)]}'
{
  "commit": "9426414f0d42f824892ecd4dccfebf8987084a41",
  "tree": "8ef299818292d567a88da80cf69905a7956a9712",
  "parents": [
    "8f5ae30d69d7543eee0d70083daf4de8fe15d585",
    "0cee64c547e3c9cda646af3e075a64f445ee8148"
  ],
  "author": {
    "name": "Christian Brauner",
    "email": "brauner@kernel.org",
    "time": "Mon Sep 15 14:50:03 2025 +0200"
  },
  "committer": {
    "name": "Christian Brauner",
    "email": "brauner@kernel.org",
    "time": "Fri Sep 19 13:11:06 2025 +0200"
  },
  "message": "Merge patch series \"writeback: Avoid lockups when switching inodes\"\n\nJan Kara \u003cjack@suse.cz\u003e says:\n\nThis patch series addresses lockups reported by users when systemd unit reading\nlots of files from a filesystem mounted with lazytime mount option exits. See\npatch 3 for more details about the reproducer.\n\nThere are two main issues why switching many inodes between wbs:\n\n1) Multiple workers will be spawned to do the switching but they all contend\non the same wb-\u003elist_lock making all the parallelism pointless and just\nwasting time.\n\n2) Sorting of wb-\u003eb_dirty list by dirtied_time_when is inherently slow.\n\nPatches 1-3 address these problems, patch 4 adds a tracepoint for better\nobservability of inode writeback switching.\n\n* patches from https://lore.kernel.org/20250912103522.2935-1-jack@suse.cz:\n  writeback: Add tracepoint to track pending inode switches\n  writeback: Avoid excessively long inode switching times\n  writeback: Avoid softlockup when switching many inodes\n  writeback: Avoid contention on wb-\u003elist_lock when switching inodes\n\nSigned-off-by: Christian Brauner \u003cbrauner@kernel.org\u003e\n",
  "tree_diff": []
}