CI build for quanta/gbs

[Andrew Geissler]

> On May 13, 2021, at 7:47 AM, George Hung (洪忠敬) <George.Hung at quantatw.com> wrote:
> 
> Hi Andrew,
>  
> Sorry to make your inconvenience, we already fix all CI build errors and also run CI build locally to make sure it can work fine.

Looks good, I added it to https://jenkins.openbmc.org/job/ci-openbmc/

Andrew

>  
> Could you help to merge gbs to CI build ?
>  
>  
> Thanks a lot.
>  
> Best Regards
> George Hung
> 
>  
> Apologies, one of my old patch that Quanta seemed to have been using was finally merged (https://gerrit.openbmc-project.xyz/c/openbmc/phosphor-hwmon/+/24337/) and it seems like that caused the patch failure you see here.
>  
> Here is the fix that should fix the problem: https://gerrit.openbmc-project.xyz/c/openbmc/openbmc/+/43046
>  
> Thanks,
> Brandon
>  
> On Mon, May 10, 2021 at 8:55 AM Andrew Geissler <geissonator at gmail.com> wrote:
> 
> 
> > On May 7, 2021, at 7:19 PM, Brandon Kim <brandonkim at google.com> wrote:
> > 
> > Hi Andrew,
> > 
> > Do you mind kicking off the docker build again for gbs? It looks like sslh bump may have  been needed from meta-google https://gerrit.openbmc-project.xyz/c/openbmc/openbmc/+/43015
> 
> The job runs daily (when new code is available), looks like we’re on to something new now?
> 
> https://jenkins.openbmc.org/job/latest-master/label=docker-builder,target=gbs/234/console
> 
> 3 out of 5 hunks FAILED -- rejects in file sensor.hpp
> Patch 0001-sensor-Implement-sensor-ASYNC_READ_TIMEOUT.patch does not apply (enforce with -f)
> 
> stderr: 
> ERROR: Logfile of failure stored in: /data0/jenkins/workspace/latest-master/label/docker-builder/target/gbs/build/work/armv7a-openbmc-linux-gnueabi/phosphor-hwmon/1.0+gitAUTOINC+6d50c3e9e1-r1/temp/log.do_patch.19118
> 
> 
> > 
> > Thanks,
> > Brandon
> > 
> > On Fri, May 7, 2021 at 11:26 AM Andrew Geissler <geissonator at gmail.com> wrote:
> > 
> > 
> >> On May 7, 2021, at 2:10 AM, George Hung (洪忠敬) <George.Hung at quantatw.com> wrote:
> >> 
> >> Hi Andrew,
> >>  
> >> We already fix the patch error for phosphor-hwmon, could you help proceed to join the quanta/gbs to CI build ?
> > 
> > Hey George, I’ve added it to our daily build, it’s still not passing for gbs though. Please take a look at https://jenkins.openbmc.org/job/latest-master/233/label=docker-builder,target=gbs/ 
> > 
> > Once I see that daily job pass I’ll add it to our official CI job.
> > 
> >>  
> >> Thanks a lot.
> >>  
> >> Best Regards
> >> George Hung
> >> 
> >>  
> >> Hi Andrew,
> >>  
> >> It looks like it's due to my patch to phosphor-hwmon (which, the repo is going through some overhaul and is causing conflicts with). The patch in question is in the process of getting upstreamed and the repo is settling down, so I'll ping you on Discord once it looks ready to be tried again (hopefully by end of the week).
> >>  
> >> Thanks,
> >> Brandon
> >>  
> >> On Thu, May 6, 2021 at 6:38 AM Andrew Geissler <geissonator at gmail.com> wrote:
> >>  
> >> 
> >> 
> >> On May 5, 2021, at 2:37 PM, Brandon Kim <brandonkim at google.com> wrote:
> >>  
> >> Hi Andrew,
> >>  
> >> `gbs` machine will provide meta-google coverage, so it would be great if `gbs` could be added in addition to `gsj` instead of as a replacement.
> >>  
> >> Is there a shortage of OpenBMC CI nodes by any chance? Google has been providing 2 of the OpenBMC CI nodes, so it would be great if adding `gbs` could count towards one of those CI nodes.
> >>  
> >> Yep, google donates two servers so sounds good. I added it to our daily build of master and it looks like it hit a build issue. We can coordinate in discord if you like but once we get this figured out, I’ll add it to the main openbmc/openbmc CI job.
> >>  
> >> https://jenkins.openbmc.org/job/latest-master/231/label=docker-builder,target=gbs/console
> >> 
> >> 
> >>  
> >> Thank you,
> >> Brandon
> >>  
> >> On Wed, May 5, 2021 at 10:54 AM Patrick Venture <venture at google.com> wrote:
> >> On Wed, May 5, 2021 at 6:25 AM Andrew Geissler <geissonator at gmail.com> wrote:
> >> >
> >> >
> >> >
> >> > > On May 5, 2021, at 1:26 AM, George Hung (洪忠敬) <George.Hung at quantatw.com> wrote:
> >> > >
> >> > > Hi Andrew Geissler,
> >> > >
> >> > > We have ported many OpenBMC features to quanta/gbs machine (Nuvoton BMC platform) and hope it could be added to CI build, could you help add quanta/gbs machine to CI build verification ? (I'm not familiar with this, if there's anything we need to do first, please let us know)
> >> >
> >> > Hey George,
> >> >
> >> > Getting a new system into CI doesn’t have the most defined process but
> >> > I think in general it has to fall under one of these due to our constraints
> >> > in compute power for CI:
> >> >
> >> > 1) Propose an existing system in CI to replace (and the benefits of that)
> >> > 2) Donate a jenkins compute node to openbmc CI
> >> > 3) Convince the community that your system provides additional meta-*
> >> >     layer coverage (or some other critical benefit) that would be worth
> >> >     the additional hit to the existing CI infrastructure.
> >> >
> >> > We currently have a “gsj” system in CI. Would it make sense to replace
> >> > it with this new “gbs” machine?
> >> 
> >> Jumping on this thread to say that I'm hoping to have quanta-gbs
> >> support in upstream qemu shortly --- so we could also enable this as a
> >> Qemu CI machine.
> >> 
> >> >
> >> >
> >> > >
> >> > >
> >> > > Thanks a lot.
> >> > >
> >> > > Best Regards
> >> > > George Hung
> >> > >
> >> >
> > 
>