re PR target/68620 (ICE on gcc.target/arm/attr-neon-fp16.c)

2016-01-26  Christophe Lyon  <christophe.lyon@linaro.org>

gcc/
PR target/68620
* config/arm/arm.c (neon_valid_immediate): Handle FP16 vectors.
* config/arm/arm_neon.h (__ARM_NUM_LANES, __arm_lane, arm_lanq):
New helper macros.
(vget_lane_f16): Handle big-endian.
(vgetq_lane_f16): Likewise.
(vset_lane_f16): Likewise.
(vsetq_lane_f16): Likewise.
* config/arm/iterators.md (VQXMOV): Add V8HF.
(VDQ): Add V4HF and V8HF.
(V_reg): Handle V4HF and V8HF.
(Is_float_mode): Likewise.
* config/arm/neon.md (movv4hf, movv8hf, neon_vdup_nv4hf,
neon_vdup_nv8hf): New patterns.
(vec_set<mode>_internal, vec_extract<mode>, neon_vld1_dup<mode>):
Use VD_LANE iterator.
(neon_vld1_dup<mode>): Use VQ2 iterator.

testsuite/
PR target/68620
* gcc.target/arm/pr68620.c: New test.

From-SVN: r232832

Remove -m32 from gcc.target/i386/pr68986-2.c

* gcc.target/i386/pr68986-2.c: Remove -m32.

From-SVN: r232831

omp-low.h (oacc_fn_attrib_kernels_p): Declare.

* omp-low.h (oacc_fn_attrib_kernels_p): Declare.
(set_oacc_fn_attrib): Add IS_KERNEL arg.
* omp-low.c (set_oacc_fn_attrib): Add IS_KERNEL arg.
(oacc_fn_attrib_kernels_p, oacc_fn_attrib_level): New.
(expand_omp_target): Pass is_kernel to set_oacc_fn_attrib.
(oacc_validate_dims): Add LEVEL arg, don't return level.
(new_oacc_loop_routine): Use oacc_fn_attrib_level, not
oacc_validate_dims.
(execute_oacc_device_lower): Adjust, add more dump output.
* tree-ssa-loop.c (gate_oacc_kernels): Use
oacc_fn_attrib_kernels_p.
* tree-parloops.c (create_parallel_loop): Adjust
set_oacc_fn_attrib call.

From-SVN: r232830

re PR lto/69254 (ICE in streamer_get_builtin_tree when using -fsanitize=shift on the compile side only)

PR lto/69254
* lto-wrapper.c (merge_and_complain): Handle -fcilkplus.
(append_compiler_options): Handle -fcilkplus.
(append_linker_options): Ignore -fno-{openmp,openacc,cilkplus}.

From-SVN: r232829

re PR target/66655 (miscompilation due to ipa-ra on MinGW)

PR target/66655
* config/i386/winnt.c (i386_pe_binds_local_p): If a function has
been marked as DECL_ONE_ONLY but we do not the means to make it
so, then do not allow it to bind locally.

From-SVN: r232828

re PR lto/69254 (ICE in streamer_get_builtin_tree when using -fsanitize=shift on the compile side only)

PR lto/69254
* opts.h (parse_sanitizer_options): New prototype.
* opts.c (sanitizer_opts): New array.
(parse_sanitizer_options): New function.
(common_handle_option): Use parse_sanitizer_options.

From-SVN: r232826

Update stack alignment in ix86_update_stack_boundary

Stack alignment adjustment for __tls_get_addr should be done in
ix86_update_stack_boundary, not ix86_compute_frame_layout.  Also
there is no need to over-align stack for __tls_get_addr and function
with __tls_get_addr call isn't a leaf function.

gcc/

PR target/68986
* config/i386/i386.c (ix86_compute_frame_layout): Move stack
alignment adjustment to ...
(ix86_update_stack_boundary): Here.  Don't over-align stack for
__tls_get_addr.
(ix86_finalize_stack_realign_flags): Use stack_alignment_needed
if __tls_get_addr is called.

gcc/testsuite/

PR target/68986
* gcc.target/i386/pr68986-1.c: New test.
* gcc.target/i386/pr68986-2.c: Likewise.
* gcc.target/i386/pr68986-3.c: Likewise.

From-SVN: r232825

remove arm_crypto_pragma test selector

From-SVN: r232824

* config/ia64/ia64.c (ia64_expand_vecint_compare): Use gen_int_mode.

From-SVN: r232822

re PR tree-optimization/69467 (Pattern X * C1 CMP 0 to X CMP 0 causes performance drop on 32-bit x86.)

2016-01-26  Richard Biener  <rguenther@suse.de>

PR middle-end/69467
* match.pd: Guard X * CST CMP 0 pattern with single_use.

From-SVN: r232821

re PR tree-optimization/69452 (gcc ICE at -O3 on x86_64-linux-gnu in with verify_ssa failed)

2016-01-26  Richard Biener  <rguenther@suse.de>

PR tree-optimization/69452
* tree-ssa-loop-im.c (move_computations_dom_walker): Remove.
(move_computations_dom_walker::before_dom_children): Rename
to ...
(move_computations_worker): This.
(move_computations): Perform an RPO rather than a DOM walk.

* gcc.dg/torture/pr69452.c: New testcase.

From-SVN: r232820

re PR target/69442 (wrong code with -Og and 64bit modulo @ armv7a)

PR target/69442
* combine.c (combine_instructions): For REG_EQUAL note with
SET_DEST being ZERO_EXTRACT, also temporarily set SET_DEST
to the underlying register.
* doc/rtl.texi (REG_EQUAL): Document the behavior of
REG_EQUAL/REG_EQUIV notes if SET_DEST is ZERO_EXTRACT.

* gcc.dg/pr69442.c: New test.

From-SVN: r232819

[PATCH] Do not set structural equality on polynomial types

gcc/ChangeLog:

PR target/67896
* config/aarch64/aarch64-builtins.c
(aarch64_init_simd_builtin_types): Do not set structural
equality to __Poly{8,16,64,128}_t types.

gcc/testsuite/ChangeLog:

PR target/67896
* gcc.target/aarch64/simd/pr67896.C: New.

