* [PATCH] Silencing 'killing requests for dead queue'
@ 2011-11-09 7:39 Hannes Reinecke
2011-11-09 11:53 ` Bart Van Assche
2011-11-10 2:10 ` Stephen Rothwell
0 siblings, 2 replies; 6+ messages in thread
From: Hannes Reinecke @ 2011-11-09 7:39 UTC (permalink / raw)
To: James Bottomley; +Cc: linux-scsi, Hannes Reinecke
When we tear down a device we try to flush all outstanding
commands in scsi_free_queue(). However the check in
scsi_request_fn() is imperfect as it only signals that
we _might start_ aborting commands, not that we've actually
aborted some.
So move the printk inside the scsi_kill_request function,
this will also give us a hint about which commands are aborted.
Signed-off-by: Hannes Reinecke <hare@suse.de>
diff --git a/drivers/scsi/scsi_lib.c b/drivers/scsi/scsi_lib.c
index 06bc265..f85cfa6 100644
--- a/drivers/scsi/scsi_lib.c
+++ b/drivers/scsi/scsi_lib.c
@@ -1409,6 +1409,8 @@ static void scsi_kill_request(struct request *req, struct request_queue *q)
blk_start_request(req);
+ scmd_printk(KERN_INFO, cmd, "killing request\n");
+
sdev = cmd->device;
starget = scsi_target(sdev);
shost = sdev->host;
@@ -1490,7 +1492,6 @@ static void scsi_request_fn(struct request_queue *q)
struct request *req;
if (!sdev) {
- printk("scsi: killing requests for dead queue\n");
while ((req = blk_peek_request(q)) != NULL)
scsi_kill_request(req, q);
return;
^ permalink raw reply related [flat|nested] 6+ messages in thread
* Re: [PATCH] Silencing 'killing requests for dead queue'
2011-11-09 7:39 [PATCH] Silencing 'killing requests for dead queue' Hannes Reinecke
@ 2011-11-09 11:53 ` Bart Van Assche
2011-11-09 12:24 ` Hannes Reinecke
2011-11-10 2:10 ` Stephen Rothwell
1 sibling, 1 reply; 6+ messages in thread
From: Bart Van Assche @ 2011-11-09 11:53 UTC (permalink / raw)
To: Hannes Reinecke; +Cc: James Bottomley, linux-scsi
On Wed, Nov 9, 2011 at 8:39 AM, Hannes Reinecke <hare@suse.de> wrote:
> When we tear down a device we try to flush all outstanding
> commands in scsi_free_queue(). However the check in
> scsi_request_fn() is imperfect as it only signals that
> we _might start_ aborting commands, not that we've actually
> aborted some.
> So move the printk inside the scsi_kill_request function,
> this will also give us a hint about which commands are aborted.
>
> Signed-off-by: Hannes Reinecke <hare@suse.de>
>
> diff --git a/drivers/scsi/scsi_lib.c b/drivers/scsi/scsi_lib.c
> index 06bc265..f85cfa6 100644
> --- a/drivers/scsi/scsi_lib.c
> +++ b/drivers/scsi/scsi_lib.c
> @@ -1409,6 +1409,8 @@ static void scsi_kill_request(struct request *req, struct request_queue *q)
>
> blk_start_request(req);
>
> + scmd_printk(KERN_INFO, cmd, "killing request\n");
> +
> sdev = cmd->device;
> starget = scsi_target(sdev);
> shost = sdev->host;
> @@ -1490,7 +1492,6 @@ static void scsi_request_fn(struct request_queue *q)
> struct request *req;
>
> if (!sdev) {
> - printk("scsi: killing requests for dead queue\n");
> while ((req = blk_peek_request(q)) != NULL)
> scsi_kill_request(req, q);
> return;
Can this message appear during LUN scanning and hence confuse users ?
Also, patch 3308511 is already present in several stable trees so
shouldn't a patch like the above CC stable@vger.kernel.org too ?
Bart.
--
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
^ permalink raw reply [flat|nested] 6+ messages in thread
* Re: [PATCH] Silencing 'killing requests for dead queue'
2011-11-09 11:53 ` Bart Van Assche
@ 2011-11-09 12:24 ` Hannes Reinecke
0 siblings, 0 replies; 6+ messages in thread
From: Hannes Reinecke @ 2011-11-09 12:24 UTC (permalink / raw)
To: Bart Van Assche; +Cc: James Bottomley, linux-scsi, stable
On 11/09/2011 12:53 PM, Bart Van Assche wrote:
> On Wed, Nov 9, 2011 at 8:39 AM, Hannes Reinecke<hare@suse.de> wrote:
>> When we tear down a device we try to flush all outstanding
>> commands in scsi_free_queue(). However the check in
>> scsi_request_fn() is imperfect as it only signals that
>> we _might start_ aborting commands, not that we've actually
>> aborted some.
>> So move the printk inside the scsi_kill_request function,
>> this will also give us a hint about which commands are aborted.
>>
>> Signed-off-by: Hannes Reinecke<hare@suse.de>
>>
>> diff --git a/drivers/scsi/scsi_lib.c b/drivers/scsi/scsi_lib.c
>> index 06bc265..f85cfa6 100644
>> --- a/drivers/scsi/scsi_lib.c
>> +++ b/drivers/scsi/scsi_lib.c
>> @@ -1409,6 +1409,8 @@ static void scsi_kill_request(struct request *req, struct request_queue *q)
>>
>> blk_start_request(req);
>>
>> + scmd_printk(KERN_INFO, cmd, "killing request\n");
>> +
>> sdev = cmd->device;
>> starget = scsi_target(sdev);
>> shost = sdev->host;
>> @@ -1490,7 +1492,6 @@ static void scsi_request_fn(struct request_queue *q)
>> struct request *req;
>>
>> if (!sdev) {
>> - printk("scsi: killing requests for dead queue\n");
>> while ((req = blk_peek_request(q)) != NULL)
>> scsi_kill_request(req, q);
>> return;
>
> Can this message appear during LUN scanning and hence confuse users ?
