gcc.git
5 years agore PR debug/90074 (wrong debug info at -O3)
Richard Biener [Mon, 15 Apr 2019 12:26:11 +0000 (12:26 +0000)]
re PR debug/90074 (wrong debug info at -O3)

2019-04-15  Richard Biener  <rguenther@suse.de>

PR debug/90074
* tree-loop-distribution.c (destroy_loop): Preserve correct
debug info.

* gcc.dg/guality/pr90074.c: New testcase.

From-SVN: r270370

5 years agore PR tree-optimization/90071 (internal compiler error: SSA corruption)
Richard Biener [Mon, 15 Apr 2019 11:59:02 +0000 (11:59 +0000)]
re PR tree-optimization/90071 (internal compiler error: SSA corruption)

2019-04-15  Richard Biener  <rguenther@suse.de>

PR tree-optimization/90071
* tree-ssa-reassoc.c (init_range_entry): Do not pick up
abnormal operands from def stmts.

* gcc.dg/torture/pr90071.c: New testcase.

From-SVN: r270369

5 years agocombine: Count auto_inc properly (PR89794)
Segher Boessenkool [Mon, 15 Apr 2019 11:33:29 +0000 (13:33 +0200)]
combine: Count auto_inc properly (PR89794)

The code that checks if an auto-increment from i0 or i1 is not lost is
a bit shaky.  The code to check the same for i2 is non-existent, and
cannot be implemented in a similar way at all.  So, this patch counts
all auto-increments, and makes sure we end up with the same number as
we started with.  This works because we still have a check that we
will not duplicate any.

We should do this some better way, but not while we are in stage 4.

PR rtl-optimization/89794
* combine.c (count_auto_inc): New function.
(try_combine): Count how many auto_inc expressions there were in the
original instructions.  Ensure we have the same number in the new
instructions.  Remove the code that tried to ensure auto_inc side
effects on i1 and i0 are not lost.

gcc/testsuite/
PR rtl-optimization/89794
* gcc.dg/torture/pr89794.c: New testcase.

From-SVN: r270368

5 years agore PR ipa/88936 (-fipa-pta breaks bash (incorrect optimisation of recursive static...
Richard Biener [Mon, 15 Apr 2019 10:09:08 +0000 (10:09 +0000)]
re PR ipa/88936 (-fipa-pta breaks bash (incorrect optimisation of recursive static function))

2019-04-15  Richard Biener  <rguenther@suse.de>

PR ipa/88936
* tree.h (auto_var_p): Declare.
* tree.c (auto_var_p): New function, split out from ...
(auto_var_in_fn_p): ... here.
* tree-ssa-structalias.c (struct variable_info): Add shadow_var_uid
member.
(new_var_info): Initialize it.
(set_uids_in_ptset): Also set the shadow variable uid if required.
(ipa_pta_execute): Postprocess points-to solutions assigning
shadow variable uids for locals that may reach their containing
function recursively.
* tree-ssa-ccp.c (fold_builtin_alloca_with_align): Do not
assert but instead check whether the points-to solution is
a singleton.

* gcc.dg/torture/pr88936-1.c: New testcase.
* gcc.dg/torture/pr88936-2.c: Likewise.
* gcc.dg/torture/pr88936-3.c: Likewise.

From-SVN: r270366

5 years agoReorganize cgraph_node::clone_of_p
Martin Jambor [Mon, 15 Apr 2019 08:30:36 +0000 (10:30 +0200)]
Reorganize cgraph_node::clone_of_p

2019-04-15  Martin Jambor  <mjambor@suse.cz>

PR ipa/pr89693
* cgraph.c (clone_of_p): Loop over clone chain for each step in
the thunk chain.

testsuite/
* g++.dg/ipa/pr89693.C: New test.

From-SVN: r270364

5 years ago[NDS32] Refine force unwind. Linux kernel only uses RT_SIGRETURN.
Monk Chiang [Mon, 15 Apr 2019 08:19:23 +0000 (08:19 +0000)]
[NDS32] Refine force unwind. Linux kernel only uses RT_SIGRETURN.

libgcc/
* config/nds32/linux-unwind.h (SIGRETURN): Remove.
(RT_SIGRETURN): Update.
(nds32_fallback_frame_state): Update.

From-SVN: r270363

5 years ago[NDS32] nds32*-*-linux* target using init_array/finit_array for ctor/dtor.
Monk Chiang [Mon, 15 Apr 2019 08:05:03 +0000 (08:05 +0000)]
[NDS32] nds32*-*-linux* target using init_array/finit_array for ctor/dtor.

gcc/
* config.gcc (nds32*-*-linux*): Set gcc_cv_initfini_array to yes.

From-SVN: r270362

5 years ago[NDS32] Rewrite PIC/TLS patterns.
Monk Chiang [Mon, 15 Apr 2019 07:59:01 +0000 (07:59 +0000)]
[NDS32] Rewrite PIC/TLS patterns.

gcc/
* config/nds32/nds32-md-auxiliary.c
(nds32_legitimize_pic_address): Use new PIC pattern.
(nds32_legitimize_tls_address): Use new TLS pattern.
(nds32_output_symrel): New.
* config/nds32/nds32-protos.h (nds32_output_symrel): Declare.
(nds32_alloc_relax_group_id): Ditto.
* config/nds32/nds32-relax-opt.c (nds32_alloc_relax_group_id): New.
(nds32_group_insns): Use nds32_alloc_relax_group_id instead of use
relax_group_id.
(nds32_group_tls_insn): Ditto.
(nds32_group_float_insns): Ditto.
* config/nds32/nds32.md (tls_le): New.
(sym_got): Ditto.

Co-Authored-By: Kito Cheng <kito.cheng@gmail.com>
Co-Authored-By: Shiva Chen <shiva0217@gmail.com>
From-SVN: r270361

5 years agore PR tree-optimization/90020 (-O2 -Os x86-64 wrong code generated for GNU Emacs)
Dominique d'Humieres [Mon, 15 Apr 2019 07:56:43 +0000 (09:56 +0200)]
re PR tree-optimization/90020 (-O2 -Os x86-64 wrong code generated for GNU Emacs)

2019-04-15  Dominique d'Humieres  <dominiq@gcc.gnu.org>

PR tree-optimization/90020
* gcc.dg/torture/pr90020.c: Add linker options for darwin.

From-SVN: r270360

5 years ago[NDS32] Provide one valid nds32 assembly instruction to check assembler debugging...
Chung-Ju Wu [Mon, 15 Apr 2019 07:39:20 +0000 (07:39 +0000)]
[NDS32] Provide one valid nds32 assembly instruction to check assembler debugging options and features.

gcc/
* configure: Add nds32 target for dwarf2 debug_line checking.
* configure.ac: Regenerated.

From-SVN: r270359

5 years agoDaily bump.
GCC Administrator [Mon, 15 Apr 2019 00:16:17 +0000 (00:16 +0000)]
Daily bump.

From-SVN: r270358

5 years agore PR lto/89358 (Combining -std=c++14 and -std=c++17 objects gives ODR warnings)
Jan Hubicka [Sun, 14 Apr 2019 20:56:45 +0000 (22:56 +0200)]
re PR lto/89358 (Combining -std=c++14 and -std=c++17 objects gives ODR warnings)

PR lto/89358
* g++.dg/lto/pr89358_0.C: New testcase.
* g++.dg/lto/pr89358_1.C: New testcase.
* ipa-devirt.c (skip_in_fields_list_p): New.
(odr_types_equivalent_p): Use it.

From-SVN: r270355

5 years agore PR fortran/85448 (Report binding label clash with a global identifyer)
Thomas Koenig [Sun, 14 Apr 2019 20:15:48 +0000 (20:15 +0000)]
re PR fortran/85448 (Report binding label clash with a global identifyer)

2019-04-14  Thomas Koenig  <tkoenig@gcc.gnu.org>

    PR fortran/85448
    * gfortran.dg/bind_c_usage_33.f90: New test and...
    * gfortran.dg/bind_c_usage_33_c.c: Additional source.

From-SVN: r270354

5 years agore PR fortran/89843 (CFI_section delivers incorrect result descriptor)
Paul Thomas [Sun, 14 Apr 2019 18:14:58 +0000 (18:14 +0000)]
re PR fortran/89843 (CFI_section delivers incorrect result descriptor)

2019-04-14  Paul Thomas  <pault@gcc.gnu.org>

PR fortran/89843
* trans-decl.c (gfc_get_symbol_decl): Assumed shape and assumed
rank dummies of bind C procs require deferred initialization.
(convert_CFI_desc): New procedure to convert incoming CFI
descriptors to gfc types and back again.
(gfc_trans_deferred_vars): Call it.
* trans-expr.c (gfc_conv_gfc_desc_to_cfi_desc): Null the CFI
descriptor pointer. Free the descriptor in all cases.

PR fortran/89846
* expr.c (is_CFI_desc): New function.
(is_subref_array): Tidy up by referencing the symbol directly.
* gfortran.h : Prototype for is_CFI_desc.
* trans_array.c (get_CFI_desc): New function.
(gfc_get_array_span, gfc_conv_scalarized_array_ref,
gfc_conv_array_ref): Use it.
* trans.c (get_array_span): Extract the span from descriptors
that are indirect references.

PR fortran/90022
* trans-decl.c (gfc_get_symbol_decl): Make sure that the se
expression is a pointer type before converting it to the symbol
backend_decl type.
* trans-expr.c (gfc_conv_gfc_desc_to_cfi_desc): Eliminate
temporary creation for intent(in).

2019-04-14  Paul Thomas  <pault@gcc.gnu.org>

PR fortran/89843
* gfortran.dg/ISO_Fortran_binding_4.f90: Modify the value of x
in ctg. Test the conversion of the descriptor types in the main
program.
* gfortran.dg/ISO_Fortran_binding_10.f90: New test.
* gfortran.dg/ISO_Fortran_binding_10.c: Called by it.

PR fortran/89846
* gfortran.dg/ISO_Fortran_binding_11.f90: New test.
* gfortran.dg/ISO_Fortran_binding_11.c: Called by it.

PR fortran/90022
* gfortran.dg/ISO_Fortran_binding_1.c: Correct the indexing for
the computation of 'ans'. Also, change the expected results for
CFI_is_contiguous to comply with standard.
* gfortran.dg/ISO_Fortran_binding_1.f90: Correct the expected
results for CFI_is_contiguous to comply with standard.
* gfortran.dg/ISO_Fortran_binding_9.f90: New test.
* gfortran.dg/ISO_Fortran_binding_9.c: Called by it.

2019-04-14  Paul Thomas  <pault@gcc.gnu.org>

PR fortran/89843
* runtime/ISO_Fortran_binding.c (cfi_desc_to_gfc_desc): Only
return immediately if the source pointer is null. Bring
forward the extraction of the gfc type. Extract the kind so
that the element size can be correctly computed for sections
and components of derived type arrays. Remove the free of the
CFI descriptor since this is now done in trans-expr.c.
(gfc_desc_to_cfi_desc): Only allocate the CFI descriptor if it
is not null.
(CFI_section): Normalise the difference between the upper and
lower bounds by the stride to correctly calculate the extents
of the section.

PR fortran/89846
* runtime/ISO_Fortran_binding.c (cfi_desc_to_gfc_desc): Use
the stride measure for the gfc span if it is not a multiple
of the element length. Otherwise use the element length.

PR fortran/90022
* runtime/ISO_Fortran_binding.c (CFI_is_contiguous) : Return
1 for true and 0 otherwise to comply with the standard. Correct
the contiguity check for rank 3 and greater by using the stride
measure of the lower dimension rather than the element length.

From-SVN: r270353

5 years agoPR d/87799 Fix D build on windows hosts
Johannes Pfau [Sun, 14 Apr 2019 09:52:22 +0000 (09:52 +0000)]
PR d/87799 Fix D build on windows hosts

PR d/87799
* d-system.h (_mkdir): Forward _mkdir on MinGW to mkdir in system.h.
* configure.ac: Remove d from unsupported languages on mingw and cygwin.
* configure: Regenerate.

From-SVN: r270349

5 years agoEnable libphobos on Solaris 11/x86
Rainer Orth [Sun, 14 Apr 2019 09:38:01 +0000 (09:38 +0000)]
Enable libphobos on Solaris 11/x86

toplevel:
* configure.ac (enable_libphobos): Check LIBPHOBOS_SUPPORTED.
* configure: Regenerate.

libphobos:
* configure.tgt (LIBPHOBOS_SUPPORTED): Default to no.
Set to yes explicitly.
* configure.ac: Handle --enable-libphobos.
(x86_64-*-solaris2.* | i?86-*-solaris2.*): Only mark supported
with gas.
(ENABLE_LIBPHOBOS): New conditional.
* configure: Regenerate.
* Makefile.am (SUBDIRS): Only set if ENABLE_LIBPHOBOS.
* Makefile.in: Regenerate.

From-SVN: r270348

5 years agoWork around lack of dlpi_tls_modid before Solaris 11.5
Rainer Orth [Sun, 14 Apr 2019 09:30:42 +0000 (09:30 +0000)]
Work around lack of dlpi_tls_modid before Solaris 11.5

2019-04-14  Rainer Orth  <ro@CeBiTec.Uni-Bielefeld.DE>
    Iain Buclaw  <ibuclaw@gdcproject.org>

PR d/88150
* m4/druntime/os.m4 (DRUNTIME_OS_DLPI_TLS_MODID): New macro.
* configure.ac: Use it.
Call AC_USE_SYSTEM_EXTENSIONS.
* configure: Regenerate.
* Makefile.in, libdruntime/Makefile.in, src/Makefile.in,
testsuite/Makefile.in: Regenerate.
* libdruntime/gcc/config.d.in (OS_Have_Dlpi_Tls_Modid): Define.
* libdruntime/gcc/sections/elf_shared.d: Import gcc.config.
(scanSegments) <PT_TLS> [OS_Have_Dlpi_Tls_Modid]: Use
dlpi_tls_modid.
[Solaris]: Use dlinfo(RTLD_DI_LINKMAP) to get rt_tlsmodid.
Otherwise clear pdso._tlsMod, pdso._tlsSize.
(getTLSRange) [Solaris && !OS_Have_Dlpi_Tls_Modid]: Readjust mod.

Co-Authored-By: Iain Buclaw <ibuclaw@gdcproject.org>
From-SVN: r270347

5 years agoWork around Solaris ld bug linking __tls_get_addr on 64-bit x86
Rainer Orth [Sun, 14 Apr 2019 09:24:26 +0000 (09:24 +0000)]
Work around Solaris ld bug linking __tls_get_addr on 64-bit x86

* m4/druntime/os.m4 (DRUNTIME_OS_LINK_SPEC): New macro.
* configure.ac: Call it.
* configure: Regenerate.
* Makefile.in, libdruntime/Makefile.in, src/Makefile.in,
testsuite/Makefile.in: Regenerate.
* src/libgphobos.spec.in (*link): Append OS_LINK_SPEC.

From-SVN: r270346

5 years agoUse gcc/sections/elf_shared.d on Solaris 11.5 (PR d/88150)
Rainer Orth [Sun, 14 Apr 2019 09:18:42 +0000 (09:18 +0000)]
Use gcc/sections/elf_shared.d on Solaris 11.5 (PR d/88150)

PR d/88150
* libdruntime/gcc/sections/elf_shared.d [Solaris] (SharedELF): Set
to true.
Import core.sys.solaris.dlfcn, core.sys.solaris.link,
core.sys.solaris.sys.elf, core.sys.solaris.sys.link.
(dummy_ref): Declare.
(initSections): Initialize dummy_ref.
(getDependencies): Set strtab.
(handleForName): Don't dlclose handle.
(findDSOInfoForAddr): Set IterateManually.
(getprogname): Declare.
(progname): Use it.
* libdruntime/gcc/sections/package.d [Solaris]: Import
gcc.sections.elf_shared instead of gcc.sections.solaris.
* libdruntime/gcc/sections/solaris.d: Remove.
* libdruntime/Makefile.am (DRUNTIME_DSOURCES): Remove
gcc/sections/solaris.d.

From-SVN: r270345

5 years agoDaily bump.
GCC Administrator [Sun, 14 Apr 2019 00:16:15 +0000 (00:16 +0000)]
Daily bump.

From-SVN: r270344

5 years agolibphobos: Move rt.sections modules to gcc.sections
Iain Buclaw [Sat, 13 Apr 2019 15:29:15 +0000 (15:29 +0000)]
libphobos: Move rt.sections modules to gcc.sections

These modules depend on a mixture between how the compiler emits
run-time module information, and what functions are exposed by the
platform to inquire about loaded global and thread-local data sections.

As the upstream implementation is written to work only with how the
reference D compiler writes out data, much of what is present does not
apply to the GCC D front-end.  So it has been moved to a non-upstream
location in the source tree, where most of it will be rewritten once
each port has been completed.

The only tested module sections/elf_shared.d has been cleaned up so that
all deprecated declarations have been removed, as well as the brittle
module collision checking, which required bss_sections.c.  All other
ports have been left unchanged apart from a commonizing of attributes.

libphobos/ChangeLog:

2019-04-13  Iain Buclaw  <ibuclaw@gdcproject.org>

* libdruntime/Makefile.am (DRUNTIME_CSOURCES): Remove bss_sections.c.
(DRUNTIME_DSOURCES): Rename rt/sections_* modules to gcc/sections/*.
* libdruntime/Makefile.in: Regenerate.
* libdruntime/gcc/sections/android.d: New file.
* libdruntime/gcc/sections/elf_shared.d: New file.
* libdruntime/gcc/sections/osx.d: New file.
* libdruntime/gcc/sections/package.d: New file.
* libdruntime/gcc/sections/solaris.d: New file.
* libdruntime/gcc/sections/win32.d: New file.
* libdruntime/gcc/sections/win64.d: New file.
* libdruntime/rt/bss_section.c: Remove.
* libdruntime/rt/sections.d: Publicly import gcc.sections.
* libdruntime/rt/sections_android.d: Remove.
* libdruntime/rt/sections_elf_shared.d: Remove.
* libdruntime/rt/sections_osx.d: Remove.
* libdruntime/rt/sections_solaris.d: Remove.
* libdruntime/rt/sections_win32.d: Remove.
* libdruntime/rt/sections_win64.d: Remove.

