Frederik Harwath [Tue, 19 May 2020 09:15:28 +0000 (11:15 +0200)]
contrib/gcc-changelog: Skip over review lines
git-check-commit.py does not know about "Reviewed-by",
"Reviewed-on", and "Signed-off-by" lines and hence it
expects those lines which follow the ChangeLog entries
to be indented by a tab.
This commit makes the script skip those lines. No further
processing is attempted because the review information
is not part of the ChangeLogs.
contrib/
2020-05-20 Frederik Harwath <frederik@codesourcery.com>
* gcc-changelog/git_commit.py: Skip over lines starting
with "Reviewed-by: ", "Reviewed-on: ", or "Signed-off-by: "
Martin Liska [Wed, 20 May 2020 08:05:05 +0000 (10:05 +0200)]
git_check_commit: shorted option name
* gcc-changelog/git_check_commit.py: Change
--allow-non-strict-mode to --non-strict-mode.
Martin Liska [Wed, 20 May 2020 07:57:05 +0000 (09:57 +0200)]
Add gcc-backport and support git cherry pick.
* gcc-changelog/git_commit.py: Support cherry pick
prefix.
* gcc-changelog/test_email.py: Test it.
* gcc-changelog/test_patches.txt: Add new patch.
* gcc-git-customization.sh: Add gcc-backport.
Richard Biener [Tue, 19 May 2020 05:58:33 +0000 (07:58 +0200)]
c/95141 - fix bogus integer overflow warning
This fixes an integer overflow warning that ultimatively happens because
of TREE_OVERFLOW propagating through transforms and the existing guard
against this,
375 if (TREE_OVERFLOW_P (ret)
376 && !TREE_OVERFLOW_P (op0)
377 && !TREE_OVERFLOW_P (op1))
378 overflow_warning (EXPR_LOC_OR_LOC (expr, input_location,
being insufficient. Rather than trying to use sth like walk_tree to
exhaustively walk operands (with the possibility of introducing
quadraticness when folding larger expressions recursively) the
following amends the above with an ad-hoc test for a binary op0
with a possibly constant op1.
2020-05-30 Richard Biener <rguenther@suse.de>
PR c/95141
gcc/c
* c-fold.c (c_fully_fold_internal): Enhance guard on
overflow_warning.
gcc/testsuite
* gcc.dg/pr95141.c: New testcase.
Kito Cheng [Tue, 14 Apr 2020 06:53:19 +0000 (14:53 +0800)]
Fix alignment for local variable [PR90811]
- The alignment for local variable was adjust during estimate_stack_frame_size,
however it seems wrong spot to adjust that, expand phase will adjust that
but it little too late to some gimple optimization, which rely on certain
target hooks need to check alignment, forwprop is an example for
that, result of simplify_builtin_call rely on the alignment on some
target like ARM or RISC-V.
- Exclude static local var and hard register var in the process of
alignment adjustment.
- This patch fix gfortran.dg/pr45636.f90 for arm and riscv.
- Regression test on riscv32/riscv64 and x86_64-linux-gnu, no new fail
introduced.
gcc/ChangeLog
PR target/90811
* Makefile.in (OBJS): Add adjust-alignment.o.
* adjust-alignment.c (pass_data_adjust_alignment): New.
(pass_adjust_alignment): New.
(pass_adjust_alignment::execute): New.
(make_pass_adjust_alignment): New.
* tree-pass.h (make_pass_adjust_alignment): New.
* passes.def: Add pass_adjust_alignment.
GCC Administrator [Wed, 20 May 2020 00:16:23 +0000 (00:16 +0000)]
Daily bump.
Jonathan Wakely [Tue, 19 May 2020 15:49:21 +0000 (16:49 +0100)]
libstdc++: Use RDRAND as fallback if RDSEED keeps failing (PR 94087)
It's not difficult for multiple threads to drain the entropy available
to the RDSEED instruction, at which point we throw an exception. This
change will try to use RDRAND after RDSEED fails repeatedly, and only
throw if RDRAND also fails repeatedly. This doesn't guarantee a random
value can always be read, but reduces the likelihood of failure when
using the RDSEED instruction.
PR libstdc++/94087
* src/c++11/random.cc (__x86_rdseed): Allow fallback function to be
passed in.
(__x86_rdseed_rdrand): New function that uses rdseed with rdrand
fallback.
(random_device::_M_init): Use __x86_rdseed_rdrand when both
instructions are available.
* testsuite/26_numerics/random/random_device/94087.cc: New test.
H.J. Lu [Tue, 19 May 2020 21:42:12 +0000 (14:42 -0700)]
x86: Add FEATURE_AVX512VP2INTERSECT and update GFNI check
Add FEATURE_AVX512VP2INTERSECT to libgcc so that enum processor_features
in libgcc matches enum processor_features in i386-builtins.c. Update
GFNI check to support processors with SSE and AVX versions of GFNI.
PR target/95212
PR target/95220
* config/i386/cpuinfo.c (get_available_features): Fix
FEATURE_GFNI check. Also check FEATURE_AVX512VP2INTERSECT.
* config/i386/cpuinfo.h (processor_features): Add
FEATURE_AVX512VP2INTERSECT.
Nathan Sidwell [Tue, 19 May 2020 20:29:19 +0000 (13:29 -0700)]
c++: Alias template instantiation template info
I discovered that the alias instantiation machinery would setup
template_info, and then sometime later overwrite that with equivalent
info. This broke modules, because the template info, once set, is
logically immutable. Let's just not do that.
* pt.c (lookup_template_class_1): Do not reinit template_info of an
alias here.
Nathan Sidwell [Tue, 19 May 2020 20:20:32 +0000 (13:20 -0700)]
preprocessor: Random cleanups
This fixes a bunch of poorly formatted decls, marks some getters as
PURE, deletes some C-relevant bool hackery, and finally uses a
passed-in location rather than deducing a closely-related but not
necessarily the same location.
* include/cpplib.h (cpp_get_otions, cpp_get_callbacks)
(cpp_get_deps): Mark as PURE.
* include/line-map.h (get_combined_adhoc_loc)
(get_location_from_adhoc_loc, get_pure_location): Reformat decls.
* internal.h (struct lexer_state): Clarify comment.
* system.h: Remove now-unneeded bool hackery.
* files.c (_cpp_find_file): Store LOC not highest_location.
Alex Coplan [Tue, 19 May 2020 19:33:20 +0000 (20:33 +0100)]
[aarch64] PR target/94591: Don't generate invalid REV64 insns
This fixes PR94591. The problem was the function aarch64_evpc_rev_local()
matching vector permutations that were not reversals. In particular, prior to
this patch, this function matched the identity permutation which led to
generating bogus REV64 insns which were rejected by the assembler.
gcc/
PR target/94591
* config/aarch64/aarch64.c (aarch64_evpc_rev_local): Don't match
identity permutation.
gcc/testsuite/
PR target/94591
* gcc.c-torture/execute/pr94591.c: New test.
Martin Liska [Tue, 19 May 2020 19:16:27 +0000 (21:16 +0200)]
Use REST API for bug titles in mklog.
* mklog.py: Use REST API for bug title downloading.
Martin Liska [Tue, 19 May 2020 19:16:10 +0000 (21:16 +0200)]
Use commit timestamp in git_update_version.py.
* gcc-changelog/git_commit.py: Add param use_commit_ts
for to_changelog_entries.
* gcc-changelog/git_update_version.py: Se use_commit_ts to True.
Martin Sebor [Tue, 19 May 2020 18:46:37 +0000 (12:46 -0600)]
PR c++/94923 - False positive -Wclass-memaccess with trivially copyable std::optional
gcc/cp/ChangeLog:
PR c++/94923
* call.c ((maybe_warn_class_memaccess): Use is_byte_access_type.
* cp-tree.h (is_dummy_object): Return bool.
(is_byte_access_type): Declare new function.
* tree.c (is_dummy_object): Return bool.
(is_byte_access_type): Define new function.
gcc/testsuite/ChangeLog:
PR c++/94923
* g++.dg/Wclass-memaccess.C: Add tests for std::byte.
Nathan Sidwell [Tue, 19 May 2020 18:28:05 +0000 (11:28 -0700)]
preprocessor: Reimplement raw string lexing [pr95149]
pr95149 is a false positive static analysis checker. But it
encouranged me to fix raw string lexing, which does contain a
complicated macro and pointers to local variables. The
reimplementation does away with that macro. Part of the complication
is we need to undo some of the fresh line processing -- trigraph notes
and escaped line continuations. But the undone characters need to go
through the raw string processing, as they can legitimately be part of
the prefix marker. however, in this reformulation we only process one
line marker at a time[*], so there's a limited number of undone
characters. We can arrange the buffering to make sure we don't split
such an append sequence, and then simply take the characters from the
append buffer.
The prefix scanner had a switch statement, which I discovered was not
optimized as well as an if of a bunch of explicit comparisons (pr
95208 filed).
Finally I adjusted the failure mode. When we get a bad prefix, we lex
up until the next '"', thus often swallowing the whole raw string.
Previously we'd bail and then the lexer would usually generate stupid
tokens, particularly when meeting the ending '"'.
libcpp/
* lex.c (struct lit_accum): New.
(bufring_append): Replace by lit_accum::append.
(lex_raw_string): Reimplement, using fragments of the old version.
(lex_string): Adjust lex_raw_string call.
gcc/testsuite/
* c-c++-common/raw-string-14.c: Adjust errors.
* c-c++-common/raw-string-16.c: Likewise.
