Re: [GIT PULL v2] PCI changes for v6.17

[Nam Cao]

Ammar Faizi <ammarfaizi2@xxxxxxxxxxx> writes:

> On Sat, Aug 09, 2025 at 05:32:16PM +0200, Nam Cao wrote:
>> So unlike what VMD doc says, it actually can have non-MSI-X children devices!
>
> If that's the conclusion, then Intel VMD doc also needs fixing :/

Other possibilities are problem with your BIOS (likely), or problem with
Linux's PCI enumeration (unlikely).

But without hardware, I cannot investigate this further. Now that we
know my commit didn't make the driver any worse, I am done here.

>> Please discard the reverts and the diff I sent you, and try the diff
>> below. I believe your machine will work now.
>
> Yes, I can confirm it's now clean. Just to verify both sides, here is
> the last result:
>
>   https://gist.github.com/ammarfaizi2/72578d2b4cc385fbdb5faee69013d530
>
> If that one fix is final, then:
>
> Tested-by: Ammar Faizi <ammarfaizi2@xxxxxxxxxxx>
>
> Thanks for the debugging work.

Thanks for running tests, it would be impossible to figure out otherwise.

> It's probably too late to get the fix in mainline before rc1. But if it
> can go upstream sooner, that would be great.

I don't think PCI maintainers are available at the moment, so I will
send the patch next Monday. Time to enjoy my weekends..

Nam