Commits · c0809f8d79045941d45c7bd60a12ddd0f6e0811a · Roger Ferrer / llvm-epi · GitLab

Sep 15, 2020
- [mlir] check for failures when packing function sigunatures in std->llvm conversion · 967c7b69
  Alex Zinenko authored Sep 14, 2020
  967c7b69
Sep 14, 2020
- Make namespace handling uniform across dialect backends. · 7d1ed69c
  Federico Lebrón authored Sep 14, 2020
  7d1ed69c
- Add support for casting elements in vectors for certain Std dialect type conversion operations. · ef7a255c
  Lubomir Litchev authored Sep 09, 2020
  ef7a255c
- [mlir] Check for type conversion success in std->llvm function conversion · 5cac85c9
  Alex Zinenko authored Sep 03, 2020
  5cac85c9
Sep 11, 2020
- [mlir] Fix some edge cases around 0-element TensorFromElementsOp · 84a6da67
  Sean Silva authored Sep 10, 2020
  84a6da67
- [mlir][spirv] OpConvertSToF support operands with different bitwidth. · aeb43143
  Xin Wang authored Sep 11, 2020
  aeb43143
- [NFC] Follow up on D87111 - Add an option for unrolling loops up to a factor - CR issues addressed. · 32062478
  Lubomir Litchev authored Sep 08, 2020
  32062478
- [mlir][Linalg] Refactor StructuredOpInterface - NFC · e6f2f17f
  Nicolas Vasilache authored Sep 11, 2020
  e6f2f17f
- [mlir][Linalg] Print both types for linalg.transpose · a0e0d30a
  Benjamin Kramer authored Sep 10, 2020
  a0e0d30a
- [mlir][Linalg] Make LinalgBaseTilingPattern not delete the original operation. · d380b582
  MaheshRavishankar authored Sep 10, 2020
  d380b582
- [mlir][Analysis] Allow Slice Analysis to work with linalg::LinalgOp · 0a391c60
  MaheshRavishankar authored Sep 10, 2020
  0a391c60
Sep 10, 2020
- Allow Dialects to be initialized via nullptr. · d867be5d
  Federico Lebrón authored Sep 10, 2020
  d867be5d
- Make struct dialects have the same field name as everything else, 'dialect'. · a3942308
  Federico Lebrón authored Sep 10, 2020
  a3942308
- Introduce linalg.vecmat · 5638df19
  Eugene Burmako authored Sep 10, 2020
  5638df19
- [MLIR][Standard] Simplify `tensor_from_elements` · 018f6936
  Frederik Gossen authored Sep 10, 2020
  018f6936
- [mlir][Linalg] Small refactoring of ConvOpVectorization · fea175b5
  Jakub Lichman authored Sep 10, 2020
  fea175b5
- [mlir][Linalg] Add Utility method to get loop ranges for a LinalgOp. · a7b2977a
  MaheshRavishankar authored Sep 09, 2020
  a7b2977a
Sep 09, 2020
- Wordsmith RegionBranchOpInterface verification errors · be35264a
  Sean Silva authored Sep 08, 2020
  be35264a
- Rename MemRefDescriptor::getElementType() to MemRefDescriptor::getElementPtrType(). · 3a577f54
  Christian Sigg authored Sep 09, 2020
  3a577f54
- [mlir] Added support for loops to BufferPlacement transformation. · feb0b9c3
  Marcel Koester authored Aug 07, 2020
  feb0b9c3
- [MLIR][Shape] Lower `shape_of` to `dynamic_tensor_from_elements` · 5106a8b8
  Frederik Gossen authored Sep 09, 2020
  5106a8b8
- [MLIR][Standard] Update `tensor_from_elements` assembly format · 133322d2
  Frederik Gossen authored Sep 09, 2020
  133322d2
- [MLIR] Add debug support for ignored patterns · 6a494e11
  Frederik Gossen authored Sep 09, 2020
  6a494e11
Sep 08, 2020
- [MLIR] Change Operation::create() methods to use Value/Type/Block ranges. · 8893d081
  Rahul Joshi authored Sep 02, 2020
  8893d081
- Add more explicit error message when creating a type or attribute for an unregistered dialect (NFC) · 97e77ac0
  Mehdi Amini authored Sep 08, 2020
  97e77ac0
- Add an option for unrolling loops up to a factor. · e2394245
  Lubomir Litchev authored Sep 03, 2020
  e2394245
- [mlir][VectorOps] Fix more GCC5 weirdness · 51d30c34
  Benjamin Kramer authored Sep 08, 2020
  51d30c34
- [mlir] Fix bug in copy removal · 4e9f4d0b
  Ehsan Toosi authored Aug 24, 2020
  4e9f4d0b
- [mlir][VectorOps] · df63eede
  Benjamin Kramer authored Sep 08, 2020
  df63eede
- [mlir] remove BufferAssignmentPlacer from BufferAssignmentOpConversionPattern · 847299d3
  Ehsan Toosi authored Aug 20, 2020
  847299d3
- [mlir][VectorOps] Clean up outdated comments. NFCI. · 307dc7b2
  Benjamin Kramer authored Sep 08, 2020
  307dc7b2
- [mlir] Fix of broken build on windows caused by using uint · 83d82d1f
  Jakub Lichman authored Sep 08, 2020
  
  83d82d1f
- [mlir][VectorOps] Redo the scalar loop emission in VectoToSCF to pad instead of clipping · 239eff50
  Benjamin Kramer authored Sep 07, 2020
  239eff50
- [mlir] Conv ops vectorization pass · 67b37f57
  Jakub Lichman authored Sep 08, 2020
  67b37f57
- [mlir][Vector] Make VectorToSCF deterministic · 9be61784
  Nicolas Vasilache authored Sep 08, 2020
  9be61784
Sep 07, 2020
- [MLIR][Shape] Merge `shape` to `std`/`scf` lowerings. · a70f2eb3
  Frederik Gossen authored Sep 07, 2020
  a70f2eb3
- Revert "[MLIR][Shape] Merge `shape` to `std`/`scf` lowerings." · 973800dc
  David Truby authored Sep 07, 2020
  973800dc
- [MLIR][Shape] Merge `shape` to `std`/`scf` lowerings. · 15acdd75
  Frederik Gossen authored Sep 07, 2020
  15acdd75
- [MLIR][Standard] Add `dynamic_tensor_from_elements` operation · 136eb79a
  Frederik Gossen authored Sep 07, 2020
  136eb79a
- [mlir][Vector] Revisit VectorToSCF. · 8d64df9f
  Nicolas Vasilache authored Sep 04, 2020
  8d64df9f