Kito Cheng [Tue, 28 Jul 2020 02:29:49 +0000 (10:29 +0800)]
RISC-V: Extend syntax for the multilib-generator
- Support expansion operator (*) in the multilib config string.
- Motivation of this patch is reduce the complexity when we deal multilib with
sub-extension, expand the combinations by hand would be very painful and
error prone, no one deserve to experience this[1] again!
[1] https://github.com/sifive/freedom-tools/blob/
f4d7facafb27d16125768c90ff1790c674e4be7a/Makefile#L348
gcc/ChangeLog:
* config/riscv/multilib-generator: Add TODO, import itertools
and functools.reduce.
Handle expantion operator.
(LONG_EXT_PREFIXES): New.
(arch_canonicalize): Update comment and improve python3
debuggability/compatibility.
(add_underline_prefix): New.
(_expand_combination): Ditto.
(unique): Ditto.
(expand_combination): Ditto.
Jakub Jelinek [Thu, 22 Oct 2020 07:34:28 +0000 (09:34 +0200)]
phiopt: Optimize x ? __builtin_clz (x) : 32 in GIMPLE fallout [PR97503]
> this broke sparc-sun-solaris2.11 bootstrap
>
> /vol/gcc/src/hg/master/local/gcc/tree-ssa-phiopt.c: In function 'bool cond_removal_in_popcount_clz_ctz_pattern(basic_block, basic_block, edge, edge, gimple*, tree, tree)':
> /vol/gcc/src/hg/master/local/gcc/tree-ssa-phiopt.c:1858:27: error: variable 'mode' set but not used [-Werror=unused-but-set-variable]
> 1858 | scalar_int_mode mode = SCALAR_INT_TYPE_MODE (TREE_TYPE (arg));
> | ^~~~
>
>
> and doubtlessly several other targets that use the defaults.h definition of
>
> #define CTZ_DEFINED_VALUE_AT_ZERO(MODE, VALUE) 0
Ugh, seems many of those macros do not evaluate the first argument.
This got broken by the change to direct_internal_fn_supported_p, previously
it used mode also in the optab test.
2020-10-22 Jakub Jelinek <jakub@redhat.com>
* tree-ssa-phiopt.c (cond_removal_in_popcount_clz_ctz_pattern):
For CLZ and CTZ tests, use type temporary instead of mode.
Jakub Jelinek [Thu, 22 Oct 2020 07:33:17 +0000 (09:33 +0200)]
openmp: Add test for OMP_TARGET_OFFLOAD=mandatory for cases where it must not fail
2020-10-22 Jakub Jelinek <jakub@redhat.com>
* testsuite/libgomp.c/target-41.c: New test.
Jakub Jelinek [Thu, 22 Oct 2020 07:32:08 +0000 (09:32 +0200)]
x86: Allow configuring with --with-arch_64=x86-64-v[234]
> + {"x86-64", PROCESSOR_K8, CPU_K8, PTA_X86_64_BASELINE, 0, P_NONE},
> + {"x86-64-v2", PROCESSOR_K8, CPU_GENERIC, PTA_X86_64_V2 | PTA_NO_TUNE,
> + 0, P_NONE},
> + {"x86-64-v3", PROCESSOR_K8, CPU_GENERIC, PTA_X86_64_V3 | PTA_NO_TUNE,
> + 0, P_NONE},
> + {"x86-64-v4", PROCESSOR_K8, CPU_GENERIC, PTA_X86_64_V4 | PTA_NO_TUNE,
> + 0, P_NONE},
> {"eden-x2", PROCESSOR_K8, CPU_K8,
> PTA_64BIT | PTA_MMX | PTA_SSE | PTA_SSE2 | PTA_SSE3 | PTA_FXSR,
> 0, P_NONE},
I have noticed that one can't configure gcc to default to these.
I've also found various other 32-bit or 64-bit -march= arguments for which
it wasn't possible to configure gcc to default to those.
The x86-64-v* the patch only allows in --with-arch_64=, because otherwise
it fails build miserably - as
./xgcc -B ./ -S -march=x86-64-v2 -m32 test.c
cc1: error: ‘x86-64-v2’ architecture level is only defined for the x86-64 psABI
when building 32-bit multilibs. Even if multilibs are disallowed, I think
the compiler still supports -m32 and so --with-arch_64= seems to be the only
option in which we can support that.
2020-10-22 Jakub Jelinek <jakub@redhat.com>
* config.gcc (x86_archs): Add samuel-2, nehemiah, c7 and esther.
(x86_64_archs): Add eden-x2, nano, nano-1000, nano-2000, nano-3000,
nano-x2, eden-x4, nano-x4, x86-64-v2, x86-64-v3 and x86-64-v4.
(i[34567]86-*-* | x86_64-*-*): Only allow x86-64-v* as argument
to --with-arch_64=.
Jakub Jelinek [Thu, 22 Oct 2020 07:31:01 +0000 (09:31 +0200)]
openmp: Change omp_get_initial_device () to match OpenMP 5.1 requirements
> Therefore, I think until omp_get_initial_device () value is changed, we
The following so far untested patch implements that change.
OpenMP 4.5 said for omp_get_initial_device:
The value of the device number is implementation defined. If it is between 0 and one less than
omp_get_num_devices() then it is valid for use with all device constructs and routines; if it is
outside that range, then it is only valid for use with the device memory routines and not in the
device clause.
and OpenMP 5.0 similarly, but OpenMP 5.1 says:
The value of the device number is the value returned by the omp_get_num_devices routine.
As the new value is compatible with what has been required earlier, I think
we can change it already now.
2020-10-22 Jakub Jelinek <jakub@redhat.com>
* icv.c (omp_get_initial_device): Remove including corresponding
ialias.
* icv-device.c (omp_get_initial_device): New function. Return
gomp_get_num_devices (). Add ialias.
* target.c (resolve_device): Don't fail with
OMP_TARGET_OFFLOAD=mandatory if device_id is equal to
gomp_get_num_devices ().
(omp_target_alloc, omp_target_free, omp_target_is_present,
omp_target_memcpy, omp_target_memcpy_rect, omp_target_associate_ptr,
omp_target_disassociate_ptr, omp_pause_resource): Use
gomp_get_num_devices () instead of GOMP_DEVICE_HOST_FALLBACK on the
first use in the functions, in uses dominated by the
gomp_get_num_devices call use num_devices_openmp instead.
* libgomp.texi (omp_get_initial_device): Document.
* config/gcn/icv-device.c (omp_get_initial_device): New function.
Add ialias.
* config/nvptx/icv-device.c (omp_get_initial_device): Likewise.
* testsuite/libgomp.c/target-40.c: New test.
Alexandre Oliva [Thu, 22 Oct 2020 05:32:13 +0000 (02:32 -0300)]
Use Aux_Long_Float for all real types on LynxOS
Its libc does not offer *f or *l elementary functions, so rely on the
C double ones only.
for gcc/ada/ChangeLog
* Makefile.rtl (LIBGNAT_TARGET_PAIRS) <lynxos178>: Rely on
Aux_Long_Float for all real types.
Alexandre Oliva [Thu, 22 Oct 2020 05:32:10 +0000 (02:32 -0300)]
vxworks float EFs not precise enough -> use long float
Some acats-4 tests that check the precision of Float elementary
functions fail with vxworks 7.2's implementations of single-precision
math functions.
This patch arranges for us to bypass the single-precision functions,
and use the Aux_Long_Float implementation, based on the double-typed
calls from the C library, for Float and Short_Float.
for gcc/ada/ChangeLog
* Makefile.rtl (LIBGNAT_TARGET_PAIRS): Use Long Float-based
variant of Aux_Short_Float and Aux_Float on vxworks targets.
* libgnat/a-nashfl__wraplf.ads: New.
* libgnat/a-nuaufl__wraplf.ads: New.
Rainer Orth [Thu, 22 Oct 2020 05:32:06 +0000 (02:32 -0300)]
Use Aux_Long_Long_Float wraplf for sparc*-sun-solaris too
Like aarch64-* and ppc*-linux-gnu, sparc*-sun-solaris has
Long_Long_Float mapped to double rather than long double, so the
intrinsics in the default version of a-nallfl.ads have mismatching
types. Adopt the wraplf workaround for it as well.
for gcc/ada/ChangeLog
* Makefile.rtl (LIBGNAT_TARGET_PAIRS) <sparc*-sun-solaris>:
Use wraplf version of a-nallfl.
Alexandre Oliva [Thu, 22 Oct 2020 05:32:04 +0000 (02:32 -0300)]
aarch64-* and ppc*-linux-gnu long long float/long double mismatch
Some platforms have failed to build because long long float is mapped
to double rather than long double, and then the attempts to import
intrinsics for long double in Aux_Long_Long_Float raise warnings
turned into errors.
This patch is a work around for the mismatch, arranging for
Aux_Long_Long_Float to map to Aux_Long_Float.
for gcc/ada/ChangeLog
* Makefile.rtl (LIBGNAT_TARGET_PAIRS): Use
a-nallfl__wraplf.ads on aarch64-* and ppc*-linux-gnu targets.
* libgnat/a-nallfl__wraplf.ads: New.
Nikhil Benesch [Sun, 18 Oct 2020 19:28:54 +0000 (19:28 +0000)]
syscall: only compile ptrace varargs shim on Linux
Only compile the __go_ptrace varargs shim on Linux to avoid compilation
failures on some other platforms. The C ptrace function is not entirely
portable (e.g., NetBSD has `int data` instead of `void* data`), and so
far Linux is the only platform that needs the varargs shim.
Additionally, make the types in the ptrace and raw_ptrace function
declarations match. This makes it more clear that the only difference
between the two is that calls via the former are allowed to block while
calls via the latter are not.
Reviewed-on: https://go-review.googlesource.com/c/gofrontend/+/263517
Jan Hubicka [Thu, 22 Oct 2020 04:34:48 +0000 (06:34 +0200)]
Free stale summaries in ipa-pure-const
* ipa-pure-const.c (funct_state_summary_t::insert): Free stale
summaries.
Jan Hubicka [Thu, 22 Oct 2020 04:33:34 +0000 (06:33 +0200)]
Move nested function info out of cgraph_node
this patch moves nested function information out of symbol table (to a summary).
This saves memory (especially at WPA time) and also makes nested function
support more contained.
gcc/ChangeLog:
2020-10-22 Jan Hubicka <hubicka@ucw.cz>
* cgraph.c: Include tree-nested.h
(cgraph_node::create): Call maybe_record_nested_function.
(cgraph_node::remove): Do not remove function from nested function
infos.
(cgraph_node::dump): Update.
(cgraph_node::unnest): Move to tree-nested.c
(cgraph_node::verify_node): Update.
(cgraph_c_finalize): Call nested_function_info::release.
* cgraph.h (struct symtab_node): Remove nested function info.
* cgraphclones.c (cgraph_node::create_clone): Do not clone nested
function info.
* cgraphunit.c (cgraph_node::analyze): Update.
(cgraph_node::expand): Do not worry about nested functions; they are
lowered.
(symbol_table::finalize_compilation_unit): Call
nested_function_info::release.
* gimplify.c: Include tree-nested.h
(unshare_body): Update.
(unvisit_body): Update.
* omp-offload.c (omp_discover_implicit_declare_target): Update.
* tree-nested.c: Include alloc-pool.h, tree-nested.h, symbol-summary.h
(nested_function_sum): New static variable.
(nested_function_info::get): New member function.
(nested_function_info::get_create): New member function.
(unnest_function): New function.
(nested_function_info::~nested_function_info): New member function.
(nested_function_info::release): New function.
(maybe_record_nested_function): New function.
(lookup_element_for_decl): Update.
(check_for_nested_with_variably_modified): Update.
(create_nesting_tree): Update.
(unnest_nesting_tree_1): Update.
(gimplify_all_functions): Update.
(lower_nested_functions): Update.
* tree-nested.h (class nested_function_info): New class.
(maybe_record_nested_function): Declare.
(unnest_function): Declare.
(first_nested_function): New inline function.
(next_nested_function): New inline function.
(nested_function_origin): New inline function.
gcc/ada/ChangeLog:
2020-10-22 Jan Hubicka <hubicka@ucw.cz>
* gcc-interface/trans.c: Include tree-nested.h
(walk_nesting_tree): Update for new nested function info.
gcc/c-family/ChangeLog:
2020-10-22 Jan Hubicka <hubicka@ucw.cz>
* c-gimplify.c: Include tree-nested.h
(c_genericize): Update for new nested function info.
gcc/d/ChangeLog:
2020-10-22 Jan Hubicka <hubicka@ucw.cz>
* decl.cc: Include tree-nested.h
(get_symbol_decl): Update for new nested function info.
liuhongt [Tue, 13 Oct 2020 07:35:29 +0000 (15:35 +0800)]
Simplify vec_select of a subreg of X to just a vec_select of X.
gcc/ChangeLog
PR rtl-optimization/97249
* simplify-rtx.c (simplify_binary_operation_1): Simplify
vec_select of a subreg of X to a vec_select of X.
gcc/testsuite/ChangeLog
* gcc.target/i386/pr97249-1.c: New test.
liuhongt [Sat, 26 Sep 2020 07:34:23 +0000 (15:34 +0800)]
Refactor implementation of *_bcst{_1,_2,_3} patterns.
Add new predicate bcst_mem_operand and corresponding constraint "Br"
to merge "$(pattern)_bcst{_1,_2,_3}" into "$(pattern)", also delete
those separate "*_bcst{_1,_2,_3}" patterns.
gcc/ChangeLog:
PR target/87767
* config/i386/constraints.md ("Br"): New special memory
constraint.
