|
| 1 | +2025-10-22 H.J. Lu <hjl.tools@gmail.com> |
| 2 | + |
| 3 | + * config/i386/i386-expand.cc (ix86_expand_set_or_cpymem): Use |
| 4 | + HOST_WIDE_INT_0U and HOST_WIDE_INT_M1U to initialize unsigned |
| 5 | + HOST_WIDE_INT. |
| 6 | + |
| 7 | +2025-10-22 Tamar Christina <tamar.christina@arm.com> |
| 8 | + |
| 9 | + * config/aarch64/aarch64-simd.md (reduc_sbool_and_scal_<mode>, |
| 10 | + reduc_sbool_ior_scal_<mode>, reduc_sbool_xor_scal_<mode>): Use SVE if |
| 11 | + available. |
| 12 | + * config/aarch64/aarch64-sve.md (*cmp<cmp_op><mode>_ptest): Rename ... |
| 13 | + (@aarch64_pred_cmp<cmp_op><mode>_ptest): ... To this. |
| 14 | + (reduc_sbool_xor_scal_<mode>): Rename ... |
| 15 | + (@reduc_sbool_xor_scal_<mode>): ... To this. |
| 16 | + |
| 17 | +2025-10-22 Tamar Christina <tamar.christina@arm.com> |
| 18 | + |
| 19 | + * config/aarch64/aarch64-simd.md (reduc_sbool_and_scal_<mode>, |
| 20 | + reduc_sbool_ior_scal_<mode>, reduc_sbool_xor_scal_<mode>): New. |
| 21 | + * config/aarch64/iterators.md (VALLI): New. |
| 22 | + |
| 23 | +2025-10-22 Tamar Christina <tamar.christina@arm.com> |
| 24 | + |
| 25 | + * config/aarch64/aarch64-sve.md (reduc_sbool_and_scal_<mode>, |
| 26 | + reduc_sbool_ior_scal_<mode>, reduc_sbool_xor_scal_<mode>): New. |
| 27 | + |
| 28 | +2025-10-22 Tamar Christina <tamar.christina@arm.com> |
| 29 | + |
| 30 | + * tree-vect-loop.cc (vectorizable_reduction): Don't always require |
| 31 | + IFN_VEC_SHL_INSERT when using reduc sbool optabs. |
| 32 | + |
| 33 | +2025-10-22 Srinath Parvathaneni <srinath.parvathaneni@arm.com> |
| 34 | + |
| 35 | + * config/aarch64/aarch64.opt.urls: Regenerate. |
| 36 | + |
| 37 | +2025-10-22 Richard Biener <rguenther@suse.de> |
| 38 | + |
| 39 | + PR tree-optimization/122364 |
| 40 | + * tree-vect-slp.cc (vect_analyze_slp_reduc_chain): Re-try |
| 41 | + linearization on a conversion source. |
| 42 | + |
| 43 | +2025-10-22 Richard Biener <rguenther@suse.de> |
| 44 | + |
| 45 | + PR tree-optimization/122370 |
| 46 | + * tree-vect-loop.cc (vect_create_epilog_for_reduction): |
| 47 | + Also update compute_vectype when demoting masks to an |
| 48 | + integer vector. |
| 49 | + |
| 50 | +2025-10-22 Richard Biener <rguenther@suse.de> |
| 51 | + |
| 52 | + PR tree-optimization/122371 |
| 53 | + * tree-vect-loop.cc (vectorize_fold_left_reduction): Get |
| 54 | + to the scalar def to replace via the scalar PHI backedge def. |
| 55 | + * tree-vect-slp.cc (vect_analyze_slp_reduc_chain): Do not |
| 56 | + re-associate to for a reduction chain if a fold-left |
| 57 | + reduction is required. |
| 58 | + |
| 59 | +2025-10-22 Richard Biener <rguenther@suse.de> |
| 60 | + |
| 61 | + PR tree-optimization/122365 |
| 62 | + * tree-vect-loop.cc (vect_create_epilog_for_reduction): |
| 63 | + Convert all inputs. Use the proper vector element sizes |
| 64 | + for the elementwise reduction. |
| 65 | + |
| 66 | +2025-10-22 Haochen Jiang <haochen.jiang@intel.com> |
| 67 | + |
| 68 | + * common/config/i386/cpuinfo.h |
| 69 | + (get_intel_cpu): Handle Nova Lake. |
| 70 | + * common/config/i386/i386-common.cc (processor_name): |
| 71 | + Add Nova Lake. |
| 72 | + (processor_alias_table): Ditto. |
| 73 | + * common/config/i386/i386-cpuinfo.h (enum processor_types): |
| 74 | + Add INTEL_COREI7_NOVALAKE. |
| 75 | + * config.gcc: Add -march=novalake. |
| 76 | + * config/i386/driver-i386.cc (host_detect_local_cpu): Handle |
| 77 | + novalake. |
| 78 | + * config/i386/i386-c.cc (ix86_target_macros_internal): Ditto. |
| 79 | + * config/i386/i386-options.cc (processor_cost_table): Ditto. |
| 80 | + (m_NOVALAKE): New. |
| 81 | + (m_CORE_HYBRID): Add novalake. |
| 82 | + * config/i386/i386.h (enum processor_type): Ditto. |
| 83 | + * doc/extend.texi: Ditto. |
| 84 | + * doc/invoke.texi: Ditto. |
| 85 | + |
| 86 | +2025-10-22 Haochen Jiang <haochen.jiang@intel.com> |
| 87 | + |
| 88 | + * config/i386/driver-i386.cc (host_detect_local_cpu): Correct |
| 89 | + the logic for unknown model number cpu guess value. |
| 90 | + |
| 91 | +2025-10-22 liuhongt <hongtao.liu@intel.com> |
| 92 | + |
| 93 | + PR target/122320 |
| 94 | + * config/i386/sse.md (*<avx512>_cmp<mode>3_dup_op): New define_insn_and_split. |
| 95 | + |
| 96 | +2025-10-22 Antoni Boucher <bouanto@zoho.com> |
| 97 | + |
| 98 | + * config/i386/i386-jit.cc: Mark new float types as supported. |
| 99 | + |
1 | 100 | 2025-10-21 Antoni Boucher <bouanto@zoho.com> |
2 | 101 |
|
3 | 102 | * config.gcc (jit_target_objs): Don't set this variable since |
|
0 commit comments