Control: tags -1 + moreinfo
Hi,
On Mon, Jul 04, 2016 at 05:15:31PM +0200, Ben Hutchings wrote:
> Control: reassign -1 src:linux 4.6.1-1
> Control: severity -1 important
> Control: tag -1 confirmed upstream
>
> On Mon, 2016-07-04 at 00:47 -0400, westlake wrote:
> > Package: linux-image-4.6.0-0.bpo.1-amd64
> > Version: 4.6.1-1~bpo8+1
> > Severity: normal
> >
> > kernels 4.5 and 4.6 breaks qemu-nbd from mapping partitions
> >
> > qemu-nbd can map partitions from VM images and raw disk images to
> > /dev/nbdNpM nodes, but proper module options have to be loaded prior so
> > that qemu-nbd performs these additional mappings (the general consensus
> > around mapping image files has users issuing "kpartx -a /dev/",
> > but kpartx is not needed at all after issuing qemu-nbd)
> >
> > nothing is specially set up other than a module option file and that's
> > pretty much it -- there's just 2 or 3 commands for the testing --- all
> > which indicates the issue is more towards kernel than it is with qemu-nbd.
>
> I can confirm this issue.
>
> > after a modprobe nbd is issued, there is no "/dev/nbd0p1"-like devices,
>
> That much is correct behaviour since no device is attached yet...
>
> > so the main problem with kernels 4.5/4.6 is that qemu-nbd is not being
> > allowed to generate any device nodes. ("partition" device nodes that is
> > -- "/dev/nbd0" and "/dev/nbd1" nodes are generated by modprobe nbd)
> [...]
>
> ...but the kernel should scan the partition table for an nbd as soon as
> it is attached to a server. Currently that isn't happening. As a
> workaround, running 'fdisk' and then re-writing the same partition
> table triggers a rescan.
>
> This is almost certainly caused by:
>
> commit 37091fdd831f28a6509008542174ed324dd645bc
> Author: Markus Pargmann
> <m...@pengutronix.de>
> Date: Mon Jul 27 07:36:49 2015 +0200
>
> nbd: Create size change events for userspace
>
> I have a rough idea of how to fix it and will update you later.
This issue has probably been fixed in meanwhile?
Regards,
Salvatore
