* src/c++11/codecvt.cc (__codecvt_utf8_base<wchar_t>::do_in)
[__SIZEOF_WCHAR_T__==2 && __BYTE_ORDER__!=__ORDER_BIG_ENDIAN__]: Set
little_endian element in bitmask.
* testsuite/22_locale/codecvt/codecvt_utf8/69703.cc: Run all tests.
* testsuite/22_locale/codecvt/codecvt_utf8/wchar_t/1.cc: New.
From-SVN: r260389
In C++11 and C++14 any odr-use of these constants requires a definition
at namespace-scope. In C++17 they are implicitly inline and so the
namespace-scope redeclarations are redundant (and allowing them is
deprecated).
PR libstdc++/85098
* include/bits/regex.h [__cplusplus < 201703L] (basic_regex::icase)
(basic_regex::nosubs, basic_regex::optimize, basic_regex::collate)
(basic_regex::ECMAScript, basic_regex::basic, basic_regex::extended)
(basic_regex::awk, basic_regex::grep, basic_regex::egrep): Add
definitions.
* include/bits/regex_automaton.h (_NFA::_M_insert_state): Adjust
whitespace.
* include/bits/regex_compiler.tcc (__INSERT_REGEX_MATCHER): Add
braces around body of do-while.
* testsuite/28_regex/basic_regex/85098.cc: New
From-SVN: r260371
We deprecated architecture versions earlier than Armv4T in GCC 6 [1].
This patch removes support for architectures lower than Armv4.
That is the -march values armv2, armv2a, armv3, armv3m are removed
with this patch. I did not remove armv4 because it's a bit more
involved code-wise and there has been some pushback on the implications
for -mcpu=strongarm support.
Removing armv3m and earlier though is pretty straightforward.
This allows us to get rid of the armv3m and mode32 feature bits
in arm-cpus.in as they can be assumed to be universally available.
Consequently the mcpu values arm2, arm250, arm3, arm6, arm60, arm600, arm610, arm620, arm7, arm7d, arm7di, arm70, arm700, arm700i, arm710, arm720, arm710c, arm7100, arm7500, arm7500fe, arm7m, arm7dm, arm7dm are now also removed.
Bootstrapped and tested on arm-none-linux-gnueabihf and on arm-none-eabi
with an aprofile multilib configuration (which builds quite a lot of library
configurations).
[1] https://gcc.gnu.org/gcc-6/changes.html#arm
* config/arm/arm-cpus.in (armv3m, mode32): Delete features.
(ARMv4): Update.
(ARMv2, ARMv3, ARMv3m): Delete fgroups.
(ARMv6m): Update.
(armv2, armv2a, armv3, armv3m): Delete architectures.
(arm2, arm250, arm3, arm6, arm60, arm600, arm610, arm620,
arm7, arm7d, arm7di, arm70, arm700, arm700i, arm710, arm720,
arm710c, arm7100, arm7500, arm7500fe, arm7m, arm7dm, arm7dmi):
Delete cpus.
* config/arm/arm.md (maddsidi4): Remove check for arm_arch3m.
(*mulsidi3adddi): Likewise.
(mulsidi3): Likewise.
(*mulsidi3_nov6): Likewise.
(umulsidi3): Likewise.
(umulsidi3_nov6): Likewise.
(umaddsidi4): Likewise.
(*umulsidi3adddi): Likewise.
(smulsi3_highpart): Likewise.
(*smulsi3_highpart_nov6): Likewise.
(umulsi3_highpart): Likewise.
(*umulsi3_highpart_nov6): Likewise.
* config/arm/arm.h (arm_arch3m): Delete.
* config/arm/arm.c (arm_arch3m): Delete.
(arm_option_override_internal): Update armv3-related comment.
(arm_configure_build_target): Delete use of isa_bit_mode32.
(arm_option_reconfigure_globals): Delete set of arm_ach3m.
(arm_rtx_costs_internal): Delete check of arm_arch3m.
* config/arm/arm-fixed.md (mulsq3): Delete check for arm_arch3m.
(mulsa3): Likewise.
(mulusa3): Likewise.
* config/arm/arm-protos.h (arm_arch3m): Delete.
* config/arm/arm-tables.opt: Regenerate.
* config/arm/arm-tune.md: Likewise.
* config/arm/t-arm-elf (all_early_nofp): Delete mentions of
deleted architectures.
* gcc.target/arm/pr62554.c: Delete.
* gcc.target/arm/pr69610-1.c: Likewise.
* gcc.target/arm/pr69610-2.c: Likewise.
From-SVN: r260363
The -march=armv5 and armv5e options have been deprecated in GCC 7 [1].
This patch removes support for them.
It's mostly mechanical stuff. The functionality that was previously
gated on arm_arch5 is now gated on arm_arch5t and the functionality
that was gated on arm_arch5e is now gated on arm_arch5te.
A path in TARGET_OS_CPP_BUILTINS for VxWorks is now unreachable and
therefore is deleted.
References to armv5 and armv5e are deleted/updated throughout the
source tree and testsuite.
Bootstrapped and tested on arm-none-linux-gnueabihf.
Also built a cc1 for arm-wrs-vxworks as a sanity check.
* config/arm/arm-cpus.in (armv5, armv5e): Delete features.
(armv5t, armv5te): New features.
(ARMv5, ARMv5e): Delete fgroups.
(ARMv5t, ARMv5te): Adjust for above changes.
(ARMv6m): Likewise.
(armv5, armv5e): Delete arches.
* config/arm/arm.md (*call_reg_armv5): Use arm_arch5t instead of
arm_arch5.
(*call_reg_arm): Likewise.
(*call_value_reg_armv5): Likewise.
(*call_value_reg_arm): Likewise.
(*call_symbol): Likewise.
(*call_value_symbol): Likewise.
(*sibcall_insn): Likewise.
(*sibcall_value_insn): Likewise.
(clzsi2): Likewise.
(prefetch): Likewise.
(define_split and define_peephole2 dependent on arm_arch5):
Likewise.
* config/arm/arm.h (TARGET_LDRD): Use arm_arch5te instead of
arm_arch5e.
(TARGET_ARM_QBIT): Likewise.
(TARGET_DSP_MULTIPLY): Likewise.
(enum base_architecture): Delete BASE_ARCH_5, BASE_ARCH_5E.
(arm_arch5, arm_arch5e): Delete.
(arm_arch5t, arm_arch5te): Declare.
* config/arm/arm.c (arm_arch5, arm_arch5e): Delete.
(arm_arch5t): Declare.
(arm_option_reconfigure_globals): Update for the above.
(arm_options_perform_arch_sanity_checks): Update comment, replace
use of arm_arch5 with arm_arch5t.
(use_return_insn): Likewise.
(arm_emit_call_insn): Likewise.
(output_return_instruction): Likewise.
(arm_final_prescan_insn): Likewise.
(arm_coproc_builtin_available): Likewise.
* config/arm/arm-c.c (arm_cpu_builtins): Replace arm_arch5 and
arm_arch5e with arm_arch5t and arm_arch5te.
* config/arm/arm-protos.h (arm_arch5, arm_arch5e): Delete.
(arm_arch5t, arm_arch5te): Declare.
* config/arm/arm-tables.opt: Regenerate.
* config/arm/t-arm-elf: Remove references to armv5, armv5e.
* config/arm/t-multilib: Likewise.
* config/arm/thumb1.md (*call_reg_thumb1_v5): Check arm_arch5t
instead of arm_arch5.
(*call_reg_thumb1): Likewise.
(*call_value_reg_thumb1_v5): Likewise.
(*call_value_reg_thumb1): Likewise.
* config/arm/vxworks.h (TARGET_OS_CPP_BUILTINS): Remove now
unreachable path.
* doc/invoke.texi (ARM Options): Remove references to armv5, armv5e.
* gcc.target/arm/pr40887.c: Update comment.
* lib/target-supports.exp: Don't generate effective target checks
and related helpers for armv5. Update comment.
* gcc.target/arm/armv5_thumb_isa.c: Delete.
* gcc.target/arm/di-longlong64-sync-withhelpers.c: Update effective
target check and options.
* config/arm/libunwind.S: Update comment relating to armv5.
From-SVN: r260362
2018-05-18 Martin Liska <mliska@suse.cz>
PR gcov-profile/84846
* gcov.c (print_usage): Add new -t option.
(process_args): Handle the option.
(generate_results): Use stdout as output when requested by
the option.
2018-05-18 Martin Liska <mliska@suse.cz>
PR gcov-profile/84846
* doc/gcov.texi: Document -t option of gcov tool.
