)]}'
{
  "commit": "8deb765d39d4710f085dd276008c541a3f7d71f9",
  "tree": "25ae2fb040e792a218eea988fcfca92326233588",
  "parents": [
    "842466f5f94e7f2a3d8f5118d9b5db984b163baa"
  ],
  "author": {
    "name": "Gerda Zsejke More",
    "email": "gerdazsejke.more@arm.com",
    "time": "Mon Oct 20 09:03:39 2025 +0200"
  },
  "committer": {
    "name": "Jonathan Wright",
    "email": "jonathan.wright@arm.com",
    "time": "Mon Nov 17 10:05:46 2025 -0800"
  },
  "message": "Optimize SVE implementation of av1_warp_affine\n\nIn case of beta \u003d\u003d 0 and alpha \u003d\u003d 0 we know filter values before\nprocessing loops so add new logic using Neon USMMLA instructions,\nkeeping the SVE implementation for the remaining cases.\n\nBy permuting the input samples and the 6-tap filter we can use the\nArmv8.6 I8MM USMMLA matrix multiply instructions to accelerate\nhorizontal 6-tap convolutions. The 2x8 by 8x2 matrix multiply\ninstruction does twice the work of the USDOT dot product\ninstructions.\n\nIn case of 8-tap filter we can replace the USDOT instruction with\nUSMMLA, to apply a 7-tap filter, and an extra multiplication.\n\nChange-Id: Ia5df8a05512525f1eb4add4da4423c67aff2ca74\n",
  "tree_diff": [
    {
      "type": "modify",
      "old_id": "10aee35b1a6f579914edfac205a30301b318caf2",
      "old_mode": 33188,
      "old_path": "av1/common/arm/warp_plane_sve.c",
      "new_id": "455e29d12492802cc363022bd153f7663da76d6b",
      "new_mode": 33188,
      "new_path": "av1/common/arm/warp_plane_sve.c"
    }
  ]
}