After hot-unplug a stressed disk, I found that rl->wait[] is not empty
while rl->count[] is empty and there are theads still sleeping on
get_request after the queue cleanup. With simple debug code, I found
there are exactly nr_sleep - nr_wakeup of theads in D state. So there
are missed wakeup.
$ dmesg | grep nr_sleep
[ 52.917115] ---> nr_sleep=1046, nr_wakeup=873, delta=173
$ vmstat 1
1 173 0 712640 24292 96172 0 0 0 0 419 757 0 0 0 100 0
To quote Tejun:
Ah, okay, freed_request() wakes up single waiter with the assumption
that after the wakeup there will at least be one successful allocation
which in turn will continue the wakeup chain until the wait list is
empty - ie. waiter wakeup is dependent on successful request
allocation happening after each wakeup. With queue marked dead, any
woken up waiter fails the allocation path, so the wakeup chaining is
lost and we're left with hung waiters. What we need is wake_up_all()
after drain completion.
This patch fixes the missed wakeup by waking up all the theads which
are sleeping on wait queue after queue drain.
Signed-off-by: Asias He <[email protected]>
---
block/blk-core.c | 11 ++++++++++-
1 file changed, 10 insertions(+), 1 deletion(-)
diff --git a/block/blk-core.c b/block/blk-core.c
index 1f61b74..1a45877 100644
--- a/block/blk-core.c
+++ b/block/blk-core.c
@@ -359,9 +359,10 @@ EXPORT_SYMBOL(blk_put_queue);
*/
void blk_drain_queue(struct request_queue *q, bool drain_all)
{
+ int i;
+
while (true) {
bool drain = false;
- int i;
spin_lock_irq(q->queue_lock);
@@ -400,6 +401,14 @@ void blk_drain_queue(struct request_queue *q, bool drain_all)
break;
msleep(10);
}
+
+ /* Wake up threads which are sleeping on get_request() */
+ spin_lock_irq(q->queue_lock);
+ for (i = 0; i < ARRAY_SIZE(q->rq.wait); i++) {
+ if (waitqueue_active(&q->rq.wait[i]))
+ wake_up_all(&q->rq.wait[i]);
+ }
+ spin_unlock_irq(q->queue_lock);
}
/**
--
1.7.10.2
On Mon, May 28, 2012 at 12:19:03PM +0800, Asias He wrote:
> diff --git a/block/blk-core.c b/block/blk-core.c
> index 1f61b74..1a45877 100644
> --- a/block/blk-core.c
> +++ b/block/blk-core.c
> @@ -359,9 +359,10 @@ EXPORT_SYMBOL(blk_put_queue);
> */
> void blk_drain_queue(struct request_queue *q, bool drain_all)
> {
> + int i;
> +
> while (true) {
> bool drain = false;
> - int i;
>
> spin_lock_irq(q->queue_lock);
>
> @@ -400,6 +401,14 @@ void blk_drain_queue(struct request_queue *q, bool drain_all)
> break;
> msleep(10);
> }
> +
> + /* Wake up threads which are sleeping on get_request() */
> + spin_lock_irq(q->queue_lock);
> + for (i = 0; i < ARRAY_SIZE(q->rq.wait); i++) {
> + if (waitqueue_active(&q->rq.wait[i]))
> + wake_up_all(&q->rq.wait[i]);
> + }
> + spin_unlock_irq(q->queue_lock);
I don't think we need waitqueue_active() optimization here. Let's
just do,
/* please explain why this is necessary */
for (i = 0; i < ARRAY_SIZE(q->rq.wait); i++)
wake_up_all(&q->rq.wait[i]);
Thanks.
--
tejun
After hot-unplug a stressed disk, I found that rl->wait[] is not empty
while rl->count[] is empty and there are theads still sleeping on
get_request after the queue cleanup. With simple debug code, I found
there are exactly nr_sleep - nr_wakeup of theads in D state. So there
are missed wakeup.
$ dmesg | grep nr_sleep
[ 52.917115] ---> nr_sleep=1046, nr_wakeup=873, delta=173
$ vmstat 1
1 173 0 712640 24292 96172 0 0 0 0 419 757 0 0 0 100 0
To quote Tejun:
Ah, okay, freed_request() wakes up single waiter with the assumption
that after the wakeup there will at least be one successful allocation
which in turn will continue the wakeup chain until the wait list is
empty - ie. waiter wakeup is dependent on successful request
allocation happening after each wakeup. With queue marked dead, any
woken up waiter fails the allocation path, so the wakeup chaining is
lost and we're left with hung waiters. What we need is wake_up_all()
after drain completion.
This patch fixes the missed wakeup by waking up all the theads which
are sleeping on wait queue after queue drain.
Changes in v2: Drop waitqueue_active() optimization
Signed-off-by: Asias He <[email protected]>
---
block/blk-core.c | 13 ++++++++++++-
1 file changed, 12 insertions(+), 1 deletion(-)
diff --git a/block/blk-core.c b/block/blk-core.c
index 1f61b74..232fe89 100644
--- a/block/blk-core.c
+++ b/block/blk-core.c
@@ -359,9 +359,10 @@ EXPORT_SYMBOL(blk_put_queue);
*/
void blk_drain_queue(struct request_queue *q, bool drain_all)
{
+ int i;
+
while (true) {
bool drain = false;
- int i;
spin_lock_irq(q->queue_lock);
@@ -400,6 +401,16 @@ void blk_drain_queue(struct request_queue *q, bool drain_all)
break;
msleep(10);
}
+
+ /*
+ * With queue marked dead, any woken up waiter will fail the
+ * allocation path, so the wakeup chaining is lost and we're
+ * left with hung waiters. We need to wake up those waiters.
