Daily bump.
authorGCC Administrator <gccadmin@gcc.gnu.org>
Fri, 14 Oct 2022 00:16:35 +0000 (00:16 +0000)
committerGCC Administrator <gccadmin@gcc.gnu.org>
Fri, 14 Oct 2022 00:16:35 +0000 (00:16 +0000)
gcc/ChangeLog
gcc/DATESTAMP
gcc/analyzer/ChangeLog
gcc/c-family/ChangeLog
gcc/cp/ChangeLog
gcc/testsuite/ChangeLog
include/ChangeLog
libgomp/ChangeLog
libitm/ChangeLog

index 7d9d899..93fd9ef 100644 (file)
@@ -1,3 +1,149 @@
+2022-10-13  Eric Botcazou  <ebotcazou@adacore.com>
+
+       * gimple-ssa-warn-access.cc (pass_waccess::check_call): Return
+       early for calls made from thunks.
+
+2022-10-13  Eric Botcazou  <ebotcazou@adacore.com>
+
+       * expr.cc (emit_group_stote): Fix handling of modes of different
+       sizes for big-endian targets in latest change and add commentary.
+
+2022-10-13  Martin Liska  <mliska@suse.cz>
+
+       * output.h (assemble_vtv_preinit_initializer): Remove.
+       * varasm.cc (assemble_vtv_preinit_initializer): Remove.
+
+2022-10-13  Andrew MacLeod  <amacleod@redhat.com>
+
+       PR tree-optimization/102540
+       PR tree-optimization/102872
+       * gimple-range-cache.cc (ranger_cache::fill_block_cache):
+       Handle partial equivs.
+       (ranger_cache::range_from_dom): Cleanup dump output.
+
+2022-10-13  Andrew MacLeod  <amacleod@redhat.com>
+
+       * range-op.cc (operator_cast::lhs_op1_relation): New.
+       (operator_bitwise_and::lhs_op1_relation): New.
+
+2022-10-13  Andrew MacLeod  <amacleod@redhat.com>
+
+       * gimple-range-cache.cc (ranger_cache::fill_block_cache): Use
+       iterator.
+       * value-relation.cc
+         (equiv_relation_iterator::equiv_relation_iterator): New.
+       (equiv_relation_iterator::next): New.
+       (equiv_relation_iterator::get_name): New.
+       * value-relation.h (class relation_oracle): Privatize some methods.
+       (class equiv_relation_iterator): New.
+       (FOR_EACH_EQUIVALENCE): New.
+       (FOR_EACH_PARTIAL_EQUIV): New.
+       (FOR_EACH_PARTIAL_AND_FULL_EQUIV): New.
+
+2022-10-13  Andrew MacLeod  <amacleod@redhat.com>
+
+       * value-relation.cc (equiv_chain::dump): Don't print empty
+       equivalences.
+       (equiv_oracle::equiv_oracle): Allocate a partial equiv table.
+       (equiv_oracle::~equiv_oracle): Release the partial equiv table.
+       (equiv_oracle::add_partial_equiv): New.
+       (equiv_oracle::partial_equiv_set): New.
+       (equiv_oracle::partial_equiv): New.
+       (equiv_oracle::query_relation): Check for partial equivs too.
+       (equiv_oracle::dump): Also dump partial equivs.
+       (dom_oracle::register_relation): Handle partial equivs.
+       (dom_oracle::query_relation): Check for partial equivs.
+       * value-relation.h (enum relation_kind_t): Add partial equivs.
+       (relation_partial_equiv_p): New.
+       (relation_equiv_p): New.
+       (class pe_slice): New.
+       (class equiv_oracle): Add prototypes.
+       (pe_to_bits): New.
+       (bits_to_pe): New.
+       (pe_min): New.
+
+2022-10-13  Richard Biener  <rguenther@suse.de>
+
+       PR tree-optimization/107247
+       * tree-vect-loop.cc (vect_create_epilog_for_reduction):
+       Reduce multi vector SLP reduction accumulators.  Check
+       the adjusted number of accumulator vectors against