From-SVN: r232818

PR 69400: Invalid 128-bit modulus result

As described in the PR, wi::divmod_internal was sign- rather than
zero-extending a modulus result in cases where the result has fewer
HWIs than the precision and the upper bit of the upper HWI was set.

This patch tries to make things more robust by getting wi_pack
to handle the canonicalisation step itself.

Tested on x86_64-linux-gnu.  I added tests to the wide-int
plugin since that seemed more direct.

gcc/
PR tree-optimization/69400
* wide-int.cc (wi_pack): Take the precision as argument and
perform canonicalization here rather than in the callers.
Use the main loop to handle all full-width HWIs.  Add a
zero HWI if in_len isn't a full result.
(wi::divmod_internal): Update accordingly.
(wi::mul_internal): Likewise.  Simplify.

gcc/testsuite/
PR tree-optimization/69400
* gcc.dg/plugin/wide-int_plugin.c (test_wide_int_mod_trunc): New
function.
(plugin_init): Call it.
* gcc.dg/torture/pr69400.c: New test.

From-SVN: r232817

pr68674.c: Check and use arm_fp effective target.

2016-01-26  Christophe Lyon  <christophe.lyon@linaro.org>

* gcc.target/arm/pr68674.c: Check and use arm_fp effective target.

From-SVN: r232816

Add missing includes in JIT frontend.

gcc/jit/
* jit-playback.c: Include pthread.h.

From-SVN: r232815

new scop schedule for isl-0.15

Keep unchanged the implementation for isl-0.14.

* graphite-poly.c (apply_poly_transforms): Simplify.
(print_isl_set): Use more readable format: ISL_YAML_STYLE_BLOCK.
(print_isl_map): Same.
(print_isl_union_map): Same.
(print_isl_schedule): New.
(debug_isl_schedule): New.
* graphite-dependences.c (scop_get_reads): Do not call
isl_union_map_add_map that is undocumented isl functionality.
(scop_get_must_writes): Same.
(scop_get_may_writes): Same.
(scop_get_original_schedule): Remove.
(scop_get_dependences): Do not call isl_union_map_compute_flow that
is deprecated in isl 0.15.  Instead, use isl_union_access_* interface.
(compute_deps): Remove.
* graphite-isl-ast-to-gimple.c (print_schedule_ast): New.
(debug_schedule_ast): New.
(translate_isl_ast_to_gimple::scop_to_isl_ast): Call set_separate_option.
(graphite_regenerate_ast_isl): Add dump.
(translate_isl_ast_to_gimple::scop_to_isl_ast): Generate code
from scop->transformed_schedule.
(graphite_regenerate_ast_isl): Add more dump.
* graphite-optimize-isl.c (optimize_isl): Set
scop->transformed_schedule.  Check whether schedules are equal.
(apply_poly_transforms): Move here.
* graphite-poly.c (apply_poly_transforms): ... from here.
(free_poly_bb): Static.
(free_scop): Static.
(pbb_number_of_iterations_at_time): Remove.
(print_isl_ast): New.
(debug_isl_ast): New.
(debug_scop_pbb): New.
* graphite-scop-detection.c (print_edge): Move.
(print_sese): Move.
* graphite-sese-to-poly.c (build_pbb_scattering_polyhedrons): Remove.
(build_scop_scattering): Remove.
(create_pw_aff_from_tree): Assert instead of bailing out.
(add_condition_to_pbb): Remove unused code, do not fail.
(add_conditions_to_domain): Same.
(add_conditions_to_constraints): Remove.
(build_scop_context): New.
(add_iter_domain_dimension): New.
(build_iteration_domains): Initialize pbb->iterators.
Call add_conditions_to_domain.
(nested_in): New.
(loop_at): New.
(index_outermost_in_loop): New.
(index_pbb_in_loop): New.
(outermost_pbb_in): New.
(add_in_sequence): New.
(add_outer_projection): New.
(outer_projection_mupa): New.
(add_loop_schedule): New.
(build_schedule_pbb): New.
(build_schedule_loop): New.
(embed_in_surrounding_loops): New.
(build_schedule_loop_nest): New.
(build_original_schedule): New.
(build_poly_scop): Call build_original_schedule.
* graphite.h: Declare print_isl_schedule and debug_isl_schedule.
(free_poly_dr): Remove.
(struct poly_bb): Add iterators.  Remove schedule, transformed, saved.
(free_poly_bb): Remove.
(debug_loop_vec): Remove.
(print_isl_ast): Declare.
(debug_isl_ast): Declare.
(scop_do_interchange): Remove.
(scop_do_strip_mine): Remove.
(scop_do_block): Remove.
(flatten_all_loops): Remove.
(optimize_isl): Remove.
(pbb_number_of_iterations_at_time): Remove.
(debug_scop_pbb): Declare.
(print_schedule_ast): Declare.
(debug_schedule_ast): Declare.
(struct scop): Remove schedule.  Add original_schedule,
transformed_schedule.
(free_gimple_poly_bb): Remove.
(print_generated_program): Remove.
(debug_generated_program): Remove.
(unify_scattering_dimensions): Remove.
* sese.c (print_edge): ... here.
(print_sese): ... here.
(debug_edge): ... here.
(debug_sese): ... here.
* sese.h (print_edge): Declare.
(print_sese): Declare.
(dump_edge): Declare.
(dump_sese): Declare.

Co-Authored-By: Sebastian Pop <s.pop@samsung.com>
From-SVN: r232812

fix PR68343: disable fuse-*.c tests for isl 0.14 or earlier

The patch disables all fuse-*.c tests when configuring gcc with isl 0.14 or earlier.