From-SVN: r270341

5 years ago* gcc.pot: Regenerate.
Jakub Jelinek [Sat, 13 Apr 2019 15:22:23 +0000 (17:22 +0200)]
* gcc.pot: Regenerate.

From-SVN: r270340

5 years agore PR target/89093 (C++ exception handling clobbers d8 VFP register)
Jakub Jelinek [Sat, 13 Apr 2019 15:20:46 +0000 (17:20 +0200)]
re PR target/89093 (C++ exception handling clobbers d8 VFP register)

PR target/89093
* config/arm/arm.c (arm_valid_target_attribute_rec): Use strcmp
instead of strncmp when checking for thumb and arm.  Formatting fixes.

* gcc.target/arm/pr89093.c: New test.

From-SVN: r270339

5 years agore PR fortran/79842 (i18n: subword translation in "Can't use the same %smodule")
Dominique d'Humieres [Sat, 13 Apr 2019 11:32:49 +0000 (13:32 +0200)]
re PR fortran/79842 (i18n: subword translation in "Can't use the same %smodule")

2019-04-13  Dominique d'Humieres  <dominiq@gcc.gnu.org>

PR fortran/79842
* module.c (gfc_use_module): use complete sentences.

From-SVN: r270338

5 years ago* gcc.pot: Regenerate.
Jakub Jelinek [Sat, 13 Apr 2019 08:16:59 +0000 (10:16 +0200)]
* gcc.pot: Regenerate.

From-SVN: r270337

5 years agocompiler: improve type handling for string concat ops on constants
Ian Lance Taylor [Sat, 13 Apr 2019 01:03:55 +0000 (01:03 +0000)]
compiler: improve type handling for string concat ops on constants

    Resolve a small problem with concatenation of string constants: in a
    string concat X + Y where X has named type and Y has abstract string
    type, insure that the result has X's type, and disable folding if the
    both sides have a concrete type that does not match.

    Fixes golang/go#31412.

    Reviewed-on: https://go-review.googlesource.com/c/gofrontend/+/171797

From-SVN: r270336

5 years agoDaily bump.
GCC Administrator [Sat, 13 Apr 2019 00:16:34 +0000 (00:16 +0000)]
Daily bump.

From-SVN: r270335

5 years agoUglify identifiers missed in previous commit(s)
Thomas Rodgers [Sat, 13 Apr 2019 00:13:46 +0000 (00:13 +0000)]
Uglify identifiers missed in previous commit(s)

       * include/pstl/algorithm_impl.h: Uglify identfiers.
       * include/pstl/numeric_impl.h:  Uglify identfiers.
       * include/pstl/parallel_backend_tbb.h: Uglify identfiers.

From-SVN: r270332

5 years agoCommit a change missed in r270326:
Martin Sebor [Fri, 12 Apr 2019 22:37:12 +0000 (16:37 -0600)]
Commit a change missed in r270326:

gcc/c-family/ChangeLog:

PR c/88383
PR c/89288
PR c/89798
PR c/89797
* c-attribs.c (type_valid_for_vector_size): Detect excessively
large sizes.
(validate_attribute): Handle DECLs and expressions.
(has_attribute): Handle types referenced by expressions.
Avoid considering array attributes in ARRAY_REF expressions .

From-SVN: r270331

5 years agolibphobos: Add subdir-objects to configure script.
Iain Buclaw [Fri, 12 Apr 2019 21:50:42 +0000 (21:50 +0000)]
libphobos: Add subdir-objects to configure script.

Running autoreconf regenerates all files cleanly with autoconf2.69.

libphobos/ChangeLog:

2019-04-12  Iain Buclaw  <ibuclaw@gdcproject.org>

* configure.ac (AM_INIT_AUTOMAKE): Add subdir-objects.
* configure: Regenerate.
* libdruntime/Makefile.in: Regenerate.

From-SVN: r270330

5 years agore PR c/89933 (ICE in merge_decls, at c/c-decl.c:2517)
Jakub Jelinek [Fri, 12 Apr 2019 21:45:54 +0000 (23:45 +0200)]
re PR c/89933 (ICE in merge_decls, at c/c-decl.c:2517)

PR c/89933
c/
* c-decl.c (merge_decls): When newdecl's type is its main variant,
don't try to remove it from the variant list, but instead assert
it has no variants.
cp/
* decl.c (duplicate_decls): When newdecl's type is its main variant,
don't try to remove it from the variant list, but instead assert
it has no variants.
testsuite/
* c-c++-common/pr89933.c: New test.

From-SVN: r270329

5 years agolibphobos: Add target-zlib to top-level configure
Iain Buclaw [Fri, 12 Apr 2019 20:13:34 +0000 (20:13 +0000)]
libphobos: Add target-zlib to top-level configure

Removes the building of zlib from within libphobos, using instead the
libz_convenience.a library.

ChangeLog:

2019-04-12  Iain Buclaw  <ibuclaw@gdcproject.org>

* configure.ac: Add target-zlib to target_libraries.
* configure: Regenerate.

gcc/ChangeLog:

2019-04-12  Iain Buclaw  <ibuclaw@gdcproject.org>

* doc/install.texi: Document --with-target-system-zlib.

libphobos/ChangeLog:

2019-04-12  Iain Buclaw  <ibuclaw@gdcproject.org>

* m4/druntime/libraries.m4 (DRUNTIME_LIBRARIES_ZLIB): Use
libz_convenience.a if not using system zlib.
* Makefile.in: Regenerate.
* configure: Regenerate.
* libdruntime/Makefile.in: Regenerate.
* src/Makefile.am: Remove ZLIB_CSOURCES and AM_CFLAGS.
* src/Makefile.in: Regenerate.
* testsuite/Makefile.in: Regenerate.

From-SVN: r270328

5 years agoextend.texi (__builtin_has_attribute): Add a clarifying note.
Martin Sebor [Fri, 12 Apr 2019 19:11:59 +0000 (13:11 -0600)]
extend.texi (__builtin_has_attribute): Add a clarifying note.

gcc/ChangeLog:
* doc/extend.texi (__builtin_has_attribute): Add a clarifying note.

From-SVN: r270327

5 years agoPR c/88383 - ICE calling __builtin_has_attribute on a reference
Martin Sebor [Fri, 12 Apr 2019 19:01:17 +0000 (19:01 +0000)]
PR c/88383 - ICE calling __builtin_has_attribute on a reference

PR c/88383 - ICE calling __builtin_has_attribute on a reference
PR c/89288 - ICE in tree_code_size, at tree.c:865
PR c/89798 - excessive vector_size silently accepted and truncated
PR c/89797 - ICE on a vector_size (1LU << 33) int variable

gcc/ChangeLog:

PR c/89797
* targhooks.c (default_vector_alignment): Avoid assuming
argument fits in SHWI.
* tree.h (TYPE_VECTOR_SUBPARTS): Avoid sign overflow in
a shift expression.
* doc/extend.texi (__builtin_has_attribute): Add a clarifying note.

gcc/c-family/ChangeLog:

PR c/88383
PR c/89288
PR c/89798
PR c/89797
* c-attribs.c (type_valid_for_vector_size): Detect excessively
large sizes.
(validate_attribute): Handle DECLs and expressions.
(has_attribute): Handle types referenced by expressions.
Avoid considering array attributes in ARRAY_REF expressions .

gcc/cp/ChangeLog:

PR c/88383
PR c/89288
* parser.c (cp_parser_has_attribute_expression): Handle assignment
expressions.

gcc/testsuite/ChangeLog:

PR c/88383
PR c/89288
PR c/89798
PR c/89797
* c-c++-common/attributes-1.c: Adjust.
* c-c++-common/builtin-has-attribute-6.c: New test.
* c-c++-common/builtin-has-attribute-7.c: New test.
* c-c++-common/builtin-has-attribute-4.c: Adjust expectations.
* c-c++-common/builtin-has-attribute-6.c: New test.
* c-c++-common/pr71574.c: Adjust.
* gcc.dg/pr25559.c: Adjust.
* gcc.dg/attr-vector_size.c: New test.

From-SVN: r270326

5 years agoAvoid ICE on pmf{} in template.
Jason Merrill [Fri, 12 Apr 2019 16:25:59 +0000 (12:25 -0400)]
Avoid ICE on pmf{} in template.

Now that we return the original CONSTRUCTOR from finish_compound_literal,
the call to null_member_pointer_value_p in tsubst_copy_and_build was getting
confused because the CONSTRUCTOR was still empty rather than a valid PMF
value.

* call.c (null_member_pointer_value_p): Handle an empty CONSTRUCTOR
of PMF type.

