)]}'
{
  "commit": "e0f79b8f1f3394bb344b7b83d6f121ac2af327de",
  "tree": "71fbe51998f2e22889bb59d361acf36898d71732",
  "parents": [
    "902d2e8ae0de29f483840ba1134af27343b9564d"
  ],
  "author": {
    "name": "Johannes Weiner",
    "email": "hannes@saeurebad.de",
    "time": "Sat Oct 18 20:26:55 2008 -0700"
  },
  "committer": {
    "name": "Linus Torvalds",
    "email": "torvalds@linux-foundation.org",
    "time": "Mon Oct 20 08:52:31 2008 -0700"
  },
  "message": "vmscan: don\u0027t accumulate scan pressure on unrelated lists\n\nDuring each reclaim scan we accumulate scan pressure on unrelated lists\nwhich will result in bogus scans and unwanted reclaims eventually.\n\nScanning lists with few reclaim candidates results in a lot of rotation\nand therefor also disturbs the list balancing, putting even more\npressure on the wrong lists.\n\nIn a test-case with much streaming IO, and therefor a crowded inactive\nfile page list, swapping started because\n\n  a) anon pages were reclaimed after swap_cluster_max reclaim\n  invocations -- nr_scan of this list has just accumulated\n\n  b) active file pages were scanned because *their* nr_scan has also\n  accumulated through the same logic.  And this in return created a\n  lot of rotation for file pages and resulted in a decrease of file\n  list priority, again increasing the pressure on anon pages.\n\nThe result was an evicted working set of anon pages while there were\ntons of inactive file pages that should have been taken instead.\n\nSigned-off-by: Johannes Weiner \u003channes@saeurebad.de\u003e\nReviewed-by: Rik van Riel \u003criel@redhat.com\u003e\nSigned-off-by: Andrew Morton \u003cakpm@linux-foundation.org\u003e\nSigned-off-by: Linus Torvalds \u003ctorvalds@linux-foundation.org\u003e\n",
  "tree_diff": [
    {
      "type": "modify",
      "old_id": "ca64e3e0c518965a81bc99516199978e16f6b469",
      "old_mode": 33188,
      "old_path": "mm/vmscan.c",
      "new_id": "412d7872fc75af2e33444470951029db21b95b6d",
      "new_mode": 33188,
      "new_path": "mm/vmscan.c"
    }
  ]
}