powerpc: Add some XFAILs to 20050603-3.c (PR68803)
In r230167 I made this testcase be tested on 64-bit as well, since it
now works. That was a tad optimistic for powerpc64le though. For now,
XFAIL it there.
PR target/68803
* gcc.target/powerpc/20050603-3.c: Add xfails for powerpc64le.
David Malcolm [Thu, 14 Jan 2016 19:10:17 +0000 (19:10 +0000)]
PR preprocessor/69177 and PR c++/68819: libcpp fallbacks and -Wmisleading-indentation
gcc/c-family/ChangeLog:
PR c++/68819
* c-indentation.c (get_visual_column): Add location_t param.
Handle the column number being zero by effectively disabling the
warning, with an "inform".
(should_warn_for_misleading_indentation): Add location_t argument
for all uses of get_visual_column.
gcc/testsuite/ChangeLog:
PR c++/68819
PR preprocessor/69177
* gcc.dg/plugin/location-overflow-test-1.c: New test case.
* gcc.dg/plugin/location-overflow-test-2.c: New test case.
* gcc.dg/plugin/location_overflow_plugin.c: New test plugin.
* gcc.dg/plugin/plugin.exp (plugin_test_list): Add the above.
libcpp/ChangeLog:
PR preprocessor/69177
* line-map.c (LINE_MAP_MAX_LOCATION_WITH_PACKED_RANGES): New
constant.
(LINE_MAP_MAX_LOCATION_WITH_COLS): Add note about unit tests
to comment.
(can_be_stored_compactly_p): Reduce threshold from
LINE_MAP_MAX_LOCATION_WITH_COLS to
LINE_MAP_MAX_LOCATION_WITH_PACKED_RANGES.
(get_combined_adhoc_loc): Likewise.
(get_range_from_loc): Likewise.
(linemap_line_start): Ensure that a new ordinary map is created
when transitioning from range-packing being enabled to disabled,
at the LINE_MAP_MAX_LOCATION_WITH_PACKED_RANGES threshold. Set
range_bits to 0 for new ordinary maps when beyond this limit.
Prevent the "increase the column bits of a freshly created map"
optimization if the range bits has reduced.
Jakub Jelinek [Thu, 14 Jan 2016 15:25:22 +0000 (16:25 +0100)]
re PR middle-end/68146 (ice in gimple_stmt_nonnegative_warnv_p with -O2)
PR middle-end/68146
PR tree-optimization/69155
* tree-complex.c: Include cfganal.h.
(phis_to_revisit): New variable.
(extract_component): Add phiarg_p argument. Assert that returned
SSA_NAME has non-NULL SSA_NAME_DEF_STMT unless phiarg_p is true.
(update_phi_components): Partly rewrite to use loop over real/imag
components instead of code duplication. If extract_component returns
SSA_NAME with NULL SSA_NAME_DEF_STMT, store SSA_NAME_VAR or
create_tmp_reg into the PHI node instead, and mention the phi triplet
in phis_to_revisit.
(tree_lower_complex): Walk bbs in rpo order. Adjust phis recorded
in phis_to_revisit at the end.
* gfortran.dg/pr68146.f: New test.
* gfortran.dg/pr69155.f90: New test.
Richard Biener [Thu, 14 Jan 2016 15:13:19 +0000 (15:13 +0000)]
re PR tree-optimization/68060 (ICE on valid code at -O3 on x86_64-linux-gnu in vect_get_vec_def_for_operand, at tree-vect-stmts.c:1413)
2016-01-14 Richard Biener <rguenther@suse.de>
PR tree-optimization/68060
* tree-vect-loop.c (vect_is_simple_reduction): Check the
outer loop reduction is only used in the inner loop before
detecting a double reduction.
* gcc.dg/torture/pr68060-1.c: New testcase.
* gcc.dg/torture/pr68060-2.c: Likewise.
