Show patches with: Submitter = Liu, Hongtao       |    State = Action Required       |    Archived = No       |   147 patches
« 1 2 »
Patch Series rb/tb S/W/F Date Submitter Delegate State
Don't vectorize when vector stmts are only vec_contruct and stores Don't vectorize when vector stmts are only vec_contruct and stores - - --2 2023-12-04 Liu, Hongtao New
Take register pressure into account for vec_construct when the components are not loaded from memor… Take register pressure into account for vec_construct when the components are not loaded from memor… - - 3-- 2023-11-28 Liu, Hongtao New
Set AVOID_256FMA_CHAINS TO m_GENERIC as it's generally good to new platforms Set AVOID_256FMA_CHAINS TO m_GENERIC as it's generally good to new platforms - - 4-- 2023-11-22 Liu, Hongtao New
[V2] Simplify vector ((VCE (a cmp b ? -1 : 0)) < 0) ? c : d to just (VCE ((a cmp b) ? (VCE c) : (VC… [V2] Simplify vector ((VCE (a cmp b ? -1 : 0)) < 0) ? c : d to just (VCE ((a cmp b) ? (VCE c) : (VC… - - 2-- 2023-11-16 Liu, Hongtao New
Simplify vector ((VCE?(a cmp b ? -1 : 0)) < 0) ? c : d to just VCE:((a cmp b) ? (VCE c) : (VCE d)). Simplify vector ((VCE?(a cmp b ? -1 : 0)) < 0) ? c : d to just VCE:((a cmp b) ? (VCE c) : (VCE d)). - - 2-- 2023-11-10 Liu, Hongtao New
Simplify vector ((VCE?(a cmp b ? -1 : 0)) < 0) ? c : d to just (VCE:a cmp VCE:b) ? c : d. Simplify vector ((VCE?(a cmp b ? -1 : 0)) < 0) ? c : d to just (VCE:a cmp VCE:b) ? c : d. - - 2-- 2023-11-10 Liu, Hongtao New
Handle bitop with INTEGER_CST in analyze_and_compute_bitop_with_inv_effect. Handle bitop with INTEGER_CST in analyze_and_compute_bitop_with_inv_effect. - - 4-- 2023-10-30 Liu, Hongtao New
[V2,2/2] Support vec_cmpmn/vcondmn for v2hf/v4hf. [V2,1/2] Pass type of comparison operands instead of comparison result to truth_type_for in build_v… - - 4-- 2023-10-26 Liu, Hongtao New
[GCC13,backport] Avoid compile time hog on vect_peel_nonlinear_iv_init for nonlinear induction vec_… [GCC13,backport] Avoid compile time hog on vect_peel_nonlinear_iv_init for nonlinear induction vec_… - - --4 2023-10-24 Liu, Hongtao New
Support vec_cmpmn/vcondmn for v2hf/v4hf. Support vec_cmpmn/vcondmn for v2hf/v4hf. - - 4-- 2023-10-23 Liu, Hongtao New
Avoid compile time hog on vect_peel_nonlinear_iv_init for nonlinear induction vec_step_op_mul when … Avoid compile time hog on vect_peel_nonlinear_iv_init for nonlinear induction vec_step_op_mul when … - - 4-- 2023-10-19 Liu, Hongtao New
Avoid compile time hog on vect_peel_nonlinear_iv_init for nonlinear induction vec_step_op_mul when … Avoid compile time hog on vect_peel_nonlinear_iv_init for nonlinear induction vec_step_op_mul when … - - 4-- 2023-10-18 Liu, Hongtao New
[2/2] Support signbit/xorsign/copysign/abs/neg/and/xor/ior/andn for V2HF/V4HF. [1/2,x86] Support smin/smax for V2HF/V4HF - - 4-- 2023-10-08 Liu, Hongtao New
[1/2,x86] Support smin/smax for V2HF/V4HF [1/2,x86] Support smin/smax for V2HF/V4HF - - 1-3 2023-10-08 Liu, Hongtao New
Remove constraint modifier % for fcmaddcph/fcmulcph since there're not commutative. Remove constraint modifier % for fcmaddcph/fcmulcph since there're not commutative. - - 3-- 2023-09-08 Liu, Hongtao New
Adjust costing of emulated vectorized gather/scatter Adjust costing of emulated vectorized gather/scatter - - 4-- 2023-08-30 Liu, Hongtao New
Refactor vector HF/BF mode iterators and patterns. Refactor vector HF/BF mode iterators and patterns. - - 4-- 2023-08-30 Liu, Hongtao New
