vries [Mon, 17 Dec 2018 21:26:49 +0000 (21:26 +0000)]
[nvptx] Unify C/Fortran routine handling in nvptx_goacc_validate_dims
The Fortran front-end has a bug (PR72741) that means what when
nvptx_goacc_validate_dims is called for a Fortran routine, the dims parameter
is not the same as it would have been if the function would have been called for
an equivalent C routine.
Work around this bug by overriding the dims parameter for routines, allowing the
function to handle routines in Fortran and C the same.
Build and reg-tested on x86_64 with nvptx accelerator.
2018-12-17 Tom de Vries <tdevries@suse.de>
* config/nvptx/nvptx.c (nvptx_goacc_validate_dims): Work around Fortran
bug PR72741 by overriding dims parameter for routines.
vries [Mon, 17 Dec 2018 21:26:39 +0000 (21:26 +0000)]
[nvptx] Rewrite nvptx_goacc_validate_dims to use predicate vars
The function nvptx_goacc_validate_dims has arguments decl and fn_level which
together describe different situations.
Introduce a predicate var for each situation, and use them, allowing to
understand what the function does in each situation without having to know the
way the situations are encoded in the args.
Build and reg-tested on x86_64 with nvptx accelerator.
2018-12-17 Tom de Vries <tdevries@suse.de>
* config/nvptx/nvptx.c (nvptx_goacc_validate_dims): Rewrite using
predicate vars.
sje [Mon, 17 Dec 2018 19:14:04 +0000 (19:14 +0000)]
2018-12-17 Steve Ellcey <sellcey@cavium.com>
* config/aarch64/aarch64-protos.h (aarch64_use_simple_return_insn_p):
New prototype.
(aarch64_epilogue_uses): Ditto.
* config/aarch64/aarch64.c (aarch64_attribute_table): New array.
(aarch64_simd_decl_p): New function.
(aarch64_reg_save_mode): New function.
(aarch64_function_ok_for_sibcall): Check for simd calls.
(aarch64_layout_frame): Check for simd function.
(aarch64_gen_storewb_pair): Handle E_TFmode.
(aarch64_push_regs): Use aarch64_reg_save_mode to get mode.
(aarch64_gen_loadwb_pair): Handle E_TFmode.
(aarch64_pop_regs): Use aarch64_reg_save_mode to get mode.
(aarch64_gen_store_pair): Handle E_TFmode.
(aarch64_gen_load_pair): Ditto.
(aarch64_save_callee_saves): Handle different mode sizes.
(aarch64_restore_callee_saves): Ditto.
(aarch64_components_for_bb): Check for simd function.
(aarch64_epilogue_uses): New function.
(aarch64_process_components): Check for simd function.
(aarch64_expand_prologue): Ditto.
(aarch64_expand_epilogue): Ditto.
(aarch64_expand_call): Ditto.
(aarch64_use_simple_return_insn_p): New function.
(TARGET_ATTRIBUTE_TABLE): New define.
* config/aarch64/aarch64.h (EPILOGUE_USES): Redefine.
(FP_SIMD_SAVED_REGNUM_P): New macro.
* config/aarch64/aarch64.md (simple_return): New define_expand.
(load_pair_dw_tftf): New instruction.
(store_pair_dw_tftf): Ditto.
(loadwb_pair<TX:mode>_<P:mode>): Ditto.
(storewb_pair<TX:mode>_<P:mode>): Ditto.
Backports a fix where a bad cast to TypeFunction resulted in memory
corruption. The logic in the function semantic has been fixed, and
casts have been replaced with a function call to always check the
front-end AST node value.
which may expand hash table even if no insertion is neeed and change hash
table traverse order. When output_macinfo_op is called, all index strings
have been added to hash table by save_macinfo_strings and we shouldn't
expand index string hash table. Otherwise find_slot_with_hash will expand
hash table when hash table has the right size and hash table traverse of
output_index_string will have a different traverse order from index_string.
PR debug/79342
* dwarf2out.c (find_AT_string_in_table): Add insert argument
defaulting to INSERT and replace INSERT.
(find_AT_string): Likewise.
(output_macinfo_op): Pass NO_INSERT to find_AT_string.
rsandifo [Mon, 17 Dec 2018 10:05:51 +0000 (10:05 +0000)]
Add a loop versioning pass
This patch adds a pass that versions loops with variable index strides
for the case in which the stride is 1. E.g.:
for (int i = 0; i < n; ++i)
x[i * stride] = ...;
becomes:
if (stepx == 1)
for (int i = 0; i < n; ++i)
x[i] = ...;
else
for (int i = 0; i < n; ++i)
x[i * stride] = ...;
This is useful for both vector code and scalar code, and in some cases
can enable further optimisations like loop interchange or pattern
recognition.
The pass gives a 7.6% improvement on Cortex-A72 for 554.roms_r at -O3
and a 2.4% improvement for 465.tonto. I haven't found any SPEC tests
that regress.
Sizewise, there's a 10% increase in .text for both 554.roms_r and
465.tonto. That's obviously a lot, but in tonto's case it's because
the whole program is written using assumed-shape arrays and pointers,
so a large number of functions really do benefit from versioning.
roms likewise makes heavy use of assumed-shape arrays, and that
improvement in performance IMO justifies the code growth.
The next biggest .text increase is 4.5% for 548.exchange2_r. I did see
a small (0.4%) speed improvement there, but although both 3-iteration runs
produced stable results, that might still be noise. There was a slightly
larger (non-noise) improvement for a 256-bit SVE model.
481.wrf and 521.wrf_r .text grew by 2.8% and 2.5% respectively, but
without any noticeable improvement in performance. No other test grew
by more than 2%.
Although the main SPEC beneficiaries are all Fortran tests, the
benchmarks we use for SVE also include some C and C++ tests that
benefit.
Using -frepack-arrays gives the same benefits in many Fortran cases.
The problem is that using that option inappropriately can force a full
array copy for arguments that the function only reads once, and so it
isn't really something we can turn on by default. The new pass is
supposed to give most of the benefits of -frepack-arrays without
the risk of unnecessary repacking.
