Symbol tables are insert only.

* cp-tree.h (default_hash_traits <lang_identifier *>): Don't
derive from pointer_hash.  Make undeletable.

From-SVN: r249266

class.c (resort_type_method_vec): Avoid potential unsigned overflow.

* class.c (resort_type_method_vec): Avoid potential unsigned
overflow.

From-SVN: r249265

Don't defer noexcept_deferred_spec.

* cp-tree.h (unevaluated_noexcept_spec): Don't declare.
* decl.c (cxx_init_decl_processing): Initialize
noexcept_deferred_spec.
* except.c (unevaluated_noexcept_spec): Delete.
* class.c (deduce_noexcept_on_destructor): Use
noexcept_deferred_spec directly.
* method.c (implicitly_declare_fn): Likewise.

From-SVN: r249264

Make keyed_classes a vector.

* cp-tree.h (CPTI_KEYED_CLASSES, keyed_classes): Delete.
(keyed_classes): Declare as vector.
* decl.c (keyed_classes): Define.
(cxx_init_decl_processing): Allocate it.
(record_key_method_defined): Use vec_safe_push.
* class.c (finish_struct_1): Likewise.
* pt.c (instantiate_class_template_1): Likewise.
* decl2.c (c_parse_final_cleanups): Reverse iterate keyed_classes.

From-SVN: r249263

Fix keyboard flub.

From-SVN: r249262

re PR libstdc++/81092 (Missing symbols for new std::wstring constructors)

PR libstdc++/81092
* config/abi/post/i486-linux-gnu/baseline_symbols.txt: Update.
* config/abi/post/powerpc-linux-gnu/baseline_symbols.txt: Update.
* config/abi/post/powerpc64-linux-gnu/32/baseline_symbols.txt: Update.

From-SVN: r249260

Make rtti lazier

Make rtti lazier
* rtti.c (enum tinfo_kind): Add TK_DERIVED_TYPES,
TK_VMI_CLASS_TYPES, TK_MAX.  Delete TK_FIXED.
(tinfo_names): New.
(typeid_ok_p): Add quotes to error messages.  Use get_tinfo_desc.
(get_tinfo_decl): Use get_tinfo_desc.
(get_pseudo_ti_init): Likewise. Adjust VMI construction.
(create_pseudo_type_info): Delete.
(get_pseudo_ti_index): Just determine the index.
(get_tinfo_desc): New.  Create all types lazily.
(create_tinfo_types): Just allocate the descriptor array.
(emit_support_tinfos): Use non-inserting type lookup.  Set builtin
location.

From-SVN: r249258

Fix std::wbuffer_convert::sync()

* include/bits/locale_conv.h (wbuffer_convert::sync): Fix condition.
* testsuite/22_locale/conversions/buffer/2.cc: New.

From-SVN: r249255

re PR tree-optimization/81090 ([graphite] ICE in loop_preheader_edge)

2017-06-16  Richard Biener  <rguenther@suse.de>

PR tree-optimization/81090
* passes.def (pass_record_bounds): Remove.
* tree-pass.h (make_pass_record_bounds): Likewise.
* tree-ssa-loop.c (pass_data_record_bounds, pass_record_bounds,
make_pass_record_bounds): Likewise.
* tree-ssa-loop-ivcanon.c (canonicalize_induction_variables): Do
not free niter estimates at the beginning but at the end.
* tree-scalar-evolution.c (scev_finalize): Free niter estimates.

* gcc.dg/graphite/pr81090.c: New testcase.

From-SVN: r249249

PR libstdc++/81092 add std::wstring symbols and bump library version

PR libstdc++/81092
* acinclude.m4: Bump libtool_VERSION.
* config/abi/post/i386-linux-gnu/baseline_symbols.txt: Update.
* config/abi/post/x86_64-linux-gnu/32/baseline_symbols.txt: Update.
* config/abi/pre/gnu.ver: Add wstring constructor symbols to
GLIBCXX_3.4.24 version and move random_device::_M_get_entropy() symbol
to new GLIBCXX_3.4.25 version.
* doc/xml/manual/abi.xml: Document new versions.
* doc/html/*: Regenerate.
* testsuite/21_strings/basic_string/cons/char/8.cc: Use base object
constructors to ensure required symbols are exported.
* testsuite/21_strings/basic_string/cons/wchar_t/8.cc: Likewise.
* testsuite/util/testsuite_abi.cc: Add new version.

From-SVN: r249246

tree-switch-conversion.c (emit_case_bit_tests): Adjust initializer to workaround ICE in host GCC 4.8.

2017-06-16  Richard Biener  <rguenther@suse.de>

* tree-switch-conversion.c (emit_case_bit_tests): Adjust
initializer to workaround ICE in host GCC 4.8.

From-SVN: r249245

ipa-inline-transform.c (update_noncloned_frequencies): Update also counts.

* ipa-inline-transform.c (update_noncloned_frequencies): Update also
counts.
(clone_inlined_nodes): Update.

From-SVN: r249244

re PR fortran/80983 ([F03] memory leak when calling procedure-pointer component with allocatable result)

2017-06-16  Janus Weil  <janus@gcc.gnu.org>

PR fortran/80983
* gfortran.dg/proc_ptr_comp_51.f90: Repair test case.

From-SVN: r249243

Update prefetch tuning parameters for qdf24xx.

* config/aarch64/aarch64.c (qdf24xx_prefetch_tune): Update
prefetch settings, and enable prefetching by default at -O3.

Change-Id: I2f0da54a8c262f6fbd0dcfde4584141bb09f1013

From-SVN: r249242

Enable -fprefetch-loop-arrays at given optimization level.

* config/aarch64/aarch64.c (aarch64_override_options_internal):
Set flag_prefetch_loop_arrays according to tuning data.

Change-Id: Id41411e671e0a55dc7268e0ad0a4e8ff1421c90a

From-SVN: r249241

Add prefetch configuration to aarch64 backend.

