Re: [Xen-devel] [PATCH] EPT: utilize GLA->GPA translation known for certain faults

[Tim Deegan <tim@xxxxxxx>]

At 10:01 +0100 on 29 Aug (1409302864), Jan Beulich wrote:
> > If that's the behaviour you want then I think we can simplify the test
> > a lot because we don't care whether this is an ifetch or not; only
> > whether the mapping is valid (assuming we model a coherent
> > icache/dcache and itlb/dtlb here).
> 
> Perhaps, but I wanted to be as restrictive as possible (and the hit
> rate of the accelerated case of over 99.9% tells me that this extra
> [possibly paranoid] care doesn't really hurt).

OK, sure.  I withdraw the objection and the patch can have my
reviewed-by: as it stands. 

> Otherwise I can't see
> why we couldn't drop the read_access check too, and similarly the
> write_access one in the write path.

We probably can drop those, yes.

Cheers,

Tim.

_______________________________________________
Xen-devel mailing list
Xen-devel@xxxxxxxxxxxxx
http://lists.xen.org/xen-devel