The patch therefore enables the pass by default at -O3.
2018-12-17 Richard Sandiford <richard.sandiford@arm.com>
Ramana Radhakrishnan <ramana.radhakrishnan@arm.com>
Kyrylo Tkachov <kyrylo.tkachov@arm.com>
gcc/
* doc/invoke.texi (-fversion-loops-for-strides): Document
(loop-versioning-group-size, loop-versioning-max-inner-insns)
(loop-versioning-max-outer-insns): Document new --params.
* Makefile.in (OBJS): Add gimple-loop-versioning.o.
* common.opt (fversion-loops-for-strides): New option.
* opts.c (default_options_table): Enable fversion-loops-for-strides
at -O3.
* params.def (PARAM_LOOP_VERSIONING_GROUP_SIZE)
(PARAM_LOOP_VERSIONING_MAX_INNER_INSNS)
(PARAM_LOOP_VERSIONING_MAX_OUTER_INSNS): New parameters.
* passes.def: Add pass_loop_versioning.
* timevar.def (TV_LOOP_VERSIONING): New time variable.
* tree-ssa-propagate.h
(substitute_and_fold_engine::substitute_and_fold): Add an optional
block parameter.
* tree-ssa-propagate.c
(substitute_and_fold_engine::substitute_and_fold): Likewise.
When passed, only walk blocks dominated by that block.
* tree-vrp.h (range_includes_p): Declare.
(range_includes_zero_p): Turn into an inline wrapper around
range_includes_p.
* tree-vrp.c (range_includes_p): New function, generalizing...
(range_includes_zero_p): ...this.
* tree-pass.h (make_pass_loop_versioning): Declare.
* gimple-loop-versioning.cc: New file.
ibuclaw [Sun, 16 Dec 2018 23:18:25 +0000 (23:18 +0000)]
libphobos: Merge common version blocks for core.sys.posix.sys.msg.
This is a continuation of simplifying C bindings so there aren't dozens
of duplicated code for each architecture. For this particular module,
it now more closely resembles how glibc arranges msq.h, fixing a couple
of targets in the process, notably X32.
kargl [Sun, 16 Dec 2018 16:29:43 +0000 (16:29 +0000)]
2018-12-16 Steven G. Kargl <kargl@gcc.gnu.org>
PR fortran/88116
PR fortran/88467
* array.c (gfc_match_array_constructor): Check return value of
gfc_convert_type(). Skip constructor elements with BT_UNKNOWN,
which need to go through resolution.
* intrinsic.c (gfc_convert_type_warn): Return early if the types
martch (i.e., no conversion is required).
* simplify.c (gfc_convert_constant): Remove a gfc_internal_error,
and return gfc_bad_expr.
2018-12-16 Steven G. Kargl <kargl@gcc.gnu.org>
PR fortran/88116
* gfortran.dg/pr88116_1.f90: New test.
* gfortran.dg/pr88116_2.f90: Ditto.
PR fortran/88467
* gfortran.dg/pr88467.f90: New test.
hubicka [Sun, 16 Dec 2018 12:05:04 +0000 (12:05 +0000)]
* ipa-fnsummary.c (analyze_function_body): Do not loeak conds and
size_time_table.
(ipa_fn_summary_generate): Add prevails parameter; do not allocate
data when symbol is not prevailing.
(inline_read_section): Likewise.
jakub [Sat, 15 Dec 2018 23:51:31 +0000 (23:51 +0000)]
PR c++/88482
* except.c (verify_library_fn): New function.
(declare_library_fn): Use it. Initialize TM even if the non-TM
library function has been user declared.
(do_end_catch): Don't set TREE_NOTHROW on error_mark_node.
(expand_start_catch_block): Don't call initialize_handler_parm
for error_mark_node.
(build_throw): Use verify_library_fn. Initialize TM even if the
non-TM library function has been user declared. Don't crash if
any library fn is error_mark_node.
* g++.dg/eh/builtin5.C: New test.
* g++.dg/eh/builtin6.C: New test.
* g++.dg/eh/builtin7.C: New test.
* g++.dg/eh/builtin8.C: New test.
* g++.dg/eh/builtin9.C: New test.
* g++.dg/eh/builtin10.C: New test.
* g++.dg/eh/builtin11.C: New test.
* g++.dg/parse/crash55.C: Adjust expected diagnostics.
* eh_cpp.cc (__cxa_throw): Change DEST argument type from
void * to void (*) (void *).
(_ITM_cxa_throw): Likewise.
* libitm.h (_ITM_cxa_throw): Likewise.
* libitm.texi (_ITM_cxa_throw): Likewise.
jason [Sat, 15 Dec 2018 22:10:18 +0000 (22:10 +0000)]
Small lambda instantiation tweak.
While looking at something else I noticed that we were passing 0 to the
"nonclass" parameter here; we might as well pass 1, since capture proxies
are always at block scope.
* pt.c (tsubst_expr) [DECL_EXPR]: Ignore class-scope bindings when
looking up a capture proxy.
jakub [Sat, 15 Dec 2018 11:04:41 +0000 (11:04 +0000)]
PR tree-optimization/88464
PR target/88498
* tree-vect-stmts.c (vect_build_gather_load_calls): For NARROWING
and mask with integral masktype, don't try to permute mask vectors,
instead emit VEC_UNPACK_{LO,HI}_EXPR. Fix up NOP_EXPR operand.
(vectorizable_store): Handle masked scatters with decl and integral
mask type.
(permute_vec_elements): Allow scalar_dest to be NULL.
* config/i386/i386.c (ix86_get_builtin)
<case IX86_BUILTIN_GATHER3ALTDIV16SF>: Use lowpart_subreg for masks.
<case IX86_BUILTIN_GATHER3ALTDIV8SF>: Don't assume mask and src have
to be the same.
* gcc.target/i386/avx512f-pr88462-1.c: Rename to ...
