David Faust [Tue, 1 Jun 2021 16:22:59 +0000 (09:22 -0700)]
BTF: Support for BTF_KIND_FLOAT
Add BTF_KIND_FLOAT, a new BTF type kind which has recently stabilized in
the linux kernel [1]. This kind is used for encoding floating point
types, and is of particular use when generating BTF for some s390
arch-specific kernel headers.
Also update some BTF tests which previously used floating point types to
check correct behavior for types with no BTF representation.
[1]: https://github.com/torvalds/linux/commit/
b1828f0b04828aa8cccadf00a702f459caefeed9
include/ChangeLog:
* btf.h (struct btf_type): Update bit usage comment.
(BTF_INFO_KIND): Update bit mask.
(BTF_KIND_FLOAT): New define.
(BTF_KIND_MAX): Update.
gcc/ChangeLog:
* btfout.c (get_btf_kind): Support BTF_KIND_FLOAT.
(btf_asm_type): Likewise.
gcc/testsuite/ChangeLog:
* gcc.dg/debug/btf/btf-float-1.c: New test.
* gcc.dg/debug/btf/btf-function-3.c: Use different unrepresentable type.
* gcc.dg/debug/btf/btf-struct-2.c: Likewise.
* gcc.dg/debug/btf/btf-variables-2.c: Likewise.
Jeff Law [Fri, 2 Jul 2021 15:07:37 +0000 (11:07 -0400)]
Preparing to use shifts to eliminate redundant test/compare insns on H8
* config/h8300/h8300-protos.h (output_a_shift): Make first argument
an array of rtx rather than a pointer to rtx. Add code argument.
(compute_a_shift_length): Similarly.
* config/h8300/h8300.c (h8300_shift_costs): Adjust now that the
shift itself isn't an operand. Create dummy operand[0] to carry
a mode and pass a suitable rtx code to compute_a_shift_length.
(get_shift_alg): Adjust operand number of clobber in output templates.
(output_a_shift): Make first argument an array of rtx rather than
a pointer to rtx. Add code argument for the type of shift.
Adjust now that the shift itself is no longer an operand.
(compute_a_shift_length): Similarly.
* config/h8300/shiftrotate.md (shiftqi, shifthi, shiftsi): Use an
iterator rather than nshift_operator.
(shiftqi_noscratch, shifthi_noscratch, shiftsi_noscratch): Likewise.
(shiftqi_clobber_flags): Adjust to API changes in output_a_shift
and compute_a_shift_length.
(shiftqi_noscratch_clobber_flags): Likewise.
(shifthi_noscratch_clobber_flags): Likewise.
(shiftsi_noscratch_clobber_flags): Likewise.
Iain Sandoe [Fri, 2 Jul 2021 08:51:57 +0000 (09:51 +0100)]
Darwin, CTF, BTF: Do not run the DWARF debug link for BTF/CTF [PR101283].
Darwin uses an efficient two-stage process for debug linking.
The static linker (ld64) notes the inputs required but does not
link the debug. When required / on demand the debug is linked
into a separate package by the debug linker (dsymutil). At
present none of the Darwin tools consume or understand BTF/CTF.
The static linker silently accepts the sections (but will not
act on them as containing anything to be processed).
However, the debug linker produces a warning that it has been
presented with input with no [DWARF] debug content:
warning: no debug symbols in executable (-arch x86_64).
This causes several testsuite fails with excess errors.
Signed-off-by: Iain Sandoe <iain@sandoe.co.uk>
PR debug/101283 - Several tests fail on Darwin with -gctf/gbtf
PR debug/101283
gcc/ChangeLog:
* config/darwin.h (DSYMUTIL_SPEC): Do not try to run
dsymutil for BTF/CTF.
Iain Sandoe [Fri, 2 Jul 2021 08:44:59 +0000 (09:44 +0100)]
Darwin, BTF: Provide a suitable section name for BTF [PR101283].
In a similar manner to r12-1960-gcc8453012f75d, this provides a
placeholder section name for BTF data. This change groups BTF
and CTF debug formats in the same segment, but keeps them in
separate sections.
As per the CTF section designation, this should be agreed or
amended to an agreed form before GCC 12 ships.
Signed-off-by: Iain Sandoe <iain@sandoe.co.uk>
PR debug/101283 - Several tests fail on Darwin with -gctf/gbtf
PR debug/101283
gcc/ChangeLog:
* config/darwin.h (CTF_INFO_SECTION_NAME): Update the
segment to include BTF.
(BTF_INFO_SECTION_NAME): New.
Jeff Law [Fri, 2 Jul 2021 14:48:26 +0000 (10:48 -0400)]
Fix m32r target specific fallout from recent int->bool changes
gcc/ChangeLog
* config/m32r/m32r-protos.h (call_operand): Adjust return type.
(small_data_operand, memreg_operand, small_insn_p): Likewise.
* config/m32r/m32r.c (call_operand): Adjust return type.
(small_data_operand, memreg_operand): Likewise.
Jeff Law [Fri, 2 Jul 2021 14:37:52 +0000 (10:37 -0400)]
Fix frv target specific fallout from recent int->bool changes
gcc/ChangeLog
* config/frv/frv-protos.h (integer_register_operand): Adjust return
type.
(frv_load_operand, gpr_or_fpr_operand, gpr_no_subreg_operand): Likewise.
(fpr_or_int6_operand, gpr_or_int_operand); Likewise.
(gpr_or_int12_operand, gpr_or_int10_operand); Likewise.
(move_source_operand, move_destination_operand): Likewise.
(condexec_source_operand, condexec_dest_operand): Likewise.
(lr_operand, gpr_or_memory_operand, fpr_or_memory_operand): Likewise.
(reg_or_0_operand, fcc_operand, icc_operand, cc_operand): Likewise.
(fcr_operand, icr_operand, cr_operand, call_operand): Likewise.
(fpr_operand, even_reg_operand, odd_reg_operand): Likewise.
(even_gpr_operand, odd_gpr_operand, quad_fpr_operand): Likewise.
(even_fpr_operand, odd_fpr_operand): Likewise.
(dbl_memory_one_insn_operand, dbl_memory_two_insn_operand): Likewise.
(int12_operand, int6_operand, int5_operand, uint5_operand): Likewise.
(uint4_operand, uint1_operand, int_2word_operand): Likewise
(upper_int16_operand, uint16_operand, symbolic_operand): Likewise.
(relational_operator, float_relational_operator): Likewise.
(ccr_eqne_operator, minmax_operator): Likewise.
(condexec_si_binary_operator, condexec_si_media_operator): Likewise.
(condexec_si_divide_operator, condexec_si_unary_operator): Likewise.
(condexec_sf_conv_operator, condexec_sf_add_operator): Likewise.
(intop_compare_operator, acc_operand, even_acc_operand): Likewise.
(quad_acc_operand, accg_operand): Likewise.
Maxim Kuvyrkov [Fri, 2 Jul 2021 14:15:35 +0000 (14:15 +0000)]
Add description of how testsuite parallelization works
... for the benefit of the next poor soul who will need to debug it.
gcc/testsuite/
* lib/gcc-defs.exp: Add a comment.
Jeff Law [Fri, 2 Jul 2021 14:31:31 +0000 (10:31 -0400)]
Fix xstormy16 target specific fallout from recent int->bool changes
gcc/
* config/stormy16/stormy16-protos.h (xstormy16_below_100_symbol): Change
return type to a bool.
(nonimmediate_nonstack_operand): Likewise.
(xstormy16_splittable_below100_operand): Likewise.
* config/stormy16/stormy16.c (xstormy16_below_100_symbol): Fix
return type.
(xstormy16_splittable_below100_operand): Likewise.
Richard Biener [Fri, 2 Jul 2021 10:57:06 +0000 (12:57 +0200)]
tree-optimization/101293 - further enhance LIMs ref canonicalization
This makes sure to handle MEM[p + 4] and MEM[p].j with j at offset 4
as the same ref in store motion. For hashing we need to be
more restrictive in what we handle since there's no poly-int
handlers for inchash. For comparison we can compare poly_offsets directly.
2021-07-02 Richard Biener <rguenther@suse.de>
PR tree-optimization/101293
* tree-ssa-loop-im.c (mem_ref_hasher::equal): Compare MEM_REF bases
with combined offsets.
(gather_mem_refs_stmt): Hash MEM_REFs as if their offset were
combined with the rest of the offset.
* gcc.dg/tree-ssa/ssa-lim-15.c: New testcase.
Jonathan Wakely [Fri, 2 Jul 2021 07:46:18 +0000 (08:46 +0100)]
libstdc++: Revert changes to std::unique_ptr<T[]>::operator[] [PR 101271]
This reverts the changes in r12-1778 which added a noexcept-specifier to
std::unique_ptr<T[]>::operator[], and the changes in r12-1844 which
tried to make it work with incomplete types (for PR 101236).
The noexcept-specifier is not required by the standard, and is causing
regressions, so just remove it.
Signed-off-by: Jonathan Wakely <jwakely@redhat.com>
libstdc++-v3/ChangeLog:
PR libstdc++/101271
* include/bits/unique_ptr.h (unique_ptr<T[],D>::operator[]):
Remove noexcept-specifier.
(unique_ptr<T[],D>::_S_nothrow_deref): Remove.
* testsuite/20_util/unique_ptr/lwg2762.cc: Remove checks for
operator[].
Eric Botcazou [Fri, 2 Jul 2021 08:21:11 +0000 (10:21 +0200)]
Change EH pointer encodings to PC relative on Windows
A big difference between ELF and PE-COFF is that, with the latter, you can
build position-independent executables or DLLs without generating PIC; as
a matter of fact, flag_pic has historically been forced to 0 for 32-bit:
/* Don't allow flag_pic to propagate since gas may produce invalid code
otherwise. */
\
do {
\
flag_pic = TARGET_64BIT ? 1 : 0; \
} while (0)
The reason is that the linker builds a .reloc section that collects the
absolute relocations in the generated binary, and the loader uses them to
relocate it at load time if need be (e.g. if --dynamicbase is enabled).
Up to binutils 2.35, the GNU linker didn't build the .reloc section for
executables and defaulted to --enable-auto-image-base for DLLs, which means
that DLLs had an essentially unique load address and, therefore, need not
be relocated by the loader in most cases.
With binutils 2.36 and later, the GNU linker builds a .reloc section for
executables (thus making them PIE), --enable-auto-image-base is disabled
and --dynamicbase is enabled by default, which means that essentially all
the binaries are relocated at load time.
This badly breaks the 32-bit compiler configured to use DWARF-2 EH because
the loader corrupts the .eh_frame section when processing the relocations
contained in the .reloc section.
gcc/
* config/i386/i386.c (asm_preferred_eh_data_format): Always use the
PIC encodings for PE-COFF targets.
Jakub Jelinek [Fri, 2 Jul 2021 08:06:56 +0000 (10:06 +0200)]
i386: Punt on broadcasts from TImode integers [PR101286]
ix86_expand_vector_init_duplicate doesn't handle TImode -> V2TImode
or TImode -> V4TImode broadcasts, so I think we should punt on TImode
inner mode in ix86_broadcast_from_integer_constant, otherwise we ICE
in ix86_expand_vector_move when ix86_broadcast_from_integer_constant
returns non-NULL and ix86_expand_vector_init_duplicate returns false.
In theory TImode element broadcasts could be handled by some permutations,
but I'm not sure it is worth it.
2021-07-02 Jakub Jelinek <jakub@redhat.com>
PR target/101286
* config/i386/i386-expand.c (ix86_broadcast_from_integer_constant):
Return nullptr for TImode inner mode.