* config/i386/i386-expand.c (ix86_binary_operator_ok): Both
source operand cannot be in memory or bcst_memory_operand.
* config/i386/i386.c (ix86_print_operand): Print bcst_mem_operand.
* config/i386/i386.h (VALID_BCST_MODE_P): New.
* config/i386/predicates.md (bcst_mem_operand): New predicate
for AVX512 embedding broadcast memory operand.
(bcst_vector_operand): New predicate, vector_operand or
bcst_mem_operand.
* config/i386/sse.md
(*<plusminus_insn><mode>3<mask_name><round_name>): Extend
predicate and constraints to handle bcst_mem_operand.
(*mul<mode>3<mask_name><round_name>): Ditto.
(<sse>_div<mode>3<mask_name><round_name>): Ditto.
(<sd_mask_codefor>fma_fmadd_<mode><sd_maskz_name><round_name>):
Ditto.
(<sd_mask_codefor>fma_fmsub_<mode><sd_maskz_name><round_name>):
Ditto.
(<sd_mask_codefor>fma_fnmadd_<mode><sd_maskz_name><round_name>):
Ditto.
(<sd_mask_codefor>fma_fnmsub_<mode><sd_maskz_name><round_name>):
Ditto.
(*<plusminus_insn><mode>3): Ditto.
(avx512dq_mul<mode>3<mask_name>): Ditto.
(*<sse4_1_avx2>_mul<mode>3<mask_name>): Ditto.
(*andnot<mode>3): Ditto.
(<mask_codefor><code><mode>3<mask_name>): Ditto.
(*sub<mode>3<mask_name>_bcst): Removed.
(*add<mode>3<mask_name>_bcst): Ditto.
(*mul<mode>3<mask_name>_bcst): Ditto.
(*<avx512>_div<mode>3<mask_name>_bcst): Ditto.
(*<sd_mask_codefor>fma_fmadd_<mode><sd_maskz_name>_bcst_1):
Ditto.
(*<sd_mask_codefor>fma_fmadd_<mode><sd_maskz_name>_bcst_2):
Ditto.
(*<sd_mask_codefor>fma_fmadd_<mode><sd_maskz_name>_bcst_3):
Ditto.
(*<sd_mask_codefor>fma_fmsub_<mode><sd_maskz_name>_bcst_1):
Ditto.
(*<sd_mask_codefor>fma_fmsub_<mode><sd_maskz_name>_bcst_2):
Ditto.
(*<sd_mask_codefor>fma_fmsub_<mode><sd_maskz_name>_bcst_3):
Ditto.
(*<sd_mask_codefor>fma_fnmadd_<mode><sd_maskz_name>_bcst_1):
Ditto.
(*<sd_mask_codefor>fma_fnmadd_<mode><sd_maskz_name>_bcst_2):
Ditto.
(*<sd_mask_codefor>fma_fnmadd_<mode><sd_maskz_name>_bcst_3):
Ditto.
(*<sd_mask_codefor>fma_fnmsub_<mode><sd_maskz_name>_bcst_1):
Ditto.
(*<sd_mask_codefor>fma_fnmsub_<mode><sd_maskz_name>_bcst_2):
Ditto.
(*<sd_mask_codefor>fma_fnmsub_<mode><sd_maskz_name>_bcst_3):
Ditto.
(*sub<mode>3_bcst): Ditto.
(*add<mode>3_bcst): Ditto.
(*avx512dq_mul<mode>3<mask_name>_bcst): Ditto.
(*avx512f_mul<mode>3<mask_name>_bcst): Ditto.
(*andnot<mode>3_bcst): Ditto.
(*<code><mode>3_bcst): Ditto.
* config/i386/subst.md (bcst_round_constraint): New subst
attribute.
(bcst_round_nimm_predicate): Ditto.
(bcst_mask_prefix3): Ditto.
(bcst_mask_prefix4): Ditto.
liuhongt [Sat, 26 Sep 2020 07:08:32 +0000 (15:08 +0800)]
Extend special_memory_constraint.
For operand with special_memory_constraint, there could be a wrapper
for memory_operand. Extract mem for operand for conditional judgement
like MEM_P, also for record_address_regs.
gcc/ChangeLog:
PR target/87767
* ira-costs.c (record_operand_costs): Extract memory operand
from recog_data.operand[i] for record_address_regs.
(record_reg_classes): Extract memory operand from OP for
conditional judgement MEM_P.
* ira.c (ira_setup_alts): Ditto.
* lra-constraints.c (extract_mem_from_operand): New function.
(satisfies_memory_constraint_p): Extract memory operand from
OP for decompose_mem_address, return false when there's no
memory operand inside OP.
(process_alt_operands): Remove MEM_P (op) since it would be
judged in satisfies_memory_constraint_p.
* recog.c (asm_operand_ok): Extract memory operand from OP for
judgement of memory_operand (OP, VOIDmode).
(constrain_operands): Don't unwrapper unary operator when
there's memory operand inside.
* rtl.h (extract_mem_from_operand): New decl.
Dennis Zhang [Thu, 22 Oct 2020 00:27:31 +0000 (01:27 +0100)]
arm: Auto-vectorization for MVE: vmin/vmax
This patch enables MVE vmin/vmax instructions for auto-vectorization.
MVE target is included in expander smin<mode>3, umin<mode>3, smax<mode>3
and umax<mode>3 for vectorization. Related insns for vmin/vmax in mve.md
are modified to use smin, umin, smax and umax expressions instead of
unspec to support the expanders.
gcc/ChangeLog:
2020-10-22 Dennis Zhang <dennis.zhang@arm.com>
* config/arm/mve.md (mve_vmaxq_<supf><mode>): Replace with ...
(mve_vmaxq_s<mode>, mve_vmaxq_u<mode>): ... these new insns to
use smax/umax instead of VMAXQ.
(mve_vminq_<supf><mode>): Replace with ...
(mve_vminq_s<mode>, mve_vminq_u<mode>): ... these new insns to
use smin/umin instead of VMINQ.
(mve_vmaxnmq_f<mode>): Use smax instead of VMAXNMQ_F.
(mve_vminnmq_f<mode>): Use smin instead of VMINNMQ_F.
* config/arm/vec-common.md (smin<mode>3): Use the new mode macros
ARM_HAVE_<MODE>_ARITH.
(umin<mode>3, smax<mode>3, umax<mode>3): Likewise.
gcc/testsuite/ChangeLog:
* gcc.target/arm/simd/mve-vminmax_1.c: New test.
Andrew MacLeod [Thu, 22 Oct 2020 00:11:16 +0000 (20:11 -0400)]
Handle a_2= &b properly in range calculations.
when processing assignments, we were using the type of b instead of type
of &b when computing a range. This was usually filtered out by FRE.
turning it off exposed it.
gcc/
PR tree-optimization/97520
* gimple-range.cc (range_of_non_trivial_assignment): Handle x = &a
by returning a non-zero range.
gcc/testsuite/
* gcc.dg/pr97520.c: New.
Dennis Zhang [Thu, 22 Oct 2020 00:09:33 +0000 (01:09 +0100)]
arm: Auto-vectorization for MVE: vmul
This patch enables MVE vmul instructions for auto-vectorization.
It includes MVE in expander mul<mode>3 to enable vectorization for MVE.
Related MVE vmul insns are modified to support the expander by using
expression 'mult' instead of unspec.
The mul<mode>3 for vectorization in vec-common.md uses mode iterator
VDQWH instead of VALLW to cover all supported modes.
The macros ARM_HAVE_NEON_<MODE>_ARITH are used to select supported
modes for different targets.
The redundant mul<mode>3 in neon.md is removed.
gcc/ChangeLog:
2020-10-22 Dennis Zhang <dennis.zhang@arm.com>
* config/arm/mve.md (mve_vmulq<mode>): New entry for vmul instruction
using expression 'mult'.
(mve_vmulq_f<mode>): Use mult instead of VMULQ_F.
* config/arm/neon.md (mul<mode>3): Removed.
* config/arm/vec-common.md (mul<mode>3): Use the new mode macros
ARM_HAVE_<MODE>_ARITH. Use mode iterator VDQWH instead of VALLW.
gcc/testsuite/ChangeLog:
* gcc.target/arm/simd/mve-vmul_1.c: New test.
Andrew MacLeod [Wed, 21 Oct 2020 23:55:28 +0000 (19:55 -0400)]
Check for undefined before not returning a constant value
Don't return UNDEFINED for a range in an unreachable block if the global
value evaluates to a constant. Return that constant instead.
PR tree-optimization/97515
* value-query.cc (range_query::value_of_expr): If the result is
UNDEFINED, check to see if the global value is a constant.
(range_query::value_on_edge): Ditto.
Nikhil Benesch [Sat, 10 Oct 2020 00:06:36 +0000 (00:06 +0000)]
syscall: import upstream code for BSD sockets and sysctls
Import some missing upstream code for BSD sockets and sysctls and
adapt it for gccgo.
Updates golang/go#38538.
Reviewed-on: https://go-review.googlesource.com/c/gofrontend/+/261137
Jan Hubicka [Wed, 21 Oct 2020 23:42:11 +0000 (01:42 +0200)]
Strenghten bound for bulitin_constant_p hint.
this patch makes builtin_constant_p hint to combine with other loop hints
we already support.
gcc/ChangeLog:
2020-10-22 Jan Hubicka <hubicka@ucw.cz>
PR ipa/97445
* ipa-inline.c (inline_insns_single): Add hint2 parameter.
(inline_insns_auto): Add hint2 parameter.
(can_inline_edge_by_limits_p): Update.
(want_inline_small_function_p): Update.
(wrapper_heuristics_may_apply): Update.
David Edelsohn [Wed, 21 Oct 2020 16:24:00 +0000 (12:24 -0400)]
testsuite: Enable p9-lxvx-stxvx testcases and fold-vec-extract
This patch enables the p9-lxvx-stxvx testcases with appropriate
requirements. It also adjusts the expected matching instructions
in the fold-vec-extract testcases for BE.
gcc/testsuite/ChangeLog:
* gcc.target/powerpc/fold-vec-extract-float.p9.c: rldicl and
subfic only for target LE.
* gcc.target/powerpc/fold-vec-extract-longlong.p9.c: xori only
for target LE. Adjust mfvsrd and add mfvsrld for BE.
* gcc.target/powerpc/fold-vec-extract-short.p9.c: vextuhrx for LE.
vextuhlx for BE.
* gcc.target/powerpc/p9-lxvx-stxvx-1.c: Remove target.
* gcc.target/powerpc/p9-lxvx-stxvx-2.c: Remove target.
* gcc.target/powerpc/p9-lxvx-stxvx-3.c: Remove target. Require
float128
Jonathan Wakely [Wed, 21 Oct 2020 20:13:41 +0000 (21:13 +0100)]
libstdc++: Simplify std::shared_ptr construction from std::weak_ptr
The _M_add_ref_lock() and _M_add_ref_lock_nothrow() members of
_Sp_counted_base are very similar, except that the former throws an
exception when the use count is zero and the latter returns false. The
former (and its callers) can be implemented in terms of the latter.
This results in a small reduction in code size, because throwing an
exception now only happens in one place.
libstdc++-v3/ChangeLog:
* include/bits/shared_ptr.h (shared_ptr(const weak_ptr&, nothrow_t)):
Add noexcept.
* include/bits/shared_ptr_base.h (_Sp_counted_base::_M_add_ref_lock):
Remove specializations and just call _M_add_ref_lock_nothrow.
(__shared_count, __shared_ptr): Use nullptr for null pointer
constants.
(__shared_count(const __weak_count&)): Use _M_add_ref_lock_nothrow
instead of _M_add_ref_lock.
(__shared_count(const __weak_count&, nothrow_t)): Add noexcept.
(__shared_ptr::operator bool()): Add noexcept.
(__shared_ptr(const __weak_ptr&, nothrow_t)): Add noexcept.
Richard Biener [Wed, 21 Oct 2020 19:28:45 +0000 (14:28 -0500)]
rs6000: MMA type causes an ICE in ranger pass due to incompatible types
PR97360 shows a problem in how we create our PXI and POI modes that cause
an ICE in the ranger pass. The problem seems to be that the extra call
to build_distinct_type_copy() also creates new TYPE_{MIN,MAX}_VALUEs that
are not compatible/the same as the base type itself. The simple "fix" is
to actually remove the unneeded build_distinct_type_copy(), since according
to richi, the types returned from make_unsigned_type() are already distinct.
gcc/
2020-10-21 Richard Biener <rguenther@suse.de>
PR target/97360
* config/rs6000/rs6000-call.c (rs6000_init_builtins): Remove call to
build_distinct_type_copy().
gcc/testsuite/
2020-10-21 Martin Liska <mliska@suse.cz>
PR target/97360
* gcc.target/powerpc/pr97360.c: New test.
Co-authored-by: Andrew MacLeod <amacleod@redhat.com>
Co-authored-by: Martin Liska <mliska@suse.cz>
Jan Hubicka [Wed, 21 Oct 2020 18:00:22 +0000 (20:00 +0200)]
Inline functions with builtin_constant_p more agressively.
This patch implements heuristics that increases inline limits (by the hints
mechanism) for inline functions that use builtin_constant_p on parameter. Those
are very likely intended to be always inlined and simplify after inlining.
