autopkgtest regression in dgit (8.3) on amd64 due to dpkg (1.19.2 to 1.19.4)

classic Classic list List threaded Threaded
5 messages Options
Reply | Threaded
Open this post in threaded view
|

autopkgtest regression in dgit (8.3) on amd64 due to dpkg (1.19.2 to 1.19.4)

Ian Jackson-2
My grep-excuses cron writes:

> autopkgtest regression
>     in dgit (8.3) on amd64
>     due to dpkg (1.19.2 to 1.19.4)
> test info
>     REGRESSION
>     https://ci.debian.net/data/autopkgtest/testing/amd64/d/dgit/1767699/log.gz
>     https://ci.debian.net/packages/d/dgit/testing/amd64
>     null
>     https://ci.debian.net/api/v1/retry/1767699
> migration excuses for dpkg
>     Maintainer: Dpkg Developers
>     Too young, only 1 of 5 days old
>     Updating dpkg fixes old bugs: #911620
>     Piuparts tested OK - https://piuparts.debian.org/sid/source/d/dpkg.html

I thought I would write and say that I am aware of this.  I briefly
looked at the log and the two failures I looked at seemed to be a
missing `use'.  I intend to fix this in dgit, probably at the weekend.

Feel free to file a bug to track this if you would like to do so.
If you do that bug should probably be RC since this is blocking your
migration.

Whether you file a bug or not, I am treating this as an RC issue.

Regards,
Ian.

--
Ian Jackson <[hidden email]>   These opinions are my own.

If I emailed you from an address @fyvzl.net or @evade.org.uk, that is
a private address which bypasses my fierce spamfilter.

Reply | Threaded
Open this post in threaded view
|

Re: autopkgtest regression in dgit (8.3) on amd64 due to dpkg (1.19.2 to 1.19.4)

Guillem Jover
On Thu, 2019-01-24 at 12:57:08 +0000, Ian Jackson wrote:

> My grep-excuses cron writes:
> > autopkgtest regression
> >     in dgit (8.3) on amd64
> >     due to dpkg (1.19.2 to 1.19.4)
> > test info
> >     REGRESSION
> >     https://ci.debian.net/data/autopkgtest/testing/amd64/d/dgit/1767699/log.gz
> >     https://ci.debian.net/packages/d/dgit/testing/amd64
> >     null
> >     https://ci.debian.net/api/v1/retry/1767699
> > migration excuses for dpkg
> >     Maintainer: Dpkg Developers
> >     Too young, only 1 of 5 days old
> >     Updating dpkg fixes old bugs: #911620
> >     Piuparts tested OK - https://piuparts.debian.org/sid/source/d/dpkg.html
>
> I thought I would write and say that I am aware of this.  I briefly
> looked at the log and the two failures I looked at seemed to be a
> missing `use'.  I intend to fix this in dgit, probably at the weekend.

Thanks for the heads up!

> Feel free to file a bug to track this if you would like to do so.
> If you do that bug should probably be RC since this is blocking your
> migration.

I don't think that's necessary.

Regards,
Guillem

Reply | Threaded
Open this post in threaded view
|

Re: autopkgtest regression in dgit (8.3) on amd64 due to dpkg (1.19.2 to 1.19.4)

Ian Jackson-2
Hi.

dgit's autopkgtests have a regression with dpkg 1.19.4.  Tracker's
view of the excuses says:

    Too young, only 2 of 5 days old
    Updating dpkg fixes old bugs: #911620
    Piuparts tested OK - https://piuparts.debian.org/sid/source/d/dpkg.html
    autopkgtest for dgit/8.3: amd64: Regression ♻
    Not considered

Previously in this situation I have seen `migration age increased'
etc.  Is that still happening ?  Has the excuses output changed ?

FTR, I have discussed this with the dpkg maintainer.  I think this is
a bug in dgit rather than in dpkg and I am treating it as an RC issue
which will be fixed soon.

But, the new dpkg ought not to migrate to testing before the fix in
dgit, because I think the bug makes dgit fairly broken with the new
dpkg.

Regards,
Ian.

--
Ian Jackson <[hidden email]>   These opinions are my own.

If I emailed you from an address @fyvzl.net or @evade.org.uk, that is
a private address which bypasses my fierce spamfilter.

Reply | Threaded
Open this post in threaded view
|

Re: autopkgtest regression in dgit (8.3) on amd64 due to dpkg (1.19.2 to 1.19.4)

Mattia Rizzolo-5
On Fri, Jan 25, 2019 at 11:23:06AM +0000, Ian Jackson wrote:
> dgit's autopkgtests have a regression with dpkg 1.19.4.  Tracker's
> view of the excuses says:
> Previously in this situation I have seen `migration age increased'
> etc.  Is that still happening ?  Has the excuses output changed ?

Yes, it changed because since January autopkgtest regressions are
completely blocking, not just delaying migration.  Therefore the
migration is just "note considered" at all, regardless of the aging.

> FTR, I have discussed this with the dpkg maintainer.  I think this is
> a bug in dgit rather than in dpkg and I am treating it as an RC issue
> which will be fixed soon.

And you ought to treat it like that :)

> But, the new dpkg ought not to migrate to testing before the fix in
> dgit, because I think the bug makes dgit fairly broken with the new
> dpkg.

Testing migration won't happen unless autopkgtest is fixed or somebody
from the release team ignores it.

--
regards,
                        Mattia Rizzolo

GPG Key: 66AE 2B4A FCCF 3F52 DA18  4D18 4B04 3FCD B944 4540      .''`.
more about me:  https://mapreri.org                             : :'  :
Launchpad user: https://launchpad.net/~mapreri                  `. `'`
Debian QA page: https://qa.debian.org/developer.php?login=mattia  `-

signature.asc (849 bytes) Download Attachment
Reply | Threaded
Open this post in threaded view
|

Re: autopkgtest regression in dgit (8.3) on amd64 due to dpkg (1.19.2 to 1.19.4)

Ian Jackson-2
Mattia Rizzolo writes ("Re: autopkgtest regression in dgit (8.3) on amd64 due to dpkg (1.19.2 to 1.19.4)"):
> Yes, it changed because since January autopkgtest regressions are
> completely blocking, not just delaying migration.  Therefore the
> migration is just "note considered" at all, regardless of the aging.

Ah, OK, good.

> Testing migration won't happen unless autopkgtest is fixed or somebody
> from the release team ignores it.

Right, that is as it should be.

> > FTR, I have discussed this with the dpkg maintainer.  I think this is
> > a bug in dgit rather than in dpkg and I am treating it as an RC issue
> > which will be fixed soon.
>
> And you ought to treat it like that :)

Indeed; that is a consequence of the above.

Thanks,
Ian.

--
Ian Jackson <[hidden email]>   These opinions are my own.

If I emailed you from an address @fyvzl.net or @evade.org.uk, that is
a private address which bypasses my fierce spamfilter.