On Mon, 19 Jul 2021 14:39:05 +0100,
Quentin Perret [off-list ref] wrote:

On Monday 19 Jul 2021 at 13:55:29 (+0100), Marc Zyngier wrote:

quoted

On Mon, 19 Jul 2021 11:47:26 +0100,
Quentin Perret [off-list ref] wrote:

quoted

The nVHE protected mode uses invalid mappings in the host stage-2
page-table to track the owner of each page in the system. In order to
allow the usage of ignored bits (a.k.a. software bits) in these
mappings, move the owner encoding away from the top bits.

But that's exactly what the current situation is allowing: the use of
the SW bits. I am guessing that what you really mean is that you want
to *preserve* the SW bits from an existing mapping and use other bits
when the mapping is invalid?

Yes, this is really just forward looking, but I think it might be useful
to allow annotating invalid mappings with both an owner id _and_
additional flags for e.g. shared pages and such. And using bits [58-55]
to store those flags just like we do for valid mappings should make
things easier, but no big deal.

Right, so maybe worth calling that out.

I see how this is going to conflict with kvm_pgtable_stage2_annotate()
from your series though, so maybe I should just drop this patch and
leave the encoding 'issue' to the caller -- the rest of the series
doesn't depend on this anyway, this was just small cleanup.

I'm not too worried about that for now. We can always rewrite one in
terms of the other, but I wanted to understand exactly this change was
about.

Thanks,

	M.

-- 
Without deviation from the norm, progress is not possible.

_______________________________________________
linux-arm-kernel mailing list
linux-arm-kernel@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-arm-kernel