For INTEGER_CST operands, the code decides if it should emit the whole
INTEGER_CST into memory, or if there are enough upper bits either all 0s
or all 1s to warrant an optimization, where we use memory for lower limbs
or even just an INTEGER_CST for least significant limb and fill in the
rest of limbs with 0s or 1s. Unfortunately when not using
bitint_min_cst_precision, the code was using tree_int_cst_sgn (op) < 0
to determine whether to fill in the upper bits with 1s or 0s. That is
incorrect for TYPE_UNSIGNED INTEGER_CSTs which have higher limbs full of
ones, we really want to check here whether the most significant bit is
set or clear.
Fixed thusly.
2024-01-12 Jakub Jelinek <jakub@redhat.com>
PR tree-optimization/113334
* gimple-lower-bitint.cc (bitint_large_huge::handle_operand): Use
wi::neg_p (wi::to_wide (op)) instead of tree_int_cst_sgn (op) < 0
to determine if number should be extended by all ones rather than zero
extended.
* gcc.dg/torture/bitint-46.c: New test.
This is the case I was talking about in
https://gcc.gnu.org/pipermail/gcc-patches/2024-January/642423.html
and Zdenek kindly found a testcase for it.
We can only create BITINT_TYPE with precision at most 65535, not 65536,
so need to punt if we'd want to create it.
2024-01-12 Jakub Jelinek <jakub@redhat.com>
PR tree-optimization/113330
* tree-sra.cc (create_access): Punt for BITINT_TYPE accesses with
too large size.
* gcc.dg/bitint-69.c: New test.
The following testcase revealed a typo in condition, as the comment
says the intent is
/* If lhs of stmt is large/huge _BitInt SSA_NAME not in m_names
it means it will be handled in a loop or straight line code
at the location of its (ultimate) immediate use, so for
vop checking purposes check these only at the ultimate
immediate use. */
but the condition was using != BITINT_TYPE rather than == BITINT_TYPE,
so e.g. it used bitint_precision_kind on non-BITINT_TYPEs (e.g. on vector
types it will crash because TYPE_PRECISION means something different there,
or on say INTEGER_TYPEs the precision will never be large enough to be >=
bitint_prec_large).
The following patch fixes that.
2024-01-12 Jakub Jelinek <jakub@redhat.com>
PR tree-optimization/113323
* gimple-lower-bitint.cc (bitint_dom_walker::before_dom_children): Fix
check for lhs being large/huge _BitInt not in m_names.
* gcc.dg/bitint-68.c: New test.
The code to assign large/huge _BitInt SSA_NAMEs to partitions intentionally
ignores uninitialized SSA_NAMEs:
/* Also ignore uninitialized uses. */
if (SSA_NAME_IS_DEFAULT_DEF (s)
&& (!SSA_NAME_VAR (s) || VAR_P (SSA_NAME_VAR (s))))
continue;
because there is no need to store them into memory, all we need is when
trying to extract some limb from them use uninitialized SSA_NAME for the
limb.
The following testcase shows this is a problem for call arguments though,
for those we need to create replacement SSA_NAMEs which are loaded from
the underlying variable. For uninitialized SSA_NAMEs because we didn't
create underlying variable for them var_to_partition doesn't work, the
following patch handles it by just creating an uninitialized replacement
SSA_NAME.
2024-01-12 Jakub Jelinek <jakub@redhat.com>
PR tree-optimization/113316
* gimple-lower-bitint.cc (bitint_large_huge::lower_call): Handle
uninitialized large/huge _BitInt arguments to calls.
* gcc.dg/bitint-67.c: New test.
> > This patch is still work in progress, but posting to show failure with
> > bitint-7 test where handle_stmt called from lower_mergeable_stmt ICE's
> > because the idx (3) is out of range for the __BitInt(135) with a limb_prec
> > of 64.
>
> I can reproduce it, will debug it momentarily.
So, the problem was that in 2 spots I was comparing TYPE_SIZE of large/huge
BITINT_TYPEs to determine if it can be handled cheaply.