* config/aarch64/aarch64-protos.h (struct cpu_prefetch_tune):
New tune structure.
(struct tune_params): Use cpu_prefetch_tune instead of cache_line_size.
[Unrelated to main purpose of the patch] Place the pointer field last
to enable type checking errors when tune structure are wrongly merged.
* config/aarch64/aarch64.c (generic_prefetch_tune,)
(exynosm1_prefetch_tune, qdf24xx_prefetch_tune,)
(thunderx2t99_prefetch_tune): New tune constants.
(tune_params *_tunings): Update all tunings (no functional change).
(aarch64_override_options_internal): Set PARAM_SIMULTANEOUS_PREFETCHES,
PARAM_L1_CACHE_SIZE, PARAM_L1_CACHE_LINE_SIZE, and PARAM_L2_CACHE_SIZE
from tunings structures.

Change-Id: I1ddbac1863dcf078a2e5b14dd904debc76a7da94

From-SVN: r249240

re PR sanitizer/81094 (-fsanitize=object-size does not instrument aggregate call arguments)

PR sanitizer/81094
* ubsan.c (instrument_null): Add T argument, use it instead
of computing it based on IS_LHS.
(instrument_object_size): Likewise.
(pass_ubsan::execute): Adjust instrument_null and
instrument_object_size callers to pass gimple_get_lhs or
gimple_assign_rhs1 result to it.  Use instrument_null instead of
calling get_base_address and instrument_mem_ref.  Handle
aggregate call arguments for object-size sanitization.

* c-c++-common/ubsan/object-size-11.c: New test.

From-SVN: r249239

Fix missing returns in libstdc++ header and tests

* include/bits/locale_conv.h (wbuffer_convert::_M_put): Add missing
return statement.
* testsuite/21_strings/basic_string_view/operations/copy/char/1.cc:
Return void.
* testsuite/21_strings/basic_string_view/operations/copy/wchar_t/1.cc:
Likewise.
* testsuite/23_containers/map/modifiers/insert_or_assign/1.cc: Add
missing return statements.
* testsuite/23_containers/unordered_map/modifiers/insert_or_assign.cc:
Likewise.
* testsuite/27_io/basic_istream/extractors_arithmetic/char/12.cc:
Return void.
* testsuite/special_functions/14_expint/pr68397.cc: Likewise.

From-SVN: r249238

re PR tree-optimization/81089 (ICE: tree check: expected ssa_name, have integer_cst in register_edge_assert_for_2, at tree-vrp.c:5023)

2017-06-16  Yury Gribov  <tetra2005@gmail.com>

PR tree-optimization/81089
* tree-vrp.c (is_masked_range_test): Validate operands of
subexpression.

From-SVN: r249237

Don't use >& for I/O redirection

* Makefile.am (check-go-tool): Don't use >& for I/O redirection.
* Makefile.in: Regenerate.

From-SVN: r249236

stl_bvector.h (__fill_bvector(_Bit_type*, unsigned int, unsigned int, bool)): Change signature.

2017-06-16  François Dumont  <fdumont@gcc.gnu.org>

* include/bits/stl_bvector.h
(__fill_bvector(_Bit_type*, unsigned int, unsigned int, bool)):
Change signature.
(std::fill(_Bit_iterator, _Bit_iterator, bool)): Adapt.
(_Bvector_impl_data): New.
(_Bvector_impl): Inherits from latter.
(_Bvector_impl(_Bit_alloc_type&&)): Delete.
(_Bvector_impl(_Bvector_impl&&)): New, default.
(_Bvector_base()): Default.
(_Bvector_base(_Bvector_base&&)): Default.
(_Bvector_base::_M_move_data(_Bvector_base&&)): New.
(vector(vector&&, const allocator_type&)): Use latter.
(vector<bool>::operator=(vector&&)): Likewise.
(vector<bool>::vector()): Default.
(vector<bool>::vector(vector&&)): Default.
(vector<bool>::assign(_InputIterator, _InputIterator)): Use
_M_assign_aux.
(vector<bool>::assign(initializer_list<bool>)): Likewise.
(vector<bool>::_M_initialize_value(bool)): New.
(vector<bool>(size_type, const bool&, const allocator_type&)): Use
latter.
(vector<bool>::_M_initialize_dispatch(_Integer, _Integer, __true_type)):
Likewise.
(vector<bool>::_M_fill_assign(size_t, bool)): Likewise.

From-SVN: r249235

PR c++/80560 - warn on undefined memory operations involving non-trivial types

gcc/c-family/ChangeLog:

PR c++/80560
* c.opt (-Wclass-memaccess): New option.

gcc/cp/ChangeLog:

PR c++/80560
* call.c (first_non_public_field, maybe_warn_class_memaccess): New
functions.
(has_trivial_copy_assign_p, has_trivial_copy_p): Ditto.
(build_cxx_call): Call maybe_warn_class_memaccess.

gcc/ChangeLog:

PR c++/80560
* dumpfile.c (dump_register): Avoid calling memset to initialize
a class with a default ctor.
* gcc.c (struct compiler): Remove const qualification.
* genattrtab.c (gen_insn_reserv): Replace memset with initialization.
* hash-table.h: Ditto.
* ipa-cp.c (allocate_and_init_ipcp_value): Replace memset with
  assignment.
* ipa-prop.c (ipa_free_edge_args_substructures): Ditto.
* omp-low.c (lower_omp_ordered_clauses): Replace memset with
default ctor.
* params.h (struct param_info): Make struct members non-const.
* tree-switch-conversion.c (emit_case_bit_tests): Replace memset
with default initialization.
* vec.h (vec_copy_construct, vec_default_construct): New helper
functions.
(vec<T>::copy, vec<T>::splice, vec<T>::reserve): Replace memcpy
with vec_copy_construct.
(vect<T>::quick_grow_cleared): Replace memset with default ctor.
(vect<T>::vec_safe_grow_cleared, vec_safe_grow_cleared): Same.
* doc/invoke.texi (-Wclass-memaccess): Document.

libcpp/ChangeLog:

PR c++/80560
* line-map.c (line_maps::~line_maps): Avoid calling htab_delete
with a null pointer.
(linemap_init): Avoid calling memset on an object of a non-trivial
type.

libitm/ChangeLog:

PR c++/80560
* beginend.cc (GTM::gtm_thread::rollback): Avoid calling memset
on an object of a non-trivial type.
(GTM::gtm_transaction_cp::commit): Use assignment instead of memcpy
to copy an object.
* method-ml.cc (orec_iterator::reinit): Avoid -Wclass-memaccess.

gcc/testsuite/ChangeLog:

PR c++/80560
* g++.dg/Wclass-memaccess.C: New test.

