Skip to content

L
linux

xiphi1978 / linux

通知 2
Star 0
Fork 0
L
linux

体验新版 GitCode，发现更多精彩内容 >>

切换分支/标签
  1. 29 7月, 2016 26 次提交
  3. 27 7月, 2016 2 次提交
  5. 21 5月, 2016 3 次提交
    • M
      mm, oom: rework oom detection · 0a0337e0
      由 Michal Hocko 提交于 
      __alloc_pages_slowpath has traditionally relied on the direct reclaim
and did_some_progress as an indicator that it makes sense to retry
allocation rather than declaring OOM.  shrink_zones had to rely on
zone_reclaimable if shrink_zone didn't make any progress to prevent from
a premature OOM killer invocation - the LRU might be full of dirty or
writeback pages and direct reclaim cannot clean those up.

zone_reclaimable allows to rescan the reclaimable lists several times
and restart if a page is freed.  This is really subtle behavior and it
might lead to a livelock when a single freed page keeps allocator
looping but the current task will not be able to allocate that single
page.  OOM killer would be more appropriate than looping without any
progress for unbounded amount of time.

This patch changes OOM detection logic and pulls it out from shrink_zone
which is too low to be appropriate for any high level decisions such as
OOM which is per zonelist property.  It is __alloc_pages_slowpath which
knows how many attempts have been done and what was the progress so far
therefore it is more appropriate to implement this logic.

The new heuristic is implemented in should_reclaim_retry helper called
from __alloc_pages_slowpath.  It tries to be more deterministic and
easier to follow.  It builds on an assumption that retrying makes sense
only if the currently reclaimable memory + free pages would allow the
current allocation request to succeed (as per __zone_watermark_ok) at
least for one zone in the usable zonelist.

This alone wouldn't be sufficient, though, because the writeback might
get stuck and reclaimable pages might be pinned for a really long time
or even depend on the current allocation context.  Therefore there is a
backoff mechanism implemented which reduces the reclaim target after
each reclaim round without any progress.  This means that we should
eventually converge to only NR_FREE_PAGES as the target and fail on the
wmark check and proceed to OOM.  The backoff is simple and linear with
1/16 of the reclaimable pages for each round without any progress.  We
are optimistic and reset counter for successful reclaim rounds.

Costly high order pages mostly preserve their semantic and those without
__GFP_REPEAT fail right away while those which have the flag set will
back off after the amount of reclaimable pages reaches equivalent of the
requested order.  The only difference is that if there was no progress
during the reclaim we rely on zone watermark check.  This is more
logical thing to do than previous 1<<order attempts which were a result
of zone_reclaimable faking the progress.

[vdavydov@virtuozzo.com: check classzone_idx for shrink_zone]
[hannes@cmpxchg.org: separate the heuristic into should_reclaim_retry]
[rientjes@google.com: use zone_page_state_snapshot for NR_FREE_PAGES]
[rientjes@google.com: shrink_zones doesn't need to return anything]
Signed-off-by: NMichal Hocko <mhocko@suse.com>
Acked-by: NHillf Danton <hillf.zj@alibaba-inc.com>
Cc: Vladimir Davydov <vdavydov@virtuozzo.com>
Cc: Johannes Weiner <hannes@cmpxchg.org>
Cc: David Rientjes <rientjes@google.com>
Cc: Joonsoo Kim <js1304@gmail.com>
Cc: Mel Gorman <mgorman@techsingularity.net>
Cc: Tetsuo Handa <penguin-kernel@I-love.SAKURA.ne.jp>
Cc: Vlastimil Babka <vbabka@suse.cz>
Signed-off-by: NAndrew Morton <akpm@linux-foundation.org>
Signed-off-by: NLinus Torvalds <torvalds@linux-foundation.org>
      0a0337e0
    • M
      vmscan: consider classzone_idx in compaction_ready · b6459cc1
      由 Michal Hocko 提交于 
      Motivation:
As pointed out by Linus [2][3] relying on zone_reclaimable as a way to
communicate the reclaim progress is rater dubious. I tend to agree,
not only it is really obscure, it is not hard to imagine cases where a
single page freed in the loop keeps all the reclaimers looping without
getting any progress because their gfp_mask wouldn't allow to get that
page anyway (e.g. single GFP_ATOMIC alloc and free loop). This is rather
rare so it doesn't happen in the practice but the current logic which we
have is rather obscure and hard to follow a also non-deterministic.

This is an attempt to make the OOM detection more deterministic and
easier to follow because each reclaimer basically tracks its own
progress which is implemented at the page allocator layer rather spread
out between the allocator and the reclaim.  The more on the
implementation is described in the first patch.

