)]}'
{
  "commit": "c9503e0fe052020e0294cd07d0ecd982eb7c9177",
  "tree": "aa0425fc1f30ebdf8d7455cf59db5c3ea2c7dfd0",
  "parents": [
    "2296e5a0136f7ba64c99f3a48a55a687aa9abcc8"
  ],
  "author": {
    "name": "Anton Blanchard",
    "email": "anton@samba.org",
    "time": "Mon Apr 27 05:42:24 2009 -0700"
  },
  "committer": {
    "name": "David S. Miller",
    "email": "davem@davemloft.net",
    "time": "Mon Apr 27 05:42:24 2009 -0700"
  },
  "message": "ipv4: Limit size of route cache hash table\n\nRight now we have no upper limit on the size of the route cache hash table.\nOn a 128GB POWER6 box it ends up as 32MB:\n\n    IP route cache hash table entries: 4194304 (order: 9, 33554432 bytes)\n\nIt would be nice to cap this for memory consumption reasons, but a massive\nhashtable also causes a significant spike when measuring OS jitter.\n\nWith a 32MB hashtable and 4 million entries, rt_worker_func is taking\n5 ms to complete. On another system with more memory it\u0027s taking 14 ms.\nEven though rt_worker_func does call cond_sched() to limit its impact,\nin an HPC environment we want to keep all sources of OS jitter to a minimum.\n\nWith the patch applied we limit the number of entries to 512k which\ncan still be overriden by using the rt_entries boot option:\n\n    IP route cache hash table entries: 524288 (order: 6, 4194304 bytes)\n\nWith this patch rt_worker_func now takes 0.460 ms on the same system.\n\nSigned-off-by: Anton Blanchard \u003canton@samba.org\u003e\nAcked-by: Eric Dumazet \u003cdada1@cosmosbay.com\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n",
  "tree_diff": [
    {
      "type": "modify",
      "old_id": "c40debe51b38ace1224d3c9b9342ad0c73513e1c",
      "old_mode": 33188,
      "old_path": "net/ipv4/route.c",
      "new_id": "c4c60e9f068acc221cb47079178df40299e3ef20",
      "new_mode": 33188,
      "new_path": "net/ipv4/route.c"
    }
  ]
}