)]}'
{
  "commit": "2494725f08d9d4ece0475b933e16a0ae34a4165a",
  "tree": "79ec0c1156e4d183d6322c18a66da0adef52665f",
  "parents": [
    "5dabc01bba1df10de6fb3c32fcf9db4aafe78f38"
  ],
  "author": {
    "name": "Jay Foad",
    "email": "jay.foad@amd.com",
    "time": "Fri Nov 13 13:30:15 2026 +0000"
  },
  "committer": {
    "name": "Jay Foad",
    "email": "jay.foad@amd.com",
    "time": "Fri Nov 13 13:35:48 2026 +0000"
  },
  "message": "[Mips] Use IMPLICIT_DEF and INSERT_SUBREG to insert an FP value into an MSA vector. NFC\n\nThe lowering of INSERT_FW, INSERT_FD and the floating-point variants of\nINSERT_*_VIDX used SUBREG_TO_REG to view the FP register as an MSA\nregister. SUBREG_TO_REG means that the instruction defining the FP\nregister also defined the rest of the MSA register, which is not the\ncase. Only element 0 is used, so the rest is undefined, which is what\nIMPLICIT_DEF and INSERT_SUBREG express. This matches the lowering of\nFILL_FW and FILL_FD.\n\nCo-Authored-By: Claude Opus 5.5 \u003cnoreply@anthropic.com\u003e\n",
  "tree_diff": [
    {
      "type": "modify",
      "old_id": "6077c3f0732d2422c32b4acf21ecb7809f604ee8",
      "old_mode": 33188,
      "old_path": "llvm/lib/Target/Mips/MipsSEISelLowering.cpp",
      "new_id": "36f98d64ddb601696dc74ab99aae37449ffe8bcf",
      "new_mode": 33188,
      "new_path": "llvm/lib/Target/Mips/MipsSEISelLowering.cpp"
    }
  ]
}