The PR is about a function that we used to inline with
--param inline-insns-single=200 but with new default of 70 for -O2 we no longer
do so. Hints are currently configured to bump the bound up twice, so we
get limit of 140 that is still not enough to inline the particular testcase
but it should help in general. I can implement a stronger bump if that seems
useful (maybe it is). The example is bit operation written as a decision chain
with 64 conditions.
This blows up the limit on number of conditions we track per funtion (which is
30) and thus the size/time estimates are not working that well.
gcc/ChangeLog:
PR ipa/97445
* ipa-fnsummary.c (ipa_dump_hints): Add INLINE_HINT_builtin_constant_p.
(ipa_fn_summary::~ipa_fn_summary): Free builtin_constant_p_parms.
(ipa_fn_summary_t::duplicate): Duplicate builtin_constant_p_parms.
(ipa_dump_fn_summary): Dump builtin_constant_p_parms.
(add_builtin_constant_p_parm): New function
(set_cond_stmt_execution_predicate): Update builtin_constant_p_parms.
(ipa_call_context::estimate_size_and_time): Set
INLINE_HINT_builtin_constant_p..
(ipa_merge_fn_summary_after_inlining): Merge builtin_constant_p_parms.
(inline_read_section): Read builtin_constant_p_parms.
(ipa_fn_summary_write): Write builtin_constant_p_parms.
* ipa-fnsummary.h (enum ipa_hints_vals): Add
INLINE_HINT_builtin_constant_p.
* ipa-inline.c (want_inline_small_function_p): Use
INLINE_HINT_builtin_constant_p.
(edge_badness): Use INLINE_HINT_builtin_constant_p.
gcc/testsuite/ChangeLog:
PR ipa/97445
* gcc.dg/ipa/inlinehint-5.c: New test.
Douglas Rupp [Mon, 2 Dec 2019 17:17:18 +0000 (09:17 -0800)]
Remove interfering default #undefs from vx-common.h
undef'ing LIB_SPEC and especially LINK_SPEC here is unneccesary and
inteferes with the definition of LINK_SPEC for the bi-arch'd ppc*-vx7r2
target which uses the linux64 LINK_SPEC
2020-10-21 Douglas Rupp <rupp@adacore.com>
gcc/
* config/vx-common.h (LINK_SPEC, LIB_SPEC): Remove #undef.
Douglas Rupp [Tue, 18 Feb 2020 09:54:58 +0000 (09:54 +0000)]
Introduce vxworks7r2 support for ppc and ppc64
This change introduces support for the most recent versions
of VxWorks on PowerPC targets, for both 32 and 64 bit thanks
to a bi-arch setup.
The system compilers are essentially configured as Linux
toolchains with only a few specificities and we replicate
that model here.
The most visible specificities are the use of secureplt by
default, the pre-definition of some macros that the system
headers still rely on (_VX_CPU and _VX_CPU_FAMILY, for example),
and of course some variations related to the so VxWorks typical
kernel vs RTP mode distinction.
In addition to the introduction of config.gcc and libgcc
configuration chunks, much inspired by the linux ones, the
change
- Reworks rs6000/vxworks.h file to feature bits common to the
Vx6 and Vx7 port then a separate section for each, where the
Vx7 part is very short as we rely on the Linux definitions
for most things.
- Adjusts the CPU macro predefinitions in CPP_SPEC to resort
to "_VX_CPU" instead of "CPU" for Vx7, to better match the
more recent system headers expectations,
- Adds a cpu definition case for e6500.
- Changes to the use SUB3TARGET_OVERRIDE_OPTIONS instead of
SUBSUBTARGET_OVERRIDE_OPTIONS for specifics, so we don't
override the Linux's version of the latter for vx7.
2020-10-20 Douglas Rupp <rupp@adacore.com>
gcc/
* config.gcc (powerpc*-wrs-vxworks7r*): New case.
* config/rs6000/vxworks.h: Rework to handle VxWorks7.
Refactor as common bits + vx6 vs vx7 ones. For the
latter, rely essentially on the Linux configuration
and adjust CPU to _VX_CPU in CPP_SPEC. Add a case
for e6500. Use SUB3TARGET_OVERRIDE_OPTIONS for specifics
to preserve the Linux SUBSUBTARGET_OVERRIDE_OPTIONS
for vx7.
libgcc/
* config.host (powerpc*-wrs-vxworks7*): New case.
* configure.ac: Handle powerpc*-*-vxworks7* as
powerpc*-*-linux* for ppc-fp_type.
* configure: Regenerate.
Co-authored-by: Olivier Hainque <hainque@adacore.com>
David Edelsohn [Wed, 21 Oct 2020 00:52:06 +0000 (20:52 -0400)]
testsuite: enable and fix swaps-p8 testcases
This patch enables swaps-p8 and a few other testcases on non-powerpc64le
systems. It also cleans up the target requirements for various testcases.
gcc/testsuite/ChangeLog:
* gcc.target/powerpc/p9-extract-1.c: Require lp64.
* gcc.target/powerpc/p9-extract-2.c: Require lp64.
* gcc.target/powerpc/p9-extract-3.c: Require lp64.
* gcc.target/powerpc/p9-permute.c: Remove target.
* gcc.target/powerpc/pr63335.c: Remove target.
* gcc.target/powerpc/pr87507.c: Remove target.
* gcc.target/powerpc/swaps-p8-1.c: Require LE and P8.
* gcc.target/powerpc/swaps-p8-10.c: Remove target. Require P8.
* gcc.target/powerpc/swaps-p8-11.c: Remove target. Require P8.
* gcc.target/powerpc/swaps-p8-12.c: Require LE and P8.
* gcc.target/powerpc/swaps-p8-13.c: Remove target. Require P8.
* gcc.target/powerpc/swaps-p8-14.c: Remove target. Require P8.
* gcc.target/powerpc/swaps-p8-15.c: Remove target. Require P8.
* gcc.target/powerpc/swaps-p8-16.c: Remove target. Require P8.
* gcc.target/powerpc/swaps-p8-17.c: Require P8.
* gcc.target/powerpc/swaps-p8-18.c: Remove target. Require P8.
* gcc.target/powerpc/swaps-p8-19.c: Remove target. Require P8.
* gcc.target/powerpc/swaps-p8-2.c: Require LE and P8.
* gcc.target/powerpc/swaps-p8-20.c: Remove target.
* gcc.target/powerpc/swaps-p8-21.c: Remove target. Require Altivec.
* gcc.target/powerpc/swaps-p8-22.c: Remove target. Require P8.
* gcc.target/powerpc/swaps-p8-23.c: Require LE and P8.
* gcc.target/powerpc/swaps-p8-24.c: Remove target. Require P8.
* gcc.target/powerpc/swaps-p8-25.c: Require LE and P8.
* gcc.target/powerpc/swaps-p8-26.c: Remove target. Require P8.
* gcc.target/powerpc/swaps-p8-27.c: Remove target. Require P8.
* gcc.target/powerpc/swaps-p8-28.c: Remove target.
* gcc.target/powerpc/swaps-p8-29.c: Remove target.
* gcc.target/powerpc/swaps-p8-3.c: Require LE and P8.
* gcc.target/powerpc/swaps-p8-30.c: Remove target.
* gcc.target/powerpc/swaps-p8-31.c: Remove target.
* gcc.target/powerpc/swaps-p8-32.c: Remove target.
* gcc.target/powerpc/swaps-p8-33.c: Remove target.
* gcc.target/powerpc/swaps-p8-34.c: Remove target.
* gcc.target/powerpc/swaps-p8-35.c: Remove target.
* gcc.target/powerpc/swaps-p8-36.c: Remove target.
* gcc.target/powerpc/swaps-p8-37.c: Remove target.
* gcc.target/powerpc/swaps-p8-38.c: Remove target.
* gcc.target/powerpc/swaps-p8-39.c: Remove target.
* gcc.target/powerpc/swaps-p8-4.c: Require LE and P8.
* gcc.target/powerpc/swaps-p8-40.c: Remove target.
* gcc.target/powerpc/swaps-p8-41.c: Remove target.
* gcc.target/powerpc/swaps-p8-42.c: Remove target.
* gcc.target/powerpc/swaps-p8-43.c: Remove target.
* gcc.target/powerpc/swaps-p8-44.c: Remove target.
* gcc.target/powerpc/swaps-p8-45.c: Remove target.
* gcc.target/powerpc/swaps-p8-46.c: Require LE.
* gcc.target/powerpc/swaps-p8-5.c: Require LE and P8.
* gcc.target/powerpc/swaps-p8-6.c: Remove target. Require P8.
* gcc.target/powerpc/swaps-p8-7.c: Remove target. Require P8.
* gcc.target/powerpc/swaps-p8-8.c: Remove target. Require P8.
* gcc.target/powerpc/swaps-p8-9.c: Remove target. Require P8.
* gcc.target/powerpc/vec-cmp.c: Require LP64.
* gcc.target/powerpc/vec-cmpne.c: Remove target.
* gcc.target/powerpc/vec-mul.c: Remove target.
* gcc.target/powerpc/vec-set-char.c: Require LP64.
* gcc.target/powerpc/vec-set-int.c: Require LP64.
* gcc.target/powerpc/vec-set-short.c: Require LP64.
* gcc.target/powerpc/vec-xxpermdi.c: Remove target. Require VSX.
* gcc.target/powerpc/vsxcopy.c: Remove target. Require VSX.
Jonathan Wakely [Wed, 21 Oct 2020 13:40:54 +0000 (14:40 +0100)]
libstdc++: Make structured bindings always work for subranges [PR 97512]
The definition of ranges::subrange was moved to the new
<bits/ranges_util.h> header so that it could be used in <algorithm>
without including the whole of <ranges>. However, the tuple-like support
that enables subrange to be used with structured bindings was left in
<ranges>. This is arguably conforming (to use a subrange you should
include <ranges>) but it's inconvenient and probably confusing.
This change makes the tuple-like support available whenever subrange
itself is available.
libstdc++-v3/ChangeLog:
PR libstdc++/97512
* include/bits/ranges_util.h (tuple_size<subrange>)
(tuple_element<I, cv subrange>): Move here from ...
* include/std/ranges: ... here.
* testsuite/std/ranges/subrange/97512.cc: New test.
Thomas Rodgers [Wed, 21 Oct 2020 13:11:28 +0000 (06:11 -0700)]
libstdc++: Rebase include/pstl to current upstream
From llvm-project/pstl @
0b2e0e80d96
libstdc++-v3/ChangeLog:
* include/pstl/algorithm_impl.h: Update file.
* include/pstl/execution_impl.h: Likewise.
* include/pstl/glue_algorithm_impl.h: Likewise.
* include/pstl/glue_memory_impl.h: Likewise.
* include/pstl/glue_numeric_impl.h: Likewise.
* include/pstl/memory_impl.h: Likewise.
* include/pstl/numeric_impl.h: Likewise.
* include/pstl/parallel_backend.h: Likewise.
* include/pstl/parallel_backend_serial.h: Likewise.
* include/pstl/parallel_backend_tbb.h: Likewise.
* include/pstl/parallel_backend_utils.h: Likewise.
* include/pstl/pstl_config.h: Likewise.
* include/pstl/unseq_backend_simd.h: Likewise.
Tobias Burnus [Wed, 21 Oct 2020 12:38:44 +0000 (14:38 +0200)]
Fortran: class.c - update vtable comment
gcc/fortran/
PR fortran/45516
* class.c: Add _deallocate to the vtable documentation
comment.
Richard Biener [Wed, 21 Oct 2020 08:54:54 +0000 (10:54 +0200)]
tree-optimization/97500 - avoid SLP backedges for inductions
Inductions are not vectorized as cycle but materialized from SCEV data.
Filling in backedge SLP nodes confuses this process.
2020-10-21 Richard Biener <rguenther@suse.de>
PR tree-optimization/97500
* tree-vect-slp.c (vect_analyze_slp_backedges): Do not
fill backedges for inductions.
* gfortran.dg/pr97500.f90: New testcase.
liuhongt [Wed, 21 Oct 2020 05:05:16 +0000 (13:05 +0800)]
Simplify trivial VEC_COND_EXPR in expander.
gcc/ChangeLog:
PR target/97506
* config/i386/i386-expand.c (ix86_expand_sse_movcc): Move
op_true to dest directly when op_true equals op_false.
gcc/testsuite/ChangeLog:
PR target/97506
* gcc.target/i386/pr97506.c: New test.
Jakub Jelinek [Wed, 21 Oct 2020 08:51:33 +0000 (10:51 +0200)]
phiopt: Optimize x ? __builtin_clz (x) : 32 in GIMPLE [PR97503]
While we have at the RTL level noce_try_ifelse_collapse combined with
simplify_cond_clz_ctz, that optimization doesn't always trigger because
e.g. on powerpc there is an define_insn to compare a reg against zero and
copy that register to another one and so we end up with a different pseudo
in the simplify_cond_clz_ctz test and punt.
For targets that define C?Z_DEFINED_VALUE_AT_ZERO to 2 for certain modes,
we can optimize it already in phiopt though, just need to ensure that
we transform the __builtin_c?z* calls into .C?Z ifns because my recent
VRP changes codified that the builtin calls are always undefined at zero,
while ifns honor C?Z_DEFINED_VALUE_AT_ZERO equal to 2.
And, in phiopt we already have popcount handling that does pretty much the
same thing, except for always using a zero value rather than the one set
by C?Z_DEFINED_VALUE_AT_ZERO.
So, this patch extends that function to handle not just popcount, but also
clz and ctz.
2020-10-21 Jakub Jelinek <jakub@redhat.com>
PR tree-optimization/97503
* tree-ssa-phiopt.c: Include internal-fn.h.
