)]}'
{
  "commit": "bbdb396a60b2ebf7de3b717991e5d3e28c8b7bbd",
  "tree": "efeca197ae8a2421940006ace6c1b719b1811a0c",
  "parents": [
    "3bb1a852ab6c9cdf211a2f4a2f502340c8c38eca"
  ],
  "author": {
    "name": "Martin Bligh",
    "email": "mbligh@google.com",
    "time": "Sat Oct 28 10:38:25 2006 -0700"
  },
  "committer": {
    "name": "Linus Torvalds",
    "email": "torvalds@g5.osdl.org",
    "time": "Sat Oct 28 11:30:51 2006 -0700"
  },
  "message": "[PATCH] Use min of two prio settings in calculating distress for reclaim\n\nIf try_to_free_pages / balance_pgdat are called with a gfp_mask specifying\nGFP_IO and/or GFP_FS, they will reclaim the requisite number of pages, and the\nreset prev_priority to DEF_PRIORITY (or to some other high (ie: unurgent)\nvalue).\n\nHowever, another reclaimer without those gfp_mask flags set (say, GFP_NOIO)\nmay still be struggling to reclaim pages.  The concurrent overwrite of\nzone-\u003eprev_priority will cause this GFP_NOIO thread to unexpectedly cease\ndeactivating mapped pages, thus causing reclaim difficulties.\n\nFix this is to key the distress calculation not off zone-\u003eprev_priority, but\nalso take into account the local caller\u0027s priority by using\nmin(zone-\u003eprev_priority, sc-\u003epriority)\n\nSigned-off-by: Martin J. Bligh \u003cmbligh@google.com\u003e\nCc: Nick Piggin \u003cnickpiggin@yahoo.com.au\u003e\nCc: \u003cstable@kernel.org\u003e\nSigned-off-by: Andrew Morton \u003cakpm@osdl.org\u003e\nSigned-off-by: Linus Torvalds \u003ctorvalds@osdl.org\u003e\n",
  "tree_diff": [
    {
      "type": "modify",
      "old_id": "b32560ead5c0c518e0b3db5c2c1bbe802c1c441c",
      "old_mode": 33188,
      "old_path": "mm/vmscan.c",
      "new_id": "518540a4a2a66a1ade20312d1d655ba6923e05b5",
      "new_mode": 33188,
      "new_path": "mm/vmscan.c"
    }
  ]
}