From-SVN: r249234

Daily bump.

From-SVN: r249233

re PR ada/81105 (system-freebsd-x86.ads renamed but referenced by gcc/ada/gcc-interface/Makefile.in)

PR ada/81105
* gcc-interface/Makefile.in (x86 kfreebsd): Adjust system.ads setting.
(i[3456]86-pc-gnu): Likewise.
(x86_64 kfreebsd): Likewise.

From-SVN: r249228

re PR fortran/80983 ([F03] memory leak when calling procedure-pointer component with allocatable result)

2017-06-15  Janus Weil  <janus@gcc.gnu.org>

PR fortran/80983
* trans-expr.c (gfc_conv_procedure_call): Deallocate the result of
scalar allocatable procedure-pointer components.

2017-06-15  Janus Weil  <janus@gcc.gnu.org>

PR fortran/80983
* gfortran.dg/proc_ptr_comp_51.f90: New test.

From-SVN: r249227

2017-06-15  François Dumont  <fdumont@gcc.gnu.org>

* src/c++98/tree.cc  [!_GLIBCXX_INLINE_VERSION]
(_Rb_tree_rotate_left, _Rb_tree_rotate_right): Delete.

From-SVN: r249226

Revert accidental checkin.

From-SVN: r249225

re PR lto/69866 (lto1: internal compiler error: in add_symbol_to_partition_1, at lto/lto-partition.c:158)

PR lto/69866
* lto-symtab.c (lto_symtab_merge_symbols): Drop useless definitions
that resolved externally.

* gcc.dg/lto/pr69866_0.c: New test.
* gcc.dg/lto/pr69866_1.c: Likewise.

From-SVN: r249224

Update comment about is_leaf

crtl->is_leaf has a comment above it referring to "Local Register
Allocation". However this is set by IRA and not LRA since the meaning
of Local Register Allocator in GCC has changed quite drastically since
1999 when this comment was introduced above the variable
current_function_is_leaf. From memory gcc in those days had "local"
and "global" register allocation - however all of this has changed
quite a bit drastically since.

This patch merely updates the comment as I did some head scratching
this morning about how this could be set before Local Register
Allocator especially given the comment in ira.c as below.

 /* Determine if the current function is a leaf before running IRA
     since this can impact optimizations done by the prologue and
     epilogue thus changing register elimination offsets.  */
  crtl->is_leaf = leaf_function_p ();

From-SVN: r249223

Fix problems with the MSP430 port's handling of persistent data.

PR target/78818
gcc * config/msp430/msp430.c (msp430_data_attr): Check that it's possible
for a variable to have a section before checking if the section has a
name.
Set section to.persistent if persistent attribute is set.
Warn if .persistent attribute is used on an automatic variable.

tests * gcc.target/msp430/pr78818-real.c: New template for tests.
* gcc.target/msp430/pr78818-auto.c: New test.
* gcc.target/msp430/pr78818-data-region.c: New test.
* gcc.target/msp430/pr78818-data-sec.c: New test.
* gcc.target/msp430/pr78818-auto-warn.c: New test.

From-SVN: r249222

re PR rtl-optimization/80474 (ipa-cp wrongly adding LO(symbol) twice)

PR rtl-optimization/80474
* reorg.c (update_block): Do not ignore instructions in a delay slot.

From-SVN: r249219

Use pthread effective-target in testsuite

* testsuite: Add dg-require-effective-target pthread to -pthread
tests.
Remove explicit target lists from dg-do and dg-options.

* testsuite/30_threads/async/forced_unwind.cc: Remove explit
target list from dg-options.
* testsuite/30_threads/packaged_task/forced_unwind.cc: Likewise.

* 30_threads/shared_mutex/cons/1.cc: Likewise.
Pass -pthread for all targets.
* 30_threads/shared_mutex/try_lock/1.cc: Likewise.
* 30_threads/shared_mutex/try_lock/2.cc: Likewise.
* 30_threads/shared_mutex/unlock/1.cc: Likewise.

* testsuite/30_threads/this_thread/57060.cc: Require c++11 via
dg-require-effective-target.

From-SVN: r249217

[ARM] Make gcc.target/arm/its.c more robust

2017-06-15  Thomas Preud'homme  <thomas.preudhomme@arm.com>

    gcc/testsuite/
    * gcc.target/arm/its.c: Check that no IT blocks has more than 2
    instructions in it rather than the number of IT blocks being 2.
    Transfer scan directive arm_thumb2 restriction to the whole
    testcase and restrict further to Thumb-only targets.

From-SVN: r249215

sdiv_costs_1.c: Use dg-add-options.

2017-06-15  Tamar Christina  <tamar.christina@arm.com>

* gcc.target/arm/sdiv_costs_1.c:
Use dg-add-options.

From-SVN: r249214

rs6000: Don't use REGNO in the add<mode>3 expander

The argument could be a subreg of reg instead, which means we have to
use reg_or_subregno instead of REGNO.

* config/rs6000/rs6000.md (add<mode>3): Use reg_or_subregno instead
of REGNO.

From-SVN: r249213

Daily bump.

From-SVN: r249212

compiler: avoid crash on erroneous type

    If there is an error constructing the backend type, the GCC backend
    will report that the size is 1.  That will then cause construction of
    the ptrmask to crash.  Avoid that case by just generating an empty
    ptrmask.

    Noticed while compiling a broken package.  The policy I've been
    following is to not commit a test case for a compiler crash on invalid
    code, so no test case.

    Reviewed-on: https://go-review.googlesource.com/45775

From-SVN: r249208

MIPS16/GCC: Emit bounds checking as RTL in `casesi'

gcc/
* config/mips/mips.md (MIPS16_T_REGNUM): Remove constant.
(casesi): Emit bounds checking as RTL.
(casesi_internal_mips16_<mode>): Remove bounds checking.

gcc/testsuite/
* gcc.target/mips/data-sym-jump.c: Adjust for whitespace changes.
* gcc.target/mips/pr51513-1.c: New test.
* gcc.target/mips/pr51513-2.c: New test.

From-SVN: r249207

Fix libstdc++ testsuite failures with -D_GLIBCXX_USE_CXX11_ABI=0

* doc/xml/manual/test.xml: Correct instructions on running tests.
* testsuite/21_strings/basic_string/cons/char/deduction.cc: Adjust to
pass when -D_GLIBCXX_USE_CXX11_ABI=0 added to RUNTESTFLAGS.
* testsuite/21_strings/basic_string/cons/wchar_t/deduction.cc:
Likewise.
* testsuite/27_io/basic_ios/copyfmt/char/1.cc: Likewise.
* testsuite/27_io/basic_ios/exceptions/char/1.cc: Likewise.
* testsuite/27_io/basic_istream/extractors_arithmetic/char/
exceptions_failbit.cc: Likewise.
* testsuite/27_io/basic_istream/extractors_arithmetic/wchar_t/
exceptions_failbit.cc: Likewise.
* testsuite/27_io/basic_istream/extractors_other/char/
exceptions_null.cc: Likewise.
* testsuite/27_io/basic_istream/extractors_other/wchar_t/
exceptions_null.cc: Likewise.
* testsuite/27_io/basic_istream/sentry/char/12297.cc: Likewise.
* testsuite/27_io/basic_istream/sentry/wchar_t/12297.cc: Likewise.
* testsuite/27_io/basic_ostream/inserters_other/char/
exceptions_null.cc: Likewise.
* testsuite/27_io/basic_ostream/inserters_other/wchar_t/
exceptions_null.cc: Likewise.
* testsuite/27_io/ios_base/storage/2.cc: Likewise.

From-SVN: r249206

Makefile.def: Add check-gotools to go check targets.

* Makefile.def: Add check-gotools to go check targets.
* Makefile.in: Rebuild.

From-SVN: r249204

Makefile.am (libgosrcdir): Define.

* Makefile.am (libgosrcdir): Define.
(check-head, check-gccgo, check-go-tool): New targets.
(CHECK_ENV): Define.
(check): New target.
(mostlyclean-local): New target.
* Makefile.in: Rebuild.

From-SVN: r249203

gcc: xtensa: allow XCHAL_* macros to be non-constant

XCHAL_* macros from the xtensa-config.h are used in a number of places
that require them to be preprocessor constants. Rewrite these places so
that non-constant XCHAL_* definitions could be used there.

2017-06-14  Max Filippov  <jcmvbkbc@gmail.com>
gcc/
* config/xtensa/xtensa.c (xtensa_option_override): Append
MASK_CONST16 to target_flags in the absence of TARGET_L32R.
(hwloop_optimize, hwloop_fail, hwloop_pattern_reg,
 xtensa_doloop_hooks): Define unconditionally.
(xtensa_reorg_loops): Only call reorg_loops in the presence of
TARGET_LOOPS.
* config/xtensa/xtensa.h (TARGET_L32R): New definition.
(TARGET_DEFAULT): Remove XCHAL_HAVE_L32R condition and account
for it in xtensa_option_override.
(HARD_FRAME_POINTER_IS_FRAME_POINTER,
 HARD_FRAME_POINTER_IS_ARG_POINTER): New definitions.

From-SVN: r249202

Recognize '-' as special -MF argument (write to stdout)

Sometimes it is useful to generate pre-processed output to a file and
the dependency information to stdout for further analysis/processing.
For example:

g++ -E -MD -fdirectives-only -o test.ii test.cxx

This will generate the dependency information to test.d (as per the
documentation). While changing this behavior is probably unwise, one
traditional (e.g., supported by -o) way to handle this is to recognize
the special '-' file name as an instruction to write to stdout:

g++ -E -MD -fdirectives-only -o test.ii -MF - test.cxx

Currently this will create a file named '-'. The included patch
changes this behavior to write to stdout.

Note also that Clang has supported this from at least version 3.5.

gcc:
2017-06-14  Boris Kolpackov  <boris@codesynthesis.com>

* doc/cppopts.texi: Document '-' special value to -MF.

gcc/c-family:
2017-06-14  Boris Kolpackov  <boris@codesynthesis.com>

* c-opts.c (c_common_finish): Handle '-' special value to -MF.

From-SVN: r249201

Improve Cortex-A53 FP scheduler

The Cortex-A53 scheduler model of FMAC bypass is not quite right
for FMAC to FMAC forwarding.  Experiments also show the latencies of
FP operations are too high as well.  Rather than adding more bypasses,
adjust the latencies of FP instructions to get a better schedule on
average.  As a result SPECFP2006 is 1.1% faster.

    gcc/
* config/arm/cortex-a53.md (cortex_a53_fpalu) Adjust latency.
(cortex_a53_fconst): Likewise.
(cortex_a53_fpmul): Likewise.
(cortex_a53_f_load_64): Likewise.
(cortex_a53_f_load_many): Likewise.
(cortex_a53_advsimd_alu): Likewise.
(cortex_a53_advsimd_alu_q): Likewise.
(cortex_a53_advsimd_mul): Likewise.
(cortex_a53_advsimd_mul_q): Likewise.
(fpmac bypass): Add new bypass for fpmac-fpmac case.
Add missing fmul, r2f_cvt and fconst cases.