ChangeLog:

* Makefile.in: Regenerate.
* Makefile.tpl: Export ISLVER.
* configure: Regenerate.
* config/isl.m4: Detect isl-0.15.

gcc/

* Makefile.in: Set ISLVER in site.exp.

gcc/testsuite/

* gcc.dg/graphite/graphite.exp: Only run the fuse-*.c tests with isl-0.15.

Co-Authored-By: Sebastian Pop <s.pop@samsung.com>
From-SVN: r232811

add missing testcase

Co-Authored-By: Sebastian Pop <s.pop@samsung.com>
From-SVN: r232810

Daily bump.

From-SVN: r232809

re PR target/69444 (recent changes broke x64_64 -> powerpc crosscompiler build)

PR target/69444
* config/rs6000/sfp-machine.h: Fix a typo in #ifndef - __NO_FPRS__
instead of ___NO_FPRS__.

From-SVN: r232805

omp-low.c (lower_omp_target): Set DECL_VALUE_EXPR of new_var even for the non-array case.

* omp-low.c (lower_omp_target) <case USE_DEVICE_PTR>: Set
DECL_VALUE_EXPR of new_var even for the non-array case.  Look
through DECL_VALUE_EXPR for expansion.

* c-c++-common/goacc/use_device-1.c: New test.

From-SVN: r232804

mips.c (mips_compute_frame_info): Skip re-computing the frame info after reload completed.

        * config/mips/mips.c (mips_compute_frame_info): Skip re-computing
        the frame info after reload completed.

From-SVN: r232803

re PR tree-optimization/69196 (code size regression with jump threading at -O2)

PR tree-optimization/69196
PR tree-optimization/68398
* tree-ssa-threadupdate.h (enum bb_dom_status): Moved here from
tree-ssa-threadupdate.c.
(determine_bb_domination_status): Prototype
* tree-ssa-threadupdate.c (enum bb_dom_status): Remove
(determine_bb_domination_status): No longer static.
(valid_jump_thread_path): Remove code to detect characteristics
of the jump thread path not associated with correctness.
* tree-ssa-threadbackward.c (fsm_find_control_statment_thread_paths):
Correct test for thread path length.  Count PHIs for real operands as
statements that need to be copied.  Do not count ASSERT_EXPRs.
Look at all the blocks in the thread path.  Compute and selectively
filter thread paths based on threading through the latch, threading
a multiway branch or crossing a multiway branch.

PR tree-optimization/69196
PR tree-optimization/68398
* gcc.dg/tree-ssa/pr66752-3.c: Update expected output
* gcc.dg/tree-ssa/pr68198.c: Likewise.

From-SVN: r232802

Avoid including all of <random> in <algorithm>

PR libstdc++/69464
* include/Makefile.am: Add new header.
* include/Makefile.in: Regenerate.
* include/bits/random.h (uniform_int_distribution): Move to
bits/uniform_int_dist.h.
* include/bits/random.tcc (uniform_int_distribution::operator(),
uniform_int_distribution::__generate_impl): Likewise.
* include/bits/uniform_int_dist.h: New header.
* include/bits/stl_algo.h [__cplusplus >= 201103L]: Include
<bits/uniform_int_dist.h> instead of <random>.
* testsuite/20_util/specialized_algorithms/uninitialized_copy/
move_iterators/1.cc: Include correct header for uninitialized_copy.
* testsuite/20_util/specialized_algorithms/uninitialized_copy_n/
move_iterators/1.cc: Likewise.
* testsuite/25_algorithms/nth_element/58800.cc: Include correct
header for vector.
* testsuite/26_numerics/random/pr60037-neg.cc: Adjust dg-error lines.

From-SVN: r232798

re PR target/69469 (test case gcc.target/powerpc/vsx-vector-2.c fails on power starting with r232632)

        PR target/69469
        * gcc.target/powerpc/vsx-vector-2.c: Adjust for Goldschmidt's
        Algorithm.

From-SVN: r232796

re PR testsuite/69380 (FAIL: g++.dg/tree-ssa/pr69336.C   scan-tree-dump-not optimized "cmap")

2016-01-25  Richard Biener  <rguenther@suse.de>

PR testsuite/69380
* g++.dg/tree-ssa/pr69336.C: Restrict to x86_64 and i?86.

From-SVN: r232795

rs6000.c (rs6000_keep_leaf_when_profiled): Add decl with __attribute__ ((unused)) annotation.

2016-01-25  Bill Schmidt  <wschmidt@linux.vnet.ibm.com>

* config/rs6000/rs6000.c (rs6000_keep_leaf_when_profiled):  Add
decl with __attribute__ ((unused)) annotation.

From-SVN: r232793

re PR target/69421 (ICE in maybe_legitimize_operand, at optabs.c:6888 with -O3)

gcc/

PR target/69421
* tree-vect-stmts.c (vectorizable_condition): Check vectype
of operands is compatible with a statement vectype.

gcc/testsuite/

PR target/69421
* gcc.dg/pr69421.c: New test.

From-SVN: r232792

extend.texi (scalar_storage_order type attribute): Fix typo and improve wording for mixed storage order support.

* doc/extend.texi (scalar_storage_order type attribute): Fix typo and
improve wording for mixed storage order support.

From-SVN: r232791

[AARCH64][ACLE][NEON] Implement vcvt*_s64_f64 and vcvt*_u64_f64 NEON intrinsics.

gcc/

* config/aarch64/arm_neon.h (vcvt_s64_f64): New intrinsic.
(vcvt_u64_f64): Likewise.
(vcvta_s64_f64): Likewise.
(vcvta_u64_f64): Likewise.
(vcvtm_s64_f64): Likewise.
(vcvtm_u64_f64): Likewise.
(vcvtn_s64_f64): Likewise.
(vcvtn_u64_f64): Likewise.
(vcvtp_s64_f64): Likewise.
(vcvtp_u64_f64): Likewise.

gcc/testsuite/

* gcc.target/aarch64/simd/vcvt_s64_f64_1.c: New.
* gcc.target/aarch64/simd/vcvt_u64_f64_1.c: Likewise.
* gcc.target/aarch64/simd/vcvta_s64_f64_1.c: Likewise.
* gcc.target/aarch64/simd/vcvta_u64_f64_1.c: Likewise.
* gcc.target/aarch64/simd/vcvtm_s64_f64_1.c: Likewise.
* gcc.target/aarch64/simd/vcvtm_u64_f64_1.c: Likewise.
* gcc.target/aarch64/simd/vcvtn_s64_f64_1.c: Likewise.
* gcc.target/aarch64/simd/vcvtn_u64_f64_1.c: Likewise.
* gcc.target/aarch64/simd/vcvtp_s64_f64_1.c: Likewise.
* gcc.target/aarch64/simd/vcvtp_u64_f64_1.c: Likewise.

