❌ FAIL: Test report for kernel 5.5.0-rc6-b260f5e.cki?(arm-next)

Will Deacon will at kernel.org
Fri Jan 17 03:13:44 PST 2020


On Fri, Jan 17, 2020 at 10:57:49AM +0000, Robin Murphy wrote:
> On 2020-01-17 10:34 am, Will Deacon wrote:
> > On Fri, Jan 17, 2020 at 05:24:17AM -0500, Veronika Kabatova wrote:
> > > ----- Original Message -----
> > > > From: "CKI Project" <cki-project at redhat.com>
> > > > To: will at kernel.org, "catalin marinas" <catalin.marinas at arm.com>, linux-arm-kernel at lists.infradead.org
> > > > Cc: "Yi Chen" <yiche at redhat.com>, "Jianwen Ji" <jiji at redhat.com>, "Hangbin Liu" <haliu at redhat.com>, "Xiumei Mu"
> > > > <xmu at redhat.com>
> > > > Sent: Friday, January 17, 2020 4:35:27 AM
> > > > Subject: ❌ FAIL: Test report for kernel 5.5.0-rc6-b260f5e.cki	(arm-next)
> > > > 
> > > > We ran automated tests on a recent commit from this kernel tree:
> > > > 
> > > >         Kernel repo:
> > > >         git://git.kernel.org/pub/scm/linux/kernel/git/arm64/linux.git
> > > >              Commit: b260f5eb4e57 - Merge branch 'for-next/core' into
> > > >              for-kernelci
> > > > 
> > > > The results of these automated tests are provided below.
> > > > 
> > > >      Overall result: FAILED (see details below)
> > > >               Merge: OK
> > > >             Compile: OK
> > > >               Tests: FAILED
> > > > 
> > > > All kernel binaries, config files, and logs are available for download here:
> > > > 
> > > >    https://artifacts.cki-project.org/pipelines/387219
> > > > 
> > > > One or more kernel tests failed:
> > > > 
> > > >      aarch64:
> > > >       ❌ Networking UDP: socket
> > > >       ❌ Networking ipsec: basic netns tunnel
> > > > 
> > > 
> > > after yesterday's discussions I thought I'd point it out specifically.
> > > These tests did really report a failure and the results are not caused
> > > by the infra issue we were talking about. I'm sure the test maintainers
> > > can help with figuring out if this is something to be concerned with.
> > 
> > Thanks, Veronika -- I was about to mail you to ask! We queued some IP
> > checksum changes yesterday, so I bet they're the culprit, since this looks
> > to be related to UDP checksumming.
> 
> Urgh... let me see if I can reproduce something locally - I had iperf3
> running fine in both directions with tx/rx offload disabled, and my
> userspace tests were showing the new routine produce the same sum as the
> generic version for the same buffer, so I wouldn't have expected UDP to
> behave any differently, but I'll dig in and double-check everything...

It could easily be one of the other things I queued yesterday, but I've
pushed out an updated for-kernelci branch with your patch reverted so keep
your UTF-8 eyes peeled to see if we get a big green tick back or not.

Will



More information about the linux-arm-kernel mailing list