03-23 Eric Botcazou <ebotcazou@adacore.com>
21795
21796 * combine.c (try_combine): Adjust comment. Do not add the set of
21797 insn #0 if the destination indirectly is set or dies in insn #2.
21798 Tidy up code to distribute a new note.
21799
23b7850d
UB		 218002013-03-22 Uros Bizjak <ubizjak@gmail.com>
21801
21802 * config/i386/i386.md (*movdi_internal): Set prefix_rex attribute
21803 also for alternatives 16 and 17.
21804
bed852cf
UB		 218052013-03-22 Uros Bizjak <ubizjak@gmail.com>
21806
21807 * config/i386/sse.md (*mov<mode>_internal): Merge with
21808 *mov<mode>_internal_rex64. Use x64 and nox64 isa attributes.
21809 Emit insn template depending on type attribute. Use
21810 HAVE_AS_IX86_INTERUNIT_MOVQ to handle broken assemblers that require
21811 movd instead of movq mnemonic for interunit moves. Rewrite mode
21812 attribute calculation. Remove unit attribute calculation.
21813 Set prefix attribute to maybe_vex for sselog1 and ssemov types.
21814 Set prefix_data16 attribute for DImode ssemov types.
21815 Use Ym instead of y for SSE-MMX conversion alternatives.
21816 Reorder operand constraints.
21817
1e8a7937
SB		 218182013-03-22 Steven Bosscher <steven@gcc.gnu.org>
21819
21820 * df.h (df_insn_delete): Adjust prototype.
21821 * emit-rtl.c (remove_insn): Pass a basic block to df_insn_delete
21822 and let it decide whether mark the basic block dirty.
21823 (set_insn_deleted): Only pass INSN_P insns to df_insn_delete.
21824 * df-scan.c (df_insn_info_delete): New helper function, split
21825 off from df_insn_delete.
21826 (df_scan_free_bb_info): Use it.
21827 (df_insn_rescan, df_insn_rescan_all, df_process_deferred_rescans):
21828 Likewise.
21829 (df_insn_delete): Likewise. Take insn rtx as argument. Verify
21830 that the insn is actually an insn and it has a non-NULL basic block.
21831 Do not mark basic block dirty if only deleting a DEBUG_INSN.
21832
e14d094c
RB		 218332013-03-22 Richard Biener <rguenther@suse.de>
21834
21835 * tree-ssa-loop-im.c (struct mem_ref): Remove indep_ref and
21836 dep_ref members.
21837 (mem_ref_alloc): Do not allocate them.
21838 (refs_independent_p): Do not query or maintain a cache.
21839
e9cf7316
RB		 218402013-03-22 Richard Biener <rguenther@suse.de>
21841
21842 * tree-ssa-loop-im.c (memory_references): Drop all_refs_in_loop.
21843 (gather_mem_refs_in_loops): Do not compute it.
21844 (analyze_memory_references): Do not allocate it.
21845 (tree_ssa_lim_finalize): Do not free it.
21846 (for_all_locs_in_loop): Do not query all_refs_in_loop.
21847
f046e81b
RB		 218482013-03-22 Richard Biener <rguenther@suse.de>
21849
21850 * is-a.h (as_a): Use gcc_checking_assert.
bed852cf 21851
a05c0ddf
IB		 218522013-03-22 Ian Bolton <ian.bolton@arm.com>
21853
21854 * config/aarch64/aarch64.c (aarch64_print_operand): New
21855 format specifier for printing a constant in hex.
21856 * config/aarch64/aarch64.md (insv_imm<mode>): Use the X
21857 format specifier for printing second operand.
21858
c00217fc
RB		 218592013-03-22 Richard Biener <rguenther@suse.de>
21860
21861 * tree-ssa-loop-im.c (memory_references): Add refs_stored_in_loop
21862 bitmaps.
21863 (gather_mem_refs_in_loops): Perform store accumulation here.
21864 (create_vop_ref_mapping_loop): Remove.
21865 (create_vop_ref_mapping): Likewise.
21866 (analyze_memory_references): Initialize refs_stored_in_loop.
bed852cf 21867 (LOOP_DEP_BIT): New define to map to bits in (in)dep_loop bitmaps.
c00217fc
RB		 21868 (record_indep_loop): Remove.
21869 (record_dep_loop): New function.
21870 (ref_indep_loop_p_1): Adjust to only walk over references
21871 in the loop, not its subloops.
21872 (ref_indep_loop_p): Rename to ...
21873 (ref_indep_loop_p_2): ... this and recurse over the loop tree,
21874 maintaining a more fine-grained cache.
21875 (ref_indep_loop_p): Wrap ref_indep_loop_p_2.
21876 (tree_ssa_lim_finalize): Free refs_stored_in_loop.
21877
15d19bf8
RB		 218782013-03-22 Richard Biener <rguenther@suse.de>
21879
21880 * tree-ssa-loop-im.c (struct mem_ref_locs): Remove.
21881 (struct mem_ref): Make accesses_in_loop a vec of a vec of
21882 aggregate mem_ref_loc.
21883 (free_mem_ref_locs): Inline into ...
21884 (memref_free): ... this and adjust.
21885 (mem_ref_alloc): Adjust.
21886 (mem_ref_locs_alloc): Remove.
21887 (record_mem_ref_loc): Adjust.
21888 (get_all_locs_in_loop): Rewrite into ...
21889 (for_all_locs_in_loop): ... this iterator.
21890 (rewrite_mem_ref_loc): New functor.
21891 (rewrite_mem_refs): Use for_all_locs_in_loop.
21892 (sm_set_flag_if_changed): New functor.
21893 (execute_sm_if_changed_flag_set): Use for_all_locs_in_loop.
21894 (ref_always_accessed): New functor.
21895 (ref_always_accessed_p): Use for_all_locs_in_loop.
21896
6f37411d
MG		 218972013-03-21 Marc Glisse <marc.glisse@inria.fr>
21898
21899 * tree-pass.h (PROP_gimple_lvec): New.
21900 * passes.c (dump_properties): Handle PROP_gimple_lvec.
21901 (init_optimization_passes): Move pass_lower_vector.
21902 * tree-vect-generic.c (gate_expand_vector_operations_ssa): Test
21903 PROP_gimple_lvec.
21904 (pass_lower_vector): Provide PROP_gimple_lvec.
21905 (pass_lower_vector_ssa): Likewise.
21906 * cfgexpand.c (pass_expand): Require PROP_gimple_lvec.
21907
511ef689
MW		 219082013-03-21 Mark Wielaard <mjw@redhat.com>
21909
21910 * dwarf2out.c (size_of_aranges): Skip DECL_IGNORED_P functions.
21911
cc1df30b
UB		 219122013-03-21 Uros Bizjak <ubizjak@gmail.com>
21913
21914 * config/i386/i386.md (*movdi_internal): Disparage slightly
21915 all MMX moves to/from memory. Use Yi instead of x for SSE-MMX
21916 conversion alternatives.
21917
c845cfe1
JJ		 219182013-03-21 Jakub Jelinek <jakub@redhat.com>
21919
37e99116
JJ		 21920 PR middle-end/48087
21921 * diagnostic.def (DK_WERROR): New kind.
21922 * diagnostic.h (werrorcount): Define.
21923 * diagnostic.c (diagnostic_report_diagnostic): For DK_WARNING
21924 promoted to DK_ERROR, increment DK_WERROR counter instead of
21925 DK_ERROR counter.
21926 * toplev.c (toplev_main): Call print_ignored_options even if
21927 just werrorcount is non-zero. Exit with FATAL_EXIT_CODE
21928 even if just werrorcount is non-zero.
21929
c845cfe1
JJ		 21930 PR debug/55608
21931 * dwarf2out.c (tree_add_const_value_attribute): Call ggc_free (array)
21932 on failure.
21933 (resolve_one_addr): Fail if referenced STRING_CST hasn't been written.
21934 (string_cst_pool_decl): New function.
21935 (optimize_one_addr_into_implicit_ptr): New function.
21936 (resolve_addr_in_expr): Optimize DWARF location expression
21937 DW_OP_addr DW_OP_stack_value where DW_OP_addr refers to some variable
21938 which doesn't live in memory, but has DW_AT_location or
21939 DW_AT_const_value, or refers to a string literal, into
21940 DW_OP_GNU_implicit_pointer.
21941 (optimize_location_into_implicit_ptr): New function.
21942 (resolve_addr): If removing DW_AT_location of a variable because
21943 it was DW_OP_addr of address of the variable, but the variable doesn't
21944 live in memory, try to emit const value attribute for the initializer.
21945
08e0cda6
MG		 219462013-03-21 Marc Glisse <marc.glisse@inria.fr>
21947
21948 * tree.h (VECTOR_TYPE_P): New macro.
21949 (VECTOR_INTEGER_TYPE_P, VECTOR_FLOAT_TYPE_P, FLOAT_TYPE_P,
21950 TYPE_MODE): Use it.
21951 * fold-const.c (fold_cond_expr_with_comparison): Use build_zero_cst.
21952 VEC_COND_EXPR cannot be lvalues.
21953 (fold_ternary_loc) <VEC_COND_EXPR>: Merge with the COND_EXPR case.
21954
d08633b4
MG		 219552013-03-21 Marc Glisse <marc.glisse@inria.fr>
21956
21957 * simplify-rtx.c (simplify_binary_operation_1) <VEC_CONCAT>:
21958 Restrict the transformation to equal modes.
21959
e6647190
RB		 219602013-03-21 Richard Biener <rguenther@suse.de>
21961
21962 PR tree-optimization/39326
21963 * tree-ssa-loop-im.c (UNANALYZABLE_MEM_ID): New define.
21964 (MEM_ANALYZABLE): Adjust.
21965 (record_mem_ref_loc): Move bitmap ops ...
21966 (gather_mem_refs_stmt): ... here. Use the shared mem-ref for
21967 unanalyzable refs, do not record locations for it.
21968 (analyze_memory_references): Allocate ref zero as shared
21969 unanalyzable ref.
21970 (refs_independent_p): Do not test for unanalyzed mems here.
21971 (ref_indep_loop_p_1): Special-case disambiguation against
21972 the unanalyzed ref.
cc1df30b 21973 (ref_indep_loop_p): Assert we are not queried for the unanalyzed mem.
e6647190 21974
65074f54
CL		 219752013-03-21 Christophe Lyon <christophe.lyon@linaro.org>
21976
21977 * config/arm/arm-protos.h (tune_params): Add
21978 prefer_neon_for_64bits field.
21979 * config/arm/arm.c (prefer_neon_for_64bits): New variable.
21980 (arm_slowmul_tune): Default prefer_neon_for_64bits to false.
21981 (arm_fastmul_tune, arm_strongarm_tune, arm_xscale_tune): Ditto.
21982 (arm_9e_tune, arm_v6t2_tune, arm_cortex_tune): Ditto.
21983 (arm_cortex_a15_tune, arm_cortex_a5_tune): Ditto.
21984 (arm_cortex_a9_tune, arm_v6m_tune, arm_fa726te_tune): Ditto.
21985 (arm_option_override): Handle -mneon-for-64bits new option.
21986 * config/arm/arm.h (TARGET_PREFER_NEON_64BITS): New macro.
21987 (prefer_neon_for_64bits): Declare new variable.
21988 * config/arm/arm.md (arch): Rename neon_onlya8 and neon_nota8 to
cc1df30b 21989 avoid_neon_for_64bits and neon_for_64bits. Remove onlya8 and nota8.
65074f54
CL		 21990 (arch_enabled): Handle new arch types. Remove support for onlya8
21991 and nota8.
21992 (one_cmpldi2): Use new arch names.
bc5faa5b 21993 (zero_extend<mode>di2, extend<mode>di2): Ditto.
65074f54
CL		 21994 * config/arm/arm.opt (mneon-for-64bits): Add option.
21995 * config/arm/neon.md (adddi3_neon, subdi3_neon, iordi3_neon)
21996 (anddi3_neon, xordi3_neon, ashldi3_neon, <shift>di3_neon): Use
21997 neon_for_64bits instead of nota8 and avoid_neon_for_64bits instead
21998 of onlya8.
21999 * doc/invoke.texi (-mneon-for-64bits): Document.
22000
5a2d2a79
RB		 220012013-03-21 Richard Biener <rguenther@suse.de>
22002
22003 PR tree-optimization/39326
22004 * tree-ssa-loop-im.c (bb_loop_postorder): New global static.
22005 (sort_bbs_in_loop_postorder_cmp): New function.
cc1df30b 22006 (gather_mem_refs_in_loops): Assign mem-ref IDs in loop postorder.
5a2d2a79 22007
5abe1e05
RB		 220082013-03-21 Richard Biener <rguenther@suse.de>
22009
22010 * tree-vect-data-refs.c (vect_update_interleaving_chain): Remove.
22011 (vect_insert_into_interleaving_chain): Likewise.
22012 (vect_drs_dependent_in_basic_block): Inline ...
22013 (vect_slp_analyze_data_ref_dependence): ... here. New function,
22014 split out from ...
22015 (vect_analyze_data_ref_dependence): ... here. Simplify.
22016 (vect_check_interleaving): Simplify.
22017 (vect_analyze_data_ref_dependences): Likewise. Split out ...
22018 (vect_slp_analyze_data_ref_dependences): ... this new function.
22019 (dr_group_sort_cmp): New function.
22020 (vect_analyze_data_ref_accesses): Compute data-reference groups
22021 here instead of in vect_analyze_data_ref_dependence. Use
22022 a more efficient algorithm.
22023 * tree-vect-slp.c (vect_slp_analyze_bb_1): Use
22024 vect_slp_analyze_data_ref_dependences. Call
22025 vect_analyze_data_ref_accesses earlier.
22026 * tree-vect-loop.c (vect_analyze_loop_2): Likewise.
22027 * tree-vectorizer.h (vect_analyze_data_ref_dependences): Adjust.
22028 (vect_slp_analyze_data_ref_dependences): New prototype.
22029
cad1735b
RB		 220302013-03-21 Richard Biener <rguenther@suse.de>
22031
22032 * tree-ssa-loop-im.c (can_sm_ref_p): Do not test whether
22033 ref is stored in the loop.
22034 (find_refs_for_sm): Walk only over all stores.
22035 (store_motion_loop): Allocate from lim_bitmap_obstack.
22036 (store_motion): Likewise.
22037
141310ef
RB		 220382013-03-21 Richard Biener <rguenther@suse.de>
22039
22040 * tree-vect-loop-manip.c (slpeel_tree_peel_loop_to_edge):
22041 Update virtual SSA form.
22042
5022315a
RO		 220432013-03-21 Rainer Orth <ro@CeBiTec.Uni-Bielefeld.DE>
22044
22045 * configure.ac (gcc_cv_ld_eh_frame_ciev3): New test.
22046 * configure: Regenerate.
22047 * config.in: Regenerate.
22048 * config/sol2.c (solaris_override_options): Only enforce DWARF 2
22049 if !HAVE_LD_EH_FRAME_CIEV3.
22050
50d4421c
RB		 220512013-03-21 Richard Biener <rguenther@suse.de>
22052
22053 * tree-cfg.c (verify_expr_no_block): New function.
22054 (verify_expr_location_1): Verify that neither DECL_DEBUG_EXPR
22055 nor DECL_VALUE_EXPR have locations with associated blocks.
22056 * tree-ssa-live.c (clear_unused_block_pointer_1): Remove.
22057 (clear_unused_block_pointer): Remove code dealing with
22058 blocks in DECL_DEBUG_EXPR locations.
22059
839b422f
RB		 220602013-03-21 Richard Biener <rguenther@suse.de>
22061
22062 * tree.h (DECL_DEBUG_EXPR_IS_FROM): Rename to ...
22063 (DECL_HAS_DEBUG_EXPR_P): ... this. Guard properly.
22064 * tree.c (copy_node_stat): Do not copy DECL_HAS_DEBUG_EXPR_P.
22065 * dwarf2out.c (add_var_loc_to_decl): Use DECL_HAS_DEBUG_EXPR_P
22066 instead of DECL_DEBUG_EXPR_IS_FROM.
22067 * gimplify.c (gimplify_modify_expr): Likewise.
22068 * tree-cfg.c (verify_expr_location_1): Likewise.
22069 * tree-complex.c (create_one_component_var): Likewise.
22070 * tree-sra.c (create_access_replacement): Likewise.
22071 * tree-ssa-live.c (clear_unused_block_pointer_1): Likewise.
22072 (clear_unused_block_pointer): Likewise.
22073 * tree-streamer-in.c (unpack_ts_decl_common_value_fields): Likewise.
22074 * tree-streamer-out.c (pack_ts_decl_common_value_fields): Likewise.
22075 * var-tracking.c (var_debug_decl): Likewise.
22076 (track_expr_p): Likewise.
22077 * tree-inline.c (add_local_variables): Likewise. Set
22078 DECL_HAS_DEBUG_EXPR_P after copying it.
22079 * tree-diagnostic.c (default_tree_printer): Use DECL_HAS_DEBUG_EXPR_P
22080 instead of DECL_DEBUG_EXPR_IS_FROM. Guard properly.
22081
fe04878d
UB		 220822013-03-21 Uros Bizjak <ubizjak@gmail.com>
22083
22084 PR bootstrap/56656
22085 * configure.ac (HAVE_AS_IX86_INTERUNIT_MOVQ): New test.
22086 * configure: Regenerate.
22087 * config.in: Regenerate.
22088 * config/i386/i386.md (*movdf_internal): Use
22089 HAVE_AS_IX86_INTERUNIT_MOVQ to handle broken assemblers that require
22090 movd instead of movq mnemonic for interunit moves.
22091 (*movdi_internal): Ditto.
22092
bd059b26 220932013-03-21 Naveen H.S <Naveen.Hurugalawadi@caviumnetworks.com>
fa2e9a58
N		 22094
22095 * config/aarch64/aarch64-simd.md (simd_fabd): New Attribute.
22096 (abd<mode>_3): New pattern.
22097 (aba<mode>_3): New pattern.
22098 (fabd<mode>_3): New pattern.
22099
bd059b26 221002013-03-21 Naveen H.S <Naveen.Hurugalawadi@caviumnetworks.com>
50ce6f88
N		 22101
22102 * config/aarch64/aarch64-elf.h (REGISTER_PREFIX): Remove.
22103 * config/aarch64/aarch64.c (aarch64_print_operand): Remove all
22104 occurrence of REGISTER_PREFIX as its empty string.
22105
4f1f78b9
JL		 221062013-03-20 Jeff Law <law@redhat.com>
22107
22108 * tree-ssa-dom.c (record_equivalences_from_incoming_edge): Record
22109 addititional equivalences for equality comparisons between an SSA_NAME
22110 and a constant where the SSA_NAME was set from a widening conversion.
22111
327a1118
WL		 221122013-03-20 Walter Lee <walt@tilera.com>
22113
22114 * config/tilegx/sync.md (atomic_test_and_set): New pattern.
22115
813e0036
UB		 221162013-03-20 Uros Bizjak <ubizjak@gmail.com>
22117
22118 * config/i386/i386.md (*movoi_internal_avx): Emit insn template
22119 depending on type attribute.
22120 (*movti_internal): Ditto.
22121 (*movtf_internal): Ditto.
22122 (*movxf_internal): Ditto.
22123 (*movdf_internal): Ditto.
22124 (*movsf_internal): Ditto.
22125
7cf34aae
UB		 221262013-03-20 Uros Bizjak <ubizjak@gmail.com>
22127
22128 * config/i386/i386.md (*movti_internal): Set prefix attribute to
22129 maybe_vex for sselog1 and ssemov types.
22130 (*movdi_internal): Reorder operand constraints.
22131 (*movsi_internal): Ditto. Set prefix attribute to
22132 maybe_vex for sselog1 and ssemov types.
22133 (*movtf_internal): Set prefix attribute to maybe_vex
22134 for sselog1 and ssemov types.
22135 (*movdf_internal): Ditto. Set prefix_data16 attribute for
22136 DImode ssemov types. Reorder operand constraints.
22137 (*movsf_internal): Set type of alternatives 3,4 to imov. Set prefix
22138 attribute to maybe_vex for sselog1 and ssemov types. Set prefix_data16
22139 attribute for SImode ssemov types. Reorder operand constraints.
22140
19321415
MJ		 221412013-03-20 Martin Jambor <mjambor@suse.cz>
22142
22143 * params.def (PARAM_IPA_CP_ARRAY_INDEX_HINT_BONUS): New parameter.
22144 * ipa-cp.c (hint_time_bonus): Add abonus for known array indices.
22145
12211b99 221462013-03-20 Pat Haugen <pthaugen@us.ibm.com>
c6d5ff83
MM		 22147
22148 * config/rs6000/predicates.md (indexed_address, update_address_mem
22149 update_indexed_address_mem): New predicates.
22150 * config/rs6000/vsx.md (vsx_extract_<mode>_zero): Set correct "type"
22151 attribute for load/store instructions.
22152 * config/rs6000/dfp.md (movsd_store): Likewise.
22153 (movsd_load): Likewise.
22154 * config/rs6000/rs6000.md (zero_extend<mode>di2_internal1): Likewise.
22155 (unnamed HI->DI extend define_insn): Likewise.
22156 (unnamed SI->DI extend define_insn): Likewise.
22157 (unnamed QI->SI extend define_insn): Likewise.
22158 (unnamed QI->HI extend define_insn): Likewise.
22159 (unnamed HI->SI extend define_insn): Likewise.
22160 (unnamed HI->SI extend define_insn): Likewise.
22161 (extendsfdf2_fpr): Likewise.
22162 (movsi_internal1): Likewise.
22163 (movsi_internal1_single): Likewise.
22164 (movhi_internal): Likewise.
22165 (movqi_internal): Likewise.
22166 (movcc_internal1): Correct mnemonic for stw insn. Set correct "type"
22167 attribute for load/store instructions.
22168 (mov<mode>_hardfloat): Set correct "type" attribute for load/store
22169 instructions.
22170 (mov<mode>_softfloat): Likewise.
22171 (mov<mode>_hardfloat32): Likewise.
22172 (mov<mode>_hardfloat64): Likewise.
22173 (mov<mode>_softfloat64): Likewise.
22174 (movdi_internal32): Likewise.
22175 (movdi_internal64): Likewise.
22176 (probe_stack_<mode>): Likewise.
22177
221782013-03-20 Michael Meissner <meissner@linux.vnet.ibm.com>
22179
22180 * config/rs6000/vector.md (VEC_R): Add 32-bit integer, binary
22181 floating point, and decimal floating point to reload iterator.
22182
22183 * config/rs6000/constraints.md (wl constraint): New constraints to
22184 return FLOAT_REGS if certain options are used to reduce the number
22185 of separate patterns that exist in the file.
22186 (wx constraint): Likewise.
22187 (wz constraint): Likewise.
22188
22189 * config/rs6000/rs6000.c (rs6000_debug_reg_global): If
22190 -mdebug=reg, print wg, wl, wx, and wz constraints.
22191 (rs6000_init_hard_regno_mode_ok): Initialize new constraints.
22192 Initialize the reload functions for 64-bit binary/decimal floating
22193 point types.
22194 (reg_offset_addressing_ok_p): If we are on a power7 or later, use
22195 LFIWZX and STFIWX to load/store 32-bit decimal types, and don't
22196 create the buffer on the stack to overcome not having a 32-bit
22197 load and store.
22198 (rs6000_emit_move): Likewise.
22199 (rs6000_secondary_memory_needed_rtx): Likewise.
22200 (rs6000_alloc_sdmode_stack_slot): Likewise.
22201 (rs6000_preferred_reload_class): On VSX, we can create SFmode 0.0f
22202 via xxlxor, just like DFmode 0.0.
22203
22204 * config/rs6000/rs6000.h (TARGET_NO_SDMODE_STACK): New macro,
22205 define as 1 if we are running on a power7 or newer.
22206 (enum r6000_reg_class_enum): Add new constraints.
22207
22208 * config/rs6000/dfp.md (movsd): Delete, combine with binary
22209 floating point moves in rs6000.md. Combine power6x (mfpgpr) moves
22210 with other moves by using conditional constraits (wg). Use LFIWZX
bd059b26 22211 and STFIWX for loading SDmode on power7. Use xxlxor to create 0.0f.
c6d5ff83
MM		 22212 (movsd splitter): Likewise.
22213 (movsd_hardfloat): Likewise.
22214 (movsd_softfloat): Likewise.
22215
22216 * config/rs6000/rs6000.md (FMOVE32): New iterators to combine
22217 binary and decimal floating point moves.
22218 (fmove_ok): New attributes to combine binary and decimal floating
22219 point moves, and to combine power6x (mfpgpr) moves along normal
22220 floating moves.
22221 (real_value_to_target): Likewise.
22222 (f32_lr): Likewise.
22223 (f32_lm): Likewise.
22224 (f32_li): Likewise.
22225 (f32_sr): Likewise.
22226 (f32_sm): Likewise.
22227 (f32_si): Likewise.
22228 (movsf): Combine binary and decimal floating point moves. Combine
22229 power6x (mfpgpr) moves with other moves by using conditional
bd059b26 22230 constraits (wg). Use LFIWZX and STFIWX for loading SDmode on power7.
c6d5ff83
MM		 22231 (mov<mode> for SFmode/SDmode); Likewise.
22232 (SFmode/SDmode splitters): Likewise.
22233 (movsf_hardfloat): Likewise.
22234 (mov<mode>_hardfloat for SFmode/SDmode): Likewise.
22235 (movsf_softfloat): Likewise.
22236 (mov<mode>_softfloat for SFmode/SDmode): Likewise.
22237
22238 * doc/md.texi (PowerPC and IBM RS6000 constraints): Document wl,
22239 wx and wz constraints.
22240
22241 * config/rs6000/constraints.md (wg constraint): New constraint to
22242 return FLOAT_REGS if -mmfpgpr (power6x) was used.
22243
22244 * config/rs6000/rs6000.h (enum r6000_reg_class_enum): Add wg
22245 constraint.
22246
22247 * config/rs6000/rs6000.c (rs6000_debug_reg_global): If
22248 -mdebug=reg, print wg, wl, wx, and wz constraints.
22249 (rs6000_init_hard_regno_mode_ok): Initialize new constraints.
22250 Initialize the reload functions for 64-bit binary/decimal floating
22251 point types.
22252 (reg_offset_addressing_ok_p): If we are on a power7 or later, use
22253 LFIWZX and STFIWX to load/store 32-bit decimal types, and don't
22254 create the buffer on the stack to overcome not having a 32-bit
22255 load and store.
22256 (rs6000_emit_move): Likewise.
22257 (rs6000_secondary_memory_needed_rtx): Likewise.
22258 (rs6000_alloc_sdmode_stack_slot): Likewise.
22259 (rs6000_preferred_reload_class): On VSX, we can create SFmode 0.0f
22260 via xxlxor, just like DFmode 0.0.
22261
c6d5ff83
MM		 22262 * config/rs6000/dfp.md (movdd): Delete, combine with binary
22263 floating point moves in rs6000.md. Combine power6x (mfpgpr) moves
22264 with other moves by using conditional constraits (wg). Use LFIWZX
22265 and STFIWX for loading SDmode on power7.
22266 (movdd splitters): Likewise.
22267 (movdd_hardfloat32): Likewise.
22268 (movdd_softfloat32): Likewise.
22269 (movdd_hardfloat64_mfpgpr): Likewise.
22270 (movdd_hardfloat64): Likewise.
22271 (movdd_softfloat64): Likewise.
22272
22273 * config/rs6000/rs6000.md (FMOVE64): New iterators to combine
22274 64-bit binary and decimal floating point moves.
22275 (FMOVE64X): Likewise.
22276 (movdf): Combine 64-bit binary and decimal floating point moves.
22277 Combine power6x (mfpgpr) moves with other moves by using
22278 conditional constraits (wg).
22279 (mov<mode> for DFmode/DDmode): Likewise.
22280 (DFmode/DDmode splitters): Likewise.
22281 (movdf_hardfloat32): Likewise.
22282 (mov<mode>_hardfloat32 for DFmode/DDmode): Likewise.
22283 (movdf_softfloat32): Likewise.
22284 (movdf_hardfloat64_mfpgpr): Likewise.
22285 (movdf_hardfloat64): Likewise.
22286 (mov<mode>_hardfloat64 for DFmode/DDmode): Likewise.
22287 (movdf_softfloat64): Likewise.
22288 (mov<mode>_softfloat64 for DFmode/DDmode): Likewise.
22289 (reload_<mode>_load): Move to later in the file so they aren't in
22290 the middle of the floating point move insns.
22291 (reload_<mode>_store): Likewise.
22292
22293 * doc/md.texi (PowerPC and IBM RS6000 constraints): Document wg
22294 constraint.
22295
22296 * config/rs6000/rs6000.c (rs6000_debug_reg_global): Print out wg
22297 constraint if -mdebug=reg.
bd059b26
UB		 22298 (rs6000_initi_hard_regno_mode_ok): Enable wg constraint if -mfpgpr.
22299 Enable using dd reload support if needed.
c6d5ff83
MM		 22300
22301 * config/rs6000/dfp.md (movtd): Delete, combine with 128-bit
22302 binary and decimal floating point moves in rs6000.md.
22303 (movtd_internal): Likewise.
22304
22305 * config/rs6000/rs6000.md (FMOVE128): Combine 128-bit binary and
22306 decimal floating point moves.
22307 (movtf): Likewise.
22308 (movtf_internal): Likewise.
22309 (mov<mode>_internal, TDmode/TFmode): Likewise.
22310 (movtf_softfloat): Likewise.
22311 (mov<mode>_softfloat, TDmode/TFmode): Likewise.
22312
22313 * config/rs6000/rs6000.md (movdi_mfpgpr): Delete, combine with
22314 movdi_internal64, using wg constraint for move direct operations.
22315 (movdi_internal64): Likewise.
22316
22317 * config/rs6000/rs6000.c (rs6000_debug_reg_global): Print
22318 MODES_TIEABLE_P for selected modes. Print the numerical value of
22319 the various virtual registers. Use GPR/FPR first/last values,
22320 instead of hard coding the register numbers. Print which modes
22321 have reload functions registered.
bd059b26
UB		 22322 (rs6000_option_override_internal): If -mdebug=reg, trace the options
22323 settings before/after setting cpu, target and subtarget settings.
22324 (rs6000_secondary_reload_trace): Improve the RTL dump for -mdebug=addr
22325 and for secondary reload failures in rs6000_secondary_reload_inner.
c6d5ff83
MM		 22326 (rs6000_secondary_reload_fail): Likewise.
22327 (rs6000_secondary_reload_inner): Likewise.
22328
22329 * config/rs6000/rs6000.md (FIRST_GPR_REGNO): Add convenience
22330 macros for first/last GPR and FPR registers.
22331 (LAST_GPR_REGNO): Likewise.
22332 (FIRST_FPR_REGNO): Likewise.
22333 (LAST_FPR_REGNO): Likewise.
22334
22335 * config/rs6000/vector.md (mul<mode>3): Use the combined macro
22336 VECTOR_UNIT_ALTIVEC_OR_VSX_P instead of separate calls to
22337 VECTOR_UNIT_ALTIVEC_P and VECTOR_UNIT_VSX_P.
22338 (vcond<mode><mode>): Likewise.
22339 (vcondu<mode><mode>): Likewise.
22340 (vector_gtu<mode>): Likewise.
22341 (vector_gte<mode>): Likewise.
22342 (xor<mode>3): Don't allow logical operations on TImode in 32-bit
bd059b26 22343 to prevent the compiler from converting DImode operations to TImode.
c6d5ff83
MM		 22344 (ior<mode>3): Likewise.
22345 (and<mode>3): Likewise.
22346 (one_cmpl<mode>2): Likewise.
22347 (nor<mode>3): Likewise.
22348 (andc<mode>3): Likewise.
22349
22350 * config/rs6000/constraints.md (wt constraint): New constraint
22351 that returns VSX_REGS if TImode is allowed in VSX registers.
22352
22353 * config/rs6000/predicates.md (easy_fp_constant): 0.0f is an easy
22354 constant under VSX.
22355
22356 * config/rs6000/rs6000-modes.def (PTImode): Define, PTImode is
22357 similar to TImode, but it is restricted to being in the GPRs.
22358
22359 * config/rs6000/rs6000.opt (-mvsx-timode): New switch to allow
22360 TImode to occupy a single VSX register.
22361
22362 * config/rs6000/rs6000-cpus.def (ISA_2_6_MASKS_SERVER): Default to
22363 -mvsx-timode for power7/power8.
22364 (power7 cpu): Likewise.
22365 (power8 cpu): Likewise.
22366
22367 * config/rs6000/rs6000.c (rs6000_hard_regno_nregs_internal): Make
22368 sure that TFmode/TDmode take up two registers if they are ever
22369 allowed in the upper VSX registers.
22370 (rs6000_hard_regno_mode_ok): If -mvsx-timode, allow TImode in VSX
22371 registers.
22372 (rs6000_init_hard_regno_mode_ok): Likewise.
22373 (rs6000_debug_reg_global): Add debugging for PTImode and wt
22374 constraint. Print if LRA is turned on.
22375 (rs6000_option_override_internal): Give an error if -mvsx-timode
22376 and VSX is not enabled.
22377 (invalid_e500_subreg): Handle PTImode, restricting it to GPRs. If
22378 -mvsx-timode, restrict TImode to reg+reg addressing, and PTImode
22379 to reg+offset addressing. Use PTImode when checking offset
22380 addresses for validity.
22381 (reg_offset_addressing_ok_p): Likewise.
22382 (rs6000_legitimate_offset_address_p): Likewise.
22383 (rs6000_legitimize_address): Likewise.
22384 (rs6000_legitimize_reload_address): Likewise.
22385 (rs6000_legitimate_address_p): Likewise.
22386 (rs6000_eliminate_indexed_memrefs): Likewise.
22387 (rs6000_emit_move): Likewise.
22388 (rs6000_secondary_reload): Likewise.
22389 (rs6000_secondary_reload_inner): Handle PTImode. Allow 64-bit
22390 reloads to fpr registers to continue to use reg+offset addressing,
22391 but 64-bit reloads to altivec registers need reg+reg addressing.
