Skip to content

Numa: Use Generic Per-cpu Variables for numa_*_id()

Lee Schermerhorn linux-arch linux-mm 1 version In review

Get series with b4
b4 am [email protected]

Versions newest first

  1. v1 Latest6 patches 10 repliesIn review

Patches in v1 No applies recorded yet

#Subject and filesSizeRepliesApplied
1/6numa: Use Generic Per-cpu Variables for numa_node_id()include/asm-generic/percpu.h include/asm-generic/topology.h include/linux/percpu.h +2 more+493 -4605 replies
2/6numa: x86_64: use generic percpu var numa_node_id() implementationarch/x86/Kconfig arch/x86/include/asm/percpu.h arch/x86/include/asm/topology.h +3 more+13 -211 reply
3/6numa: ia64: use generic percpu var numa_node_id() implementationarch/ia64/Kconfig arch/ia64/include/asm/topology.h arch/ia64/kernel/smpboot.c+10 -51 reply
4/6numa: Introduce numa_mem_id()- effective local memory node idinclude/linux/mmzone.h include/linux/topology.h mm/page_alloc.c +1 more+70 -01 reply
5/6numa: ia64: support numa_mem_id() for memoryless nodesarch/ia64/Kconfig arch/ia64/kernel/smpboot.c+5 -00 replies
6/6numa: slab: use numa_mem_id() for slab local memory nodemm/slab.c+14 -131 reply

Reviews trailers per patch

2 of 6 patches have a Reviewed-by or Acked-by.

PatchChristoph Lameter
1/6
2/6
3/6R-b
4/6R-b
5/6
6/6

Trailers given in replies are filled; those carried in the patches from earlier versions are outlined with dashes.

Read the v1 thread

Cover letter [PATCH/RFC 0/6] Numa: Use Generic Per-cpu Variables for numa_*_id()

PATCH/RFC - 00/NN - numa: Use generic per-cpu variables for numa_*_id()

In http://marc.info/?l=linux-mm&m=125683610312546&w=4 , I described a performance problem with slab and memoryless nodes that we see on some of our platforms. I proposed modifying slab to use the "effective local memory node"--the node that local mempolicy would select--as the "local" node id for slab allocation purposes. This will allow slab to cache objects from its "local memory node" on the percpu queues, effectively eliminating the problem.

Christoph Lameter suggested a more general approach using the generic percpu support: define a new interface--e.g., numa_mem_id()--that returns the "effective local memory node" for the calling context [cpu]. For nodes with memory, this will == the id of the node itself. For memoryless nodes, this will be the first node in the generic [!this_node] zonelist.

Christoph also suggested converting the current "numa_node_id()" interface to use the generic percpu infrastructure. x86[_64] supports a custom [arch- specific] per cpu variable implementation of numa_node_id(). Most other archs do a table lookup.

This series introduces a generic percpu implementation of numa_node_id() and numa_mem_id() in separate patches based on an incomplete "starter patch" from Christoph. Both of these implementations are conditional on new respective config options. I know that new config options aren't popular, but this allows other archs to adapt to the new implementations incrementally.

Additional patches provide x86_64 and ia64 arch specific changes to use the new numa_node_id() implementation, and ia64 support for the numa_mem_id() interface. Finally, I've reimplemented the "slab memoryless node 'regression' fix" patch linked above atop the new numa_mem_id() interface.

Ad hoc measurements on x86_64 using: hackbench 400 process 200

2.6.32-rc5+mmotm-091101 no patch this series
x86_64 avg of 40: 4.605 4.628 ~0.5%

Ia64 showed ~1.2% longer time with the series applied.