* c-c++-common/raw-string-5.c: Likewise.
Richard Biener [Tue, 19 May 2020 18:30:40 +0000 (20:30 +0200)]
Fix FAIL: gcc.target/i386/pr92645-4.c
This adjusts the testcase for the introduced vector promotion/demotion
support.
2020-05-19 Richard Biener <rguenther@suse.de>
* gcc.target/i386/pr92645-4.c: Adjust expected pattern.
Nathan Sidwell [Tue, 19 May 2020 13:11:22 +0000 (06:11 -0700)]
preprocessor: Fix ICE with EOF in macro args [pr95182]
This was another latent case of us losing an EOF token, but succeeding
anyway. Since my patch to make us pay more attention to EOFs it came
to light. We also need to keep the EOF if we fall off the end of the
main file. Forced includes look like regular nested includes at this
point.
PR preprocessor/95182
libcpp/
* macro.c (collect_args): Preserve EOFif we fell out of the main
file.
(cpp_get_token_1): Reformat a couple of short lines.
Jozef Lawrynowicz [Tue, 19 May 2020 11:46:17 +0000 (12:46 +0100)]
TESTSUITE: Fix tests for 16-bit targets
gcc/ChangeLog:
2020-05-19 Jozef Lawrynowicz <jozef.l@mittosystems.com>
* doc/sourcebuild.texi: Document new short_eq_int, ptr_eq_short,
msp430_small, msp430_large and size24plus DejaGNU effective
targets.
Improve grammar in descriptions for size20plus and size32plus effective
targets.
gcc/testsuite/ChangeLog:
2020-05-19 Jozef Lawrynowicz <jozef.l@mittosystems.com>
* c-c++-common/builtin-has-attribute-7.c: Require size24plus.
* c-c++-common/cpp/pr63831-1.c: Store result in _has_cpp_attribute in a
long.
* c-c++-common/pr81376.c: Skip scan-tree-dump for short_eq_int. Extend
test for short_eq_int.
* g++.dg/abi/scoped1.C: Skip dg-warning tests for short_eq_int.
* g++.dg/cpp0x/constexpr-70001-1.C: Require size24plus.
* g++.dg/cpp0x/constexpr-bitfield3.C: Require int32plus.
* g++.dg/cpp0x/enum13.C: Skip dg-warning for short_eq_int.
* g++.dg/cpp0x/initlist5.C: Add dg-error for short_eq_int.
* g++.dg/cpp0x/initlist7.C: Add dg-warning for !int32plus.
* g++.dg/cpp0x/nullptr04.C: Skip dg-error for ptr_eq_short.
* g++.dg/cpp0x/variadic-value1.C: Add typedef for int32_t.
* g++.dg/cpp1y/constexpr-arith-overflow.C: Fix test for
sizeof(int) == sizeof(short).
* g++.dg/cpp1y/digit-sep-neg.C: Add typedef for int32_t.
* g++.dg/cpp1y/pr57644.C: Add typedef for uint32_t.
* g++.dg/cpp1y/pr77321.C: Require size24plus.
* g++.dg/cpp1y/var-templ4.C: Add typedef for int32_t.
* g++.dg/cpp1z/direct-enum-init1.C: Skip dg-error for short_eq_int.
* g++.dg/delayedfold/fwrapv1.C: Skip for int16.
* g++.dg/expr/bitfield9.C: Add typedef for int32_t.
* g++.dg/ext/attribute-test-1.C: Add typedef for uint32_t.
* g++.dg/ext/bitfield1.C: Add typedef for int32_t.
* g++.dg/ext/flexary13.C: Add typedef for int32_t.
* g++.dg/ext/utf-cvt.C: Adjust dg-warning for int16.
* g++.dg/ext/vector28.C: Add typedef for int32_t.
* g++.dg/ext/vla15.C: Add typedef for int32_t.
* g++.dg/init/array11.C: Require size32plus.
* g++.dg/init/array15.C: Require size24plus.
* g++.dg/init/array4.C: Require size20plus.
* g++.dg/init/const7.C: Skip dg-message for ptr_eq_short.
* g++.dg/init/new38.C: Relax regex in dg-error.
* g++.dg/init/new44.C: Skip dg-error for msp430_small.
Adjust test for 16-bit size_t.
Add special case for msp430 -mlarge.
* g++.dg/init/value9.C: Add typedef for int32_t.
* g++.dg/ipa/pr77333.C: Add typedef for int32_t.
* g++.dg/lto/
20080908-1_0.C: Add typedef for int32_t.
* g++.dg/opt/pr55717.C: Add typedef for uint32_t.
* g++.dg/opt/pr60597.C: Add typedef for int32_t.
* g++.dg/opt/pr81715.C: Require size20plus.
* g++.dg/opt/reload3.C: Add typedef for uint32_t.
* g++.dg/opt/temp2.C: Require size20plus.
* g++.dg/opt/thunk1.C: Likewise.
* g++.dg/other/error23.C: Dont assume __SIZEOF_INT__ == 4.
* g++.dg/other/pr31078.C: Adjust typedef for 32-bit int.
* g++.dg/parse/concat1.C: Skip dg-error for size20plus.
* g++.dg/parse/defarg5.C: Add typedef for int32_t and uint32_t.
* g++.dg/pr48484.C: Add typedef for int32_t.
* g++.dg/pr53037-2.C: Likewise.
* g++.dg/pr53037-3.C: Likewise.
* g++.dg/pr66655.C: Use int32_t.
* g++.dg/pr66655.h: Add typedef for int32_t.
* g++.dg/pr66655_1.cc: Use int32_t.
* g++.dg/pr67351.C: Define 32-bit uint.
* g++.dg/template/array30.C: Add typedef for int32_t.
* g++.dg/template/constant1.C: Extend test for 8-bit and 16-bit int.
* g++.dg/template/constant2.C: Likewise.
* g++.dg/template/friend18.C: Add typedef for int32_t.
* g++.dg/template/pr68978.C: Likewise.
* g++.dg/torture/pr37421.C: Require int_eq_float.
* g++.dg/torture/pr88861.C: Handle 16-bit int.
* g++.dg/tree-ssa/pr19807.C: Likewise.
* g++.dg/tree-ssa/pr27291.C: Fix typedef for uint32_t.
* g++.dg/tree-ssa/pr49516.C: Fix typedefs for int{16,32}_t and
uint{32,64}_t.
* g++.dg/warn/Wconversion-integer.C: Add typedefs for {u,}int32_t.
* g++.dg/warn/Wconversion-null-2.C: Adjust g() declaration.
* g++.dg/warn/Wconversion-null.C: Likewise.
* g++.dg/warn/Wconversion3.C: Skip dg-warning for short_eq_int.
* g++.dg/warn/Wduplicated-branches1.C: Add dg-warning for short_eq_int.
* g++.dg/warn/Wplacement-new-size-5.C: Add typedef for int32_t.
* g++.dg/warn/Wplacement-new-size.C: Likewise.
* g++.dg/warn/Wstrict-aliasing-5.C: Add typedef for uint32_t.
* g++.dg/warn/Wstrict-aliasing-bogus-signed-unsigned.C: Add typedef for
{u,}int32_t.
* g++.dg/warn/Wtype-limits-Wextra.C: Adjust dg-warning for
short_eq_int.
* g++.dg/warn/Wtype-limits.C: Likewise.
* g++.old-deja/g++.brendan/enum11.C: Add typedef for uint32_t.
* g++.old-deja/g++.bugs/900227_01.C: Skip dg-error for ptr_eq_short.
* g++.old-deja/g++.mike/ns15.C: Require size20plus.
* g++.old-deja/g++.other/exprstmt1.C: Add typedef for uint32_t.
* g++.old-deja/g++.other/inline12.C: Adjust udword typedef.
* g++.old-deja/g++.other/new6.C: Add typedef for int32_t.
* g++.old-deja/g++.pt/crash16.C: Skip for int16.
* g++.old-deja/g++.robertl/eb76.C: Likewise.
* g++.old-deja/g++.warn/flow1.C: Add typedef for int32_t.
* gcc.dg/Walloca-14.c: Adjust -Walloca-larger-than= parameter for
!ptr32plus.
* gcc.dg/Warray-bounds-32.c: Adjust dg-warning for size20plus.
* gcc.dg/Wbuiltin-declaration-mismatch-4.c: Adjust dg-warning for
short_eq_int.
Handle case where ptrdiff_t/size_t is __int20.
* gcc.dg/concat2.c: Skip dg-error for size20plus.
* gcc.dg/fold-convmaxconv-1.c: Add typedef for {u,}int32_t.
* gcc.dg/fold-convminconv-1.c: Likewise.
* gcc.dg/graphite/scop-4.c: Require size20plus.
* gcc.dg/loop-versioning-1.c: Adjust test for small size_t.
* gcc.dg/loop-versioning-2.c: Require size20plus.
* gcc.dg/lto/
20081210-1_0.c: Adjust typedef for uintptr_t.
* gcc.dg/lto/pr85870_0.c: Add typedef for uint32_t.
* gcc.dg/lto/pr85870_1.c: Likewise.
* gcc.dg/pr36227.c: Adjust typedef for ptrcast.
* gcc.dg/pr42611.c: First check for size_t equality with void *
before trying other types.
* gcc.dg/pr59963-2.c: Skip dg-warning for int16 instead of
xfail.
* gcc.dg/pr68317.c: Add typedef for int32_t.
* gcc.dg/pr78973.c: Adjust dg-warning for int16.
* gcc.dg/pr85859.c: Cast using __INTPTR_TYPE__ instead of long.