22392 Drop test for PRE_MODIFY, since VSX loads/stores no longer support
22393 it. Treat LO_SUM like a PLUS operation.
22394 (rs6000_secondary_reload_class): If type is 64-bit, prefer to use
bd059b26 22395 FLOAT_REGS instead of VSX_RGS to allow use of reg+offset addressing.
c6d5ff83
MM		 22396 (rs6000_cannot_change_mode_class): Do not allow TImode in VSX
22397 registers to share a register with a smaller sized type, since VSX
22398 puts scalars in the upper 64-bits.
22399 (print_operand): Add support for PTImode.
22400 (rs6000_register_move_cost): Use VECTOR_MEM_VSX_P instead of
22401 VECTOR_UNIT_VSX_P to catch types that can be loaded in VSX
22402 registers, but don't have arithmetic support.
22403 (rs6000_memory_move_cost): Add test for VSX.
22404 (rs6000_opt_masks): Add -mvsx-timode.
22405
22406 * config/rs6000/vsx.md (VSm): Change to use 64-bit aligned moves
22407 for TImode.
22408 (VSs): Likewise.
22409 (VSr): Use wt constraint for TImode.
22410 (VSv): Drop TImode support.
22411 (vsx_movti): Delete, replace with versions for 32-bit and 64-bit.
22412 (vsx_movti_64bit): Likewise.
22413 (vsx_movti_32bit): Likewise.
22414 (vec_store_<mode>): Use VSX iterator instead of vector iterator.
22415 (vsx_and<mode>3): Delete use of '?' constraint on inputs, just put
22416 one '?' on the appropriate output constraint. Do not allow TImode
22417 logical operations on 32-bit systems.
22418 (vsx_ior<mode>3): Likewise.
22419 (vsx_xor<mode>3): Likewise.
22420 (vsx_one_cmpl<mode>2): Likewise.
22421 (vsx_nor<mode>3): Likewise.
22422 (vsx_andc<mode>3): Likewise.
22423 (vsx_concat_<mode>): Likewise.
22424 (vsx_xxpermdi_<mode>): Fix thinko for non V2DF/V2DI modes.
22425
22426 * config/rs6000/rs6000.h (MASK_VSX_TIMODE): Map from
22427 OPTION_MASK_VSX_TIMODE.
22428 (enum rs6000_reg_class_enum): Add RS6000_CONSTRAINT_wt.
22429 (STACK_SAVEAREA_MODE): Use PTImode instead of TImode.
22430
22431 * config/rs6000/rs6000.md (INT mode attribute): Add PTImode.
22432 (TI2 iterator): New iterator for TImode, PTImode.
22433 (wd mode attribute): Add values for vector types.
bd059b26
UB		 22434 (movti_string): Replace TI move operations with operations for TImode
22435 and PTImode. Add support for TImode being allowed in VSX registers.
c6d5ff83
MM		 22436 (mov<mode>_string, TImode/PTImode): Likewise.
22437 (movti_ppc64): Likewise.
22438 (mov<mode>_ppc64, TImode/PTImode): Likewise.
22439 (TI mode splitters): Likewise.
22440
22441 * doc/md.texi (PowerPC and IBM RS6000 constraints): Document wt
22442 constraint.
22443
1fc5eced
MG		 224442013-03-20 Marc Glisse <marc.glisse@inria.fr>
22445
22446 PR tree-optimization/56355
22447 * fold-const.c (tree_binary_nonnegative_warnv_p) <MULT_EXPR>:
22448 Also handle integers with undefined overflow.
22449
22c4c869
CM		 224502013-03-20 Catherine Moore <clm@codesourcery.com>
22451 Maciej W. Rozycki <macro@codesourcery.com>
22452 Tom de Vries <tom@codesourcery.com>
12211b99 22453 Nathan Sidwell <nathan@codesourcery.com>
22c4c869
CM		 22454 Iain Sandoe <iain@codesourcery.com>
22455 Nathan Froyd <froydnj@codesourcery.com>
12211b99 22456 Chao-ying Fu <fu@mips.com>
22c4c869
CM		 22457
22458 * doc/extend.texi: (micromips, nomicromips, nocompression):
7cf34aae 22459 Document new function attributes.
22c4c869
CM		 22460 * doc/invoke.texi (minterlink-compressed, mmicromips,
22461 m14k, m14ke, m14kec): Document new options.
22462 (minterlink-mips16): Update documentation.
22463 * doc/md.texi (ZC, ZD): Document new constraints.
22464 * configure.ac (gcc_cv_as_micromips): Check if linker
22465 supports the .set micromips directive.
22466 * configure: Regenerate.
22467 * config.in: Regenerate.
22468 * config/mips/mips-tables.opt: Regenerate.
22469 * config/mips/micromips.md: New file.
22470 * constraints.md (ZC, ZD): New constraints.
22471 * config/mips/predicates.md (movep_src_register): New predicate.
22472 (movep_src_operand): New predicate.
22473 (non_volatile_mem_operand): New predicate.
22474 * config/mips/mips.md (multimem): New type.
22475 (length): Differentiate between 17-bit and 18-bit branch offsets.
22476 (MOVEP1, MOVEP2): New mode iterator.
33159866 22477 (mov_<load>l): Use ZC constraint.
22c4c869
CM		 22478 (mov_<load>r): Likewise.
22479 (mov_<store>l): Likewise.
22480 (mov_<store>r): Likewise.
22481 (*branch_equality<mode>_inverted): Add microMIPS support.
22482 (*branch_equality<mode>): Likewise.
22483 (*jump_absolute): Likewise.
22484 (indirect_jump_<mode>): Likewise.
22485 (tablejump_<mode>): Likewise.
22486 (<optab>_internal): Likewise.
22487 (sibcall_internal): Likewise.
22488 (sibcall_value_internal): Likewise.
22489 (prefetch): Use constraint ZD.
22490 * config/mips/mips.opt (minterlink-compressed): New option.
22491 (minterlink-mips16): Now an alias for minterlink-compressed.
22492 (mmicromips): New option.
22493 * config/mips/sync.md (sync_compare_and_swap<mode>): Use ZR constraint.
22494 (compare_and_swap_12): Likewise.
22495 (sync_add<mode>): Likewise.
22496 (sync_<optab>_12): Likewise.
22497 (sync_old_<optab>_12): Likewise.
22498 (sync_new_<optab>_12): Likewise.
22499 (sync_nand_12): Likewise.
22500 (sync_old_nand_12): Likewise.
22501 (sync_new_nand_12): Likewise.
22502 (sync_sub<mode>): Likewise.
22503 (sync_old_add<mode>): Likewise.
22504 (sync_old_sub<mode>): Likewise.
22505 (sync_new_add<mode>): Likewise.
22506 (sync_new_sub<mode>): Likewise.
22507 (sync_<optab><mode>): Likewise.
22508 (sync_old_<optab><mode>): Likewise.
22509 (sync_new_<optab><mode>): Likewise.
22510 (sync_nand<mode>): Likewise.
22511 (sync_old_nand<mode>): Likewise.
22512 (sync_new_nand<mode>): Likewise.
22513 (sync_lock_test_and_set<mode>): Likewise.
22514 (test_and_set_12): Likewise.
22515 (atomic_compare_and_swap<mode>): Likewise.
22516 (atomic_exchange<mode>_llsc): Likewise.
22517 (atomic_fetch_add<mode>_llsc): Likewise.
22518 * config/mips/mips-cpus.def (m14kc, m14k): New processors.
22519 * config/mips/mips-protos.h (umips_output_save_restore): New prototype.
22520 (umips_save_restore_pattern_p): Likewise.
22521 (umips_load_store_pair_p): Likewise.
22522 (umips_output_load_store_pair): Likewise.
22523 (umips_movep_target_p): Likewise.
22524 (umips_12bit_offset_address_p): Likewise.
22525 * config/mips/mips.c (MIPS_MAX_FIRST_STEP): Update for microMIPS.
22526 (mips_base_mips16): Rename this...
22527 (mips_base_compression_flags): ...to this. Update all uses.
22528 (mips_attribute_table): Add micromips, nomicromips and nocompression.
22529 (mips_mips16_decl_p): Delete.
22530 (mips_nomips16_decl_p): Delete.
33159866 22531 (mips_get_compress_on_flags): New function.
22c4c869
CM		 22532 (mips_get_compress_off_flags): New function.
22533 (mips_get_compress_mode): New function.
22534 (mips_get_compress_on_name): New function.
22535 (mips_get_compress_off_name): New function.
22536 (mips_insert_attributes): Support multiple compression types.
22537 (mips_merge_decl_attributes): Likewise.
22538 (umips_12bit_offset_address_p): New function.
22539 (mips_start_function_definition): Emit .set micromips directive.
22540 (mips_call_may_need_jalx_p): New function.
22541 (mips_function_ok_for_sibcall): Add microMIPS support.
22542 (mips_print_operand_punctuation): Support short delay slots and
22543 compact jumps.
22544 (umips_swm_mask, umips_swm_encoding): New.
22545 (umips_build_save_restore): New function.
22546 (mips_for_each_saved_gpr_and_fpr): Add microMIPS support.
22547 (was_mips16_p): Remove.
22548 (old_compression_mode): New.
22549 (mips_set_compression_mode): New function.
22550 (mips_set_current_function): Add microMIPS support.
22551 (mips_option_override): Likewise.
22552 (umips_save_restore_pattern_p): New function.
22553 (umips_output_save_restore): New function.
22554 (umips_load_store_pair_p_1): New function.
22555 (umips_load_store_pair_p): New function.
22556 (umips_output_load_store_pair_1): New function.
22557 (umips_output_load_store_pair): New function.
22558 (umips_movep_target_p) New function.
22559 (mips_prepare_pch_save): Add microMIPS support.
22560 * config/mips/mips.h (TARGET_COMPRESSION): New.
22561 (TARGET_CPU_CPP_BUILTINS): Update macro
22562 to use new compression flags and to support microMIPS.
22563 (MIPS_ISA_LEVEL_SPEC): Add m14k processors.
22564 (MIPS_ARCH_FLOAT_SPEC): Likewise.
22565 (ISA_HAS_LWXS): Include TARGET_MICROMIPS.
22566 (ISA_HAS_LOAD_DELAY): Exclude TARGET_MICROMIPS.
22567 (ASM_SPEC): Support mmicromips and mno-micromips.
22568 (M16STORE_REG_P): New macro.
22569 (MIPS_CALL): Support TARGET_MICROMIPS.
22570 (MICROMIPS_J): New macro.
22571 (mips_base_mips16): Rename this...
22572 (mips_base_compression_flags): ...to this.
22573 (UMIPS_12BIT_OFFSET_P): New macro.
22574 * config/mips/t-sde: (MULTILIB_OPTIONS): Add microMIPS.
22575 (MULTILIB_DIRNAMES): Likewise.
92a8d7a7
RB		 225762013-03-20 Richard Biener <rguenther@suse.de>
22577
22578 PR tree-optimization/56661
22579 * tree-ssa-sccvn.c (visit_use): Only value-number calls if
22580 the result does not have to be distinct.
22581
54714c68
RB		 225822013-03-20 Richard Biener <rguenther@suse.de>
22583
22584 * tree-inline.c (copy_tree_body_r): Sync MEM_REF code with
22585 remap_gimple_op_r.
22586
cca1130d 225872013-03-20 Bill Schmidt <wschmidt@linux.vnet.ibm.com>
12211b99 22588 Steven Bosscher <steven@gcc.gnu.org>
cca1130d
BS		 22589
22590 PR rtl-optimization/56605
22591 * loop-iv.c (implies_p): Handle equal RTXs and subregs.
22592
2a930377
UB		 225932013-03-20 Uros Bizjak <ubizjak@gmail.com>
22594
22595 PR bootstrap/56656
22596 * config/i386/i386.md (*movdi_internal): Handle broken assemblers
22597 that require movd instead of movq.
22598
d6d305fe
RB		 225992013-03-20 Richard Biener <rguenther@suse.de>
22600
22601 * tree-ssa-structalias.c (struct variable_info): Add pointer
22602 to the first field of an aggregate with sub-vars. Make
22603 this and the pointer to the next subfield its ID.
22604 (vi_next): New function.
22605 (nothing_id, anything_id, readonly_id, escaped_id, nonlocal_id,
22606 storedanything_id, integer_id): Increment by one.
22607 (new_var_info, get_call_vi, lookup_call_clobber_vi,
22608 get_call_clobber_vi): Adjust.
22609 (solution_set_expand): Simplify and speedup.
22610 (solution_set_add): Inline into ...
22611 (set_union_with_increment): ... this. Adjust accordingly.
22612 (do_sd_constraint): Likewise.
22613 (do_ds_constraint): Likewise.
22614 (do_complex_constraint): Simplify.
22615 (build_pred_graph): Adjust.
22616 (solve_graph): Likewise. Simplify and speedup.
22617 (get_constraint_for_ssa_var, get_constraint_for_ptr_offset,
22618 get_constraint_for_component_ref, get_constraint_for_1,
22619 first_vi_for_offset, first_or_preceding_vi_for_offset,
22620 create_function_info_for, create_variable_info_for_1,
22621 create_variable_info_for, intra_create_variable_infos): Adjust.
22622 (init_base_vars): Push NULL for ID zero.
22623 (compute_points_to_sets): Adjust.
22624
a271b42d
RB		 226252013-03-20 Richard Biener <rguenther@suse.de>
22626
22627 * cfgloop.c (verify_loop_structure): Streamline and avoid
22628 ICEing on corrupt loop tree.
22629 * graph.c (draw_cfg_nodes_for_loop): Avoid ICEing on corrupt
22630 loop tree.
22631
ebd65954
RB		 226322013-03-20 Richard Biener <rguenther@suse.de>
22633
22634 * tree-vect-loop-manip.c (slpeel_can_duplicate_loop_p): Do not
22635 check whether an SSA update is needed.
22636
4547b7ee
RS		 226372013-03-20 Richard Sandiford <rdsandiford@googlemail.com>
22638
22639 * config/mips/constraints.md (T): Rename to...
22640 (Yf): ...this.
22641 (U): Rename to...
22642 (Yd): ...this.
22643 * config/mips/mips.md (*movdi_64bit, *movdi_64bit_mips16)
22644 (*mov<mode>_internal, *mov<mode>_mips16): Update accordingly.
22645
0b8cdc58
IB		 226462013-03-19 Ian Bolton <ian.bolton@arm.com>
22647
22648 * config/aarch64/aarch64.md (*sub<mode>3_carryin): New pattern.
22649 (*subsi3_carryin_uxtw): Likewise.
22650
5977a10d
IB		 226512013-03-19 Ian Bolton <ian.bolton@arm.com>
22652
22653 * config/aarch64/aarch64.md (*ror<mode>3_insn): New pattern.
22654 (*rorsi3_insn_uxtw): Likewise.
22655
bd83ff2c
IB		 226562013-03-19 Ian Bolton <ian.bolton@arm.com>
22657
22658 * config/aarch64/aarch64.md (*extr<mode>5_insn): New pattern.
22659 (*extrsi5_insn_uxtw): Likewise.
22660
ba65123c
RB		 226612013-03-19 Richard Biener <rguenther@suse.de>
22662
22663 PR tree-optimization/56273
22664 * passes.c (init_optimization_passes): Move second VRP after DOM.
22665
2eac0476
UB		 226662013-03-19 Uros Bizjak <ubizjak@gmail.com>
22667
22668 * config/i386/i386.md (*movti_internal): Merge from
22669 *movti_internal_rex64 and *movti_internal_sse. Use x64 isa attribute.
22670 (*movdi_internal): Merge with *movdi_internal_rex64. Use x64 and
22671 nox64 isa attributes.
22672
6e55eda7
RB		 226732013-03-18 Richard Biener <rguenther@suse.de>
22674
22675 * tree-ssa-structalias.c (find): Use gcc_checking_assert.
22676 (unite): Likewise.
22677 (merge_node_constraints): Likewise.
22678 (build_succ_graph): Likewise.
22679 (valid_graph_edge): Inline into single caller.
22680 (unify_nodes): Likewise. Use bitmap_set_bit return value
22681 and cache varinfo.
22682 (scc_visit): Fix formatting and variable use.
22683 (do_sd_constraint): Use gcc_checking_assert.
22684 (do_ds_constraint): Likewise.
22685 (do_complex_constraint): Likewise.
22686 (condense_visit): Likewise. Cleanup.
22687 (dump_pred_graph): New function.
22688 (perform_var_substitution): Dump the pred-graph before
22689 variable substitution.
22690 (find_equivalent_node): Use gcc_checking_assert.
22691 (rewrite_constraints): Guard checking loop with ENABLE_CHECKING.
22692
4bdd44c4
RB		 226932013-03-18 Richard Biener <rguenther@suse.de>
22694
22695 * tree-vect-loop-manip.c (vect_create_cond_for_alias_checks):
22696 Remove cond_expr_stmt_list argument and do not gimplify the
22697 built expression.
22698 (vect_loop_versioning): Adjust.
22699 * tree-vect-data-refs.c (vect_create_addr_base_for_vector_ref):
22700 Cleanup to use less temporaries.
22701 (vect_create_data_ref_ptr): Cleanup.
22702
38c56a5b
JJ		 227032013-03-18 Jakub Jelinek <jakub@redhat.com>
22704
22705 PR tree-optimization/56635
22706 * fold-const.c (operand_equal_p): For MEM_REF and TARGET_MEM_REF,
22707 require types_compatible_p types.
22708
20b2e6a0
NC		 227092013-03-18 Nick Clifton <nickc@redhat.com>
22710
a6178a25
NC		 22711 * config/stormy16/stormy16.c (xstormy16_expand_prologue): Remove
22712 spurious backslash.
22713
20b2e6a0
NC		 22714 * config/mn10300/mn10300.c (mn10300_get_live_callee_saved_regs):
22715 Add missing line to comment describing function.
22716
92608d0e
RB		 227172013-03-18 Richard Biener <rguenther@suse.de>
22718
22719 PR tree-optimization/56210
22720 * tree-ssa-structalias.c (find_func_aliases_for_builtin_call):
22721 Handle string / character search functions.
22722 * tree-ssa-alias.c (ref_maybe_used_by_call_p_1): Likewise.
22723
31348d52
RB		 227242013-03-18 Richard Biener <rguenther@suse.de>
22725
22726 PR middle-end/56483
22727 * cfgexpand.c (expand_gimple_cond): Inline gimple_cond_single_var_p
22728 and implement properly.
22729 * gimple.h (gimple_cond_single_var_p): Remove.
22730
fcac74a1
RB		 227312013-03-18 Richard Biener <rguenther@suse.de>
22732
22733 * tree-data-ref.h (find_data_references_in_loop): Declare.
22734 * tree-data-ref.c (get_references_in_stmt): Use a stack
22735 vector pre-allocated in the callers.
22736 (find_data_references_in_stmt): Adjust.
22737 (graphite_find_data_references_in_stmt): Likewise.
22738 (create_rdg_vertices): Likewise.
22739 (find_data_references_in_loop): Export.
22740 * tree-vect-data-refs.c (vect_analyze_data_ref_dependences):
22741 Compute dependences here...
22742 (vect_analyze_data_refs): ...not here. When we encounter
22743 a non-vectorizable data reference in basic-block vectorization
22744 truncate the data reference vector. Do not bother to
22745 fixup data-dependence information for gather loads.
22746 * tree-vect-slp.c (vect_slp_analyze_bb_1): Check the number
22747 of data references, as reported.
22748
0d5a1b56
RB		 227492013-03-18 Richard Biener <rguenther@suse.de>
22750
22751 PR tree-optimization/3713
22752 * tree-ssa-sccvn.c (visit_copy): Simplify. Always propagate
22753 has_constants and expr.
22754 (stmt_has_constants): Properly valueize SSA names when deciding
22755 whether the stmt has constants.
22756
789c34e3
RB		 227572013-03-18 Richard Biener <rguenther@suse.de>
22758
22759 * tree-ssa-loop-manip.c (find_uses_to_rename): Do not scan the
22760 whole function when there is nothing to do.
22761 * tree-ssa-loop.c (pass_vectorize): Remove TODO_update_ssa.
22762 * tree-vectorizer.c (vectorize_loops): Update virtual and
22763 loop-closed SSA once.
22764 * tree-vect-loop.c (vect_transform_loop): Do not update SSA here.
22765
076b4605
RB		 227662013-03-18 Richard Biener <rguenther@suse.de>
22767
22768 PR middle-end/56113
22769 * domwalk.c (bb_postorder): New global static.
22770 (cmp_bb_postorder): New function.
22771 (walk_dominator_tree): Replace scheme imposing an order for
22772 visiting dominator sons by one sorting them at the time they
22773 are pushed on the stack.
22774
bdb01696
RB		 227752013-03-18 Richard Biener <rguenther@suse.de>
22776
22777 PR tree-optimization/39326
22778 * tree-ssa-loop-im.c (refs_independent_p): Exploit symmetry.
22779 (struct mem_ref): Replace mem member with ao_ref typed member.
22780 (MEM_ANALYZABLE): Adjust.
22781 (memref_eq): Likewise.
22782 (mem_ref_alloc): Likewise.
22783 (gather_mem_refs_stmt): Likewise.
22784 (mem_refs_may_alias_p): Use the ao_ref to query the alias oracle.
22785 (execute_sm_if_changed_flag_set): Adjust.
22786 (execute_sm): Likewise.
22787 (ref_always_accessed_p): Likewise.
22788 (refs_independent_p): Likewise.
22789 (can_sm_ref_p): Likewise.
22790
12d2dc5e
JJ		 227912013-03-18 Jakub Jelinek <jakub@redhat.com>
22792
22793 PR c/56566
22794 * tree.c (tree_int_cst_min_precision): For integer_zerop (value)
22795 return 1 even for !unsignedp.
22796
286fb677
UB		 227972013-03-17 Uros Bizjak <ubizjak@gmail.com>
22798
22799 * config/i386/i386.md (isa): Add x64 and nox64.
22800 (enabled): Define x64 for TARGET_64BIT and nox64 for !TARGET_64BIT.
22801 (*pushtf): Enable *roF alternative for x64 isa only.
22802 (*pushxf): Merge with *pushxf_nointeger. Use Yx*r constraint. Set
22803 mode attribute of integer alternatives to DImode for TARGET_64BIT.
22804 (*pushdf): Merge with *pushdf_rex64. Use x64 and nox64 isa attributes.
22805 (*movtf_internal): Merge from *movtf_internal_rex64 and
22806 *movtf_internal_sse. Use x64 and nox64 isa attributes.
22807 (*movxf_internal): Merge with *movxf_internal_rex64. Use x64 and
22808 nox64 isa attributes.
22809 (*movdf_internal): Merge with *movdf_internal_rex64. Use x64 and
22810 nox64 isa attributes.
22811 * config/i386/constraints.md (Yd): Do not set for TARGET_64BIT.
22812
88b97037
UB		 228132013-03-17 Uros Bizjak <ubizjak@gmail.com>
22814
22815 * config/alpha/alpha.c (TARGET_LRA_P): New define.
22816
9f4f1735
JJ		 228172013-03-17 Jakub Jelinek <jakub@redhat.com>
22818
22819 PR target/56640
22820 * config/arm/arm.h (REG_CLASS_NAMES): Add "SFP_REG" and "AFP_REG"
22821 class names. Remove trailing comma after "ALL_REGS".
22822
48f4a6fa
JH		 228232013-03-16 Jan Hubicka <jh@suse.cz>
22824
22825 * cgraph.h (cgraph_get_create_real_symbol_node): Declare.
22826 * cgraph.c (cgraph_get_create_real_symbol_node): New function.
22827 * cgrpahbuild.c: Use cgraph_get_create_real_symbol_node instead
22828 of cgraph_get_create_node.
22829 * ipa-prop.c (ipa_make_edge_direct_to_target): Likewise.
22830
98e81123
JM		 228312013-03-16 Jason Merrill <jason@redhat.com>
22832
22833 PR debug/49090
22834 * dwarf2out.c (gen_generic_params_dies): Indicate default arguments
22835 with DW_AT_default_value.
22836
8a298c04
JJ		 228372013-03-16 Jakub Jelinek <jakub@redhat.com>
22838
22839 * BASE-VER: Set to 4.9.0.
22840
4323afa0
AK		 228412013-03-14 Andi Kleen <ak@linux.intel.com>
22842
22843 PR target/56619
22844 * doc/extend.texi: Document __ATOMIC_HLE_ACQUIRE,
22845 __ATOMIC_HLE_RELEASE. Document __builtin_ia32 TSX intrincs.
22846 Document _x* TSX intrinsics.
22847
b3c0d469
JJ		 228482013-03-14 Edgar E. Iglesias <edgar.iglesias@xilinx.com>
22849 David Holsgrove <david.holsgrove@xilinx.com>
8cc9a5a5
EI		 22850
22851 * configure.ac: Add MicroBlaze TLS support detection.
22852 * configure: Regenerate.
88b97037
UB		 22853 * config/microblaze/microblaze-protos.h
22854 (microblaze_cannot_force_const_mem, microblaze_tls_referenced_p,
22855 symbol_mentioned_p, label_mentioned_p): Add prototypes.
22856 * config/microblaze/microblaze.c (microblaze_address_type): Add
22857 ADDRESS_TLS and tls_reloc address types.
8cc9a5a5
EI		 22858 (microblaze_address_info): Add tls_reloc.
22859 (TARGET_HAVE_TLS): Define.
22860 (get_tls_get_addr, microblaze_tls_symbol_p, microblaze_tls_operand_p_1,
22861 microblaze_tls_referenced_p, microblaze_cannot_force_const_mem,
88b97037
UB		 22862 symbol_mentioned_p, label_mentioned_p, tls_mentioned_p,
22863 load_tls_operand, microblaze_call_tls_get_addr,
22864 microblaze_legitimize_tls_address): New functions.
8cc9a5a5
EI		 22865 (microblaze_classify_unspec): Handle UNSPEC_TLS.
22866 (get_base_reg): Use microblaze_tls_symbol_p.
22867 (microblaze_classify_address): Handle TLS.
88b97037
UB		 22868 (microblaze_legitimate_pic_operand): Use symbol_mentioned_p,
22869 label_mentioned_p and microblaze_tls_referenced_p.
8cc9a5a5
EI		 22870 (microblaze_legitimize_address): Handle TLS.
22871 (microblaze_address_insns): Handle ADDRESS_TLS.
22872 (pic_address_needs_scratch): Handle TLS.
22873 (print_operand_address): Handle TLS.
22874 (microblaze_expand_prologue): Check TLS_NEEDS_GOT.
22875 (microblaze_expand_move): Handle TLS.
88b97037
UB		 22876 (microblaze_legitimate_constant_p): Check
22877 microblaze_cannot_force_const_mem and microblaze_tls_symbol_p.
8cc9a5a5
EI		 22878 (TARGET_CANNOT_FORCE_CONST_MEM): Define.
22879 * config/microblaze/microblaze.h (TLS_NEEDS_GOT): Define
22880 (PIC_OFFSET_TABLE_REGNUM): Set.
22881 * config/microblaze/linux.h (TLS_NEEDS_GOT): Define.
22882 * config/microblaze/microblaze.md (UNSPEC_TLS): Define.
22883 (addsi3, movsi_internal2, movdf_internal): Update constraints
22884 * config/microblaze/predicates.md (arith_plus_operand): Define
88b97037
UB		 22885 (move_operand): Redefine as move_src_operand,
22886 check microblaze_tls_referenced_p.
8cc9a5a5 22887
d803a491
IB		 228882013-03-14 Ian Bolton <ian.bolton@arm.com>
22889
22890 * config/aarch64/aarch64.md: (*and<mode>3nr_compare0): Use CC_NZ.
22891 (*and_<SHIFT:optab><mode>3nr_compare0): Likewise.
22892
a8504f22
IB		 228932013-03-14 Ian Bolton <ian.bolton@arm.com>
22894
22895 * config/aarch64/aarch64.c (aarch64_select_cc_mode): Return correct
22896 CC mode for AND.
22897
df2dfaea
JJ		 228982013-03-14 Jakub Jelinek <jakub@redhat.com>
22899
fbd28bc3
JJ		 22900 PR tree-optimization/53265
22901 * common.opt (Waggressive-loop-optimizations): New option.
22902 * tree-ssa-loop-niter.c: Include tree-pass.h.
22903 (do_warn_aggressive_loop_optimizations): New function.
22904 (record_estimate): Call it. Don't add !is_exit bounds to loop->bounds
22905 if number_of_latch_executions returned constant.
22906 (estimate_numbers_of_iterations_loop): Call number_of_latch_executions
22907 early. If number_of_latch_executions returned constant, set
22908 nb_iterations_upper_bound back to it.
22909 * cfgloop.h (struct loop): Add warned_aggressive_loop_optimizations
22910 field.
22911 * Makefile.in (tree-ssa-loop-niter.o): Depend on $(TREE_PASS_H).
22912 * doc/invoke.texi (-Wno-aggressive-loop-optimizations): Document.
22913
df2dfaea
JJ		 22914 * config/aarch64/t-aarch64-linux (MULTARCH_DIRNAME): Remove.
22915 (MULTILIB_OSDIRNAMES): Set.
22916 * genmultilib: If defaultosdirname doesn't start with :: , set
22917 defaultosdirname2 instead, clear it and emit two . multilib_raw
22918 entries instead of just one.
22919
ee0d2b68
KK		 229202013-03-14 Kaz Kojima <kkojima@gcc.gnu.org>
22921
22922 * config/sh/linux.h (TARGET_DEFAULT): Remove MASK_USERMODE.
22923 (SUBTARGET_OVERRIDE_OPTIONS): Set TARGET_USERMODE as default.
22924 * config/sh/netbsd-elf.h (TARGET_DEFAULT): Remove MASK_USERMODE.
22925 (SUBTARGET_OVERRIDE_OPTIONS): New.
22926
decc676e
OE		 229272013-03-13 Oleg Endo <olegendo@gcc.gnu.org>
22928
22929 PR target/49880
22930 * config/sh/sh.opt (FPU_SINGLE_ONLY): New mask.
22931 (musermode): Convert to Var(TARGET_USERMODE).
22932 * config/sh/sh.h (SELECT_SH2A_SINGLE_ONLY, SELECT_SH4_SINGLE_ONLY,
22933 MASK_ARCH): Add MASK_FPU_SINGLE_ONLY.
22934 * config/sh/sh.c (sh_option_override): Use
22935 TARGET_FPU_DOUBLE || TARGET_FPU_SINGLE_ONLY for call-fp case.
22936 * config/sh/sh.md (udivsi3_i1, divsi3_i1): Remove ! TARGET_SH4
22937 condition.
22938 (udivsi3_i4, divsi3_i4): Use TARGET_FPU_DOUBLE condition instead of
22939 TARGET_SH4.
22940 (udivsi3_i4_single, divsi3_i4_single): Use
22941 TARGET_FPU_SINGLE_ONLY || TARGET_FPU_DOUBLE instead of TARGET_HARD_SH4.
22942
f4b719c7
DK		 229432013-03-13 Dave Korn <dave.korn.cygwin@....>
22944
22945 * config/i386/cygwin.h (SHARED_LIBGCC_SPEC): Make shared libgcc the
22946 default setting.
22947
c40eced0
RB		 229482013-03-13 Richard Biener <rguenther@suse.de>
22949
22950 PR tree-optimization/56608
22951 * tree-vect-slp.c (vect_schedule_slp): Do not remove scalar
22952 calls when vectorizing basic-blocks.
22953
1bfa5973
JJ		 229542013-03-13 Jakub Jelinek <jakub@redhat.com>
22955
22956 PR plugins/45078
22957 * config.gcc: On arm, mips, sh and sparc add vxworks-dummy.h to
22958 tm_file.
22959
6fcf5434
JJ		 229602013-03-12 Jakub Jelinek <jakub@redhat.com>
22961
22962 * doc/invoke.texi (-Waddr-space-convert): Move into the table earlier.
22963
4f38fa8c
JH		 229642013-03-11 Jan Hubicka <jh@suse.cz>
22965
22966 PR lto/56557
22967 * lto-streamer-out.c (output_symbol_p): Skip references from
22968 constructors of external variables.
22969
c5c5ba89
JH		 229702013-03-11 Jan Hubicka <jh@suse.cz>
22971
22972 PR middle-end/56571
22973 * valtrack.c (cleanup_auto_inc_dec): Unshare clobbers originating
22974 from pseudos.
22975 * emit-rtl.c (verify_rtx_sharing): Likewise.
22976 (copy_insn_1): Likewise.
22977 * rtl.c (copy_rtx): Likewise.
22978
c2a939b1
GJL		 229792013-03-11 Georg-Johann Lay <avr@gjlay.de>
22980
22981 PR target/56591
22982 * config/avr/avr.c (avr_print_operand): Add space after '%c' in
22983 output_operand_lossage message.
88b97037 22984
3c3279fb
RE		 229852013-03-11 Richard Earnshaw <rearnsha@arm.com>
22986
22987 PR target/56470
22988 * arm.c (shift_op): Validate RTL pattern on the fly.
22989 (arm_print_operand, case 'S'): Don't use shift_operator to validate
22990 the RTL.
22991
aef5ef9d
JDA		 229922013-03-10 John David Anglin <dave.anglin@nrc-cnrc.gc.ca>
22993
22994 PR target/56347
22995 * config/pa/pa.md (call_value): Check for calls to powf and direct to
22996 new call patterns that clobber %fr12.
22997 (call_val_powf, call_val_powf_pic, call_val_powf_64bit): New insn,
22998 split and postreload patterns.
22999 * config/pa/pa.c (pa_conditional_register_usage): Revert marking
23000 registers %fr12 and %fr12R as call used.
23001
456610d3
SB		 230022013-03-09 Steven Bosscher <steven@gcc.gnu.org>
23003
23004 * dse.c (delete_dead_store_insn): Respect TDF_DETAILS.
23005 (canon_address, record_store, replace_read, check_mem_read_rtx,
23006 scan_insn, dse_step1, dse_step2_init, dse_step2_spill,
23007 dse_step4, dse_step5_nospill, dse_step5_spill, dse_step6,
23008 rest_of_handle_dse): Likewise.
23009
4b1baac8
RS		 230102013-03-09 Richard Sandiford <rdsandiford@googlemail.com>
23011
23012 PR middle-end/56524
23013 * tree.h (tree_optimization_option): Rename target_optabs to optabs.
23014 Add base_optabs.
23015 (TREE_OPTIMIZATION_OPTABS): Update after previous field change.
23016 (TREE_OPTIMIZATION_BASE_OPTABS): New macro.
23017 (save_optabs_if_changed): Replace with...
23018 (init_tree_optimization_optabs): ...this.
23019 * optabs.c (save_optabs_if_changed): Rename to...
23020 (init_tree_optimization_optabs): ...this. Take the optimization node
23021 as argument. Do nothing if the base optabs are already correct.
23022 Reuse the existing TREE_OPTIMIZATION_OPTABS memory if we need
23023 to recompute optabs.
23024 * function.h (function): Remove optabs field.
23025 * function.c (invoke_set_current_function_hook): Call
23026 init_tree_optimization_optabs. Use the result to initialize
23027 this_fn_optabs.
23028
b7a78683
AH		 230292013-02-27 Aldy Hernandez <aldyh@redhat.com>
23030
23031 * trans-mem.c (expand_transaction): Do not set PR_INSTRUMENTEDCODE
23032 if GTMA_HAS_NO_INSTRUMENTATION.
23033 (generate_tm_state): Keep GTMA_HAS_NO_INSTRUMENTATION bit.
23034 (ipa_tm_transform_transaction): Set GTMA_HAS_NO_INSTRUMENTATION.
23035 * gimple.h (GTMA_HAS_NO_INSTRUMENTATION): Define.
23036 * gimple-pretty-print.c (dump_gimple_transaction): Handle
23037 GTMA_HAS_NO_INSTRUMENTATION.
23038
6384c29b
JJ		 230392013-03-08 Jakub Jelinek <jakub@redhat.com>
23040
23041 * config/gnu-user.h (LIBTSAN_EARLY_SPEC): Don't link against
23042 libasan_preinit.o.
23043
ecd4f20a
MP		 230442013-03-08 Marek Polacek <polacek@redhat.com>
23045 Jakub Jelinek <jakub@redhat.com>
23046
23047 PR tree-optimization/56478
23048 * predict.c (is_comparison_with_loop_invariant_p): Change the
23049 type of loop_step to tree.
23050 (predict_loops): Adjust.
23051 (predict_iv_comparison): Perform the computations on double_ints.
23052
64366d35
RB		 230532013-03-08 Richard Biener <rguenther@suse.de>
23054
23055 PR tree-optimization/56570
23056 * tree-cfg.c (verify_expr_location_1): Verify locations for
23057 DECL_DEBUG_EXPR.
23058 * tree-sra.c (create_access_replacement): Strip locations
23059 from DECL_DEBUG_EXPRs.
23060
a9d5a059
RB		 230612013-03-08 Richard Biener <rguenther@suse.de>
23062
23063 * tree-inline.c (expand_call_inline): Do not associate
23064 a BLOCK with the location in BLOCK_SOURCE_LOCATION.
23065 * tree-cfg.c (verify_location): Verify BLOCK_SOURCE_LOCATION.
23066
b7aa4e9a
RB		 230672013-03-08 Richard Biener <rguenther@suse.de>
23068
23069 * tree-ssa-ter.c (is_replaceable_p): Do not TER across location
23070 or block changes with -Og. Fix for location / block encoding
23071 changes and PHI arguments with locations.
23072
c4c2f9fa
SB		 230732013-03-07 Steven Bosscher <steven@gcc.gnu.org>
23074
23075 * bitmap.c (struct bitmap_descriptor_d): Use unsigned HOST_WIDEST_INT
23076 for all counters.
23077 (struct output_info): Likewise.
23078 (register_overhead): Remove bad gcc_assert.
23079 (bitmap_find_bit): If there is only a single bitmap element, do not
23080 count a miss as a search.
23081 (print_statistics): Update for counter type changes.
23082 (dump_bitmap_statistics): Likewise. Print headers such that they
23083 are properly lined up with the printed counters.
23084
5bf6606a
JJ		 230852013-03-07 Jakub Jelinek <jakub@redhat.com>
23086
23087 PR tree-optimization/56559
23088 * tree-ssa-reassoc.c (zero_one_operation): When looking at rhs2,
23089 check that it has only a single use.
23090
2c653d46
RB		 230912013-03-07 Richard Biener <rguenther@suse.de>
23092
23093 * doc/invoke.texi (fwhole-program): Discourage use in combination
23094 with -flto.
23095
a72d8780
JJ		 230962013-03-06 Jakub Jelinek <jakub@redhat.com>
23097
01a454df
JJ		 23098 * config/arm/t-arm (TM_H, OPTIONS_H_EXTRA): Add arm-cores.def.
23099
c1781047
JJ		 23100 PR tree-optimization/56539
23101 * tree-tailcall.c (adjust_return_value_with_ops): Use GSI_SAME_STMT
23102 instead of GSI_CONTINUE_LINKING as last argument to
23103 force_gimple_operand_gsi. Adjust function comment.
23104
9772c47a
JJ		 23105 * config/aarch64/t-aarch64 (TM_H, OPTIONS_H_EXTRA): Add
23106 aarch64-cores.def.
23107
a72d8780
JJ		 23108 PR middle-end/56548
23109 * expr.c (expand_cond_expr_using_cmove): When expanding cmove in
23110 promoted mode, convert the result back to the original mode.
23111
fa5556de
RB		 231122013-03-06 Richard Biener <rguenther@suse.de>
23113
23114 PR middle-end/56294
23115 * tree-into-ssa.c (insert_phi_nodes_for): Add dumping.
23116 (insert_updated_phi_nodes_compare_uids): New function.
23117 (update_ssa): Sort symbols_to_rename after UID before
23118 traversing it to insert PHI nodes.
23119
010403d1
RB		 231202013-03-06 Richard Biener <rguenther@suse.de>
23121
23122 PR middle-end/50494
23123 * tree-vect-data-refs.c (vect_can_force_dr_alignment_p):
23124 Do not adjust alignment of DECL_IN_CONSTANT_POOL decls.
23125
23126 Revert
23127 2013-02-13 Richard Biener <rguenther@suse.de>
23128
23129 PR lto/50494
23130 * varasm.c (output_constant_def_1): Get the decl representing
23131 the constant as argument.
23132 (output_constant_def): Wrap output_constant_def_1.
23133 (make_decl_rtl): Use output_constant_def_1 with the decl
23134 representing the constant.
23135 (build_constant_desc): Optionally re-use a decl already
23136 representing the constant.
23137 (tree_output_constant_def): Adjust.
23138
3c27ce4c
JY		 231392013-03-06 Joey Ye <joey.ye@arm.com>
23140
23141 PR lto/50293
23142 * gcc.c (convert_white_space): New function.
23143 (main): Handles white space in function name.
23144
8f6d1c86
OE		 231452013-03-06 Oleg Endo <olegendo@gcc.gnu.org>
23146
23147 PR target/56529
23148 * config/sh/sh.c (sh_option_override): Check for TARGET_DYNSHIFT
23149 instead of TARGET_SH2 for call-table case. Do not set sh_div_strategy
23150 to SH_DIV_CALL_TABLE for TARGET_SH2.
23151 * config.gcc (sh_multilibs): Add m2 and m2a to sh*-*-linux* multilib
23152 list.
23153 * doc/invoke.texi (SH options): Document mdiv= call-div1, call-fp,
23154 call-table options.
23155
798a209f
SA		 231562013-03-05 Sterling Augustine <saugustine@google.com>
23157 Cary Coutant <ccoutant@google.com>
23158
23159 PR debug/55364
23160 * dwarf2out.c (resolve_addr): Don't call
23161 remove_loc_list_addr_table_entries a second time for the same
23162 expression.
23163
6cfa417f 231642013-03-05 Jakub Jelinek <jakub@redhat.com>
ee61ea38 23165
6cfa417f
JJ		 23166 PR debug/56510
23167 * cfgexpand.c (expand_debug_parm_decl): Call copy_rtx on incoming.
23168 (avoid_complex_debug_insns): New function.
23169 (expand_debug_locations): Call it.
23170
23171 PR rtl-optimization/56484
23172 * ifcvt.c (noce_process_if_block): If else_bb is NULL, avoid extending
23173 lifetimes of hard registers on small register class machines.
23174
231752013-03-05 David Holsgrove <david.holsgrove@xilinx.com>
23176
23177 * config/microblaze/microblaze-protos.h: Rename
ee61ea38 23178 microblaze_is_interrupt_handler to microblaze_is_interrupt_variant.
6cfa417f 23179 * config/microblaze/microblaze.c (microblaze_attribute_table): Add
ee61ea38
DH		 23180 fast_interrupt.
23181 (microblaze_fast_interrupt_function_p): New function.
23182 (microblaze_is_interrupt_handler): Rename to
23183 microblaze_is_interrupt_variant and add fast_interrupt check.
23184 (microblaze_must_save_register): Use microblaze_is_interrupt_variant.
23185 (save_restore_insns): Likewise.
23186 (compute_frame_size): Likewise.
23187 (microblaze_function_prologue): Add FAST_INTERRUPT_NAME.
23188 (microblaze_globalize_label): Likewise.
6cfa417f
JJ		 23189 * config/microblaze/microblaze.h: Define FAST_INTERRUPT_NAME.
23190 * config/microblaze/microblaze.md: Use wrapper
ee61ea38
DH		 23191 microblaze_is_interrupt_variant.
23192
731aefac
KT		 231932013-03-05 Kai Tietz <ktietz@redhat.com>
23194
23195 * sdbout.c (sdbout_one_type): Switch to current function's section
23196 supporting cold/hot.
23197
a72d8780 231982013-03-05 David Holsgrove <david.holsgrove@xilinx.com>
e34fd5a7
DH		 23199
23200 * doc/invoke.texi (MicroBlaze): Add -mbig-endian, -mlittle-endian,
23201 -mxl-reorder.
23202
334e71e8
JJ		 232032013-03-05 Jakub Jelinek <jakub@redhat.com>
23204
0b50e654
JJ		 23205 PR middle-end/56461
23206 * ggc-common.c (gt_pch_save): For ENABLE_VALGRIND_CHECKING,
23207 if VALGRIND_GET_VBITS is defined, temporarily make object
23208 memory all defined, and restore previous valgrind addressability
23209 and definability afterwards. Free this_object at the end.
23210
4ccf8f43
JJ		 23211 PR middle-end/56461
23212 * lra.c (lra): Call lra_clear_live_ranges if live_p,
23213 right before calling lra_create_live_ranges, also call it
23214 when clearing live_p. Only call lra_clear_live_ranges
23215 at the end if live_p.
23216
334e71e8
JJ		 23217 PR middle-end/56461
23218 * sched-deps.c (delete_dep_node): Free DEP_REPLACE.
23219
9ca966ca
RB		 232202013-03-05 Richard Biener <rguenther@suse.de>
23221
23222 PR tree-optimization/56521
23223 * tree-ssa-sccvn.c (set_value_id_for_result): Always initialize
23224 value-id.
23225
d273b176
SB		 232262013-03-05 Steven Bosscher <steven@gcc.gnu.org>
23227
23228 PR c++/55135
23229 * except.h (remove_unreachable_eh_regions): New prototype.
23230 * except.c (remove_eh_handler_splicer): New function, split out
23231 of remove_eh_handler.
23232 (remove_eh_handler): Use remove_eh_handler_splicer. Add comment
23233 warning about running it on many EH regions one at a time.
23234 (remove_unreachable_eh_regions_worker): New function, walk the
23235 EH tree in depth-first order and remove non-marked regions.
23236 (remove_unreachable_eh_regions): New function.
23237 * tree-eh.c (mark_reachable_handlers): New function, split out
23238 from remove_unreachable_handlers.
23239 (remove_unreachable_handlers): Use mark_reachable_handlers and
23240 remove_unreachable_eh_regions.
23241 (remove_unreachable_handlers_no_lp): Use mark_reachable_handlers
23242 and remove_unreachable_eh_regions.
23243
5e7f1aef
RB		 232442013-03-05 Richard Biener <rguenther@suse.de>
23245
23246 PR middle-end/56525
23247 * loop-init.c (fix_loop_structure): Remove loops in two stages,
23248 not freeing them until the end.
23249
f276b762
AK		 232502013-03-05 Andreas Krebbel <Andreas.Krebbel@de.ibm.com>
23251
23252 * config/s390/s390.h: Define DWARF2_ASM_LINE_DEBUG_INFO.
23253
12a43ab8
RB		 232542013-03-05 Richard Biener <rguenther@suse.de>
23255
23256 PR tree-optimization/56270
23257 * tree-vect-slp.c (vect_schedule_slp): Clear vectorized stmts
23258 of loads after scheduling an SLP instance.
23259
93675444
JJ		 232602013-03-05 Jakub Jelinek <jakub@redhat.com>
23261
db4138e3
JJ		 23262 * Makefile.in (dg_target_exps): Add aarch64.exp, epiphany.exp and
23263 tic6x.exp.
23264 (check_gcc_parallelize): Run guality.exp as a separate job from
23265 vect.exp with unsorted.exp and $(dg_target_exps) separately from
23266 struct-layout-1.exp with stackalign.exp.
23267
dd3d1ec0
JJ		 23268 * alias.c (init_alias_analysis): Clear reg_known_equiv_p bitmap.
23269
b8d381a3
JJ		 23270 PR middle-end/56461
23271 * tree-vect-slp.c (vect_supported_load_permutation_p): Free
23272 load_index sbitmap even if some bit in it isn't set.
23273
b4f9786b
JJ		 23274 PR middle-end/56461
23275 * tree-ssa-loop-niter.c (bb_queue): Remove typedef.
23276 (discover_iteration_bound_by_body_walk): Change queues to
23277 vec<vec<basic_block> > and queue to vec<basic_block>. Fix up
23278 spelling in comment. Call safe_push on queues[bound_index] directly.
23279 Release queues[queue_index] in every iteration unconditionally.
23280 Release bounds vector.
23281
93675444
JJ		 23282 PR middle-end/56461
23283 * tree-vect-stmts.c (free_stmt_vec_info_vec): Call
23284 free_stmt_vec_info on any left-over stmt_vec_info in the vector.
23285 * tree-vect-loop.c (vect_create_epilog_for_reduction): Release
23286 inner_phis vector.
23287
3e492e9c
RB		 232882013-03-05 Richard Biener <rguenther@suse.de>
23289
23290 PR lto/56515
23291 * tree-inline.c (remap_blocks_to_null): New function.
23292 (expand_call_inline): When expanding a call stmt without
23293 an associated block inline remap all callee blocks to NULL.
23294
a5d1569a
JJ		 232952013-03-05 Jakub Jelinek <jakub@redhat.com>
23296
86efb5cd
JJ		 23297 PR rtl-optimization/56494
23298 * simplify-rtx.c (simplify_truncation): If C is narrower than A,
23299 optimize (truncate:A (subreg:B (truncate:C X) 0)) into
23300 (subreg:A (truncate:C X) 0) instead of (truncate:A X).
23301
85f5dbea
JJ		 23302 PR middle-end/56461
23303 * sel-sched-ir.c (free_sched_pools): Release
23304 succs_info_pool.stack[succs_info_pool.max_top] vectors too
23305 if succs_info_pool.max_top isn't -1.
23306
a5d1569a
JJ		 23307 PR bootstrap/56509
23308 * opts.c (opts_obstack, opts_concat): Moved to...
23309 * opts-common.c (opts_obstack, opts_concat): ... here.
23310
4432aa6c
JJ		 233112013-03-04 Jakub Jelinek <jakub@redhat.com>
23312
23313 PR middle-end/56461
23314 * diagnostic.c (diagnostic_append_note): Save and restore old prefix.
23315
f3a1fb91
MJ		 233162013-03-04 Martin Jambor <mjambor@suse.cz>
23317
23318 * tree-dfa.c (get_or_create_ssa_default_def): Use parameter fn in
23319 all appropriate places.
23320
22110e6c
EB		 233212013-01-04 Eric Botcazou <ebotcazou@adacore.com>
23322
23323 PR tree-optimization/56424
23324 * ipa-split.c (split_function): Do not set the RSO flag if result is
23325 not by reference and its type is a register type.
23326
a72d8780 233272013-03-04 David Holsgrove <david.holsgrove@xilinx.com>
91c4e421 23328
88b97037 23329 * config/microblaze/microblaze.c (microblaze_valid_pic_const): New
91c4e421 23330 (microblaze_legitimate_pic_operand): Likewise
88b97037 23331 * config/microblaze/microblaze.h (LEGITIMATE_PIC_OPERAND_P): calls
91c4e421
DH		 23332 new function microblaze_legitimate_pic_operand
23333 * config/microblaze/microblaze-protos.h
23334 (microblaze_legitimate_pic_operand): Declare.
23335
a72d8780 233362013-03-04 Edgar E. Iglesias <edgar.iglesias@gmail.com>
82f8f53a 23337
a72d8780 23338 * config/microblaze/predicates.md (call_insn_simple_operand):
82f8f53a 23339 New predicate for supported rtx code types.
a72d8780 23340 * config/microblaze/microblaze.md (call_internal1): Use
82f8f53a
EI		 23341 call_insn_simple_operand predicate.
23342
541d9ac8
JJ		 233432013-03-04 Jakub Jelinek <jakub@redhat.com>
23344
5eb010bc
JJ		 23345 PR middle-end/56461
23346 * tree-loop-distribution.c (ldist_gen): Call partition_free after each
23347 partitions.ordered_remove.
23348
30862efc
JJ		 23349 PR middle-end/56461
23350 * tree-vect-stmts.c (vectorizable_conversion): Don't call
23351 vec_oprnds0.create (1) for modifier == NONE.
23352
8930f723
JJ		 23353 PR middle-end/56461
23354 * tree-vect-stmts.c (vectorizable_shift): Don't call create methods
23355 on vec_oprnds0 or vec_oprnds1 before loop, only call it on
23356 vec_oprnds1 right before pushing anything to it for
23357 scalar_shift_arg.
23358
541d9ac8
JJ		 23359 PR middle-end/56461
23360 * tree-vect-loop.c (destroy_loop_vec_info): For !clean_stmts, just
23361 set nbbs to 0 instead of having separate code path.
23362 (vect_analyze_loop_form): Call destroy_loop_vec_info with true
23363 instead of false as last argument if returning NULL.
23364
7aa7f2e3
SL		 233652013-03-03 Sandra Loosemore <sandra@codesourcery.com>
23366
23367 * target.def (TARGET_OPTION_VALID_ATTRIBUTE_P): Update comments;
23368 the attribute is now called "target" instead of "option".
23369 (TARGET_OPTION_PRAGMA_PARSE): Likewise, for the pragma.
23370 * doc/tm.texi.in (Target Attributes): Likewise document the correct
23371 attribute/pragma name for TARGET_OPTION_VALID_P and
23372 TARGET_OPTION_PRAGMA_PARSE. Also copy-edit and correct markup.
23373 * doc/tm.texi: Regenerated.
23374
8930f723 233752013-03-02 David Holsgrove <david.holsgrove@xilinx.com>
cb8a1637 23376
8930f723 23377 * config/microblaze/microblaze.c:
cb8a1637 23378 Check mcpu, pcmp requirement and set TARGET_REORDER to 0 if not met.
8930f723
JJ		 23379 * config/microblaze/microblaze.h: Add -mxl-reorder to
23380 DRIVER_SELF_SPECS.
23381 * config/microblaze/microblaze.md: New bswapsi2 and bswaphi2.
23382 instructions emitted if TARGET_REORDER.
88b97037
UB		 23383 * config/microblaze/microblaze.opt: New option -mxl-reorder set to 1
23384 or 0 for -m/-mno case, but initialises as 2 to detect default use case
8930f723 23385 separately.
cb8a1637 23386
4dc1d68c
XDL		 233872013-03-01 Xinliang David Li <davidxl@google.com>
23388
23389 * tree-ssa-uninit.c (compute_control_dep_chain): Limit post-dom
23390 walk length.
23391
689eaba3
JJ		 233922013-03-01 Jakub Jelinek <jakub@redhat.com>
23393
9d676bf6
JJ		 23394 PR middle-end/56461
23395 * tree-ssa-loop-ivcanon.c (tree_estimate_loop_size): Release path
23396 vector even when returning true. Fix up function comment formatting.
23397
caff7edf
JJ		 23398 PR middle-end/56461
23399 * ira-build.c (ira_loop_nodes_count): New variable.
23400 (create_loop_tree_nodes): Initialize it.
23401 (finish_loop_tree_nodes): Use it instead of looking at current_loops.
23402
b6b9227d
JJ		 23403 PR middle-end/56461
23404 * tree-vect-data-refs.c (vect_permute_store_chain): Avoid using copy
23405 method on dr_chain and result_chain.
23406 * tree-vect-stmts.c (vectorizable_store): Only call
23407 result_chain.create if j == 0.
23408
689eaba3
JJ		 23409 PR middle-end/56461
23410 * tree-vect-stmts.c (vect_create_vectorized_promotion_stmts): Call
23411 vec_oprnds0->release (); rather than vec_oprnds0->truncate (0)
23412 before overwriting it.
23413
40bf31ed
TB		 234142013-03-01 Tobias Burnus <burnus@net-b.de>
23415
23416 * doc/extended.texi (C Extensions): Change order in @menu
23417 to match @node.
23418 (Other MIPS Built-in Functions): Move last MIPS entry before
23419 "picoChip Built-in Functions".
23420 (SH Built-in Functions): Move after RX Built-in Functions.
23421 * doc/gcc.texi (Introduction): Change order in @menu
23422 to match @node.
23423 * doc/md.texi (Constraints): Ditto.
23424 * gty.texi (Type Information): Ditto.
23425 (User-provided marking routines for template types): Make
23426 subsection.
23427 * doc/invoke.texi (AArch64 Options): Move before
23428 "Adapteva Epiphany Options".
23429
e664c61c
KS		 234302013-02-28 Konstantin Serebryany <konstantin.s.serebryany@gmail.com>
23431 Jakub Jelinek <jakub@redhat.com>
23432
23433 PR sanitizer/56454
23434 * asan.c (gate_asan): Lookup no_sanitize_address instead of
23435 no_address_safety_analysis attribute.
23436 * doc/extend.texi (no_address_safety_attribute): Rename to
23437 no_sanitize_address attribute, mention no_address_safety_analysis
23438 attribute as deprecated alias.
23439
37b5ec8f
JJ		 234402013-02-28 Jakub Jelinek <jakub@redhat.com>
23441
23442 PR middle-end/56461
23443 * tree-vectorizer.h (vect_get_slp_defs): Change 3rd argument
23444 type to vec<vec<tree> > *.
23445 * tree-vect-slp.c (vect_get_slp_defs): Likewise. Change vec_defs
23446 to be vec<tree> instead of vec<tree> *, set vec_defs
23447 to vNULL and call vec_defs.create (number_of_vects), adjust other
23448 uses of vec_defs.
23449 * tree-vect-stmts.c (vect_get_vec_defs, vectorizable_call,
23450 vectorizable_condition): Adjust vect_get_slp_defs callers.
23451
ba96cdfb
JG		 234522013-02-28 James Greenhalgh <james.greenhalgh@arm.com>
23453
23454 * config/aarch64/aarch64.c
23455 (aarch64_float_const_representable): Remove unused variable.
23456
6f549691
JG		 234572013-02-28 James Greenhalgh <james.greenhalgh@arm.com>
23458
23459 * config/aarch64/aarch64.c (aarch64_mangle_type): Make static.
23460
af55e82d
JG		 234612013-02-28 James Greenhalgh <james.greenhalgh@arm.com>
23462
23463 * config/aarch64/aarch64-builtins.c
23464 (aarch64_init_simd_builtins): Make static.
23465
1df3f464
JG		 234662013-02-28 James Greenhalgh <james.greenhalgh@arm.com>
23467
23468 * config/aarch64/aarch64.c
23469 (aarch64_simd_make_constant): Make static.
23470
f8f42513
MJ		 234712013-02-28 Martin Jambor <mjambor@suse.cz>
23472
23473 * tree-sra.c (load_assign_lhs_subreplacements): Do not put replacements
23474 with no initialization to the RHS of debug statements.
23475
b48b3fc4
MJ		 234762013-02-28 Martin Jambor <mjambor@suse.cz>
23477
23478 PR tree-optimization/56294
23479 * tree-sra.c (analyze_access_subtree): Create replacement declarations.
23480 Adjust dumping.
23481 (get_access_replacement): Do not call create_access_replacement.
23482 Assert a replacement exists.
23483 (get_repl_default_def_ssa_name): Create the replacement declaration
23484 itself.
23485
c3ae224c
RR		 234862013-02-28 Ramana Radhakrishnan <ramana.radhakrishnan@arm.com>
23487
23488 * config/arm/arm.c (arm_output_mi_thunk): Call final_start_function and
23489 final_end_function.
23490
45fa0eef
MP		 234912013-02-28 Marek Polacek <polacek@redhat.com>
23492
23493 PR rtl-optimization/56466
23494 * loop-unroll.c (unroll_and_peel_loops): Call fix_loop_structure
23495 if we're changing a loop.
23496 (peel_loops_completely): Likewise.
23497
502c067d
PC		 234982013-02-28 Paolo Carlini <paolo.carlini@oracle.com>
23499
23500 PR c++/55813
23501 * doc/invoke.texi ([-Wctor-dtor-privacy]): Complete.
23502
ff24afc8
GJL		 235032013-02-28 Georg-Johann Lay <avr@gjlay.de>
23504
23505 PR target/56445
23506 * config/avr/avr.c (avr_init_builtins): Use 'n' instead of empty
23507 macro parameters with: FX_FTYPE_FX, FX_FTYPE_FX_INT, INT_FTYPE_FX,
23508 INTX_FTYPE_FX, FX_FTYPE_INTX.
23509 * config/avr/builtins.def: Adjust respective DEF_BUILTIN.
23510
c7b1fc1b
GJL		 235112013-02-28 Georg-Johann Lay <avr@gjlay.de>
23512
23513 * avr/avr-mcus.def (ata5272, ata5505, attiny1634, ata6285)
23514 (ata6286, atmega8a, atmega48pa, ata5790, ata5790n, ata5795)
23515 (atmega164pa, atmega165pa, atmega168pa, atmega16hva, atmega16hvb)
23516 (atmega16hvbrevb, atmega16m1, atmega16u4, atmega26hvg, atmega32a)
23517 (atmega32a, atmega3250pa, atmega3290pa, atmega32c1, atmega32m1)
23518 (atmega32u4, atmega32u6, atmega64a, atmega6490a, atmega6490p)
23519 (atmega64c1, atmega64m1, atmega64rfa2, atmega64rfr2, atmega32hvb)
23520 (atmega32hvbrevb, atmega16hva2, atmega48hvf, at90pwm161)
23521 (atmega128a, atmega1284, atmxt112sl, atmxt224, atmxt224e)
23522 (atmxt336s, atxmega16a4u, atxmega16c4, atxmega32a4u, atxmega32c4)
23523 (atxmega32e5, atxmega64a3u, atxmega64a4u, atxmega64b1, atxmega64b3)
23524 (atxmega64c3, atxmega64d4, atxmega128a3u, atxmega128b1)
23525 (atxmega128b3, atxmega128c3, atxmega128d4, atmxt540s, atmxt540sreva)
23526 (atxmega192a3u, atxmega192c3, atxmega256a3u, atxmega256c3)
23527 (atxmega384c3, atxmega384d3, atxmega128a4u): New AVR_MCU.
23528 (avrxmega6): Increase max flash segments from 5 to 6.
23529 * config/avr/t-multilib: Regenerate.
23530 * config/avr/avr-tables.opt: Regenerate.
23531 * doc/avr-mmcu.texi: Regenerate.
23532
4a0e3cfe
GJL		 235332013-02-28 Georg-Johann Lay <avr@gjlay.de>
23534
23535 * config/avr/avr.h (device_to_arch): Rename to device_to_ld.
23536 (avr_device_to_arch): Rename to avr_device_to_ld.
23537 (avr_device_to_as): New prototype.
23538 (EXTRA_SPEC_FUNCTIONS): Add device_to_as.
23539 (ASM_SPEC): Use device_to_as to get -mmcu= and -mno-skip-bug=.
23540 * config/avr/driver-avr.c (avr_device_to_as): New.
23541 (avr_device_to_arch): Rename to avr_device_to_ld.
23542
97785e52
JJ		 235432013-02-27 Jakub Jelinek <jakub@redhat.com>
23544
3f292312
JJ		 23545 PR middle-end/56461
23546 * tree-vect-data-refs.c (vect_permute_load_chain): Avoid using copy
23547 method on dr_chain and result_chain.
23548
a344216b
JJ		 23549 PR middle-end/56461
23550 * tree-ssa-loop-niter.c (maybe_lower_iteration_bound): Call
23551 pointer_set_destroy on not_executed_last_iteration.
23552
f121ad02 23553 PR middle-end/56461
88b97037 23554 * tree-vect-loop.c (vectorizable_reduction): Release vect_defs vector.
f121ad02 23555
307f83a3
JJ		 23556 PR middle-end/56461
23557 * ipa-pure-const.c (propagate): Use FOR_EACH_FUNCTION instead of
23558 FOR_EACH_DEFINED_FUNCTION when freeing state.
23559
e19624ee
JJ		 23560 PR middle-end/56461
23561 * df-scan.c (df_insn_delete): Use df_scan_free_mws_vec before
23562 pool_free.
23563 (df_insn_rescan_debug_internal): Use df_scan_free_mws_vec before
23564 overwriting it.
23565
90e709fd
JJ		 23566 PR middle-end/56461
23567 * ipa-cp.c (decide_whether_version_node): Call vec_free on
23568 known_aggs[i].items and release known_aggs vector.
23569
97785e52
JJ		 23570 PR middle-end/56461
23571 * ipa-reference.c (propagate): Free node_info even for alias nodes.
23572
235732013-02-27 Edgar E. Iglesias <edgar.iglesias@gmail.com>
bbb3a9e2 23574
97785e52
JJ		 23575 * config/microblaze/microblaze.c (microblaze_emit_compare):
23576 Use xor for EQ/NE comparisions.
23577 * config/microblaze/microblaze.md (cstoresf4): Add constraints
23578 (cbranchsf4): Adjust operator to comparison_operator.
bbb3a9e2 23579
6fa5e0ed
JJ		 235802013-02-27 Jakub Jelinek <jakub@redhat.com>
23581
23582 PR middle-end/56461
23583 * tree-flow.h (edge_var_map_vector): Change into va_heap, vl_embed
23584 vector.
23585 * tree-ssa.c (redirect_edge_var_map_add): Use vec_safe_reserve and
23586 vec_safe_push, always update *slot.
23587 (redirect_edge_var_map_clear): Use vec_free.
23588 (redirect_edge_var_map_dup): Use vec_safe_copy and vec_safe_reserve.
23589 (free_var_map_entry): Use vec_free.
23590 * tree-cfgcleanup.c (remove_forwarder_block_with_phi): Use
23591 FOR_EACH_VEC_SAFE_ELT instead of FOR_EACH_VEC_ELT.
23592
436a956a
AB		 235932013-02-27 Andrey Belevantsev <abel@ispras.ru>
23594
23595 PR middle-end/45472
436a956a
AB		 23596 * sel-sched-ir.c (merge_expr): Also change vinsn of merged expr
23597 when the may_trap_p bit of the exprs being merged differs.
436a956a
AB		 23598 Reorder tests for speculativeness in the logical and operator.
23599
f5c2caca
JJ		 236002013-02-27 Jakub Jelinek <jakub@redhat.com>
23601
0fcb564b
JJ		 23602 * incpath.c (add_standard_paths): Use reconcat instead of concat
23603 where appropriate and avoid leaking memory.
23604
dc357798
JJ		 23605 * opts.h: Include obstack.h.
23606 (opts_concat): New prototype.
23607 (opts_obstack): New declaration.
23608 * opts.c (opts_concat): New function.
23609 (opts_obstack): New variable.
23610 (init_options_struct): Call gcc_init_obstack on opts_obstack.
23611 (finish_options): Use opts_concat instead of concat
23612 and XOBNEWVEC instead of XNEWVEC.
23613 * opts-common.c (generate_canonical_option, decode_cmdline_option,
23614 generate_option): Likewise.
23615 * Makefile.in (OPTS_H): Depend on $(OBSTACK_H).
23616 * lto-wrapper.c (main): Call gcc_init_obstack on opts_obstack.
23617
f5c2caca
JJ		 23618 PR target/56455
23619 * stmt.c (expand_switch_as_decision_tree_p): If flag_pic
23620 and ASM_OUTPUT_ADDR_DIFF_ELT isn't defined, return true.
23621
d0163673
JJ		 236222013-02-26 Jakub Jelinek <jakub@redhat.com>
23623
23624 PR middle-end/56461
23625 * lra-spills.c (lra_spill): Free spill_hard_reg at the end.
23626
b5ad2b8e
JR		 236272013-02-26 Joern Rennecke <joern.rennecke@embecosm.com>
23628
23629 * config/arm/arm.c (const_ok_for_dimode_op): Back out last change.
23630 (arm_block_move_unaligned_straight): Likewise.
23631 (arm_adjust_block_mem): Likewise.
23632
9b639e2c
JR		 236332013-02-26 Joern Rennecke <joern.rennecke@embecosm.com>
23634
23635 PR target/48901
23636 * config/lm32/lm32.c (gen_int_relational): Remove unused variables
23637 temp, cond and label.
23638 * config/lm32/lm32.md (ashlsi3): Remove unused variable one.
23639
e13a0ccb
JR		 23640 PR target/52500
23641 * config/c6x/c6x.c (dbx_register_map): Change to unsigned.
23642 * config/c6x/c6x.h (dbx_register_map): Update declaration.
23643
fbe4f171
JR		 23644 PR target/52501
23645 * config/cr16/cr16-protos.h: Move end of RTX_CODE guard below end
23646 of prologue/epilogue functions.
