Commits

Branch selector

 main

User selector

phoebewang

Datepicker

All time

Commit History

Commits on May 1, 2024

• [X86] Enable EVEX512 when host CPU has AVX512 (llvm#90479)

  Show description for a7b8b89
  

  phoebewang

  authored and

  tstellar

  committed
  a7b8b89

  Copy full SHA for a7b8b89

Commits on Apr 30, 2024

• [X86][EVEX512] Check hasEVEX512 for canExtendTo512DQ (llvm#90390)

  Show description for 58648f3
  

  phoebewang

  authored and

  tstellar

  committed
  58648f3

  Copy full SHA for 58648f3

Commits on Mar 19, 2024

• [X86] Add missing subvector_subreg_lowering for BF16 (llvm#83720)

  

  phoebewang

  authored and

  tstellar

  committed
  26a1d66

  Copy full SHA for 26a1d66

Commits on Jan 17, 2024

• [X86] Use vXi1 for k constraint in inline asm (llvm#77733)

  Show description for 8d6e82d
  phoebewang

  authored
  8d6e82d

  Copy full SHA for 8d6e82d

• [X86][BF16] Improve float -> bfloat lowering under AVX512BF16 and AVXNECONVERT (llvm#78042)

  phoebewang

  authored
  9745c13

  Copy full SHA for 9745c13

Commits on Jan 14, 2024

• [AVX10][Doc] Add documentation about AVX10 options and their attentions (llvm#77925)

  phoebewang

  authored
  0e93d04

  Copy full SHA for 0e93d04

Commits on Jan 5, 2024

• [X86] Add ABI handling for __float128 to match with GCC (llvm#75156)

  Show description for f07aba4
  phoebewang

  authored
  f07aba4

  Copy full SHA for f07aba4

• [X86][BF16] Try to use f16 for lowering (llvm#76901)

  Show description for 59af659
  phoebewang

  authored
  59af659

  Copy full SHA for 59af659

Commits on Jan 4, 2024

• [X86][BF16] Add 32-bit tests to show ABI problem, NFC

  phoebewang

  committed
  176c341

  Copy full SHA for 176c341

Commits on Dec 31, 2023

• [X86][BF16] Add subvec_zero_lowering patterns (llvm#76507)

  phoebewang

  authored
  a384cd5

  Copy full SHA for a384cd5

Commits on Dec 29, 2023

• [X86][NFC] Remove meaningless FIXME

  Show description for 6c87f46
  phoebewang

  committed
  6c87f46

  Copy full SHA for 6c87f46

Commits on Dec 28, 2023

• [X86][BF16] Support INSERT_SUBVECTOR and CONCAT_VECTORS (llvm#76485)

  phoebewang

  authored
  e499ae5

  Copy full SHA for e499ae5

• [X86][BF16] Add X86SubVBroadcastld patterns (llvm#76479)

  phoebewang

  authored
  3081bac

  Copy full SHA for 3081bac

Commits on Dec 15, 2023

• [X86][AVX10] Allow 64-bit mask register used without EVEX512 (llvm#75571)

  Show description for f5e48fe
  phoebewang

  authored
  f5e48fe

  Copy full SHA for f5e48fe

Commits on Dec 5, 2023

• [NFC] Remove duplicated message resulted from a bad merge I think

  phoebewang

  committed
  0ca80eb

  Copy full SHA for 0ca80eb

Commits on Nov 24, 2023

• Adding missing useSoftFloat check, NFCI

  phoebewang

  committed
  006f264

  Copy full SHA for 006f264

• [X86][AVX10] Allow AVX10 use VBMI2 instructions (llvm#73276)

  phoebewang

  authored
  ea81e31

  Copy full SHA for ea81e31

Commits on Nov 22, 2023

• [X86][AVX10] Add no-evex512 for MMX intrinsics, NFCI

  Show description for e5cc3da
  phoebewang

  committed
  e5cc3da

  Copy full SHA for e5cc3da

Commits on Nov 19, 2023

• [NFC] Fix typos in comments

  phoebewang

  committed
  5237193

  Copy full SHA for 5237193

Commits on Nov 14, 2023

• Reland "[X86][AVX10] Fix a bug when using -march with no-evex512 attribute (llvm#72126)"

  Show description for e96edde
  phoebewang

  committed
  e96edde

  Copy full SHA for e96edde

• Revert "[X86][AVX10] Fix a bug when using -march with no-evex512 attribute (llvm#72126)"

  Show description for 17dd0c7
  phoebewang

  committed
  17dd0c7

  Copy full SHA for 17dd0c7

• [X86][AVX10] Fix a bug when using -march with no-evex512 attribute (llvm#72126)

  Show description for 451c594
  phoebewang

  authored
  451c594

  Copy full SHA for 451c594

Commits on Nov 10, 2023

• [X86][AVX10] Permit AVX512 options/features used together with AVX10 (llvm#71318)

  Show description for f229ba4
  phoebewang

  authored
  f229ba4

  Copy full SHA for f229ba4

Commits on Nov 7, 2023

• [X86] Add a EVEX256 macro to match with GCC and MSVC (llvm#71317)

  phoebewang

  authored
  c78aeab

  Copy full SHA for c78aeab

Commits on Oct 29, 2023

• [X86] Avoid returning the same shuffle operation for broadcast (llvm#70592)

  Show description for b5281af
  phoebewang

  authored
  b5281af

  Copy full SHA for b5281af

Commits on Oct 27, 2023

• [X86][EVEX512] Do not allow 512-bit memcpy without EVEX512 (llvm#70420)

  Show description for 58d4fe2
  phoebewang

  authored
  58d4fe2

  Copy full SHA for 58d4fe2

Commits on Oct 18, 2023

• [X86][RFC] Support AVX10 options (llvm#67278)

  Show description for cfbf0a5
  phoebewang

  authored
  cfbf0a5

  Copy full SHA for cfbf0a5

Commits on Oct 17, 2023

• [X86] Enable bfloat type support in inline assembly constraints (llvm#68469)

  Show description for 3d6e416
  phoebewang

  authored
  3d6e416

  Copy full SHA for 3d6e416

Commits on Oct 16, 2023

• [X86][FP16] Do not combine to ADDSUB if target doesn't support FP16 (llvm#69109)

  Show description for 0ddca87
  phoebewang

  authored
  0ddca87

  Copy full SHA for 0ddca87

Commits on Sep 26, 2023

• [X86][FP16] Add missing handling for FP16 constrained cmp intrinsics (llvm#67400)

  phoebewang

  authored
  31631d3

  Copy full SHA for 31631d3

Commits on Sep 11, 2023

• [X86][EVEX512] Restrict attaching EVEX512 for default CPU only, NFCI (llvm#65920)

  Show description for 8a58407
  phoebewang

  authored
  8a58407

  Copy full SHA for 8a58407

Commits on Sep 8, 2023

• [X86][RFC] Add new option -m[no-]evex512 to disable ZMM and 64-bit mask instructions for AVX512 features

  Show description for 2419409
  phoebewang

  committed
  2419409

  Copy full SHA for 2419409

• [X86] Do not check alignment for VINSERTPS (llvm#65721)

  Show description for da1eb88
  phoebewang

  authored
  da1eb88

  Copy full SHA for da1eb88

• [X86] Do not directly fold for VINSERTPS (llvm#65718)

  Show description for 2e44b07
  phoebewang

  authored
  2e44b07

  Copy full SHA for 2e44b07

• [X86][NFC] Add a test case to show wrong memory folding for vinsertps

  phoebewang

  committed
  11c3b97

  Copy full SHA for 11c3b97

Pagination

Previous

Next