Re: [Stable-8.1.3 05/55] tests/docker: avoid invalid escape in Python string

[Thomas Huth]

On 03/12/2024 05.59, Michael Tokarev wrote:
> 02.12.2024 23:26, Thomas Huth wrote:
> > On 09/11/2023 14.42, Michael Tokarev wrote:
> > > From: Paolo Bonzini <pbonzini@redhat.com>
> > >
> > > This is an error in Python 3.12; fix it by using a raw string literal.
> > >
> > > Cc: qemu-stable@nongnu.org
> > > Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
>
> >   Hi Michael!
>
> Hello!
>
> > Could you add this patch to the stable-7.2 branch, too? I think it's 
> > missing there...
> Heck!  You caught me cheating! :)
>
> It's not just this one is missing from 7.2, it's whole lot
> of tests/ and python changes which aren't there and should
> be picked up.

I don't think you need each and every patch in 7.2, but I just noticed that 
a recent python is complaining when compiling the 7.2 branch on a recent 
Fedora installation.

> And this really made me curious, - why do you still use
> qemu 7.2 branch?  This is an interesting (to me) question,
> since these days, it feels like no one uses older branches, -
> even 8.2 one which has been requested by ubuntu, - they
> actually never used it, so I'm not really sure it is worth
> all the CI minutes we spend on it.  In the past, 7.2 was
> used for RHEL, now I think this usage is gone too.  So your
> email made me.. really curious :)

Don't get too exited. I just tried to see whether we could clean up that 
cirrus-ci mess in 7.2 to avoid that we are wasting too much CI cycles / 
credits in vain there.

I got it working to a point where it compiles again, but it then hangs in 
the CI until it times out after 1h ... no clue how to fix that right now. 
Maybe it's best if we completely disable cirrus-ci in the 7.2 branch to 
avoid burning our free credits there?

On the bright side, I think I fixed a bunch of the normal gitlab-CI jobs 
along the way. If you like, I can send out those fixes for 7.2 at least...?

 Thomas