Techwizz/kernel-fxtec-pro1x
1
0
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity

cfq-iosched: properly protect ioc_gone and ioc count

Browse source 
If we have multiple tasks freeing cfq_io_contexts when cfq-iosched
is being unloaded, we could complete() ioc_gone twice. Fix that by
protecting ioc_gone complete() and clearing with a spinlock for
just that purpose. Doesn't matter from a performance perspective,
since it'll only enter that path when ioc_gone != NULL (when cfq-iosched
is being rmmod'ed).

Signed-off-by: Jens Axboe <jens.axboe@oracle.com>
This commit is contained in:
Jens Axboe 2008-05-29 09:32:08 +02:00
parent c461a97311
commit 9a11b4ed0e
1 changed files with 15 additions and 3 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

16
block/cfq-iosched.c
View file

@ -48,6 +48,7 @@ static struct kmem_cache *cfq_ioc_pool;
static DEFINE_PER_CPU(unsigned long, ioc_count);
static struct completion *ioc_gone;
static DEFINE_SPINLOCK(ioc_gone_lock);
#define CFQ_PRIO_LISTS IOPRIO_BE_NR
#define cfq_class_idle(cfqq) ((cfqq)->ioprio_class == IOPRIO_CLASS_IDLE)
@ -1177,8 +1178,19 @@ static void cfq_cic_free_rcu(struct rcu_head *head)
kmem_cache_free(cfq_ioc_pool, cic);
elv_ioc_count_dec(ioc_count);
if (ioc_gone && !elv_ioc_count_read(ioc_count))
if (ioc_gone) {
/*
* CFQ scheduler is exiting, grab exit lock and check
* the pending io context count. If it hits zero,
* complete ioc_gone and set it back to NULL
*/
spin_lock(&ioc_gone_lock);
if (ioc_gone && !elv_ioc_count_read(ioc_count)) {
complete(ioc_gone);
ioc_gone = NULL;
}
spin_unlock(&ioc_gone_lock);
}
}
static void cfq_cic_free(struct cfq_io_context *cic)
@ -2317,7 +2329,7 @@ static void __exit cfq_exit(void)
* pending RCU callbacks
*/
if (elv_ioc_count_read(ioc_count))
wait_for_completion(ioc_gone);
wait_for_completion(&all_gone);
cfq_slab_kill();
}