23647
ae006543
JR		 23648 PR target/52550
23649 * config/tilegx/tilegx.c (tilegx_expand_prologue):
23650 Remove unused variable cfa_offset.
23651 * config/tilepro/tilepro.c (tilepro_expand_prologue): Likewise.
23652
c81369fa
JR		 23653 PR target/54639
23654 * config/mn10300/mn10300.c (mn10300_expand_epilogue): Avoid offset
23655 type promotion to unsigned.
23656
f8a8fea7
JR		 23657 PR target/54640
23658 * config/arm/arm.c (const_ok_for_dimode_op): Make code consistent
23659 for HOST_WIDE_INT of 32 bit / same size as int.
23660 (arm_block_move_unaligned_straight): Likewise.
23661 (arm_adjust_block_mem): Likewise.
23662
f8be5169
JR		 23663 PR target/54662
23664 * config/mep/t-mep (mep-pragma.o): Use ALL_COMPILERFLAGS instead of
23665 ALL_CFLAGS.
23666
f1ad3354
MP		 236672013-02-26 Marek Polacek <polacek@redhat.com>
23668
23669 PR tree-optimization/56426
88b97037 23670 * tree-ssa-loop.c (tree_ssa_loop_init): Always call scev_initialize.
f1ad3354 23671
a6af0f42
RB		 236722013-02-26 Richard Biener <rguenther@suse.de>
23673
23674 PR target/56444
23675 * config/mn10300/mn10300.c (mn10300_scan_for_setlb_lcc): Remove
23676 unused variable loops.
23677
aca43c6c
JJ		 236782013-02-26 Jakub Jelinek <jakub@redhat.com>
23679
cecbe5d9
JJ		 23680 PR tree-optimization/56448
23681 * fold-const.c (operand_equal_p) <case tcc_reference>: Don't look at
23682 TREE_SIDE_EFFECTS if flags contain OEP_CONSTANT_ADDRESS_OF.
23683 Clear OEP_CONSTANT_ADDRESS_OF from flags before recursing on second or
23684 later operands of the references, or even first operand for
23685 INDIRECT_REF, TARGET_MEM_REF or MEM_REF.
23686
aca43c6c
JJ		 23687 PR tree-optimization/56443
23688 * tree-vect-stmts.c (get_vectype_for_scalar_type_and_size): For
23689 overaligned types, pass TYPE_UNSIGNED (scalar_type) as second argument
23690 to type_for_mode langhook.
23691
53e2e141
MT		 236922013-02-25 Matt Turner <mattst88@gmail.com>
23693
23694 * doc/invoke.texi: Document r4700.
23695
259ee451
RB		 236962013-02-25 Richard Biener <rguenther@suse.de>
23697
23698 PR tree-optimization/56175
23699 * tree-ssa-forwprop.c (hoist_conversion_for_bitop_p): New predicate,
23700 split out from ...
23701 (simplify_bitwise_binary): ... here. Also guard the conversion
23702 of (type) X op CST to (type) (X op ((type-x) CST)) with it.
23703
6d65e8f1
CM		 237042013-02-25 Catherine Moore <clm@codesourcery.com>
23705
23706 Revert:
23707 2013-02-24 Catherine Moore <clm@codesourcery.com>
23708 Maciej W. Rozycki <macro@codesourcery.com>
23709 Tom de Vries <tom@codesourcery.com>
a72d8780 23710 Nathan Sidwell <nathan@codesourcery.com>
6d65e8f1
CM		 23711 Iain Sandoe <iain@codesourcery.com>
23712 Nathan Froyd <froydnj@codesourcery.com>
a72d8780 23713 Chao-ying Fu <fu@mips.com>
88b97037 23714
6d65e8f1 23715 * doc/extend.texi: (micromips, nomicromips, nocompression):
88b97037 23716 Document new function attributes.
6d65e8f1
CM		 23717 * doc/invoke.texi (minterlink-compressed, mmicromips,
23718 m14k, m14ke, m14kec): Document new options.
23719 (minterlink-mips16): Update documentation.
23720 * doc/md.texi (ZC, ZD): Document new constraints.
23721 * configure.ac (gcc_cv_as_micromips): Check if linker
23722 supports the .set micromips directive.
23723 * configure: Regenerate.
23724 * config.in: Regenerate.
23725 * config/mips/mips-tables.opt: Regenerate.
23726 * config/mips/micromips.md: New file.
23727 * constraints.md (ZC, AD): New constraints.
23728 * config/mips/predicates.md (movep_src_register): New predicate.
23729 (movep_src_operand): New predicate.
23730 (non_volatile_mem_operand): New predicate.
23731 * config/mips/mips.md (multimem): New type.
23732 (length): Differentiate between 17-bit and 18-bit branch offsets.
23733 (MOVEP1, MOVEP2): New mode iterator.
23734 (mov_<load>l): Use ZC constraint.
23735 (mov_<load>r): Likewise.
23736 (mov_<store>l): Likewise.
23737 (mov_<store>r): Likewise.
23738 (*branch_equality<mode>_inverted): Add microMIPS support.
23739 (*branch_equality<mode>): Likewise.
23740 (*jump_absolute): Likewise.
23741 (indirect_jump_<mode>): Likewise.
23742 (tablejump_<mode>): Likewise.
23743 (<optab>_internal): Likewise.
23744 (sibcall_internal): Likewise.
23745 (sibcall_value_internal): Likewise.
23746 (prefetch): Use constraint ZD.
23747 * config/mips/mips.opt (minterlink-compressed): New option.
23748 (minterlink-mips16): Now an alias for minterlink-compressed.
23749 (mmicromips): New option.
23750 * config/mips/sync.md (sync_compare_and_swap<mode>): Use ZR constraint.
23751 (compare_and_swap_12): Likewise.
23752 (sync_add<mode>): Likewise.
23753 (sync_<optab>_12): Likewise.
23754 (sync_old_<optab>_12): Likewise.
23755 (sync_new_<optab>_12): Likewise.
23756 (sync_nand_12): Likewise.
23757 (sync_old_nand_12): Likewise.
23758 (sync_new_nand_12): Likewise.
23759 (sync_sub<mode>): Likewise.
23760 (sync_old_add<mode>): Likewise.
23761 (sync_old_sub<mode>): Likewise.
23762 (sync_new_add<mode>): Likewise.
23763 (sync_new_sub<mode>): Likewise.
23764 (sync_<optab><mode>): Likewise.
23765 (sync_old_<optab><mode>): Likewise.
23766 (sync_new_<optab><mode>): Likewise.
23767 (sync_nand<mode>): Likewise.
23768 (sync_old_nand<mode>): Likewise.
23769 (sync_new_nand<mode>): Likewise.
23770 (sync_lock_test_and_set<mode>): Likewise.
23771 (test_and_set_12): Likewise.
23772 (atomic_compare_and_swap<mode>): Likewise.
23773 (atomic_exchange<mode>_llsc): Likewise.
23774 (atomic_fetch_add<mode>_llsc): Likewise.
23775 * config/mips/mips-cpus.def (m14kc, m14k): New processors.
23776 * config/mips/mips-protos.h (umips_output_save_restore): New prototype.
23777 (umips_save_restore_pattern_p): Likewise.
23778 (umips_load_store_pair_p): Likewise.
23779 (umips_output_load_store_pair): Likewise.
23780 (umips_movep_target_p): Likewise.
23781 (umips_12bit_offset_address_p): Likewise.
23782 * config/mips/mips.c (MIPS_MAX_FIRST_STEP): Update for microMIPS.
23783 (mips_base_mips16): Rename this...
23784 (mips_base_compression_flags): ...to this. Update all uses.
23785 (mips_attribute_table): Add micromips, nomicromips and nocompression.
23786 (mips_mips16_decl_p): Delete.
23787 (mips_nomips16_decl_p): Delete.
23788 (mips_get_compress_on_flags): New function.
23789 (mips_get_compress_off_flags): New function.
23790 (mips_get_compress_mode): New function.
23791 (mips_get_compress_on_name): New function.
23792 (mips_get_compress_off_name): New function.
23793 (mips_insert_attributes): Support multiple compression types.
23794 (mips_merge_decl_attributes): Likewise.
23795 (umips_12bit_offset_address_p): New function.
23796 (mips_start_function_definition): Emit .set micromips directive.
23797 (mips_call_may_need_jalx_p): New function.
23798 (mips_function_ok_for_sibcall): Add microMIPS support.
23799 (mips_print_operand_punctuation): Support short delay slots and
23800 compact jumps.
23801 (umips_swm_mask, umips_swm_encoding): New.
23802 (umips_build_save_restore): New function.
23803 (mips_for_each_saved_gpr_and_fpr): Add microMIPS support.
23804 (was_mips16_p): Remove.
23805 (old_compression_mode): New.
23806 (mips_set_compression_mode): New function.
23807 (mips_set_current_function): Add microMIPS support.
23808 (mips_option_override): Likewise.
23809 (umips_save_restore_pattern_p): New function.
23810 (umips_output_save_restore): New function.
23811 (umips_load_store_pair_p_1): New function.
23812 (umips_load_store_pair_p): New function.
23813 (umips_output_load_store_pair_1): New function.
23814 (umips_output_load_store_pair): New function.
23815 (umips_movep_target_p) New function.
23816 (mips_prepare_pch_save): Add microMIPS support.
23817 * config/mips/mips.h (TARGET_COMPRESSION): New.
23818 (TARGET_CPU_CPP_BUILTINS): Update macro
23819 to use new compression flags and to support microMIPS.
23820 (MIPS_ISA_LEVEL_SPEC): Add m14k processors.
23821 (MIPS_ARCH_FLOAT_SPEC): Likewise.
23822 (ISA_HAS_LWXS): Include TARGET_MICROMIPS.
23823 (ISA_HAS_LOAD_DELAY): Exclude TARGET_MICROMIPS.
23824 (ASM_SPEC): Support mmicromips and mno-micromips.
23825 (M16STORE_REG_P): New macro.
23826 (MIPS_CALL): Support TARGET_MICROMIPS.
23827 (MICROMIPS_J): New macro.
23828 (mips_base_mips16): Rename this...
23829 (mips_base_compression_flags): ...to this.
23830 (UMIPS_12BIT_OFFSET_P): New macro.
23831 * config/mips/t-sde: (MULTILIB_OPTIONS): Add microMIPS.
23832 (MULTILIB_DIRNAMES): Likewise.
23833
5e5df392
TV		 238342013-02-25 Tom de Vries <tom@codesourcery.com>
23835
23836 PR rtl-optimization/56131
23837 * insn-notes.def (INSN_NOTE_BASIC_BLOCK): Update comment.
23838 * cfgrtl.c (delete_insn): Don't reorder NOTE_INSN_DELETED_LABEL and
23839 NOTE_INSN_BASIC_BLOCK if BLOCK_FOR_INSN == NULL.
23840
3fdb53c1
TB		 238412013-02-25 Tobias Burnus <burnus@net-b.de>
23842
23843 * doc/invoke.texi (-fsanitize=): Move from optimization
23844 to debugging options.
23845
ed358aea
AB		 238462013-02-25 Andrey Belevantsev <abel@ispras.ru>
23847
23848 * sched-deps.c (sched_analyze_insn): Fix typo in comment.
23849
f45e9053
AB		 238502013-02-25 Andrey Belevantsev <abel@ispras.ru>
23851 Alexander Monakov <amonakov@ispras.ru>
23852
23853 PR middle-end/56077
23854 * sched-deps.c (sched_analyze_insn): When reg_pending_barrier,
23855 flush pending lists also on non-jumps. Adjust comment.
23856
6941b508
CM		 238572013-02-24 Catherine Moore <clm@codesourcery.com>
23858 Maciej W. Rozycki <macro@codesourcery.com>
23859 Tom de Vries <tom@codesourcery.com>
a72d8780 23860 Nathan Sidwell <nathan@codesourcery.com>
6941b508
CM		 23861 Iain Sandoe <iain@codesourcery.com>
23862 Nathan Froyd <froydnj@codesourcery.com>
a72d8780 23863 Chao-ying Fu <fu@mips.com>
6941b508
CM		 23864
23865 * doc/extend.texi: (micromips, nomicromips, nocompression):
88b97037 23866 Document new function attributes.
6941b508
CM		 23867 * doc/invoke.texi (minterlink-compressed, mmicromips,
23868 m14k, m14ke, m14kec): Document new options.
23869 (minterlink-mips16): Update documentation.
23870 * doc/md.texi (ZC, ZD): Document new constraints.
23871 * configure.ac (gcc_cv_as_micromips): Check if linker
23872 supports the .set micromips directive.
23873 * configure: Regenerate.
23874 * config.in: Regenerate.
23875 * config/mips/mips-tables.opt: Regenerate.
23876 * config/mips/micromips.md: New file.
23877 * constraints.md (ZC, AD): New constraints.
23878 * config/mips/predicates.md (movep_src_register): New predicate.
23879 (movep_src_operand): New predicate.
23880 (non_volatile_mem_operand): New predicate.
23881 * config/mips/mips.md (multimem): New type.
23882 (length): Differentiate between 17-bit and 18-bit branch offsets.
23883 (MOVEP1, MOVEP2): New mode iterator.
3fdb53c1 23884 (mov_<load>l): Use ZC constraint.
6941b508
CM		 23885 (mov_<load>r): Likewise.
23886 (mov_<store>l): Likewise.
23887 (mov_<store>r): Likewise.
23888 (*branch_equality<mode>_inverted): Add microMIPS support.
23889 (*branch_equality<mode>): Likewise.
23890 (*jump_absolute): Likewise.
23891 (indirect_jump_<mode>): Likewise.
23892 (tablejump_<mode>): Likewise.
23893 (<optab>_internal): Likewise.
23894 (sibcall_internal): Likewise.
23895 (sibcall_value_internal): Likewise.
23896 (prefetch): Use constraint ZD.
23897 * config/mips/mips.opt (minterlink-compressed): New option.
23898 (minterlink-mips16): Now an alias for minterlink-compressed.
23899 (mmicromips): New option.
23900 * config/mips/sync.md (sync_compare_and_swap<mode>): Use ZR constraint.
23901 (compare_and_swap_12): Likewise.
23902 (sync_add<mode>): Likewise.
23903 (sync_<optab>_12): Likewise.
23904 (sync_old_<optab>_12): Likewise.
23905 (sync_new_<optab>_12): Likewise.
23906 (sync_nand_12): Likewise.
23907 (sync_old_nand_12): Likewise.
23908 (sync_new_nand_12): Likewise.
23909 (sync_sub<mode>): Likewise.
23910 (sync_old_add<mode>): Likewise.
23911 (sync_old_sub<mode>): Likewise.
23912 (sync_new_add<mode>): Likewise.
23913 (sync_new_sub<mode>): Likewise.
23914 (sync_<optab><mode>): Likewise.
23915 (sync_old_<optab><mode>): Likewise.
23916 (sync_new_<optab><mode>): Likewise.
23917 (sync_nand<mode>): Likewise.
23918 (sync_old_nand<mode>): Likewise.
23919 (sync_new_nand<mode>): Likewise.
23920 (sync_lock_test_and_set<mode>): Likewise.
23921 (test_and_set_12): Likewise.
23922 (atomic_compare_and_swap<mode>): Likewise.
23923 (atomic_exchange<mode>_llsc): Likewise.
23924 (atomic_fetch_add<mode>_llsc): Likewise.
23925 * config/mips/mips-cpus.def (m14kc, m14k): New processors.
23926 * config/mips/mips-protos.h (umips_output_save_restore): New prototype.
23927 (umips_save_restore_pattern_p): Likewise.
23928 (umips_load_store_pair_p): Likewise.
23929 (umips_output_load_store_pair): Likewise.
23930 (umips_movep_target_p): Likewise.
23931 (umips_12bit_offset_address_p): Likewise.
23932 * config/mips/mips.c (MIPS_MAX_FIRST_STEP): Update for microMIPS.
23933 (mips_base_mips16): Rename this...
23934 (mips_base_compression_flags): ...to this. Update all uses.
23935 (mips_attribute_table): Add micromips, nomicromips and nocompression.
23936 (mips_mips16_decl_p): Delete.
23937 (mips_nomips16_decl_p): Delete.
3fdb53c1 23938 (mips_get_compress_on_flags): New function.
6941b508
CM		 23939 (mips_get_compress_off_flags): New function.
23940 (mips_get_compress_mode): New function.
23941 (mips_get_compress_on_name): New function.
23942 (mips_get_compress_off_name): New function.
23943 (mips_insert_attributes): Support multiple compression types.
23944 (mips_merge_decl_attributes): Likewise.
23945 (umips_12bit_offset_address_p): New function.
23946 (mips_start_function_definition): Emit .set micromips directive.
23947 (mips_call_may_need_jalx_p): New function.
23948 (mips_function_ok_for_sibcall): Add microMIPS support.
23949 (mips_print_operand_punctuation): Support short delay slots and
23950 compact jumps.
23951 (umips_swm_mask, umips_swm_encoding): New.
23952 (umips_build_save_restore): New function.
23953 (mips_for_each_saved_gpr_and_fpr): Add microMIPS support.
23954 (was_mips16_p): Remove.
23955 (old_compression_mode): New.
23956 (mips_set_compression_mode): New function.
23957 (mips_set_current_function): Add microMIPS support.
23958 (mips_option_override): Likewise.
23959 (umips_save_restore_pattern_p): New function.
23960 (umips_output_save_restore): New function.
23961 (umips_load_store_pair_p_1): New function.
23962 (umips_load_store_pair_p): New function.
23963 (umips_output_load_store_pair_1): New function.
23964 (umips_output_load_store_pair): New function.
23965 (umips_movep_target_p) New function.
23966 (mips_prepare_pch_save): Add microMIPS support.
23967 * config/mips/mips.h (TARGET_COMPRESSION): New.
23968 (TARGET_CPU_CPP_BUILTINS): Update macro
23969 to use new compression flags and to support microMIPS.
23970 (MIPS_ISA_LEVEL_SPEC): Add m14k processors.
23971 (MIPS_ARCH_FLOAT_SPEC): Likewise.
23972 (ISA_HAS_LWXS): Include TARGET_MICROMIPS.
23973 (ISA_HAS_LOAD_DELAY): Exclude TARGET_MICROMIPS.
23974 (ASM_SPEC): Support mmicromips and mno-micromips.
23975 (M16STORE_REG_P): New macro.
23976 (MIPS_CALL): Support TARGET_MICROMIPS.
23977 (MICROMIPS_J): New macro.
23978 (mips_base_mips16): Rename this...
23979 (mips_base_compression_flags): ...to this.
23980 (UMIPS_12BIT_OFFSET_P): New macro.
23981 * config/mips/t-sde: (MULTILIB_OPTIONS): Add microMIPS.
23982 (MULTILIB_DIRNAMES): Likewise.
23983
cdebbc6d
JJ		 239842013-02-24 Jakub Jelinek <jakub@redhat.com>
23985
23986 PR target/52555
23987 * target-globals.c (save_target_globals): For init_reg_sets and
23988 target_reinit remporarily set this_fn_optabs to this_target_optabs.
23989
18c63565
JG		 239902013-02-22 James Grennahlgh <james.greenhalgh@arm.com>
23991
23992 * config/aarch64/aarch64-simd-builtins.def: Add copyright header.
23993 * config/aarch64/t-aarch64
23994 (aarch64-builtins.o): Depend on aarch64-simd-builtins.def.
23995
2194f7a2
VM		 239962013-02-22 Vladimir Makarov <vmakarov@redhat.com>
23997
23998 PR inline-asm/56148
23999 * lra-constraints.c (process_alt_operands): Reload operand
24000 conflicting with earlier clobber only if no more other conflicting
24001 operands.
24002
7d613735
JJ		 240032013-02-22 Jakub Jelinek <jakub@redhat.com>
24004
24005 PR sanitizer/56393
24006 * config/gnu-user.h (LIBASAN_EARLY_SPEC): Link in libasan_preinit.o
24007 if not linking a shared library.
24008
ac8d93a7
SL		 240092013-02-22 Seth LaForge <sethml@google.com>
24010
24011 * config.gcc (arm*-*-eabi*): Treat arm*eb as big-endian.
24012
e0237780
GY		 240132013-02-22 Greta Yorsh <Greta.Yorsh@arm.com>
24014
3fdb53c1
TB		 24015 * config/arm/arm.md (split for extendsidi): Update condition.
24016 (zero_extend<mode>di2,extend<mode>di2): Add an alternative.
24017 * config/arm/iterators.md (qhs_extenddi_cstr): Likewise.
24018 (qhs_zextenddi_cstr): Likewise.
e0237780 24019
d7fde18c
JJ		 240202013-02-21 Jakub Jelinek <jakub@redhat.com>
24021
be63b77d
JJ		 24022 PR middle-end/56420
24023 * expmed.c (EXACT_POWER_OF_2_OR_ZERO_P): Do subtraction in uhwi, to
24024 avoid signed wrapping.
24025 (expand_mult): Handle properly multiplication by
24026 ((dword_type) -1) << (BITS_PER_WORD - 1). Improve multiplication by
24027 ((dword_type) 1) << (BITS_PER_WORD - 1). Avoid undefined behavior
24028 in the compiler if coeff is HOST_WIDE_INT_MIN.
24029 (expand_divmod): Don't make ext_op1 static, change it's type to
24030 uhwi. Avoid undefined behavior in -INTVAL (op1).
24031
d7fde18c
JJ		 24032 PR rtl-optimization/50339
24033 * lower-subreg.h (struct lower_subreg_choices): Add splitting_ashiftrt
24034 field.
24035 * lower-subreg.c (compute_splitting_shift): Handle ASHIFTRT.
24036 (compute_costs): Call compute_splitting_shift also for ASHIFTRT
24037 into splitting_ashiftrt field.
24038 (find_decomposable_shift_zext, resolve_shift_zext): Handle also
24039 ASHIFTRT.
24040 (dump_choices): Fix up printing LSHIFTRT choices, print ASHIFTRT
24041 choices.
24042
6aad4455
AH		 240432013-02-20 Aldy Hernandez <aldyh@redhat.com>
24044
24045 PR middle-end/56108
24046 * trans-mem.c (execute_tm_mark): Do not expand transactions that
24047 are sure to go irrevocable.
24048
38fe784d
HPN		 240492013-02-21 Hans-Peter Nilsson <hp@axis.com>
24050
24051 * doc/rtl.texi (vec_concat, vec_duplicate): Mention that
24052 scalars are valid operands.
24053
0fd44da3
MJ		 240542013-02-21 Martin Jambor <mjambor@suse.cz>
24055
24056 PR tree-optimization/56310
24057 * ipa-cp.c (agg_replacements_to_vector): New parameter index, copy
24058 only matching indices and non-negative final offsets.
24059 (intersect_aggregates_with_edge): Pass src_idx to
24060 agg_replacements_to_vector. Pass src_idx insstead of index to
24061 intersect_with_agg_replacements.
24062
7a92038b
MJ		 240632013-02-21 Martin Jambor <mjambor@suse.cz>
24064
24065 * ipa-cp.c (good_cloning_opportunity_p): Dump the real threshold
24066 instead of hard-wired defaults.
24067
c0da9c37
MR		 240682013-02-21 Maciej W. Rozycki <macro@codesourcery.com>
24069
24070 * doc/invoke.texi (MIPS Options): Update documentation of the
24071 floating-point multiply-accumulate instruction restrictions.
24072
d247ea0c 240732013-02-21 Kostya Serebryany <kcc@google.com>
7df59255
KS		 24074
24075 * config/i386/i386.c (ix86_asan_shadow_offset): Use 0x7fff8000 as
24076 asan_shadow_offset on x86_64 linux.
24077
22deefcb
RB		 240782013-02-21 Richard Biener <rguenther@suse.de>
24079
24080 PR tree-optimization/56415
24081 Revert
24082 2013-02-11 Richard Biener <rguenther@suse.de>
24083
24084 PR tree-optimization/56273
24085 * tree-vrp.c (simplify_cond_using_ranges): Disable for the
24086 first VRP run.
24087
7bcc6e75
JJ		 240882013-02-21 Jakub Jelinek <jakub@redhat.com>
24089
a0ad148f
JJ		 24090 PR bootstrap/56258
24091 * doc/invoke.texi (-fdump-rtl-pro_and_epilogue): Use @item
24092 instead of @itemx.
24093
7bcc6e75
JJ		 24094 PR inline-asm/56405
24095 * expr.c (expand_expr_real_1) <case TARGET_MEM_REF, MEM_REF>: Don't
24096 use movmisalign or extract_bit_field for EXPAND_MEMORY modifier.
24097
a0a7b611
JH		 240982013-02-20 Jan Hubicka <jh@suse.cz>
24099
24100 PR tree-optimization/56265
88b97037
UB		 24101 * ipa-prop.c (ipa_make_edge_direct_to_target): Fixup callgraph
24102 when target is referenced for first time.
a0a7b611 24103
c0e50f72
RB		 241042013-02-20 Richard Biener <rguenther@suse.de>
24105
24106 * tree-call-cdce.c (tree_call_cdce): Do not remove unused locals.
24107 * tree-ssa-forwprop.c (ssa_forward_propagate_and_combine): Likewise.
24108 * tree-ssa-dce.c (perform_tree_ssa_dce): Likewise.
24109 * tree-ssa-copyrename.c (copy_rename_partition_coalesce): Do
24110 not return anything.
24111 (rename_ssa_copies): Do not remove unused locals.
24112 * tree-ssa-ccp.c (do_ssa_ccp): Likewise.
88b97037 24113 * tree-ssanames.c (pass_release_ssa_names): Remove unused locals first.
c0e50f72
RB		 24114 * passes.c (execute_function_todo): Do not schedule unused locals
24115 removal if cleanup_tree_cfg did something.
24116 * tree-ssa-live.c (remove_unused_locals): Dump statistics
24117 about the number of removed locals.
24118
a52ca739
RB		 241192013-02-20 Richard Biener <rguenther@suse.de>
24120
24121 PR tree-optimization/56398
88b97037 24122 * tree-vect-loop-manip.c (adjust_debug_stmts): Skip SSA default defs.
a52ca739 24123
ec9c9d1b
MJ		 241242013-02-20 Martin Jambor <mjambor@suse.cz>
24125
24126 PR tree-optimization/55334
24127 * ipa-cp.c (initialize_node_lattices): Disable IPA-CP through and to
24128 restricted pointers to arrays.
24129
e91c8ed6 241302013-02-20 Richard Biener <rguenther@suse.de>
88b97037 24131 Jakub Jelinek <jakub@redhat.com>
e91c8ed6
RB		 24132
24133 PR tree-optimization/56396
24134 * tree-ssa-ccp.c (n_const_val): New static variable.
24135 (get_value): Return NULL for SSA names we don't have a lattice
24136 entry for.
24137 (ccp_initialize): Initialize n_const_val.
24138 * tree-ssa-copy.c (n_copy_of): New static variable.
24139 (init_copy_prop): Initialize n_copy_of.
24140 (get_value): Return NULL_TREE for SSA names we don't have a
24141 lattice entry for.
24142
3d916479
MJ		 241432013-02-20 Martin Jambor <mjambor@suse.cz>
24144
24145 * ipa-cp.c (initialize_node_lattices): Fix dumping condition.
24146
71a86758
RB		 241472013-02-20 Richard Biener <rguenther@suse.de>
24148
24149 * genpreds.c (write_lookup_constraint): Do not compare first
24150 letter of the constraint again.
24151
79836a12
RB		 241522013-02-20 Richard Biener <rguenther@suse.de>
24153
24154 * tree-ssa-loop-ivopts.c (alloc_use_cost_map): Use bitmap_count_bits
24155 and ceil_log2.
24156 (get_use_iv_cost): Terminate hashtable walk when coming across
24157 an empty entry.
24158
bbe4fb2c
IZ		 241592013-02-20 Igor Zamyatin <igor.zamyatin@intel.com>
24160
24161 * config/i386/i386.c (initial_ix86_tune_features): Turn on fp
24162 reassociation for avx2 targets.
24163
a72d8780 241642012-02-19 Edgar E. Iglesias <edgar.iglesias@gmail.com>
79836a12 24165
c77f83d5
EI		 24166 * config/microblaze/microblaze.c: microblaze_has_clz = 0
24167 Add version check for v8.10.a to enable microblaze_has_clz
88b97037 24168 * config/microblaze/microblaze.h: Add TARGET_HAS_CLZ as combined
c77f83d5
EI		 24169 version and TARGET_PATTERN_COMPARE check
24170 * config/microblaze/microblaze.md: New clzsi2 instruction
3fdb53c1 24171
a72d8780 241722012-02-19 Edgar E. Iglesias <edgar.iglesias@gmail.com>
959434e6 24173
a72d8780 24174 * config/microblaze/microblaze.md (call_value_intern): Check symbol is
959434e6
EI		 24175 function before branching.
24176
242387fa
AB		 241772012-02-19 Andrey Belevantsev <abel@ispras.ru>
24178
24179 * sel-sched-dump.c (dump_insn_rtx_flags): Explicitly set
24180 DUMP_INSN_RTX_UID.
24181 (dump_insn_rtx_1): Pass PATTERN (insn) to str_pattern_slim.
24182
b4979ab9
AB		 241832012-02-19 Andrey Belevantsev <abel@ispras.ru>
24184
24185 PR middle-end/55889
b4979ab9
AB		 24186 * sel-sched.c: Include ira.h.
24187 (implicit_clobber_conflict_p): New function.
24188 (moveup_expr): Use it.
88b97037 24189 * Makefile.in (sel-sched.o): Depend on ira.h.
b4979ab9 24190
24d63016
RB		 241912013-02-19 Richard Biener <rguenther@suse.de>
24192
24193 PR tree-optimization/56384
24194 * tree-ssa-sccvn.h (struct vn_phi_s): Add type member.
24195 (vn_hash_type): Split out from ...
24196 (vn_hash_constant_with_type): ... here.
24197 * tree-ssa-sccvn.c (vn_phi_compute_hash): Use vn_hash_type.
24198 (vn_phi_eq): Compare types from vn_phi_s structure.
24199 (vn_phi_lookup): Populate vn_phi_s type.
24200 (vn_phi_insert): Likewise.
24201
a475fd3d
JJ		 242022013-02-19 Jakub Jelinek <jakub@redhat.com>
24203
47cc28f5
JJ		 24204 PR tree-optimization/56350
24205 * tree-vect-loop.c (vectorizable_reduction): If orig_stmt, return false
24206 if haven't found reduction or nested cycle operand, rather than
24207 asserting we must find it.
24208
a475fd3d
JJ		 24209 PR tree-optimization/56381
24210 * tree-ssa-pre.c (create_expression_by_pieces): Fix up last argument
24211 to fold_build3.
24212
242132013-02-18 Aldy Hernandez <aldyh@redhat.com>
135204dd
AH		 24214 Jakub Jelinek <jakub@redhat.com>
24215
24216 PR target/52555
24217 * genopinit.c (raw_optab_handler): Use this_fn_optabs.
24218 (swap_optab_enable): Same.
24219 (init_all_optabs): Use argument instead of global.
88b97037 24220 * tree.h (struct tree_optimization_option): New field target_optabs.
135204dd
AH		 24221 * expr.h (init_all_optabs): Add argument to prototype.
24222 (TREE_OPTIMIZATION_OPTABS): New.
24223 (save_optabs_if_changed): Protoize.
24224 * optabs.h: Declare this_fn_optabs.
24225 * optabs.c (save_optabs_if_changed): New.
24226 Declare this_fn_optabs.
24227 (init_optabs): Add argument to init_all_optabs() call.
24228 * function.c (invoke_set_current_function_hook): Handle per
24229 function optabs.
24230 * function.h (struct function): New field optabs.
24231 * config/mips/mips.c (mips_set_mips16_mode): Handle when
24232 optimization_current_node has changed.
24233 * target-globals.h (save_target_globals_default_opts): Protoize.
24234 * target-globals.c (save_target_globals_default_opts): New.
24235
3f587ca3
JDA		 242362013-02-18 John David Anglin <dave.anglin@nrc-cnrc.gc.ca>
24237
24238 PR target/56347
24239 * config/pa/pa.c (pa_conditional_register_usage): On HP-UX, mark
24240 registers %fr12 and %fr12R as call used.
ceaca33e
JDA		 24241
24242 PR target/56214
24243 * config/pa/predicates.md (base14_operand): Except for BLKmode, QImode
88b97037
UB		 24244 and HImode, require all displacements to be an integer multiple of
24245 their mode size.
ceaca33e
JDA		 24246 * config/pa/pa.c (pa_legitimate_address_p): For REG+BASE addresses,
24247 only allow QImode and HImode when reload is in progress and strict is
24248 true. Likewise for symbolic addresses. Use base14_operand to check
24249 displacements in REG+BASE addresses.
24250
fe0b4796
RB		 242512013-02-18 Richard Biener <rguenther@suse.de>
24252
24253 PR tree-optimization/56366
24254 * tree-vect-loop.c (get_initial_def_for_induction): Properly
24255 handle sign-conversion of outer-loop initial induction value.
24256
6aaf596b
RB		 242572013-02-18 Richard Biener <rguenther@suse.de>
24258
73db8ff1 24259 PR middle-end/56349
6aaf596b
RB		 24260 * cfghooks.c (merge_blocks): If we merge a latch into another
24261 block adjust references to it.
24262 * cfgloop.c (flow_loops_find): Reset latch before recomputing it.
24263 (verify_loop_structure): Verify that a recorded latch is in fact
24264 a latch.
24265
5e97dfb6
RB		 242662013-02-18 Richard Biener <rguenther@suse.de>
24267
24268 PR tree-optimization/56321
24269 * tree-ssa-reassoc.c (propagate_op_to_single_use): Properly
24270 order SSA name release and virtual operand unlinking.
24271
825527e8
EI		 242722013-02-17 Edgar E. Iglesias <edgar.iglesias@gmail.com>
24273
24274 * config/microblaze/microblaze.md (save_stack_block): Define.
24275 (restore_stack_block): Likewise.
