* kswapd skips compaction if reclaim order drops to zero?
@ 2013-08-15 10:02 Hillf Danton
2013-08-15 10:47 ` Mel Gorman
0 siblings, 1 reply; 9+ messages in thread
From: Hillf Danton @ 2013-08-15 10:02 UTC (permalink / raw)
To: Mel Gorman; +Cc: Minchan Kim, LKML, Linux-MM
If the allocation order is not high, direct compaction does nothing.
Can we skip compaction here if order drops to zero?
--- a/mm/vmscan.c Thu Aug 15 17:47:26 2013
+++ b/mm/vmscan.c Thu Aug 15 17:48:58 2013
@@ -3034,7 +3034,7 @@ static unsigned long balance_pgdat(pg_da
* Compact if necessary and kswapd is reclaiming at least the
* high watermark number of pages as requsted
*/
- if (pgdat_needs_compaction && sc.nr_reclaimed > nr_attempted)
+ if (pgdat_needs_compaction && sc.nr_reclaimed > nr_attempted && order)
compact_pgdat(pgdat, order);
/*
--
--
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>
^ permalink raw reply [flat|nested] 9+ messages in thread
* Re: kswapd skips compaction if reclaim order drops to zero?
2013-08-15 10:02 kswapd skips compaction if reclaim order drops to zero? Hillf Danton
@ 2013-08-15 10:47 ` Mel Gorman
2013-08-15 13:41 ` Minchan Kim
0 siblings, 1 reply; 9+ messages in thread
From: Mel Gorman @ 2013-08-15 10:47 UTC (permalink / raw)
To: Hillf Danton; +Cc: Minchan Kim, LKML, Linux-MM
On Thu, Aug 15, 2013 at 06:02:53PM +0800, Hillf Danton wrote:
> If the allocation order is not high, direct compaction does nothing.
> Can we skip compaction here if order drops to zero?
>
If the allocation order is not high then
pgdat_needs_compaction == (order > 0) == false == no calling compact_pdatt
In the case where order is reset to 0 due to fragmentation then it does
call compact_pgdat but it does no work due to the cc->order check in
__compact_pgdat.
--
Mel Gorman
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>
^ permalink raw reply [flat|nested] 9+ messages in thread
* Re: kswapd skips compaction if reclaim order drops to zero?
2013-08-15 10:47 ` Mel Gorman
@ 2013-08-15 13:41 ` Minchan Kim
2013-08-15 13:56 ` Mel Gorman
0 siblings, 1 reply; 9+ messages in thread
From: Minchan Kim @ 2013-08-15 13:41 UTC (permalink / raw)
To: Mel Gorman; +Cc: Hillf Danton, LKML, Linux-MM
Hey Mel,
On Thu, Aug 15, 2013 at 11:47:27AM +0100, Mel Gorman wrote:
> On Thu, Aug 15, 2013 at 06:02:53PM +0800, Hillf Danton wrote:
> > If the allocation order is not high, direct compaction does nothing.
> > Can we skip compaction here if order drops to zero?
> >
>
> If the allocation order is not high then
>
> pgdat_needs_compaction == (order > 0) == false == no calling compact_pdatt
>
> In the case where order is reset to 0 due to fragmentation then it does
> call compact_pgdat but it does no work due to the cc->order check in
> __compact_pgdat.
>
I am looking at mmotm-2013-08-07-16-55 but couldn't find cc->order
check right before compact_zone in __comact_pgdat.
Could you pinpoint code piece?
> --
> Mel Gorman
> 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>
--
Kind regards,
Minchan Kim
--
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>
^ permalink raw reply [flat|nested] 9+ messages in thread
* Re: kswapd skips compaction if reclaim order drops to zero?
2013-08-15 13:41 ` Minchan Kim
@ 2013-08-15 13:56 ` Mel Gorman
2013-08-15 14:10 ` Minchan Kim
0 siblings, 1 reply; 9+ messages in thread
From: Mel Gorman @ 2013-08-15 13:56 UTC (permalink / raw)
To: Minchan Kim; +Cc: Hillf Danton, LKML, Linux-MM
On Thu, Aug 15, 2013 at 10:41:39PM +0900, Minchan Kim wrote:
> Hey Mel,
>
> On Thu, Aug 15, 2013 at 11:47:27AM +0100, Mel Gorman wrote:
> > On Thu, Aug 15, 2013 at 06:02:53PM +0800, Hillf Danton wrote:
> > > If the allocation order is not high, direct compaction does nothing.
