On Sun, Jan 18, 2015 at 11:12:47PM -0500, Ted Unangst wrote:
> On Mon, Jan 19, 2015 at 11:28, David Gwynne wrote:
>
> >> if you're interested in seeing the effect of freeing at different
> >> intervals, you could try the diff below. it adds kern.pool.wait_free
> >> and kern.pool.wait_gc tunables s
On Mon, Jan 19, 2015 at 11:28, David Gwynne wrote:
>> if you're interested in seeing the effect of freeing at different
>> intervals, you could try the diff below. it adds kern.pool.wait_free
>> and kern.pool.wait_gc tunables so you can set how long a page has
>> to be idle before a pool_put and t
> On 23 Dec 2014, at 11:38, David Gwynne wrote:
>
> On Mon, Dec 22, 2014 at 10:54:16AM -0500, Ted Unangst wrote:
>> On Mon, Dec 22, 2014 at 14:59, Mike Belopuhov wrote:
>>> 1) how is it different from what we have now?
>>>
>>> 2) why can't you do it when you pool_put?
>>
>> Right now, if you a
On Mon, Dec 22, 2014 at 10:54:16AM -0500, Ted Unangst wrote:
> On Mon, Dec 22, 2014 at 14:59, Mike Belopuhov wrote:
> > 1) how is it different from what we have now?
> >
> > 2) why can't you do it when you pool_put?
>
> Right now, if you allocate a bunch of things, then free them, the
> pages won
On Mon, Dec 22, 2014 at 14:59, Mike Belopuhov wrote:
> 1) how is it different from what we have now?
>
> 2) why can't you do it when you pool_put?
Right now, if you allocate a bunch of things, then free them, the
pages won't be freed because the timestamp is new. But you've already
freed everythi
On 22 December 2014 at 06:43, David Gwynne wrote:
> this introduces a global gc task that loops over all the pools
> looking for pages that havent been used for a very long time so
> they can be freed.
>
> this is the simplest way of doing this without introducing per pool
> timeouts/tasks which i
this introduces a global gc task that loops over all the pools
looking for pages that havent been used for a very long time so
they can be freed.
this is the simplest way of doing this without introducing per pool
timeouts/tasks which in turn could introduce races with pool_destroy,
or more shared
