All these tests fail when using -mpure-code:
* some force A or R profile
* some use Neon
* some use -fpic/-fPIC
all of which are not supported by this option.

2019-11-05  Christophe Lyon  <christophe.lyon@linaro.org>

	* gcc.target/arm/attr-crypto.c: Skip if -mpure-code is used.
	* gcc.target/arm/attr-neon-fp16.c: Likewise.
	* gcc.target/arm/combine-cmp-shift.c: Likewise.
	* gcc.target/arm/data-rel-1.c: Likewise.
	* gcc.target/arm/data-rel-2.c: Likewise.
	* gcc.target/arm/data-rel-3.c: Likewise.
	* gcc.target/arm/ftest-armv7a-arm.c: Likewise.
	* gcc.target/arm/ftest-armv7a-thumb.c: Likewise.
	* gcc.target/arm/ftest-armv7r-arm.c: Likewise.
	* gcc.target/arm/ftest-armv7r-thumb.c: Likewise.
	* gcc.target/arm/ftest-armv7ve-arm.c: Likewise.
	* gcc.target/arm/ftest-armv7ve-thumb.c: Likewise.
	* gcc.target/arm/ftest-armv8a-arm.c: Likewise.
	* gcc.target/arm/ftest-armv8a-thumb.c: Likewise.
	* gcc.target/arm/lceil-vcvt_1.c: Likewise.
	* gcc.target/arm/lfloor-vcvt_1.c: Likewise.
	* gcc.target/arm/lround-vcvt_1.c: Likewise.
	* gcc.target/arm/memset-inline-10.c: Likewise.
	* gcc.target/arm/mod_2.c: Likewise.
	* gcc.target/arm/mod_256.c: Likewise.
	* gcc.target/arm/pr40657-1.c: Likewise.
	* gcc.target/arm/pr44788.c: Likewise.
	* gcc.target/arm/pr50305.c: Likewise.
	* gcc.target/arm/pr51835.c: Likewise.
	* gcc.target/arm/pr51915.c: Likewise.
	* gcc.target/arm/pr52006.c: Likewise.
	* gcc.target/arm/pr53187.c: Likewise.
	* gcc.target/arm/pr58784.c: Likewise.
	* gcc.target/arm/pr59575.c: Likewise.
	* gcc.target/arm/pr59923.c: Likewise.
	* gcc.target/arm/pr60650-2.c: Likewise.
	* gcc.target/arm/pr60657.c: Likewise.
	* gcc.target/arm/pr60663.c: Likewise.
	* gcc.target/arm/pr67439_1.c: Likewise.
	* gcc.target/arm/pr68620.c: Likewise.
	* gcc.target/arm/pr7676.c: Likewise.
	* gcc.target/arm/pr79239.c: Likewise.
	* gcc.target/arm/pr81863.c: Likewise.
	* gcc.target/arm/pragma_attribute.c: Likewise.
	* gcc.target/arm/pragma_cpp_fma.c: Likewise.
	* gcc.target/arm/require-pic-register-loc.c: Likewise.
	* gcc.target/arm/thumb-ltu.c: Likewise.
	* gcc.target/arm/thumb1-pic-high-reg.c: Likewise.
	* gcc.target/arm/thumb1-pic-single-base.c: Likewise.
	* gcc.target/arm/tlscall.c: Likewise.
	* gcc.target/arm/unsigned-float.c: Likewise.
	* gcc.target/arm/vrinta-ce.c: Likewise.

From-SVN: r277828

2019-11-05  Martin Liska  <mliska@suse.cz>

	* symbol-summary.h: Use ggc_delete.

From-SVN: r277827

Here is the one-liner that fixes the incorrect
vec_perm cost for thunderx2t99 chip.
With the patch applied 526.blender of CPU2017
gets ~5% improvement with no measurable changes
for other benchmarks.

2019-11-05  Anton Youdkevitch  <anton.youdkevitch@bell-sw.com>

	* config/aarch64/aarch64.c (thunderx2t99_vector_cost):
	Change vec_perm field to 10.

From-SVN: r277826

From-SVN: r277824

2019-11-05  Martin Liska  <mliska@suse.cz>

	* symbol-summary.h: Rename allocator to m_allocator and
	add comment.

From-SVN: r277823

2019-11-05  Richard Biener  <rguenther@suse.de>

	PR tree-optimization/92324
	* tree-vect-loop.c (check_reduction_path): For MIN/MAX require
	all signed or unsigned operations.

	* gcc.dg/vect/pr92324-3.c: New testcase.

From-SVN: r277822

	* hsa-brig.c: Include alloc-pool.h
	* hsa-dump.c: Likewise.
	* hsa-gen.c: Likewise.
	* hse-regalloc.c: Likewise.
	* ipa-hsa.c: Likewise.
	* ipa-predicate.c: Likewise.
	* ipa-reference.c: Likewise.
	* ipa-sra.c: Likewise.
	* omp-expand.c: Likewise.
	* omp-general.c: Likewise.
	* omp-low.c: Likewise.
	* sumbol-summary.h (function_summary_base): Add allocator.
	(function_summary<T *>::function_summary): Update construction.
	(fast_function_summary<T *, V>::fast_function_summary): Likewise.
	(call_summary_base): Add allcator.
	(call_summary<T *>::call_summary): Update construction.
	(fast_call_summary<T *, V>::fast_call_summary): Likewise.

From-SVN: r277821

re PR tree-optimization/91945 (ICE: tree check: expected integer_cst, have var_decl in get_len, at tree.h:5837 since r274997)

	PR tree-optimization/91945
	* builtins.c (compute_objsize): For ARRAY_REF, only multiply off
	by tpsize if it is both non-NULL and INTEGER_CST, otherwise punt.
	Formatting fix.

	* gfortran.dg/pr91945.f90: New test.

From-SVN: r277820

	PR c++/92343
	* constexpr.c (potential_constant_expression_1): Return true rather
	than false for PREDICT_EXPR.

	* g++.dg/cpp2a/attr-likely6.C: New test.

From-SVN: r277819

Move vrp_set_zero_nonzero_bits from tree-vrp.c into vr-values.c, and make it use wi_set_zero_nonzero_bits.

Move vrp_set_zero_nonzero_bits from tree-vrp.c into vr-values.c, and
make it use wi_set_zero_nonzero_bits.  Remove the now redundant
wide_int_range_set_zero_nonzero_bits.

From-SVN: r277818

        gcc/
        * config/gcn/t-omp-device: New.
        Should have been part of Rev. 277797.

From-SVN: r277817

2019-11-05  Martin Liska  <mliska@suse.cz>

	PR c++/92339
	* fold-const.c (operand_compare::hash_operand): Remove
	FIELD_DECL handling.
2019-11-05  Martin Liska  <mliska@suse.cz>

	PR c++/92339
	* g++.dg/pr92339.C: New test.

From-SVN: r277816

2019-10-05  Richard Biener  <rguenther@suse.de>

        * crontab: Disable snapshots from gcc-7-branch.

From-SVN: r277815

	* decl.c (omp_declare_variant_finalize_one): Call
	declare_simd_adjust_this not just on the context, but also on the
	variant-id expression for methods.  Don't call
	cp_get_callee_fndecl_nofold, call cp_get_callee and only if it is
	safe cp_get_fndecl_from_callee.  Don't try to print as %qD
	NULL in diagnostics.
	* pt.c (tsubst_attribute): Handle "omp declare variant base"
	attribute.
	(tsubst_function_decl): Call omp_declare_variant_finalize
	if there are any "omp declare variant base" attributes left.

	* g++.dg/gomp/declare-variant-7.C: New test.
	* g++.dg/gomp/declare-variant-8.C: New test.

From-SVN: r277814

From-SVN: r277813

symbolics without dying.

From-SVN: r277812

From-SVN: r277810

2019-11-04  Kamlesh Kumar  <kamleshbhalui@gmail.com>

gcc/cp
        * cp/mangle.c (write_template_arg_literal): Handle nullptr
        mangling.
gcc
	* common.opt (-fabi-version): Document =14.
	* doc/invoke.texi (C++ Dialect Options): Likewise.
gcc/c-family
        * c-opts.c (c_common_post_options): Update
        latest_abi_version.
libiberty
        * cp-demangle.c (d_expr_primary): Handle
        nullptr demangling.
        * testsuite/demangle-expected: Added test.