Richard Biener [Thu, 14 Jan 2016 14:51:35 +0000 (14:51 +0000)]
re PR tree-optimization/66856 (ICE in compute_live_loop_exits, at tree-ssa-loop-manip.c:234)
2016-01-14 Richard Biener <rguenther@suse.de>
PR tree-optimization/66856
* tree-vect-slp.c (vect_build_slp_tree): Refactor to build
SLP node only if it built successfully.
(vect_analyze_slp_instance): Adjust.
Nick Clifton [Thu, 14 Jan 2016 12:36:31 +0000 (12:36 +0000)]
* lib/target-supports.exp
(check_effective_target_arm_neon_ok_nocache): Add an option
sequence that includes setting the ARM architecture to ARMv7-A.
* gcc.target/arm/attr-neon.c: Use dg-add-options to add the
command line options necessary to enable Neon support.
* gcc.target/arm/neon-vlshr-imm-1.c: Likewise.
* gcc.target/arm/neon-vshl-imm-1.c: Likewise.
* gcc.target/arm/neon-vshr-imm-1.c: Likewise.
* gcc.target/arm/pr69180.c: Likewise.
Jeff Law [Thu, 14 Jan 2016 07:38:18 +0000 (00:38 -0700)]
[PATCH][PR tree-optimization/69270] Exploit VRP information in DOM
PR tree-optimization/69270
* tree-ssa-dom.c (ssa_name_has_boolean_range): New function.
(record_edge_info): Use it. Convert boolean_{true,false}_node
to the type of op0.
PR tree-optimization/69270
* gcc.dg/tree-ssa/pr69270.c: New test.
gcc/
* doc/invoke.texi (Submodel Options): Rename section to
"Machine-Dependent Options" to better reflect its content.
Rewrite introductory text to remove archaic CPU names.
Update references.
gcc/
* doc/invoke.texi (Code Gen Options): Move section up in file,
before target-specific options. Update menu and option summary
to reflect the new section ordering.
Michael Meissner [Wed, 13 Jan 2016 20:07:12 +0000 (20:07 +0000)]
rs6000-builtin.def (BU_FLOAT128_2): Add support for pack/unpack functions for __ibm128.
[gcc]
2016-01-13 Michael Meissner <meissner@linux.vnet.ibm.com>
* config/rs6000/rs6000-builtin.def (BU_FLOAT128_2): Add support
for pack/unpack functions for __ibm128.
(PACK_IF): Likewise.
(UNPACK_IF): Likewise.
* config/rs6000/rs6000.c (rs6000_builtin_mask_calculate): Add
support for __ibm128 pack/unpack functions.
(rs6000_invalid_builtin): Likewise.
(rs6000_init_builtins): Likewise.
(rs6000_opt_masks): Likewise.
* config/rs6000/rs6000.h (MASK_FLOAT128): Add short name.
(RS6000_BTM_FLOAT128): Add support for __ibm128 pack/unpack
functions
(RS6000_BTM_COMMON): Likewise.
* config/rs6000/rs6000.md (f128_vsx): New mode attribute.
(unpack<mode>): Use FMOVE128_FPR iterator instead of FMOVE128, to
disallow __builtin_{pack,unpack}_longdouble if long double is IEEE
128-bit floating point. Add support for the double values to be
in Altivec registers for TF/IF packing and unpacking, but restrict
TD packing sub-fields to be FPR registers. Don't allow overlapped
register support for packing. Allow pack inputs to be memory
locations. Don't build generator functions for unpack<mode>_dm
and unpack<mode>_nodm.
(unpack<mode>_dm): Likewise.
(unpack<mode>_nodm): Likewise.
(pack<mode>): Likewise.
* doc/extend.texi (PowerPC Built-in Functions): Document
__builtin_pack_ibm128 and __builtin_unpack_ibm128.
[libgcc]
2016-01-13 Michael Meissner <meissner@linux.vnet.ibm.com>
Steven Munroe <munroesj@linux.vnet.ibm.com>
Tulio Magno Quites Machado Filho <tulioqm@br.ibm.com>
* config/rs6000/sfp-exceptions.c: New file to provide exception
support for IEEE 128-bit floating point.
* config/rs6000/float128-hw.c: New file for ISA 3.0 IEEE 128-bit
floating point hardware support.
* config/rs6000/floattikf.c: New files for IEEE 128-bit floating
point conversions.
* config/rs6000/fixunskfti.c: Likewise.
* config/rs6000/fixkfti.c: Likewise.
* config/rs6000/floatuntikf.c: Likewise.
* config/rs6000/extendkftf2-sw.c: Likewise.
* config/rs6000/trunctfkf2-sw.c: Likewise.
* config/rs6000/float128-ifunc.c: New file to pick either IEEE
128-bit floating point software emulation or use ISA 3.0 hardware
support if it is available.
* config/rs6000/quad-float128.h: New file to support IEEE 128-bit
floating point.
* config/rs6000/t-float128: New Makefile fragments to enable
building __float128 emulation support.
* config/rs6000/t-float128-hw: Likewise.
* config/rs6000/float128-sed: New file to convert TF names to KF
names for PowerPC IEEE 128-bit floating point support.
* config/rs6000/sfp-machine.h (_FP_W_TYPE_SIZE): Use 64-bit types
when building on 64-bit systems, or when VSX is enabled.
(_FP_W_TYPE): Likewise.
(_FP_WS_TYPE): Likewise.
(_FP_I_TYPE): Likewise.
(TItype): Define on 64-bit systems.
(UTItype): Likewise.
(TI_BITS): Likewise.
(_FP_MUL_MEAT_D): Add support for using 64-bit types.
(_FP_MUL_MEAT_Q): Likewise.
(_FP_DIV_MEAT_D): Likewise.
(_FP_DIV_MEAT_Q): Likewise.
(_FP_NANFRAC_D): Likewise.
(_FP_NANFRAC_Q): Likewise.
(ISA_BIT): Add exception support if we are being compiled on a
machine with hardware floating point support to build the IEEE
128-bit emulation functions.
(FP_EX_INVALID): Likewise.
(FP_EX_OVERFLOW): Likewise.
(FP_EX_UNDERFLOW): Likewise.
(FP_EX_DIVZERO): Likewise.
(FP_EX_INEXACT): Likewise.
(FP_EX_ALL): Likewise.
(__sfp_handle_exceptions): Likewise.
(FP_HANDLE_EXCEPTIONS): Likewise.
(FP_RND_NEAREST): Likewise.
(FP_RND_ZERO): Likewise.
(FP_RND_PINF): Likewise.
(FP_RND_MINF): Likewise.
(FP_RND_MASK): Likewise.
(_FP_DECL_EX): Likewise.
(FP_INIT_ROUNDMODE): Likewise.
(FP_ROUNDMODE): Likewise.
* configure.ac (powerpc*-*-linux*): Check whether the PowerPC
compiler can do __float128.
* configure: Regenerate.
* libgcc/config.host (powerpc*-*-linux*): If compiler can compile
VSX code, enable IEEE 128-bit floating point.
Bernd Schmidt [Wed, 13 Jan 2016 20:03:26 +0000 (20:03 +0000)]
Improve warning locations (PR66208)
PR c/66208
* c-common.c (check_function_nonnull): Remove unnecessary declaration.
Add new arg loc and pass it down as context.
(check_nonnull_arg): Don't mark ctx arg as unused. Use it as a pointer
to the location to use for the warning.
(check_function_arguments): New arg loc. All callers changed. Pass
it to check_function_nonnull.
* c-common.h (check_function_arguments): Adjust declaration.
testsuite/
PR c/66208
* c-c++-common/pr66208.c: New file.
Rainer Orth [Wed, 13 Jan 2016 09:48:53 +0000 (09:48 +0000)]
Introduce separate baselines for Solaris 10 and 11+ (PR libstdc++/66006)
PR libstdc++/66006
* configure.host (abi_baseline_pair): Use separate baseline for
Solaris 11+ and Solaris 10 with gld.
* config/abi/post/solaris2.11/amd64/baseline_symbols.txt: New file.
* config/abi/post/solaris2.11/baseline_symbols.txt: New file.
* config/abi/post/solaris2.11/sparcv9/baseline_symbols.txt: New file.
PR tree-optimization/pr67755
* tree-ssa-threadupdate.c (struct ssa_local_info_t): Add new field
"need_profile_correction".
(thread_block_1): Initialize new field to false by default. If we
have multiple thread paths through a common joiner to different
final targets, then set new field to true.
(compute_path_counts): Only do count adjustment when it's really
needed.
PR tree-optimization/67755
* gcc.dg/tree-ssa/pr67755.c: New test.
gcc/
* doc/invoke.texi (Option Summary) Move -no-canonical-prefixes to
Directory Options, and -specs= to Overall Options.
(Overall Options): Adjust similarly. Reorder to group related
options together. Make -specs= cross-reference the spec file details.
(Directory Options): Adjust similarly.
Olivier Hainque [Tue, 12 Jan 2016 17:53:14 +0000 (17:53 +0000)]
Prevent "undef var" errors on gcc --help or --version with self-specs
* gcc.c (spec_undefvar_allowed): New global.
(process_command): Set to true when running for --version or --help,
alone or together.
(getenv_spec_function): When the variable is not defined, use the
variable name as the variable value if we're allowed not to issue
a fatal error.
gcc/
* doc/invoke.texi (Invoking GCC): Copy-edit. Incorporate information
about name of GCC executable. Remove deleted node from menu.
(Directory Options) <-B>: Remove cross-reference to deleted node.
(Target Options): Delete section.
Jonathan Wakely [Tue, 12 Jan 2016 14:54:33 +0000 (14:54 +0000)]
Prevent recursive instantiation in std::function
PR libstdc++/69005
PR libstdc++/69222
* include/std/functional (function::_Invoke): Remove, use result_of.
(function::_Callable): Replace alias template with class template
and use partial specialization instead of _NotSelf alias template.
(function(_Functor)): Add "not self" constraint so that _Callable is
not used while type is incomplete.
* testsuite/20_util/function/69222.cc: New.
Jakub Jelinek [Tue, 12 Jan 2016 13:20:33 +0000 (14:20 +0100)]
re PR target/69198 (FAIL: gcc.target/i386/avx512vl-vmovaps-1.c scan-assembler-times vmovaps[ \\t]+[^{\n]*%xmm[0-9]+[^\n]*\\){%k[1-7]}(?:\n|[ \\t]+#) 1)
PR target/69198
* config/i386/i386.c (ix86_expand_special_args_builtin): Ensure
aligned_mem is properly set for AVX512-VL floating point masked
stores.
Richard Biener [Tue, 12 Jan 2016 08:36:24 +0000 (08:36 +0000)]
re PR tree-optimization/69168 (ICE in vect_analyze_loop_2, at tree-vect-loop.c:2182)
2016-01-12 Richard Biener <rguenther@suse.de>
PR tree-optimization/69168
* tree-vect-loop.c (vect_analyze_loop_2): Reset both main and
pattern stmt SLP type.
* tree-vect-slp.c (vect_detect_hybrid_slp_stmts): Patterns may
end up unused so cope with that case.
Richard Biener [Tue, 12 Jan 2016 08:30:44 +0000 (08:30 +0000)]
re PR tree-optimization/69157 (ICE in vect_transform_stmt, at tree-vect-stmts.c:8176)
2016-01-12 Richard Biener <rguenther@suse.de>
PR tree-optimization/69157
* tree-vect-stmts.c (vectorizable_mask_load_store): Check
stmts def type only during analyze phase.
(vectorizable_call): Likewise.
(vectorizable_simd_clone_call): Likewise.
(vectorizable_conversion): Likewise.
(vectorizable_assignment): Likewise.
(vectorizable_shift): Likewise.
(vectorizable_operation): Likewise.
(vectorizable_store): Likewise.
(vectorizable_load): Likewise.
* gcc.dg/torture/pr69157.c: New testcase.
