Commits · 747c6a0c734e618db8132b503f432d8274cc56b5 · Lorenzo Albano / LLVM bpEVL · GitLab

May 02, 2022
- [SDAG] fix miscompile when casting int->FP->int · 747c6a0c
  Sanjay Patel authored May 02, 2022
  747c6a0c
- [Object][DX] Initial DXContainer parsing support · 4070aa01
  Chris Bieneman authored Apr 28, 2022
  4070aa01
- Update movmsk-cmp.ll to match improvements made to InstCombine · b6beb590
  Amaury Séchet authored May 02, 2022
  b6beb590
- [VPlan] Do not create VPWidenCall recipes for scalar vector factors. · 0ef8ca6d
  Florian Hahn authored May 02, 2022
  0ef8ca6d
- [DeadArgElim] Set unused arguments for internal functions · e547a333
  Quentin Colombet authored Apr 29, 2022
  e547a333
- [ifs] Fix bug where exclude only excluded when outputting ifs files · 7aaad7b1
  Alex Brachet authored May 02, 2022
  7aaad7b1
- Reapply "[BuildLibCalls] Introduce getOrInsertLibFunc() for use when building · 304378fd
  Jonas Paulsson authored Apr 20, 2022
  304378fd
- [mlir][OpenMP] Add omp.cancel and omp.cancellationpoint. · c685f821
  Raghu Maddhipatla authored May 02, 2022
  c685f821
- [PowerPC] Enable CR bits support for Power8 and above. · 2534dc12
  Amy Kwan authored May 02, 2022
  2534dc12
- [GlobalOpt] Iterate over replaced values deterministically to constprop · b07aab8f
  Arthur Eubanks authored May 02, 2022
  b07aab8f
- [InstCombine] Handle non-canonical GEP index in indexed compare fold (PR55228) · 95fedfab
  Nikita Popov authored May 02, 2022
  95fedfab
- [gn build] Port 5de0a3e9 · aa69cb76
  LLVM GN Syncbot authored May 02, 2022
  
  aa69cb76
- [X86] Reduce some superfluous diffs between znver1/znver2 models. NFC · 59dc8ce9
  Simon Pilgrim authored May 02, 2022
  59dc8ce9
- [X86][AMX] combineLdSt - don't dereference dyn_cast. NFC · ce9c0fac
  Simon Pilgrim authored May 02, 2022
  ce9c0fac
- BuildLibCalls: add alloc-family attribute to many allocator functions · c7ae423e
  Augie Fackler authored Mar 17, 2022
  c7ae423e
- [LV][SLP] Add tests for vectorizing fptoi_sat intrinsics. NFC · c7d39fd6
  David Green authored May 02, 2022
  
  c7d39fd6
- BuildLibCalls: infer allocptr attribute for free and realloc() family functions · e9404565
  Augie Fackler authored Mar 16, 2022
  e9404565
- [CodeGen] Add tests for X+(Y&~X) pattern (NFC) · 2e8a5061
  Nikita Popov authored May 02, 2022
  
  2e8a5061
- [AArch64] add tests for int->FP->int casts; NFC · cb3fb085
  Sanjay Patel authored May 02, 2022
  cb3fb085
- [x86] add tests for int->FP->int casts; NFC · 8bc0272c
  Sanjay Patel authored May 02, 2022
  8bc0272c
- [x86] update test file with complete auto-generated check lines; NFC · bdc30731
  Sanjay Patel authored May 02, 2022
  bdc30731
- [X86] MOVDDUP has the same sched behaviour as MOVSHDUP/MOVSLDUP on Skylake · c7662dc3
  Simon Pilgrim authored May 02, 2022
  c7662dc3
- [InstCombine] Add tests for A+(B&~A) and A+((A&B)^B) (NFC) · 9082ebeb
  Nikita Popov authored May 02, 2022
  
  9082ebeb
- [gn build] (manually) port fb7a4354 · d21289e0
  Nico Weber authored May 02, 2022
  
  d21289e0
- [SLP][X86] Add test coverage for PR41892 · c5e875f5
  Simon Pilgrim authored May 02, 2022
  
  c5e875f5
- [AArch64] Cost modelling for fptoi_sat · 2dcb2d85
  David Green authored May 02, 2022
  2dcb2d85
- [CostModel][X86] getScalarizationOverhead - handle vXi1 extracts with MOVMSK (pre-AVX512) · 86bb7df6
  Simon Pilgrim authored May 02, 2022
  86bb7df6
- [Local] Consider atomic loads from constant global as dead · aae5f811
  Nikita Popov authored Apr 22, 2022
  aae5f811
- [ConstantFold] Don't convert getelementptr to ptrtoint+inttoptr · 597946a4
  Nikita Popov authored Apr 29, 2022
  597946a4
- [AArch64] Add more comprehensive reverse shuffle costmodel tests. NFC · 986de8f5
  David Green authored May 02, 2022
  
  986de8f5
- [ArgPromotion][Attributor] Update min-legal-vector-width when do promotion · 7c044542
  Phoebe Wang authored May 02, 2022
  7c044542
- [LoopCacheAnalysis] Use stable_sort() to avoid non-deterministic print output · 3d6fe7ac
  Congzhe Cao authored May 02, 2022
  3d6fe7ac
- [gn build] Port 3939e99a · 1790e297
  LLVM GN Syncbot authored May 01, 2022
  
  1790e297
- llvm-reduce: Fix not removing first instruction in MachineBasicBlock · aabea3b2
  Matt Arsenault authored Apr 19, 2022
  aabea3b2
- llvm-reduce: Introduce new scoring mechanism for MIR reductions · 35264e71
  Matt Arsenault authored Apr 19, 2022
  35264e71
- llvm-reduce: Do not try to delete frame instructions · 0b896b75
  Matt Arsenault authored Apr 25, 2022
  0b896b75
May 01, 2022
- llvm-reduce: Add pass to reduce IR references from MIR · 3939e99a
  Matt Arsenault authored Apr 19, 2022
  3939e99a
- [RISCV] Lower case the first letter of LowerRISCVMachineOperandToMCOperand. NFC · 2019c9b1
  Fangrui Song authored May 01, 2022
  
  2019c9b1
- doc: update of the adv build doc now that clang is in tree too · ee4ac3a8
  Sylvestre Ledru authored May 01, 2022
  ee4ac3a8
- [SimpleLoopUnswitch] Freeze individual OR/AND operands. · 5387a38c
  Florian Hahn authored May 01, 2022
  5387a38c