From-SVN: r277801

From-SVN: r277799

From-SVN: r277798

        gcc/
        * config/gcn/gcn.c (gcn_omp_device_kind_arch_isa): New function.
        (TARGET_OMP_DEVICE_KIND_ARCH_ISA): Redefine to
        gcn_omp_device_kind_arch_isa.
        * config/gcn/t-omp-device: New file.
        * configure.ac: Support gcn for omp_device_property.
        * configure: Regenerate.

From-SVN: r277797

Remove handle_pointers argument from all the vrp_val*{min,max} functions.  Always
assume pointers should be handled.

From-SVN: r277796

	* ipa-reference.c (init_function_info): Initialize
	info->global.statics_read.

From-SVN: r277795

make sure we build canonically correct ranges.

From-SVN: r277794

The SVE port now tries to register variable-length VECTOR_TYPEs
at start-up, so it's no longer possible to use the asserting
to_constant on the number of vector elements.  This patch punts
on variable element counts instead, just like we do for other
things that the frontend doesn't recognise.

The brace indentation matches the surrounding style.

2019-11-04  Richard Sandiford  <richard.sandiford@arm.com>

gcc/d/
	* d-builtins.cc (build_frontend_type): Cope with variable
	TYPE_VECTOR_SUBPARTS.

From-SVN: r277793

From-SVN: r277792

From-SVN: r277791

From-SVN: r277790

	* ipa-inline-transform.c: Include ipa-utils.h
	(inline_call): Set thunk_expansion flag.
	* ipa-utils.h (thunk_expansion): Declare.
	* ipa-devirt.c (thunk_expansion): New global var.
	(devirt_node_removal_hook): Do not invalidate cache while
	doing thunk expansion.

From-SVN: r277789

This initializes the rstmt variable with NULL and adds an assert to
check that a value has been given by one of the if cases before use.

This fixes the bootstrap failure due to -Werror.

Committed under the gcc obvious rule.

gcc/ChangeLog:

	* tree-vect-slp.c (vectorize_slp_instance_root_stmt): Initialize rstmt.

From-SVN: r277788

	PR testsuite/92302
	* gcc.target/sparc/sparc-ret-3.c: Accept more registers in address.

From-SVN: r277787

gcc/testsuite/ChangeLog:

	PR tree-optimization/92349
	* gcc.dg/Warray-bounds-50.c: New test.

gcc/ChangeLog:

	PR tree-optimization/92349
	* tree-vrp.c (vrp_prop::check_array_ref): Avoid assuming struct
	memebers have constant sizes.

From-SVN: r277786

gcc/ChangeLog:
2019-11-04  Andre Vieira  <andre.simoesdiasvieira@arm.com>

	* tree-vect-loop.c (vect_analyze_loop): Remove orig_loop_vinfo
	parameter.
	* tree-vectorizer.h (vect_analyze_loop): Update declaration.
	* tree-vectorizer.c (try_vectorize_loop_1): Update calls to
	vect_analyze_loop.

From-SVN: r277785

gcc/ChangeLog:
2019-11-04  Joel Hutton  <Joel.Hutton@arm.com>

	* expr.c (store_constructor): Modify to handle single element vectors.
	* tree-vect-slp.c (vect_analyze_slp_instance): Add case for vector
	constructors.
	(vect_slp_check_for_constructors): New function.
	(vect_slp_analyze_bb_1): Call new function to check for vector
	constructors.
	(vectorize_slp_instance_root_stmt): New function.
	(vect_schedule_slp): Call new function to vectorize root stmt of vector
	constructors.
	* tree-vectorizer.h (SLP_INSTANCE_ROOT_STMT): New field.

gcc/testsuite/ChangeLog:
2019-11-04  Joel Hutton  <Joel.Hutton@arm.com>

	* gcc.dg/vect/bb-slp-40.c: New test.
	* gcc.dg/vect/bb-slp-41.c: New test.

From-SVN: r277784

re PR tree-optimization/92345 (ICE in vec<_stmt_vec_info*, va_heap, vl_embed>::space (vect_get_and_check_slp_defs))

2019-11-04  Richard Biener  <rguenther@suse.de>

	PR tree-optimization/92345
	* tree-vect-loop.c (vect_is_simple_reduction): Return whether
	we produced a reduction chain.
	(vect_analyze_scalar_cycles_1): Do not add reduction chains to
	LOOP_VINFO_REDUCTIONS.

	* gcc.dg/torture/pr92345.c: New testcase.

