Commits · 8846557672d6a490b500b1c24e90a8effcb38901 · Roger Ferrer / llvm-epi-0.8 · GitLab

This is an archived project. Repository and other project resources are read-only.

Dec 14, 2019
- Fix maskAndClamp in gpu.all_reduce. · 88465576
  Christian Sigg authored Dec 12, 2019
  88465576
Dec 13, 2019
- NFC: Cleanup the various Op::print methods. · e7aa47ff
  River Riddle authored Dec 12, 2019
  e7aa47ff
Dec 12, 2019
- Fix logic on when to emit collective type but separate arg builder · a50cb184
  Jacques Pienaar authored Dec 12, 2019
  a50cb184
- [VectorOps] Add lowering of vector.shuffle to LLVM IR · 1c81adf3
  Aart Bik authored Dec 12, 2019
  1c81adf3
- Add type inference variant for separate params builder generated · 41a73ddc
  Jacques Pienaar authored Dec 12, 2019
  41a73ddc
- Retire !linalg.buffer type - NFC · 782ae296
  Nicolas Vasilache authored Dec 12, 2019
  782ae296
- [Linalg] Add test for fusion of GenericOp with IndexedGenericOp. · 1b579d99
  Alexander Belyaev authored Dec 12, 2019
  1b579d99
- Added lowering of `std.tanh` to llvm function call to `tanh` and `tanhf`. · f7bffad5
  Ehsan Toosi authored Dec 12, 2019
  f7bffad5
- Move cpu runner utils templates to .h · 95b5a4fd
  Nicolas Vasilache authored Dec 12, 2019
  95b5a4fd
- Automated rollback of commit f68ac464 · 9b855826
  Christian Sigg authored Dec 12, 2019
  9b855826
- Switch from shfl.bfly to shfl.down. · f68ac464
  Christian Sigg authored Dec 12, 2019
  f68ac464
- Make OpBuilder::insert virtual instead of OpBuilder::createOperation. · 851a8516
  River Riddle authored Dec 11, 2019
  851a8516
- Add std.log* and llvm.intr.log* that correspond to the LLVMIR intrinsics · 9dfa84a2
  Nicolas Vasilache authored Dec 11, 2019
  9dfa84a2
Dec 11, 2019
- Add missing CMake dependency for MLIRTestIR. · b909299d
  Mahesh Ravishankar authored Dec 11, 2019
  b909299d
- Fix OSS build · beda0b2d
  Nicolas Vasilache authored Dec 11, 2019
  beda0b2d
- Expose a convenience function to add interface attributes to a function. · 652fc261
  Mahesh Ravishankar authored Dec 11, 2019
  652fc261
- [spirv] Add lowering for std.fdiv, std.frem, std.fsub · d968f969
  Denis Khalikov authored Dec 11, 2019
  d968f969
- Continue refactoring StructuredOps utilities · 508d4e67
  Nicolas Vasilache authored Dec 11, 2019
  508d4e67
- NFC: Fix naming inconsistency: FuncOpLowering -> GPUFuncOpLowering. · c5fb4c13
  Christian Sigg authored Dec 11, 2019
  c5fb4c13
- Roll-forward initial liveness analysis including test cases. · 4b0198ac
  Alexander Belyaev authored Dec 11, 2019
  4b0198ac
- Automated rollback of commit 98fbf410 · 984fdde2
  Alexander Belyaev authored Dec 11, 2019
  984fdde2
- Add a function to get lowering patterns from GPU to NVVM. · b96f86da
  Stephan Herhut authored Dec 11, 2019
  b96f86da
- [Linalg] Add tiling for IndexedGenericOp with a region. · bae8a7a7
  Alexander Belyaev authored Dec 11, 2019
  bae8a7a7
- Add initial liveness analysis including test cases. · 98fbf410
  Marcel Koester authored Dec 11, 2019
  98fbf410
- [VectorOps] Add lowering of vector.insert to LLVM IR · 9826fe5c
  Aart Bik authored Dec 10, 2019
  9826fe5c
- Add VectorOp transform pattern which splits vector TransferReadOps to target vector unroll size. · 4d8ba886
  Andy Davis authored Dec 10, 2019
  4d8ba886
- More affine expr simplifications for floordiv and mod · 36a415bc
  Uday Bondhugula authored Dec 10, 2019
  36a415bc
Dec 10, 2019
- Move gpu.launch_func to ODS. NFC · d1213ae5
  Alex Zinenko authored Dec 10, 2019
  d1213ae5
- Fold TestLinalgTilePermutePatterns into TestLinalgTransformPatterns - NFC · 995048d7
  Nicolas Vasilache authored Dec 10, 2019
  995048d7
- Refactor the various operand/result/type iterators to use indexed_accessor_range. · 9ed22ae5
  River Riddle authored Dec 10, 2019
  9ed22ae5
- [Linalg] Add a Linalg iterator permutation transformation · b19fed54
  Jose Ignacio Gomez authored Dec 10, 2019
  b19fed54
- Uniformize Vector transforms as patterns on the model of Linalg - NFC · ad38e498
  Nicolas Vasilache authored Dec 10, 2019
  ad38e498
- Add Py API for composing an affine expression with a map. Also allows... · 8ccb3509
  MLIR Team authored Dec 10, 2019
  8ccb3509
- More convenience build methods for SPIR-V ops. · 04fdd33d
  Mahesh Ravishankar authored Dec 10, 2019
  04fdd33d
- Add a doc on guidelines for contributing a new dialect to the MLIR core repo · 90b72dd6
  Mehdi Amini authored Dec 10, 2019
  90b72dd6
- Drop Markdown style annotations · ac487332
  Alex Zinenko authored Dec 10, 2019
  ac487332
- Fix build breakage on gcc-5 · acb23ff4
  Jacques Pienaar authored Dec 09, 2019
  acb23ff4
- [VectorOps] Add a ShuffleOp to the VectorOps dialect · 1fe65688
  Aart Bik authored Dec 09, 2019
  1fe65688
- [VectorOps] Fix off-by-one error in insert/extract validation · 0e963b9c
  Aart Bik authored Dec 09, 2019
  0e963b9c
- Refactor the Block support classes. · 3f9744a6
  River Riddle authored Dec 09, 2019
  3f9744a6