Age | Commit message (Expand) | Author | Files | Lines |
33 hours | i386: Fix the order of operands in andn<MMXMODEI:mode>3 [PR117192] | Uros Bizjak | 1 | -3/+3 |
3 days | Support andn_optab for x86 | Cui, Lili | 1 | -0/+7 |
2024-09-30 | i386: Fix dot_prod backend patterns for mmx and sse targets | Victor Do Nascimento | 1 | -15/+15 |
2024-09-26 | x86: Extend AVX512 Vectorization for Popcount in Various Modes | Levy Hsu | 1 | -0/+24 |
2024-09-12 | i386: Implement SAT_ADD for signed vector integers | Uros Bizjak | 1 | -1/+1 |
2024-09-12 | i386: Enable V2BF/V4BF vec_cmp with AVX10.2 vcmppbf16 | Levy Hsu | 1 | -0/+17 |
2024-09-05 | i386: Support partial vectorized FMA for V2BF/V4BF | Levy Hsu | 1 | -0/+80 |
2024-09-05 | i386: Support partial signbit/xorsign/copysign/abs/neg/and/xor/ior/andn for V... | Levy Hsu | 1 | -35/+39 |
2024-09-03 | i386: Support partial vectorized V2BF/V4BF smaxmin | Levy Hsu | 1 | -0/+19 |
2024-09-03 | i386: Support partial vectorized V2BF/V4BF plus/minus/mult/div/sqrt | Levy Hsu | 1 | -0/+37 |
2024-08-21 | Align predicates for operands[1] between mov<mode> and *mov<mode>_internal. | liuhongt | 1 | -1/+1 |
2024-07-01 | Remove vcond{,u,eq}<mode> expanders since they will be obsolete. | liuhongt | 1 | -97/+0 |
2024-07-01 | Optimize a < 0 ? -1 : 0 to (signed)a >> 31. | liuhongt | 1 | -0/+52 |
2024-06-27 | vect: support direct conversion under x86-64-v3. | Hu, Lin1 | 1 | -2/+33 |
2024-06-27 | vect: Support v4hi -> v4qi. | Hu, Lin1 | 1 | -4/+13 |
2024-05-17 | i386: Rename sat_plusminus expanders to standard names [PR112600] | Uros Bizjak | 1 | -0/+7 |
2024-05-07 | Support dot_prod optabs for 64-bit vector. | liuhongt | 1 | -0/+195 |
2024-03-06 | i386: Fix and improve insn constraint for V2QI arithmetic/shift insns | Uros Bizjak | 1 | -10/+23 |
2024-02-27 | i386: psrlq is not used for PERM<a,{0},1,2,3,4> [PR113871] | Uros Bizjak | 1 | -2/+2 |
2024-02-14 | i386: psrlq is not used for PERM<a,{0},1,2,3,4> [PR113871] | Uros Bizjak | 1 | -0/+69 |
2024-01-03 | Update copyright years. | Jakub Jelinek | 1 | -1/+1 |
2023-12-19 | i386: Fix mmx.md signbit expanders [PR112816] | Jakub Jelinek | 1 | -2/+8 |
2023-12-11 | Support vpcmov for V4HF/V4BF/V2HF/V2BF under TARGET_XOP. | liuhongt | 1 | -0/+22 |
2023-11-21 | Support reduc_{and,ior,xor}_scal_m for V4HI/V8QI/V4QImode | liuhongt | 1 | -2/+29 |
2023-11-16 | Fix ICE of unrecognizable insn. | liuhongt | 1 | -4/+4 |
2023-11-16 | [i386] APX: Fix EGPR usage in several patterns. | Hongyu Wang | 1 | -1/+1 |
2023-11-13 | Support vec_set/vec_extract/vec_init for V4HF/V2HF. | liuhongt | 1 | -62/+209 |
2023-11-03 | i386: Handle multiple address register classes | Uros Bizjak | 1 | -14/+17 |
2023-11-02 | Support cmul{_conj}v4hf3/cmla{_conj}v4hf4 with AVX512FP16 instruction. | liuhongt | 1 | -0/+86 |
2023-10-31 | Fix wrong code due to incorrect define_split | liuhongt | 1 | -78/+34 |
2023-10-27 | Support vec_cmpmn/vcondmn for v2hf/v4hf. | liuhongt | 1 | -2/+267 |
2023-10-23 | Remove unused mmx_pinsrw. | liuhongt | 1 | -14/+0 |
2023-10-18 | Support 32/64-bit vectorization for _Float16 fma related operations. | liuhongt | 1 | -1/+151 |
2023-10-16 | Support 32/64-bit vectorization for conversion between _Float16 and integer/f... | liuhongt | 1 | -19/+145 |
2023-10-16 | Enable vectorization for V2HF/V4HF rounding operations and sqrt. | liuhongt | 1 | -1/+190 |
2023-10-11 | Refine predicate of operands[2] in divv4hf3 with register_operand. | liuhongt | 1 | -1/+1 |
2023-10-08 | Support signbit/xorsign/copysign/abs/neg/and/xor/ior/andn for V2HF/V4HF. | liuhongt | 1 | -1/+113 |
2023-10-08 | Support smin/smax for V2HF/V4HF | liuhongt | 1 | -31/+43 |
2023-10-07 | [APX EGPR] Handle vex insns that only support GPR16 (5/5) | Kong Lingling | 1 | -5/+6 |
2023-10-07 | [APX_EGPR] Handle legacy insns that only support GPR16 (4/5) | Kong Lingling | 1 | -57/+86 |
2023-08-18 | i386: Use PUNPCKL?? to implement vector extend and zero_extend for TARGET_SSE2. | Uros Bizjak | 1 | -3/+21 |
2023-08-10 | i386: Do not sanitize upper part of V2HFmode and V4HFmode reg with -fno-trapp... | liuhongt | 1 | -6/+14 |
2023-08-08 | i386: Do not sanitize upper part of V2SFmode reg with -fno-trapping-math [PR1... | Uros Bizjak | 1 | -32/+43 |
2023-08-07 | i386: Clear upper bits of XMM register for V4HFmode/V2HFmode operations [PR11... | liuhongt | 1 | -19/+90 |
2023-08-07 | x86: drop redundant "prefix_data16" attributes | Jan Beulich | 1 | -10/+0 |
2023-08-07 | x86: correct "length_immediate" in a few cases | Jan Beulich | 1 | -2/+2 |
2023-08-07 | x86: replace/correct bogus "prefix_extra" | Jan Beulich | 1 | -4/+4 |
2023-08-07 | x86: "sse4arg" adjustments | Jan Beulich | 1 | -20/+6 |
2023-07-26 | i386: Clear upper half of XMM register for V2SFmode operations [PR110762] | Uros Bizjak | 1 | -403/+475 |
2023-06-19 | Reimplement packuswb/packusdw with UNSPEC_US_TRUNCATE instead of original us_... | liuhongt | 1 | -15/+30 |