]> git.ipfire.org Git - thirdparty/gcc.git/commitdiff
Daily bump.
authorGCC Administrator <gccadmin@gcc.gnu.org>
Wed, 17 Sep 2025 00:20:31 +0000 (00:20 +0000)
committerGCC Administrator <gccadmin@gcc.gnu.org>
Wed, 17 Sep 2025 00:20:31 +0000 (00:20 +0000)
gcc/ChangeLog
gcc/DATESTAMP
gcc/ada/ChangeLog
gcc/c/ChangeLog
gcc/testsuite/ChangeLog
libstdc++-v3/ChangeLog

index 62b2a29e0e8aa99984ab65a21a4e465fef30aac9..b72586f5f09a97ffe78b81d7818ac432b60fafab 100644 (file)
@@ -1,3 +1,64 @@
+2025-09-16  Jakub Jelinek  <jakub@redhat.com>
+
+       * doc/invoke.texi (Wimplicit-fallthrough=): Document that also C23
+       provides a standard way to suppress the warning with [[fallthrough]];.
+
+2025-09-16  Takayuki 'January June' Suwa  <jjsuwa_sys3175@yahoo.co.jp>
+
+       * config/xtensa/xtensa.cc (#include):
+       Add "addresses.h".
+       * config/xtensa/xtensa.h (REGNO_OK_FOR_BASE_P):
+       Simplify to just a call to GP_REG_P().
+       (BASE_REG_P): Replace REGNO_OK_FOR_BASE_P() with the equivalent
+       call to regno_ok_for_base_p().
+
+2025-09-16  Wilco Dijkstra  <wilco.dijkstra@arm.com>
+
+       PR middle-end/66462
+       * config/aarch64/aarch64.md (isnan<mode>2): Add new expander.
+
+2025-09-16  Richard Biener  <rguenther@suse.de>
+
+       * tree-vect-stmts.cc (get_load_store_type): Get in a flag
+       whether a SLP_TREE_LOAD_PERMUTATION on the node can be
+       code generated and use it.  Fix the condition on using
+       strided gather/scatter to avoid dropping a meaningful
+       permutation.
+       (vectorizable_store): Adjust.
+       (vectorizable_load): Analyze the permutation early and
+       pass the result down to get_load_store_type.  Fix the
+       condition on when we are allowed to elide a load permutation.
+
+2025-09-16  Jennifer Schmitz  <jschmitz@nvidia.com>
+
+       PR target/121602
+       * config/aarch64/aarch64-sve-builtins.cc
+       (gimple_folder::fold_active_lanes_to): Add force_vector
+       statement.
+
+2025-09-16  Jiawei  <jiawei@iscas.ac.cn>
+
+       * config.gcc: Accept RISC-V profiles in `--with-arch`.
+       * config/riscv/arch-canonicalize: Add profile detection and
+       skip canonicalization for profiles.
+
+2025-09-16  Jiawei  <jiawei@iscas.ac.cn>
+
+       * common/config/riscv/riscv-common.cc (struct riscv_profiles): Add comments.
+       (RISCV_PROFILE): Removed.
+       * config/riscv/riscv-profiles.def: New file.
+
+2025-09-16  Dongyan Chen  <chendongyan@isrc.iscas.ac.cn>
+
+       * config/riscv/riscv-ext.def: Imply zicsr.
+
+2025-09-16  liuhongt  <hongtao.liu@intel.com>
+
+       * config/i386/predicates.md (avx_vbroadcast128_operand): New
+       predicate.
+       * config/i386/sse.md (*avx_vbroadcastf128_<mode>_perm): New
+       pre_reload splitter.
+
 2025-09-15  Alexandre Oliva  <oliva@adacore.com>
 
