[Xen-devel] Re: [PATCH 4/6] mm: New XENMEM space, XENMAPSPACE_gmfn_range

["Jan Beulich" <JBeulich@xxxxxxxx>]

>>> On 10.11.11 at 11:21, Keir Fraser <keir@xxxxxxx> wrote:
> On 10/11/2011 09:54, "Jan Beulich" <JBeulich@xxxxxxxx> wrote:
>> Further (I realize I should have commented on this earlier) I think that
>> in order to allow forward progress you should not check for preemption
>> on the very first iteration of each (re-)invocation. That would also
>> guarantee no behavioral change to the original single-page variants.
> 
> There are plenty of other examples where we check for preemption before
> doing any real work (eg. do_mmuext_op, do_mmu_update).

And I never really liked that, because of the very potential to live lock.

> I guess checking at
> the end of the loop is a little bit better maybe. I'm not very bothered
> either way.

Yes, and then only if further iterations are needed.

Jan


_______________________________________________
Xen-devel mailing list
Xen-devel@xxxxxxxxxxxxxxxxxxx
http://lists.xensource.com/xen-devel