[dm-devel] [PATCH] dm-table: delayed cleanup for dm_table_destroy()
mpatocka at redhat.com
Wed Mar 21 01:24:12 UTC 2012
On Tue, 20 Mar 2012, Hannes Reinecke wrote:
> Ah. That'll explain it.
> The actual problem I'm trying to track down is that I'm seeing an
> excessive duration for the 'resume' dm ioctl after a table update.
> I've got reports where the ioctl can take up to several seconds.
> Which (as this is multipath) causes an extremely erratic behaviour.
> And the 'msleep' here is one of the obvious culprits.
> But I'll continue debugging, maybe I'll find something else.
If it takes several second, it is not obviously problem with msleep(1) ---
if we converted msleep(1) to a wait queue, it would improve by at most
1ms, but it couldn't improve by a few seconds.
More information about the dm-devel