1 2020-01-06 Michael Meissner <meissner@linux.ibm.com>
3 * config/rs6000/vsx.md (vsx_extract_<mode>_var, VSX_D iterator):
4 Use 'Q' for doing vector extract from memory.
5 (vsx_extract_v4sf_var): Use 'Q' for doing vector extract from
7 (vsx_extract_<mode>_var, VSX_EXTRACT_I iterator): Use 'Q' for
8 doing vector extract from memory.
9 (vsx_extract_<mode>_<VS_scalar>mode_var): Use 'Q' for doing vector
12 * config/rs6000/rs6000.c (rs6000_adjust_vec_address): Add support
13 for the offset being 34-bits when -mcpu=future is used.
15 2020-01-06 John David Anglin <danglin@gcc.gnu.org>
17 * config/pa/pa.md: Revert change to use ordered_comparison_operator
18 instead of cmpib_comparison_operator in cmpib patterns.
19 * config/pa/predicates.md (cmpib_comparison_operator): Revert removal
20 of cmpib_comparison_operator. Revise comment.
22 2020-01-06 Richard Sandiford <richard.sandiford@arm.com>
24 * tree-vect-slp.c (vect_build_slp_tree_1): Require all shifts
25 in an IFN_DIV_POW2 node to be equal.
27 2020-01-06 Richard Sandiford <richard.sandiford@arm.com>
29 * tree-vect-stmts.c (vect_check_load_store_mask): Rename to...
30 (vect_check_scalar_mask): ...this.
31 (vectorizable_store, vectorizable_load): Update call accordingly.
32 (vectorizable_call): Use vect_check_scalar_mask to check the mask
33 argument in calls to conditional internal functions.
35 2020-01-06 Andrew Stubbs <ams@codesourcery.com>
37 * config/gcn/gcn-valu.md (subv64di3): Use separate alternatives for
39 (subv64di3_exec): Likewise.
41 2020-01-06 Bryan Stenson <bryan@siliconvortex.com>
43 * config/mips/mips.c (vr4130_align_insns): Fix typo.
44 * doc/md.texi (movstr): Likewise.
46 2020-01-06 Andrew Stubbs <ams@codesourcery.com>
48 * config/gcn/gcn-valu.md (vec_extract<mode><scalar_mode>): Add early
51 2020-01-06 Richard Sandiford <richard.sandiford@arm.com>
53 * config/aarch64/t-aarch64 ($(srcdir)/config/aarch64/aarch64-tune.md):
55 (s-aarch64-tune-md): ...this new stamp file. Pipe the new contents
56 to a temporary file and use move-if-change to update the real
59 2020-01-06 Richard Sandiford <richard.sandiford@arm.com>
61 * config/aarch64/aarch64-sve.md (@aarch64_sel_dup<mode>): Use Upl
62 rather than Upa for CPY /M.
64 2020-01-06 Andrew Stubbs <ams@codesourcery.com>
66 * config/gcn/gcn.c (gcn_inline_constant_p): Allow 64 as an inline
69 2020-01-06 Martin Liska <mliska@suse.cz>
71 PR tree-optimization/92860
72 * params.opt: Mark param_max_combine_insns with Optimization
75 2020-01-05 Jakub Jelinek <jakub@redhat.com>
78 * config/i386/i386.md (SWIDWI): New mode iterator.
79 (DWI, dwi): Add TImode variants.
80 (addv<mode>4): Use SWIDWI iterator instead of SWI. Use
81 <general_hilo_operand> instead of <general_operand>. Use
82 CONST_SCALAR_INT_P instead of CONST_INT_P.
83 (*addv<mode>4_1): Rename to ...
84 (addv<mode>4_1): ... this.
85 (QWI): New mode attribute.
86 (*addv<dwi>4_doubleword, *addv<dwi>4_doubleword_1): New
87 define_insn_and_split patterns.
88 (*addv<mode>4_overflow_1, *addv<mode>4_overflow_2): New define_insn
90 (uaddv<mode>4): Use SWIDWI iterator instead of SWI. Use
91 <general_hilo_operand> instead of <general_operand>.
92 (*addcarry<mode>_1): New define_insn.
93 (*add<dwi>3_doubleword_cc_overflow_1): New define_insn_and_split.
95 2020-01-03 Konstantin Kharlamov <Hi-Angel@yandex.ru>
97 * gdbinit.in (pr, prl, pt, pct, pgg, pgq, pgs, pge, pmz, pdd, pbs, pbm):
98 Use "call" instead of "set".
100 2020-01-03 Martin Jambor <mjambor@suse.cz>
103 * ipa-cp.c (print_all_lattices): Skip functions without info.
105 2020-01-03 Jakub Jelinek <jakub@redhat.com>
108 * config/i386/i386-options.c (ix86_simd_clone_adjust): If
109 TARGET_PREFER_AVX128, use prefer-vector-width=256 for 'c' and 'd'
110 simd clones. If TARGET_PREFER_AVX256, use prefer-vector-width=512
114 * config/i386/i386.opt (x_prefer_vector_width_type): Remove TargetSave
116 (mprefer-vector-width=): Add Save.