(cond_removal_in_popcount_pattern): Rename to ...
(cond_removal_in_popcount_clz_ctz_pattern): ... this. Handle not just
popcount, but also clz and ctz if it has C?Z_DEFINED_VALUE_AT_ZERO 2.
* gcc.dg/tree-ssa/pr97503.c: New test.
Richard Biener [Wed, 21 Oct 2020 07:40:09 +0000 (09:40 +0200)]
Simplify CFG copying tables
This simplifies the maps between original and new basic blocks and
loops as used for CFG copying. Instead of using a pointer hash
table to allocated mapping entries use a hash_map with int_hash,
removing the indirection and code duplication. We can use -1 and
-2 as empty/deleted values as those are not valid basic-block
indices or loop numbers.
2020-10-21 Richard Biener <rguenther@suse.de>
* cfg.c (htab_bb_copy_original_entry): Remove.
(bb_copy_hasher): Likewise.
(bb_original, bb_copy, loop_copy): Use
hash_map<int_hash<int, -1, -2>, int>.
(original_copy_bb_pool): Remove.
(initialize_original_copy_tables): Adjust.
(reset_original_copy_tables): Likewise.
(free_original_copy_tables): Likewise.
(original_copy_tables_initialized_p): Likewise.
(copy_original_table_clear): Simplify.
(copy_original_table_set): Likewise.
(get_bb_original): Likewise.
(get_bb_copy): Likewise.
(get_loop_copy): Likewise.
Richard Biener [Wed, 21 Oct 2020 07:45:50 +0000 (09:45 +0200)]
Separate new_edges compute in copy_bbs
This separates out a loop finding new_edges from edges in copy_bbs,
making its complexity cheaper overall from total number of succs in
copied bbs times num_edges to num_edges times the complexity of
find_edge.
2020-10-21 Richard Biener <rguenther@suse.de>
* cfghooks.c (copy_bbs): Split out loop computing new_edges.
Aldy Hernandez [Wed, 21 Oct 2020 07:33:06 +0000 (09:33 +0200)]
Adjust overflow for invariants in bounds_of_var_in_loop.
Invariants returned from SCEV can have TREE_OVERFLOW set. Clear the
overflow as we do with the rest of the values returned from this
function.
gcc/ChangeLog:
* gimple-range.cc (gimple_ranger::range_of_ssa_name_with_loop_info):
Remove TREE_OVERFLOW special case.
* vr-values.c (bounds_of_var_in_loop): Adjust overflow for
invariants.
Jakub Jelinek [Wed, 21 Oct 2020 08:21:52 +0000 (10:21 +0200)]
libgomp: Hopefully avoid false positive warnings in env.c on solaris
> the patch also breaks bootstrap on both i386-pc-solaris2.11 and
> sparc-sun-solaris2.11:
>
> /vol/gcc/src/hg/master/local/libgomp/env.c: In function 'initialize_env':
> /vol/gcc/src/hg/master/local/libgomp/env.c:414:16: error: 'new_offload' may be used uninitialized in this function [-Werror=maybe-uninitialized]
> 414 | *offload = new_offload;
> | ~~~~~~~~~^~~~~~~~~~~~~
> /vol/gcc/src/hg/master/local/libgomp/env.c:384:30: note: 'new_offload' was declared here
> 384 | enum gomp_target_offload_t new_offload;
> | ^~~~~~~~~~~
I can't reproduce that, but I fail to see why we need two separate
variables, one with actual value and one tracking if the value is valid.
So, I'm going with:
2020-10-21 Jakub Jelinek <jakub@redhat.com>
* env.c (parse_target_offload): Change new_offload var type to int,
preinitialize to -1, remove found var and test new_offload != -1
instead of found.
Aldy Hernandez [Tue, 20 Oct 2020 16:39:46 +0000 (18:39 +0200)]
Move simplify_cond_using_ranges_2 to tree-vrp.c
This was slated to be moved last year, as its only use is in tree-vrp.c
There are no functional changes. It's just a move and a rename.
gcc/ChangeLog:
* vr-values.h: Remove simplify_cond_using_ranges_2.
(range_fits_type_p): New.
* vr-values.c (range_fits_type_p): Remove static qualifier.
(vrp_simplify_cond_using_ranges): Move...
* tree-vrp.c (vrp_simplify_cond_using_ranges): ...to here.
Martin Liska [Wed, 21 Oct 2020 07:28:56 +0000 (09:28 +0200)]
libsanitizer: add test-case
gcc/testsuite/ChangeLog:
PR sanitizer/97414
* g++.dg/asan/pr97414.C: New test.
Martin Liska [Fri, 16 Oct 2020 13:08:52 +0000 (15:08 +0200)]
ASAN: Support detect_invalid_pointer_pairs=1 with detect_stack_use_after_return=1
Do not crash when AsanThread::GetStackVariableShadowStart does not find
a variable for a pointer on a shadow stack.
Cherry-pick from
ad2be02a833e56f7fe280797280b219eb3312621.
Differential Revision: https://reviews.llvm.org/D89552
Eric Botcazou [Sat, 27 Jun 2020 10:43:32 +0000 (12:43 +0200)]
[Ada] Support of the Ada.Text_IO hierarchy for 128-bit types
gcc/ada/
* Makefile.rtl (GNATRTL_NONTASKING_OBJS): Add a-llltio, a-lllwti,
a-lllzti and remove a-timoau, a-wtmoau and a-ztmoau.
(GNATRTL_128BIT_PAIRS): Add a-tiinio.adb, a-timoio.adb, a-wtinio.adb,
a-wtmoio.adb, a-ztinio.adb and a-ztmoio.adb.
* impunit.adb (Non_Imp_File_Names_95): Add a-llltio, a-lllwti and
a-lllzti.
* krunch.ads: Document trick for Ada.Long_Long_Long_Integer_*_IO.
* krunch.adb (Krunch): Add trick for Ada.Long_Long_Long_Integer_*_IO.
* libgnat/a-llltio.ads: Instantiate Ada.Text_IO.Integer_IO.
* libgnat/a-lllwti.ads: Instantiate Ada.Wide_Text_IO.Integer_IO.
* libgnat/a-lllzti.ads: Instantiate Ada.Wide_Wide_Text_IO.Integer_IO.
* libgnat/a-tigeau.ads (Load_Integer): New procedure.
* libgnat/a-tigeau.adb (Load_Integer): Likewise.
* libgnat/a-tiinau.ads, libgnat/a-tiinau.adb: Change to generic
package.
* libgnat/a-tiinio.adb: Instantiate it.
* libgnat/a-tiinio__128.adb: Likewise.
* libgnat/a-timoau.ads, libgnat/a-timoau.adb: Change to generic
package.
* libgnat/a-timoio.adb: Instantiate it.
* libgnat/a-timoio__128.adb: Likewise.
* libgnat/a-wtgeau.ads (Load_Integer): New procedure.
* libgnat/a-wtgeau.adb (Load_Integer): Likewise.
* libgnat/a-wtinau.ads, libgnat/a-wtinau.adb: Change to generic
package.
* libgnat/a-wtinio.adb: Instantiate it.
* libgnat/a-wtinio__128.adb: Likewise.
* libgnat/a-wtmoau.ads, libgnat/a-wtmoau.adb: Change to generic
package.
* libgnat/a-wtmoio.adb: Instantiate it.
* libgnat/a-wtmoio__128.adb: Likewise.
* libgnat/a-ztgeau.ads (Load_Integer): New procedure.
* libgnat/a-ztgeau.adb (Load_Integer): Likewise.
* libgnat/a-ztinau.ads, libgnat/a-ztinau.adb: Change to generic
package.
* libgnat/a-ztinio.adb: Instantiate it.
* libgnat/a-ztinio__128.adb: Likewise.
* libgnat/a-ztmoau.ads, libgnat/a-ztmoau.adb: Change to generic
package.
* libgnat/a-ztmoio.adb: Instantiate it.
* libgnat/a-ztmoio__128.adb: Likewise.
Eric Botcazou [Wed, 5 Aug 2020 13:50:06 +0000 (15:50 +0200)]
[Ada] Fix problematic placement of freeze node after instantiation
gcc/ada/
* sem_ch12.adb (Freeze_Subprogram_Body): Do not move the freeze
node of the package body enclosing the instance when its parent
is in the same declarative part as the freeze node of the parent.
Steve Baird [Sat, 8 Aug 2020 22:04:21 +0000 (15:04 -0700)]
[Ada] Implement missing function result finalization.
gcc/ada/
* exp_ch6.adb (Insert_Post_Call_Actions): When a function's
result type requires finalization and we decide to make copy of
a call to the function and subsequently refer only to the copy,
then don't forget to finalize the original function result
object.
Eric Botcazou [Sat, 27 Jun 2020 10:39:49 +0000 (12:39 +0200)]
[Ada] Support of attributes Image, Put_Image, Val and Width for 128-bit types
gcc/ada/
* Makefile.rtl (GNATRTL_NONTASKING_OBJS): Add s-widint,
s-widthi, s-widuns, s-widuns.
(GNATRTL_128BIT_OBJS): Add s-imglllb, s-imgllli, s-imglllu, s-imglllw,
s-valllli, s-vallllu, s-widllli, s-widlllu.
* exp_imgv.adb (Expand_Image_Attribute): Deal with 128-bit types.
(Expand_Value_Attribute): Likewise.
(Expand_Width_Attribute): Likewise.
* exp_put_image.adb (Build_Elementary_Put_Image_Call): Likewise.
* krunch.adb (Krunch): Deal with s-img, s-val and s-wid prefixes.
* rtsfind.ads (RTU_Id): Add System_Img_LLLI, System_Img_LLLU,
System_Val_LLLI, System_Val_LLL, System_Wid_Int, System_Wid_LLLI,
System_Wid_LLLU, System_Wid_Uns).
(RE_Id): Add RE_Image_Long_Long_Long_Integer,
RE_Image_Long_Long_Long_Unsigned,
RE_Put_Image_Long_Long_Long_Integer,
RE_Put_Image_Long_Long_Long_Unsigned,
RE_Long_Long_Long_Unsigned, RE_Value_Long_Long_Long_Integer,
RE_Value_Long_Long_Long_Unsigned, RE_Width_Integer,
RE_Width_Long_Long_Long_Integer, RE_Width_Long_Long_Long_Unsigned,
RE_Width_Unsigned, RE_Image_Long_Long_Long_Integer,
RE_Image_Long_Long_Long_Unsigned, RE_Put_Image_Long_Long_Long_Integer,
RE_Put_Image_Long_Long_Long_Unsigned, RE_Long_Long_Long_Unsigned,
RE_Value_Long_Long_Long_Integer, RE_Value_Long_Long_Long_Unsigned,
RE_Width_Integer, RE_Width_Long_Long_Long_Integer,
RE_Width_Long_Long_Long_Unsigned, RE_Width_Unsigned.
* libgnat/s-imageb.ads, libgnat/s-imageb.adb: New generic
package.
* libgnat/s-imagei.ads, libgnat/s-imagei.adb: Likewise.
* libgnat/s-imageu.ads, libgnat/s-imageu.adb: Likewise.
* libgnat/s-imagew.ads, libgnat/s-imagew.adb: Likewise.
* libgnat/s-imgbiu.ads: Instantiate System.Image_B.
* libgnat/s-imgbiu.adb: Add pragma No_Body.
* libgnat/s-imgint.ads: Instantiate System.Image_I.
* libgnat/s-imgint.adb: Add pragma No_Body.
* libgnat/s-imgllb.ads: Instantiate System.Image_B.
* libgnat/s-imgllb.adb: Add pragma No_Body0
* libgnat/s-imglli.ads: Instantiate System.Image_I.
* libgnat/s-imglli.adb: Add pragma No_Body.
* libgnat/s-imglllb.ads: Instantiate System.Image_B.
* libgnat/s-imgllli.ads: Instantiate System.Image_I.
* libgnat/s-imglllu.ads: Instantiate System.Image_U.
* libgnat/s-imglllw.ads: Instantiate System.Image_W.
* libgnat/s-imgllu.ads: Instantiate System.Image_U.
* libgnat/s-imgllu.adb: Add pragma No_Body.
* libgnat/s-imgllw.ads: Instantiate System.Image_W.
* libgnat/s-imgllw.adb: Add pragma No_Body.
* libgnat/s-imgrea.adb: Remove clauses for System.Unsigned_Types.
* libgnat/s-imguns.ads: Instantiate System.Image_U.
* libgnat/s-imguns.adb: Add pragma No_Body.
* libgnat/s-imgwiu.ads: Instantiate System.Image_W.
* libgnat/s-imgwiu.adb: Add pragma No_Body.
* libgnat/s-putima.ads (Long_Long_Long_Unsigned): New subtype.
(Put_Image_Long_Long_Long_Unsigned): New procedure.
* libgnat/s-putima.adb (Small): Rename to Integer_Images.
(Large): Rename to LL_Integer_Images.
(LLL_Integer_Images): New instantiation.
(Put_Image_Long_Long_Long_Integer): New renaming.
(Put_Image_Long_Long_Long_Unsigned): Likewise.
* libgnat/s-valint.ads: Instantiate System.Value_I.
* libgnat/s-valint.adb: Add pragma No_Body.
* libgnat/s-vallli.ads: Instantiate System.Value_I.
* libgnat/s-vallli.adb: Add pragma No_Body.
* libgnat/s-valllli.ads: Instantiate System.Value_I.
* libgnat/s-vallllu.ads: Instantiate System.Value_U.