From-SVN: r260361
2018-05-18 Prathamesh Kulkarni <prathamesh.kulkarni@linaro.org>
PR middle-end/85817
* ipa-pure-const.c (malloc_candidate_p): Remove the check integer_zerop
for retval and return false if all args to phi are zero.
testsuite/
* gcc.dg/tree-ssa/pr83648.c: Change scan-tree-dump to
scan-tree-dump-not for h.
From-SVN: r260358
2018-05-18 Toon Moene <toon@moene.org>
* doc/invoke.texi: Move -floop-unroll-and-jam documentation
directly after that of -floop-interchange. Indicate that both
options are enabled by default when specifying -O3.
From-SVN: r260352
We've a deficiency in our vec_set family of patterns.
We don't support directly loading a vector lane using LD1 for V2DImode and all the vector floating-point modes.
We do do it correctly for the other integer vector modes (V4SI, V8HI etc) though.
The alternatives on the relative floating-point patterns only allow a register-to-register INS instruction.
That means if we want to load a value into a vector lane we must first load it into a scalar register and then
perform an INS, which is wasteful.
There is also an explicit V2DI vec_set expander dangling around for no reason that I can see. It seems to do the
exact same things as the other vec_set expanders. This patch removes that.
It now unifies all vec_set expansions into a single "vec_set<mode>" define_expand using the catch-all VALL_F16 iterator.
With this patch we avoid loading values into scalar registers and then doing an explicit INS on them to move them into
the desired vector lanes. For example for:
typedef float v4sf __attribute__ ((vector_size (16)));
typedef long long v2di __attribute__ ((vector_size (16)));
v2di
foo_v2di (long long *a, long long *b)
{
v2di res = { *a, *b };
return res;
}
v4sf
foo_v4sf (float *a, float *b, float *c, float *d)
{
v4sf res = { *a, *b, *c, *d };
return res;
}
we currently generate:
foo_v2di:
ldr d0, [x0]
ldr x0, [x1]
ins v0.d[1], x0
ret
foo_v4sf:
ldr s0, [x0]
ldr s3, [x1]
ldr s2, [x2]
ldr s1, [x3]
ins v0.s[1], v3.s[0]
ins v0.s[2], v2.s[0]
ins v0.s[3], v1.s[0]
ret
but with this patch we generate the much cleaner:
foo_v2di:
ldr d0, [x0]
ld1 {v0.d}[1], [x1]
ret
foo_v4sf:
ldr s0, [x0]
ld1 {v0.s}[1], [x1]
ld1 {v0.s}[2], [x2]
ld1 {v0.s}[3], [x3]
ret
* config/aarch64/aarch64-simd.md (vec_set<mode>): Use VALL_F16 mode
iterator. Delete separate integer-mode vec_set<mode> expander.
(aarch64_simd_vec_setv2di): Delete.
(vec_setv2di): Delete.
(aarch64_simd_vec_set<mode>): Delete all other patterns with that name.
Use VALL_F16 mode iterator. Add LD1 alternative and use vwcore for
the "w, r" alternative.
* gcc.target/aarch64/vect-init-ld1.c: New test.
From-SVN: r260351
2018-05-18 Martin Liska <mliska@suse.cz>
* passes.def: Add pass_lower_switch and pass_lower_switch_O0.
* tree-pass.h (make_pass_lower_switch_O0): New function.
* tree-switch-conversion.c (node_has_low_bound): Remove.
(node_has_high_bound): Likewise.
(node_is_bounded): Likewise.
(class pass_lower_switch): Make it a template type and create
two instances.
(pass_lower_switch::execute): Add template argument.
(make_pass_lower_switch): New function.
(make_pass_lower_switch_O0): New function.
(do_jump_if_equal): Remove.
(emit_case_nodes): Simplify to just handle all 3 cases and leave
all the hard work to tree optimization passes.
2018-05-18 Martin Liska <mliska@suse.cz>
* gcc.dg/tree-ssa/vrp104.c: Adjust dump file that is scanned.
* gcc.dg/tree-prof/update-loopch.c: Likewise.
From-SVN: r260350
There are four optabs for various forms of fused multiply-add:
fma, fms, fnma and fnms. Of these, only fma had a direct gimple
representation. For the other three we relied on special pattern-
matching during expand, although tree-ssa-math-opts.c did have
some code to try to second-guess what expand would do.