+ */
+ spin_lock_irq(q->queue_lock);
+ for (i = 0; i < ARRAY_SIZE(q->rq.wait); i++)
+ wake_up_all(&q->rq.wait[i]);
+ spin_unlock_irq(q->queue_lock);
}
/**
--
1.7.10.2
On 05/28/2012 06:26 PM, Tejun Heo wrote:
> On Mon, May 28, 2012 at 12:19:03PM +0800, Asias He wrote:
>> diff --git a/block/blk-core.c b/block/blk-core.c
>> index 1f61b74..1a45877 100644
>> --- a/block/blk-core.c
>> +++ b/block/blk-core.c
>> @@ -359,9 +359,10 @@ EXPORT_SYMBOL(blk_put_queue);
>> */
>> void blk_drain_queue(struct request_queue *q, bool drain_all)
>> {
>> + int i;
>> +
>> while (true) {
>> bool drain = false;
>> - int i;
>>
>> spin_lock_irq(q->queue_lock);
>>
>> @@ -400,6 +401,14 @@ void blk_drain_queue(struct request_queue *q, bool drain_all)
>> break;
>> msleep(10);
>> }
>> +
>> + /* Wake up threads which are sleeping on get_request() */
>> + spin_lock_irq(q->queue_lock);
>> + for (i = 0; i< ARRAY_SIZE(q->rq.wait); i++) {
>> + if (waitqueue_active(&q->rq.wait[i]))
>> + wake_up_all(&q->rq.wait[i]);
>> + }
>> + spin_unlock_irq(q->queue_lock);
>
> I don't think we need waitqueue_active() optimization here. Let's
> just do,
>
> /* please explain why this is necessary */
> for (i = 0; i< ARRAY_SIZE(q->rq.wait); i++)
> wake_up_all(&q->rq.wait[i]);
OK. V2 is on its way.
--
Asias
On Tue, May 29, 2012 at 09:15:00AM +0800, Asias He wrote:
> After hot-unplug a stressed disk, I found that rl->wait[] is not empty
> while rl->count[] is empty and there are theads still sleeping on
> get_request after the queue cleanup. With simple debug code, I found
> there are exactly nr_sleep - nr_wakeup of theads in D state. So there
> are missed wakeup.
>
> $ dmesg | grep nr_sleep
> [ 52.917115] ---> nr_sleep=1046, nr_wakeup=873, delta=173
> $ vmstat 1
> 1 173 0 712640 24292 96172 0 0 0 0 419 757 0 0 0 100 0
>
> To quote Tejun:
>
> Ah, okay, freed_request() wakes up single waiter with the assumption
> that after the wakeup there will at least be one successful allocation
> which in turn will continue the wakeup chain until the wait list is
> empty - ie. waiter wakeup is dependent on successful request
> allocation happening after each wakeup. With queue marked dead, any
> woken up waiter fails the allocation path, so the wakeup chaining is
> lost and we're left with hung waiters. What we need is wake_up_all()
> after drain completion.
>
> This patch fixes the missed wakeup by waking up all the theads which
> are sleeping on wait queue after queue drain.
>
> Changes in v2: Drop waitqueue_active() optimization
>
> Signed-off-by: Asias He <[email protected]>
Acked-by: Tejun Heo <[email protected]>
Jens, this one wants Cc: stable.
Thanks.
--
tejun
Hello, Jens
On 05/29/2012 09:21 AM, Tejun Heo wrote:
> On Tue, May 29, 2012 at 09:15:00AM +0800, Asias He wrote:
>> After hot-unplug a stressed disk, I found that rl->wait[] is not empty
>> while rl->count[] is empty and there are theads still sleeping on
>> get_request after the queue cleanup. With simple debug code, I found
>> there are exactly nr_sleep - nr_wakeup of theads in D state. So there
>> are missed wakeup.
>>
>> $ dmesg | grep nr_sleep
>> [ 52.917115] ---> nr_sleep=1046, nr_wakeup=873, delta=173
>> $ vmstat 1
>> 1 173 0 712640 24292 96172 0 0 0 0 419 757 0 0 0 100 0
>>
>> To quote Tejun:
>>
>> Ah, okay, freed_request() wakes up single waiter with the assumption
>> that after the wakeup there will at least be one successful allocation
>> which in turn will continue the wakeup chain until the wait list is
>> empty - ie. waiter wakeup is dependent on successful request
>> allocation happening after each wakeup. With queue marked dead, any
>> woken up waiter fails the allocation path, so the wakeup chaining is
>> lost and we're left with hung waiters. What we need is wake_up_all()
>> after drain completion.
>>
>> This patch fixes the missed wakeup by waking up all the theads which
>> are sleeping on wait queue after queue drain.
>>
>> Changes in v2: Drop waitqueue_active() optimization
>>
>> Signed-off-by: Asias He<[email protected]>
>
> Acked-by: Tejun Heo<[email protected]>
>
> Jens, this one wants Cc: stable.
Ping.
--
Asias