* gcc.target/i386/avx2-pr101286.c: New test.
Richard Biener [Fri, 2 Jul 2021 06:51:43 +0000 (08:51 +0200)]
tree-optimization/101280 - re-revise interchange fix for PR101173
The following fixes up the revision of the original fix for PR101173
to properly guard all dependence checks with DDR_REVERSED_P or its
inverse.
2021-07-01 Richard Biener <rguenther@suse.de>
PR tree-optimization/101280
PR tree-optimization/101173
* gimple-loop-interchange.cc
(tree_loop_interchange::valid_data_dependences): Properly
guard all dependence checks with DDR_REVERSED_P or its
inverse.
Hongyu Wang [Wed, 30 Jun 2021 06:38:31 +0000 (14:38 +0800)]
Clear odata for aes(enc|dec)(wide)?kl intrinsics when ZF is set.
For Keylocker aesenc/aesdec intrinsics, current implementation
moves idata to odata unconditionally, which causes safety issue when
the instruction meets runtime error. So we add a branch to clear
odata when ZF is set after instruction exectution.
gcc/ChangeLog:
* config/i386/i386-expand.c (ix86_expand_builtin):
Add branch to clear odata when ZF is set for asedecenc_expand
and wideaesdecenc_expand.
gcc/testsuite/ChangeLog:
* gcc.target/i386/keylocker-aesdec128kl.c: Update test.
* gcc.target/i386/keylocker-aesdec256kl.c: Likewise.
* gcc.target/i386/keylocker-aesdecwide128kl.c: Likewise.
* gcc.target/i386/keylocker-aesdecwide256kl.c: Likewise.
* gcc.target/i386/keylocker-aesenc128kl.c: Likewise.
* gcc.target/i386/keylocker-aesenc256kl.c: Likewise.
* gcc.target/i386/keylocker-aesencwide128kl.c: Likewise.
* gcc.target/i386/keylocker-aesencwide256kl.c: Likewise.
Eugene Rozenfeld [Thu, 1 Jul 2021 23:21:36 +0000 (16:21 -0700)]
Update gen_autofdo_event.py and gcc-auto-profile.
gen_autofdo_event.py was stumbling on models with stepping so
I updated the script to handle this case similar to the code in
https://github.com/andikleen/pmu-tools/blob/
c6a5f63aede19def8886d6a8b74d7a55c38ca947/event_download.py
The second change was to tolerate cases when the CPU supports PEBS but the
perf command with /p fails. This can happen in, e.g., a virtual machine.
I regenerated gcc-auto-profile using the updated script.
contrib/ChangeLog:
* gen_autofdo_event.py: handle stepping, non-working PEBS
gcc/ChangeLog:
* config/i386/gcc-auto-profile: regenerate
liuhongt [Wed, 30 Jun 2021 09:10:44 +0000 (17:10 +0800)]
Fix typo in standard pattern name of trunc<mode><pmov_dst_4>2.
gcc/ChangeLog
* config/i386/sse.md (trunc<mode><pmov_dst_4>2): Refined to ..
(trunc<mode><pmov_dst_4_lower>2): this.
GCC Administrator [Fri, 2 Jul 2021 00:16:47 +0000 (00:16 +0000)]
Daily bump.
H.J. Lu [Thu, 1 Jul 2021 19:27:00 +0000 (12:27 -0700)]
soft-fp: Update soft-fp from glibc
This patch is updating soft-fp from glibc:
1. Add __extendhfxf2 to return an IEEE half converted to IEEE extended.
2. Add __truncxfhf2 to truncate IEEE extended into IEEE half.
These are needed by x86 _Float16 support.
* soft-fp/extendhfxf2.c: New file.
* soft-fp/truncxfhf2.c: Likewise.
David Malcolm [Thu, 1 Jul 2021 21:44:51 +0000 (17:44 -0400)]
input.c: move file caching globals to a new file_cache class
This moves some global state from input.c to a new file_cache class,
of which an instance is owned by global_dc. Various state is also
made private.
No functional change intended.
gcc/ChangeLog:
* diagnostic.h (diagnostic_context::m_file_cache): New field.
* input.c (class fcache): Rename to...
(class file_cache_slot): ...this, making most members private and
prefixing fields with "m_".
(file_cache_slot::get_file_path): New accessor.
(file_cache_slot::get_use_count): New accessor.
(file_cache_slot::missing_trailing_newline_p): New accessor.
(file_cache_slot::inc_use_count): New.
(fcache_buffer_size): Move to...
(file_cache_slot::buffer_size): ...here.
(fcache_line_record_size): Move to...
(file_cache_slot::line_record_size): ...here.
(fcache_tab): Delete, in favor of global_dc->m_file_cache.
(fcache_tab_size): Move to file_cache::num_file_slots.
(diagnostic_file_cache_init): Update for move of fcache_tab
to global_dc->m_file_cache.
(diagnostic_file_cache_fini): Likewise.
(lookup_file_in_cache_tab): Convert to...
(file_cache::lookup_file): ...this.
(diagnostics_file_cache_forcibly_evict_file): Update for move of
fcache_tab to global_dc->m_file_cache, moving most of
implementation to...
(file_cache::forcibly_evict_file): ...this new function and...
(file_cache_slot::evict): ...this new function.
(evicted_cache_tab_entry): Convert to...
(file_cache::evicted_cache_tab_entry): ...this.
(add_file_to_cache_tab): Convert to...
(file_cache::add_file): ...this, moving bulk of implementation
to...
(file_cache_slot::create): ..this new function.
(file_cache::file_cache): New.
(file_cache::~file_cache): New.
(lookup_or_add_file_to_cache_tab): Convert to...
(file_cache::lookup_or_add_file): ..this new function.
(fcache::fcache): Rename to...
(file_cache_slot::file_cache_slot): ...this, adding "m_" prefixes
to fields.
(fcache::~fcache): Rename to...
(file_cache_slot::~file_cache_slot): ...this, adding "m_" prefixes
to fields.
(needs_read): Convert to...
(file_cache_slot::needs_read_p): ...this.
(needs_grow): Convert to...
(file_cache_slot::needs_grow_p): ...this.
(maybe_grow): Convert to...
(file_cache_slot::maybe_grow): ...this.
(read_data): Convert to...
(file_cache_slot::read_data): ...this.
(maybe_read_data): Convert to...
(file_cache_slot::maybe_read_data): ...this.
(get_next_line): Convert to...
(file_cache_slot::get_next_line): ...this.
(goto_next_line): Convert to...
(file_cache_slot::goto_next_line): ...this.
(read_line_num): Convert to...
(file_cache_slot::read_line_num): ...this.
(location_get_source_line): Update for moving of globals to
global_dc->m_file_cache.
(location_missing_trailing_newline): Likewise.
* input.h (class file_cache_slot): New forward decl.
(class file_cache): New.
Signed-off-by: David Malcolm <dmalcolm@redhat.com>
Sandra Loosemore [Thu, 1 Jul 2021 20:07:59 +0000 (13:07 -0700)]
Fortran: set version field in CFI_cdesc_t to CFI_VERSION
When converting a GFC descriptor to a CFI descriptor, it was
incorrectly copying the version number from the former to the latter.
It should be using the value of CFI_VERSION instead (currently 1).
Going the other direction, the GFC version field is initialized to
zero elsewhere, so do that here too.
2021-06-30 Tobias Burnus <tobias@codesourcery.com>
Sandra Loosemore <sandra@codesourcery.com>
libgfortran/
* runtime/ISO_Fortran_binding.c (cfi_desc_to_gfc_desc):
Initialize version field to 0.
(gfc_desc_to_cfi_desc): Initialize version field to CFI_VERSION.
Michael Meissner [Thu, 1 Jul 2021 18:50:34 +0000 (14:50 -0400)]
Add IEEE 128-bit fp conditional move on PowerPC.
This patch adds the support for power10 IEEE 128-bit floating point conditional
move and for automatically generating min/max.
In this patch, I simplified things compared to previous patches. Instead of
allowing any four of the modes to be used for the conditional move comparison
and the move itself could use different modes, I restricted the conditional
move to just the same mode. I.e. you can do:
_Float128 a, b, c, d, e, r;
r = (a == b) ? c : d;
But you can't do:
_Float128 c, d, r;
double a, b;
r = (a == b) ? c : d;
or:
_Float128 a, b;
double c, d, r;
r = (a == b) ? c : d;
This eliminates a lot of the complexity of the code, because you don't have to
worry about the sizes being different, and the IEEE 128-bit types being
restricted to Altivec registers, while the SF/DF modes can use any VSX
register.
I did not modify the existing support that allowed conditional moves where
SFmode operands are compared and DFmode operands are moved (and vice versa).
I modified the test cases that I added to reflect this change. I have also
fixed the test for not equal to use '!=' instead of '=='.
2021-07-01 Michael Meissner <meissner@linux.ibm.com>
gcc/
* config/rs6000/rs6000.c (rs6000_maybe_emit_fp_cmove): Add IEEE
128-bit floating point conditional move support.
(have_compare_and_set_mask): Add IEEE 128-bit floating point
types.
* config/rs6000/rs6000.md (mov<mode>cc, IEEE128 iterator): New insn.
(mov<mode>cc_p10, IEEE128 iterator): New insn.
(mov<mode>cc_invert_p10, IEEE128 iterator): New insn.
(fpmask<mode>, IEEE128 iterator): New insn.
(xxsel<mode>, IEEE128 iterator): New insn.
gcc/testsuite/
* gcc.target/powerpc/float128-cmove.c: New test.
* gcc.target/powerpc/float128-minmax-3.c: New test.
Jonathan Wakely [Wed, 30 Jun 2021 23:30:54 +0000 (00:30 +0100)]
libstdc++: Improvements to Doxygen markup
This attempst to improve the doxygen output to work around what seems to
be some bugs in doxygen (issues 8635 and 8638).
The @addtogroup command doesn't work for entities inside a nested
namespace (see 8635) so we need to close and reopen groups on entering
and elaving nested namespaces. This fixes the problem that
chrono::duration and chrono::time_point were not documented in the
"Time" documentation group. I am unable to make the path classes appear
as part of their relevant groups (File System and Filesystem TS), nor
the contents of <exception> or <system_error>. I have made some minor
improvements to the docs for those types, including starting to address
PR 97001 by adding @since to the doxygen comments.
This change also excludes the <experimental/bits/net.h> header from
Doxygen processing, so we don't get an unwanted "Networking-ts" group
in the documentation.
Signed-off-by: Jonathan Wakely <jwakely@redhat.com>
libstdc++-v3/ChangeLog:
* doc/doxygen/doxygroups.cc: Fix docs for std::literals.
* doc/doxygen/user.cfg.in: Exclude the Networking TS header.
Add some more predefined macros.
* include/bits/fs_fwd.h: Move @addtogroup commands inside
namespaces. Add better documentation.
* include/bits/fs_path.h: Likewise.
* include/experimental/bits/fs_fwd.h: Likewise.
* include/experimental/bits/fs_path.h: Likewise.
* include/ext/throw_allocator.h: Fix typo and improve docs.
* include/std/chrono: Move @addtogroup commands.
* include/std/system_error: Move @addtogroup commands.
* libsupc++/exception: Improve documentation.
* libsupc++/exception.h: Add @since documentation.
Eric Botcazou [Thu, 1 Jul 2021 16:06:46 +0000 (18:06 +0200)]
Improve packed record layout support with -fdump-ada-spec
We cannot fully support packed record layout in -fdump-ada-spec, as packing
in C and Ada does not behave the same, so we issue a warning. But simple
cases are OK and can actually be handled without much work.
gcc/c-family/
* c-ada-spec.c (packed_layout): New global variable.