From-SVN: r270324

5 years agore PR rtl-optimization/89965 (wrong code with -O -mtune=nano-x2 -fcaller-saves -fexpe...
Jakub Jelinek [Fri, 12 Apr 2019 16:20:21 +0000 (18:20 +0200)]
re PR rtl-optimization/89965 (wrong code with -O -mtune=nano-x2 -fcaller-saves -fexpensive-optimizations -fno-tree-dce -fno-tree-ter)

PR rtl-optimization/89965
* dce.c: Include rtl-iter.h.
(struct check_argument_load_data): New type.
(check_argument_load): New function.
(find_call_stack_args): Check for loads from stack slots still tracked
in sp_bytes and punt if any is found.

* gcc.target/i386/pr89965.c: New test.

From-SVN: r270323

5 years ago* config/mips/loongson-mmiintrin.h: Fix up #error message.
Jakub Jelinek [Fri, 12 Apr 2019 16:16:35 +0000 (18:16 +0200)]
* config/mips/loongson-mmiintrin.h: Fix up #error message.

From-SVN: r270322

5 years agore PR translation/90041 (Command line option without proper quoting in translation...
Jakub Jelinek [Fri, 12 Apr 2019 16:12:49 +0000 (18:12 +0200)]
re PR translation/90041 (Command line option without proper quoting in translation message)

PR translation/90041
* exgettext: Print MissingArgError, UnknownError or Warn
*.opt argument using error or warning instead of _ to mark it
as gcc-internal-format.

* c.opt (-fhandle-exceptions): Use %< and %> around option names
in the Warn diagnostics.

From-SVN: r270321

5 years agoPR c++/87603 - constexpr functions are no longer noexcept.
Marek Polacek [Fri, 12 Apr 2019 15:29:03 +0000 (15:29 +0000)]
PR c++/87603 - constexpr functions are no longer noexcept.

* constexpr.c (is_sub_constant_expr): Remove unused function.
* cp-tree.h (is_sub_constant_expr): Remove declaration.
* except.c (check_noexcept_r): Don't consider a call to a constexpr
function noexcept.

* g++.dg/cpp0x/constexpr-noexcept.C: Adjust the expected result.
* g++.dg/cpp0x/constexpr-noexcept3.C: Likewise.
* g++.dg/cpp0x/constexpr-noexcept4.C: Likewise.
* g++.dg/cpp0x/constexpr-noexcept8.C: New test.
* g++.dg/cpp0x/inh-ctor32.C: Remove dg-message.
* g++.dg/cpp1y/constexpr-noexcept1.C: New test.

From-SVN: r270320

5 years agoexcept.c (build_noexcept_spec): Use build_converted_constant_bool_expr instead of...
Marek Polacek [Fri, 12 Apr 2019 15:26:16 +0000 (15:26 +0000)]
except.c (build_noexcept_spec): Use build_converted_constant_bool_expr instead of perform_implicit_conversion_flags.

* except.c (build_noexcept_spec): Use build_converted_constant_bool_expr
instead of perform_implicit_conversion_flags.

* g++.dg/cpp0x/noexcept30.C: Tweak dg-error.
* g++.dg/cpp0x/pr86397-1.C: Likewise.
* g++.dg/cpp0x/pr86397-2.C: Likewise.

From-SVN: r270319

5 years agoinvoke.texi (max-lto-streaming-paralellism): New --param.
Jan Hubicka [Fri, 12 Apr 2019 14:36:26 +0000 (14:36 +0000)]
invoke.texi (max-lto-streaming-paralellism): New --param.

* doc/invoke.texi (max-lto-streaming-paralellism): New --param.

From-SVN: r270318

5 years agoparams.def (PARAM_MAX_LTO_STREAMING_PARALLELISM): New parameter.
Jan Hubicka [Fri, 12 Apr 2019 14:35:48 +0000 (16:35 +0200)]
params.def (PARAM_MAX_LTO_STREAMING_PARALLELISM): New parameter.

* params.def (PARAM_MAX_LTO_STREAMING_PARALLELISM): New parameter.
* lto.c (do_stream_out): rename to ...
(stream_out): ... this one; move original code to ...
(stream_out_partitions_1, stream_out_partitions): ... these new
functions.
(lto_wpa_write_files): Honnor lto_parallelism

From-SVN: r270317

5 years ago[testsuite][arm] Update warning prune regex
Matthew Malcomson [Fri, 12 Apr 2019 13:15:52 +0000 (13:15 +0000)]
[testsuite][arm] Update warning prune regex

r269586 changed the format of some warning messages.

Each switch in the warning message is now surrounded by single quotes.

This commit updates the regex's in arm.exp dejagnu files to match the
new format and remove the extra 20+ FAILs on excess errors that are
introduced on certain variations (e.g.
arm-eabi-aem/-marm/-march=armv7-a/-mfpu=vfpv3-d16/-mfloat-abi=softfp).

Regtested arm.exp with cross-compiler arm-none-eabi

gcc/testsuite/ChangeLog:

2019-04-12  Matthew Malcomson  <matthew.malcomson@arm.com>

* g++.target/arm/arm.exp: Change format of default prune regex.
* gcc.target/arm/arm.exp: Change format of default prune regex.

From-SVN: r270316

5 years agoHandle multiple 'default' in target attribute (PR middle-end/89970).
Martin Liska [Fri, 12 Apr 2019 12:54:00 +0000 (14:54 +0200)]
Handle multiple 'default' in target attribute (PR middle-end/89970).

2019-04-12  Martin Liska  <mliska@suse.cz>

PR middle-end/89970
* multiple_target.c (create_dispatcher_calls): Wrap ifunc
in error message.
(separate_attrs): Handle multiple 'default's.
(expand_target_clones): Rework error handling code.
2019-04-12  Martin Liska  <mliska@suse.cz>

PR middle-end/89970
* gcc.target/i386/mvc15.c: New test.
* gcc.target/i386/mvc3.c: Quote target in error pattern.
* gcc.target/i386/mvc4.c: Remove duplicit 'default'.

From-SVN: r270314

5 years agore PR target/87532 (bad results from vec_extract(unsigned char, foo) dependent upon...
Kelvin Nilsen [Fri, 12 Apr 2019 12:51:58 +0000 (12:51 +0000)]
re PR target/87532 (bad results from vec_extract(unsigned char, foo) dependent upon function inline)

gcc/ChangeLog:

2019-04-12  Kelvin Nilsen  <kelvin@gcc.gnu.org>

PR target/87532
* config/rs6000/rs6000.c (rs6000_split_vec_extract_var): Use inner
mode of vector rather than mode of destination for move instruction.
* config/rs6000/vsx.md (*vsx_extract_<mode>_<VS_scalar>mode_var):
Use QI inner mode with V16QI vector mode.

gcc/testsuite/ChangeLog:

2019-04-12  Kelvin Nilsen  <kelvin@gcc.gnu.org>

PR target/87532
* gcc.target/powerpc/fold-vec-extract-char.p8.c: Adjust expected
instruction counts.
* gcc.target/powerpc/fold-vec-extract-int.p8.c: Likewise.
* gcc.target/powerpc/fold-vec-extract-short.p8.c: Likewise.

From-SVN: r270313

5 years ago* de.po, sv.po: Update.
Jakub Jelinek [Fri, 12 Apr 2019 09:07:37 +0000 (11:07 +0200)]
* de.po, sv.po: Update.

From-SVN: r270309

5 years agore PR target/52726 (Composed error message will not get translated)
Jakub Jelinek [Fri, 12 Apr 2019 08:10:51 +0000 (10:10 +0200)]
re PR target/52726 (Composed error message will not get translated)

PR target/52726
* config/tilepro/tilepro.c (tilepro_print_operand): Use just
"invalid %%t operand" in output_operand_lossage message.

From-SVN: r270308

5 years agoRemove spurious empty line from ChangeLog
Jonathan Wakely [Fri, 12 Apr 2019 08:04:56 +0000 (09:04 +0100)]
Remove spurious empty line from ChangeLog

From-SVN: r270307

5 years agoS/390: Fix a problem with the bswap vector pattern
Andreas Krebbel [Fri, 12 Apr 2019 07:41:51 +0000 (07:41 +0000)]
S/390: Fix a problem with the bswap vector pattern

arch13 introduced instructions to perform vector element-wise byte
swaps on the way from or to memory.  For a byte swap between vector
registers the vector permute instruction is required which needs a
permute pattern to be loaded into a vector register first.

With the current implementation there is a potential problem when the
decision for the reg-reg variant is made very late.  This patch is
supposed to fix that.