Mention Intel -march=gracemont for Alderlake-N. Mention Intel -march=gracemont for Alderlake-N. - - --4 2023-08-21 Liu, Hongtao New
[V2] Support -m[no-]gather -m[no-]scatter to enable/disable vectorization for all gather/scatter in… [V2] Support -m[no-]gather -m[no-]scatter to enable/disable vectorization for all gather/scatter in… - - 2-- 2023-08-11 Liu, Hongtao New
Support -m[no-]gather -m[no-]scatter to enable/disable vectorization for all gather/scatter instruc… Support -m[no-]gather -m[no-]scatter to enable/disable vectorization for all gather/scatter instruc… - - 4-- 2023-08-10 Liu, Hongtao New
Rename local variable subleaf_level to max_subleaf_level. Rename local variable subleaf_level to max_subleaf_level. - - -4- 2023-08-09 Liu, Hongtao New
[V2,X86] Workaround possible CPUID bug in Sandy Bridge. [V2,X86] Workaround possible CPUID bug in Sandy Bridge. - - -4- 2023-08-09 Liu, Hongtao New
[X86] Workaround possible CPUID bug in Sandy Bridge. [X86] Workaround possible CPUID bug in Sandy Bridge. - - 4-- 2023-08-08 Liu, Hongtao New
Optimize vlddqu + inserti128 to vbroadcasti128 Optimize vlddqu + inserti128 to vbroadcasti128 - - 3-- 2023-08-02 Liu, Hongtao New
Adjust testcase for more optimal codegen. Adjust testcase for more optimal codegen. - - -3- 2023-08-01 Liu, Hongtao New
Optimize vlddqu to vmovdqu for TARGET_AVX Optimize vlddqu to vmovdqu for TARGET_AVX - - 4-- 2023-07-20 Liu, Hongtao New
Add peephole to eliminate redundant comparison after cmpccxadd. Add peephole to eliminate redundant comparison after cmpccxadd. - - 4-- 2023-07-11 Liu, Hongtao New
Break false dependence for vpternlog by inserting vpxor or setting constraint of input operand to '… Break false dependence for vpternlog by inserting vpxor or setting constraint of input operand to '… - - 4-- 2023-07-10 Liu, Hongtao New
[V2,x86] Add pre_reload splitter to detect fp min/max pattern. [V2,x86] Add pre_reload splitter to detect fp min/max pattern. - - 4-- 2023-07-07 Liu, Hongtao New
[2/2] Adjust rtx_cost for DF/SFmode AND/IOR/XOR/ANDN operations. [1/2,x86] Add pre_reload splitter to detect fp min/max pattern. - - 4-- 2023-07-06 Liu, Hongtao New
[1/2,x86] Add pre_reload splitter to detect fp min/max pattern. [1/2,x86] Add pre_reload splitter to detect fp min/max pattern. - - -3- 2023-07-06 Liu, Hongtao New
Break false dependence for vpternlog by inserting vpxor. Break false dependence for vpternlog by inserting vpxor. - - 4-- 2023-07-04 Liu, Hongtao New
[3/3,aarch64] Adjust testcase to match assembly output after r14-2007. [1/3] Use cvt_op to save intermediate type operand instead of "subtle" vec_dest. - - 3-- 2023-06-26 Liu, Hongtao New
[2/2] Refined 256/512-bit vpacksswb/vpackssdw patterns. [1/2] Reimplement packuswb/packusdw with UNSPEC_US_TRUNCATE instead of original us_truncate. - - 2-- 2023-06-16 Liu, Hongtao New
[1/2] Reimplement packuswb/packusdw with UNSPEC_US_TRUNCATE instead of original us_truncate. [1/2] Reimplement packuswb/packusdw with UNSPEC_US_TRUNCATE instead of original us_truncate. - - --- 2023-06-16 Liu, Hongtao New
[1/2] Fold _mm{, 256, 512}_abs_{epi8, epi16, epi32, epi64} into gimple ABSU_EXPR + VCE. [1/2] Fold _mm{, 256, 512}_abs_{epi8, epi16, epi32, epi64} into gimple ABSU_EXPR + VCE. - - --- 2023-06-06 Liu, Hongtao New
Don't fold _mm{, 256}_blendv_epi8 into (mask < 0 ? src1 : src2) when -funsigned-char. Don't fold _mm{, 256}_blendv_epi8 into (mask < 0 ? src1 : src2) when -funsigned-char. - - 4-- 2023-06-06 Liu, Hongtao New
