Commits · eff0542c56c716e28a4f8c2874e9cd5222f0d7f8 · Roger Ferrer / llvm-epi · GitLab

Oct 10, 2018
- revert r344082: [InstCombine] reverse 'trunc X to <N x i1>' canonicalization · 58fc00d0
  Sanjay Patel authored Oct 10, 2018
  58fc00d0
- [LV] Add a new reduction pattern match · cb19c8e3
  Renato Golin authored Oct 10, 2018
  cb19c8e3
- Relax trivial cast requirements in CallPromotionUtils · 3759efc6
  Scott Linder authored Oct 10, 2018
  3759efc6
- Fix an ordering bug in the scalarizer. · 3d457982
  Neil Henning authored Oct 10, 2018
  3d457982
- [Analysis] Make LocationSizes carry an 'imprecise' bit · 40dc63e1
  George Burgess IV authored Oct 10, 2018
  40dc63e1
- [LV] Move test for r343954 into x86 subdirectory · 90fde0e0
  Justin Bogner authored Oct 09, 2018
  90fde0e0
Oct 09, 2018
- [FPEnv] PatternMatcher support for checking FNEG ignoring signed zeros · bea5967e
  Cameron McInally authored Oct 09, 2018
  bea5967e
- [InstCombine] reverse 'trunc X to <N x i1>' canonicalization · e9ca7ea3
  Sanjay Patel authored Oct 09, 2018
  e9ca7ea3
- [SLPVectorizer] Check that lowered type is floating point before calling isFabsFree · 81abca32
  Sam Clegg authored Oct 09, 2018
  81abca32
- [InstCombine] add tests for extract subvector shuffles; NFC · 1238e6c2
  Sanjay Patel authored Oct 09, 2018
  1238e6c2
Oct 08, 2018
- [TailCallElim] Enable marking of calls with byval as tails · 0c93ea26
  Robert Lougher authored Oct 08, 2018
  0c93ea26
- [ThinLTO] Keep non-prevailing (linkonce|weak)_odr symbols live · bfdad33b
  Xin Tong authored Oct 08, 2018
  bfdad33b
- [InstCombine] Fix incongruous GEP type addrspace · fa120cbd
  Ewan Crawford authored Oct 08, 2018
  fa120cbd
- [LV] Do not create SCEVs on broken IR in emitTransformedIndex. PR39160 · b0736965
  Max Kazantsev authored Oct 08, 2018
  b0736965
Oct 07, 2018
- [InstSimplify] add vector test for fneg+fdiv; NFC · 01daf62a
  Sanjay Patel authored Oct 07, 2018
  01daf62a
- [IAI,LV] Avoid creating interleave-groups for predicated accesse · 72f6e299
  Dorit Nuzman authored Oct 07, 2018
  72f6e299
Oct 04, 2018
- [InstCombine] drop poison flags in SimplifyVectorDemandedElts · 3436dc29
  Sanjay Patel authored Oct 04, 2018
  3436dc29
- [InstCombine] allow bitcast to/from FP for vector insert/extract transform · 3746e11a
  Sanjay Patel authored Oct 04, 2018
  3746e11a
Oct 03, 2018
- [InstCombine] allow SimplifyDemandedVectorElts to work with FP binops · cafdeb1a
  Sanjay Patel authored Oct 03, 2018
  cafdeb1a
- [InstCombine] add tests for binop undef-into-constant propagation; NFC · 6916206d
  Sanjay Patel authored Oct 03, 2018
  6916206d
- Add support for new pass manager · 9e20ade7
  Aditya Kumar authored Oct 03, 2018
  9e20ade7
- [InstCombine] add icmp+logic tests with commuted ops; NFC · abcacf97
  Sanjay Patel authored Oct 02, 2018
  abcacf97
Oct 02, 2018
- [InstCombine] add tests with undef elements; NFC · e2cd6384
  Sanjay Patel authored Oct 02, 2018
  e2cd6384
- [InstCombine] add more insert/extract vector tests with FP types; NFC · 6dbecb41
  Sanjay Patel authored Oct 02, 2018
  6dbecb41
- [InstCombine] Fold ~A - Min/Max(~A, O) -> Max/Min(A, ~O) - A · 1e44c3b6
  David Green authored Oct 02, 2018
  1e44c3b6
- [InstCombine] Tests for ~A - Min/Max(~A, O) -> Max/Min(A, ~O) - A. NFC · c066a926
  David Green authored Oct 02, 2018
  c066a926
- AMDGPU: Expand atomicrmw nand in IR · ab411933
  Matt Arsenault authored Oct 02, 2018
  ab411933
Oct 01, 2018
- [InstCombine] add inverse test for vector trunc canonical form; NFC · de5e8b93
  Sanjay Patel authored Oct 01, 2018
  de5e8b93
- [InstCombine] regenerate test checks; NFC · 746eb091
  Sanjay Patel authored Oct 01, 2018
  746eb091
- [InstCombine] Handle vector compares in foldGEPIcmp(), take 2 · c954b863
  Jesper Antonsson authored Oct 01, 2018
  c954b863
- [InstCombine] try to convert vector insert+extract to trunc; 2nd try · 31b07198
  Sanjay Patel authored Oct 01, 2018
  31b07198
- [InstCombine] add more insert-extract tests for D52439; NFC · 22ae8dab
  Sanjay Patel authored Oct 01, 2018
  22ae8dab
- Revert r343407 "[InstCombine] try to convert vector insert+extract to trunc" · a60aa913
  Hans Wennborg authored Oct 01, 2018
  a60aa913
Sep 30, 2018
- [InstCombine] try to convert vector insert+extract to trunc · 1e0f1f64
  Sanjay Patel authored Sep 30, 2018
  1e0f1f64
- [InstCombine] allow lengthening of insertelement to eliminate shuffles · 26c119a9
  Sanjay Patel authored Sep 30, 2018
  26c119a9
Sep 29, 2018
- [InstCombine] add test for vector widening of insertelements; NFC · 20c64510
  Sanjay Patel authored Sep 29, 2018
  20c64510
Sep 28, 2018
- [InstCombine] don't propagate wider shufflevector arguments to predecessors · 242f90fe
  Sanjay Patel authored Sep 28, 2018
  242f90fe
- [InstCombine] adjust shuffle undef propagation tests; NFC · 699ee504
  Sanjay Patel authored Sep 28, 2018
  699ee504
- [ScalarizeMaskedMemIntrin] Use MinAlign to calculate alignment for the scalar... · bb50c386
  Craig Topper authored Sep 28, 2018
  bb50c386
- [ScalarizeMaskedMemIntrin] Fix the alignment calculation for the scalar stores... · fdf4c76c
  Craig Topper authored Sep 28, 2018
  fdf4c76c