All of lore.kernel.org
 help / color / mirror / Atom feed
* page ownership
@ 2007-03-08 10:10 Jan Beulich
  2007-03-08 10:23 ` Keir Fraser
  0 siblings, 1 reply; 2+ messages in thread
From: Jan Beulich @ 2007-03-08 10:10 UTC (permalink / raw)
  To: xen-devel

Is there a particular reason why page_set_owner(pg, NULL) is
- not used in free_domheap_pages()?
- only used on pages 1 ... ((1 << order) - 1) in __alloc_domheap_pages()?

Thanks, Jan

^ permalink raw reply	[flat|nested] 2+ messages in thread

* Re: page ownership
  2007-03-08 10:10 page ownership Jan Beulich
@ 2007-03-08 10:23 ` Keir Fraser
  0 siblings, 0 replies; 2+ messages in thread
From: Keir Fraser @ 2007-03-08 10:23 UTC (permalink / raw)
  To: Jan Beulich, xen-devel

On 8/3/07 10:10, "Jan Beulich" <jbeulich@novell.com> wrote:

> Is there a particular reason why page_set_owner(pg, NULL) is
> - not used in free_domheap_pages()?

It's done in __alloc_domheap_pages() instead. Arguably moving it would be
cleaner.

> - only used on pages 1 ... ((1 << order) - 1) in __alloc_domheap_pages()?

Page 0 is done outside the loop, and writes directly to _domain for no
particularly good reason. I guess I implemented the first iteration outside
the loop to special-case initialisation of the tlbflush cpumask. Which
probably isn't a very worthwhile optimisation, to be honest.

I'll do a little cleanup here. It would certainly be sane to not have random
domain pointers hanging around in freed pages.

 -- Keir

^ permalink raw reply	[flat|nested] 2+ messages in thread

end of thread, other threads:[~2007-03-08 10:23 UTC | newest]

Thread overview: 2+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2007-03-08 10:10 page ownership Jan Beulich
2007-03-08 10:23 ` Keir Fraser

This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.