* libgnat/s-valllu.ads: Instantiate System.Value_U.
* libgnat/s-valllu.adb: Add pragma No_Body.
* libgnat/s-valuei.ads, libgnat/s-valuei.adb: New generic
package.
* libgnat/s-valueu.ads, libgnat/s-valueu.adb: Likewise.
* libgnat/s-valuns.ads: Instantiate System.Value_U.
* libgnat/s-valuns.adb: Add pragma No_Body.
* libgnat/s-widint.ads: Instantiate System.Width_I.
* libgnat/s-widlli.ads: Likewise.
* libgnat/s-widlli.adb: Add pragma No_Body.
* libgnat/s-widllli.ads: Instantiate System.Width_I.
* libgnat/s-widlllu.ads: Instantiate System.Width_U.
* libgnat/s-widllu.ads: Likewise.
* libgnat/s-widllu.adb: Add pragma No_Body.
* libgnat/s-widthi.ads, libgnat/s-widthi.adb: New generic
package.
* libgnat/s-widthu.ads, libgnat/s-widthu.adb: Likewise.
* libgnat/s-widuns.ads: Instantiate System.Width_U.
Eric Botcazou [Sun, 9 Aug 2020 09:48:21 +0000 (11:48 +0200)]
[Ada] Work around missing Long_Long_Long_Size entry in .atp file
gcc/ada/
* set_targ.adb (DTN): Fix oversight.
(Read_Target_Dependent_Values): Do not error out on missing
Long_Long_Long_Size entry and reuse Long_Long_Size for it.
Piotr Trojanek [Fri, 7 Aug 2020 07:13:08 +0000 (09:13 +0200)]
[Ada] Fix analysis of iterated component expression with null range
gcc/ada/
* exp_aggr.adb (Gen_Loop): Analyze copy of the expression in the
scope of the implicit loop with name of the index parameter
visible.
Eric Botcazou [Fri, 7 Aug 2020 13:41:06 +0000 (15:41 +0200)]
[Ada] Basic support for 128-bit types
gcc/ada/
* Makefile.rtl (GNATRTL_128BIT_PAIRS): New variable.
(GNATRTL_128BIT_OBJS): Likewise.
(Windows): In 64-bit mode, add the former to LIBGNAT_TARGET_PAIRS and
the latter to EXTRA_GNATRTL_NONTASKING_OBJS.
(x86_64/linux): Likewise, but unconditionally.
(GNATRTL_NONTASKING_OBJS): Add s-aridou, s-exponn, s-expont,
s-exponu.
* ada_get_targ.adb (Get_Long_Long_Long_Size): New function.
* checks.adb (Apply_Arithmetic_Overflow_Strict): Use Integer_Type_For
to find an appropriate integer type; if it does not exist and the max
integer size is larger than 64, use the 128-bit arithmetic routines.
* cstand.adb (Create_Standard): Build Standard_Long_Long_Long_Integer
and its base type. Use it for Etype of Any_Integer, Any_Modular and
Any_Numeric. Use its size for Build Standard_Long_Long_Long_Unsigned
and Universal_Integer.
(Print_Standard): Print Long_Long_Long_Integer.
* exp_attr.adb (Expand_N_Attribute_Reference) <Attribute_Mod>: Adjust
comment.
* exp_ch3.adb (Simple_Init_Initialize_Scalars_Type): Deal with 128-bit
types.
* exp_ch4.adb (Expand_Array_Comparison): Likewise.
(Expand_N_Op_Expon): Likewise.
(Narrow_Large_Operation): Likewise.
* exp_dbug.adb (Bounds_Match_Size): Handle 128-bit size.
* exp_fixd.adb (Build_Double_Divide_Code): Use RE_Double_Divide64.
* exp_intr.adb (Expand_Binary_Operator_Call): Handle 128-bit size.
* exp_pakd.ads (E_Array): Extend range to 127.
(Bits_Id): Fill in up to 127.
(Get_Id): Likewise.
(GetU_Id): Likewise.
(Set_Id): Likewise.
(SetU_Id): Likewise.
* exp_pakd.adb (Revert_Storage_Order): Handle 128-bit size.
* exp_util.adb (Integer_Type_For): Likewise.
(Small_Integer_Type_For): Likewise.
* fname.adb (Is_Predefined_File_Name): Do not return False for names
larger than 12 characters if they start with "s-".
* freeze.adb (Adjust_Esize_For_Alignment): Change the maximum value
to System_Max_Integer_Size.
(Check_Suspicious_Modulus): Adjust comment.
(Freeze_Entity): Likewise.
* get_targ.ads (Get_Long_Long_Long_Size): New function.
* get_targ.adb (Get_Long_Long_Long_Size): Likewise.
(Width_From_Size): Deal with 128-bit size.
* gnat1drv.adb (Adjust_Global_Switches): Deal with 128-bit types.
* impunit.adb (Get_Kind_Of_File): Bump buffer size. Accept files with
13 characters if they start with 's'. Compare slice of Buffer.
(Not_Impl_Defined_Unit): Accept files with 13 characters if they start
with 's'.
* krunch.ads: Document length for 128-bit support units.
* krunch.adb (Krunch): Set length to 9 for 128-bit support units.
* layout.adb (Layout_Type): Use System_Max_Integer_Size as alignment
limit.
* rtsfind.ads (RTU_Id): Add System_Arith_128,
System_Compare_Array_Signed_128, System_Compare_Array_Unsigned_128,
System_Exn_LLLI, System_Exp_LLLU, System_Pack_[65..127].
(RE_Id): Add RE_Integer_128, RE_Unsigned_128, RE_Add_With_Ovflo_Check128
RE_Multiply_With_Ovflo_Check128, RE_Subtract_With_Ovflo_Check128,
RE_Bswap_128, RE_Compare_Array_S128, RE_Compare_Array_U128,
RE_Exn_Long_Long_Long_Integer, RE_Exp_Long_Long_Long_Integer,
RE_Exp_Long_Long_Long_Unsigned, RE_Bits_[65-127], RE_Get_[65-127],
RE_Set_[65-127], RE_IS_Is16, RE_IS_Iu16, RE_Integer_128 and
RE_Unsigned_128. Rename RE_Add_With_Ovflo_Check, RE_Double_Divide,
RE_Multiply_With_Ovflo_Check, RE_Scaled_Divide and
RE_Subtract_With_Ovflo_Check. Remove RE_IS_Iz1, RE_IS_Iz2, RE_IS_Iz4,
RE_IS_Iz8, RE_Long_Unsigned, RE_Short_Unsigned, RE_Short_Short_Unsigned
(RE_Unit_Table): Likewise.
* sem_aux.adb (Corresponding_Unsigned_Type): Deal with a size equal to
that of Standard_Long_Long_Long_Integer.
(First_Subtype): Deal with Standard_Long_Long_Long_Integer'Base.
* sem_ch13.adb (Analyze_Attribute_Definition_Clause) <Attribute_Size>:
Check the size against powers of 2 and System_Max_Integer_Size for
objects as well.
(Set_Enum_Esize): Deal with 128-bit bounds.
* sem_ch3.adb (Set_Modular_Size): Handle 128-bit size.
(Modular_Type_Declaration): Deal with 128-bit types.
(Signed_Integer_Type_Declaration): Support derivation from
Standard_Long_Long_Long_Integer.
* sem_ch4.adb (Analyze_Mod): Handle 128-bit modulus.
* sem_intr.adb: Add with and use clauses for Ttypes.
(Check_Shift): Handle 128-bit size and modulus.
* sem_prag.adb (Analyze_Pragma) <Pragma_Initialize_Scalars>: Deal
with Signed_128 and Unsigned_128.
(Analyze_Integer_Value): Handle 128-bit size.
* sem_util.ads (Addressable): Adjust description.
* sem_util.adb (Addressable): Return true for 128 if the system
supports 128 bits.
(Set_Invalid_Binder_Values): Deal with Signed_128 and Unsigned_128.
* set_targ.ads (Long_Long_Long_Size): New variable.
* set_targ.adb (S_Long_Long_Long_Size): New constant.
(DTN): Add entry for S_Long_Long_Long_Size.
(DTV): Add entry for Long_Long_Long_Size.
(Set_Targ): Set Long_Long_Long_Size.
* snames.ads-tmpl (Name_Max_Integer_Size): New attribute name.
(Name_Signed_128): New scalar name.
(Name_Unsigned_128): Likewise.
(Scalar_Id): Adjust.
(Integer_Scalar_Id): Likewise.
(Attribute_Id): Add Attribute_Max_Integer_Size.
* stand.ads (Standard_Entity_Type): Add S_Long_Long_Long_Integer.
(Standard_Long_Long_Long_Integer): New renaming.
(Universal_Integer): Adjust description.
(Standard_Long_Long_Long_Unsigned): New variable.
* switch-c.adb (Scan_Front_End_Switches): Deal with -gnate128.
* ttypes.ads (Standard_Long_Long_Long_Integer_Size): New variable.
(Standard_Long_Long_Long_Integer_Width): Likewise.
(System_Max_Integer_Size): Turn into variable.
(System_Max_Binary_Modulus_Power): Likewise.
* uintp.ads (Uint_127): New constant.
* uintp.adb (UI_Power_2): Extednd to 128.
(UI_Power_10): Likewise.
(UI_Expon): Deal with exponent up to 128 specially.
* usage.adb (Write_Switch_Char): Print -gnate128 switch.
* libgnat/a-tifiio.adb (Put_Scaled): Call Scaled_Divide64.
* libgnat/interfac__2020.ads (Integer_128): New integer type.
(Unsigned_128): New modular type.
(Shift_Left, Shift_Right, Shift_Right_Arithmetic, Rotate_Left,
Rotate_Right): New intrinsic functions operating on it.
* libgnat/s-aridou.ads, libgnat/s-aridou.adb: New generic
package.
* libgnat/s-arit64.ads, libgnat/s-arit64.adb: Instantiate
System.Arithmetic_Double.
* libgnat/s-arit128.ads, libgnat/s-arit128.adb: Likewise.
* libgnat/s-bytswa.ads: Add with clause for Interfaces, use subtypes
of unsigned types defined in Interfaces and add Bswap_128.
* libgnat/s-casi128.ads, libgnat/s-casi128.adb: New package.
* libgnat/s-caun128.ads, libgnat/s-caun128.adb: Likewise.
* libgnat/s-exnint.ads: Instantiate System.Exponn.
* libgnat/s-exnint.adb: Add pragma No_Body.
* libgnat/s-exnlli.ads: Instantiate System.Exponn.
* libgnat/s-exnlli.adb: Add pragma No_Body.
* libgnat/s-exnllli.ads: Instantiate System.Exponn.
* libgnat/s-expint.ads: Likewise.
* libgnat/s-expint.adb: Add pragma No_Body.
* libgnat/s-explli.ads: Instantiate System.Exponn.
* libgnat/s-explli.adb: Add pragma No_Body.
* libgnat/s-expllli.ads: Instantiate System.Exponn.
* libgnat/s-explllu.ads: Instantiate System.Exponu.
* libgnat/s-expllu.ads: Likewise.
* libgnat/s-expllu.adb: Add pragma No_Body.
* libgnat/s-exponn.ads, libgnat/s-exponn.adb: New generic
function.
* libgnat/s-expont.ads, libgnat/s-expont.adb: Likewise.
* libgnat/s-exponu.ads, libgnat/s-exponu.adb: Likewise.
* libgnat/s-expuns.ads, libgnat/s-expuns.adb: Likewise.
* libgnat/s-pack65.ads, libgnat/s-pack65.adb: New package.
* libgnat/s-pack66.ads, libgnat/s-pack66.adb: New package.
* libgnat/s-pack67.ads, libgnat/s-pack67.adb: New package.
* libgnat/s-pack68.ads, libgnat/s-pack68.adb: New package.
* libgnat/s-pack69.ads, libgnat/s-pack69.adb: New package.
* libgnat/s-pack70.ads, libgnat/s-pack70.adb: New package.
* libgnat/s-pack71.ads, libgnat/s-pack71.adb: New package.
* libgnat/s-pack72.ads, libgnat/s-pack72.adb: New package.
* libgnat/s-pack73.ads, libgnat/s-pack73.adb: New package.
* libgnat/s-pack74.ads, libgnat/s-pack74.adb: New package.
* libgnat/s-pack75.ads, libgnat/s-pack75.adb: New package.
* libgnat/s-pack76.ads, libgnat/s-pack76.adb: New package.
* libgnat/s-pack77.ads, libgnat/s-pack77.adb: New package.
* libgnat/s-pack78.ads, libgnat/s-pack78.adb: New package.
* libgnat/s-pack79.ads, libgnat/s-pack79.adb: New package.
* libgnat/s-pack80.ads, libgnat/s-pack80.adb: New package.
* libgnat/s-pack81.ads, libgnat/s-pack81.adb: New package.
* libgnat/s-pack82.ads, libgnat/s-pack82.adb: New package.
* libgnat/s-pack83.ads, libgnat/s-pack83.adb: New package.
* libgnat/s-pack84.ads, libgnat/s-pack84.adb: New package.
* libgnat/s-pack85.ads, libgnat/s-pack85.adb: New package.
* libgnat/s-pack86.ads, libgnat/s-pack86.adb: New package.
* libgnat/s-pack87.ads, libgnat/s-pack87.adb: New package.
* libgnat/s-pack88.ads, libgnat/s-pack88.adb: New package.
* libgnat/s-pack89.ads, libgnat/s-pack89.adb: New package.
