)]}'
{
  "commit": "eb4866d0066ffd5446751c102d64feb3318d8bd1",
  "tree": "a40325308b621362938facb01b729882098e9cc2",
  "parents": [
    "38867a28a7dc9d69389990bcd42f6b7c77da3d9d"
  ],
  "author": {
    "name": "Dave Hansen",
    "email": "dave@linux.vnet.ibm.com",
    "time": "Tue Sep 20 15:19:38 2011 -0700"
  },
  "committer": {
    "name": "Linus Torvalds",
    "email": "torvalds@linux-foundation.org",
    "time": "Wed Sep 21 13:15:44 2011 -0700"
  },
  "message": "make /proc/$pid/numa_maps gather_stats() take variable page size\n\nWe need to teach the numa_maps code about transparent huge pages.  The\nfirst step is to teach gather_stats() that the pte it is dealing with\nmight represent more than one page.\n\nNote that will we use this in a moment for transparent huge pages since\nthey have use a single pmd_t which _acts_ as a \"surrogate\" for a bunch\nof smaller pte_t\u0027s.\n\nI\u0027m a _bit_ unhappy that this interface counts in hugetlbfs page sizes\nfor hugetlbfs pages and PAGE_SIZE for normal pages.  That means that to\nfigure out how many _bytes_ \"dirty\u003d1\" means, you must first know the\nhugetlbfs page size.  That\u0027s easier said than done especially if you\ndon\u0027t have visibility in to the mount.\n\nBut, that\u0027s probably a discussion for another day especially since it\nwould change behavior to fix it.  But, just in case anyone wonders why\nthis patch only passes a \u00271\u0027 in the hugetlb case...\n\nSigned-off-by: Dave Hansen \u003cdave@linux.vnet.ibm.com\u003e\nAcked-by: Hugh Dickins \u003chughd@google.com\u003e\nAcked-by: David Rientjes \u003crientjes@google.com\u003e\nSigned-off-by: Linus Torvalds \u003ctorvalds@linux-foundation.org\u003e\n",
  "tree_diff": [
    {
      "type": "modify",
      "old_id": "25b6a887adb916a93d20aa092481f67df451ff8c",
      "old_mode": 33188,
      "old_path": "fs/proc/task_mmu.c",
      "new_id": "61342a454bd94931f308af3c22ebcfdf7b92c6c0",
      "new_mode": 33188,
      "new_path": "fs/proc/task_mmu.c"
    }
  ]
}