* gcc.dg/pr86179.c: Add typedef for {u,}int32_t.
* gcc.dg/torture/
20181024-1.c: Require size32plus.
* gcc.dg/torture/pr71598-2.c: Skip for short_eq_int.
* gcc.dg/torture/pr86034.c: Add typedef for int32_t.
* gcc.dg/tree-ssa/builtin-sprintf-warn-3.c: Adjust dg-warning
for int16 and msp430 -mlarge.
* gcc.dg/tree-ssa/integer-addr.c: Use __INTPTR_MAX__ for a large
constant that is a valid address.
* gcc.dg/tree-ssa/loop-interchange-10.c: Add typedef for
int32_t.
* gcc.dg/tree-ssa/pr84436-3.c: Adjust dg-final for int16.
* gcc.dg/tree-ssa/pr84648.c: Add typedef for uint32_t.
* gcc.dg/tree-ssa/scev-8.c: Cast to char if sizeof(int) ==
sizeof(short).
* gcc.dg/tree-ssa/ssa-dom-thread-8.c: Adjust test for msp430 -mlarge.
* lib/target-supports.exp (check_effective_target_size24plus): New.
(check_effective_target_short_eq_int): New.
(check_effective_target_ptr_eq_short): New.
(check_effective_target_msp430_small): New.
(check_effective_target_msp430_large): New.
Jakub Jelinek [Tue, 19 May 2020 12:02:04 +0000 (14:02 +0200)]
openmp: Add basic library allocator support.
This patch adds very basic allocator support (omp_{init,destroy}_allocator,
omp_{alloc,free}, omp_[sg]et_default_allocator).
The plan is to use memkind (likely dlopened) for high bandwidth memory, but
that part isn't implemented yet, probably mlock for pinned memory and see
what other options there are for other kinds of memory.
For offloading targets, we need to decide if we want to support the
dynamic allocators (and on which targets), or if e.g. all we do is at compile
time replace omp_alloc/omp_free calls with constexpr predefined allocators
with something special.
And allocate directive and allocator/uses_allocators clauses are future work
too.
2020-05-19 Jakub Jelinek <jakub@redhat.com>
* allocator.c: New file.
Martin Liska [Tue, 19 May 2020 10:33:46 +0000 (12:33 +0200)]
mklog.py: improve parsing of struct names (ignore GTY).
* mklog.py: Skip GTY for struct names. Make flake8 happy.
* test_mklog.py: Add test for GTY.
Martin Liska [Tue, 19 May 2020 10:03:42 +0000 (12:03 +0200)]
Add missing changelog entry.
Martin Liska [Tue, 19 May 2020 10:01:41 +0000 (12:01 +0200)]
Fill up entries in reverse order.
contrib/ChangeLog:
* gcc-changelog/git_update_version.py:
Fill up entries in reverse order.
Jose E. Marchesi [Tue, 19 May 2020 09:46:40 +0000 (11:46 +0200)]
bpf: do not save/restore callee-saved registers in function prolog/epilog
BPF considers that every call to a function allocates a fresh set of
registers that are available to the callee, of which the first five
may have bee initialized with the function arguments. This is
implemented by both interpreter and JIT in the Linux kernel.
This is enforced by the kernel BPF verifier, which will reject any
code in which non-initialized registers are accessed before being
written. Consequently, the spill instructions generated in function
prologue were causing the verifier to reject our compiled programs.
This patch makes GCC to not save/restore callee-saved registers in
function prologue/epilogue, unless xBPF mode is enabled.
2020-05-19 Jose E. Marchesi <jose.marchesi@oracle.com>
gcc/
* config/bpf/bpf.c (bpf_compute_frame_layout): Include space for
callee saved registers only in xBPF.
(bpf_expand_prologue): Save callee saved registers only in xBPF.
(bpf_expand_epilogue): Likewise for restoring.
* doc/invoke.texi (eBPF Options): Document this is activated by
-mxbpf.
gcc/testsuite/
* gcc.target/bpf/xbpf-callee-saved-regs-1.c: New test.
* gcc.target/bpf/xbpf-callee-saved-regs-2.c: Likewise.
Jose E. Marchesi [Tue, 19 May 2020 09:46:06 +0000 (11:46 +0200)]
bpf: add support for the -mxbpf option
This patch adds support for a new option -mxbpf. This tells GCC to
generate code for an expanded version of BPF that relaxes some of the
restrictions imposed by BPF.
2020-05-19 Jose E. Marchesi <jose.marchesi@oracle.com>
gcc/
* config/bpf/bpf.opt (mxbpf): New option.
* doc/invoke.texi (Option Summary): Add -mxbpf.
(eBPF Options): Document -mxbbpf.
Martin Liska [Thu, 14 May 2020 22:44:07 +0000 (00:44 +0200)]
New mklog script.
contrib/ChangeLog:
2020-05-15 Martin Liska <mliska@suse.cz>
* gcc-git-customization.sh: Add
alias.gcc-mklog new hook.
* mklog.py: New file.
* test_mklog.py: New file.
Martin Liska [Tue, 19 May 2020 09:02:03 +0000 (11:02 +0200)]
Move 2 mklog scripts to legacy subfolder.
contrib/ChangeLog:
* legacy/mklog: Moved from mklog.
* legacy/mklog.pl: Moved from mklog.pl.
Uros Bizjak [Tue, 19 May 2020 09:28:42 +0000 (11:28 +0200)]
Add missing ChangeLog entries.
Uros Bizjak [Tue, 19 May 2020 09:25:46 +0000 (11:25 +0200)]
i386: Add missing vector zero/sign extend expanders [PR92658]
2020-05-19 Uroš Bizjak <ubizjak@gmail.com>
gcc/ChangeLog:
PR target/92658
* config/i386/sse.md (<code>v16qiv16hi2): New expander.
(<code>v32qiv32hi2): Ditto.
(<code>v8qiv8hi2): Ditto.
(<code>v16qiv16si2): Ditto.
(<code>v8qiv8si2): Ditto.
(<code>v4qiv4si2): Ditto.
(<code>v16hiv16si2): Ditto.
(<code>v8hiv8si2): Ditto.
(<code>v4hiv4si2): Ditto.
(<code>v8qiv8di2): Ditto.
(<code>v4qiv4di2): Ditto.
(<code>v2qiv2di2): Ditto.
(<code>v8hiv8di2): Ditto.
(<code>v4hiv4di2): Ditto.
(<code>v2hiv2di2): Ditto.
(<code>v8siv8di2): Ditto.
(<code>v4siv4di2): Ditto.
(<code>v2siv2di2): Ditto.
gcc/testsuite/ChangeLog:
PR target/92658
* gcc.target/i386/pr92658-sse4.c: New test.
* gcc.target/i386/pr92658-avx2.c: New test.
* gcc.target/i386/pr92658-avx512bw.c: New test.
Martin Liska [Tue, 19 May 2020 09:12:52 +0000 (11:12 +0200)]
Add missing ChangeLog entry.
Martin Liska [Tue, 19 May 2020 08:39:30 +0000 (10:39 +0200)]
Fix typo in c-parser.c.
gcc/c/ChangeLog:
* c-parser.c: Fix typo.
Jakub Jelinek [Tue, 19 May 2020 08:11:01 +0000 (10:11 +0200)]
openmp: Add basic library allocator support.
This patch adds very basic allocator support (omp_{init,destroy}_allocator,
omp_{alloc,free}, omp_[sg]et_default_allocator).
The plan is to use memkind (likely dlopened) for high bandwidth memory, but
that part isn't implemented yet, probably mlock for pinned memory and see
what other options there are for other kinds of memory.
For offloading targets, we need to decide if we want to support the
dynamic allocators (and on which targets), or if e.g. all we do is at compile
time replace omp_alloc/omp_free calls with constexpr predefined allocators
with something special.
And allocate directive and allocator/uses_allocators clauses are future work
too.
2020-05-19 Jakub Jelinek <jakub@redhat.com>
* omp.h.in (omp_uintptr_t): New typedef.
(__GOMP_UINTPTR_T_ENUM): Define.
(omp_memspace_handle_t, omp_allocator_handle_t, omp_alloctrait_key_t,
omp_alloctrait_value_t, omp_alloctrait_t): New typedefs.
(__GOMP_DEFAULT_NULL_ALLOCATOR): Define.
(omp_init_allocator, omp_destroy_allocator, omp_set_default_allocator,
omp_get_default_allocator, omp_alloc, omp_free): Declare.
* libgomp.h (struct gomp_team_state): Add def_allocator field.
(gomp_def_allocator): Declare.
* libgomp.map (OMP_5.0.1): Export omp_set_default_allocator,
omp_get_default_allocator, omp_init_allocator, omp_destroy_allocator,
omp_alloc and omp_free.
* team.c (gomp_team_start): Copy over ts.def_allocator.
* env.c (gomp_def_allocator): New variable.
(parse_wait_policy): Adjust function comment.
(parse_allocator): New function.
(handle_omp_display_env): Print OMP_ALLOCATOR.
(initialize_env): Call parse_allocator.
* Makefile.am (libgomp_la_SOURCES): Add allocator.c.
* allocator.c: New file.
* icv.c (omp_set_default_allocator, omp_get_default_allocator): New
functions.
* testsuite/libgomp.c-c++-common/alloc-1.c: New test.
* testsuite/libgomp.c-c++-common/alloc-2.c: New test.
* testsuite/libgomp.c-c++-common/alloc-3.c: New test.
* Makefile.in: Regenerated.