Fold _mm{, 256, 512}_abs_{epi8, epi16, epi32, epi64} into gimple ABSU_EXPR + VCE. Fold _mm{, 256, 512}_abs_{epi8, epi16, epi32, epi64} into gimple ABSU_EXPR + VCE. - - 4-- 2023-06-06 Liu, Hongtao New
[x86] Add missing vec_pack/unpacks patterns for _Float16 <-> int/float conversion. [x86] Add missing vec_pack/unpacks patterns for _Float16 <-> int/float conversion. - - 4-- 2023-06-05 Liu, Hongtao New
[vect] Use intermiediate integer type for float_expr/fix_trunc_expr when direct optab is not existe… [vect] Use intermiediate integer type for float_expr/fix_trunc_expr when direct optab is not existe… - - 1-- 2023-06-02 Liu, Hongtao New
i386: Add missing vector truncate patterns [PR92658]. i386: Add missing vector truncate patterns [PR92658]. - - --- 2023-06-02 Liu, Hongtao New
Provide -fcf-protection=branch,return. Provide -fcf-protection=branch,return. - - --- 2023-05-12 Liu, Hongtao New
x86: Add a new option -mdaz-ftz to enable FTZ and DAZ flags in MXCSR. x86: Add a new option -mdaz-ftz to enable FTZ and DAZ flags in MXCSR. - - --- 2023-05-10 Liu, Hongtao New
[powerpc] Add a peephole2 to eliminate redundant move from VSX_REGS to GENERAL_REGS when it's from … [powerpc] Add a peephole2 to eliminate redundant move from VSX_REGS to GENERAL_REGS when it's from … - - --- 2023-05-04 Liu, Hongtao New
[vect] Enhance NARROW FLOAT_EXPR vectorization by truncating integer to lower precision. [vect] Enhance NARROW FLOAT_EXPR vectorization by truncating integer to lower precision. - - --- 2023-04-26 Liu, Hongtao New
[2/2,i386] def_or_undef __STDCPP_FLOAT16_T__ and __STDCPP_BFLOAT16_T__ for target attribute/pragmas. [1/2,i386] Support type _Float16/__bf16 independent of SSE2. - - --- 2023-04-21 Liu, Hongtao New
[1/2,i386] Support type _Float16/__bf16 independent of SSE2. [1/2,i386] Support type _Float16/__bf16 independent of SSE2. - - --- 2023-04-21 Liu, Hongtao New
Canonicalize vec_merge when mask is constant. Canonicalize vec_merge when mask is constant. - - --- 2023-04-20 Liu, Hongtao New
[2/2] Adjust testcases after better RA decision. [1/2] Use NO_REGS in cost calculation when the preferred register class are not known yet. - - --- 2023-04-20 Liu, Hongtao New
[1/2] Use NO_REGS in cost calculation when the preferred register class are not known yet. [1/2] Use NO_REGS in cost calculation when the preferred register class are not known yet. - - --- 2023-04-20 Liu, Hongtao New
[i386] Support type _Float16/__bf16 independent of SSE2. [i386] Support type _Float16/__bf16 independent of SSE2. - - --- 2023-04-19 Liu, Hongtao New
Support vector conversion for AVX512 vcvtudq2pd/vcvttps2udq/vcvttpd2udq. Support vector conversion for AVX512 vcvtudq2pd/vcvttps2udq/vcvttpd2udq. - - --- 2023-03-30 Liu, Hongtao New
Change AVX512FP16 to AVX512-FP16 which is official name. Change AVX512FP16 to AVX512-FP16 which is official name. - - --- 2023-01-29 Liu, Hongtao New
[x86] x86: Don't add crtfastmath.o for -shared and add a new option -mdaz-ftz to enable FTZ and DAZ… [x86] x86: Don't add crtfastmath.o for -shared and add a new option -mdaz-ftz to enable FTZ and DAZ… - - --- 2022-12-14 Liu, Hongtao New
[v2,x86] Fix incorrect _mm_cvtsbh_ss. [v2,x86] Fix incorrect _mm_cvtsbh_ss. - - --- 2022-11-24 Liu, Hongtao New
[x86] Fix incorrect implementation for mm_cvtsbh_ss. [x86] Fix incorrect implementation for mm_cvtsbh_ss. - - --- 2022-11-23 Liu, Hongtao New
[x86] Some tidy up for RA related hooks. [x86] Some tidy up for RA related hooks. - - --- 2022-11-21 Liu, Hongtao New
[1/2] Implement hwasan target_hook. Support HWASAN with Intel LAM - - --- 2022-11-11 Liu, Hongtao New