On x86_64 with limb_mode == abi_limb_mode (both DImode) that works fine,
if TYPE_SIZE is equal, it means it has the same number of limbs.
But on aarch64 TYPE_SIZE of say _BitInt(135) and _BitInt(193) is the same,
both are 256-bit storage, but because DImode is used as limb_mode, the
former actually needs just 3 limbs, while the latter needs 4 limbs.
And limb_access_type was asserting that we don't try to access 4th limb
on types which actually have a precision which needs just 3 limbs.
The following patch should fix that.
Note, for the info.extended targets (currently none, but I think arm 32-bit
in the ABI is meant like that), we'll need to do something different,
because the upper bits aren't just padding and should be zero/sign extended,
so if we say have limb_mode SImode, abi_limb_mode DImode, we'll need to
treat _BitInt(135) not as 5 SImode limbs, but 6. For !info.extended targets
I think treating _BitInt(135) as 3 DImode limbs rather than 4 is fine.
2024-01-12 Jakub Jelinek <jakub@redhat.com>
* gimple-lower-bitint.cc (mergeable_op): Instead of comparing
TYPE_SIZE (t) of large/huge BITINT_TYPEs, compare
CEIL (TYPE_PRECISION (t), limb_prec).
(bitint_large_huge::handle_cast): Likewise.
gcc/jit/ChangeLog:
* dummy-frontend.cc (handle_alias_attribute): New function.
(handle_always_inline_attribute): New function.
(handle_cold_attribute): New function.
(handle_fnspec_attribute): New function.
(handle_format_arg_attribute): New function.
(handle_format_attribute): New function.
(handle_noinline_attribute): New function.
(handle_target_attribute): New function.
(handle_used_attribute): New function.
(handle_visibility_attribute): New function.
(handle_weak_attribute): New function.
(handle_alias_ifunc_attribute): New function.
* jit-playback.cc (fn_attribute_to_string): New function.
(variable_attribute_to_string): New function.
(global_new_decl): Add attributes support.
(set_variable_attribute): New function.
(new_global): Add attributes support.
(new_global_initialized): Add attributes support.
(new_local): Add attributes support.
* jit-playback.h (fn_attribute_to_string): New function.
(set_variable_attribute): New function.
* jit-recording.cc (recording::lvalue::add_attribute): New function.
(recording::function::function): New function.
(recording::function::write_to_dump): Add attributes support.
(recording::function::add_attribute): New function.
(recording::function::add_string_attribute): New function.
(recording::function::add_integer_array_attribute): New function.
(recording::global::replay_into): Add attributes support.
(recording::local::replay_into): Add attributes support.
* jit-recording.h: Add attributes support.
* libgccjit.cc (gcc_jit_function_add_attribute): New function.
(gcc_jit_function_add_string_attribute): New function.
(gcc_jit_function_add_integer_array_attribute): New function.
(gcc_jit_lvalue_add_attribute): New function.
* libgccjit.h (enum gcc_jit_fn_attribute): New enum.
(gcc_jit_function_add_attribute): New function.
(gcc_jit_function_add_string_attribute): New function.
(gcc_jit_function_add_integer_array_attribute): New function.
(enum gcc_jit_variable_attribute): New function.
(gcc_jit_lvalue_add_string_attribute): New function.
* libgccjit.map: Declare new functions.
gcc/testsuite/ChangeLog:
* jit.dg/all-non-failing-tests.h: Add new attributes tests.
* jit.dg/jit.exp: Add `jit-verify-assembler-output-not` test command.
* jit.dg/test-restrict-attribute.c: New test.
* jit.dg/test-alias-attribute.c: New test.
* jit.dg/test-always_inline-attribute.c: New test.
* jit.dg/test-cold-attribute.c: New test.
* jit.dg/test-const-attribute.c: New test.
* jit.dg/test-noinline-attribute.c: New test.
* jit.dg/test-nonnull-attribute.c: New test.
* jit.dg/test-pure-attribute.c: New test.
* jit.dg/test-used-attribute.c: New test.