From-SVN: r249200

cmd/cgo: make _cgo_flags consistent across runs

    The go tool will pass -I objdir as one of the flags, where objdir is
    the temporary build directory. Remove that from _cgo_flags: we don't
    need it, and it will be different each time.

    Sort the flags to avoid the unpredictable map iteration order.

    This matters for gccgo because for a package that uses cgo, the go
    tool when building for gccgo will store the _cgo_flags file in the
    archive. That means that we want to generate identical _cgo_flags for
    every run.

    The test for this is the cmd/go testsuite, to follow in a future CL.

    Reviewed-on: https://go-review.googlesource.com/45692

From-SVN: r249199

cmd/go: gccgo: consistent results, fix vendoring

    Pass the -fdebug-prefix-map and -gno-record-gcc-switches compiler
    options to gccgo to generate consistent results.

    Fix the vendoring code to look for /vendor/, not just /vendor, to
    avoid being confused by something like vendor/vendor.org.

    Tested by the cmd/go tests in a followup CL.

    Reviewed-on: https://go-review.googlesource.com/45695

From-SVN: r249198

cmd/go: skip tests that do not work with gccgo

    These tests fail for various reasons, most commonly because gccgo
    doesn't really have GOROOT, so things like `go build errors` fail.

    Reviewed-on: https://go-review.googlesource.com/45696

From-SVN: r249197

go/build: use GCCGOTOOLDIR to find gccgo tools

    Add the environment variable GCCGOTOOLDIR to permit overriding the default
    directory where tools like cgo are found when building with gccgo.

    This will be used by the cmd/go tests in a future CL.

    Reviewed-on: https://go-review.googlesource.com/45694

From-SVN: r249196

internal/testenv: support testing gccgo

    If GO_TESTING_GOTOOLS is set in the environment, permit tests using
    gccgo to run the go tool. Like GO_BUILDER_NAME, this should not be set
    normally. But it is needed when testing the go tool itself, and will
    be set by the gotools Makefile in a future CL.

    Reviewed-on: https://go-review.googlesource.com/45693

From-SVN: r249195

compiler: remove "DIR/../" when joining relative import path

    Otherwise if DIR does not exist, the path does not work. This matches
    what the gc cmd/compile tool does, because it calls path.Join.

    The test for this is the cmd/go tests, to be added in a follow-up CL.

    Reviewed-on: https://go-review.googlesource.com/45691

From-SVN: r249194

re PR tree-optimization/81083 (ICE: Unable to coalesce ssa_names 4 and 13 which are marked as MUST COALESCE)

2017-06-14  Richard Biener  <rguenther@suse.de>

PR tree-optimization/81083
* gcc.dg/torture/pr81083.c: Add prototypes.

From-SVN: r249193

re PR middle-end/81088 (UBSAN: false positive as a result of reassosiation)

2017-06-14  Richard Biener  <rguenther@suse.de>

PR middle-end/81088
* fold-const.c (split_tree): Drop TREE_OVERFLOW flag from
literal constants.
(fold_binary_loc): When associating do not treat pre-existing
TREE_OVERFLOW on literal constants as a reason to allow
TREE_OVERFLOW on associated literal constants.

* c-c++-common/ubsan/pr81088.c: New testcase.

From-SVN: r249192

cp-gimplify.c (cp_genericize_r): Turn most of the function into a switch (TREE_CODE (stmt)) statement from long...

* cp-gimplify.c (cp_genericize_r): Turn most of the function
into a switch (TREE_CODE (stmt)) statement from long else if
sequence.

From-SVN: r249191

sparc.h (MASK_ISA): Add MASK_LEON and MASK_LEON3.

* config/sparc/sparc.h (MASK_ISA): Add MASK_LEON and MASK_LEON3.
(MASK_FEATURES): New macro.
* config/sparc/sparc.c (sparc_option_override): Remove the special
handling of -mfpu and generalize it to all MASK_FEATURES switches.

From-SVN: r249189

simplify-rtx.c (simplify_binary_operation_1): Do not simplify a division of 0 if non-call exceptions are enabled.

* simplify-rtx.c (simplify_binary_operation_1) <UDIV>: Do not simplify
a division of 0 if non-call exceptions are enabled.

From-SVN: r249188

re PR target/71663 (aarch64 Vector initialization can be improved slightly)

PR target/71663
gcc
* config/aarch64/aarch64.c (aarch64_expand_vector_init):
Improve vector initialization code gen for only variable case.

gcc/testsuite
* gcc.target/aarch64/vect-init-1.c: Newtestcase.
* gcc.target/aarch64/vect-init-2.c: Likewise.
* gcc.target/aarch64/vect-init-3.c: Likewise.
* gcc.target/aarch64/vect-init-4.c: Likewise.
* gcc.target/aarch64/vect-init-5.c: Likewise.

Co-Authored-By: Naveen H.S <Naveen.Hurugalawadi@cavium.com>
From-SVN: r249187

re PR c++/58541 ([c++11] Bogus "error: redeclaration ... differs in ‘constexpr’")

2017-06-14  Paolo Carlini  <paolo.carlini@oracle.com>

PR c++/58541
* g++.dg/cpp0x/constexpr-58541.C: New.

From-SVN: r249186

* config/sparc/driver-sparc.c (cpu_names): Add SPARC-T5 entry.

From-SVN: r249183

re PR tree-optimization/81083 (ICE: Unable to coalesce ssa_names 4 and 13 which are marked as MUST COALESCE)

2017-06-14  Richard Biener  <rguenther@suse.de>

PR tree-optimization/81083
* tree-ssa-sccvn.c (vn_reference_lookup_3): Do not use abnormals
as values.

* gcc.dg/torture/pr81083.c: New testcase.

From-SVN: r249182

vrp101.c: Fix scan-tree-dump regex.

* gcc.dg/tree-ssa/vrp101.c: Fix scan-tree-dump regex.
* gcc.dg/tree-ssa/builtin-sprintf-2.c: Likewise.  Adjust comment.

