Commits · dd3b90e4d77b0cc874ffcaaeb28776ba0fd15032 · Roger Ferrer / llvm-epi · GitLab

Mar 19, 2022
- [AArch64] Combine ISD::SETCC into AArch64ISD::ANDS · dd3b90e4
  chenglin.bi authored Mar 19, 2022
  dd3b90e4
- [AArch64] Fix incorrect getSetCCInverse usage within trySwapVSelectOperands. · f46fe36d
  Paul Walker authored Mar 17, 2022
  f46fe36d
- [Xcore] Set Int_MemBarrier as a meta-instruction · ea9139fe
  Shengchen Kan authored Mar 19, 2022
  ea9139fe
Mar 18, 2022
- [ARM] Fix shouldExpandAtomicLoadInIR for subtargets without ldrexd. · ddca6662
  Eli Friedman authored Mar 18, 2022
  ddca6662
- Autogenerate llvm/test/CodeGen/ARM/atomic-load-store.ll · f10f16a6
  Eli Friedman authored Mar 18, 2022
  
  f10f16a6
- [slp][tests] Add missing function attributes · 3abf8ebd
  Philip Reames authored Mar 18, 2022
  3abf8ebd
- [AMDGPU] Add gfx940 run line to gfx90a mfma test. NFC. · d898c956
  Stanislav Mekhanoshin authored Mar 18, 2022
  
  d898c956
- [X86] Adjust tests to run on gfni capable cpu levels · 40c6bab0
  Simon Pilgrim authored Mar 18, 2022
  40c6bab0
- [X86] Fix typo in gfni funnel shift test · bec48f80
  Simon Pilgrim authored Mar 18, 2022
  bec48f80
- [GlobalOpt] Don't replace alias with aliasee if either alias/aliasee may be preemptible · c6692f81
  Fangrui Song authored Feb 01, 2022
  c6692f81
- [NewPM] Don't skip SCCs not in current RefSCC · ddc70237
  Arthur Eubanks authored Mar 16, 2022
  ddc70237
- [AMDGPU] Added gfx940 mfma dst constraint test. NFC. · 43c4d915
  Stanislav Mekhanoshin authored Mar 18, 2022
  
  43c4d915
- [AMDGPU] Disable some MFMA instructions on gfx940 · 4570527e
  Stanislav Mekhanoshin authored Mar 17, 2022
  4570527e
- [AMDGPU] reuse blgp as neg in 2 mfma operations on gfx940 · 0a79e1f3
  Stanislav Mekhanoshin authored Mar 14, 2022
  0a79e1f3
- [ARM] Fix ARM backend to correctly use atomic expansion routines. · 2f497ec3
  Eli Friedman authored Feb 17, 2022
  2f497ec3
- [CostModel][X86] Reduce cost of extracting bool vector elements · 5dde9c12
  Simon Pilgrim authored Mar 18, 2022
  5dde9c12
- [SCCP] do not clean up dead blocks that have their address taken · e1bae23f
  Nick Desaulniers authored Mar 18, 2022
  e1bae23f
- [HWASan] do not replace lifetime intrinsics with tagged address. · 078b5465
  Florian Mayer authored Mar 18, 2022
  078b5465
- Revert "[HWASan] do not replace lifetime intrinsics with tagged address." · dbc918b6
  Florian Mayer authored Mar 18, 2022
  dbc918b6
- [LICM] Add allowspeculation pass options. · 5ab421fb
  Florian Hahn authored Mar 18, 2022
  5ab421fb
- [HWASan] do not replace lifetime intrinsics with tagged address. · 208b923e
  Florian Mayer authored Mar 10, 2022
  208b923e
- [InstCombine] add tests for bswap with shifted operand; NFC · 1f001b25
  Sanjay Patel authored Mar 18, 2022
  
  1f001b25
- [BasicAA] Account for wrapping when using abs(VarIndex) >= abs(Scale). · 1b7ef6aa
  Florian Hahn authored Mar 18, 2022
  1b7ef6aa
- [SLP][X86] Add baseline SSE2 test run to lookahead.ll · b58413da
  Simon Pilgrim authored Mar 18, 2022
  
  b58413da
- [FPEnv][InstSimplify] Teach CannotBeNegativeZero() about constrained intrinsics. · bd050a34
  Kevin P. Neal authored Mar 18, 2022
  bd050a34
- [LowerConstantIntrinsics] Make TLI a required dependency · ab2284a6
  Nikita Popov authored Mar 18, 2022
  ab2284a6
- [LowerConstantIntrinsics] Add test for objectsize of malloc (NFC) · 28c53b38
  Nikita Popov authored Mar 18, 2022
  28c53b38
- [AArch64] Custom lower concat(v4i8 load, ...) · fe6057a2
  David Green authored Mar 18, 2022
  fe6057a2
- [CostModel][X86] Update RUN -passes=* to double quotes to appease update scripts on windows · 4455c5cd
  Simon Pilgrim authored Mar 18, 2022
  
  4455c5cd
- [SCEV] Use constant ranges when determining reachable blocks (PR54434) · 6ffb3ad6
  Nikita Popov authored Mar 18, 2022
  6ffb3ad6
- [RISCV] Add more sign-extending ops to MIR sext.w pass. · 7afa44f5
  Mohammed Nurul Hoque authored Mar 18, 2022
  7afa44f5
- [LoopSimplifyCFG] Check predecessors of exits before marking them dead. · 4a699ae9
  Florian Hahn authored Mar 18, 2022
  4a699ae9
- [AtomicExpand][PowerPC] Fix all-one mask value · 31906a60
  Kai Luo authored Mar 18, 2022
  31906a60
- [X86] Set Int_MemBarrier as a meta-instruction · 1a70febf
  Shengchen Kan authored Mar 18, 2022
  1a70febf
- [InstCombine] Sink instructions with multiple users in a successor block. · 0af3e6a2
  Andrew Wei authored Mar 18, 2022
  0af3e6a2
- Reland "Load pass plugins during option processing, so that plugin options are... · c80198b3
  Wael Yehia authored Mar 13, 2022
  c80198b3
- [AMDGPU] Use COPY_TO_REGCLASS for buffer_atomic_cmpswap selection · aee36849
  Abinav Puthan Purayil authored Mar 17, 2022
  aee36849
- Revert "[SLP] Fix lookahead operand reordering for splat loads." due to build failures · 9136145e
  Vasileios Porpodas authored Mar 17, 2022
  9136145e
- [SLP][NFC] Added a test for a followup patch that enables handling splat loads with uses. · 511fa080
  Vasileios Porpodas authored Mar 16, 2022
  
  511fa080
- [SLP] Fix lookahead operand reordering for splat loads. · 5efa7898
  Vasileios Porpodas authored Mar 09, 2022
  5efa7898