117 * config/i386/i386-options.c (ix86_target_string): Add PVW argument, print
118 -mprefer-vector-width= if non-zero. Fix up -mfpmath= comment.
119 (ix86_debug_options, ix86_function_specific_print): Adjust
120 ix86_target_string callers.
121 (ix86_valid_target_attribute_inner_p): Handle prefer-vector-width=.
122 (ix86_valid_target_attribute_tree): Likewise.
123 * config/i386/i386-options.h (ix86_target_string): Add PVW argument.
124 * config/i386/i386-expand.c (ix86_expand_builtin): Adjust
125 ix86_target_string caller.
128 * config/i386/i386.md (abs<mode>2): Use expand_simple_binop instead of
129 emitting ASHIFTRT, XOR and MINUS by hand. Use gen_int_mode with QImode
130 instead of gen_int_shift_amount + convert_modes.
132 PR rtl-optimization/93088
133 * loop-iv.c (find_single_def_src): Punt after looking through
134 128 reg copies for regs with single definitions. Move definitions
137 2020-01-02 Dennis Zhang <dennis.zhang@arm.com>
139 * config/arm/arm-c.c (arm_cpu_builtins): Define
140 __ARM_FEATURE_MATMUL_INT8, __ARM_FEATURE_BF16_VECTOR_ARITHMETIC,
141 __ARM_FEATURE_BF16_SCALAR_ARITHMETIC, and
142 __ARM_BF16_FORMAT_ALTERNATIVE when enabled.
143 * config/arm/arm-cpus.in (armv8_6, i8mm, bf16): New features.
144 * config/arm/arm-tables.opt: Regenerated.
145 * config/arm/arm.c (arm_option_reconfigure_globals): Initialize
146 arm_arch_i8mm and arm_arch_bf16 when enabled.
147 * config/arm/arm.h (TARGET_I8MM): New macro.
148 (TARGET_BF16_FP, TARGET_BF16_SIMD): Likewise.
149 * config/arm/t-aprofile: Add matching rules for -march=armv8.6-a.
150 * config/arm/t-arm-elf (all_v8_archs): Add armv8.6-a.
151 * config/arm/t-multilib: Add matching rules for -march=armv8.6-a.
152 (v8_6_a_simd_variants): New.
153 (v8_*_a_simd_variants): Add i8mm and bf16.
154 * doc/invoke.texi (armv8.6-a, i8mm, bf16): Document new options.
156 2020-01-02 Jakub Jelinek <jakub@redhat.com>
159 * predict.c (compute_function_frequency): Don't call
160 warn_function_cold on functions that already have cold attribute.
162 2020-01-01 John David Anglin <danglin@gcc.gnu.org>
165 * config/pa/pa.c (pa_elf_select_rtx_section): New. Put references to
166 COMDAT group function labels in .data.rel.ro.local section.
167 * config/pa/pa32-linux.h (TARGET_ASM_SELECT_RTX_SECTION): Define.
170 * config/pa/pa.md (scc): Use ordered_comparison_operator instead of
171 comparison_operator in B and S integer comparisons. Likewise, use
172 ordered_comparison_operator instead of cmpib_comparison_operator in
174 * config/pa/predicates.md (cmpib_comparison_operator): Remove.
176 2020-01-01 Jakub Jelinek <jakub@redhat.com>
178 Update copyright years.
180 * gcc.c (process_command): Update copyright notice dates.
181 * gcov-dump.c (print_version): Ditto.
182 * gcov.c (print_version): Ditto.
183 * gcov-tool.c (print_version): Ditto.
184 * gengtype.c (create_file): Ditto.
185 * doc/cpp.texi: Bump @copying's copyright year.
186 * doc/cppinternals.texi: Ditto.
187 * doc/gcc.texi: Ditto.
188 * doc/gccint.texi: Ditto.
189 * doc/gcov.texi: Ditto.
190 * doc/install.texi: Ditto.
191 * doc/invoke.texi: Ditto.
193 2020-01-01 Jan Hubicka <hubicka@ucw.cz>
195 * ipa.c (walk_polymorphic_call_targets): Fix updating of overall
198 2020-01-01 Jakub Jelinek <jakub@redhat.com>
200 PR tree-optimization/93098
201 * match.pd (popcount): For shift amounts, use integer_onep
202 or wi::to_widest () == cst instead of tree_to_uhwi () == cst
203 tests. Make sure that precision is power of two larger than or equal
204 to 16. Ensure shift is never negative. Use HOST_WIDE_INT_UC macro
205 instead of ULL suffixed constants. Formatting fixes.
207 Copyright (C) 2020 Free Software Foundation, Inc.
209 Copying and distribution of this file, with or without modification,
210 are permitted in any medium without royalty provided the copyright
211 notice and this notice are preserved.