* gcc.target/i386/avx512f-pr88464-1.c: ... this. Fix up PR number.
Expect 4 vectorized loops instead of 3.
(f4): New function.
* gcc.target/i386/avx512f-pr88462-2.c: Rename to ...
* gcc.target/i386/avx512f-pr88464-2.c: ... this. Fix up PR number
and #include.
(avx512f_test): Prepare arguments for f4 and check the results.
* gcc.target/i386/avx512f-pr88464-3.c: New test.
* gcc.target/i386/avx512f-pr88464-4.c: New test.
jakub [Sat, 15 Dec 2018 11:02:28 +0000 (11:02 +0000)]
PR tree-optimization/88464
PR target/88498
* tree-vect-stmts.c (vect_build_gather_load_calls): For NARROWING
and mask with integral masktype, don't try to permute mask vectors,
instead emit VEC_UNPACK_{LO,HI}_EXPR. Fix up NOP_EXPR operand.
(vectorizable_store): Handle masked scatters with decl and integral
mask type.
(permute_vec_elements): Allow scalar_dest to be NULL.
* config/i386/i386.c (ix86_get_builtin)
<case IX86_BUILTIN_GATHER3ALTDIV16SF>: Use lowpart_subreg for masks.
<case IX86_BUILTIN_GATHER3ALTDIV8SF>: Don't assume mask and src have
to be the same.
* gcc.target/i386/avx512f-pr88462-1.c: Rename to ...
* gcc.target/i386/avx512f-pr88464-1.c: ... this. Fix up PR number.
Expect 4 vectorized loops instead of 3.
(f4): New function.
* gcc.target/i386/avx512f-pr88462-2.c: Rename to ...
* gcc.target/i386/avx512f-pr88464-2.c: ... this. Fix up PR number
and #include.
(avx512f_test): Prepare arguments for f4 and check the results.
* gcc.target/i386/avx512f-pr88464-3.c: New test.
* gcc.target/i386/avx512f-pr88464-4.c: New test.
hubicka [Sat, 15 Dec 2018 10:31:37 +0000 (10:31 +0000)]
* ipa.c (cgraph_build_static_cdtor_1): Add OPTIMIZATION and TARGET
parameters.
(cgraph_build_static_cdtor): Update.
(build_cdtor): Use OPTIMIZATION and TARGET of the first real cdtor
callsed.
amodra [Sat, 15 Dec 2018 07:13:19 +0000 (07:13 +0000)]
[RS6000] Use gen_hard_reg_clobber in rs6000.c
I noticed when looking at PR88311 that rs6000_call_sysv should be
using gen_hard_reg_clobber (as the sysv call insns did prior to
introducing rs6000_call_sysv). This patch fixes that minor
regression, and other like places in rs6000.c.
* config/rs6000/rs6000.c (generate_set_vrsave, rs6000_emit_savres_rtx),
(rs6000_emit_prologue, rs6000_call_aix, rs6000_call_sysv),
(rs6000_call_darwin_1): Use gen_hard_reg_clobber.
jakub [Fri, 14 Dec 2018 23:21:10 +0000 (23:21 +0000)]
PR target/88489
* config/i386/sse.md (UNSPEC_SFIXUPIMM): New unspec enumerator.
(avx512f_sfixupimm<mode><mask_name><round_saeonly_name>): Use it
instead of UNSPEC_FIXUPIMM.
* gcc.target/i386/avx512vl-vfixupimmsd-2.c: New test.
* gcc.target/i386/avx512vl-vfixupimmss-2.c: New test.
aoliva [Fri, 14 Dec 2018 21:57:07 +0000 (21:57 +0000)]
[PR c++/87814] undefer deferred noexcept on tsubst if request
tsubst_expr and tsubst_copy_and_build are not expected to handle
DEFERRED_NOEXCEPT exprs, but if tsubst_exception_specification takes a
DEFERRED_NOEXCEPT expr with !defer_ok, it just passes the expr on for
tsubst_copy_and_build to barf.
This patch arranges for tsubst_exception_specification to combine the
incoming args with those already stored in a DEFERRED_NOEXCEPT, and
then substitute them into the pattern, when retaining a deferred
noexcept is unacceptable.
for gcc/cp/ChangeLog
PR c++/87814
* pt.c (tsubst_exception_specification): Handle
DEFERRED_NOEXCEPT with !defer_ok.
hjl [Fri, 14 Dec 2018 21:35:36 +0000 (21:35 +0000)]
x86; Add -mmanual-endbr and cf_check function attribute
Currently GCC inserts ENDBR instruction at entries of all non-static
functions, unless LTO compilation is used. Marking all functions,
which are not called indirectly with nocf_check attribute, is not
ideal since 99% of functions in a program may be of this kind.
This patch adds -mmanual-endbr and cf_check function attribute. They
can be used together with -fcf-protection such that ENDBR instruction
is inserted only at entries of functions with cf_check attribute. It
can limit number of ENDBR instructions to reduce program size.
gcc/
* config/i386/i386.c (rest_of_insert_endbranch): Insert ENDBR
at the function entry only when -mmanual-endbr isn't used or
there is cf_check function attribute.
(ix86_attribute_table): Add cf_check.
* config/i386/i386.opt: Add -mmanual-endbr.
* doc/extend.texi: Document cf_check attribute.
* doc/invoke.texi: Document -mmanual-endbr.
tschwinge [Fri, 14 Dec 2018 20:43:12 +0000 (20:43 +0000)]
Missing changes from "Adjust copy/copyin/copyout/create for OpenACC 2.5"
Most of that patch's changes were already committed as part of r261813 "Update
OpenACC data clause semantics to the 2.5 behavior", but not all of them.
libgomp/
* oacc-mem.c (acc_present_or_create): Remove definition and change
to alias of acc_create.
(acc_present_or_copyin): Remove definition and change to alias of
acc_copyin.
* oacc-parallel.c (GOACC_enter_exit_data): Call acc_create instead
of acc_present_or_create.
* testsuite/libgomp.oacc-c-c++-common/data-already-1.c: Remove.