[V2,x86] Fix incorrect digit constraint [V2,x86] Fix incorrect digit constraint - - --- 2022-10-31 Liu, Hongtao New
[x86] Fix incorrect digit constraint [x86] Fix incorrect digit constraint - - --- 2022-10-27 Liu, Hongtao New
Canonicalize vec_perm index to make the first index come from the first vector. Canonicalize vec_perm index to make the first index come from the first vector. - - --- 2022-10-18 Liu, Hongtao New
[x86] Add define_insn_and_split to support general version of "kxnor". [x86] Add define_insn_and_split to support general version of "kxnor". - - --- 2022-10-11 Liu, Hongtao New
Check nonlinear iv in vect_can_advance_ivs_p. Check nonlinear iv in vect_can_advance_ivs_p. - - --- 2022-09-29 Liu, Hongtao New
[x86] Support 2-instruction vector shuffle for V4SI/V4SF in ix86_expand_vec_perm_const_1. [x86] Support 2-instruction vector shuffle for V4SI/V4SF in ix86_expand_vec_perm_const_1. - - --- 2022-09-23 Liu, Hongtao New
Fix incorrect handle in vectorizable_induction for mixed induction type. Fix incorrect handle in vectorizable_induction for mixed induction type. - - --- 2022-09-20 Liu, Hongtao New
[x86] Don't optimize cmp mem, 0 to load mem, reg + test reg, reg [x86] Don't optimize cmp mem, 0 to load mem, reg + test reg, reg - - --- 2022-09-16 Liu, Hongtao New
Modernize ix86_builtin_vectorized_function with corresponding expanders. Modernize ix86_builtin_vectorized_function with corresponding expanders. - - --- 2022-09-16 Liu, Hongtao New
[V2] Extend vectorizer to handle nonlinear induction for neg, mul/lshift/rshift with a constant. [V2] Extend vectorizer to handle nonlinear induction for neg, mul/lshift/rshift with a constant. - - --- 2022-08-29 Liu, Hongtao New
[RFC:] Extend vectorizer to handle nonlinear induction for neg, mul/lshift/rshift with a constant. [RFC:] Extend vectorizer to handle nonlinear induction for neg, mul/lshift/rshift with a constant. - - --- 2022-08-04 Liu, Hongtao New
Move pass_cse_sincos after vectorizer. Move pass_cse_sincos after vectorizer. - - --- 2022-07-20 Liu, Hongtao New
[V2] Extend 16/32-bit vector bit_op patterns with (m, 0, i) alternative. [V2] Extend 16/32-bit vector bit_op patterns with (m, 0, i) alternative. - - --- 2022-07-19 Liu, Hongtao New
[V2,RFC] Support vectorization for Complex type. [V2,RFC] Support vectorization for Complex type. - - --- 2022-07-18 Liu, Hongtao New
Extend 16/32-bit vector bit_op patterns with (m, 0, i)(vertical) alternative. Extend 16/32-bit vector bit_op patterns with (m, 0, i)(vertical) alternative. - - --- 2022-07-18 Liu, Hongtao New
Extend 64-bit vector bit_op patterns with ?r alternative Extend 64-bit vector bit_op patterns with ?r alternative - - --- 2022-07-14 Liu, Hongtao New
[RFC] Support vectorization for Complex type. [RFC] Support vectorization for Complex type. - - --- 2022-07-11 Liu, Hongtao New
Allocate general register(memory/immediate) for 16/32/64-bit vector bit_op patterns. Allocate general register(memory/immediate) for 16/32/64-bit vector bit_op patterns. - - --- 2022-07-11 Liu, Hongtao New
Improve reg_or_subregno to return INVALID_REGNUM when the subreg of memory is processed. Improve reg_or_subregno to return INVALID_REGNUM when the subreg of memory is processed. - - --- 2022-06-23 Liu, Hongtao New
[x86] Replace REGNO with reg_or_subregno in pre_reload splitter. [x86] Replace REGNO with reg_or_subregno in pre_reload splitter. - - --- 2022-06-20 Liu, Hongtao New
Add optional __Bfloat16 support Add optional __Bfloat16 support - - --- 2022-06-10 Liu, Hongtao New
