[dm-devel] dm-mq and end_clone_request()

[Laurence Oberman]

----- Original Message -----
> From: "Bart Van Assche" <bart.vanassche at sandisk.com>
> To: "Laurence Oberman" <loberman at redhat.com>
> Cc: dm-devel at redhat.com, linux-scsi at vger.kernel.org, "Mike Snitzer" <snitzer at redhat.com>, "Johannes Thumshirn"
> <jthumshirn at suse.de>
> Sent: Tuesday, August 9, 2016 1:16:52 PM
> Subject: Re: [dm-devel] dm-mq and end_clone_request()
> 
> On 08/09/2016 10:12 AM, Laurence Oberman wrote:
> > I was talking about this patch
> >
> > --- a/drivers/scsi/scsi_scan.c
> > +++ b/drivers/scsi/scsi_scan.c
> > @@ -1890,10 +1890,11 @@ void scsi_forget_host(struct Scsi_Host *shost)
> >   restart:
> >          spin_lock_irqsave(shost->host_lock, flags);
> >          list_for_each_entry(sdev, &shost->__devices, siblings) {
> > -                if (sdev->sdev_state == SDEV_DEL)
> > +                if (sdev->sdev_state == SDEV_DEL || scsi_device_get(sdev)
> > < 0)
> >                          continue;
> >                  spin_unlock_irqrestore(shost->host_lock, flags);
> >                  __scsi_remove_device(sdev);
> > +                scsi_device_put(sdev);
> >                  goto restart;
> >          }
> >          spin_unlock_irqrestore(shost->host_lock, flags);
> 
> Hello Laurence,
> 
> Did you run your tests with that patch applied? If so, it would help if
> you could rerun your tests without that patch. If the above patch makes
> a difference it means that it can happen that __scsi_remove_device()
> does not change the device state into SDEV_DEL. That's a bug and we need
> to know whether or not __scsi_remove_device() behaves correctly.
> 
> Thanks,
> 
> Bart.
> 
Yes Sir, I ran all yesterdays tests on your kernel with that patch applied.
Of course it may well just be luck/coincidence that the host delete race is no longer happening
so I agree we need to re-run the tests so I will revert and re-run.
I will probably only get back to you tomorrow with the results.

Thanks
Laurence