Commits · 13f6cc01876c39b106f249159e2993d81d6b04cb · Lorenzo Albano / LLVM bpEVL · GitLab

Mar 29, 2019
- Run GCD test before elimination. Adds test case with rational solutions, but no integer solutions. · 13f6cc01
  MLIR Team authored Oct 25, 2018
  13f6cc01
- Implement value type abstraction for attributes. · 792d1c25
  River Riddle authored Oct 25, 2018
  792d1c25
- Move transform utilities out to their own TransformUtils library, instead of · 64d52014
  Chris Lattner authored Oct 25, 2018
  64d52014
- Refactor the bulk of the worklist driver out of the canonicalizer into its own · 92285814
  Chris Lattner authored Oct 25, 2018
  92285814
- Add a doc explaining our approach to canonicalization, which is more of a · 144795e3
  Chris Lattner authored Oct 25, 2018
  144795e3
- Introduce Fourier-Motzkin variable elimination + other cleanup/support · 80610c2f
  Uday Bondhugula authored Oct 25, 2018
  80610c2f
- Adds Gaussian Elimination to FlatAffineConstraints. · 54132393
  MLIR Team authored Oct 24, 2018
  54132393
- Change typedef to using to be consistent across the codebase · 52a0e58b
  Lei Zhang authored Oct 24, 2018
  52a0e58b
- Rename shape_cast to tensor_cast. · e8d254b9
  Alex Zinenko authored Oct 24, 2018
  e8d254b9
- Add MLIR Rationale converted to g3doc. · c1b09186
  Alex Zinenko authored Oct 24, 2018
  c1b09186
- Add MLIR specification. · 1321f6af
  Alex Zinenko authored Oct 24, 2018
  1321f6af
- Teach canonicalize pass to unique and hoist constants to the entry block. This · bd01f954
  Chris Lattner authored Oct 24, 2018
  bd01f954
- Add support to opaque elements attributes · 3d7ab2d2
  Feng Liu authored Oct 23, 2018
  3d7ab2d2
- Implement shape folding in the canonicalization pass: · 301f83f9
  Chris Lattner authored Oct 23, 2018
  301f83f9
- Move the ReturnOp type checks to ReturnOp::verify. · 991adadc
  Alex Zinenko authored Oct 23, 2018
  991adadc
- Verify that the first block of a cfgfunc does not have predecessors. · d58ffaff
  Alex Zinenko authored Oct 23, 2018
  d58ffaff
- PassResult return cleanup. · ccfe5937
  Uday Bondhugula authored Oct 22, 2018
  ccfe5937
- Add a pattern (x+0) -> x, generalize Canonicalize to CFGFunc's, address a few TODOs, · a03051b9
  Chris Lattner authored Oct 22, 2018
  a03051b9
- Random cleanups: · 50cc57e2
  Chris Lattner authored Oct 22, 2018
  50cc57e2
- introduce a memref_cast operation, refactoring common code between it and · b2f93b27
  Chris Lattner authored Oct 22, 2018
  b2f93b27
- Introduce a new Operation::erase helper to generalize some code in · 7850258c
  Chris Lattner authored Oct 21, 2018
  7850258c
- Replace the "OperationSet" abstraction with a new Dialect abstraction. This is · 9eedf6ad
  Chris Lattner authored Oct 21, 2018
  9eedf6ad
- Introduce a new PatternRewriter class to help keep the worklist in · 73a80274
  Chris Lattner authored Oct 21, 2018
  73a80274
- Use APFloat for FloatAttribute · c5a3a5e4
  Feng Liu authored Oct 20, 2018
  c5a3a5e4
- Loop bound constant folding: follow-up / address comments from cl/215997346 · 2f1103bd
  Uday Bondhugula authored Oct 19, 2018
  2f1103bd
- Rename Operation::getAs to Operation::dyn_cast · 34927e24
  Feng Liu authored Oct 19, 2018
  34927e24
- Touch an unused variable. · 8c7478d1
  MLIR Team authored Oct 19, 2018
  8c7478d1
- Fix AffineExpr printing bug: paren ellision b/117887365. · a55b2c2e
  Uday Bondhugula authored Oct 18, 2018
  a55b2c2e
- Add support to constant sparse tensor / vector attribute · 03b48999
  Feng Liu authored Oct 18, 2018
  03b48999
- Add support to constant dense vector/tensor attribute. · b5b90e54
  Feng Liu authored Oct 18, 2018
  b5b90e54
- Generalize / improve DMA transfer overlap; nested and multiple DMA support; resolve · 18e66670
  Uday Bondhugula authored Oct 18, 2018
  18e66670
- [MLIR] Basic infrastructure for vectorization test · 3013dadb
  Nicolas Vasilache authored Oct 17, 2018
  3013dadb
- Use FuncBuilder instead of MLFuncBuilder in pattern matcher. · 47e7cd33
  Jacques Pienaar authored Oct 16, 2018
  47e7cd33
- Add constant folding and binary operator reassociation to the canonicalize · 80e884a9
  Chris Lattner authored Oct 16, 2018
  80e884a9
- [opgen] Change Attr to be more intuitive to insert · 58168e47
  Jacques Pienaar authored Oct 16, 2018
  58168e47
- Make the op specification more declarative wrt properties. · d4b095b1
  Jacques Pienaar authored Oct 16, 2018
  d4b095b1
- Move Pattern and related classes to a different file · 0faf5633
  Feng Liu authored Oct 15, 2018
  0faf5633
- Adds method to AffineApplyOp which forward substitutes its results into any of... · 0114e232
  MLIR Team authored Oct 15, 2018
  0114e232
- Add op gen tool to generate C++ classes for Operations. · f37a9909
  Jacques Pienaar authored Oct 15, 2018
  f37a9909
- Fail if operation name contains null char. · d8917d80
  Jacques Pienaar authored Oct 15, 2018
  d8917d80