Commits · df9bd857b1f330a2636104cbbc228d5c0838d089 · Roger Ferrer / llvm-epi · GitLab

Mar 29, 2019
- Type system: replace Type::getBitWidth with getIntOrFloatBitWidth · df9bd857
  Alex Zinenko authored Dec 17, 2018
  df9bd857
- loop-unroll - add function callback argument for outside targets to · 4a3e4e8e
  Uday Bondhugula authored Dec 17, 2018
  4a3e4e8e
- Loop Fusion pass update: introduce utilities to perform generalized loop... · 3b69230b
  MLIR Team authored Dec 17, 2018
  3b69230b
- Fix builder getFloatAttr of double to use F64 type and use fltSemantics in FloatAttr. · 49c4d2a6
  Jacques Pienaar authored Dec 17, 2018
  49c4d2a6
- Free the name symbol in TableGen · 72159f5e
  Lei Zhang authored Dec 17, 2018
  72159f5e
- Remove duplicate code / reuse right utilities from memref-dep-check / loop-tile · dced746b
  Uday Bondhugula authored Dec 16, 2018
  dced746b
- 'memref-bound-check': extend to store op's as well · c41ee606
  Uday Bondhugula authored Dec 14, 2018
  c41ee606
- Extract vector_transfer_* Ops into a SuperVectorDialect. · bc52a639
  Alex Zinenko authored Dec 14, 2018
  bc52a639
- Fix asan failures in mlir-op-gen. · 30a30d20
  Jacques Pienaar authored Dec 14, 2018
  30a30d20
- Use dag instead of list for operands to allow named operands. · 7a62e356
  Jacques Pienaar authored Dec 14, 2018
  7a62e356
- Expression flattening improvement - reuse local expressions. · 45a0f525
  Uday Bondhugula authored Dec 13, 2018
  45a0f525
- Convert tf.FakeQuantWithMinMaxArgs/Vars to tfl.FakeQuant · b0c41e54
  Feng Liu authored Dec 13, 2018
  b0c41e54
- Define TFLite Dequantize and FakeQuant ops · a138c12c
  Feng Liu authored Dec 13, 2018
  a138c12c
- FlatAffineConstraints - complete TODOs: add method to remove duplicate / · 8365bdc1
  Uday Bondhugula authored Dec 13, 2018
  8365bdc1
- Check if the operation is already in the worklist before adding it. · 5c4f1fdd
  River Riddle authored Dec 13, 2018
  5c4f1fdd
- Fix loop unrolling test cases · 4860f0e8
  Uday Bondhugula authored Dec 12, 2018
  4860f0e8
- LLVM IR lowering: support 1D vector operations · 359835eb
  Alex Zinenko authored Dec 12, 2018
  359835eb
- ConvertToCFG: use affine_apply to implement loop steps · 97d2f3cd
  Alex Zinenko authored Dec 12, 2018
  97d2f3cd
- Add rudimentary pattern rewrite matching generation. · a2222a94
  Jacques Pienaar authored Dec 12, 2018
  a2222a94
- Remove dead code from FlatAffineConstraints · c86c4147
  Uday Bondhugula authored Dec 11, 2018
  c86c4147
- Generate another op builder with aggregated parameters · a9eb2e8f
  Lei Zhang authored Dec 11, 2018
  a9eb2e8f
- Disallow index types as elements of vector, memref and tensor types · 63261aa9
  Alex Zinenko authored Dec 11, 2018
  63261aa9
- Update/Fix LoopUtils::stmtBodySkew to handle loop step. · b9f53dc0
  Uday Bondhugula authored Dec 10, 2018
  b9f53dc0
- Fix missing check for dependent DMAs in pipeline-data-transfer · d59a95a0
  Uday Bondhugula authored Dec 10, 2018
  d59a95a0
- FlatAffineConstraints API cleanup; add normalizeConstraintsByGCD(). · 6757fb15
  Uday Bondhugula authored Dec 10, 2018
  6757fb15
- Update/fix -pipeline-data-transfer; fix b/120770946 · 2ef57806
  Uday Bondhugula authored Dec 10, 2018
  2ef57806
- Properly namespace createLowerAffineApply · 073c3ad9
  Alex Zinenko authored Dec 10, 2018
  073c3ad9
- [MLIR] Drop bug-prone global map indexed by MLFunction* · c28aeef9
  Nicolas Vasilache authored Dec 07, 2018
  c28aeef9
- Extend loop tiling utility to handle non-constant loop bounds and bounds that · 2d6478fa
  Uday Bondhugula authored Dec 07, 2018
  2d6478fa
- Generate strided DMAs from -dma-generate · dfc752e4
  Uday Bondhugula authored Dec 07, 2018
  dfc752e4
- [MLIR] Add LowerVectorTransfersPass · d9b6420f
  Nicolas Vasilache authored Dec 07, 2018
  d9b6420f
- [MLIR] Fix the name of the MaterializeVectorPass · 879be718
  Nicolas Vasilache authored Dec 07, 2018
  879be718
- [MLIR] Add composeWithUnboundedMap · db1b9f73
  Nicolas Vasilache authored Dec 07, 2018
  db1b9f73
- [MLIR] Add LangRef entries for vector_transfer ops · 692f6ffd
  Nicolas Vasilache authored Dec 07, 2018
  692f6ffd
- Return bool from all emitError methods similar to Operation::emitOpError · adca59e4
  Smit Hinsu authored Dec 07, 2018
  adca59e4
- Auto-generate op builder with TableGen · d2d7c11f
  Lei Zhang authored Dec 06, 2018
  d2d7c11f
- [MLIR] Drop assert for NYI in Vectorize.cpp · 13bc7704
  Nicolas Vasilache authored Dec 06, 2018
  13bc7704
- [MLIR] Drop assert for NYI in VectorAnalysis · 2408f0eb
  Nicolas Vasilache authored Dec 06, 2018
  2408f0eb
- [MLIR] Drop unnecessary mention of NYI. · 48d22e83
  Nicolas Vasilache authored Dec 06, 2018
  48d22e83
- [MLIR] Remove NYI assertions in LoopAnalysis.cpp · a019379c
  Nicolas Vasilache authored Dec 06, 2018
  a019379c