linux-kernel - Re: [PATCH 01/10] mm, page_alloc: Delete the zonelist

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives
Hash Suite for Android: free password hash cracker in your pocket
[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Date:	Thu, 20 Aug 2015 14:42:40 +0100
From:	Mel Gorman <mgorman@...hsingularity.net>
To:	Michal Hocko <mhocko@...nel.org>
Cc:	Linux-MM <linux-mm@...ck.org>,
	Johannes Weiner <hannes@...xchg.org>,
	Rik van Riel <riel@...hat.com>,
	Vlastimil Babka <vbabka@...e.cz>,
	David Rientjes <rientjes@...gle.com>,
	Joonsoo Kim <iamjoonsoo.kim@....com>,
	LKML <linux-kernel@...r.kernel.org>
Subject: Re: [PATCH 01/10] mm, page_alloc: Delete the zonelist_cache

On Thu, Aug 20, 2015 at 03:18:43PM +0200, Michal Hocko wrote:
> On Wed 12-08-15 11:45:26, Mel Gorman wrote:
> [...]
> > 4-node machine stutter
> > 4-node machine stutter
> >                              4.2.0-rc1             4.2.0-rc1
> >                                vanilla           nozlc-v1r20
> > Min         mmap     53.9902 (  0.00%)     49.3629 (  8.57%)
> > 1st-qrtle   mmap     54.6776 (  0.00%)     54.1201 (  1.02%)
> > 2nd-qrtle   mmap     54.9242 (  0.00%)     54.5961 (  0.60%)
> > 3rd-qrtle   mmap     55.1817 (  0.00%)     54.9338 (  0.45%)
> > Max-90%     mmap     55.3952 (  0.00%)     55.3929 (  0.00%)
> > Max-93%     mmap     55.4766 (  0.00%)     57.5712 ( -3.78%)
> > Max-95%     mmap     55.5522 (  0.00%)     57.8376 ( -4.11%)
> > Max-99%     mmap     55.7938 (  0.00%)     63.6180 (-14.02%)
> > Max         mmap   6344.0292 (  0.00%)     67.2477 ( 98.94%)
> > Mean        mmap     57.3732 (  0.00%)     54.5680 (  4.89%)
> 
> Do you have data for other leads? Because the reclaim counters look
> quite discouraging to be honest.
> 

None of the other workloads showed changes that were worth reporting.

> >                                  4.1.0       4.1.0
> >                                vanilla  nozlc-v1r4
> > Swap Ins                           838         502
> > Swap Outs                      1149395     2622895
> 
> Twice as much swapouts is a lot.
> 
> > DMA32 allocs                  17839113    15863747
> > Normal allocs                129045707   137847920
> > Direct pages scanned           4070089    29046893
> 
> 7x more scanns by direct reclaim also sounds bad.
> 

With this benchmark, the results for stutter will be highly variable as
it's hammering the system. The intent of the test was to measure stalls at
a time when desktop interactivity went to hell during IO and could stall
for several minutes. Due to it nature, there is intense reclaim *and*
compaction activity going on and there is no point drawing conclusions
from the reclaim stats that are inherently good or bad.

There will be differences in direct reclaim figures because instead of
looping in the page allocator waiting for zlc to clear, it'll enter direct
reclaim. In effect, the zlc causes processes to busy loop while kswapd
does the work. If it turns out that this is the correct behaviour then
we should do that explicitly, not rely on the broken zlc behaviour for
the same reason we no longer rely on sprinkling congestion_wait() all
over the place.

> > Kswapd pages scanned          17147837    17140694
> 
> while kswapd is doing the same amount of work so we are moving
> considerable amount of reclaim activity into the direct reclaim
> 
> > Kswapd pages reclaimed        17146691    17139601
> > Direct pages reclaimed         1888879     4886630
> > Kswapd efficiency                  99%         99%
> > Kswapd velocity              17523.721   17518.928
> > Direct efficiency                  46%         16%
> 
> which is just a wasted effort because the efficiency is really poor.
> Is this the effect of hammering a single zone which would be skipped
> otherwise while the allocation would succed from another zone?
> 

Very doubtful. It's more likely because the zlc was causing a process to
busy loop waiting for kswapd to make forward progress.

> The latencies were not very much higher to match these numbers though.
> Is it possible that other parts of the benchmark suffered? The benchmark
> has measured only mmap part AFAIU.
> 

mmap latency yes but during it, the system is getting hammered and the
latency is also affected by whether THPs were used or not.

> > Direct velocity               4159.306   29687.854
> > Percentage direct scans            19%         62%
> > Page writes by reclaim     1149395.000 2622895.000
> > Page writes file                     0           0
> > Page writes anon               1149395     2622895
> > 
> > The direct page scan and reclaim rates are noticeable. It is possible
> > this will not be a universal win on all workloads but cycling through
> > zonelists waiting for zlc->last_full_zap to expire is not the right
> > decision.
> 
> As much as I would like to see zlc go it seems that it won't be that
> easy without regressing some loads. Or the numbers

If there are regressions on a real workload then it would be worth
considering why busy looping happened to behave better and then solve it
correctly.

-- 
Mel Gorman
SUSE Labs
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/