Skip to content

Use per-cpu allocator for !irq requests and prepare for a bulk allocator v5

Mel Gorman linux-mm lkml 1 version Applied

Status from this message: [merged] mm-page_alloc-drain-per-cpu-pages-from-workqueue-context.patch removed from -mm tree

Get series with b4
b4 am [email protected]

Versions newest first

  1. v1 Latest4 patches 3 repliesApplied 4/4

Changelog since v4 from the cover letter

o Protect drain with get_online_cpus
o Micro-optimisation of stat updates
o Avoid double preparing a page free

Patches in v1 All 4 applied

#Subject and filesSizeRepliesApplied
1/4mm, page_alloc: Split buffered_rmqueuemm/page_alloc.c+78 -482 repliesApplied
2/4mm, page_alloc: Split alloc_pages_nodemaskmm/page_alloc.c+46 -291 replyApplied
3/4mm, page_alloc: Drain per-cpu pages from workqueue contextmm/page_alloc.c+37 -70 repliesApplied
4/4mm, page_alloc: Only use per-cpu allocator for irq-safe requestsmm/page_alloc.c+23 -201 replyApplied

Reviews trailers per patch

3 of 4 patches have a Reviewed-by or Acked-by.

PatchHillf DantonJesper Dangaard BrouerVlastimil Babka
1/4A-b A-b
2/4A-b A-b
3/4
4/4A-b A-b A-b

Trailers given in replies are filled; those carried in the patches from earlier versions are outlined with dashes. Scroll the table sideways for all 3 reviewers.

Read the v1 thread

Cover letter [PATCH 0/4] Use per-cpu allocator for !irq requests and prepare for a bulk allocator v5

This is rebased on top of mmotm to handle collisions with Vlastimil's series on cpusets and premature OOMs.

Changelog since v4
o Protect drain with get_online_cpus
o Micro-optimisation of stat updates
o Avoid double preparing a page free

Changelog since v3
o Debugging check in allocation path
o Make it harder to use the free path incorrectly
o Use preempt-safe stats counter
o Do not use IPIs to drain the per-cpu allocator

Changelog since v2
o Add ack's and benchmark data
o Rebase to 4.10-rc3

Changelog since v1
o Remove a scheduler point from the allocation path
o Finalise the bulk allocator and test it

This series is motivated by a conversation led by Jesper Dangaard Brouer at the last LSF/MM proposing a generic page pool for DMA-coherent pages. Part of his motivation was due to the overhead of allocating multiple order-0 that led some drivers to use high-order allocations and splitting them. This is very slow in some cases.

The first two patches in this series restructure the page allocator such that it is relatively easy to introduce an order-0 bulk page allocator. A patch exists to do that and has been handed over to Jesper until an in-kernel users is created. The third patch prevents the per-cpu allocator being drained from IPI context as that can potentially corrupt the list after patch four is merged. The final patch alters the per-cpu alloctor to make it exclusive to !irq requests. This cuts allocation/free overhead by roughly 30%.

Performance tests from both Jesper and I are included in the patch.

 mm/page_alloc.c | 282 ++++++++++++++++++++++++++++++++++++--------------------
 1 file changed, 181 insertions(+), 101 deletions(-)
Signature · 7 lines
2.11.0

--
To unsubscribe, send a message with 'unsubscribe linux-mm' in
the body to [email protected].  For more info on Linux MM,
see: http://www.linux-mm.org/ .
Don't email: <a href=mailto:"[email protected]"> [email protected] </a>