)]}'
{
  "commit": "7a7fffefc3359fa42a4d63ee6329c589dfdd095c",
  "tree": "f087612e7f49fd76dfd130f0ba50c7689bbeb8c0",
  "parents": [
    "e72a2091da9bec24b24f3814244191f95482029b"
  ],
  "author": {
    "name": "Rupert Swarbrick",
    "email": "rupert.swarbrick@argondesign.com",
    "time": "Tue Nov 14 10:58:42 2017 +0000"
  },
  "committer": {
    "name": "Debargha Mukherjee",
    "email": "debargha@google.com",
    "time": "Tue Nov 14 17:34:59 2017 +0000"
  },
  "message": "Save right # of lines in save_deblock_boundary_lines\n\nThe \"src_height\" computed in save_deblock_boundary_lines didn\u0027t match\nthe one in save_tile_row_boundary_lines, which meant that the wrapper\nfunction assumed the deblock code was saving some lines and that code\nthought that save_cdef_boundary_lines would do it.\n\nThis patch fixes up the logic to match, and also completely gets rid\nof the lines_to_save variable (after all, bad things would happen if\nlines_to_save was 1 because we\u0027ll still read both boundary lines\nlater)\n\nThe tile height gets rounded up to a multiple of 8 luma pixels in\nsave_tile_row_boundary_lines to avoid nasty corner cases. This will\nonly have any effect for rows at the bottom of the frame (where\nav1_get_tile_rect clips to the frame boundary).\n\nBUG\u003daomedia:1020\n\nChange-Id: I55adb53fa8ba9c7f97fb2fd5b328a3f2f5065464\n",
  "tree_diff": [
    {
      "type": "modify",
      "old_id": "7f15a8296df13900b37eedd86fffb772e225a97d",
      "old_mode": 33188,
      "old_path": "av1/common/restoration.c",
      "new_id": "2868d9810119559d10755984d3971d7ba4b42bdb",
      "new_mode": 33188,
      "new_path": "av1/common/restoration.c"
    }
  ]
}