mm: compaction: Avoid 100% CPU usage during compaction when a task is killed
diff mbox series

Message ID 20190718085708.GE24383@techsingularity.net
State New
Headers show
Series
  • mm: compaction: Avoid 100% CPU usage during compaction when a task is killed
Related show

Commit Message

Mel Gorman July 18, 2019, 8:57 a.m. UTC
"howaboutsynergy" reported via kernel buzilla number 204165 that
compact_zone_order was consuming 100% CPU during a stress test for
prolonged periods of time. Specifically the following command, which
should exit in 10 seconds, was taking an excessive time to finish while
the CPU was pegged at 100%.

  stress -m 220 --vm-bytes 1000000000 --timeout 10

Tracing indicated a pattern as follows

          stress-3923  [007]   519.106208: mm_compaction_isolate_migratepages: range=(0x70bb80 ~ 0x70bb80) nr_scanned=0 nr_taken=0
          stress-3923  [007]   519.106212: mm_compaction_isolate_migratepages: range=(0x70bb80 ~ 0x70bb80) nr_scanned=0 nr_taken=0
          stress-3923  [007]   519.106216: mm_compaction_isolate_migratepages: range=(0x70bb80 ~ 0x70bb80) nr_scanned=0 nr_taken=0
          stress-3923  [007]   519.106219: mm_compaction_isolate_migratepages: range=(0x70bb80 ~ 0x70bb80) nr_scanned=0 nr_taken=0
          stress-3923  [007]   519.106223: mm_compaction_isolate_migratepages: range=(0x70bb80 ~ 0x70bb80) nr_scanned=0 nr_taken=0
          stress-3923  [007]   519.106227: mm_compaction_isolate_migratepages: range=(0x70bb80 ~ 0x70bb80) nr_scanned=0 nr_taken=0
          stress-3923  [007]   519.106231: mm_compaction_isolate_migratepages: range=(0x70bb80 ~ 0x70bb80) nr_scanned=0 nr_taken=0
          stress-3923  [007]   519.106235: mm_compaction_isolate_migratepages: range=(0x70bb80 ~ 0x70bb80) nr_scanned=0 nr_taken=0
          stress-3923  [007]   519.106238: mm_compaction_isolate_migratepages: range=(0x70bb80 ~ 0x70bb80) nr_scanned=0 nr_taken=0
          stress-3923  [007]   519.106242: mm_compaction_isolate_migratepages: range=(0x70bb80 ~ 0x70bb80) nr_scanned=0 nr_taken=0

Note that compaction is entered in rapid succession while scanning and
isolating nothing. The problem is that when a task that is compacting
receives a fatal signal, it retries indefinitely instead of exiting while
making no progress as a fatal signal is pending.

It's not easy to trigger this condition although enabling zswap helps on
the basis that the timing is altered. A very small window has to be hit
for the problem to occur (signal delivered while compacting and isolating
a PFN for migration that is not aligned to SWAP_CLUSTER_MAX).

This was reproduced locally -- 16G single socket system, 8G swap, 30% zswap
configured, vm-bytes 22000000000 using Colin Kings stress-ng implementation
from github running in a loop until the problem hits). Tracing recorded the
problem occurring almost 200K times in a short window. With this patch, the
problem hit 4 times but the task existed normally instead of consuming CPU.

This problem has existed for some time but it was made worse by
cf66f0700c8f ("mm, compaction: do not consider a need to reschedule as
contention"). Before that commit, if the same condition was hit then
locks would be quickly contended and compaction would exit that way.

I haven't included a Reported-and-tested-by as the reporters real name
is unknown but this was caught and repaired due to their testing and
tracing.  If they want a tag added then hopefully they'll say so before
this gets merged.