I have tested several different scenarios but it should be clear that
testing OOM killer is quite hard to be representative.  There is usually
a tiny gap between almost OOM and full blown OOM which is often time
sensitive.  Anyway, I have tested the following 2 scenarios and I would
appreciate if there are more to test.

Testing environment: a virtual machine with 2G of RAM and 2CPUs without
any swap to make the OOM more deterministic.

1) 2 writers (each doing dd with 4M blocks to an xfs partition with 1G
   file size, removes the files and starts over again) running in
   parallel for 10s to build up a lot of dirty pages when 100 parallel
   mem_eaters (anon private populated mmap which waits until it gets
   signal) with 80M each.

   This causes an OOM flood of course and I have compared both patched
   and unpatched kernels. The test is considered finished after there
   are no OOM conditions detected. This should tell us whether there are
   any excessive kills or some of them premature (e.g. due to dirty pages):

I have performed two runs this time each after a fresh boot.

* base kernel
$ grep "Out of memory:" base-oom-run1.log | wc -l
78
$ grep "Out of memory:" base-oom-run2.log | wc -l
78

$ grep "Kill process" base-oom-run1.log | tail -n1
[   91.391203] Out of memory: Kill process 3061 (mem_eater) score 39 or sacrifice child
$ grep "Kill process" base-oom-run2.log | tail -n1
[   82.141919] Out of memory: Kill process 3086 (mem_eater) score 39 or sacrifice child

$ grep "DMA32 free:" base-oom-run1.log | sed 's@.*free:\([0-9]*\)kB.*@\1@' | calc_min_max.awk
min: 5376.00 max: 6776.00 avg: 5530.75 std: 166.50 nr: 61
$ grep "DMA32 free:" base-oom-run2.log | sed 's@.*free:\([0-9]*\)kB.*@\1@' | calc_min_max.awk
min: 5416.00 max: 5608.00 avg: 5514.15 std: 42.94 nr: 52

$ grep "DMA32.*all_unreclaimable? no" base-oom-run1.log | wc -l
1
$ grep "DMA32.*all_unreclaimable? no" base-oom-run2.log | wc -l
3

* patched kernel
$ grep "Out of memory:" patched-oom-run1.log | wc -l
78
miso@tiehlicka /mnt/share/devel/miso/kvm $ grep "Out of memory:" patched-oom-run2.log | wc -l
77

e grep "Kill process" patched-oom-run1.log | tail -n1
[  497.317732] Out of memory: Kill process 3108 (mem_eater) score 39 or sacrifice child
$ grep "Kill process" patched-oom-run2.log | tail -n1
[  316.169920] Out of memory: Kill process 3093 (mem_eater) score 39 or sacrifice child

$ grep "DMA32 free:" patched-oom-run1.log | sed 's@.*free:\([0-9]*\)kB.*@\1@' | calc_min_max.awk
min: 5420.00 max: 5808.00 avg: 5513.90 std: 60.45 nr: 78
$ grep "DMA32 free:" patched-oom-run2.log | sed 's@.*free:\([0-9]*\)kB.*@\1@' | calc_min_max.awk
min: 5380.00 max: 6384.00 avg: 5520.94 std: 136.84 nr: 77

e grep "DMA32.*all_unreclaimable? no" patched-oom-run1.log | wc -l
2
$ grep "DMA32.*all_unreclaimable? no" patched-oom-run2.log | wc -l
3

The patched kernel run noticeably longer while invoking OOM killer same
number of times. This means that the original implementation is much
more aggressive and triggers the OOM killer sooner. free pages stats
show that neither kernels went OOM too early most of the time, though. I
guess the difference is in the backoff when retries without any progress
do sleep for a while if there is memory under writeback or dirty which
is highly likely considering the parallel IO.
Both kernels have seen races where zone wasn't marked unreclaimable
and we still hit the OOM killer. This is most likely a race where
a task managed to exit between the last allocation attempt and the oom
killer invocation.

2) 2 writers again with 10s of run and then 10 mem_eaters to consume as much
   memory as possible without triggering the OOM killer. This required a lot
   of tuning but I've considered 3 consecutive runs in three different boots
   without OOM as a success.

* base kernel
size=$(awk '/MemFree/{printf "%dK", ($2/10)-(16*1024)}' /proc/meminfo)

* patched kernel
size=$(awk '/MemFree/{printf "%dK", ($2/10)-(12*1024)}' /proc/meminfo)

That means 40M more memory was usable without triggering OOM killer. The
base kernel sometimes managed to handle the same as patched but it
wasn't consistent and failed in at least on of the 3 runs. This seems
like a minor improvement.