From-SVN: r249181

vrp101.c: Fix scan-tree-dump regex.

* gcc.dg/tree-ssa/vrp101.c: Fix scan-tree-dump regex.
* gcc.dg/tree-ssa/builtin-sprintf-2.c: Likewise.  Adjust comment.

From-SVN: r249180

Daily bump.

From-SVN: r249179

re PR sanitizer/80973 (ICE with lambda and -fsanitize=undefined)

PR c++/80973
* cp-gimplify.c (cp_genericize_r): Don't instrument MEM_REF second
argument even if it has REFERENCE_TYPE.

* g++.dg/ubsan/pr80973.C: New test.

From-SVN: r249174

re PR c++/80984 (ICE with label/variable ambiguity)

PR c++/80984
* cp-gimplify.c (cp_genericize): Only look for VAR_DECLs in
BLOCK_VARS (outer) chain.
(cxx_omp_const_qual_no_mutable): Likewise.

* g++.dg/opt/nrv18.C: New test.

From-SVN: r249173

re PR go/80964 (cross-gotools are not executable on host system)

PR go/80964
* configure.ac: Set NATIVE if host_alias = target_alias.
* configure: Rebuild.

From-SVN: r249172

re PR objc/80949 (ICE in do_warn_duplicated_branches_r)

PR objc/80949
* c-warn.c (do_warn_duplicated_branches): Return if any of the
branches is null.

From-SVN: r249171

rs6000: Comment fixes + some leftovers

* config/rs6000/rs6000.c: Update all comments that mentioned SPE.
(rs6000_expand_builtin): Remove RS6000_BTC_EVSEL.
* config/rs6000/rs6000.h (RS6000_BTC_EVSEL): Delete.
* config/rs6000/vxworks.h (VXCPU_FOR_8548): Delete.  Adjust former use.
* config/rs6000/vxworksae.h (VXCPU_FOR_8548): Delete.
* config/rs6000/vxworksmils.h (VXCPU_FOR_8548): Delete.

From-SVN: r249170

rs6000: Remove VECTOR_SPE

* config/rs6000/rs6000-opts.h (enum rs6000_vector): Delete VECTOR_SPE.
* config/rs6000/rs6000.c (rs6000_debug_vector_unit): Delete VECTOR_SPE.

From-SVN: r249169

rs6000: Remove FIXED_SCRATCH

* config/rs6000/rs6000.h (FIXED_SCRATCH): Delete.

From-SVN: r249168

rs6000: Updates to t-rtems

* config/rs6000/t-rtems: Don't handle SPE.

From-SVN: r249167

rs6000: Updates to t-linux

* config/rs6000/t-linux: Don't handle SPE.

From-SVN: r249166

rs6000: Remove eabispe.h

* config/rs6000/eabispe.h: Delete file.

From-SVN: r249165

rs6000: Remove t-spe

* config/rs6000/t-spe: Delete file.

From-SVN: r249164

rs6000: Remove SPE_CONST_OFFSET_OK

* config/rs6000/rs6000.c (SPE_CONST_OFFSET_OK): Delete.
(rs6000_legitimate_offset_address_p): Return false for anything in
V2SImode or V2SFmode.

From-SVN: r249163

rs6000: Sanitize vector modes

This removes the vector modes that were only used by SPE.  It also
rearranges things so it is easier to see what is there, and for what.

* config/rs6000/rs6000-modes.def: Remove all 8-byte vector modes
except V2SF and V2SI.  Rearrange the vector modes, and add comments.
* config/rs6000/rs6000.c (rs6000_debug_reg_global): Remove V8QImode
and V4HImode.
(reg_offset_addressing_ok_p): Remove V4HImode and V1DImode.
(rs6000_legitimate_offset_address_p): Ditto.
(rs6000_emit_move): Ditto.
(rs6000_init_builtins): Remove V4HI_type_node.

From-SVN: r249162

compiler: containing small bfunction mixup in Gogo::write_globals

    Fix buglet in Gogo::write_globals-- in a couple of places the
    wrong Bfunction was being used for the containing (not target)
    function when creating calls for init functions.

    Reviewed-on: https://go-review.googlesource.com/45510

From-SVN: r249159

Implement no_sanitize function attribute

2017-06-13  Martin Liska  <mliska@suse.cz>

PR sanitize/78204
* c-c++-common/ubsan/attrib-2.c (float_cast2): Enhance the
test by adding no_sanitize attribute.
* gcc.dg/asan/use-after-scope-4.c: Likewise.
2017-06-13  Martin Liska  <mliska@suse.cz>

PR sanitize/78204
* c-attribs.c (add_no_sanitize_value): New function.
(handle_no_sanitize_attribute): Likewise.
(handle_no_sanitize_address_attribute): Use the function.
(handle_no_sanitize_thread_attribute): New function.
(handle_no_address_safety_analysis_attribute): Use
add_no_sanitize_value.
(handle_no_sanitize_undefined_attribute): Likewise.
* c-common.h: Declare new functions.
* c-ubsan.c (ubsan_instrument_division): Use sanitize_flags_p.
(ubsan_instrument_shift): Likewise.
(ubsan_instrument_bounds): Likewise.
(ubsan_maybe_instrument_array_ref): Likewise.
(ubsan_maybe_instrument_reference_or_call): Likewise.
2017-06-13  Martin Liska  <mliska@suse.cz>