This patch removes the old FMA_EXPR representation of fma and
introduces four new internal functions, one for each optab.
IFN_FMA is tied to BUILT_IN_FMA* while the other three are
independent directly-mapped internal functions. It's then
possible to do the pattern-matching in match.pd and
tree-ssa-math-opts.c (via folding) can select the exact
FMA-based operation.
The BRIG & HSA parts are a best guess, but seem relatively simple.
2018-05-18 Richard Sandiford <richard.sandiford@linaro.org>
gcc/
* doc/sourcebuild.texi (scalar_all_fma): Document.
* tree.def (FMA_EXPR): Delete.
* internal-fn.def (FMA, FMS, FNMA, FNMS): New internal functions.
* internal-fn.c (ternary_direct): New macro.
(expand_ternary_optab_fn): Likewise.
(direct_ternary_optab_supported_p): Likewise.
* Makefile.in (build/genmatch.o): Depend on case-fn-macros.h.
* builtins.c (fold_builtin_fma): Delete.
(fold_builtin_3): Don't call it.
* cfgexpand.c (expand_debug_expr): Remove FMA_EXPR handling.
* expr.c (expand_expr_real_2): Likewise.
* fold-const.c (operand_equal_p): Likewise.
(fold_ternary_loc): Likewise.
* gimple-pretty-print.c (dump_ternary_rhs): Likewise.
* gimple.c (DEFTREECODE): Likewise.
* gimplify.c (gimplify_expr): Likewise.
* optabs-tree.c (optab_for_tree_code): Likewise.
* tree-cfg.c (verify_gimple_assign_ternary): Likewise.
* tree-eh.c (operation_could_trap_p): Likewise.
(stmt_could_throw_1_p): Likewise.
* tree-inline.c (estimate_operator_cost): Likewise.
* tree-pretty-print.c (dump_generic_node): Likewise.
(op_code_prio): Likewise.
* tree-ssa-loop-im.c (stmt_cost): Likewise.
* tree-ssa-operands.c (get_expr_operands): Likewise.
* tree.c (commutative_ternary_tree_code, add_expr): Likewise.
* fold-const-call.h (fold_fma): Delete.
* fold-const-call.c (fold_const_call_ssss): Handle CFN_FMS,
CFN_FNMA and CFN_FNMS.
(fold_fma): Delete.
* genmatch.c (combined_fn): New enum.
(commutative_ternary_tree_code): Remove FMA_EXPR handling.
(commutative_op): New function.
(commutate): Use it. Handle more than 2 operands.
(dt_operand::gen_gimple_expr): Use commutative_op.
(parser::parse_expr): Allow :c to be used with non-binary
operators if the commutative operand is known.
* gimple-ssa-backprop.c (backprop::process_builtin_call_use): Handle
CFN_FMS, CFN_FNMA and CFN_FNMS.
(backprop::process_assign_use): Remove FMA_EXPR handling.
* hsa-gen.c (gen_hsa_insns_for_operation_assignment): Likewise.
(gen_hsa_fma): New function.
(gen_hsa_insn_for_internal_fn_call): Use it for IFN_FMA, IFN_FMS,
IFN_FNMA and IFN_FNMS.
* match.pd: Add folds for IFN_FMS, IFN_FNMA and IFN_FNMS.
* gimple-fold.h (follow_all_ssa_edges): Declare.
* gimple-fold.c (follow_all_ssa_edges): New function.
* tree-ssa-math-opts.c (convert_mult_to_fma_1): Use the
gimple_build interface and use follow_all_ssa_edges to fold the result.
(convert_mult_to_fma): Use direct_internal_fn_suppoerted_p
instead of checking for optabs directly.
* config/i386/i386.c (ix86_add_stmt_cost): Recognize FMAs as calls
rather than FMA_EXPRs.
* config/rs6000/rs6000.c (rs6000_gimple_fold_builtin): Create a
call to IFN_FMA instead of an FMA_EXPR.
gcc/brig/
* brigfrontend/brig-function.cc
(brig_function::get_builtin_for_hsa_opcode): Use BUILT_IN_FMA
for BRIG_OPCODE_FMA.
(brig_function::get_tree_code_for_hsa_opcode): Treat BUILT_IN_FMA
as a call.
gcc/c/
* gimple-parser.c (c_parser_gimple_postfix_expression): Remove
__FMA_EXPR handlng.
gcc/cp/
* constexpr.c (cxx_eval_constant_expression): Remove FMA_EXPR handling.
