Re: [PATCH v2] PCI/MSI: Avoid torn updates to MSI pairs

[Thomas Gleixner]

Jacob Pan <jacob.jun.pan@xxxxxxxxxxxxxxx> writes:

> It seems we could avoid force_retrigger once IR is turned on. I have tried
> this patch with IR turned on, still seeing force_retrigger=1 for MSIs w/o
> masking.

Yes. That patch was a POC and surely not a final solution. Aside of that
I'm still looking into a solution to avoid that which does not involve
horrible locking and tons of scary code.

Thanks,

        tglx