(dump_ada_declaration): Set it upon seeing a packed record type.
Do not put the "aliased" keyword if it is set.
(dump_ada_structure): Add Pack aspect if it is set and clear it.
gcc/testsuite/
* c-c++-common/dump-ada-spec-14.c: Adjust dg-warning directive.
Eric Botcazou [Thu, 1 Jul 2021 16:02:47 +0000 (18:02 +0200)]
Fix duplicate name issues in output of -fdump-ada-spec #2
This extends the type name conflict detection mechanism to variables.
gcc/c-family/
* c-ada-spec.c (check_name): Rename into...
(check_type_name_conflict): ...this. Minor tweak.
(dump_ada_function_declaration): Adjust to above renaming.
(dump_ada_array_domains): Fix oversight.
(dump_ada_declaration): Call check_type_name_conflict for variables.
Eric Botcazou [Thu, 1 Jul 2021 15:57:16 +0000 (17:57 +0200)]
Use intermediate integer type with proper signedness
This is a minor regression present on mainline and 11 branch, whereby the
value of the Enum_Rep attribute is always unsigned.
gcc/ada/
PR ada/101094
* exp_attr.adb (Get_Integer_Type): Return an integer type with the
same signedness as the input type.
Iain Sandoe [Thu, 1 Jul 2021 14:02:43 +0000 (15:02 +0100)]
Darwin: Define a suitable section name for CTF [PR101283]
This is a placeholder name ahead of any CTF implementation on
LLVM (which sets Darwin ABI). Ideally, we would get agreement
on this choice (or any replacement) before GCC12 is shipped.
PR debug/101283 - Several tests fail on Darwin with -gctf
PR debug/101283
gcc/ChangeLog:
* config/darwin.h (CTF_INFO_SECTION_NAME): New.
H.J. Lu [Mon, 7 Jun 2021 21:23:04 +0000 (14:23 -0700)]
x86: Add vec_duplicate<mode> expander
Add vec_duplicate<mode> expander for SSE2 if we can move from GPR to SSE
register directly.
* config/i386/i386-expand.c (ix86_expand_vector_init_duplicate):
Make it global.
* config/i386/i386-protos.h (ix86_expand_vector_init_duplicate):
New prototype.
* config/i386/sse.md (INT_BROADCAST_MODE): New mode iterator.
(vec_duplicate<mode>): New expander.
H.J. Lu [Wed, 2 Jun 2021 14:15:45 +0000 (07:15 -0700)]
x86: Convert CONST_WIDE_INT/CONST_VECTOR to broadcast
1. Update move expanders to convert the CONST_WIDE_INT and CONST_VECTOR
operands to vector broadcast from an integer with AVX.
2. Add ix86_gen_scratch_sse_rtx to return a scratch SSE register which
won't increase stack alignment requirement and blocks transformation by
the combine pass.
A small benchmark:
https://gitlab.com/x86-benchmarks/microbenchmark/-/tree/memset/broadcast
shows that broadcast is a little bit faster on Intel Core i7-8559U:
$ make
gcc -g -I. -O2 -c -o test.o test.c
gcc -g -c -o memory.o memory.S
gcc -g -c -o broadcast.o broadcast.S
gcc -g -c -o vec_dup_sse2.o vec_dup_sse2.S
gcc -o test test.o memory.o broadcast.o vec_dup_sse2.o
./test
memory : 147215
broadcast : 121213
vec_dup_sse2: 171366
$
broadcast is also smaller:
$ size memory.o broadcast.o
text data bss dec hex filename
132 0 0 132 84 memory.o
122 0 0 122 7a broadcast.o
$
3. Update PR 87767 tests to expect integer broadcast instead of broadcast
from memory.
4. Update avx512f_cond_move.c to expect integer broadcast.
A small benchmark:
https://gitlab.com/x86-benchmarks/microbenchmark/-/tree/vpaddd/broadcast
shows that integer broadcast is faster than embedded memory broadcast:
$ make
gcc -g -I. -O2 -march=skylake-avx512 -c -o test.o test.c
gcc -g -c -o memory.o memory.S
gcc -g -c -o broadcast.o broadcast.S
gcc -o test test.o memory.o broadcast.o
./test
memory : 425538
broadcast : 375260
$
gcc/
PR target/100865
* config/i386/i386-expand.c (ix86_expand_vector_init_duplicate):
New prototype.
(ix86_byte_broadcast): New function.
(ix86_convert_const_wide_int_to_broadcast): Likewise.
(ix86_expand_move): Convert CONST_WIDE_INT to broadcast if mode
size is 16 bytes or bigger.
(ix86_broadcast_from_integer_constant): New function.
(ix86_expand_vector_move): Convert CONST_WIDE_INT and CONST_VECTOR
to broadcast if mode size is 16 bytes or bigger.
* config/i386/i386-protos.h (ix86_gen_scratch_sse_rtx): New
prototype.
* config/i386/i386.c (ix86_gen_scratch_sse_rtx): New function.
gcc/testsuite/
PR target/100865
* gcc.target/i386/avx512f-broadcast-pr87767-1.c: Expect integer
broadcast.
* gcc.target/i386/avx512f-broadcast-pr87767-5.c: Likewise.
* gcc.target/i386/avx512vl-broadcast-pr87767-1.c: Likewise.
* gcc.target/i386/avx512vl-broadcast-pr87767-5.c: Likewise.
* gcc.target/i386/avx512f_cond_move.c: Also pass
-mprefer-vector-width=512 and expect integer broadcast.
* gcc.target/i386/pr100865-1.c: New test.
* gcc.target/i386/pr100865-2.c: Likewise.
* gcc.target/i386/pr100865-3.c: Likewise.
* gcc.target/i386/pr100865-4a.c: Likewise.
* gcc.target/i386/pr100865-4b.c: Likewise.
* gcc.target/i386/pr100865-5a.c: Likewise.
* gcc.target/i386/pr100865-5b.c: Likewise.
* gcc.target/i386/pr100865-6a.c: Likewise.
* gcc.target/i386/pr100865-6b.c: Likewise.
* gcc.target/i386/pr100865-6c.c: Likewise.
* gcc.target/i386/pr100865-7a.c: Likewise.
* gcc.target/i386/pr100865-7b.c: Likewise.
* gcc.target/i386/pr100865-7c.c: Likewise.
* gcc.target/i386/pr100865-8a.c: Likewise.
* gcc.target/i386/pr100865-8b.c: Likewise.
* gcc.target/i386/pr100865-8c.c: Likewise.
* gcc.target/i386/pr100865-9a.c: Likewise.
* gcc.target/i386/pr100865-9b.c: Likewise.
* gcc.target/i386/pr100865-9c.c: Likewise.
* gcc.target/i386/pr100865-10a.c: Likewise.
* gcc.target/i386/pr100865-10b.c: Likewise.
* gcc.target/i386/pr100865-11a.c: Likewise.
* gcc.target/i386/pr100865-11b.c: Likewise.
* gcc.target/i386/pr100865-11c.c: Likewise.
* gcc.target/i386/pr100865-12a.c: Likewise.
* gcc.target/i386/pr100865-12b.c: Likewise.
* gcc.target/i386/pr100865-12c.c: Likewise.
Uros Bizjak [Thu, 1 Jul 2021 14:57:57 +0000 (16:57 +0200)]
i386: Return true/false instead of 1/0 from predicates.
No functional changes.
2021-07-01 Uroš Bizjak <ubizjak@gmail.com>
gcc/
* config/i386/predicates.md (ix86_endbr_immediate_operand):
Return true/false instead of 1/0.
(movq_parallel): Ditto.
Uros Bizjak [Thu, 1 Jul 2021 14:52:11 +0000 (16:52 +0200)]
Return true/false instead of 1/0 from generic predicates.
No functional changes.
2021-07-01 Uroš Bizjak <ubizjak@gmail.com>
gcc/
* recog.c (general_operand): Return true/false instead of 1/0.
(register_operand): Ditto.
(immediate_operand): Ditto.
(const_int_operand): Ditto.
(const_scalar_int_operand): Ditto.
(const_double_operand): Ditto.
(push_operand): Ditto.
(pop_operand): Ditto.
(memory_operand): Ditto.
(indirect_operand): Ditto.
Uros Bizjak [Thu, 1 Jul 2021 13:18:50 +0000 (15:18 +0200)]
Change the type of predicates to bool.
The patch was tested on many targets, but some fallout is expected.
To fix the build error, simply change the function type from int to bool,
as was done in the patch for several targets.
2021-07-01 Uroš Bizjak <ubizjak@gmail.com>
gcc/
* genpreds.c (write_predicate_subfunction):
Change the type of written subfunction to bool.
(write_one_predicate_function):
Change the type of written function to bool.
(write_tm_preds_h): Ditto.
* recog.h (*insn_operand_predicate_fn): Change the type to bool.
* recog.c (general_operand): Change the type to bool.
(address_operand): Ditto.
(register_operand): Ditto.
(pmode_register_operand): Ditto.
(scratch_operand): Ditto.
(immediate_operand): Ditto.
(const_int_operand): Ditto.
(const_scalar_int_operand): Ditto.
(const_double_operand): Ditto.
(nonimmediate_operand): Ditto.
(nonmemory_operand): Ditto.
(push_operand): Ditto.
(pop_operand): Ditto.
(memory_operand): Ditto.
(indirect_operand): Ditto.
(ordered_comparison_operator): Ditto.
(comparison_operator): Ditto.
* config/i386/i386-expand.c (ix86_expand_sse_cmp):
Change the type of indirect predicate function to bool.
* config/rs6000/rs6000.c (easy_vector_constant):
Change the type to bool.
* config/mips/mips-protos.h (m16_based_address_p):
Change the type of operand 3 to bool.
Richard Biener [Thu, 1 Jul 2021 10:49:45 +0000 (12:49 +0200)]
tree-optimization/101280 - revise interchange fix for PR101173
The following revises the original fix for PR101173 to correctly
check for a reversed dependence rather than disallowing a zero
distance. It also adds a check from TSVC which asks for this
kind of interchange (but with a valid dependence).
2021-07-01 Richard Biener <rguenther@suse.de>
PR tree-optimization/101280
PR tree-optimization/101173
* gimple-loop-interchange.cc
(tree_loop_interchange::valid_data_dependences): Revert
previous change and instead correctly handle DDR_REVERSED_P
dependence.
* gcc.dg/tree-ssa/loop-interchange-16.c: New testcase.
Ankur Saini [Thu, 1 Jul 2021 11:13:24 +0000 (16:43 +0530)]
MAINTAINERS - Add myself for write after approval
2021-07-01 Ankur Saini <arsenic@sourceware.org>
* MAINTAINERS: Add myself for write after approval.
Richard Biener [Thu, 1 Jul 2021 09:16:01 +0000 (11:16 +0200)]
tree-optimization/101278 - handle self-use in DSE analysis
DSE store classification short-cuts the to-be classified stmt itself
from chaining but fails to first check whether the store uses itself
which can be the case when it is a call with the LHS also passed by
value as argument.
2021-07-01 Richard Biener <rguenther@suse.de>
PR tree-optimization/101278
* tree-ssa-dse.c (dse_classify_store): First check for
uses, then ignore stmt for chaining purposes.
* gcc.dg/torture/pr101278.c: New testcase.
Richard Biener [Thu, 1 Jul 2021 08:35:38 +0000 (10:35 +0200)]
tree-optimization/100778 - fix placement of trapping vectorized ops
This avoids placing possibly trapping vectorized operations where
the corresponding scalar operation was possibly not executed.
2021-01-07 Richard Biener <rguenther@suse.de>
PR tree-optimization/100778
* tree-vect-slp.c (vect_schedule_slp_node): Do not place trapping
vectorized ops ahead of their scalar BB.
* gcc.dg/torture/pr100778.c: New testcase.
Uros Bizjak [Thu, 1 Jul 2021 08:56:32 +0000 (10:56 +0200)]
i386: Add integer nabs instructions [PR101044]
The patch adds integer nabs "(NEG (ABS (...)))" instructions, adds STV
conversion and adjusts STV cost calculations accordingly. When CMOV
instruction is used to implement abs, the sign is determined from the
preceeding operand negation, and CMOVS is used to select between
negated and non-negated value.
To implement nabs, just reverse the condition and emit CMOVNS instead.
The STV costs are adjusted for inherent NOT of nabs insn. V2DI NOT is
somehow costly operation, since it is implemented as a load of zero,
followed by a SUB insn. OTOH, integer nabs with inherent NOT is relatively
cheap, so some STV chains became less profitable for conversion.
The patch rewrites operand scanner in compute_convert_gain to a switch
and reorders case instances in general_scalar_to_vector_candidate_p
to benefit from fallthroughs, and to remove special processing of
andnot in the later case.
gcc/
2021-07-01 Uroš Bizjak <ubizjak@gmail.com>
PR target/101044
* config/i386/i386.md (*nabs<dwi>2_doubleword):
New insn_and_split pattern.
(*nabs<dwi>2_1): Ditto.
* config/i386/i386-features.c
(general_scalar_chain::compute_convert_gain):
Handle (NEG (ABS (...))) RTX. Rewrite src code
scanner as switch statement.
(general_scalar_chain::convert_insn):
Handle (NEG (ABS (...))) RTX.
(general_scalar_to_vector_candidate_p):
Detect (NEG (ABS (...))) RTX. Reorder case statements
for (AND (NOT (...) ...)) fallthrough.
gcc/testsuite/
2021-07-01 Uroš Bizjak <ubizjak@gmail.com>
PR target/101044
* gcc.target/i386/pr101044.c: New test.
Richard Biener [Wed, 30 Jun 2021 14:28:50 +0000 (16:28 +0200)]
tree-optimization/101178 - handle VEC_PERM in SLP permute propagation
This adds handling of VEC_PERM nodes to SLP permute propagation.
Previously VEC_PERM acted as forced materialization of incoming
permutes since it is a good place to do that (with the constraint
of those only appearing for two-operator nodes). The following
patch, in addition to supporting (but not forcing) this, enables
VEC_PERM nodes acting as "any" permute on the outgoing side since
they also can consume arbitrary permutes on that side.
This again (meh) changes how we represent permutes and materialization
on the graph vertices now explicitely having the common incoming
permute as well as an outgoing permute and in case both are
different the vertex acts as materialization point of the incoming
permute.
2021-06-30 Richard Biener <rguenther@suse.de>
PR tree-optimization/101178
* tree-vect-slp.c (slpg_vertex::materialize): Remove.
(slpg::perm_in): Add.
(slpg::get_perm_in): Remove.
(slpg::get_perm_materialized): Add.
(vect_optimize_slp): Handle VEC_PERM nodes more optimally
during permute propagation and materialization.
* gcc.dg/vect/bb-slp-72.c: New testcase.
* gcc.dg/vect/bb-slp-73.c: Likewise.
* gcc.dg/vect/bb-slp-74.c: Likewise.
Jakub Jelinek [Thu, 1 Jul 2021 07:45:02 +0000 (09:45 +0200)]
dwarf2out: Handle COMPOUND_LITERAL_EXPR in loc_list_from_tree_1 [PR101266]
In this case dwarf2out_decl is called from the FEs with GENERIC but not
yet gimplified expressions in it.
As loc_list_from_tree_1 has an exhaustive list of tree codes it wants to
handle and for checking asserts no other codes makes it in, we should
handle even GENERIC trees that shouldn't be valid in GIMPLE.
The following patch handles COMPOUND_LITERAL_EXPR by hnadling it like the
underlying VAR_DECL temporary.
Verified the emitted DWARF is correct (but unoptimized, we emit
DW_OP_lit1 DW_OP_lit1 DW_OP_minus for the upper bound).
2021-07-01 Jakub Jelinek <jakub@redhat.com>
PR debug/101266
* dwarf2out.c (loc_list_from_tree_1): Handle COMPOUND_LITERAL_EXPR.
* gcc.dg/pr101266.c: New test.
Jakub Jelinek [Thu, 1 Jul 2021 06:55:49 +0000 (08:55 +0200)]
openmp - Fix up && and || reductions [PR94366]
As the testcase shows, the special treatment of && and || reduction combiners
where we expand them as omp_out = (omp_out != 0) && (omp_in != 0) (or with ||)
is not needed just for &&/|| on floating point or complex types, but for all
&&/|| reductions - when expanded as omp_out = omp_out && omp_in (not in C but
GENERIC) it is actually gimplified into NOP_EXPRs to bool from both operands,
which turns non-zero values multiple of 2 into 0 rather than 1.
This patch just treats all &&/|| the same and furthermore uses bool type
instead of int for the comparisons.
2021-07-01 Jakub Jelinek <jakub@redhat.com>
PR middle-end/94366
gcc/
* omp-low.c (lower_rec_input_clauses): Rename is_fp_and_or to
is_truth_op, set it for TRUTH_*IF_EXPR regardless of new_var's type,
use boolean_type_node instead of integer_type_node as NE_EXPR type.
(lower_reduction_clauses): Likewise.
libgomp/
* testsuite/libgomp.c-c++-common/pr94366.c: New test.
Patrick Palka [Thu, 1 Jul 2021 00:44:52 +0000 (20:44 -0400)]
c++: cxx_eval_array_reference and empty elem type [PR101194]
Here the initializer for x is represented as an empty CONSTRUCTOR due to
its empty element type. So during constexpr evaluation of the ARRAY_REF
x[0], we end up trying to value initialize the omitted element at index 0,
which fails because the element type is not default constructible.
This patch makes cxx_eval_array_reference specifically handle the case
where the element type is an empty type.
PR c++/101194
gcc/cp/ChangeLog:
* constexpr.c (cxx_eval_array_reference): When the element type
is an empty type and the corresponding element is omitted, just
return an empty CONSTRUCTOR instead of attempting value
initialization.
gcc/testsuite/ChangeLog:
* g++.dg/cpp0x/constexpr-empty16.C: New test.
Patrick Palka [Thu, 1 Jul 2021 00:21:16 +0000 (20:21 -0400)]
c++: Extend the PR96204 fix to variable templates too
r12-1829 corrected the access scope during partial specialization
matching of class templates, but overlooked the variable template case.
This patch moves the access scope adjustment to within
most_specialized_partial_spec so that all callers can benefit.
This patch also adjusts a couple of these callers to avoid always
passing the most general template of a variable template specialization,
since that'd cause us to push the wrong access scope for e.g. the second
testcase below (we'd push A<T> instead of A<int>/A<char>). We ought to
be passing the partially instantiated template instead.
PR c++/96204
gcc/cp/ChangeLog:
* pt.c (finish_template_variable): Pass the partially
instantiated template and its args to instantiate_template.
(instantiate_class_template_1): No need to call
push_nested_class and pop_nested_class around the call to
most_specialized_partial_spec.
(instantiate_template_1): Pass the partially instantiated
template to lookup_template_variable.
(most_specialized_partial_spec): Use push_access_scope_guard
to set the access scope appropriately. Use
deferring_access_check_sentinel to force access to get checked
immediately.
(instantiate_decl): Just pass the VAR_DECL to
most_specialized_partial_spec.
gcc/testsuite/ChangeLog:
* g++.dg/template/access41.C: New test.
* g++.dg/template/access41a.C: New test.
GCC Administrator [Thu, 1 Jul 2021 00:16:41 +0000 (00:16 +0000)]
Daily bump.
Jonathan Wakely [Wed, 30 Jun 2021 15:00:58 +0000 (16:00 +0100)]
libstdc++: Improve Doxygen documentation groups [PR 101258]
This defines some new Doxygen groups for C++17 variable templates and
for the contents of <experimental/type_traits>. By documenting the group
as a whole and adding each template to a group we don't need to document
them individually.
Also mark more internals with "@cond undocumented" so that Doxygen
ignores them by default. Also make Doxygen process <experimental/simd>.
For some reason, many of the class templates in <type_traits> do not
appear in the "Metaprogramming" group. For example, add_cv,
remove_extent, and all the is_xxx_constructible and is_xxx_assignable
traits. For some reason, Doxygen doesn't include them in the group,
despite doing it correctly for other traits in the same header.
Signed-off-by: Jonathan Wakely <jwakely@redhat.com>
libstdc++-v3/ChangeLog:
PR libstdc++/101258
* doc/doxygen/user.cfg.in (INPUT): Add <experimental/simd>.
(COLS_IN_ALPHA_INDEX): Remove obsolete tag.
(PREDEFINED): Add/fix some more macros that need to be expanded.
* include/bits/random.h: Stop Doxygen from documenting internal
implementation details.
* include/bits/random.tcc: Likewise.
* include/bits/this_thread_sleep.h: Fix @file name.
* include/experimental/bits/simd.h: Add to Doxygen group. Do not
document internal implementation details.
* include/experimental/bits/simd_detail.h: Do not document
internal implementation details.
* include/experimental/simd: Define Doxygen groups.
* include/experimental/type_traits: Improve documentation for
the header file. Define groups. Use @since commands.
* include/std/scoped_allocator (scoped_allocator_adaptor): Move
declaration before undocumented region.
* include/std/type_traits (true_type, false_type): Use using
declaration instead of typedef.
(is_invocable_v, is_nothrow_invocable_v, is_invocable_r_v)
(is_nothrow_invocable_r_v): Move definitions next to other C++17
variable templates.
Do not document internal implementation details. Move misplaced
group-end command. Define group for variable templates.
* include/std/variant: Do not document internal implementation
details.
* testsuite/26_numerics/random/pr60037-neg.cc: Adjust dg-error
line number.
Jonathan Wakely [Wed, 30 Jun 2021 20:35:51 +0000 (21:35 +0100)]
libstdc++: Make <experimental/simd> depend on C++17
Signed-off-by: Jonathan Wakely <jwakely@redhat.com>
libstdc++-v3/ChangeLog:
* include/experimental/simd: Do not define anything pre-C++17.
Jonathan Wakely [Wed, 30 Jun 2021 14:47:41 +0000 (15:47 +0100)]
libstdc++: Suppress redundant definitions of static members in <random>
Since C++17 the static members of the random number engines are
implicitly inline, so don't need definitions.
Signed-off-by: Jonathan Wakely <jwakely@redhat.com>
libstdc++-v3/ChangeLog:
* include/bits/random.tcc [__cpp_inline_variables]: Remove
redundant definitions of static constexpr member variables.
* testsuite/26_numerics/random/pr60037-neg.cc: Adjust dg-error
line number.
Indu Bhagat [Wed, 30 Jun 2021 22:13:50 +0000 (15:13 -0700)]
testsuite: Add missing dg-add-options to CTF testcase ctf-skip-types-4.c
The test already has the appropriate dg-require-effective-target, but requires
the dg-add-options to use the flags needed, if any. This patch fixes the
failure of this testcase on powerpc64.