* jit.dg/test-variable-attribute.c: New test.
* jit.dg/test-weak-attribute.c: New test.
gcc/jit/ChangeLog:
* docs/topics/compatibility.rst: Add documentation for LIBGCCJIT_ABI_26.
* docs/topics/functions.rst: Add documentation for new functions.
* docs/topics/expressions.rst: Add documentation for new functions.
Co-authored-by: Antoni Boucher <bouanto@zoho.com>
Signed-off-by: Guillaume Gomez <guillaume1.gomez@gmail.com>
rs6000_elf_declare_function_name () outputs Power ELF V1 ABI function
entry labels without using ASM_OUTPUT_FUNCTION_LABEL (). As a result,
.LASANPC labels are not emitted, causing linker errors.
In theory, it is possible to reuse ASM_OUTPUT_FUNCTION_LABEL () by
changing rs6000_output_function_entry () to generate label names
without outputting them, but this would be quite a large change.
Instead, factor out the .LASANPC emitting code from
ASM_OUTPUT_FUNCTION_LABEL () and call it manually.
Fixes: c659dd8bfb ("Implement ASM_DECLARE_FUNCTION_NAME using ASM_OUTPUT_FUNCTION_LABEL")
Suggested-by: Jakub Jelinek <jakub@redhat.com>
Signed-off-by: Ilya Leoshkevich <iii@linux.ibm.com>
gcc/ChangeLog:
PR sanitizer/113284
* config/rs6000/rs6000.cc (rs6000_elf_declare_function_name):
Use assemble_function_label_final () for Power ELF V1 ABI.
* output.h (assemble_function_label_final): New function.
* varasm.cc (assemble_function_label_raw): Use
assemble_function_label_final ().
(assemble_function_label_final): New function.
Andre reported on IRC that this test has weird preprocessor conditions,
obviously the intent was to test whether corresponding __*_MANT_DIG__
is equal to the expected value like earlier in the function definitions,
but somehow I've ended up with a comma expression instead, which was
always true.
2024-01-12 Jakub Jelinek <jakub@redhat.com>
* gcc.dg/bitint-31.c: Fix up #if conditions checking whether
__*_MANT_DIG__ is equal to a particular precision.
I seem to have implemented this feature based on the P2918R0 revision,
not the final P2918R2 one that was approved for C++26. This commit fixes
it.
The runtime-format-string type should not have a publicly accessible
data member, so add a constructor and make it a friend of
basic_format_string. It should also be non-copyable, so that it can only
be constructed from a prvalue via temporary materialization. Change the
basic_format_string constructor parameter to pass by value. Also add
noexcept to the constructors and runtime_format generator functions.
libstdc++-v3/ChangeLog:
PR libstdc++/113320
* include/std/format (__format::_Runtime_format_string): Add
constructor and disable copy operations.
(basic_format_string(_Runtime_format_string)): Add noexcept and
take parameter by value not rvalue reference.
(runtime_format): Add noexcept.
* testsuite/std/format/runtime_format.cc: Check noexcept. Check
that construction is only possible from prvalues, not xvalues.
Reviewed-by: Daniel Krügler <daniel.kruegler@gmail.com>
This was approved for C++23 at he June 2021 virtual meeting.
This allows more efficient construction of std::pair members when {} is
used as a constructor argument. The perfect forwarding constructor can
be used, so that the member variables are constructed from forwarded
arguments instead of being copy constructed from temporaries.
libstdc++-v3/ChangeLog:
PR libstdc++/105505
* include/bits/stl_pair.h (pair::pair(U1&&, U2&&)) [C++23]: Add
default template arguments, as per P1951R1.
* testsuite/20_util/pair/cons/default_tmpl_args.cc: New test.
As discussed on IRC, the following patch uses may_alias attribute, so that
on targets like aarch64 where abi_limb_mode != limb_mode the library
accesses the limbs (half limbs of the ABI) in the arrays with conservative
alias set.
2024-01-12 Jakub Jelinek <jakub@redhat.com>
* libgcc2.h (UBILtype): New typedef with may_alias attribute.