* testsuite/libgomp.oacc-c-c++-common/data-already-2.c: Likewise.
* testsuite/libgomp.oacc-c-c++-common/data-already-3.c: Likewise.
* testsuite/libgomp.oacc-c-c++-common/data-already-4.c: Likewise.
* testsuite/libgomp.oacc-c-c++-common/data-already-5.c: Likewise.
* testsuite/libgomp.oacc-c-c++-common/data-already-6.c: Likewise.
* testsuite/libgomp.oacc-c-c++-common/data-already-7.c: Likewise.
* testsuite/libgomp.oacc-c-c++-common/data-already-8.c: Likewise.
* testsuite/libgomp.oacc-fortran/data-already-1.f: Likewise.
* testsuite/libgomp.oacc-fortran/data-already-2.f: Likewise.
* testsuite/libgomp.oacc-fortran/data-already-3.f: Likewise.
* testsuite/libgomp.oacc-fortran/data-already-4.f: Likewise.
* testsuite/libgomp.oacc-fortran/data-already-5.f: Likewise.
* testsuite/libgomp.oacc-fortran/data-already-6.f: Likewise.
* testsuite/libgomp.oacc-fortran/data-already-7.f: Likewise.
* testsuite/libgomp.oacc-fortran/data-already-8.f: Likewise.
tschwinge [Fri, 14 Dec 2018 20:43:02 +0000 (20:43 +0000)]
[PR88495] An OpenACC async queue is always synchronized with itself
An OpenACC async queue is always synchronized with itself, so invocations like
"#pragma acc wait(0) async(0)", or "acc_wait_async (0, 0)" don't make a lot of
sense, but are still valid.
tschwinge [Fri, 14 Dec 2018 20:42:50 +0000 (20:42 +0000)]
[PR88484] OpenACC wait directive without wait argument but with async clause
We don't correctly handle "#pragma acc wait async (a)" for "a >= 0", handling
as a no-op whereas it should enqueue the appropriate wait operations on
"async (a)".
libgomp/
PR libgomp/88484
* oacc-parallel.c (GOACC_wait): Correct handling for "async >= 0".
* testsuite/libgomp.oacc-c-c++-common/asyncwait-nop-1.c: New file.
These are meant to be functionally equivalent (but no longer are), just using
different means. Also, use the OpenACC "*_async" functions recently added.
Per my reading of the OpenACC specification (and as supported by secondary
documentation, such as code examples, or presentations), it's valid to call
"acc_get_cuda_stream"/"acc_set_cuda_stream" also with "acc_async_sync",
"acc_async_noval" arguments, not just with the nonnegative values as currently
implemented.
tschwinge [Fri, 14 Dec 2018 20:41:46 +0000 (20:41 +0000)]
Repair liboffloadmic after "(Partial) OpenMP 5.0 support for GCC 9"
..., which now failed to build, as follows:
In file included from [...]/source-gcc/liboffloadmic/runtime/offload_common.h:43,
from [...]/source-gcc/liboffloadmic/runtime/dv_util.cpp:31:
[...]/source-gcc/liboffloadmic/runtime/offload.h:220:12: error: conflicting declaration of C function 'int omp_target_is_present(void*, int)'
220 | extern int omp_target_is_present(
| ^~~~~~~~~~~~~~~~~~~~~
In file included from [...]/source-gcc/liboffloadmic/runtime/offload.h:45,
from [...]/source-gcc/liboffloadmic/runtime/offload_common.h:43,
from [...]/source-gcc/liboffloadmic/runtime/dv_util.cpp:31:
./../libgomp/omp.h:166:12: note: previous declaration 'int omp_target_is_present(const void*, int)'
166 | extern int omp_target_is_present (const void *, int) __GOMP_NOTHROW;
| ^~~~~~~~~~~~~~~~~~~~~
In file included from [...]/source-gcc/liboffloadmic/runtime/offload_common.h:43,
from [...]/source-gcc/liboffloadmic/runtime/dv_util.cpp:31:
[...]/source-gcc/liboffloadmic/runtime/offload.h:236:12: error: conflicting declaration of C function 'int omp_target_memcpy(void*, void*, size_t, size_t, size_t, int, int)'
236 | extern int omp_target_memcpy(
| ^~~~~~~~~~~~~~~~~
In file included from [...]/source-gcc/liboffloadmic/runtime/offload.h:45,
from [...]/source-gcc/liboffloadmic/runtime/offload_common.h:43,
from [...]/source-gcc/liboffloadmic/runtime/dv_util.cpp:31:
./../libgomp/omp.h:167:12: note: previous declaration 'int omp_target_memcpy(void*, const void*, long unsigned int, long unsigned int, long unsigned int, int, int)'
167 | extern int omp_target_memcpy (void *, const void *, __SIZE_TYPE__,
| ^~~~~~~~~~~~~~~~~
In file included from [...]/source-gcc/liboffloadmic/runtime/offload_common.h:43,
from [...]/source-gcc/liboffloadmic/runtime/dv_util.cpp:31:
[...]/source-gcc/liboffloadmic/runtime/offload.h:262:12: error: conflicting declaration of C function 'int omp_target_memcpy_rect(void*, void*, size_t, int, const size_t*, const size_t*, const size_t*, const size_t*, const size_t*, int, int)'
262 | extern int omp_target_memcpy_rect(
| ^~~~~~~~~~~~~~~~~~~~~~
In file included from [...]/source-gcc/liboffloadmic/runtime/offload.h:45,
from [...]/source-gcc/liboffloadmic/runtime/offload_common.h:43,
from [...]/source-gcc/liboffloadmic/runtime/dv_util.cpp:31:
./../libgomp/omp.h:170:12: note: previous declaration 'int omp_target_memcpy_rect(void*, const void*, long unsigned int, int, const long unsigned int*, const long unsigned int*, const long unsigned int*, const long unsigned int*, const long unsigned int*, int, int)'
170 | extern int omp_target_memcpy_rect (void *, const void *, __SIZE_TYPE__, int,
| ^~~~~~~~~~~~~~~~~~~~~~
In file included from [...]/source-gcc/liboffloadmic/runtime/offload_common.h:43,