I was testing also GPF_REPEAT costly requests (hughetlb) with fragmented
memory and under memory pressure. The results are in patch 11 where the
logic is implemented. In short I can see huge improvement there.

I am certainly interested in other usecases as well as well as any
feedback. Especially those which require higher order requests.

This patch (of 14):

While playing with the oom detection rework [1] I have noticed that my
heavy order-9 (hugetlb) load close to OOM ended up in an endless loop
where the reclaim hasn't made any progress but did_some_progress didn't
reflect that and compaction_suitable was backing off because no zone is
above low wmark + 1 << order.

It turned out that this is in fact an old standing bug in
compaction_ready which ignores the requested_highidx and did the
watermark check for 0 classzone_idx.  This succeeds for zone DMA most
of the time as the zone is mostly unused because of lowmem protection.
As a result costly high order allocatios always report a successfull
progress even when there was none.  This wasn't a problem so far
because these allocations usually fail quite early or retry only few
times with __GFP_REPEAT but this will change after later patch in this
series so make sure to not lie about the progress and propagate
requested_highidx down to compaction_ready and use it for both the
watermak check and compaction_suitable to fix this issue.

[1] http://lkml.kernel.org/r/1459855533-4600-1-git-send-email-mhocko@kernel.org
[2] https://lkml.org/lkml/2015/10/12/808
[3] https://lkml.org/lkml/2015/10/13/597Signed-off-by: NMichal Hocko <mhocko@suse.com>
Acked-by: NVlastimil Babka <vbabka@suse.cz>
Acked-by: NHillf Danton <hillf.zj@alibaba-inc.com>
Cc: Johannes Weiner <hannes@cmpxchg.org>
Cc: Mel Gorman <mgorman@suse.de>
Cc: David Rientjes <rientjes@google.com>
Cc: Tetsuo Handa <penguin-kernel@I-love.SAKURA.ne.jp>
Cc: Joonsoo Kim <js1304@gmail.com>
Cc: Vladimir Davydov <vdavydov@virtuozzo.com>
Signed-off-by: NAndrew Morton <akpm@linux-foundation.org>
Signed-off-by: NLinus Torvalds <torvalds@linux-foundation.org>
      b6459cc1
    • R
      mm: vmscan: reduce size of inactive file list · 59dc76b0
      由 Rik van Riel 提交于 
      The inactive file list should still be large enough to contain readahead
windows and freshly written file data, but it no longer is the only
source for detecting multiple accesses to file pages.  The workingset
refault measurement code causes recently evicted file pages that get
accessed again after a shorter interval to be promoted directly to the
active list.

With that mechanism in place, we can afford to (on a larger system)
dedicate more memory to the active file list, so we can actually cache
more of the frequently used file pages in memory, and not have them
pushed out by streaming writes, once-used streaming file reads, etc.

This can help things like database workloads, where only half the page
cache can currently be used to cache the database working set.  This
patch automatically increases that fraction on larger systems, using the
same ratio that has already been used for anonymous memory.

[hannes@cmpxchg.org: cgroup-awareness]
Signed-off-by: NRik van Riel <riel@redhat.com>
Signed-off-by: NJohannes Weiner <hannes@cmpxchg.org>
Reported-by: NAndres Freund <andres@anarazel.de>
Signed-off-by: NAndrew Morton <akpm@linux-foundation.org>
Signed-off-by: NLinus Torvalds <torvalds@linux-foundation.org>
      59dc76b0
  7. 20 5月, 2016 2 次提交
    • H
      mm: update_lru_size do the __mod_zone_page_state · 9d5e6a9f
      由 Hugh Dickins 提交于 
      Konstantin Khlebnikov pointed out (nearly four years ago, when lumpy
reclaim was removed) that lru_size can be updated by -nr_taken once per
call to isolate_lru_pages(), instead of page by page.

Update it inside isolate_lru_pages(), or at its two callsites? I chose
to update it at the callsites, rearranging and grouping the updates by
nr_taken and nr_scanned together in both.

With one exception, mem_cgroup_update_lru_size(,lru,) is then used where
__mod_zone_page_state(,NR_LRU_BASE+lru,) is used; and we shall be adding
some more calls in a future commit.  Make the code a little smaller and
simpler by incorporating stat update in lru_size update.

The exception was move_active_pages_to_lru(), which aggregated the
pgmoved stat update separately from the individual lru_size updates; but
I still think this a simplification worth making.

