pKVM breakage in mainline on n1sdp

[Mark Brown]

On Mon, Feb 23, 2026 at 10:05:46AM +0000, Mark Rutland wrote:
> On Sat, Feb 21, 2026 at 01:42:52PM +0000, Mark Brown wrote:

> > It won't be vanilla v6.19, that specific log will be tip of Linus' tree
> > from yesterday morning UK time which should be 8bf22c33e7a1 ("Merge tag
> > 'net-7.0-rc1' of git://git.kernel.org/pub/scm/linux/kernel/git/netdev/net"),
> > there were a few results for commits over the day prior to that showing
> > the same behaviour.

> Are you sure? That log says:

>   CPU: 1 UID: 0 PID: 1955 Comm: qemu-system-aar Not tainted 6.19.0 #1 PREEMPT   

> ... where "6.19.0" means that's the 6.19 tag, and "#1" means that's the
> first build against that.

The UI claims that we're seeing this in 956b9cbd7f156 (Merge tag
'kbuild-fixes-7.0-1' of
git://git.kernel.org/pub/scm/linux/kernel/git/kbuild/linux) but
unfortunately the build logs don't clearly confirm this.

All the builds are fresh builds in a new container so it would be
surprising if we reported a build number other than 1.

> Is the build system overriding things such that this is misleading?

I believe that to be the case here, the builds are not building from a
git checkout but rather the build coordination is packing a tarball of
the source up then handing that off to be built on a separate system
which means that the magic to figure out the git describe and put that
into the version number during the build can't do anything and just
reports whatever is in the Makefile.

> Is there potentially a problem with the build system?

I can't 100% discount that possibility but it does seem to DTRT for
other commits and has done for a long time.
-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 488 bytes
Desc: not available
URL: <http://lists.infradead.org/pipermail/linux-arm-kernel/attachments/20260223/b659ef24/attachment.sig>