)]}'
{
  "commit": "cae2a4ba460540e89684919a50e56260bc943bbf",
  "tree": "ff0eb5de0eba86b8a55caaaa1288b0517fe2eb69",
  "parents": [
    "9f130806127d10ace29196cd880855d34b01bfed"
  ],
  "author": {
    "name": "Shakeel Butt",
    "email": "shakeelb@google.com",
    "time": "Fri Jan 12 19:47:43 2024 +0000"
  },
  "committer": {
    "name": "COS Cherry Picker",
    "email": "cloud-image-release@prod.google.com",
    "time": "Fri Jan 19 13:59:29 2024 -0800"
  },
  "message": "mm: writeback: ratelimit stat flush from mem_cgroup_wb_stats\n\nOne of our workloads (Postgres 14) has regressed when migrated from 5.10\nto 6.1 upstream kernel. The regression can be reproduced by sysbench\u0027s\noltp_write_only benchmark. It seems like the always on rstat flush in\nmem_cgroup_wb_stats() is causing the regression. So, rate limit that\nspecific rstat flush. One potential consequence would be the dirty\nthrottling might be decided on stale memcg stats. However from our\nbenchmarks and production traffic we have not observed any change in the\ndirty throttling behavior of the application.\n\nSigned-off-by: Shakeel Butt \u003cshakeelb@google.com\u003e\n\nBUG\u003db/246641795\nTEST\u003dpresubmit\nRELEASE_NOTE\u003dFixed a performance issue observed in Postgres databases.\n\ncos-patch: bug\nChange-Id: Ibc11a3db969b427fcac92a12d187c42f3410760b\nReviewed-on: https://cos-review.googlesource.com/c/third_party/kernel/+/63572\nReviewed-by: Oleksandr Tymoshenko \u003covt@google.com\u003e\nMain-Branch-Verified: Cusky Presubmit Bot \u003cpresubmit@cos-infra-prod.iam.gserviceaccount.com\u003e\nReviewed-by: Robert Kolchmeyer \u003crkolchmeyer@google.com\u003e\nTested-by: Cusky Presubmit Bot \u003cpresubmit@cos-infra-prod.iam.gserviceaccount.com\u003e\n",
  "tree_diff": [
    {
      "type": "modify",
      "old_id": "dacbaf4f7b2c4d6a763c996bd16162bdccb904d7",
      "old_mode": 33188,
      "old_path": "mm/memcontrol.c",
      "new_id": "db324006a403c39ca9c513b1df2511a18aadd1a3",
      "new_mode": 33188,
      "new_path": "mm/memcontrol.c"
    }
  ]
}