With the patch the required permute pattern is generated already in
the expander and attached to the bswap pattern as USE operand.  The
predicate in the insn_and_split pattern accepts it although the
permute constant as such is not a valid constant.  For the reg-reg
variant only the vector register constraint is used for the permute
constant forcing LRA to a) push the constant into literal pool and b)
load the literal pool constant into a vector register.

gcc/ChangeLog:

2019-04-12  Andreas Krebbel  <krebbel@linux.ibm.com>

* config/s390/predicates.md (permute_pattern_operand): New
predicate.
* config/s390/vector.md ("*vec_splats_bswap_vec<mode>"): Add USE
operand for the permute pattern.
("*vec_perm<mode>"): New insn definition.
("bswap<mode>"): Generate the permute pattern operand in the
expander and perform the operand reloads for pre arch13 level
already.
("*bswap<mode>_emu"): Rename to ...
("*bswap<mode>"): ... this. And make the splitter vxe2 only.
* config/s390/vx-builtins.md ("*vec_insert_and_zero_bswap<mode>"):
Add the USE operand for the permute pattern.
("*vec_set_bswap_vec<mode>"): Likewise.

From-SVN: r270306

5 years agore PR c/89946 (ICE in assemble_start_function, at varasm.c:1871)
Jakub Jelinek [Fri, 12 Apr 2019 07:28:35 +0000 (09:28 +0200)]
re PR c/89946 (ICE in assemble_start_function, at varasm.c:1871)

PR c/89946
* varasm.c (assemble_start_function): Don't use tree_fits_uhwi_p
and gcc_unreachable if it fails, just call tree_to_uhwi which
verifies that too.  Test TREE_CHAIN instead of list_length > 1.
Start warning message with a lower-case letter.  Formatting fixes.
c-family/
* c-attribs.c (handle_patchable_function_entry_attribute): Add
function comment.  Warn if arguments of the attribute are not positive
integer constants.
testsuite/
* c-c++-common/pr89946.c: New test.

From-SVN: r270305

5 years agore PR rtl-optimization/90026 (ICE: verify_flow_info failed (error: missing barrier...
Jakub Jelinek [Fri, 12 Apr 2019 07:27:25 +0000 (09:27 +0200)]
re PR rtl-optimization/90026 (ICE: verify_flow_info failed (error: missing barrier after block 2))

PR rtl-optimization/90026
* cfgcleanup.c (try_optimize_cfg): When removing empty bb with no
successors, look for BARRIERs inside of the whole BB_FOOTER chain
rather than just at the start of it.  If e->src BB_FOOTER is not NULL
in cfglayout mode, use emit_barrier_after_bb.

* g++.dg/opt/pr90026.C: New test.

From-SVN: r270304

5 years agolibphobos: Parallelize the libphobos testsuite
Iain Buclaw [Fri, 12 Apr 2019 06:25:34 +0000 (06:25 +0000)]
libphobos: Parallelize the libphobos testsuite

The removal of the check programs in libphobos means that all tests
will now be compiled one at a time, causing a notable slowdown in
comparison with the previous method of testing.

libphobos/ChangeLog:

2019-04-12  Iain Buclaw  <ibuclaw@gdcproject.org>

* testsuite/Makefile.am:
* testsuite/Makefile.am (AUTOMAKE_OPTIONS): Remove dejagnu.
(RUNTEST): Remove variable.
(RUNTESTDEFAULTFLAGS, check_p_subno, check_p_numbers0,
check_p_numbers1, check_p_numbers2, check_p_numbers3,
check_p_numbers4, check_p_numbers5, check_p_numbers6, check_p_numbers,
check_p_subdirs, check_DEJAGNU_libphobos_targets): New variables.
(site.exp, %/site.exp, check-DEJAGNU, check-am, clean-local): New
rules written so that all the *.exp files are ran parallelized.
(CLEANFILES): Add *.exe, *.o, and site.exp.
* testsuite/Makefile.in: Regenerate.

From-SVN: r270303

5 years agolibphobos: Replace library check programs with dg-runtest
Iain Buclaw [Fri, 12 Apr 2019 06:25:17 +0000 (06:25 +0000)]
libphobos: Replace library check programs with dg-runtest

Fixes multilib handling when running the libphobos testsuite.

libphobos/ChangeLog:

2019-04-12  Iain Buclaw  <ibuclaw@gdcproject.org>

PR d/89255
* configure.ac (ENABLE_SHARED, ENABLE_STATIC): Remove conditionals.
* configure: Regenerate.
* d_rules.am (%.t.lo, %.t.o): Remove rules.
(unittest_static_LINK, unittest_LINK, libgdruntime_t_la_LINK,
libgphobos_t_la_LINK): Remove variables.
* libdruntime/Makefile.am (DRUNTIME_TEST_LOBJECTS,
DRUNTIME_TEST_OBJECTS, check_PROGRAMS, check_LTLIBRARIES,
unittest_static_SOURCES, unittest_static_LIBTOOLFLAGS,
unittest_static_LDFLAGS, unittest_static_LDADD,
EXTRA_unittest_static_DEPENDENCIES, libgdruntime_t_la_SOURCES,
libgdruntime_t_la_LIBTOOLFLAGS, libgdruntime_t_la_LDFLAGS,
libgdruntime_t_la_LIBADD, EXTRA_libgdruntime_t_la_DEPENDENCIES,
unittest_SOURCES, unittest_LIBTOOLFLAGS, unittest_LDFLAGS,
unittest_LDADD): Remove variables.
(clean-local): Remove rule.
* libdruntime/Makefile.in: Regenerate.
* src/Makefile.am (PHOBOS_TEST_LOBJECTS,
PHOBOS_TEST_OBJECTS, check_PROGRAMS, check_LTLIBRARIES,
unittest_static_SOURCES, unittest_static_LIBTOOLFLAGS,
unittest_static_LDFLAGS, unittest_static_LDADD,
EXTRA_unittest_static_DEPENDENCIES, libgdruntime_t_la_SOURCES,
libgdruntime_t_la_LIBTOOLFLAGS, libgdruntime_t_la_LDFLAGS,
libgdruntime_t_la_LIBADD, EXTRA_libgdruntime_t_la_DEPENDENCIES,
unittest_SOURCES, unittest_LIBTOOLFLAGS, unittest_LDFLAGS,
unittest_LDADD): Remove variables.
(clean-local): Remove rule.
* src/Makefile.in: Regenerate.
* testsuite/lib/libphobos.exp (libphobos_test_name): New global.
(libphobos_init): Set libphobos_test_name.
(libphobos-dg-test): Override name if libphobos_test_name is set.
(filter_libphobos_unittests): New proc.
* testsuite/libphobos.aa/aa.exp: Don't load libphobos-dg.exp.
* testsuite/libphobos.cycles/cycles.exp: Set libphobos_test_name.
* testsuite/libphobos.druntime/druntime.exp: New file.
* testsuite/libphobos.druntime_shared/druntime_shared.exp: New file.
* testsuite/libphobos.phobos/phobos.exp: New file.
* testsuite/libphobos.phobos_shared/phobos_shared.exp: New file.
* testsuite/libphobos.shared/shared.exp: Use dg-runtest to run tests.
* testsuite/libphobos.hash/hash.exp: Don't load libphobos-dg.exp
* testsuite/libphobos.init_fini/init_fini.exp: Likewise.
* testsuite/libphobos.thread/thread.exp: Likewise.
* testsuite/libphobos.typeinfo/typeinfo.exp: Likewise.
* testsuite/libphobos.unittests/unittests.exp: Remove.
* testsuite/test_runner.d: Remove.
* testsuite/testsuite_flags.in: Add phobos to --gdcpaths.

From-SVN: r270302

5 years agod: Add -fbuilding-libphobos-tests option
Iain Buclaw [Fri, 12 Apr 2019 06:25:04 +0000 (06:25 +0000)]
d: Add -fbuilding-libphobos-tests option

Currently, the druntime and phobos unittests are compiled as a separate
check program, then ran by the libphobos.unittest/unittest.exp script.
As PR d/89255 notes, this process lacks proper multilib handling.

As a first step, a new internal option that instructs the compiler to
put the reference to all unittest functions in another symbol has been
added.  This will allow each module to be compiled separately as a
standalone program using dg-runtest without running into collisions in
the D runtime module registry.

gcc/d/ChangeLog:

2019-04-12  Iain Buclaw  <ibuclaw@gdcproject.org>

* d-tree.h (DECL_IN_UNITTEST_CONDITION_P): Define.
* decl.cc (DeclVisitor): Add in_version_unittest_ field.
(DeclVisitor::visit(ConditionalDeclaration)): New override.
(DeclVisitor::visit(FuncDeclaration)): Set
DECL_IN_UNITTEST_CONDITION_P.
* lang.opt (-fbuilding-libphobos-tests): Add option.
* modules.cc (current_testing_module): New static variable.
(build_module_tree): Generate second moduleinfo symbol to hold
reference to unittests if flag_building_libphobos_tests.
(register_module_decl): Check DECL_IN_UNITTEST_CONDITION_P to decide
which moduleinfo the decl should be registered against.

From-SVN: r270301

5 years agod/dmd: Merge upstream dmd c185f9df1
Iain Buclaw [Fri, 12 Apr 2019 06:24:53 +0000 (06:24 +0000)]
d/dmd: Merge upstream dmd c185f9df1

Adds new virtual isVersionCondition, this is so that in the code
generation pass, a ConditionDeclaration's condition can be identified
without requiring a Visitor function.