(__mulbitint3, __divmodbitint4): Use UBILtype * instead of
UWtype * and const UBILtype * instead of const UWtype *.
* libgcc2.c (bitint_reduce_prec, bitint_mul_1, bitint_addmul_1,
__mulbitint3, bitint_negate, bitint_submul_1, __divmodbitint4):
Likewise.
* soft-fp/bitint.h (UBILtype): Change define into a typedef with
may_alias attribute.
On GENERIC tcc_comparison can have int type so restrict the PR113126
fix to vector types.
PR middle-end/113344
* match.pd ((double)float CMP (double)float -> float CMP float):
Perform result type check only for vectors.
* fold-const.cc (fold_binary_loc): Likewise.
This test should pass no matter how we adjust cost model.
Remove -fno-vect-cost-model.
Committed.
gcc/testsuite/ChangeLog:
* gcc.target/riscv/rvv/autovec/fold-min-poly.c: Remove -fno-vect-cost-model
The following adds guards avoiding code generation to
expand_perm_as_a_vlbr_vstbr_candidate when d.testing_p.
PR target/112280
* config/s390/s390.cc (expand_perm_as_a_vlbr_vstbr_candidate):
Do not generate code when d.testing_p.
Exception handling on nios2-linux-gnu with -fpic has been broken since
revision 790854ea76, "Use _dl_find_object
in _Unwind_Find_FDE". For whatever reason, this doesn't work on nios2.
Nios2 uses the GOT address as the base for DW_EH_PE_datarel
relocations in PIC; see my previous fix to make this work, revision
2d33dcfe9f, "Support for GOT-relative
DW_EH_PE_datarel encoding". So this may be a horrible bug in the ABI
or in my interpretation of it or just glibc's implementation of
_dl_find_object for this target, but there's existing code out there
that does things this way; and realistically, nobody is going to
re-engineer this now that the vendor has EOL'ed the nios2
architecture. So, just skip over the code trying to use
_dl_find_object on this target and fall back to the way that works.
I plan to backport this patch to the GCC 12 and GCC 13 branches as well.
libgcc/ChangeLog
* unwind-dw2-fde-dip.c (_Unwind_Find_FDE): Do not try to use
_dl_find_object on nios2; it doesn't work.
After r14-2692-g1c6231c05bdcca, the option is defined as EnumSet and
-fcf-protection=branch won't unset any others bits since they're in
different groups. So to override -fcf-protection, an explicit
-fcf-protection=none needs to be added and then with
-fcf-protection=XXX
gcc/ChangeLog:
PR target/113039
* doc/invoke.texi (fcf-protection=): Update documents.
gcc/ChangeLog:
* config/riscv/riscv.cc (riscv_v_ext_mode_p): Update the
comments of predicate func riscv_v_ext_mode_p.
Signed-off-by: Pan Li <pan2.li@intel.com>
The length of vfloat16m8_t ABI-name should be 17.
gcc/ChangeLog:
* config/riscv/riscv-vector-builtins.def (vfloat16m8_t):
Modify ABI-name length of vfloat16m8_t
Eliminate the redundant sign extension that exists after the conditional
move when the target register is SImode.
gcc/ChangeLog:
* config/loongarch/loongarch.cc (loongarch_expand_conditional_move):
Adjust.
gcc/testsuite/ChangeLog:
* gcc.target/loongarch/sign-extend-2.c: Adjust.
We found that the current combine optimization pass in gcc cannot handle
the following redundant sign extension situations:
(insn 77 76 78 5 (set (reg:SI 143)
(plus:SI (subreg/s/u:SI (reg/v:DI 104 [ len ]) 0)
(const_int 1 [0x1]))) {addsi3}
(expr_list:REG_DEAD (reg/v:DI 104 [ len ])
(nil)))
(insn 78 77 82 5 (set (reg/v:DI 104 [ len ])
(sign_extend:DI (reg:SI 143))) {extendsidi2}
(nil))
Because reg:SI 143 is not died or set in insn 78, no replacement merge will
be performed for the insn sequence. We adjusted the add template to eliminate
redundant sign extensions during the expand pass.