from [...]/source-gcc/liboffloadmic/runtime/dv_util.cpp:31:
[...]/source-gcc/liboffloadmic/runtime/offload.h:285:12: error: conflicting declaration of C function 'int omp_target_associate_ptr(void*, void*, size_t, size_t, int)'
285 | extern int omp_target_associate_ptr(
| ^~~~~~~~~~~~~~~~~~~~~~~~
In file included from [...]/source-gcc/liboffloadmic/runtime/offload.h:45,
from [...]/source-gcc/liboffloadmic/runtime/offload_common.h:43,
from [...]/source-gcc/liboffloadmic/runtime/dv_util.cpp:31:
./../libgomp/omp.h:177:12: note: previous declaration 'int omp_target_associate_ptr(const void*, const void*, long unsigned int, long unsigned int, int)'
177 | extern int omp_target_associate_ptr (const void *, const void *, __SIZE_TYPE__,
| ^~~~~~~~~~~~~~~~~~~~~~~~
In file included from [...]/source-gcc/liboffloadmic/runtime/offload_common.h:43,
from [...]/source-gcc/liboffloadmic/runtime/dv_util.cpp:31:
[...]/source-gcc/liboffloadmic/runtime/offload.h:299:12: error: conflicting declaration of C function 'int omp_target_disassociate_ptr(void*, int)'
299 | extern int omp_target_disassociate_ptr(
| ^~~~~~~~~~~~~~~~~~~~~~~~~~~
In file included from [...]/source-gcc/liboffloadmic/runtime/offload.h:45,
from [...]/source-gcc/liboffloadmic/runtime/offload_common.h:43,
from [...]/source-gcc/liboffloadmic/runtime/dv_util.cpp:31:
./../libgomp/omp.h:179:12: note: previous declaration 'int omp_target_disassociate_ptr(const void*, int)'
179 | extern int omp_target_disassociate_ptr (const void *, int) __GOMP_NOTHROW;
| ^~~~~~~~~~~~~~~~~~~~~~~~~~~
Makefile:904: recipe for target 'liboffloadmic_target_la-dv_util.lo' failed
make[3]: *** [liboffloadmic_target_la-dv_util.lo] Error 1
make[3]: Leaving directory '[...]/build-gcc-offload-x86_64-intelmicemul-linux-gnu/x86_64-intelmicemul-linux-gnu/liboffloadmic'
Makefile:1031: recipe for target 'all-recursive' failed
make[2]: *** [all-recursive] Error 1
make[2]: Leaving directory '[...]/build-gcc-offload-x86_64-intelmicemul-linux-gnu/x86_64-intelmicemul-linux-gnu/liboffloadmic'
Makefile:12707: recipe for target 'all-target-liboffloadmic' failed
make[1]: *** [all-target-liboffloadmic] Error 2
make[1]: Leaving directory '[...]/build-gcc-offload-x86_64-intelmicemul-linux-gnu'
Makefile:941: recipe for target 'all' failed
make: *** [all] Error 2
aoliva [Fri, 14 Dec 2018 20:06:15 +0000 (20:06 +0000)]
[PR86823] retain deferred access checks from outside firewall
We used to preserve deferred access check along with resolved template
ids, but a tentative parsing firewall introduced additional layers of
deferred access checks, so that we don't preserve the checks we
want to any more.
This patch moves the deferred access checks from outside the firewall
into it.
From: Jason Merrill <jason@redhat.com>
for gcc/cp/ChangeLog
PR c++/86823
* parser.c (cp_parser_template_id): Rearrange deferred access
checks into the firewall.
From: Alexandre Oliva <aoliva@redhat.com>
for gcc/testsuite/ChangeLog
jakub [Fri, 14 Dec 2018 19:37:38 +0000 (19:37 +0000)]
PR c++/82294
PR c++/87436
* expr.h (categorize_ctor_elements): Add p_unique_nz_elts argument.
* expr.c (categorize_ctor_elements_1): Likewise. Compute it like
p_nz_elts, except don't multiply it by mult. Adjust recursive call.
Fix up COMPLEX_CST handling.
(categorize_ctor_elements): Add p_unique_nz_elts argument, initialize
it and pass it through to categorize_ctor_elements_1.
(mostly_zeros_p, all_zeros_p): Adjust categorize_ctor_elements callers.
* gimplify.c (gimplify_init_constructor): Likewise. Don't force
ctor into readonly data section if num_unique_nonzero_elements is
smaller or equal to 1/8 of num_nonzero_elements and size is >= 64
bytes.
* g++.dg/tree-ssa/pr82294.C: New test.
* g++.dg/tree-ssa/pr87436.C: New test.
jakub [Fri, 14 Dec 2018 19:36:33 +0000 (19:36 +0000)]
PR c++/82294
PR c++/87436
* init.c (build_vec_init): Change num_initialized_elts type from int
to HOST_WIDE_INT. Build a RANGE_EXPR if e needs to be repeated more
than once.
thopre01 [Fri, 14 Dec 2018 19:20:29 +0000 (19:20 +0000)]
ARM] Improve robustness of -mslow-flash-data
Current code to handle -mslow-flash-data in machine description files
suffers from a number of issues which this patch fixes:
1) The insn_and_split in vfp.md to load a generic floating-point
constant via GPR first and move it to VFP register are guarded by
!reload_completed which is forbidden explicitely in the GCC internals
documentation section 17.2 point 3;
2) A number of testcase in the testsuite ICEs under -mslow-flash-data
when targeting the hardfloat ABI [1];
3) Instructions performing load from literal pool are not disabled.
These problems are addressed by 2 separate actions:
1) Making the splitters take a clobber and changing the expanders
accordingly to generate a mov with clobber in cases where a literal
pool would be used. The splitter can thus be enabled after reload since
it does not call gen_reg_rtx anymore;
2) Adding new predicates and constraints to disable literal pool loads
in existing instructions when -mslow-flash-data is in effect.
