On Fri, 2025-01-03 at 10:31 -0400, Jason Gunthorpe wrote: > > IIRC Alex had a patch series for qemu to capture and rewrite the ath > non-standard MSI locations, so virtualization worries should not block > moving ath to use the device-specific MSI.. In the absence of the IOMMU in our shared fantasy, this seems like the best approach. Even for passed-through PCI hardware, the config space is largely a fiction. As are the MSI(-X) vectors. So although it's a device-specific quirk, this seems like the best approach. Probably best for the hypervisor to have some way to *advertise* that it's handling this though, as guests also might want to run on hypervisors which don't.
Attachment:
smime.p7s
Description: S/MIME cryptographic signature