Adjusted based on upstream comments:
https://gcc.gnu.org/pipermail/gcc-patches/2024-January/641988.html
gcc/ChangeLog:
* config/loongarch/loongarch.md (add<mode>3): Removed.
(*addsi3): New.
(addsi3): Ditto.
(adddi3): Ditto.
(*addsi3_extended): Removed.
(addsi3_extended): New.
gcc/testsuite/ChangeLog:
* gcc.target/loongarch/sign-extend.c: Moved to...
* gcc.target/loongarch/sign-extend-1.c: ...here.
* gcc.target/loongarch/sign-extend-2.c: New test.
This patch adds support for parsing general lvalues ("locator list item
types") for OpenMP "map", "to" and "from" clauses to the C front-end,
similar to the previously-posted patch for C++. Such syntax is permitted
for OpenMP 5.0 and above. It was previously posted for mainline here:
https://gcc.gnu.org/pipermail/gcc-patches/2022-December/609038.html
and for the og13 branch here:
https://gcc.gnu.org/pipermail/gcc-patches/2023-June/623355.html
2024-01-11 Julian Brown <julian@codesourcery.com>
gcc/c-family/
* c-pretty-print.cc (c_pretty_printer::postfix_expression,
c_pretty_printer::expression): Add OMP_ARRAY_SECTION support.
gcc/c/
* c-parser.cc (c_parser_braced_init, c_parser_conditional_expression):
Don't allow OpenMP array section.
(c_parser_postfix_expression): Don't allow array section in statement
expression.
(c_parser_postfix_expression_after_primary): Add support for OpenMP
array section parsing.
(c_parser_expr_list): Don't allow OpenMP array section here.
(c_parser_omp_variable_list): Change ALLOW_DEREF parameter to
MAP_LVALUE. Support parsing of general lvalues in "map", "to" and
"from" clauses.
(c_parser_omp_var_list_parens): Change ALLOW_DEREF parameter to
MAP_LVALUE. Update call to c_parser_omp_variable_list.
(c_parser_oacc_data_clause): Update calls to
c_parser_omp_var_list_parens.
(c_parser_omp_clause_reduction): Use OMP_ARRAY_SECTION tree node
instead of TREE_LIST for array sections.
(c_parser_omp_target): Allow GOMP_MAP_ATTACH.
* c-tree.h (c_omp_array_section_p): Add extern declaration.
(build_omp_array_section): Add prototype.
* c-typeck.cc (c_omp_array_section_p): Add flag.
(mark_exp_read): Support OMP_ARRAY_SECTION.
(build_omp_array_section): Add function.
(build_external_ref): Tweak error path for OpenMP array sections.
(handle_omp_array_sections_1): Use OMP_ARRAY_SECTION tree code instead
of TREE_LIST. Handle more kinds of expressions.
(c_oacc_check_attachments): Use OMP_ARRAY_SECTION instead of TREE_LIST
for array sections.
(c_finish_omp_clauses): Use OMP_ARRAY_SECTION instead of TREE_LIST.
Check for supported expression types.
gcc/testsuite/
* gcc.dg/gomp/bad-array-section-c-1.c: New test.
* gcc.dg/gomp/bad-array-section-c-2.c: New test.
* gcc.dg/gomp/bad-array-section-c-3.c: New test.
* gcc.dg/gomp/bad-array-section-c-4.c: New test.
* gcc.dg/gomp/bad-array-section-c-5.c: New test.
* gcc.dg/gomp/bad-array-section-c-6.c: New test.
* gcc.dg/gomp/bad-array-section-c-7.c: New test.
* gcc.dg/gomp/bad-array-section-c-8.c: New test.
libgomp/
* libgomp.texi: C/C++ lvalues are supported now for map/to/from.
* testsuite/libgomp.c-c++-common/ind-base-4.c: New test.
* testsuite/libgomp.c-c++-common/unary-ptr-1.c: New test.