Reviewed-on: https://github.com/dlang/dmd/pull/9591

From-SVN: r270300

5 years agoDaily bump.
GCC Administrator [Fri, 12 Apr 2019 00:16:31 +0000 (00:16 +0000)]
Daily bump.

From-SVN: r270299

5 years agolibphobos: Merge upstream phobos cf95639ff
Iain Buclaw [Thu, 11 Apr 2019 21:11:15 +0000 (21:11 +0000)]
libphobos: Merge upstream phobos cf95639ff

Backports ports committed to upstream phobos since the last sync.

Reviewed-on: https://github.com/dlang/phobos/pull/6947

From-SVN: r270296

5 years agolibphobos: Merge upstream druntime 175bf5fc
Iain Buclaw [Thu, 11 Apr 2019 21:11:03 +0000 (21:11 +0000)]
libphobos: Merge upstream druntime 175bf5fc

Backports extern(C) bindings committed to upstream druntime since the
last sync.

Reviewed-on: https://github.com/dlang/druntime/pull/2552

From-SVN: r270295

5 years agod/dmd: Merge upstream dmd d7ed327ed
Iain Buclaw [Thu, 11 Apr 2019 21:10:49 +0000 (21:10 +0000)]
d/dmd: Merge upstream dmd d7ed327ed

Fixes ICE when accessing empty array in CTFE.

Reviewed-on: https://github.com/dlang/dmd/pull/9563

From-SVN: r270294

5 years agoReplace direct PSTL uses of assert() with a macro
Thomas Rodgers [Thu, 11 Apr 2019 20:40:40 +0000 (20:40 +0000)]
Replace direct PSTL uses of assert() with a macro

This also replaces calls to __TBB_ASSERT so that there are two macro
definitions provided by c++config -
    __PSTL_ASSERT(_Condition)
    __PSTL_ASSERT_MSG(_Condition, _Message)

    * include/bits/c++config:
    Add definition for __PSTL_ASSERT.
    Add definition for __PSTL_ASSERT_MSG.
    * include/pstl/algorithm_impl.h: Replace use of assert().
    * include/pstl/numeric_impl.h: Replace use of assert().
    * include/pstl/parallel_backend_tbb.h:
    Replace use of assert().
    Replace use of __TBB_ASSERT().

    * include/pstl/parallel_backend_utils.h: Replace use of assert().

From-SVN: r270293

5 years agore PR translation/89939 (messages for translation must not contain embedded macro...
Thomas Koenig [Thu, 11 Apr 2019 20:14:27 +0000 (20:14 +0000)]
re PR translation/89939 (messages for translation must not contain embedded macro parameters)

2019-04-11  Thomas Koenig  <tkoenig@gcc.gnu.org>

PR translation/89939
* frontend-passes.c (B_ERROR): Delete macro.
(C_ERROR): Delete macro.
(B_ERROR_1): New macro.
(C_ERROR_1): New macro.
(C_ERROR_2): New macro.
(inline_matmul_assign): Use new macros.
(call_external_blas): Likewise.

From-SVN: r270292

5 years agoPR libstdc++/90046 fix build failure on epiphany-elf
Jonathan Wakely [Thu, 11 Apr 2019 19:58:14 +0000 (20:58 +0100)]
PR libstdc++/90046 fix build failure on epiphany-elf

The epiphany-elf target aligns structs to 8 bytes, which causes the
static_assert(alignof(_Chunk) == 1) to fail.

Instead of requiring _Chunks to be positionable at any alignment, ensure
new buffers are aligned to alignof(_Chunk). Because the buffer size is a
power of two, we know that both the buffer size and sizeof(_Chunk) are
multiples of alignof(_Chunk). So is p is aligned to alignof(_Chunk) then
so is (p + size - sizeof(_Chunk)). So just ensure the new buffer is
aligned to at least alignof(_Chunk), which should already be true
because the caller requests at least alignof(max_align_t).

PR libstdc++/90046
* src/c++17/memory_resource.cc
(monotonic_buffer_resource::_Chunk::allocate): Increase alignment if
needed to allow placing a _Chunk at the end of the buffer.
(monotonic_buffer_resource::_M_new_buffer): Remove static_assert.

From-SVN: r270291

5 years agore PR rtl-optimization/87763 (aarch64 target testcases fail after r265398)
Steve Ellcey [Thu, 11 Apr 2019 18:03:49 +0000 (18:03 +0000)]
re PR rtl-optimization/87763 (aarch64 target testcases fail after r265398)

2018-04-11  Steve Ellcey  <sellcey@marvell.com>

PR rtl-optimization/87763
* gcc.target/aarch64/combine_bfxil.c: Change some bfxil checks
to bfi.

From-SVN: r270289

5 years agore PR rtl-optimization/87763 (aarch64 target testcases fail after r265398)
Steve Ellcey [Thu, 11 Apr 2019 18:02:41 +0000 (18:02 +0000)]
re PR rtl-optimization/87763 (aarch64 target testcases fail after r265398)

2018-04-11  Steve Ellcey  <sellcey@marvell.com>

PR rtl-optimization/87763
* config/aarch64/aarch64.md (*aarch64_bfi<GPI:mode>4_noshift_alt):
New Instruction.

From-SVN: r270288

5 years ago[doc] Note variable shadowing at max macro using statement expression
Tom de Vries [Thu, 11 Apr 2019 15:36:59 +0000 (15:36 +0000)]
[doc] Note variable shadowing at max macro using statement expression

When suggesting to rewrite the unsafe (with respect to multiple evaluation of
arguments) macro definition:
...
  #define max(a,b) ((a) > (b) ? (a) : (b))
...
into the safe macro definition:
...
  #define maxint(a,b) \
    ({int _a = (a), _b = (b); _a > _b ? _a : _b; })
...
mention the variable shadowing problem for:
...
  #define maxint3(a, b, c) \
    ({int _a = (a), _b = (b), _c = (c); maxint (maxint (_a, _b), _c); })
...

2019-04-11  Tom de Vries  <tdevries@suse.de>

* doc/extend.texi (@node Statement Exprs): Note variable shadowing at
max macro using statement expression.

From-SVN: r270287

5 years agore PR translation/90035 (Non-translatable C++ parser diagnostics)
Jakub Jelinek [Thu, 11 Apr 2019 15:28:18 +0000 (17:28 +0200)]
re PR translation/90035 (Non-translatable C++ parser diagnostics)

PR translation/90035
* parser.h (struct cp_parser): Add
type_definition_forbidden_message_arg member.
* parser.c (cp_debug_parser): Print it.
(cp_parser_check_type_definition): Pass
parser->type_definition_forbidden_message_arg as second argument to
error.
(cp_parser_has_attribute_expression, cp_parser_sizeof_operand): Set
parser->type_definition_forbidden_message_arg and use G_() with
%qs for parser->type_definition_forbidden_message instead of
building untranslatable message using concat.

From-SVN: r270286

5 years agoxcoffout.h (xcoff_private_rodata_section_name): Declare.
David Edelsohn [Thu, 11 Apr 2019 14:14:35 +0000 (14:14 +0000)]
xcoffout.h (xcoff_private_rodata_section_name): Declare.

        * xcoffout.h (xcoff_private_rodata_section_name): Declare.
        * xcoffout.c (xcoff_private_rodata_section_name): Define.
        * config/rs6000/rs6000.c (rs6000_xcoff_asm_init_sections): Create
        read_only_private_data_section using xcoff_private_rodata_section_name.
        (rs6000_xcoff_file_start): Generate xcoff_private_rodata_section_name.

From-SVN: r270284

5 years ago[aarch64] PR90016 - aarch64: reference to undeclared N in help for command line option
Christophe Lyon [Thu, 11 Apr 2019 10:56:08 +0000 (10:56 +0000)]
[aarch64] PR90016 - aarch64: reference to undeclared N in help for command line option

2019-04-11  Christophe Lyon  <christophe.lyon@linaro.org>

PR target/90016
* config/aarch64/aarch64.opt (msve-vector-bits): Add missing final '.'.

From-SVN: r270280

5 years agore PR rtl-optimization/89965 (wrong code with -O -mtune=nano-x2 -fcaller-saves -fexpe...
Jakub Jelinek [Thu, 11 Apr 2019 09:02:47 +0000 (11:02 +0200)]
re PR rtl-optimization/89965 (wrong code with -O -mtune=nano-x2 -fcaller-saves -fexpensive-optimizations -fno-tree-dce -fno-tree-ter)

PR rtl-optimization/89965
* dce.c (sp_based_mem_offset): New function.
(find_call_stack_args): Use sp_based_mem_offset.

From-SVN: r270278

5 years agoClarify documentation for -flive-patching
Jonathan Wakely [Thu, 11 Apr 2019 08:52:22 +0000 (09:52 +0100)]
Clarify documentation for -flive-patching

* doc/invoke.texi (Optimize Options): Clarify -flive-patching docs.

From-SVN: r270276

5 years agore PR tree-optimization/90020 (-O2 -Os x86-64 wrong code generated for GNU Emacs)
Richard Biener [Thu, 11 Apr 2019 07:34:20 +0000 (07:34 +0000)]
re PR tree-optimization/90020 (-O2 -Os x86-64 wrong code generated for GNU Emacs)

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

PR tree-optimization/90020
* tree-ssa-sccvn.c (vn_reference_may_trap): New function.
* tree-ssa-sccvn.h (vn_reference_may_trap): Declare.
* tree-ssa-pre.c (compute_avail): Use it to not put
possibly trapping references after a call that might not
return into EXP_GEN.
* gcse.c (compute_hash_table_work): Do not elide
marking a block containing a call if the call might not
return.

* gcc.dg/torture/pr90020.c: New testcase.

From-SVN: r270275

5 years agore PR tree-optimization/90018 (r265453 miscompiled 527.cam4_r in SPEC CPU 2017)
Richard Biener [Thu, 11 Apr 2019 07:30:59 +0000 (07:30 +0000)]
re PR tree-optimization/90018 (r265453 miscompiled 527.cam4_r in SPEC CPU 2017)

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

PR tree-optimization/90018
* tree-vect-data-refs.c (vect_preserves_scalar_order_p):
Test both SLP and interleaving variants.

* gcc.dg/vect/pr90018.c: New testcase.

From-SVN: r270273

5 years agoS/390: Add arch13 pipeline and model number
Robin Dapp [Thu, 11 Apr 2019 07:04:03 +0000 (07:04 +0000)]
S/390: Add arch13 pipeline and model number

This patch adds the pipeline description and the CPU model number for
arch13.

From-SVN: r270272

5 years agoDaily bump.
GCC Administrator [Thu, 11 Apr 2019 00:16:32 +0000 (00:16 +0000)]
Daily bump.

From-SVN: r270271

5 years agore PR rtl-optimization/87763 (aarch64 target testcases fail after r265398)
Steve Ellcey [Wed, 10 Apr 2019 20:29:57 +0000 (20:29 +0000)]
re PR rtl-optimization/87763 (aarch64 target testcases fail after r265398)

2018-04-10  Steve Ellcey  <sellcey@marvell.com>

PR rtl-optimization/87763
* gcc.target/aarch64/combine_bfxil.c: Change some bfxil checks
to bfi.
* gcc.target/aarch64/combine_bfi_2.c: New test.

From-SVN: r270267

5 years agore PR rtl-optimization/87763 (aarch64 target testcases fail after r265398)
Steve Ellcey [Wed, 10 Apr 2019 20:28:19 +0000 (20:28 +0000)]
re PR rtl-optimization/87763 (aarch64 target testcases fail after r265398)

2018-04-10  Steve Ellcey  <sellcey@marvell.com>

PR rtl-optimization/87763
* config/aarch64/aarch64-protos.h (aarch64_masks_and_shift_for_bfi_p):
New prototype.
* config/aarch64/aarch64.c (aarch64_masks_and_shift_for_bfi_p):
New function.
* config/aarch64/aarch64.md (*aarch64_bfi<GPI:mode>5_shift):
New instruction.
(*aarch64_bfi<GPI:mode>5_shift_alt): Ditto.
(*aarch64_bfi<GPI:mode>4_noand): Ditto.
(*aarch64_bfi<GPI:mode>4_noand_alt): Ditto.
(*aarch64_bfi<GPI:mode>4_noshift): Ditto.

From-SVN: r270266

5 years agoUpdate documentation regarding bogus memory leaks in libstdc++
Jonathan Wakely [Wed, 10 Apr 2019 19:24:04 +0000 (20:24 +0100)]
Update documentation regarding bogus memory leaks in libstdc++

* doc/xml/faq.xml: Add information about emergency EH pool.
* doc/xml/manual/debug.xml: Update list of memory debugging tools.
Move outdated information on mt_allocator to a separate section.
* doc/xml/manual/evolution.xml: Clarify that GLIBCXX_FORCE_NEW
doesn't affect the default allocator.

From-SVN: r270264

5 years agoChange wording of -fipa-icf documentation
Jonathan Wakely [Wed, 10 Apr 2019 14:55:01 +0000 (15:55 +0100)]
Change wording of -fipa-icf documentation

* doc/invoke.texi (Optimize Options): Change "Nevertheless" to
"Although" in -fipa-icf documentation.

From-SVN: r270260

5 years agoFix typo in effective-target check
Jonathan Wakely [Wed, 10 Apr 2019 14:46:03 +0000 (15:46 +0100)]
Fix typo in effective-target check

* testsuite/lib/libstdc++.exp (check_v3_target_parallel_mode): Fix
typo.

From-SVN: r270259

5 years agoFix a stack exhaustion bug in libiberty's demangler when decoding a pathalogically...
Nick Clifton [Wed, 10 Apr 2019 14:44:47 +0000 (14:44 +0000)]
Fix a stack exhaustion bug in libiberty's demangler when decoding a pathalogically constructed mangled name.

PR 89394
* cp-demangle.c (cplus_demangle_fill_name): Reject negative
lengths.
(d_count_templates_scopes): Replace num_templates and num_scopes
parameters with a struct d_print_info pointer parameter.  Adjust
body of the function accordingly.  Add recursion counter and check
that the recursion limit is not reached.
(d_print_init): Pass dpi parameter to d_count_templates_scopes.
Reset recursion counter afterwards, unless the recursion limit was
reached.

From-SVN: r270258

5 years agoClearly document behaviour of multiple -g options
Jonathan Wakely [Wed, 10 Apr 2019 14:39:59 +0000 (15:39 +0100)]
Clearly document behaviour of multiple -g options

This copies the wording from the -O options to clearly state what
happens if more than one -g option is used.

* doc/invoke.texi (Debugging Options): Explicitly state the semantics
of using multiple -g options.

From-SVN: r270257

5 years agoMake gcov docs more precise (PR gcov-profile/89959).
Martin Liska [Wed, 10 Apr 2019 11:44:08 +0000 (13:44 +0200)]
Make gcov docs more precise (PR gcov-profile/89959).

2019-04-10  Martin Liska  <mliska@suse.cz>

PR gcov-profile/89959
* doc/gcov.texi: Make documentation of -x option
more precise.

From-SVN: r270251

5 years agotree-vectorizer.h (_stmt_vec_info): Remove same_dr_stmt member.
Richard Biener [Wed, 10 Apr 2019 11:25:10 +0000 (11:25 +0000)]
tree-vectorizer.h (_stmt_vec_info): Remove same_dr_stmt member.

2019-04-10  Richard Biener  <rguenther@suse.de>

* tree-vectorizer.h (_stmt_vec_info): Remove same_dr_stmt
member.
(DR_GROUP_SAME_DR_STMT): Remove.
* tree-vect-stmts.c (vectorizable_load): Remove unreachable code.
* tree-vect-data-refs.c (vect_analyze_group_access_1): Likewise,
replace with assert.
(vect_analyze_data_ref_accesses): Fix INTEGER_CST comparison.
(vect_record_grouped_load_vectors): Remove unreachable code.

From-SVN: r270250

5 years agoPR libstdc++/89851 Add testcase for std::variant equality
Jonathan Wakely [Wed, 10 Apr 2019 10:43:39 +0000 (11:43 +0100)]
PR libstdc++/89851 Add testcase for std::variant equality

Add a test for the regression introduced with r269422 and fixed with
r270056.

PR libstdc++/89851
* testsuite/20_util/variant/89851.cc: New test.

From-SVN: r270249

5 years ago[aarch64] PR90016 - aarch64: reference to undeclared N in help for command line option
Richard Earnshaw [Wed, 10 Apr 2019 09:51:16 +0000 (09:51 +0000)]
[aarch64] PR90016 - aarch64: reference to undeclared N in help for command line option

'to N' is now redundant and misleading given the earlier change to use
<number>.

Removed.

PR target/90016
* config/aarch64/aarch64.opt (msve-vector-bits): Remove redundant and
obsolete reference to N.

From-SVN: r270248

5 years agore PR middle-end/90025 (botan2 miscompilation on s390x-linux since r268957)
Jakub Jelinek [Wed, 10 Apr 2019 07:28:05 +0000 (09:28 +0200)]
re PR middle-end/90025 (botan2 miscompilation on s390x-linux since r268957)

PR middle-end/90025
* expr.c (store_expr): Set properly size on the MEM passed to
clear_storage.

* gcc.c-torture/execute/pr90025.c: New test.

From-SVN: r270247

5 years agore PR c++/90010 (valgrind error with snprintf and -Wall)
Jakub Jelinek [Wed, 10 Apr 2019 07:27:20 +0000 (09:27 +0200)]
re PR c++/90010 (valgrind error with snprintf and -Wall)

PR c++/90010
* gimple-ssa-sprintf.c (target_to_host): Fix handling of targstr
with strlen in between hostsz-3 and hostsz-1 inclusive when no
translation is needed, and when translation is needed, only append
... if the string length is hostsz or more bytes long.  Avoid using
strncpy or strcat.

* gcc.dg/pr90010.c: New test.

From-SVN: r270246

5 years agoDaily bump.
GCC Administrator [Wed, 10 Apr 2019 00:16:21 +0000 (00:16 +0000)]
Daily bump.

