Commits · ecc997807180a6e763f12e3d011f6b887db0d6a9 · Roger Ferrer / llvm-epi · GitLab

Sep 29, 2020
- [mlir][openacc] Add update operation · ecc99780
  Valentin Clement authored Sep 29, 2020
  ecc99780
- [mlir][Linalg] Refactor Linalg op initTensors support - NFC · 6b649570
  Nicolas Vasilache authored Sep 29, 2020
  6b649570
- [mlir][Linalg] Refactor Linalg creation of loops to allow passing iterArgs - NFC · 074ab233
  Nicolas Vasilache authored Sep 29, 2020
  074ab233
- [mlir][GPU] Improve constant sinking in kernel outlining · edeff6e6
  Stephan Herhut authored Sep 29, 2020
  edeff6e6
- [mlir][openacc] Add acc.data operation verifier · bbb5dc49
  Valentin Clement authored Sep 28, 2020
  bbb5dc49
- Remove dependency from LLVM Dialect on the OpenMP dialect · 9f9f89d4
  Mehdi Amini authored Sep 28, 2020
  9f9f89d4
- [mlir][Affine][VectorOps] Fix super vectorizer utility (D85869) · 93936da9
  Diego Caballero authored Sep 28, 2020
  93936da9
Sep 28, 2020
- [mlir] [VectorOps] Relaxed restrictions on vector.reduction types even more · e9628955
  Aart Bik authored Sep 28, 2020
  e9628955
- [mlir][openacc] Switch to assembly format for acc.data · 12ab4f8a
  Valentin Clement authored Sep 27, 2020
  12ab4f8a
- [mlir][openacc] Remove detach and delete operands from acc.data · 3d2bab17
  Valentin Clement authored Sep 27, 2020
  3d2bab17
Sep 25, 2020
- [NFC] Fix comment for DataOp · 06104cb9
  clementval authored Sep 25, 2020
  
  06104cb9
Sep 24, 2020
- [MLIR][SPIRV] Fixed dialect loading in deserialization · 8f72717e
  George Mitenkov authored Sep 24, 2020
  8f72717e
Sep 23, 2020
- [MLIR][NFC] Adopt use of BlockRange in place of ArrayRef<Block *> · a6ae6950
  Rahul Joshi authored Sep 22, 2020
  a6ae6950
- [MLIR][NFC] Adopt use of TypeRange in build() methods. · 08e4f078
  Rahul Joshi authored Sep 22, 2020
  08e4f078
- NFC: Remove unused variable. · 8e84972a
  Stella Laurenzo authored Sep 23, 2020
  
  8e84972a
- [mlir][openacc] Use OptionalParseResult in loop op parser instead of bool variables · bd8b50cd
  Valentin Clement authored Sep 23, 2020
  bd8b50cd
- [mlir][Linalg] Add pattern to fold linalg.tensor_reshape that add unit extent dims. · b62f9f44
  MaheshRavishankar authored Sep 22, 2020
  b62f9f44
Sep 22, 2020
- [NFC][mlir] Remove llvm:: prefix from SmallVector in parallel pretty printer. · bf202b8c
  David Truby authored Sep 21, 2020
  bf202b8c
- [MLIR][Linalg] Fix assertion in dependency analysis · 0841f717
  Frederik Gossen authored Sep 22, 2020
  0841f717
- [mlir][Linalg] Uniformize linalg.generic with named ops. · ed229132
  Nicolas Vasilache authored Sep 21, 2020
  ed229132
- [mlir] Fix -Wunused-variable in -DLLVM_ENABLE_ASSERTIONS=off build after D85869 · 91671e13
  Fangrui Song authored Sep 21, 2020
  
  91671e13
- [MLIR][Affine][VectorOps] Utility to vectorize loop nest using strategy · 14d0735d
  Diego Caballero authored Sep 21, 2020
  14d0735d
Sep 21, 2020
- [mlir][VectorOps] Loosen restrictions on vector.reduction types · 2d76274b
  Benjamin Kramer authored Sep 21, 2020
  2d76274b
- [mlir] Shape.AssumingOp implements RegionBranchOpInterface. · ffdd4a46
  Tres Popp authored Sep 18, 2020
  ffdd4a46
Sep 18, 2020
- [spirv] Move device info from resource limit into target env · 1f0b4363
  Lei Zhang authored Sep 18, 2020
  1f0b4363
- [MLIR][SPIRV] Create new ctx for deserialization in roundtrips. · 7b61b192
  ergawy authored Sep 18, 2020
  7b61b192
- [mlir][openacc] Add missing operands for acc.data operation · 88a1d402
  Valentin Clement authored Sep 18, 2020
  88a1d402
- [mlir][openacc] Support Index and AnyInteger in loop op · 22dde1f9
  Valentin Clement authored Sep 18, 2020
  22dde1f9
- [mlir][Linalg] Evolve named ops to use assembly form and support linalg on tensors. · 93fd30ba
  Nicolas Vasilache authored Sep 18, 2020
  93fd30ba
- [mlir][shape] Add `shape.cstr_require %bool` · bae63742
  Sean Silva authored Sep 17, 2020
  bae63742
Sep 17, 2020
- [MLIR][Affine] Add parametric tile size support for affine.for tiling · 0602e8f7
  Navdeep Kumar authored Sep 17, 2020
  0602e8f7
- [MLIR] Support for return values in Affine.For yield · 296e97ae
  Abhishek Varma authored Sep 17, 2020
  296e97ae
- [mlir][Vector] Add a folder for vector.broadcast · f16abe5f
  Hanhan Wang authored Sep 17, 2020
  f16abe5f
- [mlir][openacc] Change operand type from index to AnyInteger in parallel op · 6d3cabd9
  Valentin Clement authored Sep 17, 2020
  6d3cabd9
- [mlir][Standard] Canonicalize chains of tensor_cast operations · 5e0ded26
  Stephan Herhut authored Sep 16, 2020
  5e0ded26
- [mlir][Linalg] Convolution tiling added to ConvOp vectorization pass · 347d59b1
  Jakub Lichman authored Sep 17, 2020
  347d59b1
- [mlir] Remove redundant shape.cstr_broadcastable canonicalization. · b0562923
  Tres Popp authored Sep 15, 2020
  b0562923
- [mlir][spirv] Add GroupNonUniformBroadcastOp · 4ce84b0e
  Artur Bialas authored Sep 16, 2020
  4ce84b0e
Sep 16, 2020
- [mlir][openacc] Add missing operands for acc.parallel operation · d9953d15
  Valentin Clement authored Sep 16, 2020
  d9953d15
- Enable inlining for Linalg dialect · 8c0dc1e3
  Eugene Zhulenev authored Sep 16, 2020
  8c0dc1e3