As discussed, our handling of corresponding object parameters needed to
handle the using-declaration case better. And I took the opportunity to
share code between the add_method and cand_parms_match uses.
This patch specifically doesn't compare reversed parameters, but a follow-up
patch will.
PR c++/113191
gcc/cp/ChangeLog:
* class.cc (xobj_iobj_parameters_correspond): Add context parm.
(object_parms_correspond): Factor out of...
(add_method): ...here.
* method.cc (defaulted_late_check): Use it.
* call.cc (class_of_implicit_object): New.
(object_parms_correspond): Overload taking two candidates.
(cand_parms_match): Use it.
(joust): Check reversed before comparing constraints.
* cp-tree.h (object_parms_correspond): Declare.
gcc/testsuite/ChangeLog:
* g++.dg/cpp2a/concepts-memfun4.C: New test.
The API Evolution section of the manual should mention when the
libstdc++exp.a library was added.
libstdc++-v3/ChangeLog:
* doc/xml/manual/evolution.xml: Document addition of
libstdc++exp.a.
* doc/html/*: Regenerate.
Commit f4130a3eb5 changed the type of
__expected_handler in libsupc++/unwind-cxx.h to be a
std::terminate_handler to avoid a deprecated warning. However, the
definition in eh_unex_handler.cc still used the old type
(std::unexpected_handler) and thus causes a warning when compiling
libstdc++ with -Wdeprecated-declarations (which is the default, for
example, for clang).
Adapt the definition to match the declaration.
libstdc++-v3/ChangeLog:
* libsupc++/eh_unex_handler.cc: Adjust definition type to
declaration.
Reviewed-by: Jonathan Wakely <jwakely@redhat.com>
Due to the premature split optimizations for XTheadFMemIdx, GPR
is allocated when reload allocates registers, resulting in the
following insn.
(insn 66 21 64 5 (set (reg:DF 14 a4 [orig:136 <retval> ] [136])
(mem:DF (plus:SI (reg/f:SI 15 a5 [141])
(ashift:SI (reg/v:SI 10 a0 [orig:137 i ] [137])
(const_int 3 [0x3]))) [0 S8 A64])) 218 {*movdf_hardfloat_rv32}
(nil))
Since we currently do not support adjustments to th_m_mir/th_m_miu,
which will trigger ICE. So it is recommended to place the split
optimizations after reload to ensure FPR when registers are allocated.
gcc/ChangeLog:
* config/riscv/thead.md: Add limits for splits.
gcc/testsuite/ChangeLog:
* gcc.target/riscv/xtheadfmemidx-medany.c: New test.
Now that _M_Detach do not reset iterator _M_version value we need to reset it when
the iterator is attached to a new sequence, even if this sequencer is null when
assigning a value-initialized iterator. In this case _M_version shall be resetted to 0.
libstdc++-v3/ChangeLog:
PR libstdc++/112477
* src/c++11/debug.cc
(_Safe_iterator_base::_M_attach): Reset _M_version to 0 if attaching to null
sequence.
(_Safe_iterator_base::_M_attach_single): Likewise.
(_Safe_local_iterator_base::_M_attach): Likewise.
(_Safe_local_iterator_base::_M_attach_single): Likewise.
* testsuite/23_containers/map/debug/112477.cc: New test case.
Reviewed-by: Jonathan Wakely <jwakely@redhat.com>
This simplifies the operator() of the _Pipe and _Partial range adaptor
closure objects using C++23 deducing this, allowing us to condense
multiple operator() overloads into one.
The new __like_t alias template is similar to the expositional one from
P0847R6 except it's implemented in terms of forward_like instead of vice
versa, and thus ours always yields a reference so e.g. __like_t<A, char>
is char&& instead of char. For our purposes (forwarding) this shouldn't
make a difference, I think..
libstdc++-v3/ChangeLog:
* include/bits/move.h (__like_t): Define in C++23 mode.
* include/std/ranges (views::__adaptor::Partial::operator()):
Implement using C++23 deducing this when available.
(views::__adaptor::_Pipe::operator()): Likewise.
