[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [Xen-devel] [PATCH] fix memory exchange hypercall.

On 16/3/07 03:25, "Isaku Yamahata" <yamahata@xxxxxxxxxxxxx> wrote:

> The ia64 p2m already adopted the first option without warning.
> So the guest_physmap_remove_page() depends on PGC_allocated bit.
> What's the expected behavior of guset_physmap_remove_page() with
> PGC_allocated? It shouldn't touch the bit?
> Thanks a lot for your clarification.

I don't think that guest_physmap_remove_page() should need to touch the
PGC_allocated bit. All callers of that function can take responsibility for
the page they unhook (that's the case for all the callers from common code,
at least). I'm more worried about callers of guest_physmap_add_page(): it
may not be obvious to users of that function if they have to take
responsibility for what used to be mapped at that pseudophysical address!

 -- Keir

Xen-devel mailing list



Lists.xenproject.org is hosted with RackSpace, monitoring our
servers 24x7x365 and backed by RackSpace's Fanatical Support®.