However, the __mod_zone_page_state is not peculiar to mem_cgroups: so
better use the name update_lru_size, calls mem_cgroup_update_lru_size
when CONFIG_MEMCG.
Signed-off-by: NHugh Dickins <hughd@google.com>
Cc: "Kirill A. Shutemov" <kirill.shutemov@linux.intel.com>
Cc: Andrea Arcangeli <aarcange@redhat.com>
Cc: Andres Lagar-Cavilla <andreslc@google.com>
Cc: Yang Shi <yang.shi@linaro.org>
Cc: Ning Qu <quning@gmail.com>
Cc: Mel Gorman <mgorman@techsingularity.net>
Cc: Konstantin Khlebnikov <koct9i@gmail.com>
Signed-off-by: NAndrew Morton <akpm@linux-foundation.org>
Signed-off-by: NLinus Torvalds <torvalds@linux-foundation.org>
      9d5e6a9f
    • J
      mm: rename _count, field of the struct page, to _refcount · 0139aa7b
      由 Joonsoo Kim 提交于 
      Many developers already know that field for reference count of the
struct page is _count and atomic type.  They would try to handle it
directly and this could break the purpose of page reference count
tracepoint.  To prevent direct _count modification, this patch rename it
to _refcount and add warning message on the code.  After that, developer
who need to handle reference count will find that field should not be
accessed directly.

[akpm@linux-foundation.org: fix comments, per Vlastimil]
[akpm@linux-foundation.org: Documentation/vm/transhuge.txt too]
[sfr@canb.auug.org.au: sync ethernet driver changes]
Signed-off-by: NJoonsoo Kim <iamjoonsoo.kim@lge.com>
Signed-off-by: NStephen Rothwell <sfr@canb.auug.org.au>
Cc: Vlastimil Babka <vbabka@suse.cz>
Cc: Hugh Dickins <hughd@google.com>
Cc: Johannes Berg <johannes@sipsolutions.net>
Cc: "David S. Miller" <davem@davemloft.net>
Cc: Sunil Goutham <sgoutham@cavium.com>
Cc: Chris Metcalf <cmetcalf@mellanox.com>
Cc: Manish Chopra <manish.chopra@qlogic.com>
Cc: Yuval Mintz <yuval.mintz@qlogic.com>
Cc: Tariq Toukan <tariqt@mellanox.com>
Cc: Saeed Mahameed <saeedm@mellanox.com>
Signed-off-by: NAndrew Morton <akpm@linux-foundation.org>
Signed-off-by: NLinus Torvalds <torvalds@linux-foundation.org>
      0139aa7b
  9. 29 4月, 2016 2 次提交
    • V
      mm: wake kcompactd before kswapd's short sleep · fd901c95
      由 Vlastimil Babka 提交于 
      When kswapd goes to sleep it checks if the node is balanced and at first
it sleeps only for HZ/10 time, then rechecks if the node is still
balanced and nobody has woken it during the initial sleep.  Only then it
goes fully sleep until an allocation slowpath wakes it up again.

For higher-order allocations, waking up kcompactd is done only before
the full sleep.  This turns out to be an issue in case another
high-order allocation fails during the initial sleep.  It will wake
kswapd up, however kswapd considers the zone balanced from the order-0
perspective, and will just quickly try to sleep again.  So if there's a
longer stream of high-order allocations hitting the slowpath and waking
up kswapd, it might never actually wake up kcompactd, which may be
considered a regression from kswapd-based compaction.  In the worst
case, it might be that a single allocation that cannot direct
reclaim/compact itself is waking kswapd in the retry loop and preventing
kcompactd from being woken up and unblocking it.

This patch makes sure kcompactd is woken up in such situations by simply
moving the wakeup before the short initial sleep.  More efficient
solution would be to wake kcompactd immediately instead of kswapd if the
node is already order-0 balanced, but in that case we should also move
reset_isolation_suitable() call to kcompactd so it's not adding to the
allocator's latency.  Since it's late in the 4.6 cycle, let's go with
the simpler change for now.

