On Tue, Dec 10, 2024 at 02:48:52PM +0100, Bilge Aydin wrote:
> Hello Greg,
>
> Thank you very much for your prompt reply.
>
> "So why do you feel it is required in 5.15.y?"
> --> We have faced the issue in our K5.15 based platform. Our technical team has made an investigation and found your given patch. Afterwards we verified it by running a stress test. The result was positive and we did not face the kernel panic again.
That's great, then send us a backported patch (and the fix for that, you
did catch that one, right?)
> Since we are not OS experts, we fully respect Linux community's decisions. That's why we do not see ourselves in that high position to give any guidance and share any patch with you. During our investigation, we only recognized that the given patch was already integrated into the Android common kernel. Outgoing from that, I just want to share the indication with you.
>
> Would it be possible for you to start an internal communication within the Linux community for the synchronization of the K5.15 and K6.1 mainlines explicitly for iommu driver?
That is not how the stable kernel branches work at all, sorry. Please
read:
https://www.kernel.org/doc/html/latest/process/stable-kernel-rules.html
for more information.
Also, if you need the iommu code in 6.1, then just use the 6.1.y kernel
please. There's nothing keeping you from doing that, right? That way
you get more support for more hardware and more fixes that didn't happen
to get backported to the really old 5.15.y release. Also, you will have
to move off of 5.15.y in a few years anyway, so might as well do it now
when you have a very good reason to do so.
good luck!
greg k-h
