Thread (47 messages) 47 messages, 5 authors, 2016-08-24

Re: [PATCH 09/10] vhost, mm: make sure that oom_reaper doesn't reap memory read by vhost

From: Michal Hocko <mhocko@kernel.org>
Date: 2016-08-12 09:43:22

On Fri 29-07-16 16:14:10, Michael S. Tsirkin wrote:
On Fri, Jul 29, 2016 at 08:04:22AM +0200, Michal Hocko wrote:
quoted
On Thu 28-07-16 23:41:53, Michael S. Tsirkin wrote:
[...]
quoted
quoted
quoted
+	 */
+	set_bit(MMF_UNSTABLE, &mm->flags);
+
I would really prefer a callback that vhost would register
and stop all accesses. Tell me if you need help on above idea.

Well, in order to make callback workable the oom reaper would have to
synchronize with the said callback until it declares all currently
ongoing accesses done. That means oom reaper would have to block/wait
and that is something I would really like to prevent from because it
just adds another possibility of the lockup (say the get_user cannot
make forward progress because it is stuck in the page fault allocating
memory). Or do you see any other way how to implement such a callback
mechanism without blocking on the oom_reaper side?
I'll think it over and respond.
ping?
-- 
Michal Hocko
SUSE Labs

--
To unsubscribe, send a message with 'unsubscribe linux-mm' in
the body to majordomo@kvack.org.  For more info on Linux MM,
see: http://www.linux-mm.org/ .
Don't email: <a href=mailto:"dont@kvack.org"> email@kvack.org </a>
Keyboard shortcuts
hback out one level
jnext message in thread
kprevious message in thread
ldrill in
Escclose help / fold thread tree
?toggle this help