Skip to content

Upgrade of this_cpu_ops V2

Christoph Lameter lkml 1 version In review

Get series with b4
b4 am [email protected]

Versions newest first

  1. v1 Latest10 patches 25 repliesIn review

Patches in v1 No applies recorded yet

#Subject and filesSizeRepliesApplied
1/10percpucounter: Optimize __percpu_counter_add a bit through the use of this_cpu() options.lib/percpu_counter.c+3 -51 reply
2/10vmstat: Optimize zone counter modifications through the use of this cpu operationsmm/vmstat.c+32 -247 replies
3/10percpu: Generic support for this_cpu_add,sub,dec,inc_returninclude/linux/percpu.h+70 -02 replies
4/10x86: Support for this_cpu_add,sub,dec,inc_returnarch/x86/include/asm/percpu.h+50 -08 replies
5/10x86: Use this_cpu_inc_return for nmi counterarch/x86/kernel/apic/nmi.c+1 -22 replies
6/10vmstat: Use this_cpu_inc_return for vm statisticsmm/vmstat.c+8 -122 replies
7/10highmem: Use this_cpu_xx_return() operationsinclude/linux/highmem.h+4 -30 replies
8/10percpu: generic this_cpu_cmpxchg() and this_cpu_cmpxchg_double supportinclude/linux/percpu.h+178 -10 replies
9/10x86: this_cpu_cmpxchg and this_cpu_cmpxchg_double operationsarch/x86/include/asm/percpu.h arch/x86/lib/Makefile arch/x86/lib/cmpxchg16b_emu.S+177 -13 replies
10/10slub: Lockless fastpathsinclude/linux/slub_def.h mm/slub.c+149 -250 replies

Reviews trailers per patch

5 of 10 patches have a Reviewed-by or Acked-by.

PatchPekka EnbergTejun HeoMathieu Desnoyers
1/10R-b R-b R-b
2/10R-b
3/10R-b R-b
4/10
5/10R-b R-b R-b
6/10R-b
7/10
8/10
9/10
10/10

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 [thisops uV2 00/10] Upgrade of this_cpu_ops V2

A patchset that adds more this_cpu operations and in particular RMV operations that can be used in various places to avoid address calculations and memory accesses.

Also adds this_cpu_cmpxchg_double() which is a locally atomic version of cmpxchg and uses that to demo how a lockless, preemptless fastpath for memory allocation could work. Works good enough so that I can write this email with that fastpath scheme.

V2 has several enhancements and bugfixes that were suggested.