Re: [Xen-devel] [PATCH V3] tools/libxc, xen/x86: Added xc_set_mem_access_multi() [and 1 more messages]

[Ian Jackson <ian.jackson@xxxxxxxxxxxxx>]

Razvan Cojocaru writes ("Re: [PATCH V3] tools/libxc, xen/x86: Added 
xc_set_mem_access_multi()"):
> On 09/06/2016 01:16 PM, Ian Jackson wrote:
> > Could this aim be achieved with a multicall ?  (Can multicalls defer
> > the TLB flush?)
> 
> I assume your question is: could we do multiple xc_set_mem_access()
> calls and then call something like xc_tlb_flush() at the end of it,
> instead of a singke xc_set_mem_access_multi() call?

No, I was talking about the guest<->hypervisor interface.

Xen has a feature called "multicalls" which allows a single hypercall
to actually contain several hypercalls, so that a single pair of
guest<->hypervisor transitions can do more work.

Jan Beulich writes ("Re: [PATCH V3] tools/libxc, xen/x86: Added 
xc_set_mem_access_multi()"):
> On 06.09.16 at 12:16, <ian.jackson@xxxxxxxxxxxxx> wrote:
> > Could this aim be achieved with a multicall ?  (Can multicalls defer
> > the TLB flush?)
> 
> No, they can't, but it's not entirely excluded to make them do so.

AIUI from reading the patch the TLB flush is implicit, so extra code
complexity (and fragility) would be needed to implement that.

> And then iirc there are no multicalls available to HVM (and hence
> PVHv2) guests right now.

This is another problem with my suggestion.

So, thanks for the reply.  Carry on :-).

Ian.

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