{"id":"CVE-2026-98079","published":"2026-09-25T11:17:37.307","lastModified":"2026-09-25T11:17:37.307","description":"In the Linux kernel, the following vulnerability has been resolved:\n\nbtrfs: zstd: fix lost wakeup when waiting for a workspace\n\nA writer can sleep forever in zstd_get_workspace() even though a workspace\nis free.  When zstd_alloc_workspace() fails, the task is queued on\nzwsm->wait and schedules unconditionally, never re-testing the pool.\nzstd_put_workspace() publishes the workspace and then calls cond_wake_up(),\nwhich only wakes when a sleeper is already visible, so a workspace returned\nbetween the failed allocation and prepare_to_wait() wakes nobody.  The\nwindow is wide: zstd_alloc_workspace() goes through kvmalloc() and may\nenter reclaim.\n\nOnly a max level workspace triggers the wakeup and one is deliberately kept\nallocated as the fallback every waiter waits for, so once its wakeup is\nlost the writer stays in TASK_UNINTERRUPTIBLE until some other task happens\nto return one.  Re-check the pool after prepare_to_wait() has published the\nwaiter, and use the workspace if one turned up.","cvssScore":null,"cvssSeverity":null,"cvssVector":null,"cwes":[],"vendors":[],"products":[],"references":[{"url":"https://git.kernel.org/stable/c/0de9f31d447ae71ab08c7850a321682f8d2907c3","tags":[]},{"url":"https://git.kernel.org/stable/c/2acb9f3d1cc8f65dc81ed55e238cbf8e5b60bff7","tags":[]},{"url":"https://git.kernel.org/stable/c/d8f57049521948df5f50797141472fbbed3b0723","tags":[]}],"exploitRefs":[],"hasPoc":false,"ai":null}