2016-01-12 Richard Biener <rguenther@suse.de>
PR tree-optimization/69174
* tree-vect-stmts.c (vect_mark_relevant): Remove excessive vertical
space.
(vectorizable_load): Properly compute the number of loads needed
for permuted strided SLP loads and do not spuriously assign
to SLP_TREE_VEC_STMTS.
Andris Pavenis [Tue, 12 Jan 2016 04:57:10 +0000 (06:57 +0200)]
[PATCH 5/6] [DJGPP] Update DJGPP configuration related files
* config/i386/djgpp.h (PREFERRED_DEBUGGING_TYPE): Define to DWARF2
(TARGET_ASM_OUTPUT_IDENT): Define to default_asm_output_ident_directive
(MD_EXEC_PREFIX): Remove.
(MD_STARTFILE_PREFIX) Removee.
(FILE_NAME_ABSOLUTE_P): Remove.
(CPP_SPEC): Do not read macros from sys/version.h.
(LINK_COMMAND_SPEC): Remove.
(LOCAL_INCLUDE_DIR): Remove.
(TARGET_ASM_NAMED_SECTION): Define to i386_djgpp_asm_named_section
(TARGET_OS_CPP_BUILTINS): Add DJGPP (non ISO only), __DJGPP, __DJGPP__, unix.
(POST_LINK_SPEC): Define to invoke stubify after linker
(LIBSTDCXX): Remove define
(DBX_REGISTER_NUMBER): Define to svr4_dbx_register_map.
(DEFAULT_PCC_STRUCT_RETURN): Define to 1.
(SUBTARGET_OVERRIDE_OPTIONS): Remove warning about -mbnu2210.
(SUBTARGET_OVERRIDE_OPTIONS): Ignore -fPIC and generate message.
(SUBTARGET_OVERRIDE_OPTIONS): Default to DWARF2 debugging info.
(IX86_MAYBE_NO_LIBGCC_TFMODE): Remove.
(i386_djgpp_asm_named_section): Add propotype of new procedure
* config/i386/xm-djgpp.h (NATIVE_SYSTEM_HEADER_DIR): Define.
(MD_EXEC_PREFIX): Define (moved from config/i386/djgpp.h).
(STANDARD_STARTFILE_PREFIX_1): Define (moved from MD_STARTFILE_PREFIX in config/i386/djgpp.h).
(STANDARD_STARTFILE_PREFIX_2): Define identical to STANDARD_STARTFILE_PREFIX_1.
(LOCAL_INCLUDE_DIR): Define (moved from config/i386/djgpp.h).
(GCC_DRIVER_HOST_INITIALIZATION): Fix reporting fatal installation errors.
(MAX_OFILE_ALIGNMENT): Define to 128.
(HAVE_FTW_H): Undefine as DJGPP do not have nftw, but have ftw.h.
* config/i386/djgpp.c: New file. Add implementation of i386_djgpp_asm_named_section.
Bill Schmidt [Tue, 12 Jan 2016 04:49:55 +0000 (04:49 +0000)]
rs6000.c (v2df_reduction_p): New function.
[gcc]
2016-01-11 Bill Schmidt <wschmidt@linux.vnet.ibm.com>
* config/rs6000/rs6000.c (v2df_reduction_p): New function.
(rtx_is_swappable_p): Reductions are swappable.
(insn_is_swappable_p): V2DF reductions are swappable.
[gcc/testsuite]
2016-01-11 Bill Schmidt <wschmidt@linux.vnet.ibm.com>
* gcc.target/powerpc/swaps-p8-23.c: New test.
* gcc.target/powerpc/swaps-p8-24.c: Likewise.
Jakub Jelinek [Mon, 11 Jan 2016 19:06:34 +0000 (20:06 +0100)]
re PR target/69071 (ICE: in decompose, at rtl.h:2107 with -g)
PR target/69071
* lra-eliminations.c (move_plus_up): Only move plus up
if subreg of the constant can be simplified into constant
and use the simplified subreg of the constant instead of
the original constant.