Bugzilla: https://bugzilla.kernel.org/show_bug.cgi?id=204165
Fixes: cf66f0700c8f ("mm, compaction: do not consider a need to reschedule as contention")
Signed-off-by: Mel Gorman <mgorman@techsingularity.net>
CC: stable@vger.kernel.org # v5.1+
---
 mm/compaction.c | 11 +++++++----
 1 file changed, 7 insertions(+), 4 deletions(-)

Comments

howaboutsynergy@protonmail.com July 18, 2019, 11:48 a.m. UTC | #1
‐‐‐‐‐‐‐ Original Message ‐‐‐‐‐‐‐
On Thursday, July 18, 2019 10:57 AM, Mel Gorman <mgorman@techsingularity.net> wrote:

> "howaboutsynergy" reported via kernel buzilla number 204165 that
<SNIP>

> I haven't included a Reported-and-tested-by as the reporters real name
> is unknown but this was caught and repaired due to their testing and
> tracing. If they want a tag added then hopefully they'll say so before
> this gets merged.
>
nope, don't want :)

Thanks a lot for your work, time, understanding-how-things-work and conciseness(level over 9000), Mel Gorman. Much appreciated, enjoyed the read and happy to see this fixed! Hooray!

Best of luck.
Cheers!
Andrew Morton July 18, 2019, 9:22 p.m. UTC | #2
On Thu, 18 Jul 2019 11:48:10 +0000 howaboutsynergy@protonmail.com wrote:

> > "howaboutsynergy" reported via kernel buzilla number 204165 that
> <SNIP>
> 
> > I haven't included a Reported-and-tested-by as the reporters real name
> > is unknown but this was caught and repaired due to their testing and
> > tracing. If they want a tag added then hopefully they'll say so before
> > this gets merged.
> >
> nope, don't want :)

I added them:

Reported-by: <howaboutsynergy@protonmail.com>
Tested-by: <howaboutsynergy@protonmail.com>

Having a contact email address is potentially useful.  I'd be more
concerned about anonymity if it involved an actual code modification.

And thanks for your persistence and assistance.
Vlastimil Babka July 25, 2019, 1:02 p.m. UTC | #3
On 7/18/19 10:57 AM, Mel Gorman wrote:
> "howaboutsynergy" reported via kernel buzilla number 204165 that
> compact_zone_order was consuming 100% CPU during a stress test for
> prolonged periods of time. Specifically the following command, which
> should exit in 10 seconds, was taking an excessive time to finish while
> the CPU was pegged at 100%.
> 
>   stress -m 220 --vm-bytes 1000000000 --timeout 10
> 
> Tracing indicated a pattern as follows
> 
>           stress-3923  [007]   519.106208: mm_compaction_isolate_migratepages: range=(0x70bb80 ~ 0x70bb80) nr_scanned=0 nr_taken=0
>           stress-3923  [007]   519.106212: mm_compaction_isolate_migratepages: range=(0x70bb80 ~ 0x70bb80) nr_scanned=0 nr_taken=0
>           stress-3923  [007]   519.106216: mm_compaction_isolate_migratepages: range=(0x70bb80 ~ 0x70bb80) nr_scanned=0 nr_taken=0
>           stress-3923  [007]   519.106219: mm_compaction_isolate_migratepages: range=(0x70bb80 ~ 0x70bb80) nr_scanned=0 nr_taken=0
>           stress-3923  [007]   519.106223: mm_compaction_isolate_migratepages: range=(0x70bb80 ~ 0x70bb80) nr_scanned=0 nr_taken=0
>           stress-3923  [007]   519.106227: mm_compaction_isolate_migratepages: range=(0x70bb80 ~ 0x70bb80) nr_scanned=0 nr_taken=0
>           stress-3923  [007]   519.106231: mm_compaction_isolate_migratepages: range=(0x70bb80 ~ 0x70bb80) nr_scanned=0 nr_taken=0
>           stress-3923  [007]   519.106235: mm_compaction_isolate_migratepages: range=(0x70bb80 ~ 0x70bb80) nr_scanned=0 nr_taken=0
>           stress-3923  [007]   519.106238: mm_compaction_isolate_migratepages: range=(0x70bb80 ~ 0x70bb80) nr_scanned=0 nr_taken=0
>           stress-3923  [007]   519.106242: mm_compaction_isolate_migratepages: range=(0x70bb80 ~ 0x70bb80) nr_scanned=0 nr_taken=0
> 
> Note that compaction is entered in rapid succession while scanning and
> isolating nothing. The problem is that when a task that is compacting
> receives a fatal signal, it retries indefinitely instead of exiting while
> making no progress as a fatal signal is pending.
> 
> It's not easy to trigger this condition although enabling zswap helps on
> the basis that the timing is altered. A very small window has to be hit
> for the problem to occur (signal delivered while compacting and isolating
> a PFN for migration that is not aligned to SWAP_CLUSTER_MAX).
> 
> This was reproduced locally -- 16G single socket system, 8G swap, 30% zswap
> configured, vm-bytes 22000000000 using Colin Kings stress-ng implementation
> from github running in a loop until the problem hits). Tracing recorded the
> problem occurring almost 200K times in a short window. With this patch, the
> problem hit 4 times but the task existed normally instead of consuming CPU.
> 
> This problem has existed for some time but it was made worse by
> cf66f0700c8f ("mm, compaction: do not consider a need to reschedule as
> contention"). Before that commit, if the same condition was hit then
> locks would be quickly contended and compaction would exit that way.
> 
> I haven't included a Reported-and-tested-by as the reporters real name
> is unknown but this was caught and repaired due to their testing and
> tracing.  If they want a tag added then hopefully they'll say so before
> this gets merged.
> 
> Bugzilla: https://bugzilla.kernel.org/show_bug.cgi?id=204165
> Fixes: cf66f0700c8f ("mm, compaction: do not consider a need to reschedule as contention")
> Signed-off-by: Mel Gorman <mgorman@techsingularity.net>
> CC: stable@vger.kernel.org # v5.1+