2021-06-30 Indu Bhagat <indu.bhagat@oracle.com>
gcc/testsuite/
* gcc.dg/debug/ctf/ctf-skip-types-4.c: Add dg-add-options float64 and
float64x.
Gerald Pfeifer [Mon, 28 Jun 2021 22:39:15 +0000 (00:39 +0200)]
libiberty: No longer use /usr/tmp
/usr/tmp is antiquated and not present on decently modern systems.
Remove it from consideration when choosing a directory for temporary
files.
libiberty:
2021-06-29 Gerald Pfeifer <gerald@pfeifer.com>
* make-temp-file.c (usrtmp): Remove.
(choose_tmpdir): Remove use of usrtmp.
Michael Meissner [Wed, 30 Jun 2021 18:54:48 +0000 (14:54 -0400)]
Fix IEEE 128-bit min/max test.
This patch fixes the float128-minmax.c test so that it can accommodate the
generation of xsmincqp and xsmaxcqp instructions on power10. I changed the
effective target from 'float128' to 'ppc_float128_hw', since this needs the
IEEE 128-bit float hardware support. Changing to use 'ppc_float128_hw' allows
the 'lp64' test to be dropped. The 'lp64' test originally was needed because
big endian 32-bit code cannot enable the IEEE 128-bit floating point
instructions.
gcc/testsuite/
2021-06-30 Michael Meissner <meissner@linux.ibm.com>
* gcc.target/powerpc/float128-minmax.c: Adjust expected code for
power10.
* lib/target-supports.exp (check_effective_target_has_arch_pwr10):
New.
Patrick Palka [Wed, 30 Jun 2021 17:12:36 +0000 (13:12 -0400)]
c++: Fix push_access_scope and introduce RAII wrapper for it
When push_access_scope is passed a TYPE_DECL for a class type (which
can happen during e.g. satisfaction), we undesirably push only the
enclosing context of the class instead of the class itself. This
causes us to mishandle e.g. testcase below due to us not entering the
scope of A before checking its constraints.
This patch adjusts push_access_scope accordingly, and introduces an
RAII wrapper for it. We make use of this wrapper right away by
replacing the only user of push_nested_class_guard with this new
wrapper, which means we can remove push_nested_class_guard (whose
functionality is basically subsumed by the new wrapper).
gcc/cp/ChangeLog:
* constraint.cc (get_normalized_constraints_from_decl): Use
push_access_scope_guard instead of push_nested_class_guard.
* cp-tree.h (struct push_nested_class_guard): Replace with ...
(struct push_access_scope_guard): ... this.
* pt.c (push_access_scope): When the argument corresponds to
a class type, push the class instead of its context.
(pop_access_scope): Adjust accordingly.
gcc/testsuite/ChangeLog:
* g++.dg/cpp2a/concepts-access2.C: New test.
Hafiz Abid Qadeer [Fri, 28 May 2021 17:06:57 +0000 (18:06 +0100)]
[amdgcn] Add hook for DWARF address spaces.
Map GCN address spaces to the proposed DWARF address spaces defined by AMD at
https://llvm.org/docs/AMDGPUUsage.html#amdgpu-dwarf-address-class-mapping-table
gcc/
* config/gcn/gcn.c: Include dwarf2.h.
(gcn_addr_space_debug): New function.
(TARGET_ADDR_SPACE_DEBUG): New hook.
Hafiz Abid Qadeer [Mon, 21 Jun 2021 21:47:58 +0000 (22:47 +0100)]
[amdgcn] Use frame pointer for CFA expressions.
As size of address is bigger than registers in amdgcn, we are forced to use
DW_CFA_def_cfa_expression to make an expression that concatenates multiple
registers for the value of the CFA. This then prohibits us from using many
of the dwarf ops which expect CFA rule to be a single regsiter plus an offset.
Using frame pointer in the CFA rule is only real possibility as it is saved
in every frame and it is easy to unwind its value.
So unless user gives fomit-frame-pointer, we use frame pointer for the
cfi information. This options also has a different default now.
gcc/
* common/config/gcn/gcn-common.c
(gcn_option_optimization_table): Change OPT_fomit_frame_pointer to -O3.
* config/gcn/gcn.c (gcn_expand_prologue): Prefer the frame pointer
when emitting CFI.
(gcn_expand_prologue): Prefer the frame pointer when emitting CFI.
(gcn_frame_pointer_rqd): New function.
(TARGET_FRAME_POINTER_REQUIRED): New hook.
Hafiz Abid Qadeer [Fri, 28 May 2021 16:49:46 +0000 (17:49 +0100)]
[amdgcn] Update CFI configuration
Currently we don't get any call frame information for the amdgcn target.
This patch makes necessary adjustments to generate CFI that can work with
ROCGDB (ROCm 3.8+).
gcc/
* config/gcn/gcn.c (move_callee_saved_registers): Emit CFI notes for
prologue register saves.
(gcn_debug_unwind_info): Use UI_DWARF2.
(gcn_dwarf_register_number): Map DWARF_LINK_REGISTER to DWARF PC.
(gcn_dwarf_register_span): DWARF_LINK_REGISTER doesn't span.
* config/gcn/gcn.h: (DWARF_FRAME_RETURN_COLUMN): New define.
(DWARF_LINK_REGISTER): New define.
(FIRST_PSEUDO_REGISTER): Increment.
(FIXED_REGISTERS): Add entry for DWARF_LINK_REGISTER.
(CALL_USED_REGISTERS): Likewise.
(REGISTER_NAMES): Likewise.
Marek Polacek [Tue, 29 Jun 2021 18:30:51 +0000 (14:30 -0400)]
c++: DR2397 - auto specifier for * and & to arrays [PR100975]
This patch implements DR2397, which removes the restriction in
[dcl.array]p4 that the array element type may not be a placeholder
type. We don't need to worry about decltype(auto) here, so this
allows code like
int a[3];
auto (*p)[3] = &a;
auto (&r)[3] = a;
However, note that
auto (&&r)[2] = { 1, 2 };
auto arr[2] = { 1, 2 };
still doesn't work (although one day it might) and neither does
int arr[5];
auto x[5] = arr;
given that auto deduction is performed in terms of function template
argument deduction, so the array decays to *.
PR c++/100975
DR 2397
gcc/cp/ChangeLog:
* decl.c (create_array_type_for_decl): Allow array of auto.
gcc/testsuite/ChangeLog:
* g++.dg/cpp0x/auto24.C: Remove dg-error.
* g++.dg/cpp0x/auto3.C: Adjust dg-error.
* g++.dg/cpp0x/auto42.C: Likewise.
* g++.dg/cpp0x/initlist75.C: Likewise.
* g++.dg/cpp0x/initlist80.C: Likewise.
* g++.dg/diagnostic/auto1.C: Remove dg-error.
* g++.dg/cpp23/auto-array.C: New test.
Richard Biener [Wed, 30 Jun 2021 10:35:45 +0000 (12:35 +0200)]
tree-optimization/101267 - fix SLP vect with masked operations
This fixes the missed handling of external/constant mask SLP
operations, for the testcase in particular masked loads. The
patch adjusts the vect_check_scalar_mask API to reflect the
required vect_is_simple_use SLP compatible API plus adjusts
for the special handling of masked loads in SLP discovery.
The issue is likely latent.
2021-06-30 Richard Biener <rguenther@suse.de>
PR tree-optimization/101267
* tree-vect-stmts.c (vect_check_scalar_mask): Adjust
API and use SLP compatible interface of vect_is_simple_use.
Reject not vectorized SLP defs for callers that do not support
that.
(vect_check_store_rhs): Handle masked stores and pass down
the appropriate operator index.
(vectorizable_call): Adjust.
(vectorizable_store): Likewise.
(vectorizable_load): Likewise. Handle SLP pecularity of
masked loads.
(vect_is_simple_use): Remove special-casing of masked stores.
* gfortran.dg/pr101267.f90: New testcase.
David Malcolm [Wed, 30 Jun 2021 13:39:04 +0000 (09:39 -0400)]
analyzer: eliminate enum binding_key [PR95006]
I rewrote the way the analyzer's region_model tracks the state of memory
in GCC 11 (in
808f4dfeb3a95f50f15e71148e5c1067f90a126d), which
introduced a store with a binding_map class, mapping binding keys to
symbolic values.
The GCC 11 implementation of binding keys has an enum binding_kind,
which can be "default" vs "direct"; the idea being that direct
bindings take priority over default bindings, where the latter could
be used to represent e.g. a zero-fill of a buffer, and the former
expresses those subregions that have since been touched.
This doesn't work well: it doesn't express the idea of filling
different subregions with different values, or a memset that only
touches part of a buffer, leading to numerous XFAILs in the memset
test cases (and elsewhere).
As preparatory work towards tracking uninitialized values, this patch
eliminates the enum binding_kind, so that all bindings have
equal weight; the order in which they happen is all that matters.
If a write happens which partially overwrites an existing binding,
the new code can partially overwrite a binding, potentially punching a
hole so that an existing binding is split into two parts.
The patch adds some new classes:
- a new "bits_within_svalue" symbolic value to support extracting
parts of an existing value when its binding is partially clobbered
- a new "repeated_svalue" symbolic value to better express filling
a region with repeated copies of a symbolic value (e.g. constant
zero)
- a new "sized_region" region to express accessing a subregion
with a symbolic size in bytes
and it rewrites e.g. how memset is implemented, so that we can precisely
track which bits in a region have not been touched.
That said, the patch doesn't actually implement "uninitialized" values;
I'm saving that for a followup.
gcc/analyzer/ChangeLog:
PR analyzer/95006
* analyzer.h (class repeated_svalue): New forward decl.
(class bits_within_svalue): New forward decl.
(class sized_region): New forward decl.
(get_field_at_bit_offset): New forward decl.
* engine.cc (exploded_graph::get_or_create_node): Validate the
merged state.
(exploded_graph::maybe_process_run_of_before_supernode_enodes):
Validate the states at each stage.
* program-state.cc (program_state::validate): Validate
m_region_model.
* region-model-impl-calls.cc (region_model::impl_call_memset):
Replace special-case logic for handling constant sizes with
a call to fill_region of a sized_region with the given fill value.
* region-model-manager.cc (maybe_undo_optimize_bit_field_compare):
Drop DK_direct.
(region_model_manager::maybe_fold_sub_svalue): Fold element-based
subregions of an initial value into initial values of an element.
Fold subvalues of repeated svalues.
(region_model_manager::maybe_fold_repeated_svalue): New.
(region_model_manager::get_or_create_repeated_svalue): New.
(get_bit_range_for_field): New.
(get_byte_range_for_field): New.
(get_field_at_byte_range): New.
(region_model_manager::maybe_fold_bits_within_svalue): New.
(region_model_manager::get_or_create_bits_within): New.
(region_model_manager::get_sized_region): New.
(region_model_manager::log_stats): Update for addition of
m_repeated_values_map, m_bits_within_values_map, and
m_sized_regions.
* region-model.cc (region_model::validate): New.
(region_model::on_assignment): Drop enum binding_kind.
(region_model::get_initial_value_for_global): Likewise.
(region_model::get_rvalue_for_bits): Replace body with call to
get_or_create_bits_within.
(region_model::get_capacity): Handle RK_SIZED.
(region_model::set_value): Drop enum binding_kind.
(region_model::fill_region): New.
(region_model::get_representative_path_var_1): Handle RK_SIZED.
* region-model.h (visitor::visit_repeated_svalue): New.
(visitor::visit_bits_within_svalue): New.