> > > Can we skip compaction here if order drops to zero?
> > >
> >
> > If the allocation order is not high then
> >
> > pgdat_needs_compaction == (order > 0) == false == no calling compact_pdatt
> >
> > In the case where order is reset to 0 due to fragmentation then it does
> > call compact_pgdat but it does no work due to the cc->order check in
> > __compact_pgdat.
> >
>
> I am looking at mmotm-2013-08-07-16-55 but couldn't find cc->order
> check right before compact_zone in __comact_pgdat.
> Could you pinpoint code piece?
>
Thanks, I screwed up as that check happens too late. However, it still
ends up not mattering because it does this
compact_pgdat
-> __compact_pgdat
-> compact_zone
-> compaction_suitable
For order == 0, compaction_suitable will return either COMPACT_SKIPPED
(if the watermarks are not met) and COMPACT_PARTIAL otherwise. Either
way, compaction doesn't run.
--
Mel Gorman
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>
^ permalink raw reply [flat|nested] 9+ messages in thread
* Re: kswapd skips compaction if reclaim order drops to zero?
2013-08-15 13:56 ` Mel Gorman
@ 2013-08-15 14:10 ` Minchan Kim
2013-08-15 15:39 ` [PATCH] mm: compaction: Do not compact pgdat for order-0 Mel Gorman
` (2 more replies)
0 siblings, 3 replies; 9+ messages in thread
From: Minchan Kim @ 2013-08-15 14:10 UTC (permalink / raw)
To: Mel Gorman; +Cc: Hillf Danton, LKML, Linux-MM
On Thu, Aug 15, 2013 at 02:56:27PM +0100, Mel Gorman wrote:
> On Thu, Aug 15, 2013 at 10:41:39PM +0900, Minchan Kim wrote:
> > Hey Mel,
> >
> > On Thu, Aug 15, 2013 at 11:47:27AM +0100, Mel Gorman wrote:
> > > On Thu, Aug 15, 2013 at 06:02:53PM +0800, Hillf Danton wrote:
> > > > If the allocation order is not high, direct compaction does nothing.
> > > > Can we skip compaction here if order drops to zero?
> > > >
> > >
> > > If the allocation order is not high then
> > >
> > > pgdat_needs_compaction == (order > 0) == false == no calling compact_pdatt
> > >
> > > In the case where order is reset to 0 due to fragmentation then it does
> > > call compact_pgdat but it does no work due to the cc->order check in
> > > __compact_pgdat.
> > >
> >
> > I am looking at mmotm-2013-08-07-16-55 but couldn't find cc->order
> > check right before compact_zone in __comact_pgdat.
> > Could you pinpoint code piece?
> >
>
> Thanks, I screwed up as that check happens too late. However, it still
> ends up not mattering because it does this
>
> compact_pgdat
> -> __compact_pgdat
> -> compact_zone
> -> compaction_suitable
>
> For order == 0, compaction_suitable will return either COMPACT_SKIPPED
> (if the watermarks are not met) and COMPACT_PARTIAL otherwise. Either
> way, compaction doesn't run.
In compaction_suitable, it could pass first zone_watermark_ok
but failed second zone_watermark_ok while fragindex is -1000
so compaction could run.
And we shouldn't depend on such coincidence.
>
> --
> Mel Gorman
> 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>
--
Kind regards,
Minchan Kim
--
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>
^ permalink raw reply [flat|nested] 9+ messages in thread
* [PATCH] mm: compaction: Do not compact pgdat for order-0
2013-08-15 14:10 ` Minchan Kim
@ 2013-08-15 15:39 ` Mel Gorman
2013-08-16 4:37 ` Minchan Kim
2013-08-16 0:25 ` kswapd skips compaction if reclaim order drops to zero? Wanpeng Li
2013-08-16 0:25 ` Wanpeng Li
2 siblings, 1 reply; 9+ messages in thread
From: Mel Gorman @ 2013-08-15 15:39 UTC (permalink / raw)
To: Andrew Morton; +Cc: Hillf Danton, Minchan Kim, LKML, Linux-MM
If kswapd was reclaiming for a high order and resets it to 0 due to
fragmentation it will still call compact_pgdat. For the most part, this will
fail a compaction_suitable() test and not compact but it is unnecessarily
sloppy. It could be fixed in the caller but fix it in the API instead.