David Malcolm [Mon, 11 Jan 2016 18:03:15 +0000 (18:03 +0000)]
PR c++/68795: fix uninitialized close_paren_loc in cp_parser_postfix_expression
gcc/cp/ChangeLog:
PR c++/68795
* parser.c (cp_parser_postfix_expression): Initialize
close_paren_loc to UNKNOWN_LOCATION; only use it if
it has been written to by
cp_parser_parenthesized_expression_list.
(cp_parser_parenthesized_expression_list): Document the behavior
with respect to the CLOSE_PAREN_LOC param.
Jakub Jelinek [Mon, 11 Jan 2016 17:59:22 +0000 (18:59 +0100)]
re PR c++/69211 (g++ ICE on x86_64-linux-gnu (verify_gimple failed))
PR c++/69211
* cp-gimplify.c (cp_fold): If COMPOUND_EXPR or MODIFY_EXPR
folded operands have side-effects, but folding changed any of them,
build a new tree with the folded operands instead of returning the
unfolded tree.
Jakub Jelinek [Mon, 11 Jan 2016 17:57:43 +0000 (18:57 +0100)]
fold-const.c (fold_convertible_p): Don't return true for conversion of VECTOR_TYPE to same sized integral type.
* fold-const.c (fold_convertible_p): Don't return true
for conversion of VECTOR_TYPE to same sized integral type.
(fold_convert_loc): Fix up formatting. Fold conversion of
VECTOR_TYPE to same sized integral type using VIEW_CONVERT_EXPR
instead of NOP_EXPR.
Jakub Jelinek [Mon, 11 Jan 2016 17:56:39 +0000 (18:56 +0100)]
re PR tree-optimization/69214 (ICE (segfault) at -Os on x86_64-linux-gnu in "fail_abnormal_edge_coalesce")
PR tree-optimization/69214
* tree-vrp.c (simplify_cond_using_ranges): Don't propagate
innerop into a comparison if SSA_NAME_OCCURS_IN_ABNORMAL_PHI.
Formatting fix.
Ian Lance Taylor [Mon, 11 Jan 2016 17:55:39 +0000 (17:55 +0000)]
re PR go/68980 (ps -o cmd in gotest isn't portable)
PR 68980
libgo/testsuite: portable ps usage in gotest
gotest is using "ps" to list descendant sleep
processes in its timeout handling, grepping
the command name.
We are currently using the "cmd" ps output
formatter, which is non-portable. We should
use "comm" which is part of the POSIX
standard, and outputs only the command name
and not the arguments.
Jakub Jelinek [Mon, 11 Jan 2016 17:55:38 +0000 (18:55 +0100)]
re PR tree-optimization/69207 (gcc.target/aarch64/vldN_1.c ICEs at -O3)
PR tree-optimization/69207
* tree-vect-slp.c (vect_get_constant_vectors): For
VECTOR_BOOLEAN_TYPE_P, assert op has integral type instead of
fold_convertible_p to vector_type's element type, and always
use VCE for non-VECTOR_BOOLEAN_TYPE_P.
Uros Bizjak [Mon, 11 Jan 2016 15:48:40 +0000 (16:48 +0100)]
re PR middle-end/68999 (FAIL: gfortran.fortran-torture/execute/save_1.f90 execution)
PR middle-end/68999
* alias.c (base_alias_check): Move check for addresses with
alignment ANDs before the call for compare_base_decls.
(memrefs_conflict_p): Return -1 for different decls
that went through alignment adjustments.
Kyrylo Tkachov [Mon, 11 Jan 2016 14:44:22 +0000 (14:44 +0000)]
[AArch64] PR rtl-optimization/68796: Add patterns for QImode and HImode comparison with zero
PR rtl-optimization/68796
* config/aarch64/aarch64.md (*and<mode>_compare0): New pattern.
* config/aarch64/aarch64.c (aarch64_select_cc_mode): Handle HImode
and QImode comparisons against zero with CC_NZmode.
* config/aarch64/iterators.md (short_mask): New mode_attr.