(region_model_manager::get_or_create_repeated_svalue): New decl.
(region_model_manager::get_or_create_bits_within): New decl.
(region_model_manager::get_sized_region): New decl.
(region_model_manager::maybe_fold_repeated_svalue): New decl.
(region_model_manager::maybe_fold_bits_within_svalue): New decl.
(region_model_manager::repeated_values_map_t): New typedef.
(region_model_manager::m_repeated_values_map): New field.
(region_model_manager::bits_within_values_map_t): New typedef.
(region_model_manager::m_bits_within_values_map): New field.
(region_model_manager::m_sized_regions): New field.
(region_model::fill_region): New decl.
* region.cc (region::get_base_region): Handle RK_SIZED.
(region::base_region_p): Likewise.
(region::get_byte_size_sval): New.
(get_field_at_bit_offset): Make non-static.
(region::calc_offset): Move implementation of cases to
get_relative_concrete_offset vfunc implementations. Handle
RK_SIZED.
(region::get_relative_concrete_offset): New.
(decl_region::get_svalue_for_initializer): Drop enum binding_kind.
(field_region::get_relative_concrete_offset): New, from
region::calc_offset.
(element_region::get_relative_concrete_offset): Likewise.
(offset_region::get_relative_concrete_offset): Likewise.
(sized_region::accept): New.
(sized_region::dump_to_pp): New.
(sized_region::get_byte_size): New.
(sized_region::get_bit_size): New.
* region.h (enum region_kind): Add RK_SIZED.
(region::dyn_cast_sized_region): New.
(region::get_byte_size): Make virtual.
(region::get_bit_size): Likewise.
(region::get_byte_size_sval): New decl.
(region::get_relative_concrete_offset): New decl.
(field_region::get_relative_concrete_offset): New decl.
(element_region::get_relative_concrete_offset): Likewise.
(offset_region::get_relative_concrete_offset): Likewise.
(class sized_region): New.
* store.cc (binding_kind_to_string): Delete.
(binding_key::make): Drop enum binding_kind.
(binding_key::dump_to_pp): Delete.
(binding_key::cmp_ptrs): Drop enum binding_kind.
(bit_range::contains_p): New.
(byte_range::dump): New.
(byte_range::contains_p): New.
(byte_range::cmp): New.
(concrete_binding::dump_to_pp): Drop enum binding_kind.
(concrete_binding::cmp_ptr_ptr): Likewise.
(symbolic_binding::dump_to_pp): Likewise.
(symbolic_binding::cmp_ptr_ptr): Likewise.
(binding_map::apply_ctor_val_to_range): Likewise.
(binding_map::apply_ctor_pair_to_child_region): Likewise.
(binding_map::get_overlapping_bindings): New.
(binding_map::remove_overlapping_bindings): New.
(binding_cluster::validate): New.
(binding_cluster::bind): Drop enum binding_kind.
(binding_cluster::bind_compound_sval): Likewise.
(binding_cluster::purge_region): Likewise.
(binding_cluster::zero_fill_region): Reimplement in terms of...
(binding_cluster::fill_region): New.
(binding_cluster::mark_region_as_unknown): Drop enum binding_kind.
(binding_cluster::get_binding): Likewise.
(binding_cluster::get_binding_recursive): Likewise.
(binding_cluster::get_any_binding): Likewise.
(binding_cluster::maybe_get_compound_binding): Reimplement.
(binding_cluster::get_overlapping_bindings): Delete.
(binding_cluster::remove_overlapping_bindings): Reimplement in
terms of binding_map::remove_overlapping_bindings.
(binding_cluster::can_merge_p): Update for removal of
enum binding_kind.
(binding_cluster::on_unknown_fncall): Drop enum binding_kind.
(binding_cluster::maybe_get_simple_value): Likewise.
(store_manager::get_concrete_binding): Likewise.
(store_manager::get_symbolic_binding): Likewise.
(store::validate): New.
(store::set_value): Drop enum binding_kind.
(store::zero_fill_region): Reimplement in terms of...
(store::fill_region): New.
(selftest::test_binding_key_overlap): Drop enum binding_kind.
* store.h (enum binding_kind): Delete.
(binding_kind_to_string): Delete decl.
(binding_key::make): Drop enum binding_kind.
(binding_key::dump_to_pp): Make pure virtual.
(binding_key::get_kind): Delete.
(binding_key::mark_deleted): Delete.
(binding_key::mark_empty): Delete.
(binding_key::is_deleted): Delete.
(binding_key::is_empty): Delete.
(binding_key::binding_key): Delete.
(binding_key::impl_hash): Delete.
(binding_key::impl_eq): Delete.
(binding_key::m_kind): Delete.
(bit_range::get_last_bit_offset): New.
(bit_range::contains_p): New.
(byte_range::contains_p): New.
(byte_range::operator==): New.
(byte_range::get_start_byte_offset): New.
(byte_range::get_next_byte_offset): New.
(byte_range::get_last_byte_offset): New.
(byte_range::as_bit_range): New.
(byte_range::cmp): New.
(concrete_binding::concrete_binding): Drop enum binding_kind.
(concrete_binding::hash): Likewise.
(concrete_binding::operator==): Likewise.
(concrete_binding::mark_deleted): New.
(concrete_binding::mark_empty): New.
(concrete_binding::is_deleted): New.
(concrete_binding::is_empty): New.
(default_hash_traits<ana::concrete_binding>::empty_zero_p): Make false.
(symbolic_binding::symbolic_binding): Drop enum binding_kind.
(symbolic_binding::hash): Likewise.
(symbolic_binding::operator==): Likewise.
(symbolic_binding::mark_deleted): New.
(symbolic_binding::mark_empty): New.
(symbolic_binding::is_deleted): New.
(symbolic_binding::is_empty): New.
(binding_map::remove_overlapping_bindings): New decl.
(binding_map::get_overlapping_bindings): New decl.
(binding_cluster::validate): New decl.
(binding_cluster::bind): Drop enum binding_kind.
(binding_cluster::fill_region): New decl.
(binding_cluster::get_binding): Drop enum binding_kind.
(binding_cluster::get_binding_recursive): Likewise.
(binding_cluster::get_overlapping_bindings): Delete.
(store::validate): New decl.
(store::set_value): Drop enum binding_kind.
(store::fill_region): New decl.
(store_manager::get_concrete_binding): Drop enum binding_kind.
(store_manager::get_symbolic_binding): Likewise.
* svalue.cc (svalue::cmp_ptr): Handle SK_REPEATED and
SK_BITS_WITHIN.
(svalue::extract_bit_range): New.
(svalue::maybe_fold_bits_within): New.
(constant_svalue::maybe_fold_bits_within): New.
(unknown_svalue::maybe_fold_bits_within): New.
(unaryop_svalue::maybe_fold_bits_within): New.
(repeated_svalue::repeated_svalue): New.
(repeated_svalue::dump_to_pp): New.
(repeated_svalue::accept): New.
(repeated_svalue::all_zeroes_p): New.
(repeated_svalue::maybe_fold_bits_within): New.
(bits_within_svalue::bits_within_svalue): New.
(bits_within_svalue::dump_to_pp): New.
(bits_within_svalue::maybe_fold_bits_within): New.
(bits_within_svalue::accept): New.
(bits_within_svalue::implicitly_live_p): New.
(compound_svalue::maybe_fold_bits_within): New.
* svalue.h (enum svalue_kind): Add SK_REPEATED and SK_BITS_WITHIN.
(svalue::dyn_cast_repeated_svalue): New.
(svalue::dyn_cast_bits_within_svalue): New.
(svalue::extract_bit_range): New decl.
(svalue::maybe_fold_bits_within): New vfunc decl.
(region_svalue::key_t::mark_empty): Use 2 rather than NULL_TREE.
(region_svalue::key_t::is_empty): Likewise.
(default_hash_traits<region_svalue::key_t>::empty_zero_p): Make false.
(constant_svalue::maybe_fold_bits_within): New.
(unknown_svalue::maybe_fold_bits_within): New.
(poisoned_svalue::key_t::mark_empty): Use 2 rather than NULL_TREE.
(poisoned_svalue::key_t::is_empty): Likewise.
(default_hash_traits<poisoned_svalue::key_t>::empty_zero_p): Make
false.
(setjmp_svalue::key_t::mark_empty): Use 2 rather than NULL_TREE.
(setjmp_svalue::key_t::is_empty): Likewise.
(default_hash_traits<setjmp_svalue::key_t>::empty_zero_p): Make
false.
(unaryop_svalue::key_t::mark_empty): Use 2 rather than NULL_TREE.
(unaryop_svalue::key_t::is_empty): Likewise.
(unaryop_svalue::maybe_fold_bits_within): New.
(default_hash_traits<unaryop_svalue::key_t>::empty_zero_p): Make
false.
(binop_svalue::key_t::mark_empty): Use 2 rather than NULL_TREE.
(binop_svalue::key_t::is_empty): Likewise.
(default_hash_traits<binop_svalue::key_t>::empty_zero_p): Make
false.
(sub_svalue::key_t::mark_empty): Use 2 rather than NULL_TREE.
(sub_svalue::key_t::is_empty): Likewise.
(default_hash_traits<sub_svalue::key_t>::empty_zero_p): Make
false.
(class repeated_svalue): New.
(is_a_helper <const repeated_svalue *>::test): New.
(struct default_hash_traits<repeated_svalue::key_t>): New.
(class bits_within_svalue): New.
(is_a_helper <const bits_within_svalue *>::test): New.
(struct default_hash_traits<bits_within_svalue::key_t>): New.
(widening_svalue::key_t::mark_empty): Use 2 rather than NULL_TREE.
(widening_svalue::key_t::is_empty): Likewise.
(default_hash_traits<widening_svalue::key_t>::empty_zero_p): Make
false.
(compound_svalue::key_t::mark_empty): Use 2 rather than NULL_TREE.
(compound_svalue::key_t::is_empty): Likewise.
(compound_svalue::maybe_fold_bits_within): New.
(default_hash_traits<compound_svalue::key_t>::empty_zero_p): Make
false.
gcc/testsuite/ChangeLog:
PR analyzer/95006
* gcc.dg/analyzer/clobbers-1.c: New test.
* gcc.dg/analyzer/clobbers-2.c: New test.
* gcc.dg/analyzer/data-model-1.c (test_26): Mark xfail as fixed.
(test_28): Likewise.
(test_52): Likewise. Add coverage for end of buffer.
* gcc.dg/analyzer/explode-1.c: Add leak warning.
* gcc.dg/analyzer/memset-1.c (test_3): Mark xfail as fixed.
(test_4): Use char. Mark xfail as fixed.
(test_6b): New.
(test_7): Mark xfail as fixed. Add coverage for start of buffer.
(test_8): New.
(test_9): New.
* gcc.dg/analyzer/memset-CVE-2017-18549-1.c: New test.
* gcc.dg/analyzer/symbolic-8.c: New test.
Signed-off-by: David Malcolm <dmalcolm@redhat.com>
Tobias Burnus [Wed, 30 Jun 2021 12:49:48 +0000 (14:49 +0200)]
gcc.c: Add -foffload= to display_help
gcc/ChangeLog:
* common.opt (foffload): Remove help as Driver only.
* gcc.c (display_help): Add -foffload.
Christophe Lyon [Wed, 30 Jun 2021 11:44:00 +0000 (11:44 +0000)]
[testsuite]: Add missing dg-add-options float16 to gcc.dg/debug/ctf/ctf-skip-types-2.c
The test already checks dg-require-effective-target float16, but this
is not sufficient to use the flags needed, if any.
