Age | Commit message (Expand) | Author | Files | Lines |
2021-12-03 | x86: Speed up target attribute handling by using a cache | Jakub Jelinek | 1 | -1/+25 |
2021-12-01 | i386: Improve V8HI and V8HF inserts [PR102811] | Uros Bizjak | 3 | -51/+141 |
2021-12-01 | i386: Fix up some minor formatting issues and one inconsistency | Jakub Jelinek | 1 | -4/+4 |
2021-12-01 | Fix ICE in ix86_attr_length_immediate_default. | liuhongt | 1 | -0/+4 |
2021-11-30 | Remove more stray returns and gcc_unreachable ()s | Richard Biener | 1 | -1/+0 |
2021-11-30 | Fix -freorder-blocks-and-partition glitch with Windows SEH (continued) | Eric Botcazou | 2 | -4/+6 |
2021-11-30 | x86_64: PR target/100711: Splitters for pandn | Roger Sayle | 1 | -0/+32 |
2021-11-29 | i386: Fix and improve movhi_internal and movhf_internal some more. | Uros Bizjak | 1 | -83/+128 |
2021-11-29 | x86_64: Improved V1TImode rotations by non-constant amounts. | Roger Sayle | 2 | -3/+13 |
2021-11-29 | Optimize _Float16 usage for non AVX512FP16. | liuhongt | 3 | -7/+18 |
2021-11-29 | Fix regression introduced by r12-5536. | liuhongt | 3 | -18/+29 |
2021-11-27 | x86: Fix up x86_{,64_}sh{l,r}d patterns [PR103431] | Jakub Jelinek | 1 | -42/+260 |
2021-11-26 | i386: vcvtph2ps and vcvtps2ph should be used to convert _Float16 to SFmode wi... | konglin1 | 2 | -9/+70 |
2021-11-22 | Don't allow mask/sse/mmx mov in TLS code sequences. | liuhongt | 4 | -4/+40 |
2021-11-21 | i386: Fix up handling of target attribute [PR101180] | Jakub Jelinek | 1 | -1/+4 |
2021-11-19 | Darwin: Rework handling for unwinder code in libgcc_s and specs [PR80556]. | Iain Sandoe | 3 | -57/+0 |
2021-11-19 | Reduce cost of aligned sse register store. | liuhongt | 1 | -2/+2 |
2021-11-18 | x86: Add -mindirect-branch-cs-prefix | H.J. Lu | 2 | -0/+10 |
2021-11-18 | i386: Fix wrong codegen for -mrelax-cmpxchg-loop | Hongyu Wang | 1 | -1/+6 |
2021-11-18 | i386: add alias for f*mul_*ch intrinsics | konglin1 | 2 | -0/+56 |
2021-11-17 | x86: Add -mharden-sls=[none|all|return|indirect-branch] | H.J. Lu | 3 | -7/+41 |
2021-11-17 | x86: Remove "%!" before ret | H.J. Lu | 2 | -3/+3 |
2021-11-17 | i386: Redefine indirect_thunks_used as HARD_REG_SET. | Uros Bizjak | 1 | -11/+6 |
2021-11-17 | i386: Introduce LEGACY_SSE_REGNO_P predicate | Uros Bizjak | 2 | -12/+12 |
2021-11-17 | i386: Fix non-robust split condition in define_insn_and_split | Kewen Lin | 1 | -10/+10 |
2021-11-16 | x86_64: Avoid rorx rotation instructions with -Os. | Roger Sayle | 1 | -6/+16 |
2021-11-15 | PR target/103069: Relax cmpxchg loop for x86 target | Hongyu Wang | 5 | -1/+205 |
2021-11-15 | i386: Fix up x86 atomic_bit_test* expanders for !TARGET_HIMODE_MATH [PR103205] | Jakub Jelinek | 1 | -3/+3 |
2021-11-11 | x86: Update -mtune=alderlake | Cui,Lili | 5 | -30/+155 |
2021-11-11 | Extend vpcmov to handle V8HF/V16HFmode under TARGET_XOP. | liuhongt | 1 | -4/+4 |
2021-11-10 | i386: Support complex fma/conj_fma for _Float16. | konglin1 | 1 | -0/+23 |
2021-11-10 | i386: Optimization for mm512_set1_pch. | konglin1 | 1 | -0/+62 |
2021-11-08 | Fix couple of issues in large PIC model on x86-64/VxWorks | Eric Botcazou | 2 | -5/+17 |
2021-11-05 | x86: Make stringop_algs::stringop_strategy ctor constexpr [PR100246] | Jakub Jelinek | 1 | -2/+3 |
2021-11-04 | vect: Convert cost hooks to classes | Richard Sandiford | 1 | -53/+23 |
2021-11-04 | Extend vternlog define_insn_and_split to memory_operand to enable more optimz... | liuhongt | 2 | -15/+32 |
2021-11-04 | i386: Auto vectorize sdot_prod, usdot_prod with VNNI instruction. | Hongyu Wang | 1 | -8/+56 |
2021-11-04 | i386: Fix wrong result for AMX-TILE intrinsic when parsing expression. | Hongyu Wang | 1 | -3/+3 |
2021-11-02 | x86_64: Improved implementation of TImode rotations. | Roger Sayle | 1 | -2/+18 |
2021-11-02 | ia32: Disallow mode(V1TI) [PR103020] | Jakub Jelinek | 1 | -0/+4 |
2021-11-02 | x86_64: Expand ashrv1ti (and PR target/102986) | Roger Sayle | 3 | -13/+543 |
2021-10-29 | Enable vectorization for _Float16 floor/ceil/trunc/nearbyint/rint operations. | liuhongt | 5 | -6/+69 |
2021-10-28 | AVX512FP16: Optimize _Float16 reciprocal for div and sqrt | Hongyu Wang | 3 | -19/+117 |
2021-10-26 | x86_64: Implement V1TI mode shifts/rotates by a constant | Roger Sayle | 3 | -0/+209 |
2021-10-25 | Combine the FADD(A, FMA(B, C, 0)) to FMA(B, C, A) and combine FADD(A, FMUL(B,... | konglin1 | 1 | -0/+52 |
2021-10-23 | config/i386: Commentary typo fix | Bernhard Reutner-Fischer | 1 | -1/+1 |
2021-10-23 | x86_64: Add insn patterns for V1TI mode logic operations. | Roger Sayle | 1 | -0/+25 |
2021-10-21 | i386: Fix wrong codegen for V8HF move without TARGET_AVX512F | Hongyu Wang | 1 | -3/+12 |
2021-10-20 | Revert "target: support spaces in target attribute." | Martin Liska | 1 | -2/+0 |
2021-10-20 | Rename asm_out_file function arguments. | Martin Liska | 2 | -12/+12 |