(potential_constant_expression_1): Likewise.
gcc/testsuite/
* lib/target-supports.exp (check_effective_target_scalar_all_fma):
New proc.
* gcc.dg/fma-1.c: New test.
* gcc.dg/fma-2.c: Likewise.
* gcc.dg/fma-3.c: Likewise.
* gcc.dg/fma-4.c: Likewise.
* gcc.dg/fma-5.c: Likewise.
* gcc.dg/fma-6.c: Likewise.
* gcc.dg/fma-7.c: Likewise.
* gcc.dg/gimplefe-26.c: Use .FMA instead of __FMA and require
scalar_all_fma.
* gfortran.dg/reassoc_7.f: Pass -ffp-contract=off.
* gfortran.dg/reassoc_8.f: Likewise.
* gfortran.dg/reassoc_9.f: Likewise.
* gfortran.dg/reassoc_10.f: Likewise.
From-SVN: r260348
gcc/
* expr.c (do_tablejump): When converting index to Pmode, if we have a
sign extended promoted subreg, and the range does not have the sign bit
set, then do a sign extend.
* config/riscv/riscv.c (riscv_extend_comparands): In unsigned QImode
test, check for sign extended subreg and/or constant operands, and
do a sign extend in that case.
gcc/testsuite/
* gcc.target/riscv/switch-qi.c: New.
* gcc.target/riscv/switch-si.c: New.
From-SVN: r260340
Because path.cc is compiled with -std=gnu++17 the static constexpr
data member is implicitly 'inline' and so no definition gets emitted
unless it gets used in that translation unit. Other translation units
built as C++11 or C++14 still require a namespace-scope definition of
the variable, so mark the definition as used.
PR libstdc++/85818
* src/filesystem/path.cc (path::preferred_separator): Add used
attribute.
* testsuite/experimental/filesystem/path/preferred_separator.cc: New.
From-SVN: r260326
PR libstdc++/85812
* libsupc++/cxxabi_init_exception.h (__cxa_free_exception): Declare.
* libsupc++/exception_ptr.h (make_exception_ptr) [__cpp_exceptions]:
Refactor to separate non-throwing and throwing implementations.
[__cpp_rtti && !_GLIBCXX_HAVE_CDTOR_CALLABI]: Deallocate the memory
if constructing the object throws.
From-SVN: r260323
This patch gets the gimple FE to parse calls to internal functions.
The only non-obvious thing was how the functions should be written
to avoid clashes with real function names. One option would be to
go the magic number of underscores route, but we already do that for
built-in functions, and it would be good to keep them visually
distinct. In the end I borrowed the local/internal label convention
from asm and used:
x = .SQRT (y);
2018-05-17 Richard Sandiford <richard.sandiford@linaro.org>
gcc/
* internal-fn.h (lookup_internal_fn): Declare
* internal-fn.c (lookup_internal_fn): New function.
* gimple.c (gimple_build_call_from_tree): Handle calls to
internal functions.
* gimple-pretty-print.c (dump_gimple_call): Print "." before
internal function names.
* tree-pretty-print.c (dump_generic_node): Likewise.
* tree-ssa-scopedtables.c (expr_hash_elt::print): Likewise.
gcc/c/
* gimple-parser.c: Include internal-fn.h.
(c_parser_gimple_statement): Treat a leading CPP_DOT as a call.
(c_parser_gimple_call_internal): New function.
(c_parser_gimple_postfix_expression): Use it to handle CPP_DOT.
Fix typos in comment.
gcc/testsuite/
* gcc.dg/gimplefe-28.c: New test.
* gcc.dg/asan/use-after-scope-9.c: Adjust expected output for
internal function calls.
* gcc.dg/goacc/loop-processing-1.c: Likewise.
From-SVN: r260316
This patch makes the function versions of gimple_build and
gimple_simplify take combined_fns rather than built_in_codes,
so that they work with internal functions too. The old
gimple_builds were unused, so no existing callers need
to be updated.
2018-05-17 Richard Sandiford <richard.sandiford@linaro.org>
gcc/
* gimple-fold.h (gimple_build): Make the function forms take
combined_fn rather than built_in_function.
(gimple_simplify): Likewise.
* gimple-match-head.c (gimple_simplify): Likewise.
* gimple-fold.c (gimple_build): Likewise.