From-SVN: r277782

	* cgraphclones.c (cgraph_node::create_version_clone): Do not
	duplicate summaries.
	* ipa-fnsummary.c (ipa_fn_summary_alloc): Allocate size summary
	first.
	(ipa_fn_summary_t::duplicate): Use get instead of get_create to
	access call summaries.
	(dump_ipa_call_summary): Be ready for missing edge summaries.
	(analyze_function_body): Use get instead of get_create to access
	edge summary.
	(estimate_calls_size_and_time): Do not access summaries of
	inlined edges; sanity check they are missing.
	(ipa_call_context::estimate_size_and_time): Use get instead
	of get_create to access node summary.
	(inline_update_callee_summaries): Do not update depth of
	inlined edge.
	(ipa_merge_fn_summary_after_inlining): Remove inline edge from
	growth caches.
	(ipa_merge_fn_summary_after_inlining): Use get instead
	of get_create.
	* ipa-fnsummary.h (ipa_remove_from_growth_caches): Declare.
	* ipa-inline-analyssi.c (edge_growth_cache): Turn to
	fast summary.
	(initialize_growth_caches): Update.
	(do_estimate_edge_time): Remove redundant copy of context.
	(ipa_remove_from_growth_caches): New function.
	* ipa-inline.c (flatten_function): Update overall summary
	only when optimizing.
	(inline_to_all_callers): Update overall summary of function
	inlined to.
	* ipa-inline.h (edge_growth_cache): Turn to fast summary.
	* symbol-summary.h (call_summary_base): Set m_initialize_when_cloning
	to false.

From-SVN: r277780

2019-11-04  Richard Biener  <rguenther@suse.de>

	* system.h: Include malloc.h if INCLUDE_MALLOC_H and HAVE_MALLINFO.
	* ggc-common.c: Remove inclusion of malloc.h, define INCLUDE_MALLOC_H.

From-SVN: r277779

From-SVN: r277778

Configuring GCC for obsolete targets works as long as build = host.
When it isn't, --enable-obsolete is not passed down to the additional
build configure started by gcc/configure, used to generate
auto-build.h.  The build configure fails and we end up without a
auto-build.h, but the host configure proceeds, so we only get a fatal
failure much later, when make realizes auto-build.h is not there and
there's no rule to create it.

This patch gets the host configure to fail when the build configure
does, leaving the temporary build configure dir behind for
investigation.  It also arranges for --eanble-obsolete to be passed
down to the build configure.

Alas, the latter triggered a warning in the build configure because
--enable-obsolete is not a recognized configure option.  That's not
reported in the host configure because of the
--disable-option-checking passed by the top-level configure, so I
arranged for that to be passed down to the build configure as well.

Finally, since my initial suspicion when investigating this failure
was that auto-build.h had been removed after configuration and there
was no rule to rebuild it, I'm adding rules to gcc/Makefile to get it
created or updated as needed.  Since it is configure that creates it,
as run by e.g. config.status --recheck, and config.status is created
after auto-build.h, I've made config.status depend on auto-build.h,
and added a dummy rule to create auto-build.h.  This would normally
not be enough to create a header when needed, but since Makefile
depends on config.status, and make first updates Makefile, it ends up
working, as long as nothing else that Makefile depends on requires
auto-build.h but not config.status.  The config.status dependency and
the auto-build.h rule are only enabled in the cases in which
auto-build.h is actually used, namely when build != host.


for  gcc/ChangeLog

	* configure.ac: Pass --enable-obsolete=* and
	--enable-option-checking=* down to build configure, and fail
	if it fails.  AC_SUBST HAVE_AUTO_BUILD.
	* configure: Rebuild.
	* Makefile.in [HAVE_AUTO_BUILD] (auto-build.h): New rule.
	[HAVE_AUTO_BUILD] (config.status): Depend on auto-build.h.

From-SVN: r277777

	* testsuite/libgomp.fortran/pr66199-1.f90: Remove
	'dg-do run' (implies torture test) as 'dg-options "O2"' is used.
	* testsuite/libgomp.fortran/pr66199-2.f90: Ditto.
	* testsuite/libgomp.fortran/taskloop2.f90: Ditto.
	* testsuite/libgomp.fortran/taskloop3.f90: Ditto.
	* testsuite/libgomp.fortran/taskloop4.f90: Ditto.

From-SVN: r277776