+       one for the re-use in the epilogue.
+
+2022-10-13  Jakub Jelinek  <jakub@redhat.com>
+
+       * genmodes.cc (emit_mode_wider): Emit previous content of
+       mode_wider array into mode_next array and for mode_wider
+       emit always VOIDmode for !CLASS_HAS_WIDER_MODES_P classes,
+       otherwise skip through modes with the same precision.
+       * machmode.h (mode_next): Declare.
+       (GET_MODE_NEXT_MODE): New inline function.
+       (mode_iterator::get_next, mode_iterator::get_known_next): New
+       function templates.
+       (FOR_EACH_MODE_IN_CLASS): Use get_next instead of get_wider.
+       (FOR_EACH_MODE): Use get_known_next instead of get_known_wider.
+       (FOR_EACH_MODE_FROM): Use get_next instead of get_wider.
+       (FOR_EACH_WIDER_MODE_FROM): Define.
+       (FOR_EACH_NEXT_MODE): Define.
+       * expmed.cc (emit_store_flag_1): Use FOR_EACH_WIDER_MODE_FROM
+       instead of FOR_EACH_MODE_FROM.
+       * optabs.cc (prepare_cmp_insn): Likewise.  Remove redundant
+       !CLASS_HAS_WIDER_MODES_P check.
+       (prepare_float_lib_cmp): Use FOR_EACH_WIDER_MODE_FROM instead of
+       FOR_EACH_MODE_FROM.
+       * config/i386/i386-expand.cc (get_mode_wider_vector): Use
+       GET_MODE_NEXT_MODE instead of GET_MODE_WIDER_MODE.
+
+2022-10-13  Wilco Dijkstra  <wdijkstr@arm.com>
+
+       PR target/105773
+       * config/aarch64/aarch64.cc (aarch64_select_cc_mode): Allow
+       GT/LE for merging compare with zero into AND.
+       (aarch64_get_condition_code_1): Add CC_NZVmode support.
+       * config/aarch64/aarch64-modes.def: Add CC_NZV.
+       * config/aarch64/aarch64.md: Use CC_NZV in cmp+and patterns.
+
+2022-10-13  Richard Biener  <rguenther@suse.de>
+
+       PR tree-optimization/107160
+       * tree-vect-loop.cc (vect_create_epilog_for_reduction):
+       Do not register accumulator if we failed to reduce it
+       to a single vector.
+
+2022-10-13  Aldy Hernandez  <aldyh@redhat.com>
+
+       * range-op-float.cc (range_operator_float::op1_op2_relation): New.
+       (class foperator_equal): Add using.
+       (class foperator_not_equal): Same.
+       (class foperator_lt): Same.
+       (class foperator_le): Same.
+       (class foperator_gt): Same.
+       (class foperator_ge): Same.
+       * range-op.cc (range_op_handler::op1_op2_relation): New.
+       * range-op.h (range_operator_float::op1_op2_relation): New.
+
+2022-10-13  Richard Biener  <rguenther@suse.de>
+
+       * genmatch.cc (parser::parse_c_expr): Diagnose 'return'.
+       * match.pd: Replace 'return' statements in with expressions
+       with appropriate variants.
+
+2022-10-13  Andre Vieira  <andre.simoesdiasvieira@arm.com>
+
+       PR tree-optimization/107229
+       * tree-if-conv.cc (get_bitfield_rep): Fix bitposition calculation.
+
+2022-10-13  Andre Vieira  <andre.simoesdiasvieira@arm.com>
+
+       PR tree-optimization/107226
+       * tree-vect-patterns.cc (vect_recog_bitfield_ref_pattern): Reject
+       BITFIELD_REF's with non integral typed first operands.
+
+2022-10-13  Joseph Myers  <joseph@codesourcery.com>
+
+       * ginclude/float.h (FLT_IS_IEC_60559, DBL_IS_IEC_60559)
+       (LDBL_IS_IEC_60559): Update comment.
+
 2022-10-12  Aldy Hernandez  <aldyh@redhat.com>
 