24276
debd11d9
EI		 242772013-02-16 Edgar E. Iglesias <edgar.iglesias@gmail.com>
24278
24279 * config/microblaze/linux.h (TARGET_SUPPORTS_PIC): Define as 1.
24280 * config/microblaze/microblaze.h (TARGET_SUPPORTS_PIC): Define as 1.
24281 * config/microblaze/microblaze.c (microblaze_option_override):
24282 Bail out early for PIC modes when target does not support PIC.
24283
8ec77be0
EI		 242842013-02-16 Edgar E. Iglesias <edgar.iglesias@gmail.com>
24285
88b97037 24286 * config/microblaze/microblaze.c (microblaze_asm_trampoline_template):
8ec77be0
EI		 24287 Replace with a microblaze version.
24288 (microblaze_trampoline_init): Adapt for microblaze.
a752b80f 24289 * config/microblaze/microblaze.h (TRAMPOLINE_SIZE): Adapt for
8ec77be0
EI		 24290 microblaze.
24291
b41288b3
JJ		 242922013-02-16 Jakub Jelinek <jakub@redhat.com>
24293 Dodji Seketeli <dodji@redhat.com>
24294
24295 PR asan/56330
88b97037 24296 * asan.c (get_mem_refs_of_builtin_call): White space and style cleanup.
b41288b3
JJ		 24297 (instrument_mem_region_access): Do not forget to always put
24298 instrumentation of the of 'base' and 'base + len' in a "if (len !=
24299 0) statement, even for cases where either 'base' or 'base + len'
24300 are not instrumented -- because they have been previously
24301 instrumented. Simplify the logic by putting all the statements
24302 instrument 'base + len' inside a sequence, and then insert that
24303 sequence right before the current insertion point. Then, to
24304 instrument 'base + len', just get an iterator on that statement.
24305 And do not forget to update the pointer to iterator the function
24306 received as argument.
24307
47918951
VM		 243082013-02-15 Vladimir Makarov <vmakarov@redhat.com>
24309
24310 PR rtl-optimization/56348
24311 * lra-assigns.c (reload_pseudo_compare_func): Prefer bigger pseudos.
24312
3fb7c699
SB		 243132013-02-15 Steven Bosscher <steven@gcc.gnu.org>
24314
24315 * graph.c (start_graph_dump): Print dumpfile base as digraph label.
24316 (clean_graph_dump_file): Pass base to start_graph_dump.
24317
576fe41a
RH		 243182013-02-14 Richard Henderson <rth@redhat.com>
24319
24320 PR target/55941
24321 * lower-subreg.c (simple_move): Check dest mode instead of src mode.
24322
bb3accfa
SB		 243232013-02-14 Steven Bosscher <steven@gcc.gnu.org>
24324
24325 * collect2-aix.h: Define F_LOADONLY.
24326
03143140
RB		 243272013-02-14 Richard Biener <rguenther@suse.de>
24328
24329 PR lto/50494
24330 * varasm.c (output_constant_def_1): Get the decl representing
24331 the constant as argument.
24332 (output_constant_def): Wrap output_constant_def_1.
24333 (make_decl_rtl): Use output_constant_def_1 with the decl
24334 representing the constant.
24335 (build_constant_desc): Optionally re-use a decl already
24336 representing the constant.
24337 (tree_output_constant_def): Adjust.
24338
8fb06726
DS		 243392013-02-14 Dodji Seketeli <dodji@redhat.com>
24340
24341 Fix an asan crash
24342 * asan.c (instrument_builtin_call): Really put the length of the
24343 second source argument into src1_len.
24344
c4bfe8bf
JJ		 243452013-02-13 Jakub Jelinek <jakub@redhat.com>
24346
24347 * asan.c (create_cond_insert_point): Add create_then_fallthru_edge
24348 argument. If it is false, don't create edge from then_bb to
24349 fallthru_bb.
24350 (insert_if_then_before_iter): Pass true to it.
24351 (build_check_stmt): Pass false to it.
24352 (transform_statements): Flush hash table only on extended basic
24353 block boundaries, rather than at the beginning of every bb.
24354 Don't flush hash table on nonfreeing_call_p calls.
24355 * tree-flow.h (nonfreeing_call_p): New prototype.
24356 * tree-ssa-phiopt.c (nonfreeing_call_p): No longer static.
24357
7afe2801
DM		 243582013-02-13 David S. Miller <davem@davemloft.net>
24359
24360 * expmed.c (expand_shift_1): Only strip scalar integer subregs.
24361
70cc3288
VM		 243622013-02-13 Vladimir Makarov <vmakarov@redhat.com>
24363
24364 PR target/56184
24365 * ira.c (max_regno_before_ira): Move from ...
24366 (ira): ... here.
24367 (fix_reg_equiv_init): Use max_regno_before_ira instead of
24368 vec_safe_length.
24369
6422242b
JJ		 243702013-02-13 Jakub Jelinek <jakub@redhat.com>
24371
24372 * config/i386/i386.c (ix86_asan_shadow_offset): Revert last change.
24373
976a81ee
RB		 243742013-02-13 Richard Biener <rguenther@suse.de>
24375
24376 PR lto/56295
24377 * gimple-streamer-out.c (output_gimple_stmt): Undo wrapping
24378 globals in MEM_REFs.
24379
c1874a87
RB		 243802013-02-13 Richard Biener <rguenther@suse.de>
24381
24382 * loop-init.c (loop_optimizer_init): Clear loop state when
24383 re-initializing preserved loops.
24384 * loop-unswitch.c (unswitch_single_loop): Return whether
24385 we unswitched the loop. Do not verify loop state here.
88b97037 24386 (unswitch_loops): When we unswitched a loop discover new loops.
c1874a87 24387
b4ab7d34
KS		 243882013-02-13 Kostya Serebryany <kcc@google.com>
24389
88b97037
UB		 24390 * config/i386/i386.c: Use 0x7fff8000 as asan_shadow_offset
24391 on x86_64 linux.
b4ab7d34
KS		 24392 * sanitizer.def: Rename __asan_init to __asan_init_v1.
24393
bdcbe80c
DS		 243942013-02-12 Dodji Seketeli <dodji@redhat.com>
24395
24396 Avoid instrumenting duplicated memory access in the same basic block
24397 * Makefile.in (asan.o): Add new dependency on hash-table.h
24398 * asan.c (struct asan_mem_ref, struct mem_ref_hasher): New types.
24399 (asan_mem_ref_init, asan_mem_ref_get_end, get_mem_ref_hash_table)
24400 (has_stmt_been_instrumented_p, empty_mem_ref_hash_table)
24401 (free_mem_ref_resources, has_mem_ref_been_instrumented)
24402 (has_stmt_been_instrumented_p, update_mem_ref_hash_table)
24403 (get_mem_ref_of_assignment): New functions.
24404 (get_mem_refs_of_builtin_call): Extract from
24405 instrument_builtin_call and tweak a little bit to make it fit with
24406 the new signature.
24407 (instrument_builtin_call): Use the new
24408 get_mem_refs_of_builtin_call. Use gimple_call_builtin_p instead
24409 of is_gimple_builtin_call.
24410 (instrument_derefs, instrument_mem_region_access): Insert the
24411 instrumented memory reference into the hash table.
24412 (maybe_instrument_assignment): Renamed instrument_assignment into
24413 this, and change it to advance the iterator when instrumentation
24414 actually happened and return true in that case. This makes it
24415 homogeneous with maybe_instrument_assignment, and thus give a
24416 chance to callers to be more 'regular'.
24417 (transform_statements): Clear the memory reference hash table
24418 whenever we enter a new BB, when we cross a function call, or when
24419 we are done transforming statements. Use
24420 maybe_instrument_assignment instead of instrumentation. No more
24421 need to special case maybe_instrument_assignment and advance the
24422 iterator after calling it; it's now handled just like
24423 maybe_instrument_call. Update comment.
24424
4861a1f7
RB		 244252013-02-13 Richard Biener <rguenther@suse.de>
24426
24427 * config/mn10300/mn10300.c (mn10300_scan_for_setlb_lcc):
24428 Fix loop discovery code.
24429
1c86bd80
VM		 244302013-02-12 Vladimir Makarov <vmakarov@redhat.com>
24431
24432 PR inline-asm/56148
24433 * lra-constraints.c (process_alt_operands): Match early clobber
70cc3288 24434 operand with itself. Check conflicts with earlyclobber only if
1c86bd80
VM		 24435 the operand is not reloaded. Prefer to reload conflicting operand
24436 if earlyclobber and matching operands are the same.
24437
7cbda518
RB		 244382013-02-12 Richard Biener <rguenther@suse.de>
24439
24440 PR lto/56297
24441 * lto-streamer-out.c (write_symbol): Do not output symbols
24442 for hard register variables.
24443
e68a4ef6
GJL		 244442013-02-12 Georg-Johann Lay <avr@gjlay.de>
24445
24446 PR target/54222
24447 * config/avr/avr-dimode.md (umulsidi3, mulsidi3): New expanders.
24448 (umulsidi3_insn, mulsidi3_insn): New insns.
24449
a72d8780 244502013-02-12 Christophe Lyon <christophe.lyon@linaro.org>
2597da22
CL		 24451
24452 * config/arm/arm-protos.h (struct cpu_vec_costs): New struct type.
24453 (struct tune_params): Add vec_costs field.
24454 * config/arm/arm.c (arm_builtin_vectorization_cost)
24455 (arm_add_stmt_cost): New functions.
24456 (TARGET_VECTORIZE_BUILTIN_VECTORIZATION_COST)
24457 (TARGET_VECTORIZE_ADD_STMT_COST): Define.
24458 (arm_default_vec_cost): New struct of type cpu_vec_costs.
24459 (arm_slowmul_tune, arm_fastmul_tune, arm_strongarm_tune)
24460 (arm_xscale_tune, arm_9e_tune, arm_v6t2_tune, arm_cortex_tune)
24461 (arm_cortex_a15_tune, arm_cortex_a5_tune, arm_cortex_a9_tune)
24462 (arm_v6m_tune, arm_fa726te_tune): Define new vec_costs field.
24463
43320568
RB		 244642013-02-12 Richard Biener <rguenther@suse.de>
24465
24466 PR lto/56295
24467 * gimple-streamer-in.c (input_gimple_stmt): Strip MEM_REFs off
24468 decls again if possible.
24469
a011aa39
RB		 244702013-02-12 Richard Biener <rguenther@suse.de>
24471
24472 PR middle-end/56288
24473 * tree-ssa.c (verify_ssa_name): Fix check, move
24474 SSA_NAME_IN_FREE_LIST check up.
24475
6da26889
JJ		 244762013-02-12 Jakub Jelinek <jakub@redhat.com>
24477 Steven Bosscher <steven@gcc.gnu.org>
24478
24479 PR rtl-optimization/56151
24480 * optabs.c (add_equal_note): Don't return 0 if target is a MEM,
24481 equal to op0 or op1, and last_insn pattern is CODE operation
24482 with MEM dest and one of the operands matches that MEM.
24483
f80e0faf
ST		 244842013-02-11 Sriraman Tallam <tmsriramgoogle.com>
24485
24486 * doc/extend.texi: Document Function Multiversioning and "default"
24487 parameter string to target attribute.
24488 * config/i386/i386.c (get_builtin_code_for_version): Return 0 if
24489 target attribute parameter is "default".
24490 (ix86_compare_version_priority): Remove checks for target attribute.
24491 (ix86_mangle_function_version_assembler_name): Change error to sorry.
24492 Remove check for target attribute equal to NULL. Add assert.
24493 (ix86_generate_version_dispatcher_body): Change error to sorry.
24494
6c59ffd1
IS		 244952013-02-11 Iain Sandoe <iain@codesourcery.com>
24496 Jack Howarth <howarth@bromo.med.uc.edu>
24497 Patrick Marlier <patrick.marlier@gmail.com>
24498
24499 PR libitm/55693
24500 * config/darwin.h: Replace ENDFILE_SPEC with TM_DESTRUCTOR and
24501 define ENDFILE_SPEC as TM_DESTRUCTOR.
24502 * config/i386/darwin.h (ENDFILE_SPEC): Use TM_DESTRUCTOR.
24503
a72d8780 245042013-02-11 Alexander Potapenko <glider@google.com>
6a52ccb3
AP		 24505 Jack Howarth <howarth@bromo.med.uc.edu>
24506 Jakub Jelinek <jakub@redhat.com>
24507
24508 PR sanitizer/55617
24509 * config/darwin.c (cdtor_record): Rename ctor_record.
24510 (sort_cdtor_records): Rename sort_ctor_records.
24511 (finalize_dtors): New routine to sort destructors by
24512 priority before use in assemble_integer.
24513 (machopic_asm_out_destructor): Use finalize_dtors if needed.
24514
b63fe007
UB		 245152013-02-11 Uros Bizjak <ubizjak@gmail.com>
24516
24517 PR rtl-optimization/56275
24518 * simplify-rtx.c (avoid_constant_pool_reference): Check that
24519 offset is non-negative and less than cmode size before
24520 calling simplify_subreg.
24521
8e89b5b5
RB		 245222013-02-11 Richard Biener <rguenther@suse.de>
24523
24524 PR tree-optimization/56264
24525 * cfgloop.h (fix_loop_structure): Adjust prototype.
24526 * loop-init.c (fix_loop_structure): Return the number of
24527 newly discovered loops.
24528 * tree-cfgcleanup.c (repair_loop_structures): When new loops
24529 are discovered, do a full loop-closed SSA rewrite.
24530
b4a4b56d
RB		 245312013-02-11 Richard Biener <rguenther@suse.de>
24532
24533 PR tree-optimization/56273
24534 * tree-vrp.c (simplify_cond_using_ranges): Disable for the
24535 first VRP run.
24536 (check_array_ref): Fix missing newline in dumps.
24537 (search_for_addr_array): Likewise.
24538
0c885229
DE		 245392013-02-09 David Edelsohn <dje.gcc@gmail.com>
24540
24541 * config/rs6000/aix61.h (OS_MISSING_ALTIVEC): Undefine.
24542
59ac9a55
JJ		 245432013-02-09 Jakub Jelinek <jakub@redhat.com>
24544
24545 PR target/56256
24546 * config/rs6000/rs6000.h (ASSEMBLER_DIALECT): Define.
24547
25bb0bb5
VM		 245482013-02-08 Vladimir Makarov <vmakarov@redhat.com>
24549
24550 PR rtl-optimization/56246
0c885229 24551 * lra-constraints.c (simplify_operand_subreg): Try to reuse
25bb0bb5
VM		 24552 reload pseudo.
24553 * lra.c (lra): Clear lra_optional_reload_pseudos only when all
24554 constraints are satisfied.
24555
a698cc03
JL		 245562013-02-08 Jeff Law <law@redhat.com>
24557
24558 PR debug/53948
24559 * emit-rtl.c (reg_is_parm_p): New function.
24560 * regs.h (reg_is_parm_p): New prototype.
24561 * ira-conflicts.c (ira_build_conflicts): Allow parameters in
24562 callee-clobbered registers.
24563
e1122ddd
MM		 245642013-02-08 Michael Meissner <meissner@linux.vnet.ibm.com>
24565
24566 PR target/56043
24567 * config/rs6000/rs6000.c (rs6000_builtin_vectorized_libmass):
24568 If there is no implicit builtin declaration, just return NULL.
24569
19c5f6e6
UB		 245702013-02-08 Uros Bizjak <ubizjak@gmail.com>
24571
24572 * config/i386/sse.md (FMAMODEM): New mode iterator.
24573 (fma<mode>4, fms<mode>4, fnma<mode>4, fnms<mode>4): Use FMAMODEM
24574 mode iterator. Do not use TARGET_SSE_MATH in insn constraint.
24575
2480f2ca 245762013-02-08 Uros Bizjak <ubizjak@gmail.com>
76ef61fb 24577
2480f2ca
UB		 24578 * config/i386/gnu-user.h (TARGET_CAN_SPLIT_STACK): Define only
24579 when HAVE_GAS_CFI_PERSONALITY_DIRECTIVE is set.
24580 * config/i386/gnu-user64.h (TARGET_CAN_SPLIT_STACK): Ditto.
24581
245822013-02-08 Edgar E. Iglesias <edgar.iglesias@gmail.com>
24583
24584 * config.gcc (microblaze*-linux*): Add TARGET_BIG_ENDIAN_DEFAULT.
24585 (microblaze*-*-elf): Likewise.
24586 * config/microblaze/linux.h: Add -mbig-endian / -mlittle-endian to
24587 LINK_SPEC.
24588 * config/microblaze/microblaze-c.c: Add builtin defines for
24589 _LITTLE_ENDIAN and _BIG_ENDIAN.
24590 * config/microblaze/microblaze.h: Add TARGET_ENDIAN_DEFAULT and
24591 add to TARGET_DEFAULT flags.
76ef61fb 24592 Expand ASM_SPEC and LINK_SPEC.
2480f2ca
UB		 24593 Update BYTES_BIG_ENDIAN and WORDS_BIG_ENDIAN.
24594 * config/microblaze/microblaze.md: Update extendsidi2 and
24595 movdi_internal instructions to use low-order / high-order reg
24596 print_operands.
24597 * config/microblaze/microblaze.opt: Add mbig-endian and mlittle-endian
24598 options and inversemask / mask of LITTLE_ENDIAN.
24599 * config/microblaze/t-microblaze: Expand multilib options to
24600 include mlittle-endian (le) and update exceptions patterns.
24601
600a5961
JJ		 246022013-02-08 Jakub Jelinek <jakub@redhat.com>
24603
5df81313
JJ		 24604 PR rtl-optimization/56195
24605 * lra-constraints.c (get_reload_reg): Don't reuse regs
24606 if they have smaller mode than requested, if they have
24607 wider mode than requested, try to return a SUBREG.
24608
600a5961
JJ		 24609 PR tree-optimization/56250
24610 * fold-const.c (extract_muldiv_1) <case NEGATE_EXPR>: Don't optimize
24611 if type is unsigned and code isn't MULT_EXPR.
24612
ff544649
GJL		 246132013-02-08 Georg-Johann Lay <avr@gjlay.de>
24614
24615 PR tree-optimization/56064
24616 * fixed-value.c (fixed_from_double_int): Sign/zero extend payload
24617 bits according to mode.
24618 * fixed-value.h (fixed_from_double_int)
24619 (const_fixed_from_double_int): Adjust comments.
24620
e45cde98
RB		 246212013-02-08 Richard Biener <rguenther@suse.de>
24622
24623 PR lto/56231
24624 * lto-streamer.h (struct data_in): Remove current_file, current_line
24625 and current_col members.
24626 * lto-streamer-out.c (lto_output_location): Stream changed bits
24627 en-block for efficiency.
24628 * lto-streamer-in.c (clear_line_info): Remove.
24629 (lto_input_location): Cache current file, line and column
24630 globally via local statics. Read changed bits en-block.
24631 (input_function): Do not call clear_line_info.
24632 (lto_read_body): Likewise.
24633 (lto_input_toplevel_asms): Likewise.
24634
c1ca73d8
MM		 246352013-02-08 Michael Matz <matz@suse.de>
24636
24637 PR tree-optimization/52448
24638 * tree-ssa-phiopt.c (struct name_to_bb): Add phase member.
24639 (nt_call_phase): New static.
24640 (add_or_mark_expr): Only mark accesses with newer phase than any
24641 call seen.
24642 (nonfreeing_call_p): New.
24643 (nt_init_block): Update nt_call_phase, mark blocks as visited.
24644 (nt_fini_block): Keep blocks marked as visited.
24645 (get_non_trapping): Initialize nt_call_phase, and reset aux pointer.
24646
57548aa2
RB		 246472013-02-08 Richard Biener <rguenther@suse.de>
24648
24649 * ira.c (ira): Free broken dominator information.
24650
8e10366f
UB		 246512013-02-08 Uros Bizjak <ubizjak@gmail.com>
24652
24653 * config/i386/i386.c (ix86_spill_class): Use INTEGER_CLASS_P macro.
24654
f64fb0fa
MP		 246552013-02-08 Marek Polacek <polacek@redhat.com>
24656
8e10366f 24657 * cfgloop.c (verify_loop_structure): Add more checking of headers.
f64fb0fa 24658
0375167b
RB		 246592013-02-08 Richard Biener <rguenther@suse.de>
24660
24661 PR middle-end/56181
24662 * cfgloop.h (flow_loops_find): Adjust.
24663 (bb_loop_header_p): Declare.
24664 * cfgloop.c (bb_loop_header_p): New function split out from ...
24665 (flow_loops_find): ... here. Adjust function signature,
24666 support incremental loop structure update.
24667 (verify_loop_structure): Cleanup. Verify a loop is a loop.
24668 * cfgloopmanip.c (fix_loop_structure): Move ...
24669 * loop-init.c (fix_loop_structure): ... here.
24670 (apply_loop_flags): Split out from ...
24671 (loop_optimizer_init): ... here.
24672 (fix_loop_structure): Use apply_loop_flags. Use flow_loops_find
24673 in incremental mode, only remove dead loops here.
24674
85d768f3
GJL		 246752013-02-08 Georg-Johann Lay <avr@gjlay.de>
24676
24677 PR target/54222
24678 * config/avr/avr.md (unspec) <UNSPEC_ROUND>: Add.
24679 * config/avr/avr-fixed.md (ALL4QA, ALL124QA): New mode iterators.
24680 (round<mode>3, round<mode>3_const): New expanders for fixed-mode.
24681 (*round<mode>3.libgcc): New insns for fixed-modes.
24682 * config/avr/builtins.def (ABSxx): Use a non-NULL LIBNAME.
24683 (ROUNDxx, COUNTLSxx, BITSxx, xxBITS): New DEF_BUILTINs.
24684 (ROUNDFX, COUNTLSFX, ABSFX): New DEF_BUILTINs.
24685 * config/avr/stdfix.h (absFX, bitsFX, FXbits): Remove inline
24686 implementations. Define to __builtin_avr_absFX,
24687 __builtin_avr_bitsFX, __builtin_avr_FXbits, respectively.
24688 (roundFX, countlsFX): Define to __builtin_avr_roundFX,
24689 __builtin_avr_countlsFX, respectively.
24690 * config/avr/avr-c.c (target.h): Include it.
24691 (enum avr_builtin_id): New enum.
24692 (avr_resolve_overloaded_builtin): New static function.
24693 (avr_register_target_pragmas): Use it to set
24694 targetm.resolve_overloaded_builtin.
24695 * config/avr/avr.c (avr_init_builtins): Supply myriads of local
24696 tree nodes used by DEF_BUILTIN.
24697 (avr_expand_builtin) <AVR_BUILTIN_ROUNDxx>: Sanity-check them.
24698 (avr_fold_builtin) <AVR_BUILTIN_BITSxx>: Fold to VIEW_COVERT_EXPR.
24699 <AVR_BUILTIN_xxBITS>: Same.
24700
661bc682
RB		 247012013-02-08 Richard Biener <rguenther@suse.de>
24702
24703 * cfgloop.c (verify_loop_structure): Properly handle
24704 a loop exiting to another loop header.
24705 * ira-int.h (ira_loops): Remove.
24706 * ira.c (ira_loops): Remove.
24707 (ira): Use loop_optimizer_init and loop_optimizer_finalize.
24708 (do_reload): Use loop_optimizer_finalize.
24709 * ira-build.c (create_loop_tree_nodes): Use get_loops and
24710 number_of_loops to access the loop tree.
24711 (more_one_region_p): Likewise.
24712 (finish_loop_tree_nodes): Likewise.
24713 (rebuild_regno_allocno_maps): Likewise.
24714 (mark_loops_for_removal): Likewise.
24715 (mark_all_loops_for_removal): Likewise.
24716 (remove_unnecessary_regions): Likewise.
24717 (ira_build): Likewise.
24718 * ira-emit.c (setup_entered_from_non_parent_p): Likewise.
24719
0d5049b2
RB		 247202013-02-08 Richard Biener <rguenther@suse.de>
24721
24722 * Makefile.in (tree-tailcall.o): Add $(CFGLOOP_H) dependency.
24723 * ipa-pure-const.c (analyze_function): Avoid calling
24724 mark_irreducible_loops twice.
8e10366f 24725 * tree-tailcall.c (tree_optimize_tail_calls_1): Mark loops for fixup.
0d5049b2 24726
499f32e8
DM		 247272013-02-07 David S. Miller <davem@davemloft.net>
24728
24729 * dwarf2out.c (based_loc_descr): Perform leaf register remapping
24730 on 'reg'.
24731 * var-tracking.c (vt_add_function_parameter): Test the presence of
24732 HAVE_window_save properly and do not remap argument registers when
24733 we have a leaf function.
24734
6edc3e32
UB		 247352013-02-07 Uros Bizjak <ubizjak@gmail.com>
24736
24737 PR bootstrap/56227
24738 * ggc-page.c (ggc_print_statistics): Use HOST_LONG_LONG_FORMAT
24739 instead of "ll".
24740 * config/i386/i386.c (ix86_print_operand): Ditto.
24741
5306401f
VM		 247422013-02-07 Vladimir Makarov <vmakarov@redhat.com>
24743
6edc3e32 24744 * lra-constraints.c (process_alt_operands): Fix recently added comment.
5306401f 24745
027ece11
VM		 247462013-02-07 Vladimir Makarov <vmakarov@redhat.com>
24747
24748 PR rtl-optimization/56225
24749 * lra-constraints.c (process_alt_operands): Check that reload hard
24750 reg can hold value for strict_low_part.
24751
f980dfdb
JJ		 247522013-02-07 Jakub Jelinek <jakub@redhat.com>
24753
24754 PR debug/56154
24755 * dwarf2out.c (dwarf2_debug_hooks): Set end_function hook to
24756 dwarf2out_end_function.
24757 (in_first_function_p, maybe_at_text_label_p,
24758 first_loclabel_num_not_at_text_label): New variables.
24759 (dwarf2out_var_location): In the first function find out
6edc3e32 24760 lowest loclabel_num N where .LVLN is known not to be equal to .Ltext0.
f980dfdb
JJ		 24761 (find_empty_loc_ranges_at_text_label, dwarf2out_end_function): New
24762 functions.
24763
73dd3123
EB		 247642013-02-07 Eric Botcazou <ebotcazou@adacore.com>
24765
24766 PR rtl-optimization/56178
24767 * cse.c (cse_insn): Do not create a REG_EQUAL note if the source is a
24768 SUBREG of a register. Tidy up related block of code.
24769 * fwprop.c (forward_propagate_and_simplify): Do not create a REG_EQUAL
24770 note if the source is a register or a SUBREG of a register.
24771
e3936f47
JJ		 247722013-02-07 Jakub Jelinek <jakub@redhat.com>
24773
24774 PR target/56228
24775 * config/rs6000/rs6000.md (ptrm): New mode attr.
24776 (call_indirect_aix<ptrsize>, call_indirect_aix<ptrsize>_nor11,
24777 call_value_indirect_aix<pttrsize>,
24778 call_value_indirect_aix<pttrsize>_nor11): Use <ptrm> instead of
24779 m in constraints.
24780
d96d674b
MH		 247812013-02-07 Michael Haubenwallner <michael.haubenwallner@salomon.at>
24782
24783 * collect2.c (main): Set aix64_flag for -G and -bsvr4 too, disable
24784 if -bnortl. Convert to strcmp and strncmp.
24785
921f2dee
AM		 247862013-02-07 Alan Modra <amodra@gmail.com>
24787
24788 PR target/54009
24789 * config/rs6000/rs6000.c (mem_operand_gpr): Check that LO_SUM
24790 addresses won't wrap when offsetting.
24791 (rs6000_secondary_reload): Provide secondary reloads needed for
24792 wrapping LO_SUM addresses.
24793
d09c7dba
TS		 247942013-02-06 Thomas Schwinge <thomas@codesourcery.com>
24795
24796 * config/gnu.h (GNU_USER_TARGET_OS_CPP_BUILTINS): Never define
24797 MACH, just __MACH__.
24798
a44bbd48
RB		 247992013-02-06 Richard Biener <rguenther@suse.de>
24800
24801 * tracer.c (tracer): Mark loops with LOOPS_NEED_FIXUP
24802 instead of calling fix_loop_structure.
24803
6d840d99
JJ		 248042013-02-06 Jakub Jelinek <jakub@redhat.com>
24805
24806 PR middle-end/56217
24807 * omp-low.c (use_pointer_for_field): Return false if
24808 lower_send_shared_vars doesn't generate any copy-out code.
24809
0f33baa9
TV		 248102013-02-06 Tom de Vries <tom@codesourcery.com>
24811
24812 PR rtl-optimization/56131
24813 * cfgrtl.c (delete_insn): Use NOTE_BASIC_BLOCK instead of BLOCK_FOR_INSN
24814 to get the bb of a NOTE_INSN_BASIC_BLOCK. Handle the case that the bb
24815 of the label is NULL. Add comment.
24816
11f1e3ab
JJ		 248172013-02-05 Jakub Jelinek <jakub@redhat.com>
24818
31502f9f
JJ		 24819 * tree.h (struct tree_decl_with_vis): Remove thread_local field.
24820
a74db9bd
JJ		 24821 PR sanitizer/55374
24822 * config/gnu-user.h (LIBTSAN_EARLY_SPEC): Define.
24823 (STATIC_LIBTSAN_LIBS): Likewise.
24824 * gcc.c (ADD_STATIC_LIBTSAN_LIBS, LIBTSAN_EARLY_SPEC): Define.
24825 (LIBTSAN_SPEC): Add ADD_STATIC_LIBTSAN_LIBS, if LIBTSAN_EARLY_SPEC
24826 is defined, don't add anything else beyond that.
24827 (SANITIZER_EARLY_SPEC, SANITIZER_SPEC): Define.
24828 (LINK_COMMAND_SPEC): Use them.
24829
11f1e3ab
JJ		 24830 PR tree-optimization/56205
24831 * tree-stdarg.c (check_all_va_list_escapes): Return true if
24832 there are any PHI nodes that set non-va_list_escape_vars SSA_NAME
24833 and some va_list_escape_vars SSA_NAME appears in some PHI argument.
24834
14ac6aa2
RB		 248352013-02-05 Richard Biener <rguenther@suse.de>
24836
24837 PR tree-optimization/53342
24838 PR tree-optimization/53185
24839 * tree-vectorizer.h (vect_check_strided_load): Remove.
24840 * tree-vect-data-refs.c (vect_enhance_data_refs_alignment): Do
24841 not disallow peeling for vectorized strided loads.
24842 (vect_check_strided_load): Make static and simplify.
24843 (vect_analyze_data_refs): Adjust.
24844 * tree-vect-stmts.c (vectorizable_load): Handle peeled loops
24845 correctly when vectorizing strided loads.
24846
6f22445a
RB		 248472013-02-05 Richard Biener <rguenther@suse.de>
24848
24849 * doc/install.texi: Refer to ISL, not PPL.
24850
39f9719e
JH		 248512013-02-05 Jan Hubicka <jh@suse.cz>
24852
ec4224ac
JH		 24853 PR tree-optimization/55789
24854 * params.def (PARAM_EARLY_INLINER_MAX_ITERATIONS): Drop to 1.
24855
248562013-02-05 Jan Hubicka <jh@suse.cz>
24857
24858 PR tree-optimization/55789
39f9719e
JH		 24859 * cgraphclones.c (cgraph_remove_node_and_inline_clones): Remove
24860 the dead call anyway.
24861
956d3b33
EB		 248622013-02-05 Eric Botcazou <ebotcazou@adacore.com>
24863
24864 PR sanitizer/55374
24865 * config/gnu-user.h (LIBASAN_EARLY_SPEC): Add missing guard.
24866
248672013-02-04 Alexander Potapenko <glider@google.com>
24868 Jack Howarth <howarth@bromo.med.uc.edu>
ae526fe5
AP		 24869 Jakub Jelinek <jakub@redhat.com>
24870
24871 PR sanitizer/55617
24872 * config/darwin.c (sort_ctor_records): Stabilized qsort
24873 on constructor priority by using original position.
24874 (finalize_ctors): New routine to sort constructors by
24875 priority before use in assemble_integer.
24876 (machopic_asm_out_constructor): Use finalize_ctors if needed.
24877
7ac3af38
JJ		 248782013-02-04 Jakub Jelinek <jakub@redhat.com>
24879
24880 PR libstdc++/54314
24881 * config/i386/winnt.c (i386_pe_assemble_visibility): Don't warn
24882 about visibility on artificial decls.
24883 * config/sol2.c (solaris_assemble_visibility): Likewise.
24884
152689dc
KT		 248852013-02-04 Kai Tietz <ktietz@redhat.com>
24886
24887 PR target/56186
24888 * config/i386/i386.c (function_value_ms_64): Add additional valtype
24889 argument and improve checking of return-argument types for 16-byte
24890 modes.
24891 (ix86_function_value_1): Add additional valtype argument on call
24892 of function_value_64.
24893 (return_in_memory_ms_64): Sync 16-byte sized mode handling with
24894 handling infunction_value_64 function.
24895
2b5987b5
MGD		 248962013-02-04 Matthew Gretton-Dann <matthew.gretton-dann@linaro.org>
24897
7ac3af38 24898 * reload.c (subst_reloads): Fix DEBUG_RELOAD build issue.
2b5987b5 24899
14597080
RB		 249002013-02-04 Richard Biener <rguenther@suse.de>
24901
24902 PR tree-optimization/56188
24903 * tree-ssa-structalias.c (label_visit): Consider case with
24904 initially non-empty points-to set.
24905 (perform_var_substitution): Dump node mapping and clean up.
24906
ed73881e
RG		 249072013-02-04 Richard Guenther <rguenther@suse.de>
24908
24909 PR lto/56168
24910 * lto-symtab.c (lto_symtab_merge_decls_1): Make non-builtin
24911 node prevail as last resort.
24912 (lto_symtab_merge_decls): Remove guard on LTRANS here.
6edc3e32 24913 (lto_symtab_prevailing_decl): Builtins are their own prevailing decl.
ed73881e 24914
139a0707
RB		 249152013-02-04 Richard Biener <rguenther@suse.de>
24916
24917 PR tree-optimization/56113
24918 * tree-ssa-structalias.c (equiv_class_lookup, equiv_class_add):
24919 Merge into ...
24920 (equiv_class_lookup_or_add): ... this.
24921 (label_visit): Adjust and fix error in previous patch.
24922 (perform_var_substitution): Adjust.
24923
50fe8924
OE		 249242013-02-03 Oleg Endo <olegendo@gcc.gnu.org>
24925
24926 * config/sh/divtab.c: Fix formatting and comments throughout the file.
24927 * config/sh/sh4-300.md: Likewise.
24928 * config/sh/sh4a.md: Likewise.
24929 * config/sh/constraints.md: Likewise.
24930 * config/sh/sh.md: Likewise.
24931 * config/sh/netbsd-elf.h: Likewise.
24932 * config/sh/predicates.md: Likewise.
24933 * config/sh/sh-protos.h: Likewise.
24934 * config/sh/ushmedia.h: Likewise.
24935 * config/sh/linux.h: Likewise.
24936 * config/sh/sh.c: Likewise.
24937 * config/sh/superh.h: Likewise.
24938 * config/sh/elf.h: Likewise.
24939 * config/sh/sh4.md: Likewise.
24940 * config/sh/sh.h: Likewise.
24941
1a04ac2b
JDA		 249422013-02-03 John David Anglin <dave.anglin@nrc-cnrc.gc.ca>
24943
24944 * config/pa/constraints.md: Adjust unused letters. Change "T"
24945 constraint to match_test floating_point_store_memory_operand().
24946 * config/pa/predicates.md (reg_plus_base_memory_operand): New.
24947 (base14_operand): New.
24948 (floating_point_store_memory_operand): New.
24949 (integer_store_memory_operand): Revise to use base14_operand and
24950 reg_plus_base_memory_operand.
24951 (move_dest_operand): Allow symbolic_memory_operands.
24952 (symbolic_memory_operand): Check for LO_SOM.
24953 (symbolic_operand): Change default case to break.
24954 * config/pa/pa.md: Remove unamed DFmode and SFmode patterns to force
24955 CONST_DOUBLE values to be reloaded by putting them into memory when
24956 the destination is a floating point register.
24957 (movdf): Remove code to handle CONST_DOUBLE.
24958 (movsf): Likewise.
24959 (reload_indf_r1): New.
24960 (reload_insf_r1): New.
24961 Consistently use "Q" and "T" constraints with integer and floating
24962 point move instructions, respectively.
24963 (movdi): Remove FAIL.
24964 Change predicate for source operand unamed DImode move from
24965 general_operand to move_src_operand.
24966 (umulsidi3): Change predicate for destination operand to
24967 register_operand.
24968 Likewise for similar unamed patterns.
24969 * config/pa/pa-protos.h (pa_legitimize_reload_address): Declare.
24970 * config/pa/pa.c (pa_symbolic_expression_p): Remove extra parenthesis.
24971 (hppa_legitimize_address): Simplify mask calculation.
24972 (pa_emit_move_sequence): Revised handling of secondary reloads from
24973 REG+D addresses for floating point loads and stores. Directly handle
24974 loading CONST0_RTX (mode) to a floating point register.
24975 (pa_secondary_reload): Handle reloading DF and SFmode constant values
24976 to floating point registers. Don't restrict secondary reloads to
24977 floating point registers to integer modes. Revise some comments and
24978 cleanup some code.
24979 (TARGET_LEGITIMATE_ADDRESS_P): Define.
24980 (pa_legitimate_address_p): New.
24981 (pa_legitimize_reload_address): New.
24982 * config/pa/pa.h (STRICT_REG_OK_FOR_INDEX_P): New.
24983 (STRICT_REG_OK_FOR_BASE_P): New.
24984 (GO_IF_LEGITIMATE_ADDRESS): Delete. Update some related comments.
24985 (LEGITIMIZE_RELOAD_ADDRESS): Revise to use pa_legitimize_reload_address.
24986
3bf65591
DE		 249872013-02-03 David Edelsohn <dje.gcc@gmail.com>
24988 Andrew Dixie <andrewd@gentrack.com>
24989
24990 * collect2.c (GCC_CHECK_HDR): Do not scan objects with F_LOADONLY
24991 flag set.
24992
c0a8a3e6
RS		 249932013-02-03 Richard Sandiford <rdsandiford@googlemail.com>
24994
24995 * expmed.c (extract_bit_field_1): Pass the full width of the
24996 structure to get_best_reg_extraction_insn.
24997
99113dff
DE		 249982013-02-01 David Edelsohn <dje.gcc@gmail.com>
24999
25000 PR target/54601
25001 * configure.ac (use_cxa_atexit): Add AIX.
25002 * configure: Regenerate.
25003
25004 * config/rs6000/aix61.h (STARTFILE_SPEC): Add crtcxa.o.
25005
ddd84654
JJ		 250062013-02-01 Jakub Jelinek <jakub@redhat.com>
25007
25008 PR debug/54793
25009 * final.c (need_profile_function): New variable.
25010 (final_start_function): Drop ATTRIBUTE_UNUSED from first argument.
25011 If first of NOTE_INSN_BASIC_BLOCK or NOTE_INSN_FUNCTION_BEG
25012 is only preceeded by NOTE_INSN_VAR_LOCATION or NOTE_INSN_DELETED
25013 notes, targetm.asm_out.function_prologue doesn't emit anything,
25014 HAVE_prologue and profiler should be emitted before prologue,
25015 set need_profile_function instead of emitting it.
25016 (final_scan_insn): If need_profile_function, emit
25017 profile_function on the first NOTE_INSN_BASIC_BLOCK or
25018 NOTE_INSN_FUNCTION_BEG note.
25019
250202013-02-01 Richard Henderson <rth@redhat.com>
749af8ee
RH		 25021
25022 * config/rs6000/rs6000.md (smulditi3): New.
25023 (umulditi3): New.
25024
ff2a9d88
RH		 25025 * config/alpha/alpha.md (umulditi3): New.
25026
14d52b90
DE		 250272013-02-01 David Edelsohn <dje.gcc@gmail.com>
25028
25029 * config/rs6000/xcoff.h (ASM_OUTPUT_ALIGNED_COMMON): Use floor_log2.
25030 (ASM_OUTPUT_ALIGNED_LOCAL): New.
25031
8c7ca45c
RB		 250322013-02-01 Richard Biener <rguenther@suse.de>
25033
25034 PR tree-optimization/56113
25035 * tree-ssa-structalias.c (label_visit): Reduce work for
25036 single-predecessor nodes.
25037
9f419393
EB		 250382013-02-01 Eric Botcazou <ebotcazou@adacore.com>
25039
25040 * fold-const.c (make_range_step) <TRUTH_NOT_EXPR>: Bail out if the
25041 range isn't testing for zero.
25042
c5f4be84
SB		 250432013-01-31 Steven Bosscher <steven@gcc.gnu.org>
25044
25045 PR middle-end/56113
25046 * fwprop.c (fwprop_init): Set up loops without CFG modifications.
25047
dbdbd982
NC		 250482013-01-31 Hiroyuki Ono <hiroyuki.ono.jc@renesas.com>
25049 Nick Clifton <nickc@redhat.com>
25050
25051 * config/v850/constraints.md (Q): Define as a memory constraint.
25052 * config/v850/predicates.md (label_ref_operand): New predicate.
25053 (e3v5_shift_operand): New predicate.
25054 (ior_operator): New predicate.
25055 * config/v850/t-v850: Add e3v5 multilib.
25056 * config/v850/v850-protos.h (v850_adjust_insn_length): Prototype.
25057 (v850_gen_movdi): Prototype.
25058 * config/v850/v850.c: Add support for e3v5 architecture.
25059 Rename all uses of TARGET_V850E || TARGET_V850E2_ALL to
25060 TARGET_V850E_UP.
25061 (construct_save_jarl): Add e3v5 long JARL support.
25062 (v850_adjust_insn_length): New function. Adjust length of call
25063 insns when using e3v5 instructions.
25064 (v850_gen_movdi): New function: Generate instructions to move a
25065 DImode value.
25066 * config/v850/v850.h (TARGET_CPU_v850e3v5): Define.
25067 (CPP_SPEC): Define __v850e3v5__ as appropriate.
25068 (TARGET_USE_FPU): Enable for e3v5.
25069 (CONST_OK_FOR_W): New macro.
25070 (ADJUST_INSN_LENGTH): Define.
25071 * config/v850/v850.md (UNSPEC_LOOP): Define.
25072 (attr cpu): Add v850e3v5.
25073 Rename all uses of TARGET_V850E2 to TARGET_V850E2V3_UP.
25074 (movdi): New pattern.
25075 (movdi_internal): New pattern.
25076 (cbranchsf4): Conditionalize on TARGET_USE_FPU.
25077 (cbranchdf4): Conditionalize on TARGET_USE_FPU.
25078 (cstoresf4): Likewise.
25079 (cstoredf4): Likewise.
25080 (insv): New pattern.
25081 (rotlso3_a): New pattern.
25082 (rotlsi3_b): New pattern
25083 (rotlsi3_v850e3v5): New pattern.
25084 (doloop_begin): New pattern.
25085 (fix_loop_counter): New pattern.
25086 (doloop_end): New pattern.
25087 (branch_normal): Add e3v5 long branch support.
25088 (branch_invert): Likewise.
25089 (branch_z_normal): Likewise.
25090 (branch_z_invert): Likewise.
25091 (branch_nz_normal): Likewise.
25092 (branch_nz_invert): Likewise.
25093 (call_internal_short): Add e3v5 register-indirect JARL support.
25094 (call_internal_long): Likewise.
25095 (call_value_internal_short): Likewise.
25096 (call_value_internal_long): Likewise.
25097 * config/v850/v850.opt (mv850e3v5, mv850e2v4): New options.
25098 (mloop): New option.
25099 * config.gcc: Add support for configuring v840e3v5 target.
6edc3e32 25100 * doc/invoke.texi: Document new v850 specific command line options.
dbdbd982 25101
73861a41
PK		 251022013-01-31 Paul Koning <ni1d@arrl.net>
25103
25104 PR debug/55059
25105 PR debug/54508
25106 * dwarf2out.c (prune_unused_types_mark): Mark all of parent's
25107 children if parent is a class.
25108 (prune_unused_types_prune): Don't add DW_AT_declaration.
25109
e44978dc
RB		 251102013-01-31 Richard Biener <rguenther@suse.de>
25111
25112 PR tree-optimization/56157
25113 * tree-vect-slp.c (vect_get_slp_defs): More thoroughly try to
25114 match up operand with SLP child.
25115
90b10dec
JM		 251162013-01-31 Jason Merrill <jason@redhat.com>
25117
04d2dadd 25118 PR debug/54410
90b10dec
JM		 25119 * dwarf2out.c (gen_struct_or_union_type_die): Always schedule template
25120 parameters the first time.
25121 (gen_scheduled_generic_parms_dies): Check completeness here.
25122
6e616110
RB		 251232013-01-31 Richard Biener <rguenther@suse.de>
25124
25125 PR middle-end/53073
25126 * common.opt (faggressive-loop-optimizations): New flag,
25127 enabled by default.
25128 * doc/invoke.texi (faggressive-loop-optimizations): Document.
25129 * tree-ssa-loop-niter.c (estimate_numbers_of_iterations_loop): Guard
25130 infer_loop_bounds_from_undefined by it.
25131
636f59cf
RB		 251322013-01-31 Richard Biener <rguenther@suse.de>
25133
25134 PR tree-optimization/56150
25135 * tree-ssa-loop-manip.c (find_uses_to_rename_stmt): Do not
25136 visit virtual operands.
25137 (find_uses_to_rename_bb): Likewise.
25138
251392013-01-31 Richard Biener <rguenther@suse.de>
fcfa87ac
RB		 25140
25141 PR tree-optimization/56150
25142 * tree-ssa-tail-merge.c (gimple_equal_p): Properly handle
25143 mixed store non-store stmts.
25144
32887460
JJ		 251452013-01-30 Jakub Jelinek <jakub@redhat.com>
25146
e60e09a0
JJ		 25147 PR sanitizer/55374
25148 * gcc.c (LIBASAN_SPEC): Define just to ADD_STATIC_LIBASAN_LIBS if
25149 LIBASAN_EARLY_SPEC is defined.
25150 (LIBASAN_EARLY_SPEC): Define to empty string if not already defined.
25151 (LINK_COMMAND_SPEC): Add LIBASAN_EARLY_SPEC for -fsanitize=address,
25152 before %o.
25153 * config/gnu-user.h (LIBASAN_EARLY_SPEC): Define.
25154
32887460
JJ		 25155 PR c++/55742
25156 * config/i386/i386.c (ix86_valid_target_attribute_inner_p): Diagnose
25157 invalid args instead of ICEing on it.
25158 (ix86_valid_target_attribute_tree): Return error_mark_node if
25159 ix86_valid_target_attribute_inner_p failed.
25160 (ix86_valid_target_attribute_p): Return false only if
25161 ix86_valid_target_attribute_tree returned error_mark_node. Allow
25162 target("default") attribute.
25163 (sorted_attr_string): Change argument from const char * to tree,
25164 merge in all target attribute arguments rather than just one.
25165 Formatting fix. Use XNEWVEC instead of xmalloc and XDELETEVEC
25166 instead of free. Avoid using strcat.
25167 (ix86_mangle_function_version_assembler_name): Mangle
25168 target("default") as if no target attribute is present. Adjust
25169 sorted_attr_string caller. Avoid leaking memory. Use XNEWVEC
25170 instead of xmalloc and XDELETEVEC instead of free.
25171 (ix86_function_versions): Don't return true if one of the decls
25172 doesn't have target attribute. If they don't and one of the decls
25173 is DECL_FUNCTION_VERSIONED, report an error. Adjust
25174 sorted_attr_string caller. Use XDELETEVEC instead of free.
25175 (ix86_supports_function_versions): Remove.
25176 (make_name): Fix up formatting.
25177 (make_dispatcher_decl): Remove resolver_name and its initialization.
25178 Avoid leaking memory.
25179 (is_function_default_version): Return true if there is
25180 target("default") attribute rather than no target attribute at all.
25181 (make_resolver_func): Avoid leaking memory.
25182 (ix86_generate_version_dispatcher_body): Likewise.
25183 (TARGET_OPTION_SUPPORTS_FUNCTION_VERSIONS): Remove.
25184 * target.def (supports_function_versions): Remove.
25185 * doc/tm.texi.in (SUPPORTS_FUNCTION_VERSIONS): Remove.
25186 * doc/tm.texi: Regenerated.
25187
73cca0cc
VM		 251882013-01-30 Vladimir Makarov <vmakarov@redhat.com>
25189
25190 PR rtl-optimization/56144
25191 * lra-constraints.c (get_reload_reg): Don't reuse reload pseudo
25192 for values with side effects.
25193
11452e7b
RB		 251942013-01-30 Richard Biener <rguenther@suse.de>
25195
25196 * sparseset.h (sparseset_bit_p): Use gcc_checking_assert.
25197 (sparseset_pop): Likewise.
25198 * cfganal.c (compute_idf): Likewise. Increase work-stack size
25199 to be able to use quick_push in the worker loop.
25200
01cb1ef5
MP		 252012013-01-30 Marek Polacek <polacek@redhat.com>
25202
25203 * cfgcleanup.c (cleanup_cfg): Don't mark affected BBs.
25204
e892936e
RB		 252052013-01-30 Richard Biener <rguenther@suse.de>
25206
25207 PR lto/56147
6edc3e32 25208 * lto-symtab.c (lto_symtab_merge_decls_1): Guard DECL_BUILT_IN check.
e892936e 25209
cc06c01d
GJL		 252102013-01-30 Georg-Johann Lay <avr@gjlay.de>
25211
25212 PR tree-optimization/56064
25213 * fixed-value.c (fixed_from_double_int): New function.
25214 * fixed-value.h (fixed_from_double_int): New prototype.
25215 (const_fixed_from_double_int): New static inline function.
25216 * fold-const.c (native_interpret_fixed): New static function.
25217 (native_interpret_expr) <FIXED_POINT_TYPE>: Use it.
25218 (can_native_interpret_type_p) <FIXED_POINT_TYPE>: Return true.
25219 (native_encode_fixed): New static function.
25220 (native_encode_expr) <FIXED_CST>: Use it.
25221 (native_interpret_int): Move double_int worker code to...
25222 * double-int.c (double_int::from_buffer): ...this new static method.
25223 * double-int.h (double_int::from_buffer): Prototype it.
25224
d394a308
RB		 252252013-01-30 Richard Biener <rguenther@suse.de>
25226
25227 * tree-ssa-structalias.c (final_solutions, final_solutions_obstack):
25228 New pointer-map and obstack.
25229 (init_alias_vars): Allocate pointer-map and obstack.
25230 (delete_points_to_sets): Free them.
25231 (find_what_var_points_to): Cache result.
25232 (find_what_p_points_to): Adjust for changed interface of
25233 find_what_var_points_to.
25234 (compute_points_to_sets): Likewise.
25235 (ipa_pta_execute): Likewise.
25236
20804d96
RO		 252372013-01-30 Rainer Orth <ro@CeBiTec.Uni-Bielefeld.DE>
25238
25239 * configure.ac (HAVE_AS_SPARC_NOBITS): New test.
25240 * configure: Regenerate.
25241 * config.in: Regenerate.
25242 * config/sparc/sparc.c (sparc_solaris_elf_asm_named_section): Emit
25243 #nobits/#progbits if supported.
25244
a7ad88a2
OE		 252452013-01-29 Oleg Endo <olegendo@gcc.gnu.org>
25246
25247 PR target/56121
25248 * config/sh/sh.md (bclr_m2a, bset_m2a, bst_m2a, bld_m2a, bldsign_m2a,
25249 bld_reg, *bld_regqi, band_m2a, bandreg_m2a, bor_m2a, borreg_m2a,
25250 bxor_m2a, bxorreg_m2a): Add satisfies_constraint_K03 condition.
25251
7af79f92
GY		 252522013-01-29 Greta Yorsh <Greta.Yorsh@arm.com>
25253
91bfca59
OE		 25254 * config/arm/cortex-a7.md (cortex_a7_neon, cortex_a7_all): Remove.
25255 (cortex_a7_idiv): Use cortex_a7_both instead of cortex_a7_all.
7af79f92 25256
753bcf7b
GY		 252572013-01-29 Greta Yorsh <Greta.Yorsh@arm.com>
25258
91bfca59
OE		 25259 * config/arm/arm.c (cortexa7_younger): Return true for TYPE_CALL.
25260 * config/arm/cortex-a7.md (cortex_a7_call): Update required units.
753bcf7b 25261
8cbc2ea8
GY		 252622013-01-29 Greta Yorsh <Greta.Yorsh@arm.com>
25263
91bfca59
OE		 25264 * config/arm/arm-protos.h (arm_mac_accumulator_is_result): New
25265 declaration.
25266 * config/arm/arm.c (arm_mac_accumulator_is_result): New function.
25267 * config/arm/cortex-a7.md: New bypasses using
25268 arm_mac_accumulator_is_result.
8cbc2ea8 25269
697a3325
GY		 252702013-01-29 Greta Yorsh <Greta.Yorsh@arm.com>
25271
91bfca59 25272 * config/arm/cortex-a7.md (cortex_a7_neon_mul): New reservation.
697a3325 25273 (cortex_a7_neon_mla): Likewise.
91bfca59
OE		 25274 (cortex_a7_fpfmad): New reservation.
25275 (cortex_a7_fpmacs): Use ffmas and update required units.
25276 (cortex_a7_fpmuld): Update required units and latency.
25277 (cortex_a7_fpmacd): Likewise.
25278 (cortex_a7_fdivs, cortex_a7_fdivd): Likewise.
25279 (cortex_a7_neon). Likewise.
25280 (bypass) Update participating units.
697a3325 25281
29637783
GY		 252822013-01-29 Greta Yorsh <Greta.Yorsh@arm.com>
25283
91bfca59
OE		 25284 * config/arm/arm.md (type): Add ffmas and ffmad to "type" attribute.
25285 * config/arm/vfp.md (fma,fmsub,fnmsub,fnmadd): Change type
25286 from fmac to ffma.
25287 * config/arm/vfp11.md (vfp_farith): Use ffmas.
25288 (vfp_fmul): Use ffmad.
25289 * config/arm/cortex-r4f.md (cortex_r4_fmacs): Use ffmas.
25290 (cortex_r4_fmacd): Use ffmad.
25291 * config/arm/cortex-m4-fpu.md (cortex_m4_fmacs): Use ffmas.
25292 * config/arm/cortex-a9.md (cortex_a9_fmacs): Use ffmas.
25293 (cortex_a9_fmacd): Use ffmad.
25294 * config/arm/cortex-a8-neon.md (cortex_a8_vfp_macs): Use ffmas.
25295 (cortex_a8_vfp_macd): Use ffmad.
25296 * config/arm/cortex-a5.md (cortex_a5_fpmacs): Use ffmas.
25297 (cortex_a5_fpmacd): Use ffmad.
25298 * config/arm/cortex-a15-neon.md (cortex_a15_vfp_macs) Use ffmas.
25299 (cortex_a15_vfp_macd): Use ffmad.
25300 * config/arm/arm1020e.md (v10_fmul): Use ffmas and ffmad.
29637783 25301
2ee8a2d5
JM		 253022013-01-29 Jason Merrill <jason@redhat.com>
25303
25304 PR libstdc++/54314
25305 * varasm.c (default_assemble_visibility): Don't warn about
25306 visibility on artificial decls.
25307
78d087bc
RB		 253082013-01-29 Richard Biener <rguenther@suse.de>
25309
25310 PR tree-optimization/56113
25311 * tree-ssa-structalias.c (equiv_class_lookup): Also return
25312 the bitmap leader.
25313 (label_visit): Free duplicate bitmaps and record the leader instead.
25314 (perform_var_substitution): Adjust.
25315
83ba4d6f
RB		 253162013-01-29 Richard Biener <rguenther@suse.de>
25317
25318 PR tree-optimization/55270
25319 * tree-ssa-dom.c (eliminate_degenerate_phis): If we changed
25320 the CFG, schedule loops for fixup.
25321
66dfe13f
NC		 253222013-01-29 Nick Clifton <nickc@redhat.com>
25323
25324 * config/rl78/rl78.c (rl78_regno_mode_code_ok_for_base_p): Allow
25325 SP_REG.
25326
5a579c3b
LE		 253272013-01-28 Leif Ekblad <leif@rdos.net>
25328
25329 * config.gcc (i[34567]86-*-rdos*, x86_64-*-rdos*): New targets.
25330 * config/i386/i386.h (TARGET_RDOS): New macro.
25331 (DEFAULT_LARGE_SECTION_THRESHOLD): New macro.
25332 * config/i386/i386.c (ix86_option_override_internal): For 64bit
25333 TARGET_RDOS, set ix86_cmodel to CM_MEDIUM_PIC and flag_pic to 1.
25334 * config/i386/i386.opt (mlarge-data-threshold): Initialize to
25335 DEFAULT_LARGE_SECTION_THRESHOLD.
25336 * config/i386/i386.md (R14_REG, R15_REG): New constants.
25337 * config/i386/rdos.h: New file.
25338 * config/i386/rdos64.h: New file.
25339
51e44392
BS		 253402013-01-28 Bernd Schmidt <bernds@codesourcery.com>
25341
25342 PR other/54814
25343 * reload.c (find_valid_class_1): Use in_hard_reg_set_p instead of
25344 TEST_HARD_REG_BIT.
25345
db1fb332
JJ		 253462013-01-28 Jakub Jelinek <jakub@redhat.com>
25347
25348 PR rtl-optimization/56117
25349 * sched-deps.c (sched_analyze_2) <case PREFETCH>: For use_cselib
25350 call cselib_lookup_from_insn on the MEM before calling
25351 add_insn_mem_dependence.
25352
16917761
RB		 253532013-01-28 Richard Biener <rguenther@suse.de>
25354
25355 * tree-inline.c (remap_gimple_stmt): Do not assing a BLOCK
25356 to a stmt that didn't have one.
25357 (copy_phis_for_bb): Likewise for PHI arguments.
25358 (copy_debug_stmt): Likewise for debug stmts.
25359
b9fc0497
RB		 253602013-01-28 Richard Biener <rguenther@suse.de>
25361
25362 PR tree-optimization/56034
6edc3e32 25363 * tree-loop-distribution.c (enum partition_kind): Add PKIND_REDUCTION.
b9fc0497
RB		 25364 (partition_builtin_p): Adjust.
25365 (generate_code_for_partition): Handle PKIND_REDUCTION. Assert
25366 it is the last partition.
25367 (rdg_flag_uses): Check SSA_NAME_IS_DEFAULT_DEF before looking
25368 up the vertex for the definition.
25369 (classify_partition): Classify whether a partition is a
25370 PKIND_REDUCTION, thus has uses outside of the loop.
25371 (ldist_gen): Inherit PKIND_REDUCTION when merging partitions.
25372 Merge all PKIND_REDUCTION partitions into the last partition.
25373 (tree_loop_distribution): Seed partitions from reductions as well.
25374
aa710d25
JJ		 253752013-01-28 Jakub Jelinek <jakub@redhat.com>
25376
0bfbca58
JJ		 25377 PR tree-optimization/56125
25378 * tree-ssa-math-opts.c (gimple_expand_builtin_pow): Don't optimize
25379 pow(x,c) into sqrt(x) * powi(x, n/2) or
25380 1.0 / (sqrt(x) * powi(x, abs(n/2))) if c is an integer or when
25381 optimizing for size.
25382 Don't optimize pow(x,c) into powi(x, n/3) * powi(cbrt(x), n%3) or
25383 1.0 / (powi(x, abs(n)/3) * powi(cbrt(x), abs(n)%3)) if 2c is an
25384 integer.
25385
aa710d25
JJ		 25386 PR tree-optimization/56094
25387 * gimplify.c (force_gimple_operand_1): Temporarily set input_location
25388 to UNKNOWN_LOCATION while gimplifying expr.
25389
77dc5297
UB		 253902013-01-27 Uros Bizjak <ubizjak@gmail.com>
25391
25392 PR target/56114
25393 * config/i386/i386.md (*movabs<mode>_1): Add square brackets around
25394 operand 0 in movabs insn template for -masm=intel asm alternative.
25395 (*movabs<mode>_2): Ditto for operand 1.
25396
0bfbca58 253972013-01-26 David Holsgrove <david.holsgrove@xilinx.com>
fbf0cf90
DH		 25398
25399 PR target/54663
25400 * config.gcc (microblaze*-linux*): Add tmake_file to allow building
25401 of microblaze-c.o
25402
0bfbca58 254032013-01-26 Edgar E. Iglesias <edgar.iglesias@gmail.com>
6f14eef2
EI		 25404
25405 * config.gcc (microblaze*-*-*): Rename microblaze*-*-elf, update
25406 tm_file.
25407
0bfbca58 254082013-01-25 Naveen H.S <Naveen.Hurugalawadi@caviumnetworks.com>
706b2314
N		 25409
25410 * config/aarch64/aarch64.c (TARGET_FIXED_CONDITION_CODE_REGS):
25411 Undef to avoid warning.
25412
478f60f9
MH		 254132013-01-25 Michael Haubenwallner <michael.haubenwallner@salomon.at>
25414
25415 * configure.ac (gcc_cv_ld_static_dynamic): Define for AIX native ld.
25416 * configure: Regenerate.
25417
d7fa6ee2
JJ		 254182013-01-25 Jakub Jelinek <jakub@redhat.com>
25419
25420 PR tree-optimization/56098
25421 * tree-ssa-phiopt.c (nt_init_block): Don't call add_or_mark_expr
25422 for stmts with volatile ops.
25423 (cond_store_replacement): Don't optimize if assign has volatile ops.
25424 (cond_if_else_store_replacement_1): Don't optimize if either
25425 then_assign or else_assign have volatile ops.
25426 (hoist_adjacent_loads): Don't optimize if either def1 or def2 have
25427 volatile ops.
25428
f8fe87bd
GJL		 254292013-01-25 Georg-Johann Lay <avr@gjlay.de>
25430
25431 * doc/invoke.texi (AVR Built-in Macros): Document __XMEGA__.
25432
28398d0d
GJL		 254332013-01-25 Georg-Johann Lay <avr@gjlay.de>
25434
25435 * doc/extend.texi (Example of asm with clobbered asm reg): Fix
25436 missing ':' in asm example.
25437
b7d7d917
TB		 254382013-01-25 Tejas Belagod <tejas.belagod@arm.com>
25439
25440 * config/aarch64/aarch64-simd-builtins.def: Separate sq<r>dmulh_lane
25441 entries into lane and laneq entries.
77dc5297
UB		 25442 * config/aarch64/aarch64-simd.md (aarch64_sq<r>dmulh_lane<mode>):
25443 Remove AdvSIMD scalar modes.
b7d7d917
TB		 25444 (aarch64_sq<r>dmulh_laneq<mode>): New.
25445 (aarch64_sq<r>dmulh_lane<mode>): New RTL pattern for Scalar AdvSIMD
25446 modes.
25447 * config/aarch64/arm_neon.h: Fix all the vq<r>dmulh_lane* intrinsics'
25448 builtin implementations to relfect changes in RTL in aarch64-simd.md.
25449 * config/aarch64/iterators.md (VCOND): New.
25450 (VCONQ): New.
25451
556f9906
GJL		 254522013-01-25 Georg-Johann Lay <avr@gjlay.de>
25453
25454 PR target/54222
25455 * config/avr/builtins.def (DEF_BUILTIN): Add LIBNAME argument.
25456 Add NULL LIBNAME argument to existing definitions.
25457 (ABSHR, ABSR, ABSLR, ABSLLR, ABSHK, ABSK, ABSLK, ABSLLK): New.
25458 * config/avr/avr-c.c (DEF_BUILTIN): Add LIBNAME argument.
25459 * config/avr/avr.c (DEF_BUILTIN): Same.
25460 (avr_init_builtins): Pass down LIBNAME to add_builtin_function.
25461 (avr_expand_builtin): Expand to a vanilla call if a libgcc
25462 implementation is available (DECL_ASSEMBLER_NAME is set).
25463 (avr_fold_absfx): New static function.
25464 (avr_fold_builtin): Use it to handle: AVR_BUILTIN_ABSHR,
25465 AVR_BUILTIN_ABSR, AVR_BUILTIN_ABSLR, AVR_BUILTIN_ABSLLR,
25466 AVR_BUILTIN_ABSHK, AVR_BUILTIN_ABSK, AVR_BUILTIN_ABSLK,
25467 AVR_BUILTIN_ABSLLK.
25468 * config/avr/stdfix.h (abshr, absr, abslr, absllr)
25469 (abshk, absk, abslk, absllk): Provide as static inline functions.
25470
1f546bbb
MP		 254712013-01-25 Marek Polacek <polacek@redhat.com>
25472
25473 PR tree-optimization/56035
25474 * cfgloopmanip.c (fix_loop_structure): Remove redundant condition.
25475
7e184bd7
UB		 254762012-01-24 Uros Bizjak <ubizjak@gmail.com>
25477
25478 * config/i386/i386.md (*movti_internal_rex64): Add (o,e) alternative.
25479 (*movtf_internal_rex64): Add (!o,C) alternative
25480 (*movxf_internal_rex64): Ditto.
25481 (*movdf_internal_rex64): Add (?r,C) and (?m,C) alternatives.
25482
f5ba49ac
SM		 254832013-01-24 Shenghou Ma <minux.ma@gmail.com>
25484
25485 * doc/invoke.texi: fix typo.
25486 * doc/objc.texi: fix typo.
25487
a6343728
RS		 254882013-01-24 Richard Sandiford <rdsandiford@googlemail.com>
25489
25490 * config/mips/mips.md (*and<mode>3_mips16): Use the "W" constraint
25491 for the first two alternatives.
25492
cd030c07
DN		 254932013-01-24 Diego Novillo <dnovillo@google.com>
25494
77dc5297 25495 * Makefile.in (GGC): Remove. Replace all instances with ggc-page.o.
cd030c07
DN		 25496 (ggc-zone.o): Remove.
25497 * configure.ac: Remove option --with-gc.
25498 * configure: Re-generate.
25499 * doc/install.texi: Remove documentation for --with-gc.
25500 * gengtype.c (write_enum_defn): Remove. Update all users.
25501 (write_Types_process_field): Remove generation of gt_e_* argument.
25502 (output_type_enum): Remove. Update all users.
25503 (write_enum_defn): Remove. Update all users.
25504 (enum alloc_zone): Remove. Update all users.
77dc5297 25505 (write_splay_tree_allocator_def): Remove generation of gt_e_* argument.
cd030c07
DN		 25506 * ggc-common.c (ggc_splay_alloc): Remove first argument.
25507 Update all callers.
25508 (struct ptr_data): Remove field TYPE. Update all users.
77dc5297 25509 (gt_pch_note_object): Remove argument TYPE. Update all users.
cd030c07
DN		 25510 * ggc-internal.h (ggc_pch_alloc_object): Remove last argument.
25511 Update all users.
aaf1e810 25512 * ggc-none.c (ggc_alloc_typed_stat): Remove.
cd030c07
DN		 25513 (struct alloc_zone): Remove.
25514 (ggc_internal_alloc_zone_stat): Remove.
25515 (ggc_internal_cleared_alloc_zone_stat): Remove.
25516 * ggc-page.c (ggc_alloc_typed_stat): Remove.
77dc5297
UB		 25517 (ggc_pch_count_object): Remove last argument. Update all users.
25518 (ggc_pch_alloc_object): Remove last argument. Update all users.
cd030c07
DN		 25519 (struct alloc_zone): Remove.
25520 * ggc-zone.c: Remove.
77dc5297 25521 * ggc.h (gt_pch_note_object): Remove last argument. Update all users.
cd030c07
DN		 25522 (struct alloc_zone): Remove.
25523 (ggc_alloc_typed_stat): Remove.
25524 (ggc_alloc_typed): Remove.
25525 (ggc_splay_alloc): Remove first argument.
25526 (rtl_zone): Remove. Update all users.
25527 (tree_zone): Remove. Update all users.
