[PATCH v1 1/2] KVM: arm64: determine memory type from VMA

Jason Gunthorpe jgg at nvidia.com
Tue Mar 19 06:38:44 PDT 2024


On Sun, Mar 10, 2024 at 03:49:36AM +0000, Ankit Agrawal wrote:
> Bringing this to the fore.
> 
> >> Where does that leave us for this patch? We check the VM_MTE_ALLOWED
> >> and check for ZONE_MOVABLE struct pages as one of the conditions for
> >> NORMAL?
> >
> > I think we should keep it as simple as possible and, looking at it
> > again, maybe even ignore vm_page_prot. Two questions though:
> >
> > 1. Does VM_IO imply vm_page_prot never having MT_NORMAL or
> >    MT_NORMAL_TAGGED?

Drivers do crazy things, so I would not be surprised if something in
the tree does this (wrong) thing. We can check the pgprot in the vma
as well as the VM_IO to make the decision.

> > 2. Do all I/O ranges (side-effects, non-RAM) mapped into a guest (and
> >   which end up in user_mem_abort()) imply VM_IO?

Yes, this is definately true for VFIO.

Jason



More information about the linux-arm-kernel mailing list