Simplify (B * v + C) * D -> BD* v + CD when B, C, D are all INTEGER_CST. Simplify (B * v + C) * D -> BD* v + CD when B, C, D are all INTEGER_CST. - - --- 2022-06-07 Liu, Hongtao New
Simplify (B * v + C) * D -> BD* v + CD when B, C, D are all INTEGER_CST. Simplify (B * v + C) * D -> BD* v + CD when B, C, D are all INTEGER_CST. - - --- 2022-06-02 Liu, Hongtao New
Add a bit dislike for separate mem alternative when op is REG_P. Add a bit dislike for separate mem alternative when op is REG_P. - - --- 2022-05-25 Liu, Hongtao New
[i386] recognize bzhi pattern when there's zero_extendsidi. [i386] recognize bzhi pattern when there's zero_extendsidi. - - --- 2022-05-17 Liu, Hongtao New
Optimize vpermtiw/b to vpunpcklqdq for certain cases. Optimize vpermtiw/b to vpunpcklqdq for certain cases. - - --- 2022-05-13 Liu, Hongtao New
[i386] Implement permutation with pslldq + psrldq + por when pshufb is not available. [i386] Implement permutation with pslldq + psrldq + por when pshufb is not available. - - --- 2022-05-09 Liu, Hongtao New
[Middle-end] Enhance final_value_replacement_loop to handle bitwise induction. [Middle-end] Enhance final_value_replacement_loop to handle bitwise induction. - - --- 2022-05-09 Liu, Hongtao New
[v2] Strip of a vector load which is only used partially. [v2] Strip of a vector load which is only used partially. - - --- 2022-05-09 Liu, Hongtao New
Expand __builtin_memcmp_eq with ptest for OImode. Expand __builtin_memcmp_eq with ptest for OImode. - - --- 2022-05-07 Liu, Hongtao New
Expand __builtin_memcmp_eq with ptest for OI/TImode. Expand __builtin_memcmp_eq with ptest for OI/TImode. - - --- 2022-05-05 Liu, Hongtao New
Strip of a vector load which is only used partially. Strip of a vector load which is only used partially. - - --- 2022-05-05 Liu, Hongtao New
Refine and/ior/xor/andn masked patterns for V*HFmode. Refine and/ior/xor/andn masked patterns for V*HFmode. - - --- 2022-04-06 Liu, Hongtao New
Split vector load from parm_del to elemental loads to avoid STLF stalls. Split vector load from parm_del to elemental loads to avoid STLF stalls. - - --- 2022-04-01 Liu, Hongtao New
Split vector load from parm_del to elemental loads to avoid STLF stalls. Split vector load from parm_del to elemental loads to avoid STLF stalls. - - --- 2022-03-31 Liu, Hongtao New
[i386] Add extra cost for unsigned_load which may have stall forward issue. [i386] Add extra cost for unsigned_load which may have stall forward issue. - - --- 2022-03-16 Liu, Hongtao New
[i386] Add extra cost for unsigned_load which may have stall forward issue. [i386] Add extra cost for unsigned_load which may have stall forward issue. - - --- 2022-03-16 Liu, Hongtao New
[i386] Add extra cost for unsigned_load which may have stall forward issue. [i386] Add extra cost for unsigned_load which may have stall forward issue. - - --- 2022-03-15 Liu, Hongtao New
[i386] Prevent vectorization for load from parm_decl at O2 to avoid STF issue. [i386] Prevent vectorization for load from parm_decl at O2 to avoid STF issue. - - --- 2022-03-04 Liu, Hongtao New
[i386] Optimize v4si broadcast for noavx512vl. [i386] Optimize v4si broadcast for noavx512vl. - - --- 2022-03-04 Liu, Hongtao New
[i386] Replace ix86_gen_scratch_sse_rtx with gen_reg_rtx. [i386] Replace ix86_gen_scratch_sse_rtx with gen_reg_rtx. - - --- 2022-03-01 Liu, Hongtao New
[i386] Fix typo in <code>v1ti3. [i386] Fix typo in <code>v1ti3. - - --- 2022-02-23 Liu, Hongtao New
« 1 2 »