* testsuite/std/ranges/adaptors/100577.cc: Adjust testcase to
accept new "no match for call" errors issued in C++23 mode.
* testsuite/std/ranges/adaptors/lazy_split_neg.cc: Likewise.
The problem here is after the recent vectorizer improvements, we end up
with a comparison against a vector bool 0 which then tries expand_single_bit_test
which is not expecting vector comparisons at all.
The IR was:
vector(4) <signed-boolean:1> mask_patt_5.13;
_Bool _12;
mask_patt_5.13_44 = vect_perm_even_41 != { 0.0, 1.0e+0, 2.0e+0, 3.0e+0 };
_12 = mask_patt_5.13_44 == { 0, 0, 0, 0 };
and we tried to call expand_single_bit_test for the last comparison.
Rejecting the vector comparison is needed.
Bootstrapped and tested on x86_64-linux-gnu with no regressions.
PR middle-end/113322
gcc/ChangeLog:
* expr.cc (do_store_flag): Don't try single bit tests with
comparison on vector types.
gcc/testsuite/ChangeLog:
* gcc.c-torture/compile/pr113322-1.c: New test.
Signed-off-by: Andrew Pinski <quic_apinski@quicinc.com>
Since currently ranger does not work with the complexity of COND_EXPR in
some cases so delaying the simplification of `1/x` for signed types
help code generation.
tree-ssa/divide-8.c is a new testcase where this can help.
Bootstrapped and tested on x86_64-linux-gnu with no regressions.
PR tree-optimization/113301
gcc/ChangeLog:
* match.pd (`1/x`): Delay signed case until late.
gcc/testsuite/ChangeLog:
* gcc.dg/tree-ssa/divide-8.c: New test.
Signed-off-by: Andrew Pinski <quic_apinski@quicinc.com>
libstdc++-v3/ChangeLog:
* python/libstdcxx/v6/printers.py (StdIntegralConstantPrinter):
Add printer for std::integral_constant.
* testsuite/libstdc++-prettyprinters/cxx11.cc: Test it.
As described in PR libstdc++/113258 there are old versions of tcmalloc
which replace malloc and related APIs, but do not repalce aligned_alloc
because it didn't exist at the time they were released. This means that
when operator new(size_t, align_val_t) uses aligned_alloc to obtain
memory, it comes from libc's aligned_alloc not from tcmalloc. But when
operator delete(void*, size_t, align_val_t) uses free to deallocate the
memory, that goes to tcmalloc's replacement version of free, which
doesn't know how to free it.
If we give preference to the older posix_memalign instead of
aligned_alloc then we're more likely to use a function that will be
compatible with the replacement version of free. Because posix_memalign
has been around for longer, it's more likely that old third-party malloc
replacements will also replace posix_memalign alongside malloc and free.
libstdc++-v3/ChangeLog:
PR libstdc++/113258
* libsupc++/new_opa.cc: Prefer to use posix_memalign if
available.
DejaGNU changed its header line from "Test Run By" to "Test run by"
around 2016. This patch makes it so that both alternatives are
correcly detected.
contrib/ChangeLog:
* dg-extract-results.py: Make the test_run regex case
insensitive.
Some GCC tests on m68K port of LRA is failed on `maximum number of
generated reload insns per insn achieved`. The problem is in that for
subreg reload LRA can not narrow reg class more from ALL_REGS to
GENERAL_REGS and then to data regs or address regs. The patch permits
narrowing reg class from reload insns if this results in successful
matching of reg operand. This is the second version of the patch to
fix the PR. This version adds matching with and without narrowing reg
class and preferring match without narrowing classes.
gcc/ChangeLog:
PR rtl-optimization/112918
* lra-constraints.cc (SMALL_REGISTER_CLASS_P): Move before in_class_p.
(in_class_p): Restrict condition for narrowing class in case of
allow_all_reload_class_changes_p.
(process_alt_operands): Try to match operand without and with
narrowing reg class. Discourage narrowing the class. Finish insn
matching only if there is no class narrowing.