The patch also rework the splitter for DFmode slightly to generate an
intermediate DI load instead of 2 intermediate SI loads, thus relying on
the existing DI splitters instead of redoing their job. At last, the
patch adds some missing arm_fp_ok effective target to some of the
slow-flash-data testcases.
[1]
c-c++-common/Wunused-var-3.c
gcc.c-torture/compile/pr72771.c
gcc.c-torture/compile/vector-5.c
gcc.c-torture/compile/vector-6.c
gcc.c-torture/execute/20030914-1.c
gcc.c-torture/execute/20050316-1.c
gcc.c-torture/execute/pr59643.c
gcc.dg/builtin-tgmath-1.c
gcc.dg/debug/pr55730.c
gcc.dg/graphite/interchange-7.c
gcc.dg/pr56890-2.c
gcc.dg/pr68474.c
gcc.dg/pr80286.c
gcc.dg/torture/pr35227.c
gcc.dg/torture/pr65077.c
gcc.dg/torture/pr86363.c
g++.dg/torture/pr81112.C
g++.dg/torture/pr82985.C
g++.dg/warn/Wunused-var-7.C
and a lot more in libstdc++ in special_functions/*_comp_ellint_* and
special_functions/*_ellint_* directories.
2018-12-14 Thomas Preud'homme <thomas.preudhomme@arm.com>
gcc/
* config/arm/arm.md (arm_movdi): Split if -mslow-flash-data and
source is a constant that would be loaded by literal pool.
(movsf expander): Generate a no_literal_pool_sf_immediate insn if
-mslow-flash-data is present, targeting hardfloat ABI and source is a
float constant that cannot be loaded via vmov.
(movdf expander): Likewise but generate a no_literal_pool_df_immediate
insn.
(arm_movsf_soft_insn): Split if -mslow-flash-data and source is a
float constant that would be loaded by literal pool.
(softfloat constant movsf splitter): Splitter for the above case.
(movdf_soft_insn): Split if -mslow-flash-data and source is a float
constant that would be loaded by literal pool.
(softfloat constant movdf splitter): Splitter for the above case.
* config/arm/constraints.md (Pz): Document existing constraint.
(Ha): Define constraint.
(Tu): Likewise.
* config/arm/predicates.md (hard_sf_operand): New predicate.
(hard_df_operand): Likewise.
* config/arm/thumb2.md (thumb2_movsi_insn): Split if
-mslow-flash-data and constant would be loaded by literal pool.
* constant/arm/vfp.md (thumb2_movsi_vfp): Likewise and disable constant
load in VFP register.
(movdi_vfp): Likewise.
(thumb2_movsf_vfp): Use hard_sf_operand as predicate for source to
prevent match for a constant load if -mslow-flash-data and constant
cannot be loaded via vmov. Adapt constraint accordingly by
using Ha instead of E for generic floating-point constant load.
(thumb2_movdf_vfp): Likewise using hard_df_operand predicate instead.
(no_literal_pool_df_immediate): Add a clobber to use as the
intermediate general purpose register and also enable it after reload
but disable it constant is a valid FP constant. Add constraints and
generate a DI intermediate load rather than 2 SI loads.
(no_literal_pool_sf_immediate): Add a clobber to use as the
intermediate general purpose register and also enable it after
reload.
2018-11-14 Thomas Preud'homme <thomas.preudhomme@arm.com>
vries [Fri, 14 Dec 2018 13:48:56 +0000 (13:48 +0000)]
[offloading] Error on missing symbols
When compiling an OpenMP or OpenACC program containing a reference in the
offloaded code to a symbol that has not been included in the offloaded code,
the offloading compiler may ICE in lto1.
Fix this by erroring out instead, mentioning the problematic symbol:
...
error: variable 'var' has been referenced in offloaded code but hasn't
been marked to be included in the offloaded code
lto1: fatal error: errors during merging of translation units
compilation terminated.
...
Build x86_64 with nvptx accelerator and reg-tested libgomp.
Build x86_64 and reg-tested libgomp.
2018-12-14 Tom de Vries <tdevries@suse.de>
* lto-cgraph.c (verify_node_partition): New function.
(input_overwrite_node, input_varpool_node): Use verify_node_partition.
* testsuite/libgomp.c-c++-common/function-not-offloaded-aux.c: New test.
* testsuite/libgomp.c-c++-common/function-not-offloaded.c: New test.
* testsuite/libgomp.c-c++-common/variable-not-offloaded.c: New test.
* testsuite/libgomp.oacc-c-c++-common/function-not-offloaded.c: New test.
* testsuite/libgomp.oacc-c-c++-common/variable-not-offloaded.c: New test.
hjl [Fri, 14 Dec 2018 12:38:04 +0000 (12:38 +0000)]
x86: Don't use get_frame_size when finalizing stack frame
get_frame_size () returns used stack slots during compilation, which
may be optimized out later. Since ix86_find_max_used_stack_alignment
is called by ix86_finalize_stack_frame_flags to check if stack frame
is required, there is no need to call get_frame_size () which may give
inaccurate final stack frame size.
Tested on AVX512 machine configured with
--with-arch=native --with-cpu=native
gcc/
PR target/88483
* config/i386/i386.c (ix86_finalize_stack_frame_flags): Don't
use get_frame_size ().
gcc/testsuite/
PR target/88483
* gcc.target/i386/stackalign/pr88483.c: New test.
ams [Fri, 14 Dec 2018 11:49:28 +0000 (11:49 +0000)]
Fix LRA bug
This patch fixes an ICE building libgfortran/random.c.
The problem was an adddi3 instruction that had an eliminable frame pointer.
GCN adddi3 includes a match_scratch, which LRA substitutes with a REG, and
checks if it can be converted back to a scratch afterwards. In the meantime,
the add was converted to a move, meaning that the instruction pattern
completely changed, thus causing a segfault when the instruction is revisited
in restore_scratches.
