[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Bug#1011365: bullseye-pu: package nvidia-cuda-toolkit/11.2.2-3+deb11u2



On Mon, 2022-05-30 at 17:31 +0200, Andreas Beckmann wrote:
> On 29/05/2022 16.16, Adam D. Barratt wrote:
> > Unfortunately the amd64 upload got rejected:
> > 
> > Version check failed:
> > Your upload included the binary package nvidia-openjdk-8-jre,
> > version
> > 9.+8u332-ga-1~deb9u1~11.2.2-3+deb11u2, for amd64,
> > however experimental already has version 9.+8u332-ga-1~11.5.1-2.
> > Uploads to proposed-updates must have a lower version than present
> > in
> > experimental.
> 
> Oh yes. 1~d   < 1
>          1~d~1 > 1~1
> 
> Is there a similar constraint for sid as well? I would have expected
> that to trigger first ... and probably s/already/only/ ...
> 

Surprisingly not. I have a feeling there once was, but it was removed.
I'm not entirely sure on that though; memory can play tricks.

The current checks are:

adsb@coccia:$ dak admin v-c list-suite proposed-updates
proposed-updates MustBeNewerThan oldoldoldstable
proposed-updates MustBeNewerThan oldoldstable
proposed-updates MustBeNewerThan stable
proposed-updates Enhances stable
proposed-updates MustBeNewerThan oldstable
proposed-updates MustBeOlderThan experimental

> Before I upload a fix, I'd like you to double check that these
> versions do not validate the ordering rules:
> 
> nvidia-openjdk-8-jre_9.+8u332-ga-1~~deb9u1~11.2.2-3+deb11u3_amd64.deb
> nvidia-openjdk-8-jre_9.+8u312-b07-1~11.2.2+8u302-b08-1~11.2.2-
> 3+deb11u3_ppc64el.deb
> 

nvidia-openjdk-8-jre | 9.+8u332-ga-1~11.4.3-3 | testing/non-free          | amd64, ppc64el
nvidia-openjdk-8-jre | 9.+8u332-ga-1~11.4.3-3 | unstable/non-free         | amd64, ppc64el
nvidia-openjdk-8-jre | 9.+8u332-ga-1~11.5.1-2 | experimental/non-free     | amd64, ppc64el

$ dpkg --compare-versions "9.+8u332-ga-1~~deb9u1~11.2.2-3+deb11u3" lt "9.+8u332-ga-1~11.4.3-3" && echo y
y

$ dpkg --compare-versions "9.+8u312-b07-1~11.2.2+8u302-b08-1~11.2.2-3+deb11u3" lt "9.+8u332-ga-1~11.4.3-3" && echo y
y

Both look OK so far as I can see, and already sort below the version in
testing so don't present any issues with prop-ups etc.

Regards,

Adam


Reply to: