v2 changes:
- dropped Patch 6 as adjusting to Joonsoo's objection would be too
complicated and the results didn't justify it
- don't check for compound order > 0 in patches 4 and 5 as suggested by
Michal Nazarewicz. Negative values are handled by converting to unsinged
int, the pfn calculations work fine with 0 and it's unlikely to see 0
due to a race when we just checked PageCompound().
This series is partly the cleanups that were posted as part of the RFC for
changing initial scanning positions [1] and partly new relatively simple
scanner optimizations (yes, they are still possible). I've resumed working
on the bigger scanner changes, but that will take a while, so no point in
delaying these smaller patches.
The interesting patches are 4 and 5. In 4, skipping of compound pages in single
iteration is improved for migration scanner, so it works also for !PageLRU
compound pages such as hugetlbfs, slab etc. Patch 5 introduces this kind of
skipping in the free scanner. The trick is that we can read compound_order()
without any protection, if we are careful to filter out values larger than
MAX_ORDER. The only danger is that we skip too much. The same trick was
already used for reading the freepage order in the migrate scanner.
To demonstrate improvements of Patches 4 and 5 I've run stress-highalloc from
mmtests, set to simulate THP allocations (including __GFP_COMP) on a 4GB system
where 1GB was occupied by hugetlbfs pages. I'll include just the relevant
stats:
Patch 3 Patch 4 Patch 5
Compaction stalls 7523 7529 7515
Compaction success 323 304 322
Compaction failures 7200 7224 7192
Page migrate success 247778 264395 240737
Page migrate failure 15358 33184 21621
Compaction pages isolated 906928 980192 909983
Compaction migrate scanned 2005277 1692805 1498800
Compaction free scanned 13255284 11539986 9011276
Compaction cost 288 305 277
With 5 iterations per patch, the results are still noisy, but we can see that
Patch 4 does reduce migrate_scanned by 15% thanks to skipping the hugetlbfs
pages at once. Interestingly, free_scanned is also reduced and I have no idea
why. Patch 5 further reduces free_scanned as expected, by 15%. Other stats
are unaffected modulo noise.
[1] https://lkml.org/lkml/2015/1/19/158
Vlastimil Babka (5):
mm, compaction: more robust check for scanners meeting
mm, compaction: simplify handling restart position in free pages
scanner
mm, compaction: encapsulate resetting cached scanner positions
mm, compaction: always skip compound pages by order in migrate scanner
mm, compaction: skip compound pages by order in free scanner
mm/compaction.c | 134 ++++++++++++++++++++++++++++++++++++--------------------
1 file changed, 86 insertions(+), 48 deletions(-)
--
2.4.6
--
To unsubscribe, send a message with 'unsubscribe linux-mm' in
the body to majordomo@xxxxxxxxx. For more info on Linux MM,
see: http://www.linux-mm.org/ .
Don't email: <a href=mailto:"dont@xxxxxxxxx";> email@xxxxxxxxx </a>