{"id":"CVE-2022-49702","summary":"btrfs: fix hang during unmount when block group reclaim task is running","details":"In the Linux kernel, the following vulnerability has been resolved:\n\nbtrfs: fix hang during unmount when block group reclaim task is running\n\nWhen we start an unmount, at close_ctree(), if we have the reclaim task\nrunning and in the middle of a data block group relocation, we can trigger\na deadlock when stopping an async reclaim task, producing a trace like the\nfollowing:\n\n[629724.498185] task:kworker/u16:7   state:D stack:    0 pid:681170 ppid:     2 flags:0x00004000\n[629724.499760] Workqueue: events_unbound btrfs_async_reclaim_metadata_space [btrfs]\n[629724.501267] Call Trace:\n[629724.501759]  \u003cTASK\u003e\n[629724.502174]  __schedule+0x3cb/0xed0\n[629724.502842]  schedule+0x4e/0xb0\n[629724.503447]  btrfs_wait_on_delayed_iputs+0x7c/0xc0 [btrfs]\n[629724.504534]  ? prepare_to_wait_exclusive+0xc0/0xc0\n[629724.505442]  flush_space+0x423/0x630 [btrfs]\n[629724.506296]  ? rcu_read_unlock_trace_special+0x20/0x50\n[629724.507259]  ? lock_release+0x220/0x4a0\n[629724.507932]  ? btrfs_get_alloc_profile+0xb3/0x290 [btrfs]\n[629724.508940]  ? do_raw_spin_unlock+0x4b/0xa0\n[629724.509688]  btrfs_async_reclaim_metadata_space+0x139/0x320 [btrfs]\n[629724.510922]  process_one_work+0x252/0x5a0\n[629724.511694]  ? process_one_work+0x5a0/0x5a0\n[629724.512508]  worker_thread+0x52/0x3b0\n[629724.513220]  ? process_one_work+0x5a0/0x5a0\n[629724.514021]  kthread+0xf2/0x120\n[629724.514627]  ? kthread_complete_and_exit+0x20/0x20\n[629724.515526]  ret_from_fork+0x22/0x30\n[629724.516236]  \u003c/TASK\u003e\n[629724.516694] task:umount          state:D stack:    0 pid:719055 ppid:695412 flags:0x00004000\n[629724.518269] Call Trace:\n[629724.518746]  \u003cTASK\u003e\n[629724.519160]  __schedule+0x3cb/0xed0\n[629724.519835]  schedule+0x4e/0xb0\n[629724.520467]  schedule_timeout+0xed/0x130\n[629724.521221]  ? lock_release+0x220/0x4a0\n[629724.521946]  ? lock_acquired+0x19c/0x420\n[629724.522662]  ? trace_hardirqs_on+0x1b/0xe0\n[629724.523411]  __wait_for_common+0xaf/0x1f0\n[629724.524189]  ? usleep_range_state+0xb0/0xb0\n[629724.524997]  __flush_work+0x26d/0x530\n[629724.525698]  ? flush_workqueue_prep_pwqs+0x140/0x140\n[629724.526580]  ? lock_acquire+0x1a0/0x310\n[629724.527324]  __cancel_work_timer+0x137/0x1c0\n[629724.528190]  close_ctree+0xfd/0x531 [btrfs]\n[629724.529000]  ? evict_inodes+0x166/0x1c0\n[629724.529510]  generic_shutdown_super+0x74/0x120\n[629724.530103]  kill_anon_super+0x14/0x30\n[629724.530611]  btrfs_kill_super+0x12/0x20 [btrfs]\n[629724.531246]  deactivate_locked_super+0x31/0xa0\n[629724.531817]  cleanup_mnt+0x147/0x1c0\n[629724.532319]  task_work_run+0x5c/0xa0\n[629724.532984]  exit_to_user_mode_prepare+0x1a6/0x1b0\n[629724.533598]  syscall_exit_to_user_mode+0x16/0x40\n[629724.534200]  do_syscall_64+0x48/0x90\n[629724.534667]  entry_SYSCALL_64_after_hwframe+0x44/0xae\n[629724.535318] RIP: 0033:0x7fa2b90437a7\n[629724.535804] RSP: 002b:00007ffe0b7e4458 EFLAGS: 00000246 ORIG_RAX: 00000000000000a6\n[629724.536912] RAX: 0000000000000000 RBX: 00007fa2b9182264 RCX: 00007fa2b90437a7\n[629724.538156] RDX: 0000000000000000 RSI: 0000000000000000 RDI: 0000555d6cf20dd0\n[629724.539053] RBP: 0000555d6cf20ba0 R08: 0000000000000000 R09: 00007ffe0b7e3200\n[629724.539956] R10: 0000000000000000 R11: 0000000000000246 R12: 0000000000000000\n[629724.540883] R13: 0000555d6cf20dd0 R14: 0000555d6cf20cb0 R15: 0000000000000000\n[629724.541796]  \u003c/TASK\u003e\n\nThis happens because:\n\n1) Before entering close_ctree() we have the async block group reclaim\n   task running and relocating a data block group;\n\n2) There's an async metadata (or data) space reclaim task running;\n\n3) We enter close_ctree() and park the cleaner kthread;\n\n4) The async space reclaim task is at flush_space() and runs all the\n   existing delayed iputs;\n\n5) Before the async space reclaim task calls\n   btrfs_wait_on_delayed_iputs(), the block group reclaim task which is\n   doing the data block group relocation, creates a delayed iput at\n   replace_file_extents() (called when COWing leaves that have file extent\n   items pointing to relocated data exten\n---truncated---","modified":"2026-04-02T08:27:50.218245Z","published":"2025-02-26T02:24:22.030Z","database_specific":{"cna_assigner":"Linux","osv_generated_from":"https://github.com/CVEProject/cvelistV5/tree/main/cves/2022/49xxx/CVE-2022-49702.json"},"references":[{"type":"WEB","url":"https://git.kernel.org/stable/c/31e70e527806c546a72262f2fc3d982ee23c42d3"},{"type":"WEB","url":"https://git.kernel.org/stable/c/341d33128a940c6634175dcb6ca92dc454cfa7d2"},{"type":"WEB","url":"https://git.kernel.org/stable/c/9fadb11f1295289e0da4d3342ecb6b92c1c99540"},{"type":"ADVISORY","url":"https://github.com/CVEProject/cvelistV5/tree/main/cves/2022/49xxx/CVE-2022-49702.json"},{"type":"ADVISORY","url":"https://nvd.nist.gov/vuln/detail/CVE-2022-49702"},{"type":"PACKAGE","url":"https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git"}],"affected":[{"ranges":[{"type":"GIT","repo":"https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git","events":[{"introduced":"18bb8bbf13c1839b43c9e09e76d397b753989af2"},{"fixed":"341d33128a940c6634175dcb6ca92dc454cfa7d2"},{"fixed":"9fadb11f1295289e0da4d3342ecb6b92c1c99540"},{"fixed":"31e70e527806c546a72262f2fc3d982ee23c42d3"}]}],"database_specific":{"source":"https://storage.googleapis.com/cve-osv-conversion/osv-output/CVE-2022-49702.json"}}],"schema_version":"1.7.5"}