Re: [PATCH 1/2] amd_nb: Add PCI ID for family 19h model 78h

From: Thorsten Leemhuis
Date: Thu May 11 2023 - 16:09:21 EST

On 11.05.23 21:51, Limonciello, Mario wrote:
> [AMD Official Use Only - General]
> +stable, Sasha
>>> Together with this patch there are now at least two regressions if
>>> -rc1 whch could have been avoided and may impact testability on
>>> affected systems.
>> Are you saying that this patch which fixes s2idle on some random box
>> should've gone to Linus *immediately*?
>> And read my mail again:
>> "Some fixes need longer testing because there have been cases where
>> a fix breaks something else."
>> So yes, I disagree with rushing fixes immediately. If they're obvious
>> - whatever that means - then sure but not all of them are such.
>> --
> Unfortunately, it looks like the broken commit got backported into 6.1.28,
> but the fix still isn't in Linus' tree.
> Sasha,
> Can you please pick up
> for 6.1.29 to fix the regression?

FWIW, the stable team afaics usually does not fix anything in stable
trees before it's fixed in mainline. IOW: that fix now quickly needs to
go to Linus to get it quickly fixed in 6.1.y.

Side note: I'll soon post a rewritten section of 'Prioritize work on
fixing regressions' which is part of
Documentation/process/handling-regressions.rst. It among others will
cover this case:

* Whenever you want to swiftly resolve a regression that recently also
made it into a proper mainline, stable, or longterm release, fix it
quickly in mainline; when appropriate thus involve Linus to fast-track
the fix. That's because the stable team normally does neither revert nor
fix any changes that cause a regression in mainline, too.

Ciao, Thorsten