Martin Liska [Tue, 19 May 2020 07:19:18 +0000 (09:19 +0200)]
Add gcc-verify alias.
* gcc-git-customization.sh: Add gcc-verify alias
that uses contrib/gcc-changelog/git_check_commit.py.
Kito Cheng [Fri, 10 Apr 2020 09:20:19 +0000 (17:20 +0800)]
RISC-V: Handle implied extension for -march parser.
- Implied rule are introduced into latest RISC-V ISA spec.
- Only implemented D implied F-extension. Zicsr and Zifence are not
implement yet, so the rule not included in this patch.
- Pass preprocessed arch string to arch.
- Verified with binutils 2.30 and 2.34.
gcc/ChangeLog
* common/config/riscv/riscv-common.c (riscv_implied_info_t): New.
(riscv_implied_info): New.
(riscv_subset_list): Add handle_implied_ext.
(riscv_subset_list::to_string): New parameter version_p to
control output format.
(riscv_subset_list::handle_implied_ext): New.
(riscv_subset_list::parse_std_ext): Call handle_implied_ext.
(riscv_arch_str): New parameter version_p to control output format.
(riscv_expand_arch): New.
* config/riscv/riscv-protos.h (riscv_arch_str): New parameter,
version_p.
* config/riscv/riscv.h (riscv_expand_arch): New,
(EXTRA_SPEC_FUNCTIONS): Define.
(ASM_SPEC): Transform -march= via riscv_expand_arch.
gcc/testsuite/ChangeLog
* gcc.target/riscv/arch-6.c: New.
* gcc.target/riscv/attribute-11.c: New.
* gcc.target/riscv/attribute-12.c: New.
Kito Cheng [Fri, 10 Apr 2020 09:20:18 +0000 (17:20 +0800)]
RISC-V: Update march parser
- The arch string rule has changed in latest spec, it introduced new
multi-letter extension prefix with 'h' and 'z', and drop `sx`. also
adjust parsing order for 's' and 'x'.
gcc/ChangeLog
* riscv-common.c (parse_sv_or_non_std_ext): Rename to
parse_multiletter_ext.
(parse_multiletter_ext): Add parsing `h` and `z`, drop `sx`,
adjust parsing order for 's' and 'x'.
gcc/testsuite/ChangeLog
* gcc.target/riscv/arch-3.c: Adjust option.
* gcc.target/riscv/arch-5.c: New.
* gcc.target/riscv/attribute-9.c: Adjust option and test
condition.
Richard Biener [Mon, 18 May 2020 14:05:00 +0000 (16:05 +0200)]
cost invariant nodes from vect_slp_analyze_node_operations SLP walk
2020-05-19 Richard Biener <rguenther@suse.de>
* tree-vectorizer.h (_slp_tree::vectype): Add field.
(SLP_TREE_VECTYPE): New.
* tree-vect-slp.c (vect_create_new_slp_node): Initialize
SLP_TREE_VECTYPE.
(vect_create_new_slp_node): Likewise.
(vect_prologue_cost_for_slp): Move here from tree-vect-stmts.c
and simplify.
(vect_slp_analyze_node_operations): Walk nodes children for
invariant costing.
(vect_get_constant_vectors): Use local scope op variable.
* tree-vect-stmts.c (vect_prologue_cost_for_slp_op): Remove here.
(vect_model_simple_cost): Adjust.
(vect_model_store_cost): Likewise.
(vectorizable_store): Likewise.
Patrick Palka [Tue, 19 May 2020 03:50:32 +0000 (23:50 -0400)]
c++: Enable spec_hasher table sanitization [PR87847]
It looks like hash table sanitization is now safe to enable for the
decl_specializations and type_specializations tables, probably ever
since PR94454 was fixed.
gcc/cp/ChangeLog:
PR c++/87847
* pt.c (init_template_processing): Enable sanitization for
decl_specializations and type_specializations.
Patrick Palka [Tue, 19 May 2020 03:50:14 +0000 (23:50 -0400)]
c++: Explain fn template argument type/value mismatches [PR66439]
In fn_type_unifcation, we are passing NULL_TREE as the 'in_decl'
parameter to coerce_template_parms, and this is causing template
type/value mismatch error messages to get suppressed regardless of the
value of 'complain'.
This means that when substitution into a function template fails due to
a type/value mismatch between a template parameter and the provided
template argument, we just say "template argument deduction/substitution
failed:" without a followup explanation of the failure.
Fix this by passing 'fn' instead of NULL_TREE to coerce_template_parms.
gcc/cp/ChangeLog:
PR c++/66439
* pt.c (fn_type_unification): Pass 'fn' instead of NULL_TREE as
the 'in_decl' parameter to coerce_template_parms.
gcc/testsuite/ChangeLog:
PR c++/66439
* g++.dg/cpp2a/concepts-ts4.C: Expect a "type/value mismatch"
diagnostic.
* g++.dg/cpp2a/concepts-ts6.C: Likewise.
* g++.dg/template/error56.C: Likewise.
* g++.dg/template/error59.C: New test.
libstdc++-v3/ChangeLog:
PR c++/66439
* testsuite/20_util/pair/astuple/get_neg.cc: Prune "type/value
mismatch" messages.
* testsuite/20_util/tuple/element_access/get_neg.cc: Likewise.
GCC Administrator [Tue, 19 May 2020 00:16:28 +0000 (00:16 +0000)]
Daily bump.
Marek Polacek [Wed, 6 May 2020 19:53:33 +0000 (15:53 -0400)]
c++: ICE when shortening right shift [PR94955]
Since r10-6527 fold_for_warn calls maybe_constant_value, which means it
can fold more than it previously could. In this testcase it means that
cp_build_binary_op/RSHIFT_EXPR set short_shift because now we were able
to fold op1 to an INTEGER_CST. But then when actually performing the
shortening we crashed because cp_fold_rvalue wasn't able to fold as much
as f_f_w and so tree_int_cst_sgn crashed on a NOP_EXPR. Therefore the
calls should probably match.
PR c++/94955
* typeck.c (cp_build_binary_op): Use fold_for_warn instead of
cp_fold_rvalue.
* g++.dg/cpp0x/constexpr-shift2.C: New test.
Marek Polacek [Wed, 6 May 2020 23:24:58 +0000 (19:24 -0400)]
c++: ICE with -Wall and constexpr if [PR94937]
An ICE arises here because we call cp_get_callee_fndecl_nofold in a
template, and we've got a CALL_EXPR whose CALL_EXPR_FN is a BASELINK.
This tickles the INDIRECT_TYPE_P assert in cp_get_fndecl_from_callee.
Fixed by turning the assert into a condition and returning NULL_TREE
in that case.
PR c++/94937
* cvt.c (cp_get_fndecl_from_callee): Return NULL_TREE if the function
type is not INDIRECT_TYPE_P.
* decl.c (omp_declare_variant_finalize_one): Call
cp_get_callee_fndecl_nofold instead of looking for the function decl
manually.
* g++.dg/cpp1z/constexpr-if34.C: New test.
* g++.dg/cpp2a/is-constant-evaluated10.C: New test.
Martin Sebor [Mon, 18 May 2020 22:31:13 +0000 (16:31 -0600)]
PR middle-end/92815 - spurious -Wstringop-overflow writing into a flexible array of an extern struct
Adjust test to avoid failures in ILP32 mode.
gcc/testsuite/ChangeLog:
PR middle-end/92815
* gcc.dg/builtin-object-size-20.c: Adjust to avoid failures in
ILP32 mode.
Marek Polacek [Fri, 8 May 2020 01:10:42 +0000 (21:10 -0400)]
c++: Sorry about type-dependent arg for __builtin_has_attribute [PR90915]
Until 92104 is fixed, let's sorry rather than crash.
PR c++/90915
* parser.c (cp_parser_has_attribute_expression): Sorry on a
type-dependent argument.
* g++.dg/ext/builtin-has-attribute.C: New test.
Martin Sebor [Mon, 18 May 2020 21:24:12 +0000 (15:24 -0600)]
PR middle-end/92815 - spurious -Wstringop-overflow writing into a flexible array of an extern struct
gcc/ChangeLog:
PR middle-end/92815
* tree-object-size.c (decl_init_size): New function.
(addr_object_size): Call it.
* tree.h (last_field): Declare.
(first_field): Add attribute nonnull.
gcc/testsuite/ChangeLog:
PR middle-end/92815
* gcc.dg/Warray-bounds-56.c: Remove xfails.
* gcc.dg/builtin-object-size-20.c: New test.
* gcc.dg/builtin-object-size-21.c: New test.
Martin Sebor [Mon, 18 May 2020 21:07:48 +0000 (15:07 -0600)]
PR middle-end/94940 - spurious -Warray-bounds for a zero length array member of union
gcc/testsuite/ChangeLog:
PR middle-end/94940
* gcc.dg/Warray-bounds-61.c: New test.
gcc/ChangeLog:
PR middle-end/94940
* tree-vrp.c (vrp_prop::check_mem_ref): Remove unreachable code.
* tree.c (component_ref_size): Correct the handling or array members
of unions.
Drop a pointless test.
Rename a local variable.
Joseph Myers [Mon, 18 May 2020 20:50:35 +0000 (20:50 +0000)]
Update gcc sv.po.
* sv.po: Update.
Marek Polacek [Wed, 13 May 2020 19:52:42 +0000 (15:52 -0400)]
c++: Implement DR 1512, Pointer comparison vs qual convs [PR87699]
This patch resolves DR 1512 (and, by turn, DR 583). This entails:
1) Relational pointer comparisons against null pointer constants have
been made ill-formed:
void f(char *p) {
if (p > 0)
// ...