* libgnat/s-pack90.ads, libgnat/s-pack90.adb: New package.
* libgnat/s-pack91.ads, libgnat/s-pack91.adb: New package.
* libgnat/s-pack92.ads, libgnat/s-pack92.adb: New package.
* libgnat/s-pack93.ads, libgnat/s-pack93.adb: New package.
* libgnat/s-pack94.ads, libgnat/s-pack94.adb: New package.
* libgnat/s-pack95.ads, libgnat/s-pack95.adb: New package.
* libgnat/s-pack96.ads, libgnat/s-pack96.adb: New package.
* libgnat/s-pack97.ads, libgnat/s-pack97.adb: New package.
* libgnat/s-pack98.ads, libgnat/s-pack98.adb: New package.
* libgnat/s-pack99.ads, libgnat/s-pack99.adb: New package.
* libgnat/s-pack100.ads, libgnat/s-pack100.adb: New package.
* libgnat/s-pack101.ads, libgnat/s-pack101.adb: New package.
* libgnat/s-pack102.ads, libgnat/s-pack102.adb: New package.
* libgnat/s-pack103.ads, libgnat/s-pack103.adb: New package.
* libgnat/s-pack104.ads, libgnat/s-pack104.adb: New package.
* libgnat/s-pack105.ads, libgnat/s-pack105.adb: New package.
* libgnat/s-pack106.ads, libgnat/s-pack106.adb: New package.
* libgnat/s-pack107.ads, libgnat/s-pack107.adb: New package.
* libgnat/s-pack108.ads, libgnat/s-pack108.adb: New package.
* libgnat/s-pack109.ads, libgnat/s-pack109.adb: New package.
* libgnat/s-pack110.ads, libgnat/s-pack110.adb: New package.
* libgnat/s-pack111.ads, libgnat/s-pack111.adb: New package.
* libgnat/s-pack112.ads, libgnat/s-pack112.adb: New package.
* libgnat/s-pack113.ads, libgnat/s-pack113.adb: New package.
* libgnat/s-pack114.ads, libgnat/s-pack114.adb: New package.
* libgnat/s-pack115.ads, libgnat/s-pack115.adb: New package.
* libgnat/s-pack116.ads, libgnat/s-pack116.adb: New package.
* libgnat/s-pack117.ads, libgnat/s-pack117.adb: New package.
* libgnat/s-pack118.ads, libgnat/s-pack118.adb: New package.
* libgnat/s-pack119.ads, libgnat/s-pack119.adb: New package.
* libgnat/s-pack120.ads, libgnat/s-pack120.adb: New package.
* libgnat/s-pack121.ads, libgnat/s-pack121.adb: New package.
* libgnat/s-pack122.ads, libgnat/s-pack122.adb: New package.
* libgnat/s-pack123.ads, libgnat/s-pack123.adb: New package.
* libgnat/s-pack124.ads, libgnat/s-pack124.adb: New package.
* libgnat/s-pack125.ads, libgnat/s-pack125.adb: New package.
* libgnat/s-pack126.ads, libgnat/s-pack126.adb: New package.
* libgnat/s-pack127.ads, libgnat/s-pack127.adb: New package.
* libgnat/s-rannum.ads (Random): New function returning 128-bit.
* libgnat/s-rannum.adb (Random): Implement it.
* libgnat/s-scaval.ads: Add with clause for Interfaces, use subtypes
of unsigned types defined in Interfaces.
* libgnat/s-scaval.adb: Add use clause for Interfaces.
* libgnat/s-scaval__128.ads, libgnat/s-scaval__128.adb: New
package.
* libgnat/s-unstyp.ads (Long_Long_Long_Unsigned): New modular type.
(Shift_Left, Shift_Right, Shift_Right_Arithmetic, Rotate_Left,
Rotate_Right): New intrinsic functions operating on it.
gcc/testsuite/
* gnat.dg/multfixed.adb: Update expected exception message.
Eric Botcazou [Fri, 7 Aug 2020 13:23:23 +0000 (15:23 +0200)]
[Ada] Document Long_Long_Long_Integer in implementation characteristics
gcc/ada/
* doc/gnat_rm/implementation_defined_characteristics.rst: Add
Long_Long_Long_Integer to the list of predefined integer types.
* gnat_rm.texi: Regenerate.
Yannick Moy [Fri, 7 Aug 2020 10:20:12 +0000 (12:20 +0200)]
[Ada] Fix target configuration file used for CodePeer/SPARK for new ints
gcc/ada/
* ada_get_targ.adb (Width_From_Size): Add case for 128 bits.
Reorder declarations in the same order as get_targ.adb to
facilitate diffs.
Piotr Trojanek [Thu, 6 Aug 2020 20:19:18 +0000 (22:19 +0200)]
[Ada] Refactor repeated calls to Etype in aggregate expansion
gcc/ada/
* exp_aggr.adb (Expand_N_Aggregate): Refactor repeated calls to
Etype (N).
(Build_Array_Aggr_Code): Fix whitespace.
Dmitriy Anisimkov [Thu, 6 Aug 2020 05:54:48 +0000 (11:54 +0600)]
[Ada] Fix bootstrap with old GCC
gcc/ada/
* adaint.c (__gnat_file_time): Use regular arithmetic instead of
__builtin_*_overflow routines if GCC version 4 or less and
compiler is g++.
Piotr Trojanek [Wed, 5 Aug 2020 20:17:02 +0000 (22:17 +0200)]
[Ada] Fix crash on illegal OTHERS in iterated_component_association
gcc/ada/
* sem_aggr.adb (Resolve_Array_Aggregate): Use Choice_List, which
internally calls either Choice or Discrete_Choices, depending on
the context.
Piotr Trojanek [Tue, 4 Aug 2020 15:47:39 +0000 (17:47 +0200)]
[Ada] Use index parameter for iterated_component_association
gcc/ada/
* sem_aggr.adb (Resolve_Iterated_Component_Association): Use
existing defining identifier for index parameter.
Javier Miranda [Thu, 6 Aug 2020 17:56:50 +0000 (13:56 -0400)]
[Ada] Wrong detection of potentially blocking call in protected object
gcc/ada/
* exp_ch9.adb (Build_Task_Activation_Call): Do not generate a
call to activate tasks if we are within the scope of a protected
type and pragma Detect_Blocking is active.
Liaiss Merzougue [Mon, 13 Jan 2020 12:07:26 +0000 (13:07 +0100)]
[Ada] Codepeer remarks take into account
gcc/ada/
* libgnat/s-carsi8.adb (Compare_Array_S8): Add pragma Assert to
avoid warning concerning Left_Len and RighLen value regarding
Bytes_Compared_As_Words.
* libgnat/s-carun8.adb (Compare_Array_U8): Likewise.
* libgnat/s-geveop.adb (Binary_Operation, Unary_Operation): Add
pragma Assert concerning divide by 0 warning.
* libgnat/s-imgcha.adb (Image_Character): Code update to prevent
constant operation warning.
(Image_Character): Add pragma Assert concerning the unchecked
String size.
* libgnat/s-imgdec.adb
(Round): Upate loop code to prevent warning concerning
Digs'First access.
(Round): Add pragma assert.
(Set): Add pragma Assert for the unchecked string size.
(Set_Digits): Add pragma Assert for the input range.
(Set_Decimal_Digits): Add pragma Assert.
(Set_Blank_And_Sign): Add pragma Assert for the input range.
* libgnat/s-arit64.adb (DoubleDivide): Add pragma Assert
concerning Du /= 0.
(Multiply_With_Ovflo_Check): Add pragma Annotate to avoid
warning concerning unsigned -> signed conversion.
* libgnat/s-imguns.adb (Set_Image_Unsigned): Add pragma Assert
to prevent overflow check warning. Add pragma Assert for
controlling S'First = 1.
* libgnat/s-imgrea.adb (Image_Floating_Point, Set, Set_Digs,
Set_Special_Fill, Convert_Integer): Add pragma Annotate to
prevent overflow check warning.
(Set_Image_Real): Add pragma Annotate to avoid dead code warning
on float check. Add pragma Assert to prevent overflow check
warning.
* libgnat/s-imgwiu.adb (Set_Digits, Set_Image_Width_Unsigned):
Add pragma assert to prevent overflow check warning.
* libgnat/s-imgllu.adb (Set_Image_Long_Long_Unsigned): Add
pragma assert to prevent overflow check warning.
* libgnat/s-imgint.adb (Set_Digits): Add Assert for input
constraint and to prevent overflow check warning, create
Non_Positive subtype, and change the T parameter as Non_Positive
instead Integer.
(Set_Image_Integer): Add pragma assert to prevent overflow check
warning.
* libgnat/s-imglli.adb (Set_Digits): Add Assert for input
constraint and to prevent overflow check warning, create
Non_Positive subtype, and change the T parameter as Non_Positive
instead Integer.
(Set_Image_Long_Long_Integer): Add pragma assert to prevent
overflow check warning.
* libgnat/s-fatgen.adb (Decompose, Pred, Succ): Add pragma
Annotate to prevent dead code due to invalid float check.
* libgnat/s-imenne.adb (Image_Enumeration_8,
Image_Enumeration_16, Image_Enumeration_32): Add pragma Assert
to prevent overflow check warning. Add Names_Index subtype for
restricting Index_table content.
Gary Dismukes [Wed, 5 Aug 2020 20:29:30 +0000 (16:29 -0400)]
[Ada] Compiler crash on prefixed call to controlled function with invariant check
gcc/ada/
* exp_ch6.adb (Insert_Post_Call_Actions): Test for
N_Explicit_Dereference as part of the existing test for function
calls.
Eric Botcazou [Tue, 4 Aug 2020 13:08:56 +0000 (15:08 +0200)]
[Ada] Fix bogus error for bit-packed array with volatile component
gcc/ada/
* freeze.adb (Check_Strict_Alignment): Do not set the flag for
a bit-packed array type, even if it is a by-reference type.
Dmitriy Anisimkov [Mon, 3 Aug 2020 06:18:20 +0000 (12:18 +0600)]
[Ada] Improve precision of Ada.Directories.Modification_Time
gcc/ada/
* adaint.c (__gnat_file_time): New routine.
(__gnat_copy_attribs): Copy timestamps in nanoseconds.
* libgnat/a-direct.adb (C_Modification_Time): Bind to
__gnat_file_time.
(Modification_Time): Call to C_Modification_Time.
Piotr Trojanek [Mon, 3 Aug 2020 20:54:18 +0000 (22:54 +0200)]
[Ada] Fix crash with iterated_component_association and -gnatc
gcc/ada/
* sem_aggr.adb (Resolve_Iterated_Component_Association):
Expression's copy and now has the same parent as the original
expression.
(Resolve_Array_Aggregate): Add ??? comment about a still
existing minor issue that led to discovery of the above crash.
Javier Miranda [Mon, 3 Aug 2020 17:50:03 +0000 (13:50 -0400)]
[Ada] Spurious error in current instance used as formal package
gcc/ada/
* sem_ch12.adb (Install_Parents_Of_Generic_Context): Simplify
functionality; collect and install parents of generic child
package.
(Remove_Parents_Of_Generic_Context): Code cleanup.
(Instantiate_Package_Body): Hide parents of generic context from
visibility before installing the parent instance; restore their
visibility when the instance is analyzed
Doug Rupp [Mon, 3 Aug 2020 16:26:09 +0000 (09:26 -0700)]
[Ada] Remove -mthreads from Linker_Options pragma for x86-lynx178e
gcc/ada/
* libgnarl/s-osinte__lynxos178e.ads: Remove -mthreads switch.
Patrick Bernardi [Thu, 30 Jul 2020 23:06:13 +0000 (19:06 -0400)]
[Ada] Use VxWorks 7 APIs for accessing environment variables in kernel mode
gcc/ada/
* env.c (__gnat_setenv): call setenv for VxWorks 7 kernel mode.
(__gnat_environ): envGet takes an int instead of a NULL pointer.
(__gnat_unsetenv): call unsetenv for VxWorks 7 kernel mode.
(__gnat_clearenv): use __gnat_unsetenv to clear environment
variables.
Eric Botcazou [Sun, 2 Aug 2020 12:14:50 +0000 (14:14 +0200)]
[Ada] Use helper function in Freeze_Subprogram_Body
gcc/ada/
* sem_ch12.adb (Freeze_Subprogram_Body): Call
Package_Freeze_Node to retrieve the freeze node for the
enclosing body of the generic.
Justin Squirek [Thu, 30 Jul 2020 18:54:42 +0000 (14:54 -0400)]
[Ada] Incorrect associations for extra accessibility parameters
gcc/ada/
* exp_ch6.adb (Expand_Call_Helper): Modify calls to
Add_Extra_Actual to use Extra_Accessibility instead of
Get_Accessibility for the EF parameter.
Piotr Trojanek [Thu, 30 Jul 2020 10:02:19 +0000 (12:02 +0200)]
[Ada] Make minimum accessibility level a constant object
gcc/ada/
* exp_ch6.adb (Expand_Actuals): Whitespace cleanup.
* sem_ch6.adb (Analyze_Subprogram_Body_Helper): Make minimum
accessibility a constant object.
Patrick Bernardi [Tue, 28 Jul 2020 15:36:08 +0000 (11:36 -0400)]
[Ada] Simplify the VxWorks implementation of __gnat_environ
gcc/ada/
* env.c (__gnat_environ): For VxWorks kernel simply return the
result of the envGet call. Do this for VxWorks 6 and 7 as they
both support the same API.
