workqueue: fix is_chained_work() regression
authorTejun Heo <tj@kernel.org>
Thu, 14 Feb 2013 03:29:07 +0000 (19:29 -0800)
committerTejun Heo <tj@kernel.org>
Thu, 14 Feb 2013 03:29:07 +0000 (19:29 -0800)
c9e7cf273f ("workqueue: move busy_hash from global_cwq to
worker_pool") incorrectly converted is_chained_work() to use
get_gcwq() inside for_each_gcwq_cpu() while removing get_gcwq().

As cwq might not exist for all possible workqueue CPUs, @cwq can be
NULL and the following cwq deferences can lead to oops.

Fix it by using for_each_cwq_cpu() instead, which is the better one to
use anyway as we only need to check pools that the wq is associated
with.

Signed-off-by: Tejun Heo <tj@kernel.org>
kernel/workqueue.c

index d6fdce12ca7ef0dc06189aaf73ae710560c2b381..0d26ab3aee59c36e47e7bf10a455ebd49b91989f 100644 (file)
@@ -1167,7 +1167,7 @@ static bool is_chained_work(struct workqueue_struct *wq)
        unsigned long flags;
        unsigned int cpu;
 
-       for_each_wq_cpu(cpu) {
+       for_each_cwq_cpu(cpu, wq) {
                struct cpu_workqueue_struct *cwq = get_cwq(cpu, wq);
                struct worker_pool *pool = cwq->pool;
                struct worker *worker;