}
was always invalid in C but was -- accidentally -- allowed in C++.
2) This was ill-formed:
bool foo(int** x, const int** y) {
return x < y;
}
because 'int**' couldn't be converted to 'const int**'. This was
fixed by re-defining a generic composite pointer type. The composite
type of these two pointers will be 'const int *const *', to which
both pointers can be converted.
3) The overload descriptions for built-in operators were adjusted,
because objects of type std::nullptr_t cannot be used with relational
operators any more.
I fixed 1) by adjusting cp_build_binary_op; we already had a warning
for it so made it a hard error now.
Then 2) required tweaking composite_pointer_type_r. [expr.type] defines
the composite pointer type by using the "cv-combined type." We didn't
implement the [conv.qual]/3.3 part; previously the composite type of
'int**' and 'const int**' was 'const int**', so this didn't compile:
void f(const int **p, int **q) {
true ? p : q;
}
I wrote a more extensive test for this which uses decltype and some
template magic to check the composite type, see composite-ptr-type.C.
We still don't handle everything that [expr.type] requires us to,
but it's pretty close.
And finally 3) was handled in add_builtin_candidate. Turned out we
weren't creating built-in operator candidates when the type was
std::nullptr_t at all. We should, for == and !=. Tested in builtin4.C.
In passing, I'm fixing some of the comments too.
DR 1512
PR c++/87699
* call.c (add_builtin_candidate) <case EQ_EXPR>: Create candidate
operator functions when type is std::nullptr_t for ==/!=.
* typeck.c (composite_pointer_type_r): Add bool a * parameter. Use it
to maybe add "const" to the pointer type.
(composite_pointer_type): Update the call to composite_pointer_type_r.
(cp_build_binary_op): Turn two warning_at into error_at. Print the
types.
* g++.dg/cpp0x/constexpr-array-ptr10.C: Change dg-warning to dg-error
and adjust the expected messages in dg-error.
* g++.dg/expr/composite-ptr-type.C: New test.
* g++.dg/expr/ptr-comp1.C: New test.
* g++.dg/expr/ptr-comp2.C: New test.
* g++.dg/expr/ptr-comp3.C: New test.
* g++.dg/overload/builtin4.C: New test.
* g++.dg/warn/Wextra-3.C: Change dg-warning to dg-error.
Jason Merrill [Wed, 15 Jan 2020 03:55:59 +0000 (22:55 -0500)]
c++: Create fewer SAVE_EXPR.
In a couple of places in build_over_call we were calling
cp_stabilize_reference but only using the result once, so it isn't needed.
gcc/cp/ChangeLog
2020-05-18 Jason Merrill <jason@redhat.com>
* call.c (build_over_call): Remove unnecessary
cp_stabilize_reference.
Marek Polacek [Fri, 15 May 2020 14:59:01 +0000 (10:59 -0400)]
c++: Don't add built-in operator for ++ on bool.
This feels extremely obscure but at least it's an opportunity to fix the
comments. P0002R1 removed deprecated operator++(bool) in C++17 so let's
avoid adding a builtin overload candidate for ++ when the type is bool.
* call.c (add_builtin_candidate): Don't create a builtin overload
candidate for ++ when type is bool in C++17.
* g++.dg/overload/builtin5.C: New test.
Marek Polacek [Fri, 15 May 2020 21:54:05 +0000 (17:54 -0400)]
c++: Regenerate cp/cfns.h.
Current cfns.h includes register-qualified variables and that wouldn't
play well when bootstrapping with GCC that uses the C++17 dialect,
because 'register' was removed in C++17. Regenerating it using the
command specified in cfns.h luckily cleaned this up.
* cfns.h: Regenerated.
Douglas Rupp [Mon, 18 May 2020 18:43:48 +0000 (11:43 -0700)]
Require powerpc_vsx_ok in gcc.target/powerpc/pr71763.c
We're getting an error when running this test on PowerPC VxWorks 7,
due to an unexpected warning:
| Excess errors:
| cc1: warning: '-mvsx' and '-mno-altivec' are incompatible
The warning comes from a combination of factors:
- The test itself uses -mvsx explicitly via the following directive:
// { dg-options "-O1 -mvsx" }
- Our toolchain was configured so as to make -mno-altivec
the default;
- These two options are mutually exclusive.
This commit adds a powerpc_vsx_ok dg-require-effective-target directive
to that test, and thus making it UNSUPPORTED instead.
Tested on PowerPC VxWorks 7. Also tested on PowerPC ELF as well,
a platform where we do not make -mno-altivec the default, to verify
that the test continues to run as usual in that case.
gcc/testsuite/
* gcc.target/powerpc/pr71763.c: Require powerpc_vsx_ok.
Jason Merrill [Mon, 18 May 2020 18:28:16 +0000 (14:28 -0400)]
bootstrap: Update requirement to C++11.
There was general agreement last November that we would move to allowing
C++11 features to be used in GCC 11; this patch implements that direction.
ChangeLog
2020-05-18 Jason Merrill <jason@redhat.com>
* configure.ac: Update bootstrap dialect to -std=c++11.
config/ChangeLog
2020-05-18 Jason Merrill <jason@redhat.com>
* ax_cxx_compile_stdcxx.m4: Import from autoconf archive with
an adjustment to try the default mode.
gcc/ChangeLog
2020-05-18 Jason Merrill <jason@redhat.com>
* aclocal.m4: Add ax_cxx_compile_stdcxx.m4.
* configure.ac: Use AX_CXX_COMPILE_STDCXX(11).
Harald Anlauf [Mon, 18 May 2020 18:27:29 +0000 (20:27 +0200)]
PR fortran/95053 - division by zero constants
Partially revert the fix for PR93499. Replace by checks for valid
expressions in the declaration of array shape and PDT KIND and LEN
expressions at a later stage.
gcc/fortran/
2020-05-18 Harald Anlauf <anlauf@gmx.de>
PR fortran/95053
* arith.c (gfc_divide): Revert hunk introduced by patch for
PR93499.
* decl.c (variable_decl): Generate error for array shape not being
an INTEGER constant.
(gfc_get_pdt_instance): Generate error if KIND or LEN expressions
in declaration of a PDT instance do not simplify to INTEGER
constants.
gcc/testsuite/
2020-05-18 Harald Anlauf <anlauf@gmx.de>
PR fortran/95053
* gfortran.dg/dec_structure_23.f90: Adjust to new error messages.
* gfortran.dg/pr93499.f90: Adjust to new error messages.
* gfortran.dg/pr95053_2.f90: New test.
* gfortran.dg/pr95053_3.f90: New test.
Stefan Schulze Frielinghaus [Tue, 5 May 2020 17:44:19 +0000 (19:44 +0200)]
tree-optimization: Fix use of uninitialized variables warnings [PR94952]
While bootstrapping GCC on S/390 with --enable-checking=release several
warnings about use of uninitialized variables bitpos, bitregion_start, and
bitregion_end of function pass_store_merging::process_store are raised.
According to PR94952 these seem to be false positives which are silenced by
initialising the mentioned variables.
Bootstrapped on S/390. Ok for master and releases/gcc-10 assuming that
regtest succeeds (still running but I don't see a reason why it
should fail)?
gcc/ChangeLog:
2020-05-18 Stefan Schulze Frielinghaus <stefansf@linux.ibm.com>
PR tree-optimization/94952
* gimple-ssa-store-merging.c (pass_store_merging::process_store):
Initialize variables bitpos, bitregion_start, and bitregion_end in
order to silence warnings about use of uninitialized variables.
Marek Polacek [Mon, 18 May 2020 17:50:39 +0000 (13:50 -0400)]
c++: Add test for c++/95143
Already fixed by r10-8124-gceae6a13366d9646e172fc943fe8e221b70f0920.
PR c++/95143
* g++.dg/cpp0x/sfinae66.C: New test.
Carl Love [Wed, 29 Apr 2020 15:23:11 +0000 (10:23 -0500)]
pr94833, fix vec_first_match_index for nulls
gcc/ChangeLog
2020-04-30 Carl Love <cel@us.ibm.com>
PR target/94833
* config/rs6000/vsx.md (define_expand): Fix instruction generation for
first_match_index_<mode>.
* testsuite/gcc.target/powerpc/builtins-8-p9-runnable.c (main): Add
additional test cases with zero vector elements.
Uros Bizjak [Mon, 18 May 2020 15:52:14 +0000 (17:52 +0200)]
i386: Avoid reversing a non-trapping comparison to a trapping one [PR95169]
gcc/ChangeLog:
PR target/95169
* config/i386/i386-expand.c (ix86_expand_int_movcc):
Avoid reversing a non-trapping comparison to a trapping one.
testsuite/ChangeLog:
PR target/95169
* gcc.target/i386/pr95169.c: New test.
Alex Coplan [Mon, 18 May 2020 15:29:04 +0000 (16:29 +0100)]
[arm] Don't generate invalid LDRD insns
This fixes a bug in the arm backend where GCC generates invalid LDRD
instructions. The LDRD instruction requires the first transfer register to be
even, but GCC attempts to use odd registers here. For example, with the
following C code:
struct c {
double a;
} __attribute((aligned)) __attribute((packed));
struct c d;
struct c f(struct c);
void e() { f(d); }
The struct d is passed in registers r1 and r2 to the function f, and GCC
attempted to do this with a LDRD instruction when compiling with -march=armv7-a
on a soft float toolchain.