Fixes: accf6242 ("mm, kswapd: replace kswapd compaction with waking up kcompactd")
Signed-off-by: NVlastimil Babka <vbabka@suse.cz>
Cc: Andrea Arcangeli <aarcange@redhat.com>
Cc: "Kirill A. Shutemov" <kirill.shutemov@linux.intel.com>
Cc: Rik van Riel <riel@redhat.com>
Cc: Joonsoo Kim <iamjoonsoo.kim@lge.com>
Cc: Mel Gorman <mgorman@techsingularity.net>
Cc: David Rientjes <rientjes@google.com>
Cc: Michal Hocko <mhocko@suse.com>
Cc: Johannes Weiner <hannes@cmpxchg.org>
Signed-off-by: NAndrew Morton <akpm@linux-foundation.org>
Signed-off-by: NLinus Torvalds <torvalds@linux-foundation.org>
      fd901c95
    • M
      mm: vmscan: reclaim highmem zone if buffer_heads is over limit · 7bf52fb8
      由 Minchan Kim 提交于 
      We have been reclaimed highmem zone if buffer_heads is over limit but
commit 6b4f7799 ("mm: vmscan: invoke slab shrinkers from
shrink_zone()") changed the behavior so it doesn't reclaim highmem zone
although buffer_heads is over the limit.  This patch restores the logic.

Fixes: 6b4f7799 ("mm: vmscan: invoke slab shrinkers from shrink_zone()")
Signed-off-by: NMinchan Kim <minchan@kernel.org>
Cc: Johannes Weiner <hannes@cmpxchg.org>
Cc: <stable@vger.kernel.org>
Signed-off-by: NAndrew Morton <akpm@linux-foundation.org>
Signed-off-by: NLinus Torvalds <torvalds@linux-foundation.org>
      7bf52fb8
  11. 18 3月, 2016 4 次提交
    • J
      mm: introduce page reference manipulation functions · fe896d18
      由 Joonsoo Kim 提交于 
      The success of CMA allocation largely depends on the success of
migration and key factor of it is page reference count.  Until now, page
reference is manipulated by direct calling atomic functions so we cannot
follow up who and where manipulate it.  Then, it is hard to find actual
reason of CMA allocation failure.  CMA allocation should be guaranteed
to succeed so finding offending place is really important.

In this patch, call sites where page reference is manipulated are
converted to introduced wrapper function.  This is preparation step to
add tracepoint to each page reference manipulation function.  With this
facility, we can easily find reason of CMA allocation failure.  There is
no functional change in this patch.

In addition, this patch also converts reference read sites.  It will
help a second step that renames page._count to something else and
prevents later attempt to direct access to it (Suggested by Andrew).
Signed-off-by: NJoonsoo Kim <iamjoonsoo.kim@lge.com>
Acked-by: NMichal Nazarewicz <mina86@mina86.com>
Acked-by: NVlastimil Babka <vbabka@suse.cz>
Cc: Minchan Kim <minchan@kernel.org>
Cc: Mel Gorman <mgorman@techsingularity.net>
Cc: "Kirill A. Shutemov" <kirill.shutemov@linux.intel.com>
Cc: Sergey Senozhatsky <sergey.senozhatsky.work@gmail.com>
Cc: Steven Rostedt <rostedt@goodmis.org>
Signed-off-by: NAndrew Morton <akpm@linux-foundation.org>
Signed-off-by: NLinus Torvalds <torvalds@linux-foundation.org>
      fe896d18
    • V
      mm: vmscan: pass root_mem_cgroup instead of NULL to memcg aware shrinker · 0fc9f58a
      由 Vladimir Davydov 提交于 
      It's just convenient to implement a memcg aware shrinker when you know
that shrink_control->memcg != NULL unless memcg_kmem_enabled() returns
false.
Signed-off-by: NVladimir Davydov <vdavydov@virtuozzo.com>
Acked-by: NJohannes Weiner <hannes@cmpxchg.org>
Cc: Michal Hocko <mhocko@kernel.org>
Signed-off-by: NAndrew Morton <akpm@linux-foundation.org>
Signed-off-by: NLinus Torvalds <torvalds@linux-foundation.org>
      0fc9f58a
    • V
      mm, kswapd: replace kswapd compaction with waking up kcompactd · accf6242
      由 Vlastimil Babka 提交于 
      Similarly to direct reclaim/compaction, kswapd attempts to combine
reclaim and compaction to attempt making memory allocation of given
order available.

The details differ from direct reclaim e.g. in having high watermark as
a goal.  The code involved in kswapd's reclaim/compaction decisions has
evolved to be quite complex.

Testing reveals that it doesn't actually work in at least one scenario,
and closer inspection suggests that it could be greatly simplified
without compromising on the goal (make high-order page available) or
efficiency (don't reclaim too much).  The simplification relieas of
doing all compaction in kcompactd, which is simply woken up when high
watermarks are reached by kswapd's reclaim.