This patch makes the test pass on arm.
2021-06-30 Christophe Lyon <christophe.lyon@foss.st.com>
gcc/testsuite/
* gcc.dg/debug/ctf/ctf-skip-types-2.c: Add dg-add-options float16.
Tobias Burnus [Wed, 30 Jun 2021 11:17:54 +0000 (13:17 +0200)]
gcc.c's check_offload_target_name: Fixes to inform hints
gcc/ChangeLog:
* gcc.c (close_at_file, execute): Replace alloca by XALLOCAVEC.
(check_offload_target_name): Fix splitting OFFLOAD_TARGETS into
a candidate list; better inform no offload target is configured
and fix hint extraction when passed target is not '\0' at [len].
* common.opt (foffload): Add tailing '.'.
(foffload-options): Likewise; fix flag name in the help string.
prathamesh.kulkarni [Wed, 30 Jun 2021 09:42:06 +0000 (15:12 +0530)]
arm/66791: Gate comparison in vca intrinsics on __FAST_MATH__.
gcc/ChangeLog:
PR target/66791
* config/arm/arm_neon.h: Move vabs intrinsics before vcage_f32.
(vcage_f32): Gate comparison on __FAST_MATH__.
(vcageq_f32): Likewise.
(vcale_f32): Likewise.
(vcaleq_f32): Likewise.
(vcagt_f32): Likewise.
(vcagtq_f32): Likewise.
(vcalt_f32): Likewise.
(vcaltq_f32): Likewise.
(vcage_f16): Likewise.
(vcageq_f16): Likewise.
(vcale_f16): Likewise.
(vcaleq_f16): Likewise.
(vcagt_f16): Likewise.
(vcagtq_f16): Likewise.
(vcalt_f16): Likewise.
(vcaltq_f16): Likewise.
Richard Biener [Wed, 30 Jun 2021 07:48:37 +0000 (09:48 +0200)]
tree-optimization/101264 - rework SLP "any" permute forward prop
This integrates the forward propagation of SLP "any" permutes into
the main propagation stage as a separate single-pass propagation
didn't work out.
2021-06-30 Richard Biener <rguenther@suse.de>
PR tree-optimization/101264
* tree-vect-slp.c (vect_optimize_slp): Propagate the
computed perm_in to all "any" permute successors
we cannot de-duplicate immediately.
* gfortran.dg/pr101264.f90: New testcase.
Xi Ruoyao [Wed, 30 Jun 2021 02:47:17 +0000 (10:47 +0800)]
MAINTAINERS: Add myself for write after approval
ChangeLog:
* MAINTAINERS (Write After Approval): Add myself.
Xi Ruoyao [Mon, 28 Jun 2021 05:54:58 +0000 (13:54 +0800)]
fixinc: don't "fix" machine names in __has_include(...) [PR91085]
fixincludes/
PR other/91085
* fixfixes.c (check_has_inc): New static function.
(machine_name_fix): Don't replace header names in
__has_include(...).
* inclhack.def (machine_name): Adjust test.
* tests/base/testing.h: Update.
liuhongt [Mon, 28 Jun 2021 11:27:23 +0000 (19:27 +0800)]
The upper bits of FIXUPIMMS{S,D} should come from src1 not dest.
gcc/ChangeLog:
PR target/101248
* config/i386/sse.md
(avx512f_sfixupimm<mode><sd_maskz_name><round_saeonly_name>):
Refined to ..
(avx512f_sfixupimm<mode><maskz_scalar_name><round_saeonly_name>):
this.
(avx512f_sfixupimm<mode>_mask<round_saeonly_name>"): Refined.
* config/i386/subst.md (maskz_scalar): New define_subst.
(maskz_scalar_name): New subst_attr.
(maskz_scalar_op5): Ditto.
(round_saeonly_maskz_scalar_op5): Ditto.
(round_saeonly_maskz_scalar_operand5): Ditto.
gcc/testsuite/ChangeLog
PR target/101248
* gcc.target/i386/pr101248.c: New test.
David Edelsohn [Wed, 12 May 2021 01:07:19 +0000 (21:07 -0400)]
aix: align text CSECTs to at least 32 bytes.
gcc/ChangeLog:
* config/rs6000/rs6000.c (rs6000_xcoff_section_type_flags):
Increase code CSECT alignment to at least 32 bytes.
* config/rs6000/xcoff.h (TEXT_SECTION_ASM_OP): Add 32 byte
alignment designation.
GCC Administrator [Wed, 30 Jun 2021 00:16:52 +0000 (00:16 +0000)]
Daily bump.
Sergei Trofimovich [Tue, 29 Jun 2021 21:11:39 +0000 (22:11 +0100)]
docs: Fix s/net yet/not yet/ typo
gcc/ChangeLog:
* doc/generic.texi: Fix s/net yet/not yet/ typo.
Jason Merrill [Tue, 29 Jun 2021 19:11:25 +0000 (15:11 -0400)]
c++: don't treat member var as var template
While looking at a partial instantiation issue I noticed that we were
wrongly hitting the partial instantiation code when instantiating a static
data member of a class template. I don't think this broke anything, but we
don't need to do that (small) extra work.
gcc/cp/ChangeLog:
* pt.c (instantiate_decl): Only consider partial specializations of
actual variable templates.
Ian Lance Taylor [Tue, 29 Jun 2021 19:53:02 +0000 (12:53 -0700)]
compiler: don't generate temporaries for composite literals
We were generating temporaries for composite literals when a
conversion to interface type was required. However, Cherry's
https://golang.org/cl/176459 changed the compiler to insert explicit
type conversions. And those explicit type conversions insert the
required temporaries in Type_conversion_expression::do_flatten.
So in practice the composite literal do_flatten methods would never
insert temporaries, as the values they see would always be multi_eval_safe.
So just remove the unnecessary do_flatten methods.
Reviewed-on: https://go-review.googlesource.com/c/gofrontend/+/331691
Ian Lance Taylor [Tue, 29 Jun 2021 19:12:16 +0000 (12:12 -0700)]
go-gcc: set DECL_NAMELESS for temporary variables
* go-gcc.cc (Gcc_backend::static_chain_variable): Set
DECL_NAMELESS on the new decl.
(Gcc_backend::temporary_variable): Likewise.
(Gcc_backend::function): Set DECL_NAMELESS on the result decl.
Andrew MacLeod [Tue, 29 Jun 2021 14:52:58 +0000 (10:52 -0400)]
Fix MINUS_EXPR relations.
Flesh out and correct relations for both wrapping and non-wrapping values.
gcc/
PR tree-optimization/101254
* range-op.cc (operator_minus::op1_op2_relation_effect): Check for
wrapping/non-wrapping when setting the result range.
gcc/testsuite
* gcc.dg/pr101254.c: New.
Andrew MacLeod [Fri, 25 Jun 2021 19:31:39 +0000 (15:31 -0400)]
Allow PHIs to pick up global values.
We can also apply known global values to PHI nodes in EVRP.
* value-query.cc (gimple_range_global): Allow phis.
Andrew MacLeod [Fri, 25 Jun 2021 15:24:30 +0000 (11:24 -0400)]
Add stmt context in simplify_using_ranges.
There were places simplify_using_ranges was not utilzing the stmt context.
* vr-values.c (vr_values::vrp_stmt_computes_nonzero): Use stmt.
(simplify_using_ranges::op_with_boolean_value_range_p): Add a
statement for location context.
(check_for_binary_op_overflow): Ditto.
(simplify_using_ranges::get_vr_for_comparison): Ditto.
(simplify_using_ranges::compare_name_with_value): Ditto.
(simplify_using_ranges::compare_names): Ditto.
(vrp_evaluate_conditional_warnv_with_ops_using_ranges): Ditto.
(simplify_using_ranges::simplify_truth_ops_using_ranges): Ditto.
(simplify_using_ranges::simplify_min_or_max_using_ranges): Ditto.
(simplify_using_ranges::simplify_internal_call_using_ranges): Ditto.
(simplify_using_ranges::two_valued_val_range_p): Ditto.
(simplify_using_ranges::simplify): Ditto.
* vr-values.h: Adjust prototypes.
Ian Lance Taylor [Mon, 28 Jun 2021 23:47:55 +0000 (16:47 -0700)]
compiler: in composite literals use temps only for interfaces
For a composite literal we only need to introduce a temporary variable
if we may be converting to an interface type, so only do it then.
This saves over 80% of compilation time when using gccgo to compile
cmd/internal/obj/x86, as the GCC middle-end spends a lot of time
pointlessly computing interactions between temporary variables.
For PR debug/101064
For golang/go#46600
Reviewed-on: https://go-review.googlesource.com/c/gofrontend/+/331513
Uros Bizjak [Tue, 29 Jun 2021 17:14:35 +0000 (19:14 +0200)]
i386: Add V2SFmode vec_addsub pattern [PR95046]
gcc/
2021-06-21 Uroš Bizjak <ubizjak@gmail.com>
PR target/95046
* config/i386/mmx.md (vec_addsubv2sf3): New insn pattern.
gcc/testsuite/
2021-06-21 Uroš Bizjak <ubizjak@gmail.com>
PR target/95046
* gcc.target/i386/pr95046-9.c: New test.
Julian Brown [Fri, 11 Jun 2021 21:44:36 +0000 (14:44 -0700)]
Fortran: Re-enable 128-bit integers for AMD GCN
This patch reverts the part of Tobias's patch for PR target/96306 that
disables 128-bit integer support for AMD GCN.
2021-06-18 Julian Brown <julian@codesourcery.com>
libgfortran/
PR target/96306
* configure.ac: Remove stanza that removes KIND=16 integers for AMD GCN.
* configure: Regenerate.
Julian Brown [Wed, 9 Jun 2021 13:18:23 +0000 (06:18 -0700)]
amdgcn: Enable support for TImode for AMD GCN
This patch enables support for TImode for AMD GCN, the lack of which
is currently causing a number of test failures for the target and which
is also needed to support "omp_depend_kind" for OpenMP 5.0, since that
is implemented as a 128-bit integer.
Several libgcc support routines are built by default for the "word size"
of a machine, and also for "2 * word size" of the machine. The libgcc
build for AMD GCN is changed so that it builds for a "word size" of 64
bits, in order to better match the (64-bit) host compiler. However it
isn't really true that we have 64-bit words -- GCN has 32-bit registers,
so changing UNITS_PER_WORD unconditionally would be the wrong thing to do.
Changing this setting for libgcc (only) means that support routines
are built for "single word" operations that are DImode (64 bits), and
those for "double word" operations are built for TImode (128 bits).
That leaves some gaps regarding previous operations that were built
for a "single word" size of 32 bits and a "double word" size of 64 bits
(generic code doesn't cover both alternatives for all operations that
might be needed). Those gaps are filled in by this patch, or by the
preceding patches in the series.
2021-06-18 Julian Brown <julian@codesourcery.com>
gcc/
* config/gcn/gcn.c (gcn_init_libfuncs): New function.
(TARGET_INIT_LIBFUNCS): Define target hook using above function.
* config/gcn/gcn.h (UNITS_PER_WORD): Define to 8 for IN_LIBGCC2, 4
otherwise.
(LIBGCC2_UNITS_PER_WORD, BITS_PER_WORD): Remove definitions.
(MAX_FIXED_MODE_SIZE): Change to 128.
libgcc/
* config/gcn/lib2-bswapti2.c: New file.
* config/gcn/lib2-divmod-di.c: New file.
* config/gcn/lib2-gcn.h (DItype, UDItype, TItype, UTItype): Add
typedefs.
