Re: [linux-3.14 bisection] complete test-amd64-i386-xl-qcow2

[Ian Campbell <ian.campbell@citrix.com>]

[resending to correct stable address, sorry folks]

TL;DR: Any backport of 30b03d05e074 to earlier than commit 1401c00e59e
("xen/gntdev: convert priv->lock to a mutex", which was added in v4.0)
needs $something doing to it, either s/mutex/spinlock/ or (more likely)
backporting of 1401c00e59e too.

Looking at LTS:

3.18.y:	  Backported both.
3.16.y:	  Has backported neither
3.14.y:	* Only backported 30b03d05e074
3.12.y:	  Has backported neither
3.10.y:	* Only backported 30b03d05e074
3.4.y:	  Has backported neither
3.2.y:	  Has backported neither

So AFAICT 3.14.y and 3.10.y need fixes, probably following 3.18 and
backporting 1401c00e59e.

3.16/12/4/2 might need to be careful if they subsequently pick up 30b03d05.

See below for the build log error.

On Tue, 2015-09-01 at 11:05 +0100, Ian Campbell wrote:
> On Tue, 2015-09-01 at 10:57 +0100, Ian Jackson wrote:
> > Ian Campbell writes ("Re: [linux-3.14 bisection] complete test-amd64
> > -i386
> > -xl-qcow2"):
> > > On Wed, 2015-08-26 at 20:02 +0000, osstest service owner wrote:
> > > >   commit 9e6c072a69d87100808d16279d60e9f857291340
> > > >   Author: Marek Marczykowski-Górecki <
> > > > marmarek@invisiblethingslab.com
> > > > > 
> > > >   Date:   Fri Jun 26 03:28:24 2015 +0200
> > > >   
> > > >       xen/gntdevt: Fix race condition in gntdev_release()
> > > 
> > > I'm not sure what to make of this.
> > > 
> > > The qcow2 test is one of the only ones I'd expect to be exercising 
> > > gntdev
> > > (most tests use LVM+blkback), which explains why this particular 
> > > commit 
> > > is
> > > apparently seeing issues due to this particular change.
> > 
> > (You mean `which explains why this particular _test_ is [failing]',
> > I think.)
> 
> Indeed.
> 
> > The host serial log in one of the confirmation tests of 9e6c072a shows
> > serious trouble:
> > 
> >  http://logs.test-lab.xenproject.org/osstest/logs/60893/test-amd64-i386
> > -xl-qcow2/serial-huxelrebe1.log
> > 
> >  Aug 26 19:36:51.841068 [  738.050547] BUG: unable to handle kernel
> >  NULL pointer dereference at 00000014
> > 
> >  Aug 26 19:36:56.753068 [  738.050594] IP: []
> >  __mmu_notifier_invalidate_range_start+0x33/0x70
> > 
> > And the immediately preceding confirmation flight, which got a pass on
> > 9e6c072a~1, seems fine:
> > 
> >  http://logs.test-lab.xenproject.org/osstest/logs/60892/test-amd64-i386
> > -xl-qcow2/serial-huxelrebe1.log
> > 
> > But, it's difficult to see how that gntdev fix would be responsible
> > for the bug.  Perhaps it changes the order in which certain things
> > happen so as to expose another bug.
> 
> Or perhaps there was a fix and/or change in behaviour in the mmunotifier
> stuff which the patch relied on but which isn't in 3.14?

Looking at http://logs.test-lab.xenproject.org/osstest/logs/60949/'s build
jobs:
http://logs.test-lab.xenproject.org/osstest/logs/60949/build-amd64
-pvops/5.ts-kernel-build.log contains:

drivers/xen/gntdev.c: In function ‘gntdev_release’:
drivers/xen/gntdev.c:532:2: warning: passing argument 1 of ‘mutex_lock’
from incompatible pointer type [enabled by default]
In file included from include/linux/notifier.h:13:0,
                 from include/linux/memory_hotplug.h:6,
                 from include/linux/mmzone.h:821,
                 from include/linux/gfp.h:5,
                 from include/linux/kmod.h:22,
                 from include/linux/module.h:13,
                 from drivers/xen/gntdev.c:24:
include/linux/mutex.h:157:13: note: expected ‘struct mutex *’ but
argument is of type ‘struct spinlock_t *’
drivers/xen/gntdev.c:539:2: warning: passing argument 1 of
‘mutex_unlock’ from incompatible pointer type [enabled by default]
In file included from include/linux/notifier.h:13:0,
                 from include/linux/memory_hotplug.h:6,
                 from include/linux/mmzone.h:821,
                 from include/linux/gfp.h:5,
                 from include/linux/kmod.h:22,
                 from include/linux/module.h:13,
                 from drivers/xen/gntdev.c:24:
include/linux/mutex.h:174:13: note: expected ‘struct mutex *’ but
argument is of type ‘struct spinlock_t *’

Which is somehow a warning and not a build failure.

Ian.