[PATCH] Use list_splice for blk_abort_queue

[PATCH] Use list_splice for blk_abort_queue

[Hannes Reinecke]

blk_abort_queue takes the queue_lock with spinlock_irqsave and walks
the timer_list with list_for_each_entry_safe. Since a path to a SCSI
device just failed, the rport state is FC_PORTSTATE_BLOCKED. This
rport state triggers blk_add_timer(), which in turn calls
list_add_tail() to move the request to the end of timer_list. Thus,
the list_for_each_XXX loop never reaches the end of the timer_list.
By using list_splice_init() we're breaking the loop and process only
the requests present on the timeout_list when blk_abort_queue() is
called.

Signed-off-by: Hannes Reinecke <hare@suse.de>
Tested-by: Christof Schmitt <christof.schmitt@de.ibm.com>
---
 block/blk-timeout.c |    5 ++++-
 1 files changed, 4 insertions(+), 1 deletions(-)

diff --git a/block/blk-timeout.c b/block/blk-timeout.c
index a095353..67bcc3f 100644
--- a/block/blk-timeout.c
+++ b/block/blk-timeout.c
@@ -209,12 +209,15 @@ void blk_abort_queue(struct request_queue *q)
 {
 	unsigned long flags;
 	struct request *rq, *tmp;
+	LIST_HEAD(list);
 
 	spin_lock_irqsave(q->queue_lock, flags);
 
 	elv_abort_queue(q);
 
-	list_for_each_entry_safe(rq, tmp, &q->timeout_list, timeout_list)
+	list_splice_init(&q->timeout_list, &list);
+
+	list_for_each_entry_safe(rq, tmp, &list, timeout_list)
 		blk_abort_request(rq);
 
 	spin_unlock_irqrestore(q->queue_lock, flags);
-- 
1.5.3.2

Re: [PATCH] Use list_splice for blk_abort_queue

[Jens Axboe]

On Fri, Feb 27 2009, Hannes Reinecke wrote:
> 
> blk_abort_queue takes the queue_lock with spinlock_irqsave and walks
> the timer_list with list_for_each_entry_safe. Since a path to a SCSI
> device just failed, the rport state is FC_PORTSTATE_BLOCKED. This
> rport state triggers blk_add_timer(), which in turn calls
> list_add_tail() to move the request to the end of timer_list. Thus,
> the list_for_each_XXX loop never reaches the end of the timer_list.
> By using list_splice_init() we're breaking the loop and process only
> the requests present on the timeout_list when blk_abort_queue() is
> called.
> 
> Signed-off-by: Hannes Reinecke <hare@suse.de>
> Tested-by: Christof Schmitt <christof.schmitt@de.ibm.com>
> ---
>  block/blk-timeout.c |    5 ++++-
>  1 files changed, 4 insertions(+), 1 deletions(-)
> 
> diff --git a/block/blk-timeout.c b/block/blk-timeout.c
> index a095353..67bcc3f 100644
> --- a/block/blk-timeout.c
> +++ b/block/blk-timeout.c
> @@ -209,12 +209,15 @@ void blk_abort_queue(struct request_queue *q)
>  {
>  	unsigned long flags;
>  	struct request *rq, *tmp;
> +	LIST_HEAD(list);
>  
>  	spin_lock_irqsave(q->queue_lock, flags);
>  
>  	elv_abort_queue(q);
>  
> -	list_for_each_entry_safe(rq, tmp, &q->timeout_list, timeout_list)
> +	list_splice_init(&q->timeout_list, &list);
> +
> +	list_for_each_entry_safe(rq, tmp, &list, timeout_list)
>  		blk_abort_request(rq);
>  
>  	spin_unlock_irqrestore(q->queue_lock, flags);
> -- 
> 1.5.3.2
> 

Hannes, this fix has been in mainline for about 10 days.

-- 
Jens Axboe

Re: [PATCH] Use list_splice for blk_abort_queue

[Hannes Reinecke]

Hi Jens,

Jens Axboe wrote:
> On Fri, Feb 27 2009, Hannes Reinecke wrote:
>> blk_abort_queue takes the queue_lock with spinlock_irqsave and walks
>> the timer_list with list_for_each_entry_safe. Since a path to a SCSI
>> device just failed, the rport state is FC_PORTSTATE_BLOCKED. This
>> rport state triggers blk_add_timer(), which in turn calls
>> list_add_tail() to move the request to the end of timer_list. Thus,
>> the list_for_each_XXX loop never reaches the end of the timer_list.
>> By using list_splice_init() we're breaking the loop and process only
>> the requests present on the timeout_list when blk_abort_queue() is
>> called.
>>
>> Signed-off-by: Hannes Reinecke <hare@suse.de>
>> Tested-by: Christof Schmitt <christof.schmitt@de.ibm.com>
>> ---
>>  block/blk-timeout.c |    5 ++++-
>>  1 files changed, 4 insertions(+), 1 deletions(-)
>>
>> diff --git a/block/blk-timeout.c b/block/blk-timeout.c
>> index a095353..67bcc3f 100644
>> --- a/block/blk-timeout.c
>> +++ b/block/blk-timeout.c
>> @@ -209,12 +209,15 @@ void blk_abort_queue(struct request_queue *q)
>>  {
>>  	unsigned long flags;
>>  	struct request *rq, *tmp;
>> +	LIST_HEAD(list);
>>  
>>  	spin_lock_irqsave(q->queue_lock, flags);
>>  
>>  	elv_abort_queue(q);
>>  
>> -	list_for_each_entry_safe(rq, tmp, &q->timeout_list, timeout_list)
>> +	list_splice_init(&q->timeout_list, &list);
>> +
>> +	list_for_each_entry_safe(rq, tmp, &list, timeout_list)
>>  		blk_abort_request(rq);
>>  
>>  	spin_unlock_irqrestore(q->queue_lock, flags);
>> -- 
>> 1.5.3.2
>>
> 
> Hannes, this fix has been in mainline for about 10 days.
> 
Ah, sorry. Missed that one.
I'm just about to catch up with unread mails, and you said something
about 'the original author should resend the patch'.
Which I did.