[dhillf@gmail.com: Pointed out that it was a potential problem]
Signed-off-by: Mel Gorman <mgorman@suse.de>
---
mm/compaction.c | 3 +++
1 file changed, 3 insertions(+)
diff --git a/mm/compaction.c b/mm/compaction.c
index 05ccb4c..c437893 100644
--- a/mm/compaction.c
+++ b/mm/compaction.c
@@ -1131,6 +1131,9 @@ void compact_pgdat(pg_data_t *pgdat, int order)
.sync = false,
};
+ if (!order)
+ return;
+
__compact_pgdat(pgdat, &cc);
}
--
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>
^ permalink raw reply related [flat|nested] 9+ messages in thread
* Re: kswapd skips compaction if reclaim order drops to zero?
2013-08-15 14:10 ` Minchan Kim
2013-08-15 15:39 ` [PATCH] mm: compaction: Do not compact pgdat for order-0 Mel Gorman
2013-08-16 0:25 ` kswapd skips compaction if reclaim order drops to zero? Wanpeng Li
@ 2013-08-16 0:25 ` Wanpeng Li
2 siblings, 0 replies; 9+ messages in thread
From: Wanpeng Li @ 2013-08-16 0:25 UTC (permalink / raw)
To: Minchan Kim; +Cc: Mel Gorman, Hillf Danton, LKML, Linux-MM
Hi Minchan,
On Thu, Aug 15, 2013 at 11:10:04PM +0900, Minchan Kim wrote:
>On Thu, Aug 15, 2013 at 02:56:27PM +0100, Mel Gorman wrote:
>> On Thu, Aug 15, 2013 at 10:41:39PM +0900, Minchan Kim wrote:
>> > Hey Mel,
>> >
>> > On Thu, Aug 15, 2013 at 11:47:27AM +0100, Mel Gorman wrote:
>> > > On Thu, Aug 15, 2013 at 06:02:53PM +0800, Hillf Danton wrote:
>> > > > If the allocation order is not high, direct compaction does nothing.
>> > > > Can we skip compaction here if order drops to zero?
>> > > >
>> > >
>> > > If the allocation order is not high then
>> > >
>> > > pgdat_needs_compaction == (order > 0) == false == no calling compact_pdatt
>> > >
>> > > In the case where order is reset to 0 due to fragmentation then it does
>> > > call compact_pgdat but it does no work due to the cc->order check in
>> > > __compact_pgdat.
>> > >
>> >
>> > I am looking at mmotm-2013-08-07-16-55 but couldn't find cc->order
>> > check right before compact_zone in __comact_pgdat.
>> > Could you pinpoint code piece?
>> >
>>
>> Thanks, I screwed up as that check happens too late. However, it still
>> ends up not mattering because it does this
>>
>> compact_pgdat
>> -> __compact_pgdat
>> -> compact_zone
>> -> compaction_suitable
>>
>> For order == 0, compaction_suitable will return either COMPACT_SKIPPED
>> (if the watermarks are not met) and COMPACT_PARTIAL otherwise. Either
>> way, compaction doesn't run.
>
>In compaction_suitable, it could pass first zone_watermark_ok
>but failed second zone_watermark_ok while fragindex is -1000
>so compaction could run.
>
I'm not sure why you said that second zone_watermark_ok failed while
fragindex is -1000, actually they are the same check against order 0.
First:
watermark = low_wmark_pages(zone) + (2UL << order);
zone_watermark_ok(zone, 0, watermark, 0, 0);
Second:
fragindex == -1000 && zone_watermark_ok(zone, order, watermark, 0, 0)
Both COMPACT_SKIPPED and COMPACT_PARTIAL will fail compaction.
Regards,
Wanpeng Li
>And we shouldn't depend on such coincidence.
>
>
>>
>> --
>> Mel Gorman
>> 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>
>
>--
>Kind regards,
>Minchan Kim
>
>--
>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>
--
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>
^ permalink raw reply [flat|nested] 9+ messages in thread
* Re: kswapd skips compaction if reclaim order drops to zero?
2013-08-15 14:10 ` Minchan Kim
2013-08-15 15:39 ` [PATCH] mm: compaction: Do not compact pgdat for order-0 Mel Gorman
@ 2013-08-16 0:25 ` Wanpeng Li
2013-08-16 0:25 ` Wanpeng Li
2 siblings, 0 replies; 9+ messages in thread
From: Wanpeng Li @ 2013-08-16 0:25 UTC (permalink / raw)
To: Minchan Kim; +Cc: Mel Gorman, Hillf Danton, LKML, Linux-MM
Hi Minchan,
On Thu, Aug 15, 2013 at 11:10:04PM +0900, Minchan Kim wrote:
>On Thu, Aug 15, 2013 at 02:56:27PM +0100, Mel Gorman wrote:
>> On Thu, Aug 15, 2013 at 10:41:39PM +0900, Minchan Kim wrote:
>> > Hey Mel,
>> >
>> > On Thu, Aug 15, 2013 at 11:47:27AM +0100, Mel Gorman wrote:
>> > > On Thu, Aug 15, 2013 at 06:02:53PM +0800, Hillf Danton wrote:
>> > > > If the allocation order is not high, direct compaction does nothing.