        * config/rs6000/vxworks.h (TARGET_CMODEL, SET_CMODEL): Define.
index fed526bc14b6477f898f710841df9fc539fe8011..a9ffd35574ba5d8c298e45b6f774441da4e2592d 100644 (file)
@@ -1 +1 @@
-20250916
+20250917
index 59b890b0b4846532e539b4fcdc32447b956a0642..379816ed6be356a74ccaa5dbb4b89aac9a7a0696 100644 (file)
@@ -1,3 +1,42 @@
+2025-09-16  Ronan Desplanques  <desplanques@adacore.com>
+
+       * exp_strm.ads (Get_Primitives): New function.
+       * exp_strm.adb (Get_Primitives): Likewise.
+       (Build_Elementary_Input_Call, Build_Elementary_Write_Call): use
+       Get_Primitives.
+       (Has_Stream_Standard_Rep): Add formal parameter and rename to...
+       (Is_Stream_Standard_Rep): New function.
+       * sem_ch13.adb (Analyze_Attribute_Definition_Clause): Fix error
+       emission.
+
+2025-09-16  Gary Dismukes  <dismukes@adacore.com>
+
+       Revert:
+       2025-09-15  Gary Dismukes  <dismukes@adacore.com>
+
+       * exp_ch6.ads (Make_Build_In_Place_Call_In_Allocator): Simplify comment.
+       * exp_ch6.adb (Make_Build_In_Place_Call_In_Allocator): Remove obsolete
+       comment about not being able to allocate fixed-size controlled results
+       on the caller side, and replace another obsolete comment with a simpler
+       comment. Call Build_Allocate_Deallocate_Proc when the function doesn't
+       need a BIPalloc formal to ensure that function results with controlled
+       parts allocated on the caller side will be chained for finalization.
+       (Make_Build_In_Place_Call_In_Object_Declaration): Call Needs_BIP_Collection
+       on the function's Entity_Id rather than the function call.
+       (Needs_BIP_Collection): If a BIP function doesn't need a BIPalloc formal
+       then it doesn't need a BIP collection either; return False in that case.
+       (Needs_BIP_Alloc_Form): Remove test of Needs_BIP_Collection.
+       * exp_ch7.adb (Expand_Cleanup_Actions): Move test of Uses_Sec_Stack
+       to be the first conjunct in setting of Needs_Sec_Stack_Mark, and put
+       the other tests in a disjunction subsidiary to that. Improve preceding
+       comment.
+
+2025-09-16  Iain Sandoe  <iain@sandoe.co.uk>
+
+       PR ada/114065
+       * libgnarl/s-osinte__darwin.adb: Add and reference clause
+       for Interfaces.C, remove clause for Interfaces.C.Extensions.
+
 2025-09-15  Eric Botcazou  <ebotcazou@adacore.com>
 
        PR ada/114065
index 5c6d709bbae65301d4f6b875b26f9b240dbd25a7..8c681302fffc23a37ad56cc847bb738bda4a8993 100644 (file)
@@ -1,3 +1,9 @@
+2025-09-16  Andrew Pinski  <andrew.pinski@oss.qualcomm.com>
+
+       PR c/121421
+       * c-parser.cc (c_parser_declaration_or_fndef): Error out for gimple
+       and rtl functions as nested functions.
+
 2025-09-13  Joseph Myers  <josmyers@redhat.com>
 