* gcc.target/aarch64/tst_5.c: New test.
* gcc.target/aarch64/tst_6.c: Likewise.
When link_ptx runs, a CUDA device is already bound to current thread, so the
driver library knows the target architecture. There isn't any benefit from
forcing a specific target here; on the contrary, hardcoding sm_30 breaks
offloading on later (Maxwell, sm_5x) devices.
* plugin/plugin-nvptx.c (link_ptx): Do not set CU_JIT_TARGET.
Kyrylo Tkachov [Mon, 11 Jan 2016 12:13:50 +0000 (12:13 +0000)]
[RTL-ifcvt] PR rtl-optimization/68841: Make sure one basic block doesn't clobber CC reg usage of the other
PR rtl-optimization/68841
* ifcvt.c (struct noce_if_info): Add orig_x field.
(bbs_ok_for_cmove_arith): Add to_rename parameter.
Don't record conflicts on to_rename if it's present.
Allow memory destinations in sets.
(noce_try_cmove_arith): Call bbs_ok_for_cmove_arith even on simple
blocks, passing orig_x to the checks.
(noce_process_if_block): Set if_info->orig_x appropriately.
* gcc.dg/pr68841.c: New test.
* gcc.c-torture/execute/pr68841.c: New test.
Yuri Rumyantsev [Mon, 11 Jan 2016 12:07:31 +0000 (12:07 +0000)]
re PR rtl-optimization/68920 (Undesirable if-conversion for a rarely taken branch)
gcc/
2016-01-11 Yuri Rumyantsev <ysrumyan@gmail.com>
PR rtl-optimization/68920
* config/i386/i386.c (ix86_option_override_internal): Restrict number
of conditional moves for RTL if-conversion to 1 for
TARGET_ONE_IF_CONV_INSN.
* config/i386/i386.h (TARGET_ONE_IF_CONV_INSN): New macros.
* config/i386/x86-tune.def (X86_TUNE_ONE_IF_CONV_INSN): New macros.
* params.def (PARAM_MAX_RTL_IF_CONVERSION_INSNS) : Introduce new
parameter to restirct number of conditional moves for
RTL if-conversion.
* doc/invoke.texi (max-rtl-if-conversion-insns): Document it.
* ifcvt.c (bb_ok_for_noce_convert_multiple_sets): Limit number of
conditionl moves.
gcc/testsuite/
2016-01-11 Yuri Rumyantsev <ysrumyan@gmail.com>
PR rtl-optimization/68920
* gcc.dg/ifcvt-4.c: Add "--param max-rtl-if-conversion-insns=3" option
for ix86 targets.
* gcc.dg/ifcvt-5.c: New test.
Alexandre Oliva [Mon, 11 Jan 2016 10:40:33 +0000 (10:40 +0000)]
[PR69123] fix handling of MEMs in VTA to avoid dataflow oscillation
The problem arises because we used to drop overwritten MEMs from loc
lists of VALUEs, but not of other onepart variables, and it just so
happens that, by doing so, block 6 in the testcase has no D#5 in its
output in the first pass, because the MEM holding its (previous) value
was correctly dropped from value 88:88, but gains it in the second
pass because D#5 has the MEM location incoming directly in its loc
list, rather than indirectly in a VALUE.
This incorrect binding enables other blocks to believe they have a
tentative binding for D#5 in some cycles, but others, still operating
on the early conclusion, believe there isn't, and they oscillate from
that.
Since we check for escaping MEMs in clobbers, we won't lose anything
relevant by dropping call-clobbered or overwritten MEMs in all onepart
variables, and this ensures the loc intersection operation in onepart
vars won't let a MEM through that wasn't present in earlier
iterations.
for gcc/ChangeLog
PR bootstrap/69123
* var-tracking.c (drop_overlapping_mem_locs): Operate on all
onepart vars. Fix typo in comment. Fix reversed condition in
unshare test.
(dataflow_set_remove_mem_locs): Operate on all onepart vars.
projects / thirdparty / gcc.git / log