2018-12-14 Andrew Stubbs <ams@codesourcery.com>
gcc/
* gcc/lra-int.h (lra_register_new_scratch_op): Add third parameter.
* gcc/lra-remat.c (update_scratch_ops): Pass icode to
lra_register_new_scratch_op.
* gcc/lra.c (struct sloc): Add icode field.
(lra_register_new_scratch_op): Add icode parameter.
Use icode to skip insns that have changed beyond recognition.
ebotcazou [Fri, 14 Dec 2018 11:08:15 +0000 (11:08 +0000)]
* gcc-interface/decl.c (rm_size): Take into account the padding in
the case of a record type containing a template.
* gcc-interface/trans.c (Attribute_to_gnu) <Attr_Size>: Likewise.
Do not subtract the padded size for Max_Size_In_Storage_Elements.
<Attr_Descriptor_Size>: Tweak comment.
ebotcazou [Fri, 14 Dec 2018 10:53:41 +0000 (10:53 +0000)]
* dwarf2out.c (analyze_discr_in_predicate): Simplify.
(analyze_variants_discr): Deal with naked boolean discriminants.
ada/
* gcc-interface/decl.c (choices_to_gnu): Directly use a naked boolean
discriminant if the value is the boolean true.
amodra [Fri, 14 Dec 2018 09:03:50 +0000 (09:03 +0000)]
[RS6000] PR88311, mlongcall indirections are optimised away
Masking CALL_LONG from the cookie was done in order to simplify and
correct length attribute calculations for indirect calls at one point
in my call series tidy when the indirect patterns used alternatives
"0,n" on the cookie operand. (Leaving the CALL_LONG in place
calculated the wrong length for long calls without fp args.)
This is no longer necessary now that the indirect sysv call patterns
explicitly test for the fp arg bits in their length attribute
expressions. And without the CALL_LONG to disable insns like
call_value_local_svsv, combine merrily replaces the indirect long call
sequence with a direct call. As it should. This patch reinstates
the CALL_LONG bit.
PR rtl-optimization/88311
* config/rs6000/rs6000.c (rs6000_call_sysv): Do not mask cookie.
(rs6000_sibcall_sysv): Likewise.
jakub [Thu, 13 Dec 2018 23:33:57 +0000 (23:33 +0000)]
PR rtl-optimization/88470
* cfgcleanup.c (outgoing_edges_match): If the function is
shrink-wrapped and bb1 ends with a JUMP_INSN with a single fake
edge to EXIT, return false.
mpolacek [Thu, 13 Dec 2018 20:47:32 +0000 (20:47 +0000)]
PR c++/88216 - ICE with class type in non-type template parameter.
* mangle.c (write_expression): Handle TARGET_EXPR and
VIEW_CONVERT_EXPR.
* pt.c (convert_nontype_argument): Don't call
get_template_parm_object for value-dependent expressions.
redi [Thu, 13 Dec 2018 20:34:10 +0000 (20:34 +0000)]
Fix handling of POSIX paths containing a root-name
Fix path appending and concatenating to work correctly for a leading
root-name. Check a new macro, SLASHSLASH_IS_ROOT_NAME, instead of making
the behaviour depend directly on __CYGWIN__.
* src/filesystem/std-path.cc (SLASHSLASH_IS_ROOT_NAME): New macro to
control whether interpret paths with two slashes as a root-name.
(path::operator/=(const path&)) [SLASHSLASH_IS_ROOT_NAME]: Add a
root-directory when appending to a root-name.
(path::_M_append(basic_string_view<value_type>))
[SLASHSLASH_IS_ROOT_NAME]: Likewise.
(path::operator/=(const path&)) [SLASHSLASH_IS_ROOT_NAME]: Likewise.
(path::_M_concat(basic_string_view<value_type>))
[SLASHSLASH_IS_ROOT_NAME]: Likewise.
(path::lexically_normal()) [SLASHSLASH_IS_ROOT_NAME]: Use += instead
of /= to add a root-directory to the result.
* testsuite/27_io/filesystem/path/decompose/root_directory.cc: Fix
expected result for Cygwin.
redi [Thu, 13 Dec 2018 20:33:55 +0000 (20:33 +0000)]
PR libstdc++/71044 optimize std::filesystem::path construction
This new implementation has a smaller footprint than the previous
implementation, due to replacing std::vector<_Cmpt> with a custom pimpl
type that only needs a single pointer. The _M_type enumeration is also
combined with the pimpl type, by using a tagged pointer, reducing
sizeof(path) further still.
Construction and modification of paths is now done more efficiently, by
splitting the input into a stack-based buffer of string_view objects
instead of a dynamically-allocated vector containing strings. Once the
final size is known only a single allocation is needed to reserve space
for it. The append and concat operations no longer require constructing
temporary path objects, nor re-parsing the entire native pathname.
This results in algorithmic improvements to path construction, and
working with large paths is much faster.
PR libstdc++/71044
* include/bits/fs_path.h (path::path(path&&)): Add noexcept when
appropriate. Move _M_cmpts instead of reparsing the native pathname.
(path::operator=(const path&)): Do not define as defaulted.
(path::operator/=, path::append): Call _M_append.
(path::concat): Call _M_concat.
(path::path(string_type, _Type): Change type of first parameter to
basic_string_view<value_type>.
(path::_M_append(basic_string_view<value_type>)): New member function.
(path::_M_concat(basic_string_view<value_type>)): New member function.
(_S_convert(value_type*, __null_terminated)): Return string view.
(_S_convert(const value_type*, __null_terminated)): Return string view.
(_S_convert(value_type*, value_type*))
(_S_convert(const value_type*, const value_type*)): Add overloads for
pairs of pointers.
(_S_convert(_InputIterator, __null_terminated)): Construct string_type
explicitly, for cases where _S_convert returns a string view.
(path::_S_is_dir_sep): Replace with non-member is_dir_sep.
(path::_M_trim, path::_M_add_root_name, path::_M_add_root_dir)
(path::_M_add_filename): Remove.
(path::_M_type()): New member function to replace _M_type data member.