        * c-typeck.cc (c_mark_addressable): New parameter
index 42dd4f474ee5db1042c884dbf34ea784359e0045..f73eb8946464dfc48dc478470c6436640b68896c 100644 (file)
@@ -1,3 +1,61 @@
+2025-09-16  Andrew Pinski  <andrew.pinski@oss.qualcomm.com>
+
+       PR c/121421
+       * gcc.dg/gimplefe-error-16.c: New test.
+
+2025-09-16  Wilco Dijkstra  <wilco.dijkstra@arm.com>
+
+       PR middle-end/66462
+       * gcc.target/aarch64/pr66462.c: Update test.
+
+2025-09-16  Jennifer Schmitz  <jschmitz@nvidia.com>
+
+       PR target/121602
+       * gcc.target/aarch64/sve/acle/asm/mul_s16.c: New test.
+       * gcc.target/aarch64/sve/acle/asm/mul_s32.c: Likewise.
+       * gcc.target/aarch64/sve/acle/asm/mul_s64.c: Likewise.
+       * gcc.target/aarch64/sve/acle/asm/mul_s8.c: Likewise.
+       * gcc.target/aarch64/sve/acle/asm/mul_u16.c: Likewise.
+       * gcc.target/aarch64/sve/acle/asm/mul_u32.c: Likewise.
+       * gcc.target/aarch64/sve/acle/asm/mul_u64.c: Likewise.
+       * gcc.target/aarch64/sve/acle/asm/mul_u8.c: Likewise.
+
+2025-09-16  Haochen Jiang  <haochen.jiang@intel.com>
+
+       * gcc.target/i386/avx512bw-vpackssdw-1.c:
+       Fix xmm/ymm mask tests.
+       * gcc.target/i386/avx512bw-vpacksswb-1.c: Ditto.
+       * gcc.target/i386/avx512bw-vpackusdw-1.c: Ditto.
+       * gcc.target/i386/avx512bw-vpackuswb-1.c: Ditto.
+       * gcc.target/i386/avx512bw-vpermw-1.c: Test xmm.
+       * gcc.target/i386/avx512bw-vpmulhw-1.c:
+       Fix xmm/ymm mask tests.
+       * gcc.target/i386/avx512f-vec-init.c: Remove duplicate test.
+       * gcc.target/i386/avx512fp16-13.c: Fix test for aligned load.
+       * gcc.target/i386/avx512fp16-conjugation-1.c: Revise the test
+       to test more precisely on masks.
+       * gcc.target/i386/avx512fp16vl-conjugation-1.c: Ditto.
+       * gcc.target/i386/avx512vbmi-vpermb-1.c: Test xmm.
+       * gcc.target/i386/avx512vl-vcvtpd2ps-1.c: Fix scan asm.
+       * gcc.target/i386/avx512vl-vinsert-1.c: Fix typo.
+       * gcc.target/i386/avx512vl-vpmulld-1.c:
+       Fix xmm/ymm mask tests.
+       * gcc.target/i386/avx512vl-vptestmd-1.c: Ditto.
+       * gcc.target/i386/bitwise_mask_op-1.c: Fix typo.
+       * gcc.target/i386/cond_op_shift_q-1.c: Test both vpsra{,v}
+       and vpsll{,v}.
+       * gcc.target/i386/cond_op_shift_ud-1.c: Ditto.
+       * gcc.target/i386/cond_op_shift_uq-1.c: Ditto.
+       * gcc.target/i386/memcpy-pr95886.c: Fix the wrong const int.
+       * gcc.target/i386/part-vect-sqrtph-1.c: Remove duplicate test.
+       * gcc.target/i386/pr107432-7.c: Test vpmov{s,z}xbw instead of
+       vpmov{s,z}xbd.
+       * gcc.target/i386/pr88828-0.c: Fix pblendw scan asm.
+
+2025-09-16  liuhongt  <hongtao.liu@intel.com>
+
+       * gcc.target/i386/avx_vbroadcastf128.c: New test.
+
 2025-09-15  ClĂ©ment Chigot  <chigot@adacore.com>
 
        * gcc.target/aarch64/pr113356.C: Move to ...
index 7bf2f4f272ffb6dd5e9ae02655ddb6595b43b037..3bce6ff862b031236351f0cae5ebaad120fadb43 100644 (file)
@@ -1,3 +1,45 @@
+2025-09-16  Jonathan Wakely  <jwakely@redhat.com>
+
+       * include/std/ranges (__detail::__repeated_tuple): Use
+       __make_tuple helper alias directly, instead of doing overload
+       resolution on std::tuple_cat.
+       * include/std/tuple (__make_tuple_impl): Remove.
+       (__do_make_tuple): Replace recursion with _Index_tuple and pack
+       expansion.
+       (__make_tuple): Adjust to new __do_make_tuple definition.
+       (__combine_tuples<tuple<T1s...>, tuple<T2s...>, Rem...>): Replace
+       with a partial specialization for exactly two tuples and a
+       partial specialization for three or more tuples.
+
+2025-09-16  Jonathan Wakely  <jwakely@redhat.com>
+
+       PR libstdc++/121913
+       * include/bits/ranges_algo.h (__rotate_fn::operator()): Use
+       auto_value_t<_Iter> instead of deduced type.
+
+2025-09-16  Jonathan Wakely  <jwakely@redhat.com>
+
+       PR libstdc++/121956
+       * include/std/ranges (adjacent_view::_Iterator::value_type):
+       Always define as std::tuple<T, N>, not std::pair<T, T>.
+       * testsuite/std/ranges/adaptors/adjacent/1.cc: Check value type
+       of views::pairwise.
+
+2025-09-16  Jonathan Wakely  <jwakely@redhat.com>
+
+       * include/bits/stl_uninitialized.h (uninitialized_move)
+       (uninitialized_move_n): Replace _GLIBCXX_MAKE_MOVE_ITERATOR with
+       std::make_move_iterator.
+
+2025-09-16  Jonathan Wakely  <jwakely@redhat.com>
+
+       PR libstdc++/119820
+       * include/bits/ranges_algo.h (__shuffle_fn): Use
+       ranges::distance to get difference type value to add to
+       iterator.
+       * include/std/format (__formatter_str::_M_format_range):
+       Use ranges::next to increment iterator by a size_t value.
+
 2025-09-13  Patrick Palka  <ppalka@redhat.com>
 
        PR libstdc++/121917