Code

D2M Pass 4: Tensor Layout#2205

Code

Draft

jdesousa-TT wants to merge 12 commits intomaintenstorrent/tt-mlir:mainfrom jdesousa/ttir-tensor-layouttenstorrent/tt-mlir:jdesousa/ttir-tensor-layoutCopy head branch name to clipboard

+278

Commits

Commits on Feb 24, 2025

• wip

  jdesousa-TT

  committed
  b161fb2

  Copy full SHA for b161fb2

• wip

  jdesousa-TT

  committed
  541b5c6

  Copy full SHA for 541b5c6

• wip

  jdesousa-TT

  committed
  f1c9264

  Copy full SHA for f1c9264

• wip

  jdesousa-TT

  committed
  a5f3147

  Copy full SHA for a5f3147

• wip

  jdesousa-TT

  committed
  b94330b

  Copy full SHA for b94330b

• pre-commit fix

  jdesousa-TT

  committed
  ee6c4ed

  Copy full SHA for ee6c4ed

• Add tensor layout reduction test

  jdesousa-TT

  committed
  af1515f

  Copy full SHA for af1515f

• cleanup

  jdesousa-TT

  committed
  a44e67c

  Copy full SHA for a44e67c

• Group initial 3 tenor layout passes

  jdesousa-TT

  committed
  f861afa

  Copy full SHA for f861afa

• Use modifyOpInPlace for MemrefRewriter

  jdesousa-TT

  committed
  2c70b77

  Copy full SHA for 2c70b77

• Use modifyOpInPlace for GenericTensorLayout and FuncReturnTesnsorLayout passes

  jdesousa-TT

  committed
  16e4063

  Copy full SHA for 16e4063

• Update applyPatternsAndFoldGreedily after llvm uplift

  jdesousa-TT

  committed
  654de15

  Copy full SHA for 654de15