The scenario where kswapd compaction doesn't work was found with mmtests
test stress-highalloc configured to attempt order-9 allocations without
direct reclaim, just waking up kswapd.  There was no compaction attempt
from kswapd during the whole test.  Some added instrumentation shows
what happens:

 - balance_pgdat() sets end_zone to Normal, as it's not balanced
 - reclaim is attempted on DMA zone, which sets nr_attempted to 99, but
   it cannot reclaim anything, so sc.nr_reclaimed is 0
 - for zones DMA32 and Normal, kswapd_shrink_zone uses testorder=0, so
   it merely checks if high watermarks were reached for base pages.
   This is true, so no reclaim is attempted.  For DMA, testorder=0
   wasn't used, as compaction_suitable() returned COMPACT_SKIPPED
 - even though the pgdat_needs_compaction flag wasn't set to false, no
   compaction happens due to the condition sc.nr_reclaimed >
   nr_attempted being false (as 0 < 99)
 - priority-- due to nr_reclaimed being 0, repeat until priority reaches
   0 pgdat_balanced() is false as only the small zone DMA appears
   balanced (curiously in that check, watermark appears OK and
   compaction_suitable() returns COMPACT_PARTIAL, because a lower
   classzone_idx is used there)

Now, even if it was decided that reclaim shouldn't be attempted on the
DMA zone, the scenario would be the same, as (sc.nr_reclaimed=0 >
nr_attempted=0) is also false.  The condition really should use >= as
the comment suggests.  Then there is a mismatch in the check for setting
pgdat_needs_compaction to false using low watermark, while the rest uses
high watermark, and who knows what other subtlety.  Hopefully this
demonstrates that this is unsustainable.

Luckily we can simplify this a lot.  The reclaim/compaction decisions
make sense for direct reclaim scenario, but in kswapd, our primary goal
is to reach high watermark in order-0 pages.  Afterwards we can attempt
compaction just once.  Unlike direct reclaim, we don't reclaim extra
pages (over the high watermark), the current code already disallows it
for good reasons.

After this patch, we simply wake up kcompactd to process the pgdat,
after we have either succeeded or failed to reach the high watermarks in
kswapd, which goes to sleep.  We pass kswapd's order and classzone_idx,
so kcompactd can apply the same criteria to determine which zones are
worth compacting.  Note that we use the classzone_idx from
wakeup_kswapd(), not balanced_classzone_idx which can include higher
zones that kswapd tried to balance too, but didn't consider them in
pgdat_balanced().

Since kswapd now cannot create high-order pages itself, we need to
adjust how it determines the zones to be balanced.  The key element here
is adding a "highorder" parameter to zone_balanced, which, when set to
false, makes it consider only order-0 watermark instead of the desired
higher order (this was done previously by kswapd_shrink_zone(), but not
elsewhere).  This false is passed for example in pgdat_balanced().
Importantly, wakeup_kswapd() uses true to make sure kswapd and thus
kcompactd are woken up for a high-order allocation failure.

The last thing is to decide what to do with pageblock_skip bitmap
handling.  Compaction maintains a pageblock_skip bitmap to record
pageblocks where isolation recently failed.  This bitmap can be reset by
three ways:

1) direct compaction is restarting after going through the full deferred cycle

2) kswapd goes to sleep, and some other direct compaction has previously
   finished scanning the whole zone and set zone->compact_blockskip_flush.
   Note that a successful direct compaction clears this flag.

3) compaction was invoked manually via trigger in /proc

The case 2) is somewhat fuzzy to begin with, but after introducing
kcompactd we should update it.  The check for direct compaction in 1),
and to set the flush flag in 2) use current_is_kswapd(), which doesn't
work for kcompactd.  Thus, this patch adds bool direct_compaction to
compact_control to use in 2).  For the case 1) we remove the check
completely - unlike the former kswapd compaction, kcompactd does use the
deferred compaction functionality, so flushing tied to restarting from
deferred compaction makes sense here.

Note that when kswapd goes to sleep, kcompactd is woken up, so it will
see the flushed pageblock_skip bits.  This is different from when the
former kswapd compaction observed the bits and I believe it makes more
sense.  Kcompactd can afford to be more thorough than a direct
compaction trying to limit allocation latency, or kswapd whose primary
goal is to reclaim.