From-SVN: r232789

[ARC] Add basic support for double load and store instructions

gcc/
2016-01-25  Claudiu Zissulescu  <claziss@synopsys.com>

* config/arc/arc.c (TARGET_DWARF_REGISTER_SPAN): Define.
(arc_init): Check validity mll64 option.
(arc_save_restore): Use double load/store instruction.
(arc_expand_movmem): Likewise.
(arc_split_move): Don't split if we have double load/store
instructions. Returns a boolean.
(arc_process_double_reg_moves): Change function to return boolean
instead of a sequence of instructions.
(arc_dwarf_register_span): New function.
* config/arc/arc-protos.h (arc_split_move): Change prototype.
* config/arc/arc.h (TARGET_CPU_CPP_BUILTINS): Define __ARC_LL64__.
* config/arc/arc.md (*movdi_insn): Emit ldd/std instructions.
(*movdf_insn): Likewise.
* config/arc/arc.opt (mll64): New option.
* config/arc/predicates.md (even_register_operand): New predicate.
* doc/invoke.texi (ARC Options): Add mll64 documentation.

From-SVN: r232788

re PR lto/69393 (ICE in dwarf2out_finish, at dwarf2out.c:27175 with LTO)

2016-01-25  Richard Biener  <rguenther@suse.de>

PR lto/69393
* dwarf2out.c (is_naming_typedef_decl): Not when DECL_NAMELESS.
* tree-streamer-out.c (pack_ts_base_value_fields): Stream
DECL_NAMELESS.
* tree-streamer-in.c (unpack_ts_base_value_fields): Likewise.

* testsuite/libgomp.c++/pr69393.C: New testcase.

From-SVN: r232787

re PR tree-optimization/69376 (wrong code at -Os and above on x86_64-linux-gnu)

2016-01-25  Richard Biener  <rguenther@suse.de>

PR tree-optimization/69376
* tree-ssa-sccvn.h (struct vn_ssa_aux): Add range_info_anti_range_p
flag.
(VN_INFO_ANTI_RANGE_P): New inline.
(VN_INFO_RANGE_TYPE): Likewise.
* tree-ssa-sccvn.c (set_ssa_val_to): Also record and copy
SSA_NAME_ANTI_RANGE_P.
(free_scc_vn): Restore SSA_NAME_ANTI_RANGE_P.
* tree-ssa-pre.c (eliminate_dom_walker::before_dom_children):
Properly query VN_INFO_RANGE_TYPE.

* gcc.dg/torture/pr69376.c: New testcase.

From-SVN: r232786

re PR target/66655 (miscompilation due to ipa-ra on MinGW)

PR target/66655
* config/i386/cygming.h (MAKE_DECL_ONE_ONLY): Remove definition.

From-SVN: r232785

Daily bump.

From-SVN: r232783

[multiple changes]

2016-01-23  Jerry DeLisle  <jvdelisle@gcc.gnu.org>

PR fortran/69397
PR fortran/68442
* interface.c (gfc_arglist_matches_symbol): Replace assert with
a return false if not a procedure.
* resolve.c (resolve_generic_f): Test if we are resolving an
initialization expression and adjust error message accordingly.

2016-01-24  Jerry DeLisle  <jvdelisle@gcc.gnu.org>

PR fortran/69397
PR fortran/68442
* gfortran.dg/interface_38.f90: New test.
* gfortran.dg/interface_39.f90: New test.

From-SVN: r232780

re PR fortran/66094 (Handle transpose(A) in inline matmul)

2016-01-24  Thomas Koenig  <tkoenig@gcc.gnu.org>

PR fortran/66094
* frontend-passes.c (matmul_lhs_realloc):  Add
forgotten break statement.

From-SVN: r232779

Revert "Fix the remaining PR c++/24666 blockers"

gcc/cp/ChangeLog:

Revert:
2016-01-18  Patrick Palka  <ppalka@gcc.gnu.org>

PR c++/11858
PR c++/24663
PR c++/24664
* decl.c (grokdeclarator): Don't decay array parameter type to
a pointer type if it's dependent.
(grokparms): Invoke strip_top_quals instead of directly invoking
cp_build_qualified_type.
* pt.c (decay_dependent_array_parm_type): New static function.
(type_unification_real): Call decay_dependent_array_parm_type
to decay a dependent array parameter type to its corresponding
pointer type before unification.
(more_specialized_fn): Likewise.
(get_bindings): Likewise.
* tree.c (cp_build_qualified_type): Trivial typofix in
documentation.

gcc/testsuite/ChangeLog:

Revert:
2016-01-18  Patrick Palka  <ppalka@gcc.gnu.org>

PR c++/11858
PR c++/24663
PR c++/24664
* g++.dg/template/pr11858.C: New test.
* g++.dg/template/pr24663.C: New test.
* g++.dg/template/unify12.C: New test.
* g++.dg/template/unify13.C: New test.
* g++.dg/template/unify14.C: New test.
* g++.dg/template/unify15.C: New test.
* g++.dg/template/unify16.C: New test.
* g++.dg/template/unify17.C: New test.

From-SVN: r232778

Pass LSAN_OPTIONS to Makefiles in subdirectories

PR bootstrap/69329
* Makefile.tpl (BASE_FLAGS_TO_PASS): Add LSAN_OPTIONS.
* Makefile.in: Regenerate.

