On Wed, Feb 04, 2015 at 03:39:25PM +0100, Vlastimil Babka wrote:
> On 02/03/2015 06:07 PM, Joonsoo Kim wrote:
> >2015-02-04 0:51 GMT+09:00 Vlastimil Babka :
> >>Ah, I think I see where the misunderstanding comes from now. So to clarify,
> >>let's consider
> >>
> >>1. single compaction run - single
On 02/03/2015 06:07 PM, Joonsoo Kim wrote:
2015-02-04 0:51 GMT+09:00 Vlastimil Babka :
Ah, I think I see where the misunderstanding comes from now. So to clarify,
let's consider
1. single compaction run - single invocation of compact_zone(). It can start
from cached pfn's from previous run, or
2015-02-04 0:51 GMT+09:00 Vlastimil Babka :
> On 02/03/2015 04:00 PM, Joonsoo Kim wrote:
>> 2015-02-03 18:05 GMT+09:00 Vlastimil Babka :
>>> On 02/03/2015 07:49 AM, Joonsoo Kim wrote:
On Mon, Jan 19, 2015 at 11:05:15AM +0100, Vlastimil Babka wrote:
Hello,
I don't have any e
On 02/03/2015 04:00 PM, Joonsoo Kim wrote:
> 2015-02-03 18:05 GMT+09:00 Vlastimil Babka :
>> On 02/03/2015 07:49 AM, Joonsoo Kim wrote:
>>> On Mon, Jan 19, 2015 at 11:05:15AM +0100, Vlastimil Babka wrote:
>>>
>>> Hello,
>>>
>>> I don't have any elegant idea, but, have some humble opinion.
>>>
>>> T
2015-02-03 18:05 GMT+09:00 Vlastimil Babka :
> On 02/03/2015 07:49 AM, Joonsoo Kim wrote:
>> On Mon, Jan 19, 2015 at 11:05:15AM +0100, Vlastimil Babka wrote:
>>
>> Hello,
>>
>> I don't have any elegant idea, but, have some humble opinion.
>>
>> The point is that migrate scanner should scan whole zo
On 02/03/2015 07:49 AM, Joonsoo Kim wrote:
> On Mon, Jan 19, 2015 at 11:05:15AM +0100, Vlastimil Babka wrote:
>
> Hello,
>
> I don't have any elegant idea, but, have some humble opinion.
>
> The point is that migrate scanner should scan whole zone.
> Although your pivot approach makes some sense
On Mon, Jan 19, 2015 at 11:05:15AM +0100, Vlastimil Babka wrote:
> Even after all the patches compaction received in last several versions, it
> turns out that its effectivneess degrades considerably as the system ages
> after reboot. For example, see how success rates of stress-highalloc from
> mm
On 01/19/2015 11:05 AM, Vlastimil Babka wrote:
> Preliminary testing with THP-like allocations has shown similar improvements,
> which is somewhat surprising, because THP allocations do not use sync
> and thus do not defer compaction; but changing the pivot is currently tied
> to restarting from de
Looks like wrapping got busted in the results part. Let's try again:
-
Even after all the patches compaction received in last several versions, it
turns out that its effectivneess degrades considerably as the system ages
after reboot. For example, see how success rates of stress-highalloc from
Even after all the patches compaction received in last several versions, it
turns out that its effectivneess degrades considerably as the system ages
after reboot. For example, see how success rates of stress-highalloc from
mmtests degrades when we re-execute it several times, first time being afte
10 matches
Mail list logo
