[dm-devel] v4.3-rc2 dm-mq bug

[Mike Snitzer]

On Thu, Sep 24 2015 at  5:27pm -0400,
Bart Van Assche <bart.vanassche at sandisk.com> wrote:

> On 09/24/2015 02:18 PM, Mike Snitzer wrote:
> >p.s. Bart, why is your test destroying the DM mpath device while it is
> >issuing IO to the device? ;)
> 
> Removing SCSI device instances created by the SRP initiator while
> I/O is being submitted is a great way to stress-test the SRP
> initiator driver but apparently the test I'm running triggers bugs
> in other drivers than the SRP initiator driver ...

I'm not referring to the removal of a DM mpath device's underlying SRP
paths.  The __dm_destroy() path is only used when destroying a DM
device.  The only DM device used by multipath is the DM mpath device.

So somehow your test is running the equivalent of:
 dmsetup remove <toplevel_DM_mpath_device>

While IO is being issued directly to that device.