From-SVN: r232777

re PR fortran/68283 (ice: gfc_variable_attr(): Bad array reference)

2016-01-24  Dominique d'Humieres  <dominiq@lps.ens.fr>

PR fortran/68283
gfortran.dg/pr68283.f90: New test.

From-SVN: r232776

[multiple changes]

2016-01-24  Dominique d'Humieres <dominiq@lps.ens.fr>

PR fortran/68283
* primary.c (gfc_variable_attr): revert revision r221955,
call gfc_internal_error only if there is no error.

2016-01-24  Dominique d'Humieres  <dominiq@lps.ens.fr>

PR fortran/68283
gfortran.dg/pr68283.f90: New test.

From-SVN: r232775

re PR fortran/66094 (Handle transpose(A) in inline matmul)

2016-01-24  Thomas Koenig  <tkoenig@gcc.gnu.org>

PR fortran/66094
* frontend-passes.c (enum matrix_case):  Add case A2B2T for
MATMUL(A,TRANSPoSE(B)) where A and B are rank 2.
(inline_limit_check):  Also add A2B2T.
(matmul_lhs_realloc):  Handle A2B2T.
(check_conjg_variable):  Rename to
(check_conjg_transpose_variable):  and also count TRANSPOSE.
(inline_matmul_assign):  Handle A2B2T.

2016-01-24  Thomas Koenig  <tkoenig@gcc.gnu.org>

PR fortran/66094
* gfortran.dg/inline_matmul_13.f90:  New test.
* gfortran.dg/matmul_bounds_8.f90:  New test.
* gfortran.dg/matmul_bounds_9.f90:  New test.
* gfortran.dg/matmul_bounds_10.f90:  New test.

From-SVN: r232774

Daily bump.

From-SVN: r232773

re PR libfortran/68744 (FAIL: gfortran.dg/backtrace_1.f90   -O0  execution test)

PR libfortran/68744
* runtime/backtrace.c: Include gthr.h.
(show_backtrace): Use __gthread_active_p() to determine whether threads
are active.  Return if lbstate is NULL.

From-SVN: r232770

re PR libstdc++/69446 (cow-stdexcept.cc:374:1: error: alias d efinitions not supported in this configuration)

PR libstdc++/69446
* config/os/hpux/os_defines.h (_GLIBCXX_USE_WEAK_REF): Define.

From-SVN: r232769

Fix clobber removal in parloops

2016-01-23  Tom de Vries  <tom@codesourcery.com>

PR tree-optimization/69426
* tree-parloops.c (eliminate_local_variables_stmt): Handle vdef of
removed clobber.

* gcc.dg/autopar/pr69426.c: New test.

From-SVN: r232767

PR c++/58109 - alignas() fails to compile with constant expression

PR c++/58109 - alignas() fails to compile with constant expression
PR c++/69022 - attribute vector_size ignored with dependent bytes

gcc/testsuite/ChangeLog:
2016-01-23  Martin Sebor  <msebor@redhat.com>

PR c++/58109
PR c++/69022
* g++.dg/cpp0x/alignas5.C: New test.
* g++.dg/ext/vector29.C: Same.

gcc/cp/ChangeLog:
2016-01-23  Martin Sebor  <msebor@redhat.com>

PR c++/58109
PR c++/69022
* decl2.c (is_late_template_attribute): Handle dependent argument
  to attribute align and attribute vector_size.

From-SVN: r232766

tree-ssanames.c (release_free_names_and_compact_live_names): Replace "the the" with "the" in the comments.

* tree-ssanames.c (release_free_names_and_compact_live_names): Replace
"the the" with "the" in the comments.
* ipa-devirt.c (build_type_inheritance_graph,
update_type_inheritance_graph): Likewise.
* tree.c (build_function_type_list_1): Likewise.
* cfgloopmanip.c (scale_loop_profile): Likewise.
* tree-ssa-loop-ivopts.c (get_shiftadd_cost): Likewise.
* gimple-ssa-split-paths.c
(find_block_to_duplicate_for_splitting_paths): Likewise.
* tree-sra.c (init_subtree_with_zero, clobber_subtree): Likewise.
* expr.c (convert_move): Likewise.
* var-tracking.c (vt_stack_adjustments): Likewise.
* tree-vect-data-refs.c (vect_enhance_data_refs_alignment): Likewise.
* tree-vrp.c (test_for_singularity): Likewise.

From-SVN: r232765

tree-vect-stmts.c (vectorizable_condition): Build a VEC_COND_EXPR directly instead of building a temporary tree.

* tree-vect-stmts.c (vectorizable_condition): Build a VEC_COND_EXPR
directly instead of building a temporary tree.

From-SVN: r232764

re PR bootstrap/69434 (bootstrap failed with --with-arch=corei7 --with-cpu=corei7)

PR bootstrap/69434
* genrecog.c: Define INCLUDE_ALGORITHM before including system.h,
remove <algorithm> include.

From-SVN: r232763

chkp-strlen-2.c: Define _GNU_SOURCE.

* gcc.target/i386/chkp-strlen-2.c: Define _GNU_SOURCE.
* gcc.target/i386/chkp-strlen-4.c: Move the definition of
_GNU_SOURCE from compile flags to the source.
* gcc.target/i386/chkp-stropt-12.c: Ditto.
* gcc.target/i386/chkp-stropt-16.c: Ditto.
* gcc.target/i386/chkp-stropt-4.c: Ditto.
* gcc.target/i386/chkp-stropt-8.c: Ditto.

From-SVN: r232762

Daily bump.

From-SVN: r232760

TR29124 C++ Special Math - <math.h> pulls funcs into global namespace.

