)]}'
{
  "commit": "dcce240ead802d42b1e45ad2fcb2ed4a399cb255",
  "tree": "f9531f06ad6e0987d3c6782cfe0a7ed335d82e48",
  "parents": [
    "b36930dd508e00f0c5083bcd57d25de6d0375c76"
  ],
  "author": {
    "name": "Steven Whitehouse",
    "email": "swhiteho@redhat.com",
    "time": "Fri Nov 12 12:12:29 2010 +0000"
  },
  "committer": {
    "name": "David Teigland",
    "email": "teigland@redhat.com",
    "time": "Fri Nov 12 11:08:03 2010 -0600"
  },
  "message": "dlm: Use cmwq for send and receive workqueues\n\nSo far as I can tell, there is no reason to use a single-threaded\nsend workqueue for dlm, since it may need to send to several sockets\nconcurrently. Both workqueues are set to WQ_MEM_RECLAIM to avoid\nany possible deadlocks, WQ_HIGHPRI since locking traffic is highly\nlatency sensitive (and to avoid a priority inversion wrt GFS2\u0027s\nglock_workqueue) and WQ_FREEZABLE just in case someone needs to do\nthat (even though with current cluster infrastructure, it doesn\u0027t\nmake sense as the node will most likely land up ejected from the\ncluster) in the future.\n\nSigned-off-by: Steven Whitehouse \u003cswhiteho@redhat.com\u003e\nCc: Tejun Heo \u003ctj@kernel.org\u003e\nSigned-off-by: David Teigland \u003cteigland@redhat.com\u003e\n",
  "tree_diff": [
    {
      "type": "modify",
      "old_id": "77720f89c879a091af3cc027f9bdb295f3fcc339",
      "old_mode": 33188,
      "old_path": "fs/dlm/lowcomms.c",
      "new_id": "1d4e644c6589323b1e9a94fa295d604fd67d50ee",
      "new_mode": 33188,
      "new_path": "fs/dlm/lowcomms.c"
    }
  ]
}