Reviewed-by: Vlastimil Babka <vbabka@suse.cz>

> ---
>  mm/compaction.c | 11 +++++++----
>  1 file changed, 7 insertions(+), 4 deletions(-)
> 
> diff --git a/mm/compaction.c b/mm/compaction.c
> index 9e1b9acb116b..952dc2fb24e5 100644
> --- a/mm/compaction.c
> +++ b/mm/compaction.c
> @@ -842,13 +842,15 @@ isolate_migratepages_block(struct compact_control *cc, unsigned long low_pfn,
>  
>  		/*
>  		 * Periodically drop the lock (if held) regardless of its
> -		 * contention, to give chance to IRQs. Abort async compaction
> -		 * if contended.
> +		 * contention, to give chance to IRQs. Abort completely if
> +		 * a fatal signal is pending.
>  		 */
>  		if (!(low_pfn % SWAP_CLUSTER_MAX)
>  		    && compact_unlock_should_abort(&pgdat->lru_lock,
> -					    flags, &locked, cc))
> -			break;
> +					    flags, &locked, cc)) {
> +			low_pfn = 0;
> +			goto fatal_pending;
> +		}
>  
>  		if (!pfn_valid_within(low_pfn))
>  			goto isolate_fail;
> @@ -1060,6 +1062,7 @@ isolate_migratepages_block(struct compact_control *cc, unsigned long low_pfn,
>  	trace_mm_compaction_isolate_migratepages(start_pfn, low_pfn,
>  						nr_scanned, nr_isolated);
>  
> +fatal_pending:
>  	cc->total_migrate_scanned += nr_scanned;
>  	if (nr_isolated)
>  		count_compact_events(COMPACTISOLATED, nr_isolated);
>

Patch
diff mbox series

diff --git a/mm/compaction.c b/mm/compaction.c
index 9e1b9acb116b..952dc2fb24e5 100644
--- a/mm/compaction.c
+++ b/mm/compaction.c
@@ -842,13 +842,15 @@  isolate_migratepages_block(struct compact_control *cc, unsigned long low_pfn,
 
 		/*
 		 * Periodically drop the lock (if held) regardless of its
-		 * contention, to give chance to IRQs. Abort async compaction
-		 * if contended.
+		 * contention, to give chance to IRQs. Abort completely if
+		 * a fatal signal is pending.
 		 */
 		if (!(low_pfn % SWAP_CLUSTER_MAX)
 		    && compact_unlock_should_abort(&pgdat->lru_lock,
-					    flags, &locked, cc))
-			break;
+					    flags, &locked, cc)) {
+			low_pfn = 0;
+			goto fatal_pending;
+		}
 
 		if (!pfn_valid_within(low_pfn))
 			goto isolate_fail;
@@ -1060,6 +1062,7 @@  isolate_migratepages_block(struct compact_control *cc, unsigned long low_pfn,
 	trace_mm_compaction_isolate_migratepages(start_pfn, low_pfn,
 						nr_scanned, nr_isolated);
 
+fatal_pending:
 	cc->total_migrate_scanned += nr_scanned;
 	if (nr_isolated)
 		count_compact_events(COMPACTISOLATED, nr_isolated);