From-SVN: r270245

5 years agoc-ada-spec.c (print_destructor): Deal with deleting destructors.
Eric Botcazou [Tue, 9 Apr 2019 19:29:17 +0000 (19:29 +0000)]
c-ada-spec.c (print_destructor): Deal with deleting destructors.

* c-ada-spec.c (print_destructor): Deal with deleting destructors.
(dump_ada_declaration) <FUNCTION_DECL>: Likewise.

From-SVN: r270239

5 years agoAdd comments and style fixes to <variant>
Jonathan Wakely [Tue, 9 Apr 2019 18:50:48 +0000 (19:50 +0100)]
Add comments and style fixes to <variant>

* include/std/variant: Adjust whitespace. Add comments.
(_Multi_array): Leave primary template undefined.
(_Multi_array<_Tp>): Define partial specialization for base case of
recursion.
(__gen_vtable_impl, __gen_vtable): Remove redundant && from type
which is always a reference.
(__gen_vtable::_S_apply()): Remove function, inline body into
default member initializer.
* testsuite/20_util/variant/visit.cc: Test with noncopyable types.

From-SVN: r270238

5 years agoFix std::visit to support arbitrary callables
Jonathan Wakely [Tue, 9 Apr 2019 18:50:43 +0000 (19:50 +0100)]
Fix std::visit to support arbitrary callables

The __visitor_result_type helper didn't use std::invoke and so didn't
compile when the visitor was a pointer-to-member rather than a function
object. Use std::invoke_result instead.

* include/std/variant (__variant_idx_cookie): Add member type.
(__visitor_result_type): Remove.
(__do_visit): Use invoke_result instead of __visitor_result_type.
* testsuite/20_util/variant/visit.cc: New test.

From-SVN: r270237

5 years agoPR libstdc++/90008 remove unused capture from variant rel ops
Jonathan Wakely [Tue, 9 Apr 2019 18:50:39 +0000 (19:50 +0100)]
PR libstdc++/90008 remove unused capture from variant rel ops

PR libstdc++/90008
* include/std/variant (_VARIANT_RELATION_FUNCTION_TEMPLATE): Remove
unused capture.
* testsuite/20_util/variant/90008.cc: New test.

From-SVN: r270236

5 years ago* gcc.target/i386/ifcvt-onecmpl-abs-1.c
Uros Bizjak [Tue, 9 Apr 2019 18:16:14 +0000 (20:16 +0200)]
* gcc.target/i386/ifcvt-onecmpl-abs-1.c
(dg-options): Use -O2 -fdump-rtl-ce1.
(dg-final): Scan ce1 RTL dump instead of asm dump.

From-SVN: r270234

5 years agoAdd PSTL internal namespace qualifications
Thomas Rodgers [Tue, 9 Apr 2019 15:45:26 +0000 (15:45 +0000)]
Add PSTL internal namespace qualifications

* include/pstl/algorithm_impl.h: Add namespace qualification.
* include/pstl/execution_defs.h: Add namespace qualification.
* include/pstl/execution_impl.h: Add namespace qualification.
* include/pstl/numeric_impl.h: Add namespace qualification.
* include/pstl/parallel_backend_tbb.h: Add namespace qualification.
* include/pstl/unseq_backend_simd.h: Add namespace qualification.
* include/pstl/parallel_backend_utils.h: Include <cassert>.

From-SVN: r270231

5 years agore PR translation/90011 (trailing space in diagnostic)
Jakub Jelinek [Tue, 9 Apr 2019 13:19:16 +0000 (15:19 +0200)]
re PR translation/90011 (trailing space in diagnostic)

PR translation/90011
* typeck2.c (check_narrowing): Remove trailing space from diagnostics.

From-SVN: r270229

5 years agoHi there,
Matthew Malcomson [Tue, 9 Apr 2019 11:39:59 +0000 (11:39 +0000)]
Hi there,

The "*neon_mov<mode>" patterns for 128 bit sized quantities uses the "Dn"
constraint to match vmov.f32 and vmov.i<vec-width> patterns.
This constraint boils down to using the `neon_immediate_valid` function.
Once the constraint has matched, the output C statement asserts that function
passes.

The output C statement calls `neon_immediate_valid` with the mode taken from the
iterator, while the constraint takes the mode from the operand.
This can cause a discrepency when the operand is a CONST_INT, as the constraint
passes VOIDmode which `neon_immediate_valid` treats as DImode, while the C
statement passes the mode of the iterator which can be TImode.
When this happens, the `neon_immediate_valid` can fail in the second call (if
e.g. the CONST_INT is a valid immediate in DImode but not TImode) which would
trigger the assertion.

The testcase added with this patch triggers this when compiled with an arm cross
compiler using the command line below.
gcc -march=armv8-a -c neon-immediate-timode.c -O1 -mfloat-abi=hard -mfpu=neon-fp-armv8

This patch splits the original "Dn" constraint into three new constraints, "DN"
for TImode CONST_INT, "Dn" for DImode CONST_INT, and "Dm" for CONST_VECTOR.
Splitting things up this way requires using one extra alternative in the
"*neon_mov<mode>" patterns, but makes it clear from the constraint what mode is
being used.

We also remove the behaviour of treating VOIDmode as DImode in
`neon_valid_immediate` since the original "Dn" constraint was the only place
that functionality was used.  VOIDmode is now never passed to that function.
An assertion has been added to the function to ensure this problem is caught
earlier on.

Bootstrapped on arm-none-linux-gnueabihf
Regtested on cross-compiler arm-none-eabi

gcc/ChangeLog:

2019-04-09  Matthew Malcomson  <matthew.malcomson@arm.com>

PR target/90024
* config/arm/arm.c (neon_valid_immediate): Disallow VOIDmode parameter.
* config/arm/constraints.md (Dm, DN, Dn): Split previous Dn constraint
into three.
* config/arm/neon.md (*neon_mov<mode>): Account for TImode and DImode
differences directly.
(*smax<mode>3_neon, vashl<mode>3, vashr<mode>3_imm): Use Dm constraint.

gcc/testsuite/ChangeLog:

2019-04-09  Matthew Malcomson  <matthew.malcomson@arm.com>

PR target/90024
* gcc.dg/torture/neon-immediate-timode.c: New test.

From-SVN: r270226

5 years agore PR translation/90011 (trailing space in diagnostic)
Jakub Jelinek [Tue, 9 Apr 2019 10:27:14 +0000 (12:27 +0200)]
re PR translation/90011 (trailing space in diagnostic)

PR translation/90011
* ipa-devirt.c (compare_virtual_tables): Remove two trailing spaces
from diagnostics.
* config/arm/freebsd.h (LINK_SPEC): Remove trailing space from -p
diagnostics.
* config/riscv/freebsd.h (LINK_SPEC): Likewise.
* config/aarch64/aarch64-freebsd.h (FBSD_TARGET_LINK_SPEC): Likewise.
* config/darwin.h (DRIVER_SELF_SPECS, ASM_FINAL_SPEC): Remove
trailing space from -gsplit-dwarf diagnostics.

From-SVN: r270225

5 years agore PR middle-end/89998 (ICE: verify_gimple failed in printf-return-value)
Jakub Jelinek [Tue, 9 Apr 2019 10:26:13 +0000 (12:26 +0200)]
re PR middle-end/89998 (ICE: verify_gimple failed in printf-return-value)

PR tree-optimization/89998
* gimple-ssa-sprintf.c (try_substitute_return_value): Use lhs type
instead of integer_type_node if possible, don't add ranges if return
type is not compatible with int.
* gimple-fold.c (gimple_fold_builtin_sprintf,
gimple_fold_builtin_snprintf): Use lhs type instead of hardcoded
integer_type_node.

* gcc.c-torture/compile/pr89998-1.c: New test.
* gcc.c-torture/compile/pr89998-2.c: New test.

From-SVN: r270224

5 years agoCome up with bootstrap-lto-lean config.
Martin Liska [Tue, 9 Apr 2019 08:49:14 +0000 (10:49 +0200)]
Come up with bootstrap-lto-lean config.

2019-04-09  Martin Liska  <mliska@suse.cz>

* Makefile.in: Regenerate.
* Makefile.tpl: Pass GENERATOR_CFLAGS
in all stages.
2019-04-09  Martin Liska  <mliska@suse.cz>

* bootstrap-lto-lean.mk: New file.
2019-04-09  Martin Liska  <mliska@suse.cz>

* Makefile.in: Use GENERATOR_CFLAGS for all generators.
* doc/install.texi: Document the new config.

From-SVN: r270223

5 years agoSVE fallout from PR90006
Richard Sandiford [Tue, 9 Apr 2019 07:50:51 +0000 (07:50 +0000)]
SVE fallout from PR90006

2019-04-08  Richard Sandiford  <richard.sandiford@arm.com>

gcc/
* tree-vect-data-refs.c (vect_get_smallest_scalar_type): Always
use gimple_expr_type for load and store calls.  Skip over the
condition argument in a conditional internal function.
Protect use of TREE_INT_CST_LOW.

From-SVN: r270222