Ghjuvan Lacambre [Thu, 23 Jul 2020 07:51:43 +0000 (09:51 +0200)]
[Ada] Disable warnings on entities when building finalizers
gcc/ada/
* exp_ch7.adb (Build_Finalizer): Disable warnings on referenced
entity.
Piotr Trojanek [Mon, 27 Jul 2020 20:06:18 +0000 (22:06 +0200)]
[Ada] Fix typos and style in comments related to Default_Initial_Condition
gcc/ada/
* einfo.ads, sem_ch3.adb, sem_util.adb: Fix comments.
Javier Miranda [Sat, 25 Jul 2020 17:16:32 +0000 (13:16 -0400)]
[Ada] Spurious error in taft ammendment derivation of CPP type
gcc/ada/
* sem_ch3.adb (Build_Derived_Type): Propagate convention of
class-wide parent.
Claire Dross [Tue, 21 Jul 2020 15:37:23 +0000 (17:37 +0200)]
[Ada] Raise Capacity_Error on formal vector insertion
gcc/ada/
* libgnat/a-cofove.adb (Copy): Add explanation in case of
Capacity_Error.
(Insert_Space): Raise Capacity_Error if the new length is
greater than the capacity.
(Reserve_Capacity): Raise Capacity_Error instead of
Constraint_Error.
Andrea Corallo [Wed, 21 Oct 2020 06:48:16 +0000 (08:48 +0200)]
aarch64: [testsuite] Fix typo in diagnostic message
gcc/testsuite/ChangeLog
2020-10-21 Andrea Corallo <andrea.corallo@arm.com>
* gcc.target/aarch64/advsimd-intrinsics/vstX_lane.c (CMT):
Adopt the same style used in the rest of the file.
Aldy Hernandez [Wed, 21 Oct 2020 06:47:03 +0000 (08:47 +0200)]
Fix discrepancy in Walloca test on 32-bit systems.
There is a discrepancy in the way we report -Walloca-larger-than=
errors on 32-bit versus 64-bit architectures, due to the nature of
ranges derived from a cast operation.
For the Walloca-1 tests on 64-bits we get:
int num.0_1;
long unsigned int _2;
<bb 2> [local count:
1073741824]:
num.0_1 = num;
_2 = (long unsigned int) num.0_1;
s_8 = __builtin_alloca (_2);
Because of the cast of a 32-bit quantity into a 64-bit quantity in _2,
ranger calculates its range as:
long unsigned int [0,
2147483647][
18446744071562067968, +INF]
Thus excluding the numbers that can't exist in _2.
This causes the Walloca pass to report that the argument to alloca may be
too large.
However, for -m32 on x86, the gimple is:
int num.0_1;
unsigned int num.1_2;
<bb 2> [local count:
1073741824]:
num.0_1 = num;
num.1_2 = (unsigned int) num.0_1;
s_8 = __builtin_alloca (num.1_2);
Since num.0_1 and num.1_2 are of the same size, we cannot determine
any useful range, so we return VARYING. In the Walloca pass, VARYING
basically means "unbounded" (no known bounds for the alloca call argument).
So on 32-bits, the error message issued is slightly different:
warning: unbounded use of ‘alloca’
versus on 64-bits, where due to the cast, it is:
warning: argument to ‘alloca’ may be too large
In reality both versions of the IL show an unbounded call, but in one
version (64-bits) we can exclude some values so we assume the range
was provided, but it was out of bounds.
I've mentioned various times that all these diagnostics passes
(alloca, restrict, printf, etc), could benefit from less specific error
messages since what we have can potentially confuse the user. However,
no consensus has been reached on how to report these things.
In the meantime, this patch adjusts the testcase to accept both variants.
gcc/testsuite/ChangeLog:
* gcc.dg/Walloca-1.c: Adjust for 32-bits.
GCC Administrator [Wed, 21 Oct 2020 00:16:36 +0000 (00:16 +0000)]
Daily bump.
Jeff Law [Tue, 20 Oct 2020 23:28:34 +0000 (17:28 -0600)]
Handle 16 bit targets in diagnostic pruning
gcc/testsuite
* gcc.dg/Wbuiltin-declaration-mismatch-9.c: Improve pruning of
invalid scanf call messages.
Andrew MacLeod [Tue, 20 Oct 2020 20:55:14 +0000 (16:55 -0400)]
Temporarily disable trap in in extract_range_builtin check.
Until we figure out how to adjust ubsan for symbolics, disable the trap.
gcc/ChangeLog:
PR tree-optimization/97505
* vr-values.c (vr_values::extract_range_basic): Trap if
vr_values version disagrees with range_of_builtin_call.
Nikhil Benesch [Mon, 19 Oct 2020 02:07:34 +0000 (02:07 +0000)]
libgo: adjust NetBSD-specific types for stable syscall API
The backwards-compatibility guarantees of the syscall package
require some munging of the C API inferred by mksysinfo.sh.
Specifically, the RTM_RESOLVE constant must be added if it is
missing, and the stat_t struct must use the suffix "timespec"
rather than "tim" for its time-related fields.
Reviewed-on: https://go-review.googlesource.com/c/gofrontend/+/263519
David Edelsohn [Mon, 19 Oct 2020 17:11:56 +0000 (13:11 -0400)]
rs6000: Allow -mabi=no-altivec override on command line for AIX.
The override options machinery to set rs6000_altivec_abi for AIX
doesn't check if the option is set on the command line. This patch
checks for a command line option to not override it.
gcc/ChangeLog:
* config/rs6000/rs6000.c (rs6000_option_override_internal):
Don't implcitly enable Altivec ABI if set on the command line.
François Dumont [Sun, 4 Oct 2020 16:06:11 +0000 (18:06 +0200)]
libstdc++: Refactor _Hashtable to prepare for custom pointers
Limit usage of node pointers in _Hashtable implementation details so that when
we move to allocator custom pointer type we do not need to add an _Alloc
template parameter everywhere which would impact abi.
This is done by reviewing node type definition. It is now based on new basic
types which are not pointer dependant. The _Hashtable helper
types are now using those new node base types and do not receive node pointers.
libstdc++-v3/ChangeLog
* include/bits/hashtable_policy.h
(_Hash_node_value_base<>): Remove _Hash_node_base inheritance.
(_Hash_node_code_cache<bool _Cache_hash_code>): New.
(_Hash_node_value<typename _Value, bool _Cache_hash_code>): New.
(_Hash_node<>): Inherits _Hash_node_base<> and _Hash_node_value<>.
(_Map_base<>::__node_type): Remove.
(_Map_base<>::iterator): Remove.
(_Insert_base<>::__hash_cached): New.
(_Insert_base<>::__constant_iterators): New.
(_Insert_base<>::__hashtable_alloc): New.
(_Insert_base<>::__node_type): Remove.
(_Insert_base<>::__node_ptr): New.
(_Hash_code_base<>): Remove specializations.
(_Hash_code_base<>::__node_type): Remove.
(_Hash_code_base<>::_M_bucket_index(const __node_type*, size_t)):
Replace by...
(_Hash_code_base<>::_M_bucket_index(const _Hash_node_value<>&, size_t)):
...this.
(_Hash_code_base<>::_M_store_code(__node_type*, __hash_code)):
Replace by...
(_Hash_code_base<>::_M_store_code(_Hash_node_code_cache<>&, __hash_code)):
...this.
(_Hash_code_base<>::_M_copy_code(__node_type*, const __node_type*)):
Replace by...
(_Hash_code_base<>::_M_copy_code(_Hash_node_code_cache<>&,
const _Hash_node_code_base<>&)): ...this.
(_Hashtable_base<>::__constant_iterators): Remove.
(_Hashtable_base<>::__unique_keys): Remove.
(_Hashtable_base<>::__node_type): Remove.
(_Hashtable_base<>::iterator): Remove.
(_Hashtable_base<>::const_iterator): Remove.
(_Hashtable_base<>::local_iterator): Remove.
(_Hashtable_base<>::const_local_iterator): Remove.
(_Hashtable_base<>::__ireturn_type): Remove.
(_Hashtable_base<>::_Equal_hash_code<>::_S_equals): Replace by...
(_Hashtable_base<>::_S_equals(__hash_code, const _Hash_node_code_hash<>&)):
...this.
(_Hashtable_base<>::_Equal_hash_code<>::_S_node_equals): Replace by...
(_Hashtable_base<>::_S_node_equals(__hash_code,
const _Hash_node_code_hash<>&)): ...this.
(_Hashtable_base<>::_Equal_hash_code<>): Remove.
(_Hashtable_base<>::_M_equals): Adapt.
(_Hashtable_baxe<>::_M_node_equals): Adapt.
(_Equality<>::_M_equal): Adapt.
(_Hashtable_alloc<>::__node_ptr): New.
(_Hashtable_alloc<>::__bucket_type): Rename into...
(_Hashtable_alloc<>::__node_base_ptr): ...this.
(_Hashtable_alloc<>::__bucket_alloc_type): Rename into...
(_Hashtable_alloc<>::__buckets_alloc_type): ...this.
(_Hashtable_alloc<>::__bucket_alloc_traits): Rename into...
(_Hashtable_alloc<>::__buckets_alloc_traits): ...this.
(_Hashtable_alloc<>::__buckets_ptr): New.
(_Hashtable_alloc<>::_M_allocate_node): Adapt.
(_Hashtable_alloc<>::_M_deallocate_node): Adapt.
(_Hashtable_alloc<>::_M_deallocate_node_ptr): Adapt.
(_Hashtable_alloc<>::_M_deallocate_nodes): Adapt.
(_Hashtable_alloc<>::_M_allocate_buckets): Adapt.
(_Hashtable_alloc<>::_M_deallocate_buckets): Adapt.
* include/bits/hashtable.h (_Hashtable<>): Adapt.
Nathan Sidwell [Tue, 20 Oct 2020 19:17:21 +0000 (12:17 -0700)]
c++: block-scope extern decl with default args
In adding the DECL_LOCAL_DECL handling, I'd forgotten that the
parm-decls also need cloning -- and resetting of their DECL_CONTEXT.
Also, any default args need droping when adding an alias, as those are
not propagated. The std's not totally clear on this latter point when
there's no exising namespace decl, but that seems like the right thing
and is what clang does.
gcc/cp/
* name-lookup.c (push_local_extern_decl_alias): Reconstextualize
alias' parm decls. Drop any default args.
gcc/testsuite/
* g++.dg/lookup/local-extern.C: New.
Ian Lance Taylor [Tue, 20 Oct 2020 18:47:35 +0000 (11:47 -0700)]
libbacktrace: use __attribute__((__fallthrough__))
Use an attribute rather than a comment when falling through a switch case.
* internal.h (ATTRIBUTE_FALLTHROUGH): Define.
* elf.c (elf_zlib_inflate): Use ATTRIBUTE_FALLTHROUGH.
Aldy Hernandez [Mon, 19 Oct 2020 15:03:44 +0000 (17:03 +0200)]
Convert -Wrestrict pass to ranger.
There is one adjustment to a C++ test which now gives a false positive.
After talking with Martin Sebor, we've concluded this is expected. There
is no way to communicate that libstdc++ allocated objects are always
less than PTRDIFF_MAX.
gcc/ChangeLog:
* calls.c (get_size_range): Adjust to work with ranger.
* calls.h (get_size_range): Add ranger argument to prototype.
* gimple-ssa-warn-restrict.c (class wrestrict_dom_walker): Remove.
(check_call): Pull out of wrestrict_dom_walker into a
static function.
(wrestrict_dom_walker::before_dom_children): Rename to...
(wrestrict_walk): ...this.
(pass_wrestrict::execute): Instantiate ranger.
(class builtin_memref): Add stmt and query fields.
(builtin_access::builtin_access): Add range_query field.
(builtin_memref::builtin_memref): Same.
(builtin_memref::extend_offset_range): Same.
(builtin_access::builtin_access): Make work with ranger.
(wrestrict_dom_walker::check_call): Pull out into...
(check_call): ...here.
(check_bounds_or_overlap): Add range_query argument.
* gimple-ssa-warn-restrict.h (check_bounds_or_overlap):
Add range_query and gimple stmt arguments.
gcc/testsuite/ChangeLog:
* gcc.dg/Wrestrict-22.c: New test.
* g++.dg/torture/pr92421.C: Adjust for ranger.
libstdc++-v3/ChangeLog:
* testsuite/21_strings/basic_string/capacity/1.cc: Pass
-Wno-stringop-overflow to test.
Aldy Hernandez [Mon, 19 Oct 2020 14:52:51 +0000 (16:52 +0200)]
Convert -Walloca pass to ranger.
gcc/ChangeLog:
* gimple-ssa-warn-alloca.c (enum alloca_type): Remove
ALLOCA_BOUND_UNKNOWN and ALLOCA_CAST_FROM_SIGNED.
(warn_limit_specified_p): New.
(alloca_call_type_by_arg): Remove.
(cast_from_signed_p): Remove.
(is_max): Remove.
(alloca_call_type): Remove heuristics and replace with call into
ranger.
(pass_walloca::execute): Instantiate ranger.
gcc/testsuite/ChangeLog:
* gcc.dg/Walloca-1.c: Adjust for ranger.
* gcc.dg/Walloca-12.c: Same.
* gcc.dg/Walloca-13.c: Same.
* gcc.dg/Walloca-2.c: Same.
* gcc.dg/Walloca-3.c: Same.
* gcc.dg/Walloca-6.c: Same.
* gcc.dg/Wvla-larger-than-2.c: Same.