25528 (tree_id_zone): Remove. Update all users.
25529 (ggc_internal_zone_alloc_stat): Remove. Update all users.
77dc5297 25530 (ggc_internal_zone_cleared_alloc_stat): Remove. Update all users.
cd030c07 25531 (ggc_internal_zone_vec_alloc_stat): Remove. Update all users.
77dc5297 25532 * tree-ssanames.c: Remove references to zone allocator in comments.
cd030c07 25533
a861ffa4
GJL		 255342013-01-24 Georg-Johann Lay <avr@gjlay.de>
25535
25536 * config/avr/avr.c (avr_out_fract): Make register numbers that
25537 might be outside of source operand signed.
25538
593c0ddd
UB		 255392013-01-24 Uros Bizjak <ubizjak@gmail.com>
25540
25541 * config/i386/constraints.md (Yf): New constraint.
25542 * config/i386/i386.md (*movdf_internal_rex64): Use Yf*f instead
25543 of f constraint to conditionaly disable x87 register preferences.
25544 (*movdf_internal): Ditto.
25545 (*movsf_internal): Ditto.
25546
e86c0101
SB		 255472013-01-24 Steven Bosscher <steven@gcc.gnu.org>
25548
25549 PR inline-asm/55934
25550 * lra-assigns.c (assign_by_spills): Throw away the pattern of asms
25551 that have operands with impossible constraints.
25552 Add a FIXME for a speed-up opportunity.
25553 * lra-constraints.c (process_alt_operands): Verify that a class
25554 selected from constraints on asms is valid for the operand mode.
25555 (curr_insn_transform): Remove incorrect comment.
25556
f6fee35f
DE		 255572013-01-23 David Edelsohn <dje.gcc@gmail.com>
25558
25559 * config/rs6000/rs6000.c (rs6000_delegitimize_address): Check that
25560 TOC operand is a valid symbol ref in the constant pool.
25561
255622013-01-23 Edgar E. Iglesias <edgar.iglesias@gmail.com>
c7729184 25563
aaf1e810 25564 * config/microblaze/linux.h: Add TARGET_OS_CPP_BUILTINS
c7729184 25565
dc62d7d1
GJL		 255662013-01-23 Georg-Johann Lay <avr@gjlay.de>
25567
25568 PR target/54222
25569 * config/avr/stdfix.h: New file.
25570 * t-avr (stdfix-gcc.h): New rule to build it.
25571 (EXTRA_HEADERS): Set it to install stdfix.h, stdfix-gcc.h.
25572
2660d12d
KS		 255732013-01-23 Kostya Serebryany <kcc@google.com>
25574
77dc5297
UB		 25575 * config/darwin.h: remove dependency on
25576 CoreFoundation (asan on Mac OS).
2660d12d 25577
a70418fc
JJ		 255782013-01-23 Jakub Jelinek <jakub@redhat.com>
25579
25580 PR target/49069
25581 * config/arm/arm.md (cbranchdi4, cstoredi4): Use s_register_operand
25582 instead of cmpdi_operand for first comparison operand.
25583 Don't assert that comparison operands aren't both constants.
25584
47876a2a
JW		 255852013-01-22 Jonathan Wakely <jwakely.gcc@gmail.com>
25586
25587 * doc/install.texi (Downloading the Source): Update references to
25588 downloading separate components.
25589
255902013-01-22 Jonathan Wakely <jwakely.gcc@gmail.com>
4ebd56a5
JW		 25591
25592 * doc/extend.texi (__int128): Improve grammar.
25593
47876a2a 255942013-01-22 Uros Bizjak <ubizjak@gmail.com>
32fc5b8a
UB		 25595
25596 PR target/56028
25597 * config/i386/i386.md (*movti_internal_rex64): Change (o,riF)
25598 alternative to (o,r).
25599 (*movdi_internal_rex64): Remove (!o,n) alternative.
25600 (DImode immediate->memory splitter): Remove.
25601 (DImode immediate->memory peephole2): Remove.
25602 (movtf): Enable for TARGET_64BIT || TARGET_SSE.
25603 (*movtf_internal_rex64): Rename from *movtf_internal. Change (!o,F*r)
25604 alternative to (!o,*r).
25605 (*movtf_internal_sse): New pattern.
25606 (*movxf_internal_rex64): New pattern.
25607 (*movxf_internal): Disable for TARGET_64BIT.
25608 (*movdf_internal_rex64): Remove (!o,F) alternative.
25609
3a984f10
JJ		 256102013-01-22 Jakub Jelinek <jakub@redhat.com>
25611
502498d5
JJ		 25612 PR middle-end/56074
25613 * dumpfile.c (dump_loc): Only print loc if LOCATION_LOCUS (loc)
25614 isn't UNKNOWN_LOCATION nor BUILTINS_LOCATION.
25615 * tree-vect-loop-manip.c (find_loop_location): Also ignore
25616 stmt locations where LOCATION_LOCUS of the stmt location is
25617 UNKNOWN_LOCATION or BUILTINS_LOCATION.
25618
3a984f10
JJ		 25619 PR target/55686
25620 * config/i386/i386.md (UNSPEC_STOS): New.
25621 (strset_singleop, *strsetdi_rex_1, *strsetsi_1, *strsethi_1,
25622 *strsetqi_1): Add UNSPEC_STOS.
25623
fa817f7f
PC		 256242013-01-22 Paolo Carlini <paolo.carlini@oracle.com>
25625
25626 PR c++/56067
25627 * doc/invoke.texi: Remove left over -Wsynth example.
25628
8f498c1b
JJ		 256292013-01-21 Jakub Jelinek <jakub@redhat.com>
25630
25631 PR tree-optimization/56051
25632 * fold-const.c (fold_binary_loc): Don't fold
25633 X < (cast) (1 << Y) into (X >> Y) != 0 if cast is either
25634 a narrowing conversion, or widening conversion from signed
25635 to unsigned.
25636
47876a2a 256372013-01-21 Uros Bizjak <ubizjak@gmail.com>
35d59da7
UB		 25638
25639 PR rtl-optimization/56023
25640 * haifa-sched.c (fix_inter_tick): Do not update ticks of instructions,
25641 dependent on debug instruction.
25642
5b9db1bc
MJ		 256432013-01-21 Martin Jambor <mjambor@suse.cz>
25644
25645 PR middle-end/56022
25646 * function.c (allocate_struct_function): Call
25647 invoke_set_current_function_hook earlier.
25648
e8bb7d68
JJ		 256492013-01-21 Jakub Jelinek <jakub@redhat.com>
25650
25651 * reload1.c (init_reload): Only initialize reload_obstack
25652 during the first call.
25653
616a4e32
MP		 256542013-01-21 Marek Polacek <polacek@redhat.com>
25655
25656 * cfgloop.c (verify_loop_structure): Fix up grammar.
25657
4401981b
YHH		 256582013-01-21 Yi-Hsiu Hsu <ahsu@marvell.com>
25659
25660 * config/arm/marvell-pj4.md (pj4_shift_conds, pj4_alu_shift,
25661 pj4_alu_shift_conds, pj4_shift): Handle simple_alu_shift.
25662
8e87740b
RR		 256632013-01-21 Ramana Radhakrishnan <ramana.radhakrishnan@arm.com>
25664
25665 PR target/56058
25666 * config/arm/marvell-pj4.md: Update copyright year.
25667 Fix up use of alu to alu_reg and simple_alu_imm.
25668
47876a2a 256692013-01-21 Uros Bizjak <ubizjak@gmail.com>
1f6bc337
UB		 25670
25671 * config/i386/i386.md (enabled): Do not disable fma4 for TARGET_FMA.
25672
89d56d79
VM		 256732013-01-20 Vladimir Makarov <vmakarov@redhat.com>
25674
25675 PR target/55433
25676 * lra-constraints.c (curr_insn_transform): Don't reuse original
1f6bc337 25677 insn for secondary memory move when memory mode should be different.
89d56d79 25678
fe603553
JDA		 256792013-01-20 John David Anglin <dave.anglin@nrc-cnrc.gc.ca>
25680
25681 * config/pa/pa.md (atomic_loaddi, atomic_loaddi_1, atomic_storedi,
25682 atomic_storedi_1): New patterns.
25683
01284895
VK		 256842013-01-20 Venkataramanan Kumar <venkataramanan.kumar@amd.com>
25685
25686 btver2 pipeline descriptions.
25687 * config/i386/i386.c: Enable CPU_BTVER2 to use btver2 pipeline
25688 descriptions.
25689 * config/i386/i386.md (btver2_decode): New type attributes.
1f6bc337 25690 * config/i386/sse.md (btver2_decode, btver2_sse_attr): New
01284895
VK		 25691 type attributes.
25692 * config/i386/btver2.md: New file describing btver2 pipelines.
25693
5630e3e1
JL		 256942013-01-19 Andrew Pinski <apinski@cavium.com>
25695
25696 PR tree-optimization/52631
25697 * tree-ssa-sccvn (visit_use): Before looking up the original
25698 statement, try looking up the simplified expression.
25699
650ae806
AG		 257002013-01-19 Anthony Green <green@moxielogic.com>
25701
25702 * config/moxie/moxie.c (moxie_expand_prologue): Set
25703 current_function_static_stack_size.
25704
e300ec2d
JJ		 257052013-01-18 Jakub Jelinek <jakub@redhat.com>
25706
25707 PR tree-optimization/56029
25708 * tree-phinodes.c (reserve_phi_args_for_new_edge): Set
25709 gimple_phi_arg_location for the new arg to UNKNOWN_LOCATION.
25710
a3d7af04
SS		 257112013-01-18 Sharad Singhai <singhai@google.com>
25712
25713 PR tree-optimization/55995
25714 * dumpfile.c (dump_loc): Print location only if available.
25715 * tree-vectorizer.c (increase_alignment): Intialize vect_location.
25716
66aa7879
VM		 257172013-01-18 Vladimir Makarov <vmakarov@redhat.com>
25718
25719 PR target/55433
25720 * lra-constraints.c (curr_insn_transform): Reuse original insn for
25721 secondary memory move.
25722 (inherit_reload_reg): Use rclass instead of cl for
25723 check_secondary_memory_needed_p.
25724
3f0fee7b
JJ		 257252013-01-18 Jakub Jelinek <jakub@redhat.com>
25726
25727 PR middle-end/56015
25728 * expr.c (expand_expr_real_2) <case COMPLEX_EXPR>: Handle
1f6bc337 25729 the case where writing real complex part of target modifies op1.
3f0fee7b 25730
70c67693
JG		 257312013-01-18 James Greenhalgh <james.greenhalgh@arm.com>
25732
25733 * config/aarch64/aarch64-simd.md
25734 (aarch64_vcond_internal<mode>): Handle unordered cases.
25735 * config/aarch64/iterators.md (v_cmp_result): New.
25736
df8de9b3
YHH		 257372013-01-18 Yi-Hsiu Hsu <ahsu@marvell.com>
25738 Ramana Radhakrishnan <ramana.radhakrishnan@arm.com>
25739
25740 * config/arm/marvell-pj4.md: New file.
25741 * config/arm/arm.c (arm_issue_rate): Add marvell_pj4.
25742 * config/arm/arm.md (generic_sched): Add marvell_pj4.
25743 (generic_vfp): Likewise.
25744 * config/arm/arm-cores.def: Add marvell-pj4.
25745 * config/arm/arm-tune.md: Regenerate.
25746 * config/arm/arm-tables.opt: Regenerate.
25747 * config/arm/bpabi.h (BE8_LINK_SPEC): Add marvell_pj4.
25748 * doc/invoke.texi: Document marvell-pj4.
25749
be30c356
TB		 257502013-01-18 Tejas Belagod <tejas.belagod@arm.com>
25751
25752 * config/aarch64/arm_neon.h: Map scalar types to standard types.
25753
0bfbca58 257542013-01-18 Alexandre Oliva <aoliva@redhat.com>
af6236c1
AO		 25755
25756 PR debug/54114
25757 PR debug/54402
25758 PR debug/49888
25759 * var-tracking.c (negative_power_of_two_p): New.
25760 (global_get_addr_cache, local_get_addr_cache): New.
25761 (get_addr_from_global_cache, get_addr_from_local_cache): New.
25762 (vt_canonicalize_addr): Rewrite using the above. Adjust the
25763 heading comment.
25764 (vt_stack_offset_p): Remove.
25765 (vt_canon_true_dep): Always canonicalize loc's address.
25766 (clobber_overlapping_mems): Make sure we have a MEM.
25767 (local_get_addr_clear_given_value): New.
25768 (val_reset): Clear local cached entries.
25769 (compute_bb_dataflow): Create and release the local cache.
25770 Disable duplicate MEMs clobbering.
25771 (emit_notes_in_bb): Clobber MEMs likewise.
25772 (vt_emit_notes): Create and release the local cache.
25773 (vt_initialize, vt_finalize): Create and release the global
25774 cache, respectively.
1f6bc337 25775 * alias.c (rtx_equal_for_memref_p): Compare operands of ENTRY_VALUEs.
af6236c1 25776
0bfbca58 257772013-01-18 Alexandre Oliva <aoliva@redhat.com>
7ff37ffa
AO		 25778
25779 PR libmudflap/53359
25780 * tree-mudflap.c (mudflap_finish_file): Skip deferred decls
25781 not found in the symtab.
25782
0bfbca58 257832013-01-18 Alexandre Oliva <aoliva@redhat.com>
3aa03517 25784
c350ba53 25785 PR debug/56006
3aa03517
AO		 25786 PR rtl-optimization/55547
25787 PR rtl-optimization/53827
25788 PR debug/53671
25789 PR debug/49888
25790 * alias.c (offset_overlap_p): New, factored out of...
25791 (memrefs_conflict_p): ... this. Use absolute sizes. Retain
25792 the conservative special case for symbolic constants. Don't
25793 adjust zero sizes on alignment.
25794
c664546f
JL		 257952013-01-18 Bernd Schmidt <bernds@codesourcery.com>
25796
25797 PR rtl-optimization/52573
25798 * regrename.c (build_def_use): Ignore REG_DEAD notes if there is a
25799 REG_UNUSED for the same register.
25800
1bd3f750
MP		 258012013-01-17 Richard Biener <rguenther@suse.de>
25802 Marek Polacek <polacek@redhat.com>
25803
25804 PR rtl-optimization/55833
25805 * loop-unswitch.c (unswitch_loops): Move loop verification...
25806 (unswitch_single_loop): ...here. Call mark_irreducible_loops.
25807 * cfgloopmanip.c (fix_loop_placement): Add IRRED_INVALIDATED parameter.
25808 Set it to true when we're removing a loop from hierarchy tree in
25809 an irreducible region.
25810 (fix_bb_placements): Adjust caller.
25811 (fix_loop_placements): Likewise.
25812
e52a8b71
GJL		 258132013-01-17 Georg-Johann Lay <avr@gjlay.de>
25814
25815 * config/avr/builtins.def (DEF_BUILTIN): Factor out
25816 "__builtin_avr_" from NAME, turn NAME to an uppercase identifier.
25817 Factor out 'CODE_FOR_' from ICODE, use 'nothing' instead of '-1'.
25818 Remove ID. Adjust comments.
25819 * config/avr/avr-c.c (avr_builtin_name): Remove.
25820 (avr_cpu_cpp_builtins): Use DEF_BUILTIN instead of for-loop.
25821 * config/avr/avr.c (avr_tolower): New static function.
25822 (DEF_BUILTIN): Remove parameter ID. Prefix ICODE by 'CODE_FOR_'.
25823 Stringify NAME, prefix it with "__builtin_avr_" and lowercase it.
25824 (avr_expand_builtin): Assert insn_code != CODE_FOR_nothing for
25825 default expansion.
25826
8386a7ea
JH		 258272013-01-17 Jan Hubicka <jh@suse.cz>
25828
610fb637 25829 PR tree-optimization/55273
8386a7ea
JH		 25830 * loop-iv.c (iv_number_of_iterations): Consider zero iteration case.
25831
47876a2a 258322013-01-17 Uros Bizjak <ubizjak@gmail.com>
8abaebcd
UB		 25833
25834 PR target/55981
25835 * config/i386/sync.md (atomic_store<mode>): Always generate SWImode
25836 store through atomic_store<mode>_1.
25837 (atomic_store<mode>_1): Macroize insn using SWI mode iterator.
25838
8222c37e
MJ		 258392013-01-17 Martin Jambor <mjambor@suse.cz>
25840
25841 PR tree-optimizations/55264
25842 * ipa-inline-transform.c (can_remove_node_now_p_1): Never return true
25843 for virtual methods.
25844 * ipa.c (symtab_remove_unreachable_nodes): Never return true for
25845 virtual methods before inlining is over.
25846 * cgraph.h (cgraph_only_called_directly_or_aliased_p): Return false for
25847 virtual functions.
25848 * cgraphclones.c (cgraph_create_virtual_clone): Mark clones as
25849 non-virtual.
25850
79f01c76
VM		 258512013-01-16 Vladimir Makarov <vmakarov@redhat.com>
25852
25853 PR rtl-optimization/56005
25854 * sched-deps.c (sched_analyze_2): Check deps->readonly for adding
25855 pending reads for prefetch.
25856
d0b6bb1b
IB		 258572013-01-16 Ian Bolton <ian.bolton@arm.com>
25858
aaf1e810 25859 * config/aarch64/aarch64.md
d0b6bb1b
IB		 25860 (*cstoresi_neg_uxtw): New pattern.
25861 (*cmovsi_insn_uxtw): New pattern.
25862 (*<optab>si3_uxtw): New pattern.
25863 (*<LOGICAL:optab>_<SHIFT:optab>si3_uxtw): New pattern.
25864 (*<optab>si3_insn_uxtw): New pattern.
25865 (*bswapsi2_uxtw): New pattern.
25866
cb9cf03b
RB		 258672013-01-16 Richard Biener <rguenther@suse.de>
25868
25869 * tree-inline.c (tree_function_versioning): Remove set but
25870 never used variable.
25871
2cfc56b9
RB		 258722013-01-16 Richard Biener <rguenther@suse.de>
25873
25874 PR tree-optimization/55964
25875 * tree-flow.h (rename_variables_in_loop): Remove.
25876 (rename_variables_in_bb): Likewise.
25877 * tree-loop-distribution.c (update_phis_for_loop_copy): Remove.
25878 (copy_loop_before): Adjust and delete update-ssa status.
25879 * tree-vect-loop-manip.c (rename_variables_in_bb): Make static.
8abaebcd 25880 (rename_variables_in_bb): Likewise. Properly walk over predecessors.
2cfc56b9
RB		 25881 (rename_variables_in_loop): Remove.
25882 (slpeel_update_phis_for_duplicate_loop): Likewise.
25883 (slpeel_tree_duplicate_loop_to_edge_cfg): Handle nested loops,
25884 use available cfg machinery instead of duplicating it.
25885 Update PHI nodes and perform poor-mans SSA update here.
25886 (slpeel_tree_peel_loop_to_edge): Adjust.
25887
c25a0c60
RB		 258882013-01-16 Richard Biener <rguenther@suse.de>
25889
25890 PR tree-optimization/54767
25891 PR tree-optimization/53465
25892 * tree-vrp.c (vrp_meet_1): Revert original fix for PR53465.
25893 (vrp_visit_phi_node): For PHI arguments coming via backedges
25894 drop all symbolical range information.
25895 (execute_vrp): Compute backedges.
25896
04b535af
RB		 258972013-01-16 Richard Biener <rguenther@suse.de>
25898
25899 * doc/install.texi: Update CLooG and ISL requirements to
25900 0.18.0 and 0.11.1.
25901
8b0a1e0b
CB		 259022013-01-16 Christian Bruel <christian.bruel@st.com>
25903
25904 PR target/55301
25905 * config/sh/sh.c (sh_expand_prologue): Postpone new_stack mem symbol.
25906 (broken_move): Handle UNSPECV_SP_SWITCH_B.
25907 * config/sh/sh.md (sp_switch_1): Use set (reg:SI SP_REG).
25908
259092013-01-16 DJ Delorie <dj@redhat.com>
25910
25911 * config/sh/sh.md (UNSPECV_SP_SWITCH_B): New.
25912 (UNSPECV_SP_SWITCH_E): New.
25913 (sp_switch_1): Change to an unspec.
25914 (sp_switch_2): Change to an unspec. Don't use post-inc when we
25915 replace $r15.
25916
259172013-01-16 Uros Bizjak <ubizjak@gmail.com>
88e784e6
UB		 25918
25919 * emit-rtl.c (need_atomic_barrier_p): Mask memory model argument
25920 with MEMMODEL_MASK before comparing with MEMMODEL_* memory types.
25921 * optabs.c (maybe_emit_sync_lock_test_and_set): Ditto.
25922 (expand_mem_thread_fence): Ditto.
25923 (expand_mem_signal_fence): Ditto.
25924 (expand_atomic_load): Ditto.
25925 (expand_atomic_store): Ditto.
25926
0bfbca58 259272013-01-16 Alexandre Oliva <aoliva@redhat.com>
5147bf6a
AO		 25928
25929 PR rtl-optimization/55547
25930 PR rtl-optimization/53827
25931 PR debug/53671
25932 PR debug/49888
25933 * alias.c (memrefs_conflict_p): Set sizes to negative after
25934 AND adjustments.
25935
305e3ac1
JJ		 259362013-01-15 Jakub Jelinek <jakub@redhat.com>
25937
25938 PR target/55940
25939 * function.c (thread_prologue_and_epilogue_insns): Always
25940 add crtl->drap_reg to set_up_by_prologue.set, even if
25941 stack_realign_drap is false.
25942
f78ac4f2
JBG		 259432013-01-15 Jan-Benedict Glaw <jbglaw@lug-owl.de>
25944
25945 * config/vax/vax.md (add<mode>3, sub<mode>3, mul<mode>3, div<mode>3,
25946 and<mode>3, *and<mode>_const_int, ior<mode>3, xor<mode>3, ashrsi3,
25947 *call): Fix indention.
25948
a78a8cc4
TV		 259492013-01-15 Tom de Vries <tom@codesourcery.com>
25950
25951 PR target/55876
25952 * optabs.c (widen_operand): Use gen_lowpart instead of gen_rtx_SUBREG.
25953 Update comment.
25954
0e80383f
VM		 259552013-01-15 Vladimir Makarov <vmakarov@redhat.com>
25956
305e3ac1 25957 PR rtl-optimization/55153
0e80383f
VM		 25958 * sched-deps.c (sched_analyze_2): Add pending reads for prefetch.
25959
207b5956
MJ		 259602013-01-15 Martin Jambor <mjambor@suse.cz>
25961
25962 PR tree-optimization/55920
25963 * tree-sra.c (analyze_access_subtree): Do not mark non-removable
25964 accesses as grp_to_be_debug_replaced.
25965
a7818b54
JJ		 259662013-01-15 Jakub Jelinek <jakub@redhat.com>
25967
25968 PR tree-optimization/55920
25969 * tree-sra.c (sra_modify_assign): If for lacc->grp_to_be_debug_replaced
25970 there is non-useless type conversion needed from debug rhs to lhs,
25971 use build_debug_ref_for_model and/or VIEW_CONVERT_EXPR.
25972
b0fe107e
JM		 259732013-01-15 Joseph Myers <joseph@codesourcery.com>
25974 Mikael Pettersson <mikpe@it.uu.se>
25975
25976 PR target/43961
25977 * config/arm/arm.h (ADDR_VEC_ALIGN): Align SImode jump tables for
25978 Thumb.
25979 (ASM_OUTPUT_CASE_LABEL): Remove.
25980 (ASM_OUTPUT_BEFORE_CASE_LABEL): Define to empty.
25981 * final.c (shorten_branches): Update alignment of labels before
25982 jump tables if CASE_VECTOR_SHORTEN_MODE.
25983
34ab62ee
RB		 259842013-01-15 Richard Biener <rguenther@suse.de>
25985
25986 PR bootstrap/55961
25987 * system.h: Do not include gmp.h for building host tools.
25988
783a3a05
RB		 259892013-01-15 Richard Biener <rguenther@suse.de>
25990
25991 PR middle-end/55882
25992 * emit-rtl.c (set_mem_attributes_minus_bitpos): Correctly
25993 account for bitpos when computing alignment.
25994
3a579e09
VY		 259952013-01-15 Vladimir Yakovlev <vladimir.b.yakovlev@intel.com>
25996
25997 * config/i386/i386-c.c (ix86_target_macros_internal): New case.
25998 (ix86_target_macros_internal): Likewise.
25999
26000 * config/i386/i386.c (m_CORE2I7): Removed.
26001 (m_CORE_HASWELL): New macro.
26002 (m_CORE_ALL): Likewise.
26003 (initial_ix86_tune_features): m_CORE2I7 is replaced by m_CORE_ALL.
26004 (initial_ix86_arch_features): Likewise.
26005 (processor_target_table): Initializations for Core avx2.
26006 (cpu_names): New names "core-avx2".
26007 (ix86_option_override_internal): Changed PROCESSOR_COREI7 by
26008 PROCESSOR_CORE_HASWELL.
26009 (ix86_issue_rate): New case.
26010 (ia32_multipass_dfa_lookahead): Likewise.
26011 (ix86_sched_init_global): Likewise.
26012
26013 * config/i386/i386.h (TARGET_HASWELL): New macro.
26014 (target_cpu_default): New TARGET_CPU_DEFAULT_haswell.
26015 (processor_type): New PROCESSOR_HASWELL.
26016
955f5a07
JJ		 260172013-01-15 Jakub Jelinek <jakub@redhat.com>
26018
ff784829
JJ		 26019 PR tree-optimization/55955
26020 * tree-vect-loop.c (vectorizable_reduction): Give up early on
26021 *SHIFT_EXPR and *ROTATE_EXPR codes.
26022
955f5a07
JJ		 26023 PR tree-optimization/48766
26024 * opts.c (common_handle_option): For -fwrapv disable -ftrapv, for
26025 -ftrapv disable -fwrapv.
26026
aeb8b4e9
GJL		 260272013-01-14 Georg-Johann Lay <avr@gjlay.de>
26028
26029 PR target/55974
26030 * config/avr/avr-c.c (avr_cpu_cpp_builtins): Define __FLASH
26031 etc. to 1 and not to __flash.
26032 Use LL suffix for __INT24_MAX__ with -mint8.
26033 Use ULL suffix for __UINT24_MAX__ with -mint8.
26034
1c494c6a
GJL		 260352013-01-14 Georg-Johann Lay <avr@gjlay.de>
26036
26037 * config/avr/avr-arch.h
26038 (struct base_arch_s): Use typedef avr_arch_t instead.
26039 (struct arch_info_s): Use typedef avr_arch_info_t instead.
26040 (struct mcu_type_s): Use typedef avr_mcu_t instead.
26041 * config/avr/avr.c: Same.
26042 * config/avr/avr-devices.c: Same.
26043 * config/avr/driver-avr.c: Same.
26044 * config/avr/gen-avr-mmcu-texi.c: Same.
26045 * config/avr/avr-mcus.def: Adjust comment.
26046
a50344cb
TB		 260472013-01-14 Tejas Belagod <tejas.belagod@arm.com>
26048
88e784e6
UB		 26049 * config/aarch64/aarch64-simd.md (*aarch64_simd_ld1r<mode>): New.
26050 * config/aarch64/iterators.md (VALLDI): New.
a50344cb 26051
47876a2a 260522013-01-14 Uros Bizjak <ubizjak@gmail.com>
e6f0e052
UB		 26053 Andi Kleen <ak@linux.intel.com>
26054
26055 PR target/55948
26056 * config/i386/sync.md (atomic_store<mode>_1): New pattern.
26057 (atomic_store<mode>): Call atomic_store<mode>_1 for IX86_HLE_RELEASE
26058 memmodel flag.
26059
00892272
GJL		 260602013-01-14 Georg-Johann Lay <avr@gjlay.de>
26061
26062 * config/avr/avr-stdint.h: Remove trailing blanks.
26063 * config/avr/avr-log.h: Same.
26064 * config/avr/avr-arch.h: Same.
26065 * config/avr/avr-devices.c: Same.
26066 * config/avr/avr-dimode.md: Same.
26067 * config/avr/predicates.md: Same.
26068 * config/avr/avr-c.c: Same. And fix typo.
e6f0e052 26069
00892272
GJL		 26070 * config/avr/avr-protos.h: Same. And:
26071 (function_arg_regno_p): Rename to avr_function_arg_regno_p.
26072 (init_cumulative_args): Rename to avr_init_cumulative_args.
26073 (expand_prologue): Rename to avr_expand_prologue.
26074 (expand_epilogue): Rename to avr_expand_epilogue.
26075 (adjust_insn_length): Rename to avr_adjust_insn_length.
26076 (notice_update_cc): Rename to avr_notice_update_cc.
26077 (final_prescan_insn): Rename to avr_final_prescan_insn.
26078 * config/avr/avr.c: Same.
26079 * config/avr/avr.h: Same.
26080 * config/avr/avr.md: Remove trailing blanks.
26081 (prologue): Use avr_expand_prologue.
26082 (epilogue, sibcall_epilogue): Use avr_expand_epilogue.
26083
7bb01996
RB		 260842013-01-14 Richard Biener <rguenther@suse.de>
26085
26086 * tree-cfg.c (verify_expr_location, verify_expr_location_1,
26087 verify_location, collect_subblocks): New functions.
26088 (verify_gimple_in_cfg): Verify that locations only reference
26089 BLOCKs in the functions BLOCK tree.
26090
2724573f
RB		 260912013-01-14 Richard Biener <rguenther@suse.de>
26092
26093 * tree-cfgcleanup.c (remove_forwarder_block): Unshare propagated
26094 PHI argument.
26095 * graphite-sese-to-poly.c (insert_out_of_ssa_copy): Properly
26096 unshare reference.
26097 (insert_out_of_ssa_copy_on_edge): Likewise.
26098 (rewrite_close_phi_out_of_ssa): Likewise.
26099 * tree-ssa.c (insert_debug_temp_for_var_def): Properly unshare
26100 debug expressions.
26101 * tree-ssa-pre.c (insert_into_preds_of_block): Properly unshare
26102 propagated constants.
26103 * tree-cfg.c (tree_node_can_be_shared): Handled component-refs
26104 can not be shared.
26105
9a0bbab6
GJL		 261062013-01-14 Georg-Johann Lay <avr@gjlay.de>
26107
26108 * config/avr/avr-modes.def: Add GPL copyright notice.
26109
45805f17
UB		 261102013-01-13 Uros Bizjak <ubizjak@gmail.com>
26111
26112 * config/i386/sync.md (mem_thread_fence): Mask operands[0] with
26113 MEMMODEL_MASK to determine memory model.
26114 (atomic_store<mode>): Ditto from operands[2].
26115 * config/i386/i386.c (ix86_memmodel_check): Declare "strong" as bool.
26116
9d60be38
JJ		 261172013-01-13 Jakub Jelinek <jakub@redhat.com>
26118
26119 PR fortran/55935
45805f17 26120 * gimple-fold.c (get_symbol_constant_value): Call unshare_expr.
9d60be38
JJ		 26121 (fold_gimple_assign): Don't call unshare_expr here.
26122 (fold_ctor_reference): Call unshare_expr.
26123
e7f49d92
TG		 261242013-01-13 Terry Guo <terry.guo@arm.com>
26125
aaf1e810
EB		 26126 * Makefile.in (s-mlib): New argument MULTILIB_REUSE.
26127 * doc/fragments.texi: Document MULTILIB_REUSE.
26128 * gcc.c (multilib_reuse): New internal spec.
e7f49d92 26129 (set_multilib_dir): Also search multilib from multilib_reuse.
aaf1e810 26130 * genmultilib (tmpmultilib3): Refactor code.
e7f49d92
TG		 26131 (tmpmultilib4): Ditto.
26132 (multilib_reuse): New multilib argument.
26133
fbd03a27
RS		 261342013-01-13 Richard Sandiford <rdsandiford@googlemail.com>
26135
26136 * Makefile.in: Update copyright.
26137
b3681f13
TV		 261382013-01-12 Tom de Vries <tom@codesourcery.com>
26139
26140 PR middle-end/55890
26141 * calls.c (expand_call): Check if arg_nr is valid.
26142
3f287e4b
MM		 261432013-01-11 Michael Meissner <meissner@linux.vnet.ibm.com>
26144
26145 * doc/extend.texi (X86 Built-in Functions): Add whitespace in
26146 __builtin_ia32_paddb256 and __builtin_ia32_pavgb256
5bc35113
MM		 26147 documentation. Add missing '__' in front of
26148 __builtin_ia32_packssdw256.
3f287e4b 26149
1abcd5eb
AK		 261502013-01-11 Andreas Krebbel <Andreas.Krebbel@de.ibm.com>
26151
26152 PR target/55719
26153 * config/s390/s390.c (s390_preferred_reload_class): Do not return
26154 NO_REGS for larl operands.
26155 (s390_reload_larl_operand): Use s390_load_address instead of
26156 emit_move_insn.
26157
980d0812
RB		 261582013-01-11 Richard Biener <rguenther@suse.de>
26159
26160 * tree-cfg.c (verify_node_sharing_1): Split out from ...
26161 (verify_node_sharing): ... here.
26162 (verify_gimple_in_cfg): Use verify_node_sharing_1 for walk_tree.
26163
7cb9fd07
EB		 261642013-01-11 Eric Botcazou <ebotcazou@adacore.com>
26165
26166 * configure.ac (Tree checking): Set TREECHECKING to yes if enabled.
26167 Substitute TREECHECKING.
26168 * configure: Regenerate.
26169 * Makefile.in (TREECHECKING): New.
26170
47876a2a 261712013-01-11 Richard Guenther <rguenther@suse.de>
c7ab2530
RG		 26172
26173 PR tree-optimization/44061
26174 * tree-vrp.c (extract_range_basic): Compute zero as
26175 value-range for __builtin_constant_p of function parameters.