PR sanitize/78204
* asan.c (asan_sanitize_stack_p): Use sanitize_flags_p.
(gate_asan): Likewise.
* asan.h (asan_no_sanitize_address_p): Remove the function.
(sanitize_flags_p): New function.
* builtins.def: Fix coding style.
* common.opt: Use renamed enum value.
* convert.c (convert_to_integer_1): Use sanitize_flags_p.
* doc/extend.texi: Document no_sanitize attribute.
* flag-types.h (enum sanitize_code): Rename SANITIZE_NONDEFAULT
to SANITIZE_UNDEFINED_NONDEFAULT.
* gcc.c (sanitize_spec_function): Use the renamed enum value.
* gimple-fold.c (optimize_atomic_compare_exchange_p):
Use sanitize_flags_p.
* gimplify.c (gimplify_function_tree): Likewise.
* ipa-inline.c (sanitize_attrs_match_for_inline_p): Likewise.
* opts.c (parse_no_sanitize_attribute): New function.
(common_handle_option): Use renamed enum value.
* opts.h (parse_no_sanitize_attribute): Declare.
* tree.c (sanitize_flags_p): New function.
* tree.h: Declared here.
* tsan.c: Use sanitize_flags_p.
* ubsan.c (ubsan_expand_null_ifn): Likewise.
(instrument_mem_ref): Likewise.
(instrument_bool_enum_load): Likewise.
(do_ubsan_in_current_function): Remove the function.
(pass_ubsan::execute): Use sanitize_flags_p.
* ubsan.h: Remove do_ubsan_in_current_function
* tree-cfg.c (print_no_sanitize_attr_value): New function.
(dump_function_to_file): Use it here.
2017-06-13  Martin Liska  <mliska@suse.cz>

PR sanitize/78204
* class.c (build_base_path): Use sanitize_flags_p.
* cp-gimplify.c (cp_genericize_r): Likewise.
(cp_genericize_tree): Likewise.
(cp_genericize): Likewise.
* cp-ubsan.c (cp_ubsan_instrument_vptr_p): Likewise.
* decl.c (compute_array_index_type): Likewise.
(start_preparsed_function): Likewise.
* decl2.c (one_static_initialization_or_destruction): Likewise.
* init.c (finish_length_check): Likewise.
* lambda.c (maybe_add_lambda_conv_op): Likewise.
* typeck.c (cp_build_binary_op): Likewise.
(build_static_cast_1): Likewise.
2017-06-13  Martin Liska  <mliska@suse.cz>

PR sanitize/78204
* c-convert.c (convert): Use sanitize_flags_p.
* c-decl.c (grokdeclarator): Likewise.
* c-typeck.c (convert_for_assignment): Likewise.
(c_finish_return): Likewise.
(build_binary_op): Likewise.

From-SVN: r249158

[ARM][Testsuite]make dump_stack function call conditional in cold-lc.c

After r249013, die () and dump_stack () are both in cold section. This makes
the compiler generate bl instruction for the function call, instead of
honoring the -mlong-calls option.

This patch changes the dump_stack function call conditional, which fixes the
regression.

gcc/testsuite/

* gcc.target/arm/cold-lc.c: Update coding style, call dump_stack
conditionally.

From-SVN: r249157

runtime: don't always show frames with no function in traceback

    If there is no function name, the traceback is generally
    uninformative.  In earlier versions we did not show such frames.
    Restore that behavior.  These frames can be seen with GOTRACEBACK=system.

    Reviewed-on: https://go-review.googlesource.com/45431

From-SVN: r249156

Fixed spelling mistake in gcc/ChangeLog

From-SVN: r249155

[PR80803 2/2] Diligent queuing in SRA grp_write prop

2017-06-13  Martin Jambor  <mjambor@suse.cz>

PR tree-optimization/80803
PR tree-optimization/81063
* tree-sra.c (subtree_mark_written_and_enqueue): Move up in the file.
(propagate_subaccesses_across_link): Enqueue subtree whneve necessary
instead of relying on the caller.

testsuite/
gcc.dg/tree-ssa/pr80803.c: New test.
gcc.dg/tree-ssa/pr81063.c: Likewise.

From-SVN: r249154

[PR80803 1/2] Streamline SRA access enqueuing

2017-06-13  Martin Jambor  <mjambor@suse.cz>

* tree-sra.c (add_access_to_work_queue): Only enqueue accesses
that have a first_link.
(sort_and_splice_var_accesses): Do not check first_link before
enquing.
(subtree_mark_written_and_enqueue): Likewise.
(propagate_all_subaccesses): Likewise and do not stop at first
parent with a first_link.

From-SVN: r249153

Trivial fix in dump_access_tree_1

2017-06-13  Martin Jambor  <mjambor@suse.cz>

* tree-sra.c (dump_access_tree_1): Fix accidental dumping to stderr
instead of f.

From-SVN: r249152

match.pd: New pattern.

2017-06-13  Yury Gribov  <tetra2005@gmail.com>

gcc/
* match.pd: New pattern.

gcc/testsuite/
* c-c++-common/fold-masked-cmp-3.c: New test.

From-SVN: r249151

tree-vrp.c (is_masked_range_test): New function.

2017-06-13  Yury Gribov  <tetra2005@gmail.com>

gcc/
* tree-vrp.c (is_masked_range_test): New function.
(register_edge_assert_for): Determine ranges for
some bit tests.

From-SVN: r249150

re PR tree-optimization/67328 (range test rather than single bit test for code testing enum values)

2017-06-13  Yury Gribov  <tetra2005@gmail.com>

gcc/
PR tree-optimization/67328
* fold-const.c (maskable_range_p): New function.
(build_range_check): Generate bittests if possible.

gcc/testsuite/
PR tree-optimization/67328
* c-c++-common/fold-masked-cmp-1.c: New test.
* c-c++-common/fold-masked-cmp-2.c: Likewise.
* gcc.dg/pr46309.c: Fix pattern.
* gcc.dg/pr46309-2.c: Likewise.

From-SVN: r249149

sdiv_costs_1.c: Require arm_arch_v8a_ok and add march option.

2017-06-13  Tamar Christina  <tamar.christina@arm.com>

* gcc.target/arm/sdiv_costs_1.c:
Require arm_arch_v8a_ok and add march option.

From-SVN: r249148

Enhance dump_probability function.

