Skip to content

Commit ff88df6

Browse files
Yihan Dingl0kod
authored andcommitted
landlock: Serialize TSYNC thread restriction
syzbot found a deadlock in landlock_restrict_sibling_threads(). When multiple threads concurrently call landlock_restrict_self() with sibling thread restriction enabled, they can deadlock by mutually queueing task_works on each other and then blocking in kernel space (waiting for the other to finish). Fix this by serializing the TSYNC operations within the same process using the exec_update_lock. This prevents concurrent invocations from deadlocking. We use down_write_trylock() and restart the syscall if the lock cannot be acquired immediately. This ensures that if a thread fails to get the lock, it will return to userspace, allowing it to process any pending TSYNC task_works from the lock holder, and then transparently restart the syscall. Fixes: 42fc7e6 ("landlock: Multithreading support for landlock_restrict_self()") Reported-by: [email protected] Closes: https://syzkaller.appspot.com/bug?extid=7ea2f5e9dfd468201817 Suggested-by: Günther Noack <[email protected]> Suggested-by: Tingmao Wang <[email protected]> Tested-by: Justin Suess <[email protected]> Signed-off-by: Yihan Ding <[email protected]> Tested-by: Günther Noack <[email protected]> Reviewed-by: Günther Noack <[email protected]> Link: https://lore.kernel.org/r/[email protected] Signed-off-by: Mickaël Salaün <[email protected]>
1 parent f8e2019 commit ff88df6

1 file changed

Lines changed: 11 additions & 1 deletion

File tree

security/landlock/tsync.c

Lines changed: 11 additions & 1 deletion
Original file line numberDiff line numberDiff line change
@@ -485,6 +485,16 @@ int landlock_restrict_sibling_threads(const struct cred *old_cred,
485485
shared_ctx.new_cred = new_cred;
486486
shared_ctx.set_no_new_privs = task_no_new_privs(current);
487487

488+
/*
489+
* Serialize concurrent TSYNC operations to prevent deadlocks when
490+
* multiple threads call landlock_restrict_self() simultaneously.
491+
* If the lock is already held, we gracefully yield by restarting the
492+
* syscall. This allows the current thread to process pending
493+
* task_works before retrying.
494+
*/
495+
if (!down_write_trylock(&current->signal->exec_update_lock))
496+
return restart_syscall();
497+
488498
/*
489499
* We schedule a pseudo-signal task_work for each of the calling task's
490500
* sibling threads. In the task work, each thread:
@@ -594,6 +604,6 @@ int landlock_restrict_sibling_threads(const struct cred *old_cred,
594604
wait_for_completion(&shared_ctx.all_finished);
595605

596606
tsync_works_release(&works);
597-
607+
up_write(&current->signal->exec_update_lock);
598608
return atomic_read(&shared_ctx.preparation_error);
599609
}

0 commit comments

Comments
 (0)