26176
d1e082c2
RS		 261772013-01-10 Richard Sandiford <rdsandiford@googlemail.com>
26178
45805f17 26179 Update copyright years.
d1e082c2 26180
f85021b0
VM		 261812013-01-10 Vladimir Makarov <vmakarov@redhat.com>
26182
9d60be38 26183 PR rtl-optimization/55672
0160442c
VM		 26184 * lra-eliminations.c (mark_not_eliminable): Permit addition with
26185 const to be eliminable.
f85021b0 26186
7a8b1ec4
DE		 261872013-01-10 David Edelsohn <dje.gcc@gmail.com>
26188
26189 * configure.ac (HAVE_AS_TLS): Add check for powerpc-ibm-aix.
26190 * configure: Regenerate.
26191
ad2e5b71
RB		 261922013-01-10 Richard Biener <rguenther@suse.de>
26193
ddf9322d 26194 * builtins.c (expand_builtin_init_trampoline): Use set_mem_attributes.
ad2e5b71 26195
ffc5b2bb
RB		 261962013-01-10 Richard Biener <rguenther@suse.de>
26197
26198 PR bootstrap/55792
26199 * tree-into-ssa.c (rewrite_add_phi_arguments): Do not set
26200 locations for virtual PHI arguments.
26201 (rewrite_update_phi_arguments): Likewise.
26202
e1f674e4
JS		 262032013-01-10 Joel Sherrill <joel.sherrill@OARcorp.com>
26204
26205 * config/v850/rtems.h (ASM_SPEC): Pass -m8byte-align and -mgcc-abi
26206 on to assembler.
26207
a57fc743
JJ		 262082013-01-10 Jakub Jelinek <jakub@redhat.com>
26209
26210 PR tree-optimization/55921
26211 * tree-complex.c (expand_complex_asm): New function.
26212 (expand_complex_operations_1): Call it for GIMPLE_ASM.
26213
0ff4390d
AK		 262142013-01-10 Andreas Krebbel <Andreas.Krebbel@de.ibm.com>
26215
26216 PR target/55718
26217 * config/s390/s390.c (s390_symref_operand_p)
26218 (s390_loadrelative_operand_p): Merge the two functions.
26219 (s390_check_qrst_address, print_operand_address): Add parameters
26220 to s390_loadrelative_operand_p invokation.
26221 (s390_check_symref_alignment): Use s390_loadrelative_operand_p.
26222 (s390_reload_larl_operand, s390_secondary_reload): Use
26223 s390_loadrelative_operand_p instead of s390_symref_operand_p.
26224 (legitimize_pic_address): Handle @GOTENT and @PLT + addend.
26225
6f557e0e
MS		 262262013-01-09 Mike Stump <mikestump@comcast.net>
26227
26228 * dse.c (record_store): Remove unnecessary assert.
26229
7770c9e9
JH		 262302013-01-09 Jan Hubicka <jh@suse.cz>
26231
26232 PR tree-optimization/55569
26233 * cfgloopmanip.c (scale_loop_profile): Make ITERATION_BOUND gcov_type.
26234 * cfgloop.h (scale_loop_profile): Likewise.
26235
a19b1432
JH		 262362013-01-09 Jan Hubicka <jh@suse.cz>
26237
26238 PR lto/45375
ddf9322d
UB		 26239 * ipa-inline.c (ipa_inline): Remove extern inlines and virtual
26240 functions.
a19b1432
JH		 26241 * cgraphclones.c (cgraph_clone_node): Cpoy also LTO file data.
26242
9a002da8
RS		 262432013-01-09 Richard Sandiford <rdsandiford@googlemail.com>
26244
26245 PR middle-end/55114
26246 * expr.h (maybe_emit_group_store): Declare.
26247 * expr.c (maybe_emit_group_store): New function.
26248 * builtins.c (expand_builtin_int_roundingfn): Call it.
26249 (expand_builtin_int_roundingfn_2): Likewise.
26250
511dcace
VM		 262512013-01-09 Vladimir Makarov <vmakarov@redhat.com>
26252
e1f2b729 26253 PR rtl-optimization/55829
511dcace
VM		 26254 * lra-constraints.c (match_reload): Add code for absent output.
26255 (curr_insn_transform): Add code for reloads of matched inputs
26256 without output.
26257
7b0fe4f4
UB		 262582013-01-09 Uros Bizjak <ubizjak@gmail.com>
26259
26260 * config/i386/sse.md (*vec_interleave_highv2df): Change mode
26261 attribute of movddup insn to DF.
26262 (*vec_interleave_lowv2df): Ditto.
26263 (vec_dupv2df): Ditto.
26264
870ca331
JH		 262652013-01-09 Jan Hubicka <jh@suse.cz>
26266
26267 PR tree-optimiation/55875
26268 * tree-ssa-loop-niter.c (number_of_iterations_cond): Add
26269 EVERY_ITERATION parameter.
7b0fe4f4 26270 (number_of_iterations_exit): Check if exit is executed every iteration.
870ca331
JH		 26271 (idx_infer_loop_bounds): Similarly here.
26272 (n_of_executions_at_most): Simplify
26273 to only test for cases where statement is dominated by the
7b0fe4f4 26274 particular bound; handle correctly the "postdominance" test.
870ca331
JH		 26275 (scev_probably_wraps_p): Use max loop iterations info
26276 as a global bound first.
26277
6f575fe4 262782013-01-09 Nguyen Duy Dat <dat.nguyen.yn@rvc.renesas.com>
3cd232b5
NDD		 26279 Nick Clifton <nickc@redhat.com>
26280
26281 * config/v850/v850.md (cbranchsf4): New pattern.
26282 (cstoresf4): New pattern.
26283 (cbranchdf4): New pattern.
26284 (cstoredf4): New pattern.
26285 (movsicc): Disallow floating point comparisons.
26286 (cmpsf_le_insn): Fix order of operators.
26287 (cmpsf_lt_insn): Likewise.
26288 (cmpsf_eq_insn): Likewise.
26289 (cmpdf_le_insn): Likewise.
26290 (cmpdf_lt_insn): Likewise.
26291 (cmpdf_eq_insn): Likewise.
26292 (cmpsf_ge_insn): Use LE comparison.
26293 (cmpdf_ge_insn): Likewise.
26294 (cmpsf_gt_insn): Use LT comparison.
26295 (cmpdf_gt_insn): Likewise.
26296 (cmpsf_ne_insn): Delete pattern.
26297 (cmpdf_ne_insn): Delete pattern.
26298 * config/v850/v850.c (v850_gen_float_compare): Use
26299 gen_cmpdf_eq_insn for NE comparison.
26300 (v850_float_z_comparison_operator)
26301 (v850_float_nz_comparison_operator): Move from here ...
26302 * config/v850/predicates.md: ... to here. Move GT and GE
26303 comparisons into v850_float_z_comparison_operator.
26304 * config/v850/v850-protos.h (v850_float_z_comparison_operator):
26305 Delete prototype.
26306 (v850_float_nz_comparison_operator): Likewise.
26307
f0d54148
JDA		 263082013-01-09 John David Anglin <dave.anglin@nrc-cnrc.gc.ca>
26309
26310 * config/pa/pa.c (pa_emit_move_sequence): Replace calls to gen_insv
26311 with calls to gen_insvsi/gen_insvdi.
26312
8f01beca
VK		 263132013-01-09 Venkataramanan Kumar <venkataramanan.kumar@amd.com>
26314
26315 * config/i386/i386.c (initial_ix86_tune_features): Set up
26316 X86_TUNE_AVX128_OPTIMAL for m_BTVER2.
26317
2aa579ad
JJ		 263182013-01-09 Steven Bosscher <steven@gcc.gnu.org>
26319 Jakub Jelinek <jakub@redhat.com>
26320
26321 PR tree-optimization/48189
26322 * predict.c (predict_loops): If max is 0, don't call compare_tree_int.
26323 If nitercst is 0, don't predict the exit edge.
26324
6edc3e32 263252013-01-08 Naveen H.S <Naveen.Hurugalawadi@caviumnetworks.com>
01a3a324
N		 26326
26327 * config/aarch64/aarch64.c (aarch64_print_operand): Replace %r
26328 in asm_fprintf with reg_names.
26329 (aarch64_print_operand_address): Likewise.
26330 (aarch64_return_addr): Likewise.
26331 * config/aarch64/aarch64.h (ASM_FPRINTF_EXTENSIONS): Remove.
26332
f6f94d94
JDA		 263332013-01-08 John David Anglin <dave.anglin@nrc-cnrc.gc.ca>
26334
26335 * config/pa/pa.h (VAL_U6_BITS_P): Define.
26336 (INT_U6_BITS): Likewise.
26337 * config/pa/predicates.md (uint6_operand): New predicate.
26338 (shift5_operand, shift6_operand): Likewise.
26339 * config/pa/pa.md (lshrsi3, rotrsi3): Use shift5_operand instead of
26340 arith32_operand.
26341 (lshrdi3): Use shift6_operand.
26342 (shrpsi4, shrpdi4): New insn patterns.
26343 (extzv): Delete expander.
26344 (extzvsi, extzvdi): New expanders. Use uint5_operand and uint6_operand
26345 predicates in unamed zero extract patterns. Tighten common constraint.
26346 (extv): Delete expander.
26347 (extvsi, extvdi): New expanders. Use uint5_operand and uint6_operand
26348 predicates in unamed sign extract patterns. Tighten common constraint.
26349 (insv): Delete expander.
26350 (insvsi, insvdi): New expanders. Use uint5_operand and uint6_operand
26351 predicates in unamed insert patterns. Tighten common constraint.
26352 Change uint32_operand predicate to uint6_operand predicate in unamed
26353 DImode pattern to insert constant values of type 1...1xxxx.
26354
36b72910
JH		 263552013-01-04 Jan Hubicka <jh@suse.cz>
26356
26357 PR tree-optimization/55823
7b0fe4f4
UB		 26358 * ipa-prop.c (update_indirect_edges_after_inlining): Fix ordering
26359 issue.
36b72910 26360
47876a2a 263612013-01-08 Jakub Jelinek <jakub@redhat.com>
c6d851b9
JJ		 26362 Uros Bizjak <ubizjak@gmail.com>
26363
26364 PR rtl-optimization/55845
26365 * df-problems.c (can_move_insns_across): Stop scanning at
26366 volatile_insn_p source instruction or give up if
26367 across_from .. across_to range contains any volatile_insn_p
26368 instructions.
26369
4369c11e
TB		 263702013-01-08 Tejas Belagod <tejas.belagod@arm.com>
26371
7b0fe4f4
UB		 26372 * config/aarch64/aarch64-simd.md (vec_init<mode>): New.
26373 * config/aarch64/aarch64-protos.h (aarch64_expand_vector_init):
26374 Declare.
4369c11e 26375 * config/aarch64/aarch64.c (aarch64_simd_dup_constant,
7b0fe4f4 26376 aarch64_simd_make_constant, aarch64_expand_vector_init): New.
4369c11e 26377
aeb7e7c1
JJ		 263782013-01-08 Jakub Jelinek <jakub@redhat.com>
26379
26380 PR fortran/55341
26381 * asan.c (asan_clear_shadow): New function.
26382 (asan_emit_stack_protection): Use it.
26383
a02ad1aa
TB		 263842013-01-08 Tejas Belagod <tejas.belagod@arm.com>
26385
26386 * config/aarch64/aarch64-simd.md (aarch64_simd_vec_<su>mult_lo_<mode>,
26387 aarch64_simd_vec_<su>mult_hi_<mode>): Separate instruction and operand
26388 with tab instead of space.
26389
f90d87f5
NC		 263902013-01-08 Nick Clifton <nickc@redhat.com>
26391
26392 * config/rl78/rl78.c (rl78_expand_prologue): Always select
26393 register bank 0 at the start of an interrupt handler.
83ffd964
NC		 26394 * config/rl78/rl78.md (mulsi3_g13): Correct values for MDBL and
26395 MDBH registers.
f90d87f5 26396
385eb93d
JG		 263972013-01-08 James Greenhalgh <james.greenhalgh@arm.com>
26398
26399 * config/aarch64/aarch64-simd.md
26400 (aarch64_simd_bsl<mode>_internal): Add floating-point modes.
26401 (aarch64_simd_bsl): Likewise.
26402 (aarch64_vcond_internal<mode>): Likewise.
26403 (vcond<mode><mode>): Likewise.
26404 (aarch64_cm<cmp><mode>): Fix constraints, add new modes.
26405 * config/aarch64/iterators.md (V_cmp_result): Add V2DF.
26406
4dcd1054
JG		 264072013-01-08 James Greenhalgh <james.greenhalgh@arm.com>
26408
26409 * config/aarch64/aarch64-builtins.c
26410 (aarch64_builtin_vectorized_function): Handle sqrt, sqrtf.
26411
4267a4a6
MJ		 264122013-01-08 Martin Jambor <mjambor@suse.cz>
26413
26414 PR debug/55579
26415 * tree-sra.c (analyze_access_subtree): Return true also after
26416 potentially creating a debug-only replacement.
26417
5f4e6de3
JJ		 264182013-01-08 Jakub Jelinek <jakub@redhat.com>
26419
3138f224
JJ		 26420 PR middle-end/55890
26421 * tree-ssa-ccp.c (evaluate_stmt): Use gimple_call_builtin_p.
26422
5f4e6de3
JJ		 26423 PR tree-optimization/54120
26424 * tree-vrp.c (range_fits_type_p): Don't allow
26425 src_precision < precision from signed vr to unsigned_p
26426 if vr->min or vr->max is negative.
26427 (simplify_float_conversion_using_ranges): Test can_float_p
26428 against CODE_FOR_nothing.
26429
f3ef18ff
JJ		 264302013-01-08 Jakub Jelinek <jakub@redhat.com>
26431 Richard Biener <rguenther@suse.de>
26432
26433 PR middle-end/55851
26434 * fold-const.c (int_binop_types_match_p): Allow all INTEGRAL_TYPE_P
26435 types instead of just INTEGER_TYPE types.
26436
25c210f9
MK		 264372013-01-07 Mark Kettenis <kettenis@openbsd.org>
26438
26439 * config/i386/openbsdelf.h (LIBGCC2_HAS_TF_MODE, LIBGCC2_TF_CEXT,
26440 TF_SIZE): Define.
f90d87f5 26441
81826a7b
SE		 264422013-01-07 Steve Ellcey <sellcey@mips.com>
26443
26444 PR target/42661
26445 * config/mips/mips.opt: Change mad to mmad to match documentation.
26446
35678514
GJL		 264472013-01-07 Georg-Johann Lay <avr@gjlay.de>
26448
26449 PR target/55897
26450 * doc/extend.texi (AVR Named Address Spaces): __memx goes into
26451 .progmemx.data now.
26452
e5669488
GJL		 264532013-01-07 Georg-Johann Lay <avr@gjlay.de>
26454
26455 PR target/55897
26456 * config/avr/avr.h (ADDR_SPACE_COUNT): New enum.
26457 (avr_addrspace_t): Add .section_name field.
26458 * config/avr/avr.c (progmem_section): Use ADDR_SPACE_COUNT as
26459 array size.
26460 (avr_addrspace): Same. Initialize .section_name. Remove last
26461 NULL entry. Put __memx into .progmemx.data.
26462 (progmem_section_prefix): Remove.
26463 (avr_asm_init_sections): No need to initialize progmem_section.
26464 (avr_asm_named_section): Use avr_addrspace[].section_name to get
26465 section name prefix.
26466 (avr_asm_select_section): Ditto. And use get_unnamed_section to
26467 retrieve the progmem section.
26468 * avr-c.c (avr_cpu_cpp_builtins): Use ADDR_SPACE_COUNT as loop
26469 boundary to run over avr_addrspace[].
26470 (avr_register_target_pragmas): Ditto.
26471
656e6f37
JJ		 264722013-01-06 Jakub Jelinek <jakub@redhat.com>
26473
26474 * varasm.c (output_constant_def_contents): For asan_protect_global
26475 protected strings, adjust DECL_ALIGN if needed, before testing for
26476 anchored symbols.
26477 (place_block_symbol): Adjust size for asan protected STRING_CSTs if
26478 TREE_CONSTANT_POOL_ADDRESS_P. Increase alignment for asan protected
26479 normal decls.
26480 (output_object_block): For asan protected decls, emit asan padding
26481 after their contents.
26482 * asan.c (asan_protect_global): Don't check TREE_ASM_WRITTEN here.
26483 (asan_finish_file): Test it here instead.
26484
6edc3e32
UB		 264852013-01-07 Nick Clifton <nickc@redhat.com>
26486 Matthias Klose <doko@debian.org>
26487 Doug Kwan <dougkwan@google.com>
26488 H.J. Lu <hongjiu.lu@intel.com>
b352afba
NC		 26489
26490 PR driver/55470
26491 * collect2.c (main): Support -fuse-ld=bfd and -fuse-ld=gold.
26492
26493 * common.opt: Add fuse-ld=bfd and fuse-ld=gold.
26494
26495 * gcc.c (LINK_COMMAND_SPEC): Pass -fuse-ld=* to collect2.
26496
7b0fe4f4 26497 * opts.c (comman_handle_option): Ignore -fuse-ld=bfd and -fuse-ld=gold.
b352afba
NC		 26498
26499 * doc/invoke.texi: Document -fuse-ld=bfd and -fuse-ld=gold.
26500
2a095093
GJL		 265012013-01-07 Georg-Johann Lay <avr@gjlay.de>
26502
f3b503f4 26503 PR target/54461
2a095093
GJL		 26504 * doc/install.texi (Cross-Compiler-Specific Options): Document
26505 --with-avrlibc.
26506
383f9b34
TB		 265072013-01-07 Tejas Belagod <tejas.belagod@arm.com>
26508
26509 * config/aarch64/arm_neon.h (vmovn_high_is16, vmovn_high_s32,
26510 vmovn_high_s64, vmovn_high_u16, vmovn_high_u32, vmovn_high_u64,
26511 vqmovn_high_s16, vqmovn_high_s32, vqmovn_high_s64, vqmovn_high_u16,
26512 vqmovn_high_u32, vqmovn_high_u64, vqmovun_high_s16, vqmovun_high_s32,
26513 vqmovun_high_s64): Fix source operand number and update copyright.
26514
3626621a
RB		 265152013-01-07 Richard Biener <rguenther@suse.de>
26516
26517 PR middle-end/55890
26518 * gimple.h (gimple_call_builtin_p): New overload.
26519 * gimple.c (validate_call): New function.
26520 (gimple_call_builtin_p): Likewise.
26521 * tree-ssa-structalias.c (find_func_aliases_for_builtin_call):
26522 Use gimple_call_builtin_p.
26523 (find_func_clobbers): Likewise.
26524 * tree-ssa-strlen.c (adjust_last_stmt): Likewise.
26525 (strlen_optimize_stmt): Likewise.
26526
8b2ea410
JG		 265272013-01-07 James Greenhalgh <james.greenhalgh@arm.com>
26528
26529 * config/aarch64/arm_neon.h (vld1_dup_*): Make argument const.
26530 (vld1q_dup_*): Likewise.
26531 (vld1_*): Likewise.
26532 (vld1q_*): Likewise.
26533 (vld1_lane_*): Likewise.
26534 (vld1q_lane_*): Likewise.
26535
9713d329
RB		 265362013-01-07 Richard Biener <rguenther@suse.de>
26537
26538 * lto-streamer.h (LTO_minor_version): Bump to 2.
26539
3520f7cc
JG		 265402013-01-07 James Greenhalgh <james.greenhalgh@arm.com>
26541
26542 * config/aarch64/aarch64-protos.h
26543 (aarch64_const_double_zero_rtx_p): Rename to...
26544 (aarch64_float_const_zero_rtx_p): ...this.
26545 (aarch64_float_const_representable_p): New.
26546 (aarch64_output_simd_mov_immediate): Likewise.
26547 * config/aarch64/aarch64-simd.md (*aarch64_simd_mov<mode>): Refactor
26548 move immediate case.
26549 * config/aarch64/aarch64.c
26550 (aarch64_const_double_zero_rtx_p): Rename to...
26551 (aarch64_float_const_zero_rtx_p): ...this.
26552 (aarch64_print_operand): Allow printing of new constants.
26553 (aarch64_valid_floating_const): New.
26554 (aarch64_legitimate_constant_p): Check for valid floating-point
26555 constants.
26556 (aarch64_simd_valid_immediate): Likewise.
26557 (aarch64_vect_float_const_representable_p): New.
26558 (aarch64_float_const_representable_p): Likewise.
26559 (aarch64_simd_imm_zero_p): Also allow for floating-point 0.0.
26560 (aarch64_output_simd_mov_immediate): New.
26561 * config/aarch64/aarch64.md (*movsf_aarch64): Add new alternative.
26562 (*movdf_aarch64): Likewise.
26563 * config/aarch64/constraints.md (Ufc): New.
26564 (Y): call aarch64_float_const_zero_rtx.
26565 * config/aarch64/predicates.md (aarch64_fp_compare_operand): New.
26566
e167c04d
RB		 265672013-01-07 Richard Biener <rguenther@suse.de>
26568
26569 PR tree-optimization/55888
26570 PR tree-optimization/55862
26571 * tree-ssa-pre.c (phi_translate_1): Revert previous change.
26572 (valid_in_sets): Check if a NAME has a leader in AVAIL_OUT,
26573 not if it is contained therein.
26574
0139aaab
GJL		 265752013-01-07 Georg-Johann Lay <avr@gjlay.de>
26576
26577 * config/avr/t-avr: Typo.
26578
4a176b23
GJL		 265792013-01-07 Georg-Johann Lay <avr@gjlay.de>
26580
26581 PR55243
26582 * config/avr/t-avr: Don't automatically rebuild
26583 $(srcdir)/config/avr/t-multilib
26584 $(srcdir)/config/avr/avr-tables.opt
26585 $(srcdir)/doc/avr-mmcu.texi
26586 (avr-mcus): New phony target to build them on request.
26587 (s-avr-mlib, s-avr-mmcu-texi): Remove.
26588 * avr/avr-mcus.def: Adjust comments.
26589
c7afdc98
UB		 265902013-01-07 Uros Bizjak <ubizjak@gmail.com>
26591
26592 * config/i386/i386.c (DEFAULT_PCC_STRUCT_RETURN): Remove.
26593
1ab05c31
RS		 265942013-01-06 Richard Sandiford <rdsandiford@googlemail.com>
26595
26596 * file-find.c, file-find.h, realmpfr.c: Add FSF as copyright holder.
26597
488e3acc
RS		 265982013-01-06 Richard Sandiford <rdsandiford@googlemail.com>
26599
26600 * config/tilepro/gen-mul-tables.cc: Put copyright on one line.
26601
a89599d2
DE		 266022013-01-05 David Edelsohn <dje.gcc@gmail.com>
26603
26604 * config/rs6000/aix53.h (LIB_SPEC): Add -lpthreads when compiling
26605 to generate profiling.
26606 * config/rs6000/aix64.h (LIB_SPEC): Same.
26607
70f09188
AP		 266082013-01-04 Andrew Pinski <apinski@cavium.com>
26609
26610 * config/aarch64/aarch64.c (aarch64_fixed_condition_code_regs):
26611 New function.
26612 (TARGET_FIXED_CONDITION_CODE_REGS): Define.
26613
918d445f
UB		 266142013-01-04 Uros Bizjak <ubizjak@gmail.com>
26615
26616 * config/i386/i386.c (ix86_legitimize_address): Call convert_to_mode
26617 unconditionally.
26618 (ix86_expand_move): Ditto.
26619 (ix86_zero_extend_to_Pmode): Ditto.
26620 (ix86_expand_call): Ditto.
26621 (ix86_expand_special_args_builtin): Ditto.
26622 (ix86_expand_builtin): Ditto.
26623
361618ec
RB		 266242013-01-04 Richard Biener <rguenther@suse.de>
26625
26626 PR tree-optimization/55862
26627 * tree-ssa-pre.c (phi_translate_1): Valueize SSA names after
26628 translating them through PHI nodes.
26629
87eab554
MJ		 266302013-01-04 Martin Jambor <mjambor@suse.cz>
26631
26632 PR tree-optimization/55755
26633 * tree-sra.c (sra_modify_assign): Do not check that an access has no
26634 children when trying to avoid producing a VIEW_CONVERT_EXPR.
26635
33879b9f
MP		 266362013-01-04 Marek Polacek <polacek@redhat.com>
26637
26638 PR middle-end/55859
26639 * opts.c (default_options_optimization): Clarify error message.
26640
3068819a
RB		 266412013-01-04 Richard Biener <rguenther@suse.de>
26642
26643 PR middle-end/55863
26644 * fold-const.c (split_tree): Undo -X - 1 to ~X folding for
26645 reassociation.
26646
0e4ae794
JDA		 266472013-01-03 John David Anglin <dave.anglin@nrc-cnrc.gc.ca>
26648
26649 PR target/53789
26650 * config/pa/pa.md (movsi): Revert previous change.
e58d2e48
JDA		 26651 * config/pa/pa.c (pa_legitimate_constant_p): Reject all TLS symbol
26652 references.
0e4ae794 26653
65c88cef
RH		 266542013-01-03 Richard Henderson <rth@redhat.com>
26655
26656 * config/i386/i386.c (ix86_expand_move): Always assign to op1
26657 after eliminating TLS symbols.
26658
8ac16127
MG		 266592013-01-03 Marc Glisse <marc.glisse@inria.fr>
26660
26661 PR bootstrap/50167
26662 * graphite-interchange.c (pdr_stride_in_loop): Use gmp_fprintf.
26663 * graphite-poly.c (debug_gmp_value): Likewise.
26664
bb664f09
UB		 266652013-01-03 Uros Bizjak <ubizjak@gmail.com>
26666
26667 PR target/55712
26668 * config/i386/i386-c.c (ix86_target_macros_internal): Depending on
26669 selected code model, define __code_mode_small__, __code_model_medium__,
26670 __code_model_large__, __code_model_32__ or __code_model_kernel__.
26671 * config/i386/cpuid.h (__cpuid, __cpuid_count) [__i386__]: Prefix
26672 xchg temporary register with %k. Declare temporary register as
26673 early clobbered.
26674 [__x86_64__]: For medium and large code models, preserve %rbx register.
26675
bcf1ef00
RB		 266762013-01-03 Richard Biener <rguenther@suse.de>
26677
0506634a 26678 * tree-data-ref.c (dump_conflict_function): Use less vertical spacing.
bcf1ef00
RB		 26679 (dump_subscript): Adjust.
26680 (finalize_ddr_dependent): Do not dump redundant info.
26681 (analyze_siv_subscript): Adjust.
26682 (subscript_dependence_tester): Likewise.
26683 (compute_affine_dependence): Likewise.
26684
59fd17e3
RB		 266852013-01-03 Richard Biener <rguenther@suse.de>
26686
26687 Revert
26688 2013-01-03 Richard Biener <rguenther@suse.de>
26689
26690 PR tree-optimization/55857
26691 * tree-vect-stmts.c (vectorizable_load): Do not setup
26692 re-alignment for invariant loads.
26693
26694 2013-01-02 Richard Biener <rguenther@suse.de>
26695
26696 * tree-vect-stmts.c (vectorizable_load): When vectorizing an
0506634a 26697 invariant load do not generate a vector load from the scalar location.
59fd17e3 26698
595c2679
RB		 266992013-01-03 Richard Biener <rguenther@suse.de>
26700
26701 * tree-vect-loop.c (vect_analyze_loop_form): Clarify reason
26702 for not vectorizing.
26703 * tree-vect-data-refs.c (vect_create_addr_base_for_vector_ref): Do
26704 not build INDIRECT_REFs, call get_name once only.
26705 (vect_create_data_ref_ptr): Likewise. Dump base object kind
26706 based on DR_BASE_OBJECT, not DR_BASE_ADDRESS.
26707
90bb7d7a
RB		 267082013-01-03 Richard Biener <rguenther@suse.de>
26709
26710 PR tree-optimization/55857
26711 * tree-vect-stmts.c (vectorizable_load): Do not setup
26712 re-alignment for invariant loads.
26713
f09b77ca
RB		 267142013-01-03 Richard Biener <rguenther@suse.de>
26715
26716 PR lto/55848
26717 * lto-symtab.c (lto_symtab_merge_decls_1): As last resort, always
26718 prefer a built-in decl.
26719
3461a16e
JJ		 267202013-01-03 Jakub Jelinek <jakub@redhat.com>
26721
df8e2b4f
JJ		 26722 * gcc.c (process_command): Update copyright notice dates.
26723 * gcov.c (print_version): Likewise.
26724 * gcov-dump.c (print_version): Likewise.
26725
3461a16e
JJ		 26726 PR rtl-optimization/55838
26727 * loop-iv.c (iv_number_of_iterations): Call lowpart_subreg on
26728 iv0.step, iv1.step and step.
26729
8b5546d6
JJ		 267302013-01-03 Jakub Jelinek <jakub@redhat.com>
26731 Marc Glisse <marc.glisse@inria.fr>
26732
26733 PR tree-optimization/55832
26734 * fold-const.c (fold_binary_loc): For ABS_EXPR<x> >= 0 and
26735 ABS_EXPR<x> < 0 folding use constant_boolean_node instead of
26736 integer_{one,zero}_node.
26737
8ab1d2e9
JJ		 267382013-01-03 Jakub Jelinek <jakub@redhat.com>
26739
26740 PR debug/54402
26741 * params.def (PARAM_MAX_VARTRACK_REVERSE_OP_SIZE): New param.
26742 * var-tracking.c (reverse_op): Don't add reverse ops to
26743 VALUEs that have already
0506634a 26744 PARAM_VALUE (PARAM_MAX_VARTRACK_REVERSE_OP_SIZE) or longer locs list.
8ab1d2e9 26745
5b9ad1d4
GP		 267462013-01-02 Gerald Pfeifer <gerald@pfeifer.com>
26747
26748 * doc/contrib.texi: Note years as release manager for Mark Mitchell.
26749
e25a6711
TJ		 267502013-01-02 Teresa Johnson <tejohnson@google.com>
26751
26752 * dumpfile.c (dump_loc): Print filename with location.
26753 * tree-ssa-loop-ivcanon.c (try_unroll_loop_completely): Use
26754 new location_t parameter to emit complete unroll message with
26755 new dump framework.
26756 (canonicalize_loop_induction_variables): Compute loops location
26757 and pass to try_unroll_loop_completely.
26758 * loop-unroll.c (report_unroll_peel): New function.
26759 (peel_loops_completely): Use new dump format with location
26760 for main dumpfile message, and invoke report_unroll_peel on success.
26761 (decide_unrolling_and_peeling): Ditto.
26762 (decide_peel_once_rolling): Remove old dumpfile message subsumed
26763 by report_unroll_peel.
26764 (decide_peel_completely): Ditto.
26765 (decide_unroll_constant_iterations): Ditto.
26766 (decide_unroll_runtime_iterations): Ditto.
26767 (decide_peel_simple): Ditto.
26768 (decide_unroll_stupid): Ditto.
26769 * cfgloop.c (get_loop_location): New function.
26770 * cfgloop.h (get_loop_location): Declare.
26771
77878621
ST		 267722013-01-02 Sriraman Tallam <tmsriram@google.com>
26773
26774 * config/i386/i386.c (fold_builtin_cpu): Remove unnecessary checks for
26775 NULL.
26776
9e65d03e
JDA		 267772013-01-02 John David Anglin <dave.anglin@nrc-cnrc.gc.ca>
26778
26779 PR middle-end/55198
26780 * expr.c (expand_expr_real_1): Don't use bitfield extraction for non
26781 BLKmode objects when EXPAND_MEMORY is specified.
26782
6a7da30f
ST		 267832013-01-02 Sriraman Tallam <tmsriram@google.com>
26784
26785 * config/i386/i386.c (ix86_get_function_versions_dispatcher): Fix bug
26786 in loop predicate.
26787 (fold_builtin_cpu): Do not share cpu model decls across statements.
26788
e78167f2
JM		 267892013-01-02 Jason Merrill <jason@redhat.com>
26790
26791 PR c++/55804
26792 * tree.c (build_array_type_1): Revert earlier change.
26793
8c075fb4
YZ		 267942013-01-02 Yufeng Zhang <yufeng.zhang@arm.com>
26795
26796 * config/aarch64/aarch64-cores.def: Add entries for "cortex-a53" and
26797 "cortex-a57".
26798 * config/aarch64/aarch64-tune.md: Re-generate.
26799
0682ed3e
RB		 268002013-01-02 Richard Biener <rguenther@suse.de>
26801
26802 * tree-vect-stmts.c (vectorizable_load): When vectorizing an
0506634a 26803 invariant load do not generate a vector load from the scalar location.
0682ed3e 26804
b8f6a302
RB		 268052013-01-02 Richard Biener <rguenther@suse.de>
26806
26807 PR bootstrap/55784
26808 * configure.ac: Add $GMPINC to CFLAGS/CXXFLAGS.
26809 * configure: Regenerate.
26810
04b80dbb
RS		 268112013-01-02 Richard Sandiford <rdsandiford@googlemail.com>
26812
26813 * builtins.c (expand_builtin_mathfn, expand_builtin_mathfn_2)
26814 (expand_builtin_mathfn_ternary, expand_builtin_mathfn_3)
26815 (expand_builtin_int_roundingfn_2): Keep the original target around
26816 for the fallback case.
26817
635b0b0c
RS		 268182013-01-02 Richard Sandiford <rdsandiford@googlemail.com>
26819
26820 * tree-vrp.c (range_fits_type_p): Require the MSB of the double_int
26821 to be clear for sign changes.
26822
42d57399
JH		 268232013-01-01 Jan Hubicka <jh@suse.cz>
26824
26825 * ipa-inline-analysis.c: Fix formatting.
26826
5bb6669d
JJ		 268272013-01-01 Jakub Jelinek <jakub@redhat.com>
26828
26829 PR tree-optimization/55831
26830 * tree-vect-loop.c (get_initial_def_for_induction): Use
26831 gsi_after_labels instead of gsi_start_bb.
ad41bd84 26832\f
86a2db33 26833Copyright (C) 2013 Free Software Foundation, Inc.
ad41bd84
JM		 26834
26835Copying and distribution of this file, with or without modification,
26836are permitted in any medium without royalty provided the copyright
26837notice and this notice are preserved.
Mirror of https://gcc.gnu.org/git/gcc.git