forked from luck/tmp_suning_uos_patched
task_work: add a scheduling point in task_work_run()
It seems commit4a9d4b024a
("switch fput to task_work_add") re- introduced the problem addressed in944be0b224
("close_files(): add scheduling point") If a server process with a lot of files (say 2 million tcp sockets) is killed, we can spend a lot of time in task_work_run() and trigger a soft lockup. Signed-off-by: Eric Dumazet <edumazet@google.com> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
This commit is contained in:
parent
ec5da7f8dc
commit
f341861fb0
|
@ -75,6 +75,7 @@ void task_work_run(void)
|
|||
p = q->next;
|
||||
q->func(q);
|
||||
q = p;
|
||||
cond_resched();
|
||||
}
|
||||
}
|
||||
}
|
||||
|
|
Loading…
Reference in New Issue
Block a user