The fix is analogous to the corresponding one for STRD in the same function:
https://gcc.gnu.org/git/?p=gcc.git;a=commit;h=
52057dc4ac5295caebf83147f688d769c93cbc8d
2020-05-18 Alex Coplan <alex.coplan@arm.com>
gcc/:
* config/arm/arm.c (output_move_double): Fix codegen when loading into
a register pair with an odd base register.
gcc/testsuite/:
* gcc.c-torture/compile/packed-aligned-1.c: New test.
* gcc.c-torture/execute/packed-aligned.c: New test.
Uros Bizjak [Mon, 18 May 2020 15:25:39 +0000 (17:25 +0200)]
i386: Improve vector mode and TFmode ABS and NEG patterns
gcc/ChangeLog:
2020-05-18 Uroš Bizjak <ubizjak@gmail.com>
* config/i386/i386-expand.c (ix86_expand_fp_absneg_operator):
Do not emit FLAGS_REG clobber for TFmode.
* config/i386/i386.md (*<code>tf2_1): Rewrite as
define_insn_and_split. Mark operands 1 and 2 commutative.
(*nabstf2_1): Ditto.
(absneg SSE splitter): Use MODEF mode iterator instead of SSEMODEF.
Do not swap memory operands. Simplify RTX generation.
(neg abs SSE splitter): Ditto.
* config/i386/sse.md (*<code><mode>2): Mark operands 1 and 2
commutative. Do not swap operands. Simplify RTX generation.
(*nabs<mode>2): Ditto.
Richard Biener [Fri, 15 May 2020 11:13:38 +0000 (13:13 +0200)]
fixup BB vectorization constant generation place
This adjusts the way we compute the stmt insert location for
invariants in BB vectorization context to deal with eventually
sharing invariant SLP nodes for multiple uses. We can no longer
use a single use stmt location then but there's a simple way out.
2020-05-18 Richard Biener <rguenther@suse.de>
* tree-vect-slp.c (vect_slp_bb): Start after labels.
(vect_get_constant_vectors): Really place init stmt after scalar defs.
* tree-vect-stmts.c (vect_init_vector_1): Insert before
region begin.
H.J. Lu [Mon, 18 May 2020 12:35:27 +0000 (05:35 -0700)]
x86: Update Intel processor detection
Add cpu model numbers for Intel Airmont, Tremont, Comet Lake, Ice Lake
and Tiger Lake processor families.
* config/i386/driver-i386.c (host_detect_local_cpu): Support
Intel Airmont, Tremont, Comet Lake, Ice Lake and Tiger Lake
processor families.
Alex Coplan [Mon, 18 May 2020 11:21:17 +0000 (12:21 +0100)]
MAINTAINERS: Add myself for write after approval.
2020-05-18 Alex Coplan <alex.coplan@arm.com>
* MAINTAINERS (Write After Approval): Add myself.
Richard Biener [Mon, 18 May 2020 06:51:23 +0000 (08:51 +0200)]
middle-end/95171 - inlining of trapping compare into non-call EH fn
This fixes always-inlining across -fnon-call-exception boundaries
for conditions which we do not allow to throw.
2020-05-18 Richard Biener <rguenther@suse.de>
PR middle-end/95171
* tree-inline.c (remap_gimple_stmt): Split out trapping compares
when inlining into a non-call EH function.
* gcc.dg/pr95171.c: New testcase.
Richard Biener [Mon, 18 May 2020 07:17:24 +0000 (09:17 +0200)]
tree-optimization/95172 - avoid mixing conditionalized and ordered SM
The following testcase shows a missed optimization that then leads to
wrong-code when issueing SMed stores on exits. When we were able to
compute an ordered sequence of stores for an exit we need to emit
that in the correct order and we can emit it disregarding to any
conditional for whether a store actually happened (we know it did).
We can also improve detection as of whether we need conditional
processing at all. Both parts fix the testcase.
2020-05-18 Richard Biener <rguenther@suse.de>
PR tree-optimization/95172
* tree-ssa-loop-im.c (execute_sm): Get flag whether we
eventually need the conditional processing.
(execute_sm_exit): When processing an orderd sequence
avoid doing any conditional processing.
(hoist_memory_references): Pass down whether all edges
have ordered processing for a ref to execute_sm.
* gcc.dg/torture/pr95172.c: New testcase.
GCC Administrator [Mon, 18 May 2020 00:16:18 +0000 (00:16 +0000)]
Daily bump.
Iain Sandoe [Sun, 17 May 2020 11:26:19 +0000 (12:26 +0100)]
coroutines: Avoid a maybe used uninitialized warning. NFC.
This avoids a (bogus) warning that occurs with some bootstrap
compilers.
gcc/cp/ChangeLog:
2020-05-17 Iain Sandoe <iain@sandoe.co.uk>
* coroutines.cc (morph_fn_to_coro): Initialize the
gro variable.
Jeff Law [Sun, 17 May 2020 17:20:39 +0000 (13:20 -0400)]
Use pc_or_label_operand to collapse a couple more patterns in preparation for the cc0->CC_REG transition.
* config/h8300/predicates.md (pc_or_label_operand): New predicate.
* config/h8300/jumpcall.md (branch_true, branch_false): Consolidate
into a single pattern using pc_or_label_operand.
* config/h8300/combiner.md (bit branch patterns): Likewise.
* config/h8300/peepholes.md (HImode and SImode branches): Likewise.
H.J. Lu [Sun, 17 May 2020 17:10:34 +0000 (10:10 -0700)]
x86: Allow V1TI vector register pushes
Add V1TI vector register push and split it after reload to a sequence
of:
(set (reg:P SP_REG) (plus:P SP_REG) (const_int -8)))
(set (match_dup 0) (match_dup 1))
so that STV pass can convert TI mode integer push to V1TI vector register
push. Rename has_non_address_hard_reg to pseudo_reg_set, combine calls
of single_set and has_non_address_hard_reg to pseudo_reg_set, to ignore
pseudo register push.
Remove c-c++-common/dfp/func-vararg-mixed-2.c since it is compiled with
-mpreferred-stack-boundary=2 and leads to segfault:
Dump of assembler code for function __bid_nesd2:
0x08049210 <+0>: endbr32
0x08049214 <+4>: push %esi
0x08049215 <+5>: push %ebx
0x08049216 <+6>: call 0x8049130 <__x86.get_pc_thunk.bx>
0x0804921b <+11>: add $0x8de5,%ebx
0x08049221 <+17>: sub $0x20,%esp
0x08049224 <+20>: mov 0x30(%esp),%esi
0x08049228 <+24>: pushl 0x2c(%esp)
0x0804922c <+28>: call 0x804e600 <__bid32_to_bid64>
0x08049231 <+33>: mov %esi,(%esp)
0x08049234 <+36>: movd %edx,%xmm1
0x08049238 <+40>: movd %eax,%xmm0
0x0804923c <+44>: punpckldq %xmm1,%xmm0
=> 0x08049240 <+48>: movaps %xmm0,0x10(%esp)
0x08049245 <+53>: call 0x804e600 <__bid32_to_bid64>
0x0804924a <+58>: push %edx
0x0804924b <+59>: push %eax
0x0804924c <+60>: pushl 0x1c(%esp)
0x08049250 <+64>: pushl 0x1c(%esp)
0x08049254 <+68>: call 0x804b260 <__bid64_quiet_not_equal>
0x08049259 <+73>: add $0x34,%esp
0x0804925c <+76>: pop %ebx
0x0804925d <+77>: pop %esi
0x0804925e <+78>: ret
when libgcc is compiled with -msse2. According to GCC manual:
'-mpreferred-stack-boundary=NUM'
Attempt to keep the stack boundary aligned to a 2 raised to NUM
byte boundary. If '-mpreferred-stack-boundary' is not specified,
the default is 4 (16 bytes or 128-bits).
*Warning:* If you use this switch, then you must build all modules
with the same value, including any libraries. This includes the
system libraries and startup modules.
c-c++-common/dfp/func-vararg-mixed-2.c, which was added by
commit
3b2488ca6ece182f2136a20ee5fa0bb92f935b0f
Author: H.J. Lu <hongjiu.lu@intel.com>
Date: Wed Jul 30 19:24:02 2008 +0000
func-vararg-alternate-d128-2.c: New.
2008-07-30 H.J. Lu <hongjiu.lu@intel.com>
Joey Ye <joey.ye@intel.com>
* gcc.dg/dfp/func-vararg-alternate-d128-2.c: New.
* gcc.dg/dfp/func-vararg-mixed-2.c: Likewise.
isn't expected to work with libgcc.
gcc/
PR target/95021
* config/i386/i386-features.c (has_non_address_hard_reg):
Renamed to ...
(pseudo_reg_set): This. Return the SET expression. Ignore
pseudo register push.
(general_scalar_to_vector_candidate_p): Combine single_set and
has_non_address_hard_reg calls to pseudo_reg_set.
(timode_scalar_to_vector_candidate_p): Likewise.
* config/i386/i386.md (*pushv1ti2): New pattern.
gcc/testsuite/
PR target/95021
* c-c++-common/dfp/func-vararg-mixed-2.c: Removed.
* gcc.target/i386/pr95021-1.c: New test.
* gcc.target/i386/pr95021-2.c: Likewise.
* gcc.target/i386/pr95021-3.c: Likewise.
* gcc.target/i386/pr95021-4.c: Likewise.
* gcc.target/i386/pr95021-5.c: Likewise.