2016-01-22  Edward Smith-Rowland  <3dw4rd@verizon.net>

TR29124 C++ Special Math - <math.h> pulls funcs into global namespace.
* include/c_compatibility/math.h: Import the TR29124 functions
into the global namespace.
* testsuite/special_functions/01_assoc_laguerre/compile_2.cc: Remove
xfail and make compile-only.
* testsuite/special_functions/02_assoc_legendre/compile_2.cc: Ditto.
* testsuite/special_functions/03_beta/compile_2.cc: Ditto.
* testsuite/special_functions/04_comp_ellint_1/compile_2.cc: Ditto.
* testsuite/special_functions/05_comp_ellint_2/compile_2.cc: Ditto.
* testsuite/special_functions/06_comp_ellint_3/compile_2.cc: Ditto.
* testsuite/special_functions/07_cyl_bessel_i/compile_2.cc: Ditto.
* testsuite/special_functions/08_cyl_bessel_j/compile_2.cc: Ditto.
* testsuite/special_functions/09_cyl_bessel_k/compile_2.cc: Ditto.
* testsuite/special_functions/10_cyl_neumann/compile_2.cc: Ditto.
* testsuite/special_functions/11_ellint_1/compile_2.cc: Ditto.
* testsuite/special_functions/12_ellint_2/compile_2.cc: Ditto.
* testsuite/special_functions/13_ellint_3/compile_2.cc: Ditto.
* testsuite/special_functions/14_expint/compile_2.cc: Ditto.
* testsuite/special_functions/15_hermite/compile_2.cc: Ditto.
* testsuite/special_functions/16_laguerre/compile_2.cc: Ditto.
* testsuite/special_functions/17_legendre/compile_2.cc: Ditto.
* testsuite/special_functions/18_riemann_zeta/compile_2.cc: Ditto.
* testsuite/special_functions/19_sph_bessel/compile_2.cc: Ditto.
* testsuite/special_functions/20_sph_legendre/compile_2.cc: Ditto.
* testsuite/special_functions/21_sph_neumann/compile_2.cc: Ditto.

From-SVN: r232755

re PR other/69432 (ICE in connect_traces, at dwarf2cfi.c with -O3 -m32 -minline-stringops-dynamically)

PR target/69432
* config/i386/i386.c: Include dojump.h.
(expand_small_movmem_or_setmem,
expand_set_or_movmem_prologue_epilogue_by_misaligned_moves): Spelling
fixes.
(ix86_expand_set_or_movmem): Call do_pending_stack_adjust () early
if dynamic_check != -1.

* g++.dg/opt/pr69432.C: New test.

From-SVN: r232754

Fix ChangeLog entry in wrong file

From-SVN: r232750

Constrain std::valarray functions and operators

PR libstdc++/69116
* include/bits/valarray_before.h (__fun, __fun_with_valarray): Only
define result_type for types which can be safely used with valarrays.
* testsuite/26_numerics/valarray/69116.cc: New.

From-SVN: r232748

re PR c++/69392 (G++ can't capture 'this' pointer to templated type using init-capture)

PR c++/69392
* lambda.c (lambda_capture_field_type): Handle 'this' specially
for init-capture, too.

From-SVN: r232746

re PR tree-optimization/69347 (excessive compile time with -O2)

PR middle-end/69347
* tree-ssa-dom.c (back_propagate_equivalences): Factored out of
record_temporary_equivalences.  Rewritten to avoid unnecessary calls
into dominated_by_p.
(cprop_into_successor_phis): Avoid unnecessary tests.

From-SVN: r232745

pr63354c: Adjust target selector to include 32-bit targets testing an -m64 multilib.

2016-01-22  Bill Schmidt  <wschmidt@linux.vnet.ibm.com>

* gcc.target/powerpc/pr63354c: Adjust target selector to include
32-bit targets testing an -m64 multilib.

From-SVN: r232744

- fix a ChangeLog entry

From-SVN: r232743

re PR target/69416 (Nonsense rtl checking failure)

PR target/69416

  * config/aarch64/aarch64.md (UNSPEC_NZCV): New.
  (ccmp<mode>, fccmp<mode>, fccmpe<mode>): Use it.

From-SVN: r232737

system.h (string, algorithm): Include only conditionally.

* system.h (string, algorithm): Include only conditionally.
(new): Include always under C++.
* bb-reorder.c (toplevel): Define INCLUDE_ALGORITHM.
* final.c (toplevel): Ditto.
* ipa-chkp.c (toplevel): Define INCLUDE_STRING.
* genconditions.c (write_header): Make gencondmd.c define
INCLUDE_STRING.
* mem-stats.h (mem_usage::print_dash_line): Don't use std::string.

* config/aarch64/aarch64.c (toplevel): Define INCLUDE_STRING.
* common/config/aarch64/aarch64-common.c (toplevel): Ditto.

From-SVN: r232736

libitm: Fix HTM fastpath.

* beginend.cc (GTM::gtm_thread::serial_lock): Put on cacheline
boundary.
(htm_fastpath): Remove.
(gtm_thread::begin_transaction): Fix HTM fastpath.
(_ITM_commitTransaction): Adapt.
(_ITM_commitTransactionEH): Adapt.
* libitm/config/linux/rwlock.h (gtm_rwlock): Add htm_fastpath member
and accessors.
* libitm/config/posix/rwlock.h (gtm_rwlock): Likewise.
* libitm/config/posix/rwlock.cc (gtm_rwlock::gtm_rwlock): Adapt.
* libitm/config/x86/sjlj.S (_ITM_beginTransaction): Fix HTM fastpath.
* libitm/libitm_i.h (htm_fastpath): Remove declaration.
* libitm/method-serial.cc (htm_mg): Adapt.
(gtm_thread::serialirr_mode): Adapt.
* libitm/query.cc (_ITM_inTransaction, _ITM_getTransactionId): Adapt.

From-SVN: r232735

pr63354.c: Restrict to Linux targets with 64-bit support.

2016-01-22  Bill Schmidt  <wschmidt@linux.vnet.ibm.com>

* gcc.target/powerpc/pr63354.c: Restrict to Linux targets with
64-bit support.