>> > > > Can we skip compaction here if order drops to zero?
>> > > >
>> > >
>> > > If the allocation order is not high then
>> > >
>> > > pgdat_needs_compaction == (order > 0) == false == no calling compact_pdatt
>> > >
>> > > In the case where order is reset to 0 due to fragmentation then it does
>> > > call compact_pgdat but it does no work due to the cc->order check in
>> > > __compact_pgdat.
>> > >
>> >
>> > I am looking at mmotm-2013-08-07-16-55 but couldn't find cc->order
>> > check right before compact_zone in __comact_pgdat.
>> > Could you pinpoint code piece?
>> >
>>
>> Thanks, I screwed up as that check happens too late. However, it still
>> ends up not mattering because it does this
>>
>> compact_pgdat
>> -> __compact_pgdat
>> -> compact_zone
>> -> compaction_suitable
>>
>> For order == 0, compaction_suitable will return either COMPACT_SKIPPED
>> (if the watermarks are not met) and COMPACT_PARTIAL otherwise. Either
>> way, compaction doesn't run.
>
>In compaction_suitable, it could pass first zone_watermark_ok
>but failed second zone_watermark_ok while fragindex is -1000
>so compaction could run.
>
I'm not sure why you said that second zone_watermark_ok failed while
fragindex is -1000, actually they are the same check against order 0.
First:
watermark = low_wmark_pages(zone) + (2UL << order);
zone_watermark_ok(zone, 0, watermark, 0, 0);
Second:
fragindex == -1000 && zone_watermark_ok(zone, order, watermark, 0, 0)
Both COMPACT_SKIPPED and COMPACT_PARTIAL will fail compaction.
Regards,
Wanpeng Li
>And we shouldn't depend on such coincidence.
>
>
>>
>> --
>> Mel Gorman
>> 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>
>
>--
>Kind regards,
>Minchan Kim
>
>--
>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>
--
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>
^ permalink raw reply [flat|nested] 9+ messages in thread
* Re: [PATCH] mm: compaction: Do not compact pgdat for order-0
2013-08-15 15:39 ` [PATCH] mm: compaction: Do not compact pgdat for order-0 Mel Gorman
@ 2013-08-16 4:37 ` Minchan Kim
0 siblings, 0 replies; 9+ messages in thread
From: Minchan Kim @ 2013-08-16 4:37 UTC (permalink / raw)
To: Mel Gorman; +Cc: Andrew Morton, Hillf Danton, LKML, Linux-MM
On Thu, Aug 15, 2013 at 04:39:27PM +0100, Mel Gorman wrote:
> If kswapd was reclaiming for a high order and resets it to 0 due to
> fragmentation it will still call compact_pgdat. For the most part, this will
> fail a compaction_suitable() test and not compact but it is unnecessarily
> sloppy. It could be fixed in the caller but fix it in the API instead.
>
> [dhillf@gmail.com: Pointed out that it was a potential problem]
> Signed-off-by: Mel Gorman <mgorman@suse.de>
Acked-by: Minchan Kim <minchan@kernel.org>
--
Kind regards,
Minchan Kim
--
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>
^ permalink raw reply [flat|nested] 9+ messages in thread
end of thread, other threads:[~2013-08-16 4:37 UTC | newest]
Thread overview: 9+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2013-08-15 10:02 kswapd skips compaction if reclaim order drops to zero? Hillf Danton
2013-08-15 10:47 ` Mel Gorman
2013-08-15 13:41 ` Minchan Kim
2013-08-15 13:56 ` Mel Gorman
2013-08-15 14:10 ` Minchan Kim
2013-08-15 15:39 ` [PATCH] mm: compaction: Do not compact pgdat for order-0 Mel Gorman
2013-08-16 4:37 ` Minchan Kim
2013-08-16 0:25 ` kswapd skips compaction if reclaim order drops to zero? Wanpeng Li
2013-08-16 0:25 ` Wanpeng Li
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).