For testing, I used stress-highalloc configured to do order-9
allocations with GFP_NOWAIT|__GFP_HIGH|__GFP_COMP, so they relied just
on kswapd/kcompactd reclaim/compaction (the interfering kernel builds in
phases 1 and 2 work as usual):

stress-highalloc
                        4.5-rc1+before          4.5-rc1+after
                             -nodirect              -nodirect
Success 1 Min          1.00 (  0.00%)         5.00 (-66.67%)
Success 1 Mean         1.40 (  0.00%)         6.20 (-55.00%)
Success 1 Max          2.00 (  0.00%)         7.00 (-16.67%)
Success 2 Min          1.00 (  0.00%)         5.00 (-66.67%)
Success 2 Mean         1.80 (  0.00%)         6.40 (-52.38%)
Success 2 Max          3.00 (  0.00%)         7.00 (-16.67%)
Success 3 Min         34.00 (  0.00%)        62.00 (  1.59%)
Success 3 Mean        41.80 (  0.00%)        63.80 (  1.24%)
Success 3 Max         53.00 (  0.00%)        65.00 (  2.99%)

User                          3166.67        3181.09
System                        1153.37        1158.25
Elapsed                       1768.53        1799.37

                            4.5-rc1+before   4.5-rc1+after
                                 -nodirect    -nodirect
Direct pages scanned                32938        32797
Kswapd pages scanned              2183166      2202613
Kswapd pages reclaimed            2152359      2143524
Direct pages reclaimed              32735        32545
Percentage direct scans                1%           1%
THP fault alloc                       579          612
THP collapse alloc                    304          316
THP splits                              0            0
THP fault fallback                    793          778
THP collapse fail                      11           16
Compaction stalls                    1013         1007
Compaction success                     92           67
Compaction failures                   920          939
Page migrate success               238457       721374
Page migrate failure                23021        23469
Compaction pages isolated          504695      1479924
Compaction migrate scanned         661390      8812554
Compaction free scanned          13476658     84327916
Compaction cost                       262          838

After this patch we see improvements in allocation success rate
(especially for phase 3) along with increased compaction activity.  The
compaction stalls (direct compaction) in the interfering kernel builds
(probably THP's) also decreased somewhat thanks to kcompactd activity,
yet THP alloc successes improved a bit.

Note that elapsed and user time isn't so useful for this benchmark,
because of the background interference being unpredictable.  It's just
to quickly spot some major unexpected differences.  System time is
somewhat more useful and that didn't increase.

Also (after adjusting mmtests' ftrace monitor):

Time kswapd awake               2547781     2269241
Time kcompactd awake                  0      119253
Time direct compacting           939937      557649
Time kswapd compacting                0           0
Time kcompactd compacting             0      119099

The decrease of overal time spent compacting appears to not match the
increased compaction stats.  I suspect the tasks get rescheduled and
since the ftrace monitor doesn't see that, the reported time is wall
time, not CPU time.  But arguably direct compactors care about overall
latency anyway, whether busy compacting or waiting for CPU doesn't
matter.  And that latency seems to almost halved.

It's also interesting how much time kswapd spent awake just going
through all the priorities and failing to even try compacting, over and
over.

We can also configure stress-highalloc to perform both direct
reclaim/compaction and wakeup kswapd/kcompactd, by using
GFP_KERNEL|__GFP_HIGH|__GFP_COMP:

stress-highalloc
                        4.5-rc1+before         4.5-rc1+after
                               -direct               -direct
Success 1 Min          4.00 (  0.00%)        9.00 (-50.00%)
Success 1 Mean         8.00 (  0.00%)       10.00 (-19.05%)
Success 1 Max         12.00 (  0.00%)       11.00 ( 15.38%)
Success 2 Min          4.00 (  0.00%)        9.00 (-50.00%)
Success 2 Mean         8.20 (  0.00%)       10.00 (-16.28%)
Success 2 Max         13.00 (  0.00%)       11.00 (  8.33%)
Success 3 Min         75.00 (  0.00%)       74.00 (  1.33%)
Success 3 Mean        75.60 (  0.00%)       75.20 (  0.53%)
Success 3 Max         77.00 (  0.00%)       76.00 (  0.00%)

User                          3344.73       3246.04
System                        1194.24       1172.29
Elapsed                       1838.04       1836.76

                            4.5-rc1+before  4.5-rc1+after
                                   -direct     -direct
Direct pages scanned               125146      120966
Kswapd pages scanned              2119757     2135012
Kswapd pages reclaimed            2073183     2108388
Direct pages reclaimed             124909      120577
Percentage direct scans                5%          5%
THP fault alloc                       599         652
THP collapse alloc                    323         354
THP splits                              0           0
THP fault fallback                    806         793
THP collapse fail                      17          16
Compaction stalls                    2457        2025
Compaction success                    906         518
Compaction failures                  1551        1507
Page migrate success              2031423     2360608
Page migrate failure                32845       40852
Compaction pages isolated         4129761     4802025
Compaction migrate scanned       11996712    21750613
Compaction free scanned         214970969   344372001
Compaction cost                      2271        2694