(curr_insn_transform): Pass true to in_class_p for reg operand win.
Vectorization of bit-precision inductions isn't implemented but we
don't check this, instead we ICE during transform.
PR tree-optimization/112505
* tree-vect-loop.cc (vectorizable_induction): Reject
bit-precision induction.
* gcc.dg/vect/pr112505.c: New testcase.
The following makes sure the resulting boolean type is the same
when eliding a float extension.
PR tree-optimization/113126
* match.pd ((double)float CMP (double)float -> float CMP float):
Make sure the boolean type is the same.
* fold-const.cc (fold_binary_loc): Likewise.
* gcc.dg/torture/pr113126.c: New testcase.
The following avoids a mismatch between an early query for maximum
number of iterations of a loop and a late one when through ranger
we'd get iterations estimated. Instead make sure we compute niters
before querying the iteration bound.
PR tree-optimization/112636
* tree-ssa-loop-ch.cc (ch_base::copy_headers): Call
estimate_numbers_of_iterations before querying
get_max_loop_iterations_int.
(pass_ch::execute): Initialize SCEV and loops appropriately.
* gcc.dg/pr112636.c: New testcase.
libstdc++-v3/ChangeLog:
* src/filesystem/ops-common.h (stat_type): Use using.
Signed-off-by: Ken Matsui <kmatsui@gcc.gnu.org>
Reviewed-by: Jonathan Wakely <jwakely@redhat.com>
This patch made std::filesystem::equivalent correctly throw an exception
when either path does not exist as per [fs.op.equivalent]/4.
PR libstdc++/113250
libstdc++-v3/ChangeLog:
* src/c++17/fs_ops.cc (fs::equivalent): Use || instead of &&.
* src/filesystem/ops.cc (fs::equivalent): Likewise.
* testsuite/27_io/filesystem/operations/equivalent.cc: Handle
error codes.
* testsuite/experimental/filesystem/operations/equivalent.cc:
Likewise.
Signed-off-by: Ken Matsui <kmatsui@gcc.gnu.org>
Reviewed-by: Jonathan Wakely <jwakely@redhat.com>
LTO option streaming and target attributes both require per-function
target configuration, which is achieved via option save/restore.
We implement TARGET_OPTION_{SAVE,RESTORE} to switch the la_target
context in addition to other automatically maintained option states
(via the "Save" option property in the .opt files).
Tested on loongarch64-linux-gnu without regression.
PR target/113233
gcc/ChangeLog:
* config/loongarch/genopts/loongarch.opt.in: Mark options with
the "Save" property.
* config/loongarch/loongarch.opt: Same.
* config/loongarch/loongarch-opts.cc: Refresh -mcmodel= state
according to la_target.
* config/loongarch/loongarch.cc: Implement TARGET_OPTION_{SAVE,
RESTORE} for the la_target structure; Rename option conditions
to have the same "la_" prefix.
* config/loongarch/loongarch.h: Same.
The insert_var_expansion_initialization depends on the
HONOR_SIGNED_ZEROS to initialize the unrolling variables
to +0.0f when -0.0f and no-signed-option. Unfortunately,
we should always keep the -0.0f here because:
* The -0.0f is always the correct initial value.
* We need to support the target that always honor signed zero.
Thus, we need to leverage MODE_HAS_SIGNED_ZEROS when initialize
instead of HONOR_SIGNED_ZEROS. Then the target/backend can
decide to honor the no-signed-zero or not.
We also removed the testcase pr30957-1.c, as it makes undefined behavior
whether the return value is positive or negative.
The below tests are passed for this patch:
* The riscv regression tests.
* The aarch64 regression tests.
* The x86 bootstrap and regression tests.
gcc/ChangeLog:
* loop-unroll.cc (insert_var_expansion_initialization): Leverage
MODE_HAS_SIGNED_ZEROS for expansion variable initialization.
gcc/testsuite/ChangeLog:
* gcc.dg/pr30957-1.c: Remove.
Signed-off-by: Pan Li <pan2.li@intel.com>