2017-06-13  Martin Liska  <mliska@suse.cz>

* gimple-pretty-print.c (dump_probability): Add new argument.
(dump_edge_probability): Dump both probability and count.
(dump_gimple_label): Likewise.
(dump_gimple_bb_header): Likewise.
2017-06-13  Martin Liska  <mliska@suse.cz>

* gcc.dg/tree-ssa/builtin-sprintf-2.c: Adjust scanned pattern.
* gcc.dg/tree-ssa/dump-2.c: Likewise.
* gcc.dg/tree-ssa/vrp101.c: Likewise.

From-SVN: r249147

re PR target/81072 ([avr] Support some XMEGA devices with flash seen in RAM address space.)

PR target/81072
* config/avr/avr-devices.c: Fix indentation.
* config/avr/gen-avr-mmcu-specs.c: Dito.

From-SVN: r249146

tree-vect-loop.c (vect_model_reduction_cost): Do not fail, instead get vector type from stmt_info.

2017-06-13  Richard Biener  <rguenther@suse.de>

* tree-vect-loop.c (vect_model_reduction_cost): Do not fail,
instead get vector type from stmt_info.
(vectorizable_reduction): Adjust.  Remove dead code.

From-SVN: r249145

re PR sanitizer/81065 (UBSAN: false positive as a result of distribution involving different types)

2017-06-13  Richard Biener  <rguenther@suse.de>

PR middle-end/81065
* fold-const.c (extract_muldiv_1): Remove bogus distribution
case of C * (x * C2 + C3).
(fold_addr_of_array_ref_difference): Properly fold index difference.

* c-c++-common/ubsan/pr81065.c: New testcase.

From-SVN: r249144

runtime: clear isSystemGoroutine in goexit

    Otherwise it may be set when the g struct is reused via gfput/gfget.

    Test is golang.org/x/net/http2 with GOMAXPROCS=12.

    Reviewed-on: https://go-review.googlesource.com/45430

From-SVN: r249143

Daily bump.

From-SVN: r249142

runtime: ignore _Gscan bit when checking status in CgocallDone

    Also always access the atomicstatus field atomically.

    The effect of not checking the _Gscan bit is that if the GC decides to
    scan the stack just as the goroutine is leaving the system call, the
    goroutine might fail to call exitsyscall.  Then then typically causes
    a runtime assertion failure later on.  If we do call exitsyscall as we
    should, it will stall (in casgstatus) until the _Gscan bit is cleared.

    No separate test.  I've observed causing sporadic failures running the
    misc/cgo tests, but we don't currently have a way to run those
    routinely for gccgo.  I should fix that.

    Reviewed-on: https://go-review.googlesource.com/45392

From-SVN: r249138

Finish implementing P0426R1 "Constexpr for std::char_traits" for C++17

As discussed in PR c++/80265 ("__builtin_{memcmp,memchr,strlen} are
not usable in constexpr functions"), use __builtin_constant_p to tell
whether we can defer to a constexpr algorithm.

I used __always_inline__ just to be thorough.  It isn't really really
necessary as far as I could determine.

Changes like these:

 if (__n == 0)
   return 0;
 - return wmemcmp(__s1, __s2, __n);
 + else
 +   return wmemcmp(__s1, __s2, __n);

are necessary otherwise G++ complains that we're calling a
non-constexpr function, which looks like a a manifestation of PR67026
to me.

libstdc++-v3:
2017-06-12  Pedro Alves  <palves@redhat.com>

* doc/xml/manual/status_cxx2017.xml: Update C++17 constexpr
char_traits status.
* doc/html/*: Regenerate.

* include/bits/char_traits.h (_GLIBCXX_ALWAYS_INLINE): Define if
not already defined.
(__cpp_lib_constexpr_char_traits): Uncomment.
(__constant_string_p, __constant_char_array_p): New.
(std::char_traits<char>, std::char_traits<wchar_t>): Add
_GLIBCXX17_CONSTEXPR on compare, length and find and use
__constant_string_p, __constant_char_array_p and
__builtin_constant_p to defer to __gnu_cxx::char_traits at compile
time.

* testsuite/21_strings/char_traits/requirements/
constexpr_functions_c++17.cc: Uncomment
__cpp_lib_constexpr_char_traits tests.  Uncomment
test_compare<char>, test_length<char>, test_find<char>,
test_compare<wchar_t>, test_length<wchar_t> and test_find<wchar_t>
static_assert tests.

From-SVN: r249137

stl_tree.h (_Rb_tree_impl()): Restore _Node_allocator default init.

2017-06-12  François Dumont  <fdumont@gcc.gnu.org>

* include/bits/stl_tree.h (_Rb_tree_impl()): Restore _Node_allocator
default init.
* testsuite/util/testsuite_allocator.h
(__gnu_test::default_init_allocator<>) New.
* testsuite/23_containers/set/allocator/default_init.cc: New.
* testsuite/23_containers/map/allocator/default_init.cc: New.

From-SVN: r249136

More refinements to fixing sparc's PR target/80968.

gcc/

PR target/80968
* config/sparc/sparc.md (return expander): Emit frame blockage if
function uses alloca.

From-SVN: r249134

compiler: avoid orphaning Bexpressions when processing conversions

    The method Type_conversion_expression::do_get_backend was (in some
    circumstances) creating a Bexpression for the source expression of the
    conversion and then throwing it away before using it. Fix up this
    method to insure that the call to get_backend() on the source
    expression is only made when the result will be used.

    Reviewed-on: https://go-review.googlesource.com/45350

From-SVN: r249131

PR libstdc++/55917 do not handle exceptions in std::thread

PR libstdc++/55917
* src/c++11/thread.cc (execute_native_thread_routine): Remove
try-block so that exceptions propagate out of the thread and terminate
is called by the exception-handling runtime.
(execute_native_thread_routine_compat): Likewise.
* testsuite/30_threads/thread/cons/terminate.cc: New.

From-SVN: r249130

split-1.c: Require split_stack, don't require freorder.

* gcc.dg/tree-prof/split-1.c: Require split_stack, don't require
freorder.  Update comment to explain test.

From-SVN: r249129