>
Precisely.
Which incidentally also tells us that none of the maintainer tested
the original patch on SCSI parallel machines :-)
> Also, patch 3308511 is already present in several stable trees so
> shouldn't a patch like the above CC stable@vger.kernel.org too ?
>
Oh. Of course.
Cheers,
Hannes
--
Dr. Hannes Reinecke zSeries & Storage
hare@suse.de +49 911 74053 688
SUSE LINUX Products GmbH, Maxfeldstr. 5, 90409 Nürnberg
GF: J. Hawn, J. Guild, F. Imendörffer, 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
^ permalink raw reply [flat|nested] 6+ messages in thread
* Re: [PATCH] Silencing 'killing requests for dead queue'
2011-11-09 7:39 [PATCH] Silencing 'killing requests for dead queue' Hannes Reinecke
2011-11-09 11:53 ` Bart Van Assche
@ 2011-11-10 2:10 ` Stephen Rothwell
2011-11-10 11:57 ` Bart Van Assche
1 sibling, 1 reply; 6+ messages in thread
From: Stephen Rothwell @ 2011-11-10 2:10 UTC (permalink / raw)
To: James Bottomley; +Cc: Hannes Reinecke, linux-scsi
[-- Attachment #1: Type: text/plain, Size: 923 bytes --]
Hi James,
On Wed, 9 Nov 2011 08:39:24 +0100 Hannes Reinecke <hare@suse.de> wrote:
>
> When we tear down a device we try to flush all outstanding
> commands in scsi_free_queue(). However the check in
> scsi_request_fn() is imperfect as it only signals that
> we _might start_ aborting commands, not that we've actually
> aborted some.
> So move the printk inside the scsi_kill_request function,
> this will also give us a hint about which commands are aborted.
>
> Signed-off-by: Hannes Reinecke <hare@suse.de>
Tested-by: Stephen Rothwell <sfr@canb.auug.org.au>
[I realise that you have already committed this fix, but I thought I
should test it anyway.]
ipr_init is now down to 2543495 usecs (down from 494857993 usecs without
this patch) and the machine boots as well as can be expected.
--
Cheers,
Stephen Rothwell sfr@canb.auug.org.au
http://www.canb.auug.org.au/~sfr/
[-- Attachment #2: Type: application/pgp-signature, Size: 836 bytes --]
^ permalink raw reply [flat|nested] 6+ messages in thread
* Re: [PATCH] Silencing 'killing requests for dead queue'
2011-11-10 2:10 ` Stephen Rothwell
@ 2011-11-10 11:57 ` Bart Van Assche
2011-11-10 12:13 ` Stephen Rothwell
0 siblings, 1 reply; 6+ messages in thread
From: Bart Van Assche @ 2011-11-10 11:57 UTC (permalink / raw)
To: Stephen Rothwell; +Cc: James Bottomley, Hannes Reinecke, linux-scsi
On Thu, Nov 10, 2011 at 3:10 AM, Stephen Rothwell <sfr@canb.auug.org.au> wrote:
> ipr_init is now down to 2543495 usecs (down from 494857993 usecs without
> this patch) and the machine boots as well as can be expected.
These timing differences are puzzling me. Have you been able to find
out what is causing them ? Are kernel messages on that system perhaps
sent to a serial console ?
Bart.
^ permalink raw reply [flat|nested] 6+ messages in thread
* Re: [PATCH] Silencing 'killing requests for dead queue'
2011-11-10 11:57 ` Bart Van Assche
@ 2011-11-10 12:13 ` Stephen Rothwell
0 siblings, 0 replies; 6+ messages in thread
From: Stephen Rothwell @ 2011-11-10 12:13 UTC (permalink / raw)
To: Bart Van Assche; +Cc: James Bottomley, Hannes Reinecke, linux-scsi
[-- Attachment #1: Type: text/plain, Size: 723 bytes --]
Hi Bart,
On Thu, 10 Nov 2011 12:57:37 +0100 Bart Van Assche <bvanassche@acm.org> wrote:
>
> On Thu, Nov 10, 2011 at 3:10 AM, Stephen Rothwell <sfr@canb.auug.org.au> wrote:
> > ipr_init is now down to 2543495 usecs (down from 494857993 usecs without
> > this patch) and the machine boots as well as can be expected.
>
> These timing differences are puzzling me. Have you been able to find
> out what is causing them ? Are kernel messages on that system perhaps
> sent to a serial console ?
Yeah, effectively (not actually serial, but close enough). And there
were a very large number of messages.
--
Cheers,
Stephen Rothwell sfr@canb.auug.org.au
http://www.canb.auug.org.au/~sfr/
[-- Attachment #2: Type: application/pgp-signature, Size: 836 bytes --]
^ permalink raw reply [flat|nested] 6+ messages in thread
end of thread, other threads:[~2011-11-10 12:13 UTC | newest]
Thread overview: 6+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2011-11-09 7:39 [PATCH] Silencing 'killing requests for dead queue' Hannes Reinecke
2011-11-09 11:53 ` Bart Van Assche
2011-11-09 12:24 ` Hannes Reinecke
2011-11-10 2:10 ` Stephen Rothwell
2011-11-10 11:57 ` Bart Van Assche
2011-11-10 12:13 ` Stephen Rothwell
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox