)]}'
{
  "commit": "71a8c87fb300b601eacf7a86cc6c6322fe827bfd",
  "tree": "9f0e12933f6b2d74a452b69438693527853486c8",
  "parents": [
    "f1ccd2955157e1aff992f6aaaba0944209076220"
  ],
  "author": {
    "name": "Lachlan McIlroy",
    "email": "lachlan@sgi.com",
    "time": "Fri Sep 26 12:17:57 2008 +1000"
  },
  "committer": {
    "name": "Lachlan McIlroy",
    "email": "lachlan@redback.melbourne.sgi.com",
    "time": "Fri Sep 26 12:17:57 2008 +1000"
  },
  "message": "[XFS] Remove xfs_iext_irec_compact_full()\n\nYet another bug was found in xfs_iext_irec_compact_full() and while the\nsource of the bug was found it wasn\u0027t an easy task to track it down\nbecause the conditions are very difficult to reproduce.\n\nA HUGE thank-you goes to Russell Cattelan and Eric Sandeen for their\nsignificant effort in tracking down the source of this corruption.\n\nxfs_iext_irec_compact_full() and xfs_iext_irec_compact_pages() are almost\nidentical - they both compact indirect extent lists by moving extents from\nsubsequent buffers into earlier ones. xfs_iext_irec_compact_pages() only\nmoves extents if all of the extents in the next buffer will fit into the\nempty space in the buffer before it. xfs_iext_irec_compact_full() will go\na step further and move part of the next buffer if all the extents wont\nfit. It will then shift the remaining extents in the next buffer up to the\nstart of the buffer. The bug here was that we did not update er_extoff and\nthis caused extent list corruption.\n\nIt does not appear that this extra functionality gains us much. Calling\nxfs_iext_irec_compact_pages() instead will do a good enough job at\ncompacting the indirect list and will be quicker too.\n\nFor the case in xfs_iext_indirect_to_direct() the total number of extents\nin the indirect list will fit into one buffer so we will never need the\nextra functionality of xfs_iext_irec_compact_full() there.\n\nAlso xfs_iext_irec_compact_pages() doesn\u0027t need to do a memmove() (the\nbuffers will never overlap) so we don\u0027t want the performance hit that can\nincur.\n\nSGI-PV: 987159\n\nSGI-Modid: xfs-linux-melb:xfs-kern:32166a\n\nSigned-off-by: Lachlan McIlroy \u003clachlan@sgi.com\u003e\nSigned-off-by: Eric Sandeen \u003csandeen@sandeen.net\u003e\n",
  "tree_diff": [
    {
      "type": "modify",
      "old_id": "419cfc2eacb3ae522b00e7ce0750252947d8ab8b",
      "old_mode": 33188,
      "old_path": "fs/xfs/xfs_inode.c",
      "new_id": "dbd9cef852ece3d6d8b3b5b930626000c88bed06",
      "new_mode": 33188,
      "new_path": "fs/xfs/xfs_inode.c"
    }
  ]
}