(__divdi3, __moddi3, __udivdi3, __umoddi3): Add prototypes.
* config/gcn/t-amdgcn (LIB2ADD): Add lib2-divmod-di.c and
lib2-bswapti2.c.
Julian Brown [Thu, 17 Jun 2021 17:52:51 +0000 (10:52 -0700)]
amdgcn: Add clrsbsi2/clrsbdi2 implementation
This patch adds an open-coded implementation of the clrsb<mode>2
(count leading redundant sign bit) standard names using the GCN flbit_i*
instructions for SImode and DImode. Those don't count exactly as we need,
so we need a couple of other instructions to fix up the result afterwards.
These patterns are lost from libgcc if we build it for DImode/TImode
rather than SImode/DImode, a change we make in a later patch in this
series.
2021-06-18 Julian Brown <julian@codesourcery.com>
gcc/
* config/gcn/gcn.md (UNSPEC_FLBIT_INT): New unspec constant.
(s_mnemonic): Add clrsb.
(gcn_flbit<mode>_int): Add insn pattern for SImode/DImode.
(clrsb<mode>2): Add expander for SImode/DImode.
Julian Brown [Tue, 29 Jun 2021 10:57:31 +0000 (03:57 -0700)]
amdgcn: Add [us]mulsid3/muldi3 patterns
This patch improves 64-bit multiplication for AMD GCN: patterns for
unsigned and signed 32x32->64 bit multiplication have been added, and
also 64x64->64 bit multiplication is now open-coded rather than calling
a library function (which may be a win for code size as well as speed:
the function calling sequence isn't particularly concise for GCN).
This version of the patch uses define_insn_and_split in order to keep
multiply operations together during RTL optimisations up to register
allocation: this appears to produce more compact code via inspection on
small test cases than the previous approach using an expander.
The DImode multiply implementation is lost from libgcc if we build it
for DImode/TImode rather than SImode/DImode, a change we make in a later
patch in this series.
2021-06-29 Julian Brown <julian@codesourcery.com>
gcc/
* config/gcn/gcn.md (<su>mulsidi3, <su>mulsidi3_reg, <su>mulsidi3_imm,
muldi3): Add patterns.
Julian Brown [Thu, 17 Jun 2021 18:02:55 +0000 (11:02 -0700)]
amdgcn: Add [us]mulsi3_highpart SGPR alternatives
This patch splits the <su>mulsi3_highpart pattern into an expander and
register/immediate alternatives (to avoid meaningless sign/zero_extends on
constants), and adds alternatives for SGPR high-part multiply instructions
on GCN5+.
2021-06-29 Julian Brown <julian@codesourcery.com>
gcc/
* config/gcn/gcn.md (<su>mulsi3_highpart): Change to expander.
(<su>mulsi3_highpart_reg, <su>mulsi3_highpart_imm): New patterns.
Julian Brown [Mon, 28 Jun 2021 13:58:52 +0000 (06:58 -0700)]
amdgcn: Mark s_mulk_i32 as clobbering SCC
The s_mulk_i32 instruction sets the SCC status register according to
whether the multiplication overflows, but that is not currently modelled
in the GCN backend. AFAIK this is a latent bug and hasn't been noticed
"in the wild", but it should be fixed.
2021-06-29 Julian Brown <julian@codesourcery.com>
gcc/
* config/gcn/gcn.md (mulsi3): Make s_mulk_i32 variant clobber SCC.
Julian Brown [Thu, 17 Jun 2021 17:55:48 +0000 (10:55 -0700)]
amdgcn: Use unsigned types for udivsi3/umodsi3 libgcc helper args/return
This patch changes the argument and return types for the libgcc __udivsi3
and __umodsi3 helper functions for GCN to USItype instead of SItype.
This is probably just cosmetic in practice.
2021-06-18 Julian Brown <julian@codesourcery.com>
libgcc/
* config/gcn/lib2-divmod.c (__udivsi3, __umodsi3): Change argument and
return types to USItype.
* config/gcn/lib2-gcn.h (__udivsi3, __umodsi3): Update prototypes.
Joseph Myers [Tue, 29 Jun 2021 15:10:17 +0000 (15:10 +0000)]
bootstrap: Include memmodel.h in btfout.c and ctfout.c before tm_p.h
This fixes a "use of enum 'memmodel' without previous declaration"
error in sparc-protos.h.
Minimally tested that this fixes the build-many-glibcs.py compilers
build for sparc64-linux-gnu.
* btfout.c, ctfout.c: Include "memmodel.h".
Tobias Burnus [Tue, 29 Jun 2021 14:35:42 +0000 (16:35 +0200)]
gcc.c: Silence warning in check_offload_target_name
gcc/ChangeLog:
* gcc.c (check_offload_target_name): Cast len argument to
%q.*s to 'int'; avoid -Wstringop-truncation warning.
Richard Kenner [Wed, 21 Apr 2021 12:35:24 +0000 (08:35 -0400)]
[Ada] Make copies of entities being declared when copying block
gcc/ada/
* sem_util.adb (Visit_Node): Add handling for N_Block_Statement
with declarations.
Piotr Trojanek [Tue, 20 Apr 2021 08:53:06 +0000 (10:53 +0200)]
[Ada] Expose symmetry between Known_ and Unknown_ query routines
gcc/ada/
* einfo-utils.adb
(Unknown_Alignment): Simply negate the Known_ counterpart.
(Unknown_Component_Bit_Offset): Likewise.
(Unknown_Esize): Likewise.
(Unknown_Normalized_First_Bit): Likewise.
(Unknown_Normalized_Position): Likewise.
(Unknown_Normalized_Position_Max): Likewise.
(Unknown_RM_Size): Likewise.
Boris Yakobowski [Mon, 19 Apr 2021 10:17:07 +0000 (12:17 +0200)]
[Ada] Do not catch 'N mod -1' in CodePeer_Mode
gcc/ada/
* exp_ch4.adb (Expand_N_Op_Mod): Remove special case for mod -1
in CodePeer_Mode.
Aleksandra Pasek [Wed, 14 Apr 2021 14:44:04 +0000 (10:44 -0400)]
[Ada] Add support for Aarch64 in System.Object_Reader
gcc/ada/
* libgnat/s-objrea.adb (EM_AARCH64): New Constant.
(Initialize): Handle EM_AARCH64 case.
(Read_Address): Handle AARCH64 case.
* libgnat/s-objrea.ads (Object_Arch): Add AARCH64 record
component.
Bob Duff [Sun, 18 Apr 2021 13:45:04 +0000 (09:45 -0400)]
[Ada] Fix the -gnatyr switch so it works in record rep clauses
gcc/ada/
* sem_ch13.adb (Analyze_Record_Representation_Clause): Call
Set_Entity_With_Checks instead of Set_Entity, so we perform the
check for correct casing.
* style.adb (Check_Identifier): Minor comment improvement.
Cleanup overly complicated code.
Ed Schonberg [Fri, 16 Apr 2021 16:52:50 +0000 (12:52 -0400)]
[Ada] Crash on limited array object with address clause
gcc/ada/
* exp_aggr.adb (Convert_Aggr_In_Object_Decl): After expansion of
the aggregate, the expression can be removed from the
declaration, except if the object is class-wide, in which case
the aggregate provides the actual type. In other cases the
presence of the expression may lead to spurious freezing issue.
* exp_ch3.adb (Expand_N_Object_Declaration): If the expression
in the declaration is an aggregate with delayed expansion (as is
the case for objects of a limited type, or a subsequent address
specification) the aggregate must be resolved at this point.
This resolution must not include expansion, because the
expansion of the enclosing declaration will construct the
necessary aggregate expansion.
Piotr Trojanek [Thu, 15 Apr 2021 16:00:52 +0000 (18:00 +0200)]
[Ada] Remove an obsolete variant of Adjust_Name_Case used only by SPARK
gcc/ada/
* errout.ads (Adjust_Name_Case): Remove obsolete and now unused
variant.
* errout.adb (Adjust_Name_Case): Likewise; fix variant that uses
a custom buffer to also use it for names in Standard_Location.
Eric Botcazou [Thu, 15 Apr 2021 15:22:46 +0000 (17:22 +0200)]
[Ada] Small tweak in a couple of comments
gcc/ada/
* sem_ch12.adb (Freeze_Subprogram_Body): Add missing "freeze".
(Install_Body): Likewise.
Piotr Trojanek [Thu, 15 Apr 2021 12:17:48 +0000 (14:17 +0200)]
[Ada] Remove a special case for forking-for-expect from ordinary spawn
gcc/ada/
* adaint.c (__gnat_portable_spawn): Revert change that
introduced setting of __gnat_in_child_after_fork.
Patrick Bernardi [Wed, 14 Apr 2021 03:50:57 +0000 (13:50 +1000)]
[Ada] Ensure System.Tasking.Debug.Known_Tasks component access is atomic
gcc/ada/
* libgnarl/s-tasdeb.ads (Known_Tasks): Add Atomic_Components
aspect.
Doug Rupp [Mon, 12 Apr 2021 19:03:48 +0000 (12:03 -0700)]
[Ada] vx7-shared-libs: Unused variable __gnat_user_int_connect
gcc/ada/
* Makefile.rtl (x86_64-vx7r2) [EXTRA_GNATRTL_TASKING_OBJS]: Move
i-vxinco.o out of RTP runtime.
Claire Dross [Mon, 12 Apr 2021 07:48:48 +0000 (09:48 +0200)]
[Ada] Disable Pre/Post in formal containers
gcc/ada/
* libgnat/a-cfdlli.ads: Use pragma Assertion_Policy to disable
pre and postconditions.
* libgnat/a-cfhama.ads: Likewise.
* libgnat/a-cfhase.ads: Likewise.
* libgnat/a-cfinve.ads: Likewise.
* libgnat/a-cforma.ads: Likewise.
* libgnat/a-cforse.ads: Likewise.
* libgnat/a-cofove.ads: Likewise.
Piotr Trojanek [Tue, 6 Apr 2021 11:11:09 +0000 (13:11 +0200)]
[Ada] Fix style in Get_Fullest_View
gcc/ada/
* sem_util.ads (Get_Fullest_View): Refill comment; remove extra
extra after period.
* sem_util.adb (Get_Fullest_View): Fix style.
Piotr Trojanek [Thu, 8 Apr 2021 19:19:54 +0000 (21:19 +0200)]
[Ada] Remove redundant check for empty list
gcc/ada/
* sem_ch3.adb (Analyze_Declarations): Remove explicit check for
missing, because a subsequent call to Is_Empty_List will detect
them anyway.
Piotr Trojanek [Thu, 8 Apr 2021 19:17:33 +0000 (21:17 +0200)]
[Ada] Reuse Is_Subprogram_Or_Entry where possible
gcc/ada/
* freeze.adb (Freeze_All): Simplify by reusing
Is_Subprogram_Or_Entry.
* sem_ch11.adb (Analyze_Handled_Statement): Likewise.
Piotr Trojanek [Mon, 12 Apr 2021 14:56:23 +0000 (16:56 +0200)]
[Ada] More robust guard against cascaded errors with overlapping actuals
gcc/ada/
* sem_warn.adb (Warn_On_Overlapping_Actuals): Prevent cascaded
errors once for the subprogram call, not for every pair of
actual parameters.
Piotr Trojanek [Tue, 6 Apr 2021 11:10:01 +0000 (13:10 +0200)]
[Ada] Simplify detection of local types
gcc/ada/
* sem_ch3.adb (Is_Local_Type): Simplify by reusing Scope_Within.