        * range-op-float.cc (class foperator_negate): New.
index 0c497e6..66947c0 100644 (file)
@@ -1 +1 @@
-20221013
+20221014
index 33477da..f753d1a 100644 (file)
@@ -1,3 +1,9 @@
+2022-10-13  David Malcolm  <dmalcolm@redhat.com>
+
+       PR analyzer/107210
+       * svalue.cc (constant_svalue::maybe_fold_bits_within): Only
+       attempt to extract individual bits when tree_fits_uhwi_p.
+
 2022-10-07  David Malcolm  <dmalcolm@redhat.com>
 
        PR analyzer/105783
index 1fade0a..df4dcf2 100644 (file)
@@ -1,3 +1,8 @@
+2022-10-13  Joseph Myers  <joseph@codesourcery.com>
+
+       * c-cppbuiltin.cc (builtin_define_float_constants): Do not use
+       value 2 for *_IS_IEC_60559.
+
 2022-10-12  Lewis Hyatt  <lhyatt@gmail.com>
 
        PR preprocessor/60014
index 320b465..24bb33f 100644 (file)
@@ -1,3 +1,23 @@
+2022-10-13  Jason Merrill  <jason@redhat.com>
+
+       * cp-tree.h: Fix whitespace.
+       * parser.h: Fix whitespace.
+       * decl.cc: Fix whitespace.
+       * parser.cc: Fix whitespace.
+       * pt.cc: Fix whitespace.
+
+2022-10-13  Martin Liska  <mliska@suse.cz>
+
+       * vtable-class-hierarchy.cc (vtv_generate_init_routine): Emit
+       an artificial variable that would be put into .preinit_array
+       section.
+
+2022-10-13  Marek Polacek  <polacek@redhat.com>
+
+       PR c++/106925
+       * except.cc (maybe_splice_retval_cleanup): Check current_function_decl.
+       Make the bool const.
+
 2022-10-12  Marek Polacek  <polacek@redhat.com>
 
        * call.cc (standard_conversion): Remove LOOKUP_PREFER_RVALUE code.
index 737ee21..a985478 100644 (file)
@@ -1,3 +1,45 @@
+2022-10-13  Eric Botcazou  <ebotcazou@adacore.com>
+
+       * gcc.dg/Wstringop-overflow-89.c: New test.
+
+2022-10-13  David Malcolm  <dmalcolm@redhat.com>
+
+       PR analyzer/107210
+       * gfortran.dg/analyzer/pr107210.f90: New test.
+
+2022-10-13  Andrew MacLeod  <amacleod@redhat.com>
+
+       * gcc.dg/pr102540.c: New.
+       * gcc.dg/pr102872.c: New.
+
+2022-10-13  Marek Polacek  <polacek@redhat.com>
+
+       PR c++/106925
+       * g++.dg/cpp0x/initlist-defarg3.C: New test.
+
+2022-10-13  Wilco Dijkstra  <wdijkstr@arm.com>
+
+       PR target/105773
+       * gcc.target/aarch64/ands_2.c: Test for ANDS.
+       * gcc.target/aarch64/bics_2.c: Test for BICS.
+       * gcc.target/aarch64/tst_2.c: Test for TST.
+       * gcc.target/aarch64/tst_imm_split_1.c: Fix test.
+
+2022-10-13  Richard Biener  <rguenther@suse.de>
+
+       PR tree-optimization/107160
+       * gcc.dg/vect/pr107160.c: New testcase.
+
+2022-10-13  Andre Vieira  <andre.simoesdiasvieira@arm.com>
+
+       * gcc.dg/vect/pr107229-1.c: New test.
+       * gcc.dg/vect/pr107229-2.c: New test.
+       * gcc.dg/vect/pr107229-3.c: New test.
+
+2022-10-13  Joseph Myers  <joseph@codesourcery.com>
+
+       * gcc.dg/c2x-float-10.c: Do not expect value 2 for *_IS_IEC_60559.
+
 2022-10-12  Lewis Hyatt  <lhyatt@gmail.com>
 
        PR preprocessor/60014
index e9fb5b3..2cec065 100644 (file)
@@ -1,3 +1,9 @@
+2022-10-13  Xi Ruoyao  <xry111@xry111.site>
+
+       * longlong.h [__loongarch__] (count_leading_zeros): Define.
+       [__loongarch__] (count_trailing_zeros): Likewise.
+       [__loongarch__] (COUNT_LEADING_ZEROS_0): Likewise.
+
 2022-09-27  Jakub Jelinek  <jakub@redhat.com>
 
        PR c++/106652
index d06c05e..a339a7d 100644 (file)
@@ -1,3 +1,13 @@
+2022-10-13  Tobias Burnus  <tobias@codesourcery.com>
+
+       * testsuite/libgomp.fortran/task-7.f90: New test.
+       * testsuite/libgomp.fortran/task-8.f90: New test.
+       * testsuite/libgomp.fortran/task-in-explicit-1.f90: New test.
+       * testsuite/libgomp.fortran/task-in-explicit-2.f90: New test.
+       * testsuite/libgomp.fortran/task-in-explicit-3.f90: New test.
+       * testsuite/libgomp.fortran/task-reduction-17.f90: New test.
+       * testsuite/libgomp.fortran/task-reduction-18.f90: New test.
+
 2022-10-12  Jakub Jelinek  <jakub@redhat.com>
 
        * libgomp.texi (OpenMP 5.2): Fix up allocator -> allocate directive
index a7243a9..552d29b 100644 (file)
@@ -1,3 +1,8 @@
+2022-10-13  Lulu Cheng  <chenglulu@loongson.cn>
+
+       * config/loongarch/sjlj.S: Add a soft floating point condition to the
+       register recovery part of the code.
+
 2022-10-12  Martin Liska  <mliska@suse.cz>
 
        * configure: Regenerate.