Iain Buclaw [Sun, 17 May 2020 16:49:19 +0000 (18:49 +0200)]
libphobos: Merge upstream druntime
5cc061a8, phobos
64ed4684f
- core.cpuid has been fixed to not use i7 detection on AMD processors.
- std.net.curl has been fixed to correctly handle HTTP/2 status lines.
- std.zip has had a test fixed to not rely on unzip being installed.
Fixes: PR d/95166
PR d/95167
PR d/95168
Reviewed-on: https://github.com/dlang/druntime/pull/3107
https://github.com/dlang/phobos/pull/7486
H.J. Lu [Sun, 17 May 2020 13:52:02 +0000 (06:52 -0700)]
x86: Add gcc.target/i386/strncmp-1.c
Add a strncmp test for the cmpstrn pattern with neither of the strings
is a constant string. We can expand the cmpstrn pattern to "repz cmpsb"
only if one of the strings is a constant so that expand_builtin_strncmp()
can write the length argument to be the minimum of the const string
length and the actual length argument. Otherwise, "repz cmpsb" may pass
the 0 byte.
* gcc.target/i386/strncmp-1.c: New test.
Aldy Hernandez [Sun, 17 May 2020 11:56:55 +0000 (13:56 +0200)]
Revert previous patch:
2020-05-17 Aldy Hernandez <aldyh@redhat.com>
* tree-vrp.c (operand_less_p): Move to...
* vr-values.c (operand_less_p): ...here.
* tree-vrp.h (operand_less_p): Remove.
Aldy Hernandez [Fri, 8 May 2020 11:36:32 +0000 (13:36 +0200)]
Move operand_less_p to vr-values.c.
Aldy Hernandez [Sun, 17 May 2020 11:40:09 +0000 (13:40 +0200)]
Remove vrp_insert::live_on_edge declaration.
* tree-vrp.c (class vrp_insert): Remove prototype for
live_on_edge.
Aldy Hernandez [Sat, 16 May 2020 18:56:19 +0000 (20:56 +0200)]
More refactoring of tree-vrp.c.
New class live_names to maintain the set of SSA names live.
Fix whitespace in vrp_insert.
Move a few more methods related to ASSERT_EXPR insertion into vrp_insert.
Aldy Hernandez [Tue, 5 May 2020 16:40:44 +0000 (18:40 +0200)]
Move array bounds checking out of vrp_prop and into its own class.
GCC Administrator [Sun, 17 May 2020 00:16:17 +0000 (00:16 +0000)]
Daily bump.
Iain Sandoe [Sat, 16 May 2020 18:23:19 +0000 (19:23 +0100)]
coroutines: Implicitly movable objects should use move CTORs for co_return.
This is a case where the standard contains conflicting information.
after discussion between implementators, the accepted intent is of
[class.copy.elision]. This amends the handling of co_return statements
to follow that.
gcc/cp/ChangeLog:
2020-05-16 Iain Sandoe <iain@sandoe.co.uk>
* coroutines.cc (finish_co_return_stmt): Implement rules
from [class.copy.elision] /3.
gcc/testsuite/ChangeLog:
2020-05-16 Iain Sandoe <iain@sandoe.co.uk>
* g++.dg/coroutines/co-return-syntax-10-movable.C: New test.
Jeff Law [Sat, 16 May 2020 04:47:47 +0000 (00:47 -0400)]
Consolidate a couple peepholes and improve peepholes that combine stack allocations with stack stores.
* config/h8300/h8300.md (SFI iterator): New iterator for
SFmode and SImode.
* config/h8300/peepholes.md (memory comparison): Use mode
iterator to consolidate 3 patterns into one.
(stack allocation and stack store): Handle SFmode. Handle
8 byte allocations.
GCC Administrator [Sat, 16 May 2020 00:16:18 +0000 (00:16 +0000)]
Daily bump.
Patrick Palka [Fri, 15 May 2020 22:51:11 +0000 (18:51 -0400)]
c++: decltype of invalid non-dependent expr [PR57943]
We sometimes fail to reject an invalid non-dependent operand to decltype
when inside a template, because finish_decltype_type resolves the
decltype to the TREE_TYPE of the operand before we ever instantiate and
fully process the operand. Fix this by adding a call to
instantiate_non_dependent_expr_sfinae in finish_decltype_type.
gcc/cp/ChangeLog:
PR c++/57943
* semantics.c (finish_decltype_type): Call
instantiate_non_dependent_expr_sfinae on the expression.
gcc/testsuite/ChangeLog:
PR c++/57943
* g++.dg/cpp0x/decltype76.C: New test.
Joseph Myers [Fri, 15 May 2020 22:40:40 +0000 (22:40 +0000)]
Update cpplib sv.po.
* sv.po: Update.
Ian Lance Taylor [Fri, 15 May 2020 17:50:57 +0000 (10:50 -0700)]
libgo: only build syscall test with -static if it works
Test whether -static works, and use it if possible.
This time for sure.
For PR go/95061
Reviewed-on: https://go-review.googlesource.com/c/gofrontend/+/234024
Jason Merrill [Fri, 15 May 2020 21:59:49 +0000 (17:59 -0400)]
c++: Enable coroutines with -std=c++20.
Now that GCC 10 is out it seems time. People can still choose to disable
coroutines with -fno-coroutines.
This also switches the coroutines testsuite to run in C++20 mode. The
change to coro.h is only necessary for co-await-11-forwarding.C; we could
alternatively #include <utility> just in that file.
gcc/c-family/ChangeLog
2020-05-15 Jason Merrill <jason@redhat.com>
* c-opts.c (set_std_cxx20): Set flag_coroutines.
gcc/testsuite/ChangeLog
2020-05-15 Jason Merrill <jason@redhat.com>
* g++.dg/coroutines/coro.h: Always #include <utility>.
* g++.dg/coroutines/coroutines.exp (DEFAULT_COROFLAGS): Use
-std=c++20.
Jason Merrill [Fri, 15 May 2020 21:27:15 +0000 (17:27 -0400)]
analyzer: Remove stray semicolon.
Segher Boessenkool [Fri, 15 May 2020 18:18:57 +0000 (18:18 +0000)]
rs6000: BU_FUTURE_MISC_2 requires powerpc64
BU_FUTURE_MISC_2 is (currently) only used for instructions that require
64-bit registers.
2020-05-15 Segher Boessenkool <segher@kernel.crashing.org>
* config/rs6000/rs6000-builtin.def (BU_FUTURE_MISC_2): Also require
RS6000_BTM_POWERPC64.
Segher Boessenkool [Fri, 15 May 2020 18:15:26 +0000 (18:15 +0000)]
rs6000/testsuite: Use the int128 selector where needed
Tests that use the __int128 type need to use the int128 selector.
2020-05-15 Segher Boessenkool <segher@kernel.crashing.org>
gcc/testsuite/
* gcc.target/powerpc/vec-gnb-0.c: Use int128 effective target.
* gcc.target/powerpc/vec-gnb-1.c: Ditto.
* gcc.target/powerpc/vec-gnb-2.c: Ditto.
* gcc.target/powerpc/vec-ternarylogic-8.c: Ditto.
* gcc.target/powerpc/vec-ternarylogic-9.c: Ditto.
* gcc.target/powerpc/vec-ternarylogic-10.c: Ditto.
Segher Boessenkool [Fri, 15 May 2020 18:12:42 +0000 (18:12 +0000)]
rs6000/testsuite: Use lp64 in cnttzdm-0.c
2020-05-15 Segher Boessenkool <segher@kernel.crashing.org>
gcc/testsuite/
* gcc.target/powerpc/cnttzdm-0.c: Use lp64.
Segher Boessenkool [Fri, 15 May 2020 16:41:28 +0000 (16:41 +0000)]
rs6000/testsuite: Don't use powerpc64 effective target
The powerpc64 effective target unfortunately does not mean the target
has 64-bit instructions enabled (i.e., -mpowerpc64): instead, it means
that the assembler supports it.
Let's use the lp64 effective target instead for these tests.
2020-05-15 Segher Boessenkool <segher@kernel.crashing.org>
gcc/testsuite/
* gcc.target/powerpc/cntlzdm-0.c: Use lp64 instead of powerpc64.
* gcc.target/powerpc/cntlzdm-1.c: Ditto.
* gcc.target/powerpc/cnttzdm-1.c: Ditto.
* gcc.target/powerpc/pdep-0.c: Ditto.
* gcc.target/powerpc/pdep-1.c: Ditto.
* gcc.target/powerpc/pextd-0.c: Ditto.
* gcc.target/powerpc/pextd-1.c: Ditto.
Segher Boessenkool [Fri, 15 May 2020 16:33:20 +0000 (16:33 +0000)]
rs6000/testsuite: Use -mdejagnu-cpu= instead of -mcpu=
A bunch of new cases snuck in.
2020-05-15 Segher Boessenkool <segher@kernel.crashing.org>
gcc/testsuite/
* gcc.target/powerpc/pdep-0.c: Change -mcpu= to -mdejagnu-cpu=.
* gcc.target/powerpc/pdep-1.c: Ditto.
* gcc.target/powerpc/pextd-0.c: Ditto.
* gcc.target/powerpc/pextd-1.c: Ditto.
* gcc.target/powerpc/pr90763.c: Ditto.
* gcc.target/powerpc/pr91275.c: Ditto.
* gcc.target/powerpc/pr92796.c: Ditto.
* gcc.target/powerpc/pr93658.c: Ditto.
* gcc.target/powerpc/pr93800.c: Ditto.
* gcc.target/powerpc/setbceq.c: Ditto.