From-SVN: r232732

[Patch Obvious] gcc.dg/vect/bb-slp-pr68892.c requires vectorization of doubles

gcc/testsuite/

* gcc.dg/vect/bb-slp-pr68892.c: Require vect_double.

From-SVN: r232731

target.c (gomp_get_target_fn_addr): Allow host fallback if target function wasn't mapped to the device with...

libgomp/
* target.c (gomp_get_target_fn_addr): Allow host fallback if target
function wasn't mapped to the device with non-shared memory.

From-SVN: r232729

[multiple changes]

2016-01-22  Christian Bruel  <christian.bruel@st.com>

PR target/68674
* expr.c (expand_expr_real_1): Reset DECL_MODE if VECTOR_TYPE_P changed.

// testsuite
2016-01-21  Christian Bruel  <christian.bruel@st.com>

PR target/68674
* gcc.target/i386/pr68674.c
* gcc.target/aarch64/pr68674.c
* gcc.target/arm/pr68674.c

From-SVN: r232728

[ARM] Fix PR target/69403: Bug in thumb2_ior_scc_strict_it pattern

PR target/69403
* config/arm/thumb2.md (*thumb2_ior_scc_strict_it): Convert to
define_insn_and_split.  Ensure operands[1] and operands[0] do not
get assigned the same register.

* gcc.c-torture/execute/pr69403.c: New test.

From-SVN: r232727

Allow _GLIBCXX_HAVE_OBSOLETE_ISINF_ISNAN to be overridden

PR libstdc++/69413
* config/os/gnu-linux/os_defines.h: Define
_GLIBCXX_NO_OBSOLETE_ISINF_ISNAN_DYNAMIC.
* include/c_global/cmath (isinf, isnan): Check it.
* doc/xml/manual/internals.xml: Document it.
* doc/html/*: Regenerate.

From-SVN: r232726

ipa-prop.c (ipa_set_jf_constant): Remove redundant unshare_expr.

gcc/ChangeLog:

2016-01-22  Kugan Vivekanandarajah  <kuganv@linaro.org>

* ipa-prop.c (ipa_set_jf_constant): Remove redundant unshare_expr.

From-SVN: r232725

missed : in log

From-SVN: r232724

2016-01-22  Christian Bruel  <christian.bruel@st.com>

* config/arm/arm-c.c (arm_pragma_target_parse)
Remove warn_builtin_macro_redefined overwrite.

From-SVN: r232723

ipa-inline.c (can_inline_edge_p): Back out overzealous check on flag_non_call_exceptions compatibility.

* ipa-inline.c (can_inline_edge_p): Back out overzealous check on
flag_non_call_exceptions compatibility.

From-SVN: r232722

re PR c++/55843 (ICE after exceeding template instantiation depth)

2016-01-22  Paolo Carlini  <paolo.carlini@oracle.com>

PR c++/55843
* g++.dg/template/pr55843.C: New.

From-SVN: r232721

[Testsuite] Fix PR66877

* gcc.dg/vect/vect-over-widen-3-big-array.c: Only look for 1
vect_recog_over_widening_pattern in dump if we have vect_widen_shift.

From-SVN: r232720

re PR debug/66668 (FAIL: gcc.dg/debug/dwarf2/stacked-qualified-types-3.c scan-assembler-times DIE \\([^\n]*\\) DW_TAG_(?:const|volatile|atomic|restrict)_type 8)

PR debug/66668
* dwarf2out.c (add_child_die_after): New function.
(dwarf_qual_info_t): New type.
(dwarf_qual_info): New variable.
(qualified_die_p): New function.
(modified_type_die): For -fdebug-types-section, ensure
canonical order of qualifiers.  Put qualified DIEs adjacent
to the corresponding non-qualified type DIE and search there
for existing qualified DIEs.

From-SVN: r232719

extend.texi (scalar_storage_order type attribute): Document restriction on type punning and aliasing, and remove future tense.

* doc/extend.texi (scalar_storage_order type attribute): Document
restriction on type punning and aliasing, and remove future tense.

From-SVN: r232718

re PR testsuite/67489 (FAIL: gcc.target/powerpc/p8vector-builtin-8.c (test for excess errors))

2016-01-21  Bill Schmidt  <wschmidt@linux.vnet.ibm.com>

PR testsuite/67489
* gcc.target/powerpc/p8vector-builtin-8.c: Remove { target int128
} from dg-do compile directive, and instead add {
dg-require-effective-target int128 }.

From-SVN: r232717

Daily bump.

From-SVN: r232716

PR c/69405 - [6 Regression] ICE in c_tree_printer on an invalid

PR c/69405 - [6 Regression] ICE in c_tree_printer on an invalid
    __atomic_fetch_add

gcc/testsuite/ChangeLog:
2016-01-20  Martin Sebor  <msebor@redhat.com>

        PR c/69405
        * gcc.dg/sync-fetch.c: New test.

gcc/c-family/ChangeLog:
2016-01-20  Martin Sebor  <msebor@redhat.com>

        PR c/69405
        * c-common.c (sync_resolve_size): Avoid printing diagnostic about
        an incompatible argument when the argument isn't a valid tree node.

From-SVN: r232713

re PR target/69252 (gcc.dg/vect/vect-iv-9.c FAILs with -Os -fmodulo-sched -fmodulo-sched-allow-regmoves -fsched-pressure)

PR target/69252
* modulo-sched.c (optimize_sc): Allow branch-scheduling to add a new
first stage.

PR target/69252
* gcc.target/powerpc/pr69252.c: New test.

From-SVN: r232712

[PATCH] [PR tree-optimization/69347] Fix memory consumption in threader & minor speed improvement

PR middle-end/69347
* tree-ssa-dom.c (dom_opt_dom_walker::thread_across_edge): Avoid
useless call to record_temporary_equivalences.
* tree-ssa-threadbackward.c (find_jump_threads_backwards): Just
allocate 10 slots in the bb_path vector and let it grow as needed.
(fsm_find_control_statement_thread_paths): Similarly for the next_path
vector.

