Commits · 1f109f9d9cddbc90d97b50c154a8474e7e623356 · Roger Ferrer / llvm-epi · GitLab

May 06, 2021
- Fix array attribute in bindings for linalg.init_tensor · 1f109f9d
  Denys Shabalin authored May 06, 2021
  1f109f9d
- [mlir][NFC] Fix warning in VectorTransforms.cpp · 933551ea
  thomasraoux authored May 06, 2021
  
  933551ea
- [mlir][vector] add pattern to cast away lead unit dimension for broadcast op · 0b303da6
  thomasraoux authored May 05, 2021
  0b303da6
- [mlir] Add support for ops with regions in 'gpu-async-region' rewriter. · a0d019fc
  Christian Sigg authored May 05, 2021
  a0d019fc
- [MLIR][GPU][NVVM] Add warp synchronous matrix-multiply accumulate ops · 875eb523
  Navdeep Kumar authored May 06, 2021
  875eb523
- [mlir] Check generated IR of math_polynomial_approx.mlir · 3c952ab2
  Emilio Cota authored May 05, 2021
  3c952ab2
- [mlir][Linalg] Fix test to use new reshape op form. · 4b2d7ef3
  MaheshRavishankar authored May 05, 2021
  4b2d7ef3
- [mlir][Linalg] Fix element type of results when folding reshapes. · b6060b76
  MaheshRavishankar authored May 05, 2021
  b6060b76
May 05, 2021
- [mlir] Add polynomial approximation for math::ExpM1 · 0edc4bc8
  Emilio Cota authored May 05, 2021
  0edc4bc8
- [mlir][tosa] Add tosa.depthwise lowering to existing linalg.depthwise_conv · 7abb56c7
  Rob Suderman authored May 05, 2021
  7abb56c7
- [MC] Untangle MCContext and MCObjectFileInfo · 632ebc4a
  Philipp Krones authored May 05, 2021
  632ebc4a
- [mlir][ArmSVE] Add masked arithmetic operations · 95861216
  Javier Setoain authored Apr 19, 2021
  95861216
- [mlir][Affine][Vector] Support vectorizing reduction loops · d80b04ab
  Sergei Grechanik authored May 05, 2021
  d80b04ab
- [mlir][linalg] Fix bug in the fusion on tensors index op handling. · 4a6ee23d
  Tobias Gysi authored May 05, 2021
  4a6ee23d
- [MLIR] Rename free function `verify` on OffsetSizeAndStrideOpInterface · 62851ea7
  Uday Bondhugula authored May 05, 2021
  62851ea7
- [mlir] Use ReassociationIndices instead of affine maps in linalg.reshape. · 2865d114
  Alexander Belyaev authored May 05, 2021
  2865d114
- [mlir][ArmSVE] Add basic arithmetic operations · 001d601a
  Javier Setoain authored May 05, 2021
  001d601a
- [MLIR][SCF] Combine adjacent scf.if with same condition · f4a2dbfe
  William S. Moses authored May 03, 2021
  f4a2dbfe
- [mlir][sparse] Introduce proper sparsification passes · a2c9d4bb
  Aart Bik authored May 03, 2021
  a2c9d4bb
May 04, 2021
- [mlir] Fix region successor bug in forward dataflow analysis · c1c1df63
  River Riddle authored May 04, 2021
  c1c1df63
- [mlir][tosa] Fix tosa.concat by inserting linalg.fill after linalg.init · 1f7adf8c
  Rob Suderman authored May 04, 2021
  1f7adf8c
- [MLIR] Add not icmp canonicalization documentation · cb395b84
  William S. Moses authored May 04, 2021
  cb395b84
- [MLIR][SCF] Assume uses of condition in the body of scf.while is true · 8e211bf1
  William S. Moses authored May 03, 2021
  8e211bf1
- [MLIR] Replace a not of a comparison with appropriate comparison · 93297e4b
  William S. Moses authored May 02, 2021
  93297e4b
- [mlir][linalg] Always lower index operations during loop lowering. · 05d2297b
  Tobias Gysi authored May 04, 2021
  05d2297b
- [mlir] Add lowering from math.expm1 to LLVM. · 93537fab
  Adrian Kuegel authored Feb 16, 2021
  93537fab
- [mlir] Fix bug in TransferOpReduceRank when all dims are broadcasts · aa582819
  Matthias Springer authored May 04, 2021
  aa582819
- [mlir][tosa] Add lowerings for tosa.equal and tosa.arithmetic_right_shift · 07ce5c99
  natashaknk authored May 03, 2021
  07ce5c99
- [mlir] Linalg: add vector transfer lowering patterns to the contraction lowering · 9b67096f
  Eugene Zhulenev authored May 03, 2021
  9b67096f
- [mlir] Add polynomial approximation for math::Log1p · 1c0374e7
  Emilio Cota authored May 03, 2021
  1c0374e7
May 03, 2021
- [mlir][Linalg] Add a utility method to get reassociations maps for reshape. · a6e09391
  MaheshRavishankar authored May 03, 2021
  a6e09391
- [mlir][sparse] fixed typo: sparse -> sparse_tensor · 90d18e10
  Aart Bik authored May 03, 2021
  90d18e10
- [mlir][Linalg] Use rank-reduced versions of subtensor and subtensor insert when possible. · fd15e2b8
  MaheshRavishankar authored May 03, 2021
  fd15e2b8
- [mlir][linalg] Fix vectorization bug in vector transfer indexing map calculation · 9621c1ef
  thomasraoux authored May 03, 2021
  9621c1ef
- [MLIR][Linalg] Avoid forward declaration in `Loops.cpp` · 456efbc0
  Frederik Gossen authored May 03, 2021
  456efbc0
- [MLIR][Linalg] Lower `linalg.tiled_loop` in a separate pass · ec339163
  Frederik Gossen authored May 03, 2021
  ec339163
- Move MLIR python sources to mlir/python. · 9f3f6d7b
  Stella Laurenzo authored Apr 28, 2021
  9f3f6d7b
- [mlir][spirv] Add support to convert std.splat op · d51275cb
  thomasraoux authored May 03, 2021
  d51275cb
- [mlir][vector] Extend vector transfer unrolling to support permutations and broadcast · f44c76d6
  thomasraoux authored May 03, 2021
  f44c76d6
- [mlir][vector] Add canonicalization for extract/insert -> shapecast · 7417541f
  thomasraoux authored May 03, 2021
  7417541f