In this scenario, this patch doesn't change the overall success rate as
direct compaction already tries all it can.  There's however significant
reduction in direct compaction stalls (that is, the number of
allocations that went into direct compaction).  The number of successes
(i.e.  direct compaction stalls that ended up with successful
allocation) is reduced by the same number.  This means the offload to
kcompactd is working as expected, and direct compaction is reduced
either due to detecting contention, or compaction deferred by kcompactd.
In the previous version of this patchset there was some apparent
reduction of success rate, but the changes in this version (such as
using sync compaction only), new baseline kernel, and/or averaging
results from 5 executions (my bet), made this go away.

Ftrace-based stats seem to roughly agree:

Time kswapd awake               2532984     2326824
Time kcompactd awake                  0      257916
Time direct compacting           864839      735130
Time kswapd compacting                0           0
Time kcompactd compacting             0      257585
Signed-off-by: NVlastimil Babka <vbabka@suse.cz>
Cc: Andrea Arcangeli <aarcange@redhat.com>
Cc: "Kirill A. Shutemov" <kirill.shutemov@linux.intel.com>
Cc: Rik van Riel <riel@redhat.com>
Cc: Joonsoo Kim <iamjoonsoo.kim@lge.com>
Cc: Mel Gorman <mgorman@techsingularity.net>
Cc: David Rientjes <rientjes@google.com>
Cc: Michal Hocko <mhocko@suse.com>
Cc: Johannes Weiner <hannes@cmpxchg.org>
Signed-off-by: NAndrew Morton <akpm@linux-foundation.org>
Signed-off-by: NLinus Torvalds <torvalds@linux-foundation.org>
      accf6242
    • V
      mm, kswapd: remove bogus check of balance_classzone_idx · 81c5857b
      由 Vlastimil Babka 提交于 
      During work on kcompactd integration I have spotted a confusing check of
balance_classzone_idx, which I believe is bogus.

The balanced_classzone_idx is filled by balance_pgdat() as the highest
zone it attempted to balance.  This was introduced by commit dc83edd9
("mm: kswapd: use the classzone idx that kswapd was using for
sleeping_prematurely()").

The intention is that (as expressed in today's function names), the
value used for kswapd_shrink_zone() calls in balance_pgdat() is the same
as for the decisions in kswapd_try_to_sleep().

An unwanted side-effect of that commit was breaking the checks in
kswapd() whether there was another kswapd_wakeup with a tighter (=lower)
classzone_idx.  Commits 215ddd66 ("mm: vmscan: only read
new_classzone_idx from pgdat when reclaiming successfully") and
d2ebd0f6 ("kswapd: avoid unnecessary rebalance after an unsuccessful
balancing") tried to fixed, but apparently introduced a bogus check that
this patch removes.

Consider zone indexes X < Y < Z, where:
- Z is the value used for the first kswapd wakeup.
- Y is returned as balanced_classzone_idx, which means zones with index higher
  than Y (including Z) were found to be unreclaimable.
- X is the value used for the second kswapd wakeup

The new wakeup with value X means that kswapd is now supposed to balance
harder all zones with index <= X.  But instead, due to Y < Z, it will go
sleep and won't read the new value X.  This is subtly wrong.

The effect of this patch is that kswapd will react better in some
situations, where e.g.  the first wakeup is for ZONE_DMA32, the second is
for ZONE_DMA, and due to unreclaimable ZONE_NORMAL.  Before this patch,
kswapd would go sleep instead of reclaiming ZONE_DMA harder.  I expect
these situations are very rare, and more value is in better
maintainability due to the removal of confusing and bogus check.
Signed-off-by: NVlastimil Babka <vbabka@suse.cz>
Cc: Andrea Arcangeli <aarcange@redhat.com>
Cc: "Kirill A. Shutemov" <kirill.shutemov@linux.intel.com>
Cc: Rik van Riel <riel@redhat.com>
Cc: Joonsoo Kim <iamjoonsoo.kim@lge.com>
Cc: Mel Gorman <mgorman@techsingularity.net>
Cc: David Rientjes <rientjes@google.com>
Cc: Michal Hocko <mhocko@suse.com>
Cc: Johannes Weiner <hannes@cmpxchg.org>
Signed-off-by: NAndrew Morton <akpm@linux-foundation.org>
Signed-off-by: NLinus Torvalds <torvalds@linux-foundation.org>
      81c5857b
  13. 16 3月, 2016 1 次提交