(path::_List): Define new struct type instead of using std::vector.
(path::_Cmpt::_Cmpt(string_type, _Type, size_t)): Change type of
first parameter to basic_string_view<value_type>.
(path::operator+=(const path&)): Do not define inline.
(path::operator+=(const string_type&)): Call _M_concat.
(path::operator+=(const value_type*)): Likewise.
(path::operator+=(value_type)): Likewise.
(path::operator+=(basic_string_view<value_type>)): Likewise.
(path::operator/=(const path&)): Do not define inline.
(path::_M_append(path)): Remove.
* python/libstdcxx/v6/printers.py (StdPathPrinter): New printer that
understands the new path::_List type.
* src/filesystem/std-path.cc (is_dir_sep): New function to replace
path::_S_is_dir_sep.
(path::_Parser): New helper class to parse strings as paths.
(path::_List::_Impl): Define container type for path components.
(path::_List): Define members.
(path::operator=(const path&)): Define explicitly, to provide the
strong exception safety guarantee.
(path::operator/=(const path&)): Implement manually by processing
each component of the argument, rather than using _M_split_cmpts
to parse the entire string again.
(path::_M_append(string_type)): Likewise.
(path::operator+=(const path&)): Likewise.
(path::_M_concat(string_type)): Likewise.
(path::remove_filename()): Perform trim directly instead of calling
_M_trim().
(path::_M_split_cmpts()): Rewrite in terms of _Parser class.
(path::_M_trim, path::_M_add_root_name, path::_M_add_root_dir)
(path::_M_add_filename): Remove.
* testsuite/27_io/filesystem/path/append/source.cc: Test appending a
string view that aliases the path.
testsuite/27_io/filesystem/path/concat/strings.cc: Test concatenating
a string view that aliases the path.
Libgomp test-case libgomp.c/target-5.c is failing to compile when building for
x86_64 with nvptx accelerator due to missing:
- getpid
- gethostname
- isatty (pulled in by fwrite)
in the nvptx newlib.
This patch fixes the build failure by:
- adding a function gomp_print_string which limits the use of fwrite to a single
location (in affinity-fmt.c), and
- creating an nvptx version of affinity-fmt.c, which:
- overrides the configure test results HAVE_GETPID and HAVE_GETHOSTNAME, and
- implements fwrite using write.
Build and reg-tested on x86_64 with nvptx accelerator.
2018-12-13 Tom de Vries <tdevries@suse.de>
* affinity-fmt.c (gomp_print_string): New function, factored out of ...
(omp_display_affinity, gomp_display_affinity_thread): ... here, and ...
* fortran.c (omp_display_affinity_): ... here.
* libgomp.h (gomp_print_string): Declare.
* config/nvptx/affinity-fmt.c: New file. Include affinity-fmt.c,
undefining HAVE_GETPID and HAVE_GETHOSTNAME, and mapping fwrite to
write.
jakub [Thu, 13 Dec 2018 17:01:50 +0000 (17:01 +0000)]
PR tree-optimization/88464
* tree-vect-stmts.c (vect_build_gather_load_calls): Handle INTEGER_TYPE
masktype if mask is non-NULL.
(vectorizable_load): Don't reject masked gather loads if masktype
in the decl is INTEGER_TYPE.
* gcc.target/i386/avx512f-pr88462-1.c: New test.
* gcc.target/i386/avx512f-pr88462-2.c: New test.
jakub [Thu, 13 Dec 2018 12:53:19 +0000 (12:53 +0000)]
PR libgomp/88460
* testsuite/libgomp.c++/for-24.C (results): Include it in
omp declare target region.
(main): Use map (always, tofrom: results) instead of
map (tofrom: results).
jakub [Thu, 13 Dec 2018 12:52:11 +0000 (12:52 +0000)]
PR testsuite/88454
* gcc.dg/tree-ssa/split-path-5.c (__ctype_ptr__): Change type from
const char * to const signed char *.
(bmhi_init): Change pattern parameter's type the same. Use
__builtin_strlen instead of undeclared strlen.
krebbel [Thu, 13 Dec 2018 11:57:16 +0000 (11:57 +0000)]
S/390: Fix zvector vec_double builtin
The (unsigned) long int to double vector conversion instructions
expect 2 immediate parameters. One for the inexact suppression
control and another one for the rounding mode. However, the
vec_double builtin has just the vector source operand. The 2
addtional operands need to be added with an intermediate expander.
The expanders were already there but unfortunately not wired up
correctly to the builtin.
gcc/ChangeLog:
2018-12-13 Andreas Krebbel <krebbel@linux.ibm.com>
* config/s390/s390-builtins.def (s390_vec_double_s64): Map to
s390_vec_double_s64 instead of s390_vcdgb.
(s390_vec_double_u64): Map to s390_vec_double_u64 instead of
s390_vcdlgb.
gcc/testsuite/ChangeLog:
2018-12-13 Andreas Krebbel <krebbel@linux.ibm.com>
* gcc.target/s390/zvector/vec-double-1.c: New test.
* gcc.target/s390/zvector/vec-double-2.c: New test.
redi [Thu, 13 Dec 2018 11:01:03 +0000 (11:01 +0000)]
Disable new tests for configurations with no libstdc++fs.a
* testsuite/27_io/filesystem/path/construct/80762.cc: Skip test if
the Filesystem TS support is not configured.
* testsuite/experimental/filesystem/path/construct/80762.cc: Likewise.
We've received reports about the -msve-vector-bits=128 bits being somewhat ambiguous.
It isn't clear whether -msve-vector-bits=128 forces vector-length-agnostic code or whether
-msve-vector-bits=scalable forces 128-bit vector-lengh-specific code.
The latter is a, perhaps unintuitive, reading that we want to exclude.
This patch makes it more explicit that -msve-vector-bits=128 is special and produces
vector-length *agnostic* code. In the end, I've rewritten the whole option documentation.
Checked make pdf that the output looks reasonable.
projects / thirdparty / gcc.git / log