Oh well, sorry for the noise.

Cheers,

Hannes
-- 
Dr. Hannes Reinecke		      zSeries & Storage
hare@suse.de			      +49 911 74053 688
SUSE LINUX Products GmbH, Maxfeldstr. 5, 90409 Nürnberg
GF: Markus Rex, HRB 16746 (AG Nürnberg)
--
To unsubscribe from this list: send the line "unsubscribe linux-scsi" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Re: [PATCH] Use list_splice for blk_abort_queue

[Jens Axboe]

On Fri, Feb 27 2009, Hannes Reinecke wrote:
> Hi Jens,
>
> Jens Axboe wrote:
>> On Fri, Feb 27 2009, Hannes Reinecke wrote:
>>> blk_abort_queue takes the queue_lock with spinlock_irqsave and walks
>>> the timer_list with list_for_each_entry_safe. Since a path to a SCSI
>>> device just failed, the rport state is FC_PORTSTATE_BLOCKED. This
>>> rport state triggers blk_add_timer(), which in turn calls
>>> list_add_tail() to move the request to the end of timer_list. Thus,
>>> the list_for_each_XXX loop never reaches the end of the timer_list.
>>> By using list_splice_init() we're breaking the loop and process only
>>> the requests present on the timeout_list when blk_abort_queue() is
>>> called.
>>>
>>> Signed-off-by: Hannes Reinecke <hare@suse.de>
>>> Tested-by: Christof Schmitt <christof.schmitt@de.ibm.com>
>>> ---
>>>  block/blk-timeout.c |    5 ++++-
>>>  1 files changed, 4 insertions(+), 1 deletions(-)
>>>
>>> diff --git a/block/blk-timeout.c b/block/blk-timeout.c
>>> index a095353..67bcc3f 100644
>>> --- a/block/blk-timeout.c
>>> +++ b/block/blk-timeout.c
>>> @@ -209,12 +209,15 @@ void blk_abort_queue(struct request_queue *q)
>>>  {
>>>  	unsigned long flags;
>>>  	struct request *rq, *tmp;
>>> +	LIST_HEAD(list);
>>>
>>>  	spin_lock_irqsave(q->queue_lock, flags);
>>>
>>>  	elv_abort_queue(q);
>>>
>>> -	list_for_each_entry_safe(rq, tmp, &q->timeout_list, timeout_list)
>>> +	list_splice_init(&q->timeout_list, &list);
>>> +
>>> +	list_for_each_entry_safe(rq, tmp, &list, timeout_list)
>>>  		blk_abort_request(rq);
>>>
>>>  	spin_unlock_irqrestore(q->queue_lock, flags);
>>> --
>>> 1.5.3.2
>>>
>>
>> Hannes, this fix has been in mainline for about 10 days.
>>
> Ah, sorry. Missed that one.
> I'm just about to catch up with unread mails, and you said something
> about 'the original author should resend the patch'.
> Which I did.

Yep, I did say that. But after a few days I figured you were probably
away or something, so I hacked up a changelog myself and put it in
there. It was already tested good and it makes sense.

-- 
Jens Axboe

Re: [PATCH] Use list_splice for blk_abort_queue

[Hannes Reinecke]

Jens Axboe wrote:
> On Fri, Feb 27 2009, Hannes Reinecke wrote:
>> Hi Jens,
>>
>> Jens Axboe wrote:
[ .. ]
>>> Hannes, this fix has been in mainline for about 10 days.
>>>
>> Ah, sorry. Missed that one.
>> I'm just about to catch up with unread mails, and you said something
>> about 'the original author should resend the patch'.
>> Which I did.
> 
> Yep, I did say that. But after a few days I figured you were probably
> away or something, so I hacked up a changelog myself and put it in
> there. It was already tested good and it makes sense.
> 
Yes, indeed 'something'. Thanks.

Cheers,

Hannes
-- 
Dr. Hannes Reinecke		      zSeries & Storage
hare@suse.de			      +49 911 74053 688
SUSE LINUX Products GmbH, Maxfeldstr. 5, 90409 Nürnberg
GF: Markus Rex, HRB 16746 (AG Nürnberg)
--
To unsubscribe from this list: send the line "unsubscribe linux-scsi" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html