)]}'
{
  "commit": "e96e3c004011550c6b4ecc1520e0f022984c4b17",
  "tree": "7789c2739d0ba2b030ccb617d06ad5d230567dee",
  "parents": [
    "aef6e1fc8fe85f4a57bc06b0272ada18975494fb"
  ],
  "author": {
    "name": "Stanley Winata",
    "email": "68087699+raikonenfnu@users.noreply.github.com",
    "time": "Thu Oct 24 16:56:38 2024 -0700"
  },
  "committer": {
    "name": "GitHub",
    "email": "noreply@github.com",
    "time": "Thu Oct 24 16:56:38 2024 -0700"
  },
  "message": "[VectorLayout] Fix insertion of new constOp for non dominate issue. (#18894)\n\nMain motivation of this patch is to resolve issue where we have the same\r\nconstOp being used by multiple operations.\r\n\r\nBut with a twist where first time the constOp needs a layout is on a op\r\nthat topologically comes after other ops that use constOp. This will\r\ngenerate a copy of constOp in the location right before the latter op,\r\nwhich is problematic because this constOp will be used by other ops\r\nbefore it.\r\n\r\nPreviously for the test added we get this error:\r\n```\r\nwithin split at contraction_layout.mlir:1 offset :24:10: note: see current operation: %9 \u003d \"arith.addf\"(%8, %6) \u003c{fastmath \u003d #arith.fastmath\u003cnone\u003e}\u003e : (vector\u003c96x64xf16\u003e, vector\u003c96x64xf16\u003e) -\u003e vector\u003c96x64xf16\u003e\r\nwithin split at contraction_layout.mlir:1 offset :22:19: error: operand #1 does not dominate this use\r\n    %scaled_rhs \u003d arith.mulf %read_1, %cst_1 : vector\u003c96x64xf16\u003e\r\n```\r\n\r\nWhile minor, this is also problematic because this error seem to stopped\r\nlayout analysis (but not fatally) S.T it fails to vector distribute in\r\nsome cases, making it hard to debug.\r\n\r\nSigned-off-by: Stanley Winata \u003cstanley.winata@amd.com\u003e",
  "tree_diff": [
    {
      "type": "modify",
      "old_id": "da40806ac73c729ee59cc94a476aa5aecf621f52",
      "old_mode": 33188,
      "old_path": "compiler/src/iree/compiler/Codegen/Common/GPU/test/gpu_vector_distribution.mlir",
      "new_id": "cf47ca9d47b5d422d18f3e887a4a96ce2ab6e76d",
      "new_mode": 33188,
      "new_path": "compiler/src/iree/compiler/Codegen/Common/GPU/test/gpu_vector_distribution.mlir"
    },
    {
      "type": "modify",
      "old_id": "fa3786caf61dfd63d9d4fd8d5a838f91a18d45b1",
      "old_mode": 33188,
      "old_path": "compiler/src/iree/compiler/Codegen/Common/VectorLayoutAnalysis.cpp",
      "new_id": "deabc58165fb18abed9a76b6c38f7aa862ed0ea2",
      "new_mode": 33188,
      "new_path": "compiler/src/iree/compiler/Codegen/Common/VectorLayoutAnalysis.cpp"
    },
    {
      "type": "modify",
      "old_id": "09c4d2787bf120cdc622481aad3c173d9d9b8bbb",
      "old_mode": 33188,
      "old_path": "compiler/src/iree/compiler/Codegen/Common/test/vector_layout_analysis.mlir",
      "new_id": "28e8ab01d89fa78ee33499d6ab1433f9134665fd",
      "new_mode": 33188,
      "new_path": "compiler/src/iree/compiler/Codegen/Common/test/vector_layout_analysis.mlir"
    }
  ]
}