On 04/24/2013 12:24 AM, Anton Vorontsov wrote:
On Tue, Apr 23, 2013 at 12:22:08PM +0400, Glauber Costa wrote:
From: Glauber Costa <glommer@xxxxxxxxxxxxx>
This patch extends that to also support in-kernel users. Events that
should be generated for in-kernel consumption will be marked as such,
and for those, we will call a registered function instead of triggering
an eventfd notification.
Just a couple more questions... :-)
[...]
@@ -238,14 +244,16 @@ void vmpressure(gfp_t gfp, struct mem_cgroup *memcg,
* through vmpressure_prio(). But so far, keep calm.
*/
if (!scanned)
- return;
+ goto schedule;
mutex_lock(&vmpr->sr_lock);
vmpr->scanned += scanned;
vmpr->reclaimed += reclaimed;
+ vmpr->notify_userspace = true;
Setting the variable on every event seems a bit wasteful... does it make
sense to set it in vmpressure_register_event()? We'll have to make it a
counter, but the good thing is that we won't need any additional locks for
the counter.
Yes, vmpressure_register_event would be a better place for it. I will
change and keep the acks, since it does not change the spirit of the
patch too much.
I will also apply the cosmetics you attached. Thanks
--
To unsubscribe from this list: send the line "unsubscribe cgroups" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