From-SVN: r232711

re PR fortran/65996 (gfortran ICE with -dH)

2016-01-21  Jerry DeLisle  <jvdelisle@gcc.gnu.org>

PR fortran/65996
* error.c (gfc_error): Save the state of abort_on_error and set
it to false for buffered errors to allow normal processing.
Restore the state before leaving.

2016-01-21  Jerry DeLisle  <jvdelisle@gcc.gnu.org>

PR fortran/65996
gfortran.dg/pr65996.f90: New test.

From-SVN: r232707

re PR c++/68810 (FAIL: g++.dg/cpp0x/constexpr-reinterpret1.C  -- test for errors -- -m32)

PR c++/68810
* g++.dg/cpp0x/constexpr-reinterpret1.C: Fix line number that is
expected to generate an error.

From-SVN: r232705

re PR c++/59281 (attribute((constructor)) accepts enum class as integer constant)

PR c++/59281
* g++.dg/ext/attr-constructor1.C: New.

From-SVN: r232704

re PR c++/65687 (Inconsistent behavior for __attribute__((__deprecated__)) between C and C++.)

PR c++/65687
* decl.c (type_is_deprecated): Don't look into a typedef.

From-SVN: r232703

re PR c++/40751 (G++ never packs typedef'd enums)

PR c++/40751
PR c++/64987
* decl.c (copy_type_enum): Respect TYPE_USER_ALIGN.

From-SVN: r232702

re PR c++/43407 (Specifying visibility attribute of C++0x enum class emits warning)

PR c++/43407
* decl.c (start_enum): Add attributes parameter.
* parser.c (cp_parser_enum_specifier): Pass it.
* pt.c (lookup_template_class_1): Pass it.
* cp-tree.h: Adjust.

From-SVN: r232701

Record configure regenerate

From-SVN: r232698

configure.ac (gcc_cv_as_powerpc_mfcrf, [...]): Detangle.

        * configure.ac (gcc_cv_as_powerpc_mfcrf, gcc_cv_as_machine_directive):
        Detangle.

From-SVN: r232697

Add missing file

From-SVN: r232695

aix71.h (ASM_CPU_SPEC): Add entry for Power9.

* config/rs6000/aix71.h (ASM_CPU_SPEC): Add entry for Power9.
* config/rs6000/driver-rs6000.c (struct asm_names): Likewise.

From-SVN: r232694

libitm: Disable testing transaction-safe exceptions on Darwin and AIX.

* testsuite/libitm.c++/libstdc++-safeexc.C: Not supported on darwin
or AIX.

From-SVN: r232693

Fix missed comment fix in last change.

From-SVN: r232690

Fix PR66178, ICE due to misexpansion of constant expressions involving labels.

PR middle-end/66178
* expr.c (expand_expr_real_2) [PLUS_EXPR, MINUS_EXPR]: Don't
drop EXPAND_INITIALIZER.
* rtl.h (contains_symbolic_reference_p): Declare.
* rtlanal.c (contains_symbolic_reference_p): New function.
* simplify-rtx.c (simplify_binary_operation_1): Don't turn
a subtraction into a NOT if symbolic constants are involved.

testsuite/
PR middle-end/66178
gcc.dg/torture/pr66178.c: New test.

From-SVN: r232689

float128-sed: New files to convert TF names to KF names for PowerPC IEEE 128-bit floating...

2016-01-21  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/float128-sed: New files to convert TF names to KF
names for PowerPC IEEE 128-bit floating point support.
* config/rs6000/float128-sed-hw: Likewise.

* config/rs6000/float128-hw.c: New file for ISA 3.0 IEEE 128-bit
floating point hardware support.

* 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/extendkftf2-sw.c: New file, convert IEEE 128-bit
floating point to IBM extended double.

* config/rs6000/trunctfkf2-sw.c: New file, convert IBM extended
double to 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/sfp-exceptions.c: New file to provide exception
support for IEEE 128-bit floating point.

* config/rs6000/floattikf.c: New files for converting between IEEE
128-bit floating point and signed/unsigned 128-bit integers.
* config/rs6000/fixunskfti.c: Likewise.
* config/rs6000/fixkfti.c: Likewise.
* config/rs6000/floatuntikf.c: Likewise.

* 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.

* libgcc/config.host (powerpc*-*-linux*): If compiler can compile
VSX code, enable IEEE 128-bit floating point.  If the compiler can
compile IEEE 128-bit floating point code with ISA 3.0 IEEE 128-bit
floating point hardware instructions and it supports declaring
functions with the ifunc attribute, enable ifunc functions to
switch between software and hardware support.
* configure.ac (powerpc*-*-linux*): Likewise.
* configure: Regenerate.

Co-Authored-By: Steven Munroe <munroesj@linux.vnet.ibm.com>
Co-Authored-By: Tulio Magno Quites Machado Filho <tulioqm@br.ibm.com>
From-SVN: r232685

re PR target/63354 (gcc -pg -mprofile-kernel creates unused stack frames on leaf functions on ppc64le)

[gcc]

2016-01-21  Anton Blanchard  <anton@samba.org>
    Bill Schmidt  <wschmidt@linux.vnet.ibm.com>

PR target/63354
* config/rs6000/linux64.h (TARGET_KEEP_LEAF_WHEN_PROFILED): New
#define.
* config/rs6000/rs6000.c (rs6000_keep_leaf_when_profiled): New
function.

[gcc/testsuite]

2016-01-21  Anton Blanchard  <anton@samba.org>
    Bill Schmidt  <wschmidt@linux.vnet.ibm.com>

PR target/63354
* gcc.target/powerpc/pr63354.c:  New test.

Co-Authored-By: Bill Schmidt <wschmidt@linux.vnet.ibm.com>
From-SVN: r232684

Issue wic prefetch instruction at call site.
This optimization is enabled with  microblaze target flag mxl-prefetch.

From-SVN: r232683

Mark r20 as fixed.  Mark r21 not fixed & allocatable in callee.

From-SVN: r232682