* tree-vect-loop.c (get_initial_def_for_reduction): Use gimple_build
rather than gimple_build_call_internal.
(get_initial_defs_for_reduction): Likewise.
(vect_create_epilog_for_reduction): Likewise.
(vectorizable_live_operation): Likewise.
From-SVN: r260315
2018-05-17 Martin Liska <mliska@suse.cz>
* gimple-ssa-sprintf.c (format_directive): Do not use
space in between 'G_' and '('.
2018-05-17 Martin Liska <mliska@suse.cz>
* c-warn.c (overflow_warning): Do not use
space in between 'G_' and '('.
2018-05-17 Martin Liska <mliska@suse.cz>
* gcc.dg/plugin/ggcplug.c (plugin_init): Do not use
space in between 'G_' and '('.
From-SVN: r260314
PR target/85323
* config/i386/i386.c (ix86_fold_builtin): Handle masked shifts
even if the mask is not all ones.
* gcc.target/i386/pr85323-7.c: New test.
* gcc.target/i386/pr85323-8.c: New test.
* gcc.target/i386/pr85323-9.c: New test.
From-SVN: r260313
* config/i386/avx512fintrin.h (_mm512_set_epi16, _mm512_set_epi8,
_mm512_setzero): New intrinsics.
* gcc.target/i386/avx512f-set-v32hi-1.c: New test.
* gcc.target/i386/avx512f-set-v32hi-2.c: New test.
* gcc.target/i386/avx512f-set-v32hi-3.c: New test.
* gcc.target/i386/avx512f-set-v32hi-4.c: New test.
* gcc.target/i386/avx512f-set-v32hi-5.c: New test.
* gcc.target/i386/avx512f-set-v64qi-1.c: New test.
* gcc.target/i386/avx512f-set-v64qi-2.c: New test.
* gcc.target/i386/avx512f-set-v64qi-3.c: New test.
* gcc.target/i386/avx512f-set-v64qi-4.c: New test.
* gcc.target/i386/avx512f-set-v64qi-5.c: New test.
* gcc.target/i386/avx512f-setzero-1.c: New test.
From-SVN: r260310
In the testcase in this patch we create an SLP vector with only two
elements. Our current vector initialisation code will first duplicate
the first element to both lanes, then overwrite the top lane with a new
value.
This duplication can be clunky and wasteful.
Better would be to simply use the fact that we will always be
overwriting the remaining bits, and simply move the first element to the corrcet
place (implicitly zeroing all other bits).
This reduces the code generation for this case, and can allow more
efficient addressing modes, and other second order benefits for AArch64
code which has been vectorized to V2DI mode.
Note that the change is generic enough to catch the case for any vector
mode, but is expected to be most useful for 2x64-bit vectorization.
Unfortunately, on its own, this would cause failures in
gcc.target/aarch64/load_v2vec_lanes_1.c and
gcc.target/aarch64/store_v2vec_lanes.c , which expect to see many more
vec_merge and vec_duplicate for their simplifications to apply. To fix
this,
add a special case to the AArch64 code if we are loading from two memory
addresses, and use the load_pair_lanes patterns directly.
We also need a new pattern in simplify-rtx.c:simplify_ternary_operation
to catch:
(vec_merge:OUTER
(vec_duplicate:OUTER x:INNER)
(subreg:OUTER y:INNER 0)
(const_int N))
And simplify it to:
(vec_concat:OUTER x:INNER y:INNER) or (vec_concat y x)
This is similar to the existing patterns which are tested in this
function, without requiring the second operand to also be a vec_duplicate.
* config/aarch64/aarch64.c (aarch64_expand_vector_init): Modify
code generation for cases where splatting a value is not useful.
* simplify-rtx.c (simplify_ternary_operation): Simplify
vec_merge across a vec_duplicate and a paradoxical subreg forming a vector
mode to a vec_concat.
* gcc.target/aarch64/vect-slp-dup.c: New.
Co-Authored-By: Kyrylo Tkachov <kyrylo.tkachov@arm.com>
From-SVN: r260309
2018-05-17 Richard Biener <rguenther@suse.de>
PR tree-optimization/85757
* tree-ssa-dse.c (dse_classify_store): Record a PHI def and
remove defs that only feed that PHI from further processing.
* gcc.dg/tree-ssa/ssa-dse-34.c: New testcase.
From-SVN: r260306