Nikhil Benesch [Mon, 19 Oct 2020 02:19:04 +0000 (02:19 +0000)]
syscall: remove Sendfile on NetBSD
NetBSD does not support the sendfile syscall.
Reviewed-on: https://go-review.googlesource.com/c/gofrontend/+/263521
Tobias Burnus [Tue, 20 Oct 2020 17:41:44 +0000 (19:41 +0200)]
lto-wrapper: Fix -save-temps filename commit
Fix for 'Improve -save-temps filename' patch,
commit
b3032d1b84b8ab683f2d7345b6a2d9f783fe946d
gcc/ChangeLog
* lto-wrapper.c (run_gcc): Use proper variable for
%u.ltrans_args dump suffix.
gcc/testsuite/ChangeLog
* gcc.misc-tests/outputs.exp: Add ltrans_args dump files
for 'lto save-temps'.
zhengnannan [Tue, 20 Oct 2020 16:53:04 +0000 (17:53 +0100)]
AArch64: Add FLAG for get/set reg intrinsics [PR94442]
2020-10-20 Zhiheng Xie <xiezhiheng@huawei.com>
Nannan Zheng <zhengnannan@huawei.com>
gcc/ChangeLog:
* config/aarch64/aarch64-simd-builtins.def: Add proper FLAG
for get/set reg intrinsics.
Aldy Hernandez [Fri, 9 Oct 2020 09:26:33 +0000 (11:26 +0200)]
Refactor range handling of builtins in vr_values and ranger.
This sets things up so we can share range handling of builtins between
vr_values and ranger. It is meant to refactor the code so that we can
verify that both implementations yield the same results.
First, we abstract out gimple_ranger::range_of_builtin_call into an externally
visible counterpart that can be called from vr_values. It will take a
range_query since both ranger and vr_values inherit from this base class.
Then we abstract out all the builtin handling in vr_values into a separate
method that is easier to compare against.
Finally, we call the ranger version from vr_values and compare it with the
vr_values version. Since this proves both versions return the same,
we can remove vr_values::extract_range_builtin in a follow-up patch.
The vr_values::range_of_expr change brings the vr_values version up to par
with the ranger version. It should've handled non-SSA's. This was
a small oversight that went unnoticed because the vr_value version isn't
stressed nearly as much as the ranger version. The change is needed because
the ranger code handling builtins calls, may call it for integer arguments
in range_of_builtin_ubsan_call.
There should be no change in functionality.
gcc/ChangeLog:
* gimple-range.cc (gimple_ranger::range_of_builtin_ubsan_call):
Make externally visble...
(range_of_builtin_ubsan_call): ...here. Add range_query argument.
(gimple_ranger::range_of_builtin_call): Make externally visible...
(range_of_builtin_call): ...here. Add range_query argument.
* gimple-range.h (range_of_builtin_call): Move out from class and
make externally visible.
* vr-values.c (vr_values::extract_range_basic): Abstract out
builtin handling to...
(vr_values::range_of_expr): Handle non SSAs.
(vr_values::extract_range_builtin): ...here.
* vr-values.h (class vr_values): Add extract_range_builtin.
(range_of_expr): Rename NAME to EXPR.
Aldy Hernandez [Tue, 20 Oct 2020 13:25:20 +0000 (15:25 +0200)]
Saturate overflows return from SCEV in ranger.
bounds_of_var_in_loop is returning an overflowed int, which is causing
us to create a range for which we can't compare the bounds causing
an ICE in verify_range.
Overflowed bounds cause compare_values() to return -2, which we
don't handle in verify_range.
We don't represent overflowed ranges in irange, so this patch just
saturates any overflowed end-points to MIN or MAX.
gcc/ChangeLog:
PR tree-optimization/97501
* gimple-range.cc (gimple_ranger::range_of_ssa_name_with_loop_info):
Saturate overflows returned from SCEV.
gcc/testsuite/ChangeLog:
* gcc.dg/pr97501.c: New test.
Nathan Sidwell [Tue, 20 Oct 2020 14:51:40 +0000 (07:51 -0700)]
preprocessor: Further fix for EOF in macro args [PR97471]
My previous attempt at fixing this was incorrect. The problem occurs
earlier in that _cpp_lex_direct processes the unwinding EOF needs in
collect_args mode. This patch changes it not to do that, in the same
way as directive parsing works. Also collect_args shouldn't push_back
such fake EOFs, and neither should funlike_invocation_p.
libcpp/
* lex.c (_cpp_lex_direct): Do not complete EOF processing when
parsing_args.
* macro.c (collect_args): Do not unwind fake EOF.
(funlike_invocation_p): Do not unwind fake EOF.
(cpp_context): Replace abort with gcc_assert.
gcc/testsuite/
* gcc.dg/cpp/endif.c: Move to ...
* c-c++-common/cpp/endif.c: ... here.
* gcc.dg/cpp/endif.h: Move to ...
* c-c++-common/cpp/endif.h: ... here.
* c-c++-common/cpp/eof-2.c: Adjust diagnostic.
* c-c++-common/cpp/eof-3.c: Adjust diagnostic.
Jakub Jelinek [Tue, 20 Oct 2020 14:38:24 +0000 (16:38 +0200)]
libgomp: Fix up bootstrap in libgomp/target.c due to false positive warning
> On 10/20/20 2:11 PM, Tobias Burnus wrote:
>
> > Unfortunately, the committed patch
> > (r11-4121-g1bfc07d150790fae93184a79a7cce897655cb37b)
> > causes build errors.
> >
> > The error seems to be provoked by function cloning – as the code
> > itself looks fine:
> > ...
> > struct gomp_device_descr *devices_s
> > = malloc (num_devices * sizeof (struct gomp_device_descr));
> > ...
> > for (i = 0; i < num_devices; i++)
> > if (!(devices[i].capabilities & GOMP_OFFLOAD_CAP_OPENMP_400))
> > devices_s[num_devices_after_openmp++] = devices[i];
>
> gomp_target_init.part.0 ()
> {
> ...
> <bb 2>
> devices_s_1 = malloc (0);
> ...
> num_devices.16_67 = num_devices;
> ...
> if (num_devices.16_67 > 0)
> goto <bb 3>; [89.00%]
> else
> goto <bb 18>; [11.00%]
>
> Which seems to have an ordering problem.
This patch fixes the warning that breaks the bootstrap.
2020-10-20 Jakub Jelinek <jakub@redhat.com>
* target.c (gomp_target_init): Inside of the function, use automatic
variables corresponding to num_devices, num_devices_openmp and devices
global variables and update the globals only at the end of the
function.
Marek Polacek [Tue, 20 Oct 2020 14:15:41 +0000 (10:15 -0400)]
c++: Add fixed test [PR82239]
This test was fixed by r256550 but that commit was fixing another issue,
and just happened to fix this too.
gcc/testsuite/ChangeLog:
PR c++/82239
* g++.dg/cpp0x/static_assert16.C: New test.
Aldy Hernandez [Tue, 20 Oct 2020 07:17:12 +0000 (09:17 +0200)]
Simplify and split irange::copy_legacy_range into two functions.
copy_legacy_range was a small but complex function. It was tricky
to understand, and easy to introduce bugs into it. It also did
unnecessary work on some code paths.
This patch splits the function into two functions that are more
efficient and easier to read (copy_to_legacy and
copy_legacy_to_multi_range).
gcc/ChangeLog:
* value-range.cc (irange::operator=): Split up call to
copy_legacy_range into...
(irange::copy_to_legacy): ...this.
(irange::copy_legacy_to_multi_range): ...and this.
(irange::copy_legacy_range): Remove.
* value-range.h: Remove copoy_legacy_range.
Add copy_legacy_to_multi_range and copy_to_legacy.
Tobias Burnus [Tue, 20 Oct 2020 13:32:58 +0000 (15:32 +0200)]
gcc/doc/invoke.texi: Fix @table syntax
gcc/
* doc/invoke.texi (NVPTX options): Use @item not @itemx.
Richard Biener [Tue, 20 Oct 2020 10:52:31 +0000 (12:52 +0200)]
Avoid changing PHIs in GIMPLE split_edge
Previously I've changed gimple_split_edge to avoid PHI node
re-allocation, but this introduced swapping of PHI arguments
due to the way edge redirection works. This is now a problem
for me and which can be solved with the following approach
reducing the overhead of split_edge even more. We can simply
pretend there are no PHI nodes if we can make sure the
new fallthru will have the same dest_idx as the old edge
into the destination.
2020-10-20 Richard Biener <rguenther@suse.de>
* tree-cfg.c (reinstall_phi_args): Remove.
(gimple_split_edge): Remove PHIs around the edge redirection
to avoid touching them at all.
Richard Biener [Tue, 20 Oct 2020 11:15:45 +0000 (13:15 +0200)]
Fix latch PHI arg lookup in vectorizable_reduction for double-reduction
We were using the wrong loop to figure the latch arg of a
double-reduction PHI. Which isn't a problem in case ->dest_idx
match up with the outer loop edges - but that's of course not guaranteed.
2020-10-20 Richard Biener <rguenther@suse.de>
* tree-vect-loop.c (vectorizable_reduction): Use the correct
loops latch edge for the PHI arg lookup.
Kwok Cheung Yeung [Tue, 20 Oct 2020 11:15:59 +0000 (04:15 -0700)]
openmp: Implement support for OMP_TARGET_OFFLOAD environment variable
This implements support for the OMP_TARGET_OFFLOAD environment variable
introduced in the OpenMP 5.0 standard, which controls how offloading
is handled. It may be set to MANDATORY (abort if offloading cannot be
performed), DISABLED (no offloading to devices) or DEFAULT (offload to
device if possible, fall back to host if not).
2020-10-20 Kwok Cheung Yeung <kcy@codesourcery.com>
libgomp/
* env.c (gomp_target_offload_var): New.
(parse_target_offload): New.
(handle_omp_display_env): Print value of OMP_TARGET_OFFLOAD.
(initialize_env): Parse OMP_TARGET_OFFLOAD.
* libgomp.h (gomp_target_offload_t): New.
(gomp_target_offload_var): New.
* libgomp.texi (OMP_TARGET_OFFLOAD): New section.
* target.c (resolve_device): Generate error if device not found and
offloading is mandatory.
(gomp_target_fallback): Generate error if offloading is mandatory.
(GOMP_target): Add argument in call to gomp_target_fallback.
(GOMP_target_ext): Likewise.
(gomp_target_data_fallback): Generate error if offloading is mandatory.
(GOMP_target_data): Add argument in call to gomp_target_data_fallback.
(GOMP_target_data_ext): Likewise.
(gomp_target_task_fn): Add argument in call to gomp_target_fallback.
(gomp_target_init): Return early if offloading is disabled.
Jonathan Wakely [Tue, 20 Oct 2020 10:19:58 +0000 (11:19 +0100)]
libstdc++: Define noop coroutine details private and inline [PR 95917]
This moves the __noop_coro_frame type, the __noop_coro_fr global
variable, and the __dummy_resume_destroy function from namespace scope,
replacing them with private members of the specialization
coroutine_handle<noop_coroutine_promise>.
The function and variable are also declared inline, so that they
generate no code unless used.
libstdc++-v3/ChangeLog:
PR libstdc++/95917
* include/std/coroutine (__noop_coro_frame): Replace with
noop_coroutine_handle::__frame.
(__dummy_resume_destroy): Define inline in __frame.
(__noop_coro_fr): Replace with noop_coroutine_handle::_S_fr
and define as inline.
* testsuite/18_support/coroutines/95917.cc: New test.
Jonathan Wakely [Tue, 20 Oct 2020 10:18:35 +0000 (11:18 +0100)]
libstdc++: Remove inheritance from std::coroutine_handle<> [LWG 3460]
This removes the coroutine_handle<> base class from the primary template
and the noop_coroutine_promise explicit specialization. To preserve the
API various members are added, as they are no longer inherited from the
base class.
I've also tweaked some indentation and formatting, and replaced
subclause numbers from the standard with stable names like
[coroutine.handle.con].
libstdc++-v3/ChangeLog:
* include/std/coroutine (coroutine_handle<_Promise>): Remove
base class. Add constructors, conversions, accessors etc. as
proposed for LWG 3460.
(coroutine_handle<noop_coroutine_promise>): Likewise.
* testsuite/18_support/coroutines/lwg3460.cc: New test.
Jozef Lawrynowicz [Tue, 20 Oct 2020 10:26:20 +0000 (11:26 +0100)]
MSP430: Support a memory operand for op1 of andneghi3
This fixes an ICE caused by an unrecognizeable insn generated when
compiling gcc.c-torture/execute/pr97386-1.c at -O0.
gcc/ChangeLog:
* config/msp430/msp430.md (andneghi3): Allow general operand for
op1 and update output assembler template.
Tobias Burnus [Tue, 20 Oct 2020 10:14:03 +0000 (12:14 +0200)]
collect-utils.c, lto-wrapper + mkoffload: Improve -save-temps filename
gcc/ChangeLog:
* collect-utils.c (collect_execute, fork_execute): Add at-file suffix
argument.
* collect-utils.h (collect_execute, fork_execute): Update prototype.
* collect2.c (maybe_run_lto_and_relink, do_link, main, do_dsymutil):
Update calls by passing NULL.
* config/i386/intelmic-mkoffload.c (compile_for_target,
generate_host_descr_file, prepare_target_image, main): Likewise.
* config/gcn/mkoffload.c (compile_native, main): Pass at-file suffix.
* config/nvptx/mkoffload.c (compile_native, main): Likewise.
* lto-wrapper.c (compile_offload_image): Likewise.