* gcc.target/powerpc/setbcge.c: Ditto.
* gcc.target/powerpc/setbcgt.c: Ditto.
* gcc.target/powerpc/setbcle.c: Ditto.
* gcc.target/powerpc/setbclt.c: Ditto.
* gcc.target/powerpc/setbcne.c: Ditto.
* gcc.target/powerpc/setnbceq.c: Ditto.
* gcc.target/powerpc/setnbcge.c: Ditto.
* gcc.target/powerpc/setnbcgt.c: Ditto.
* gcc.target/powerpc/setnbcle.c: Ditto.
* gcc.target/powerpc/setnbclt.c: Ditto.
* gcc.target/powerpc/setnbcne.c: Ditto.
* gcc.target/powerpc/xxgenpc-runnable.c: Ditto.
Patrick Palka [Fri, 15 May 2020 18:50:17 +0000 (14:50 -0400)]
c++: Revert unnecessary parts of fix for [PR90996]
The process_init_constructor_array part of my PR90996 patch turns out to
be neither necessary nor sufficient to make the pr90996.C testcase work,
and I wasn't able to come up with a testcase that demonstrates this part
is ever necessary.
gcc/cp/ChangeLog:
Revert:
2020-04-07 Patrick Palka <ppalka@redhat.com>
PR c++/90996
* typeck2.c (process_init_constructor_array): Propagate
CONSTRUCTOR_PLACEHOLDER_BOUNDARY up from each element
initializer to the array initializer.
gcc/testsuite/ChangeLog:
PR c++/90996
* g++.dg/cpp1y/pr90996.C: Turn into execution test to verify
that each PLACEHOLDER_EXPR gets correctly resolved.
Jason Merrill [Fri, 15 May 2020 18:06:48 +0000 (14:06 -0400)]
PR c++/93286 - ICE with __is_constructible and variadic template.
My GCC 10 patch for 93286 fixed the missing piece in tsubst's handling of
lists vs. that in tsubst_copy_and_build, but it would be better to share the
code between them.
gcc/cp/ChangeLog
2020-05-15 Jason Merrill <jason@redhat.com>
PR c++/93286 - ICE with __is_constructible and variadic template.
* pt.c (tsubst_tree_list): New.
(tsubst, tsubst_copy_and_build): Use it.
* decl2.c (is_late_template_attribute): Handle error_mark_node args.
H.J. Lu [Fri, 15 May 2020 16:06:50 +0000 (09:06 -0700)]
x86: Also check if -fcf-protection works
When defaulting CET run-time support to auto, check if -fcf-protection
works. Even if the stage1 GCC doesn't support -fcf-protection, since
the final GCC does, CET run-time support will be enabled by default if
binutils support CET.
config/
PR bootstrap/95147
* cet.m4 (GCC_CET_FLAGS): Also check if -fcf-protection works
when defaulting to auto.
libatomic/
PR bootstrap/95147
* configure: Regenerated.
libbacktrace/
PR bootstrap/95147
* configure: Regenerated.
libgcc/
PR bootstrap/95147
* configure: Regenerated.
libgfortran/
PR bootstrap/95147
* configure: Regenerated.
libgomp/
PR bootstrap/95147
* configure: Regenerated.
libitm/
PR bootstrap/95147
* configure: Regenerated.
libobjc/
PR bootstrap/95147
* configure: Regenerated.
libphobos/
PR bootstrap/95147
* configure: Regenerated.
libquadmath/
PR bootstrap/95147
* configure: Regenerated.
libsanitizer/
PR bootstrap/95147
* configure: Regenerated.
libssp/
PR bootstrap/95147
* configure: Regenerated.
libstdc++-v3/
PR bootstrap/95147
* configure: Regenerated.
libvtv/
PR bootstrap/95147
* configure: Regenerated.
zlib/
PR bootstrap/95147
* configure: Regenerated.
Tobias Burnus [Fri, 15 May 2020 14:40:34 +0000 (16:40 +0200)]
[Fortran] OpenMP 5 – permit more sharing clauses for SIMD (PR94690)
gcc/fortran/
PR fortran/94690
* openmp.c (resolve_omp_do): Permit more clauses for SIMD
iteration variables.
gcc/testsuite/
PR fortran/94690
* gfortran.dg/gomp/openmp-simd-4.f90: New test.
Uros Bizjak [Fri, 15 May 2020 14:22:19 +0000 (16:22 +0200)]
i386: Allow SI, DI and TImode pushes from XMM registers
Also change XMM register constraint from "x" to "v" in FP push insns.
gcc/ChangeLog:
2020-05-15 Uroš Bizjak <ubizjak@gmail.com>
* config/i386/i386.md (SWI48DWI): New mode iterator.
(*push<mode>2): Allow XMM registers.
(*pushdi2_rex64): Ditto.
(*pushsi2_rex64): Ditto.
(*pushsi2): Ditto.
(push XMM reg splitter): New splitter
(*pushdf) Change "x" operand constraint to "v".
(*pushsf_rex64): Ditto.
(*pushsf): Ditto.
Nathan Sidwell [Fri, 15 May 2020 13:34:20 +0000 (06:34 -0700)]
c++: Fix thinkos in template_args_equal change.
Arseny Solokha noticed I'd flubbed this patch, and it was not saying
what I thought it was saying. Unfortunately that didn't break
anything (otherwise I'd've noticed). Fixed thusly.
* pt.c (template_args_equal): Fix thinkos in previous 'cleanup'.
Richard Biener [Fri, 15 May 2020 09:14:53 +0000 (11:14 +0200)]
tree-optimization/92260 - improve fix
This improves the fix for PR92260 changing the number of vector
computation to the canonical one, not needing to look at the
using stmt.
2020-05-15 Richard Biener <rguenther@suse.de>
PR tree-optimization/92260
* tree-vect-slp.c (vect_get_constant_vectors): Compute
the number of vector stmts in a canonical way.
Martin Liska [Fri, 15 May 2020 10:34:12 +0000 (12:34 +0200)]
Fix clang [-Wmisleading-indentation] in hsa-gen.c.
* hsa-gen.c (get_symbol_for_decl): Fix misleading indentation
warning.
Andrew Stubbs [Tue, 28 Apr 2020 19:48:51 +0000 (20:48 +0100)]
WIP amdgcn: use unsigned extend for lshiftrt
This fixes a wrong-code logic error in a previous patch.
Detected by gcc.c-torture/execute/pr53645-2.c.
2020-05-15 Andrew Stubbs <ams@codesourcery.com>
gcc/
* config/gcn/gcn-valu.md (v<expander><mode>3): Fix unsignedp.
Andrew Burgess [Mon, 11 May 2020 21:32:35 +0000 (22:32 +0100)]
contrib: Handle GDB specific test result types
This commit is for the benefit of GDB, but as the binutils-gdb
repository shares the contrib/ directory with gcc, this commit must
first be applied to gcc then copied back to binutils-gdb.
This commit extends the two scripts contrib/dg-extract-results.{py,sh}
to handle some new, GDB specific test result types. These test
results types should never appear in GCC, or any other tool that
shares the contrib/ directly, so this change should be harmless.
In this patch series:
https://sourceware.org/pipermail/gdb-patches/2020-April/167847.html
changes were made in GDB's use of Dejagnu so that two additional
conditions could be detected, these are:
1. Test names that contain either the build or source paths. Such
test names make it difficult to compare the results of two test runs
of GDB from two different directories, and
2. Duplicate test names. Duplicates make it difficult to track down
exactly which test has failed.
When running Dejagnu on GDB we can now (sometimes) see two additional
test result types matching the above conditions, these are '# of paths
in test names' and '# of duplicate test names'.
If the test is run in parallel mode (make -j...) then these extra test
results will appear in the individual test summary files, but are not
merged into the final summary file.
Additionally, within the summary file there are now two new types of
test summary line, these are 'PATH: ...' and 'DUPLICATE: ...', these
allow users to quickly search the test summary to track down where the
offending test names are. These lines are similarly not merged into
the unified gdb.sum file after a parallel test run.
This commit extends the dg-extract-results.* scripts to calculate the
totals for the two new result types, and to copy the new test summary
lines into the unified summary file.
contrib/ChangeLog:
* dg-extract-results.py: Handle GDB specific test types.
* dg-extract-results.sh: Likewise.
Richard Biener [Fri, 15 May 2020 07:38:54 +0000 (09:38 +0200)]
tree-optimization/95133 - avoid abnormal edges in path splitting
When path splitting tries to detect a CFG diamond make sure it
is composed of normal (non-EH, not abnormal) edges. Otherwise
CFG manipulation later may fail.
2020-05-15 Richard Biener <rguenther@suse.de>
PR tree-optimization/95133
* gimple-ssa-split-paths.c
(find_block_to_duplicate_for_splitting_paths): Check for
normal edges.
* gcc.dg/pr95133.c: New testcase.
Christophe Lyon [Mon, 4 May 2020 13:42:03 +0000 (13:42 +0000)]
arm: Add support for interrupt routines to reg_needs_saving_p
reg_needs_saving_p is only used when dealing with non-interrupt
routines, but it makes sense to extend it to support that context too,
and make arm_compute_save_reg0_reg12_mask use it.
Save only live registers for non-leaf functions, but assume a callee
could clobber any register.
2020-05-15 Christophe Lyon <christophe.lyon@linaro.org>
gcc/
* config/arm/arm.c (reg_needs_saving_p): Add support for interrupt
routines.
(arm_compute_save_reg0_reg12_mask): Use reg_needs_saving_p.