)]}'
{
  "commit": "04636dc410b163c2243e66c3813dd4900a50a4ed",
  "tree": "8d583d058d1ee45c9d58568aca28bb9913a9a9f3",
  "parents": [
    "4dd7b8d30cfa1aebee547958db27efd581a58d9b"
  ],
  "author": {
    "name": "Vladimir Sementsov-Ogievskiy",
    "email": "vsementsov@parallels.com",
    "time": "Tue Dec 30 13:04:16 2014 +0300"
  },
  "committer": {
    "name": "Stefan Hajnoczi",
    "email": "stefanha@redhat.com",
    "time": "Tue Jan 13 11:47:56 2015 +0000"
  },
  "message": "migration/block: fix pending() return value\n\nBecause of wrong return value of .save_live_pending() in\nmigration/block.c, migration finishes before the whole disk is\ntransferred. Such situation occurs when the migration process is fast\nenough, for example when source and dest are on the same host.\n\nIf in the bulk phase we return something \u003c max_size, we will skip\ntransferring the tail of the device. Currently we have \"set pending to\nBLOCK_SIZE if it is zero\" for bulk phase, but there no guarantee, that\nit will be \u003c max_size.\n\nTrue approach is to return, for example, max_size+1 when we are in the\nbulk phase.\n\nSigned-off-by: Vladimir Sementsov-Ogievskiy \u003cvsementsov@parallels.com\u003e\nMessage-id: 1419933856-4018-2-git-send-email-vsementsov@parallels.com\nSigned-off-by: Stefan Hajnoczi \u003cstefanha@redhat.com\u003e\n",
  "tree_diff": [
    {
      "type": "modify",
      "old_id": "a0f908cf98259cc216ae6f5071ff8c333f7d5b1c",
      "old_mode": 33188,
      "old_path": "migration/block.c",
      "new_id": "0c7610600be4953a077f8ec2069db39a576478a2",
      "new_mode": 33188,
      "new_path": "migration/block.c"
    }
  ]
}