Skip to content

Refactor spinlock inlining

Thomas Gleixner linux-arch lkml 1 version In review

Get series with b4
b4 am [email protected]

Versions newest first

  1. v1 Latest2 patches 5 repliesIn review

Patches in v1 No applies recorded yet

#Subject and filesSizeRepliesApplied
1/2locking: Make inlining decision Kconfig basedarch/s390/Kconfig arch/s390/include/asm/spinlock.h include/linux/spinlock_api_smp.h +3 more+284 -1044 replies
2/2locking: Reduce ifdefs in kernel/spinlock.ckernel/spinlock.c+128 -1321 reply

Reviews trailers per patch

All 2 patches have a Reviewed-by or Acked-by.

PatchPeter ZijlstraHeiko CarstensIngo Molnar
1/2A-b A-b A-b A-b R-b
2/2A-b A-b R-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 0/2] Refactor spinlock inlining

While moving preempt-rt to 2.6.32-rc I tripped over the spinlock inlining changes and had a hard time to figure out when and which lock function is inlined as the __always_inline__ defines are not directly controlling the inlining because there are overrides depending on various config options.

Merging the -rt modifications into this results in a horrible mess and I tried to find a solution which simplifies the inlining decisions and makes it more obvious when the decision is taken.

To allow easier intergration of preempt-rt I reworked the inline decision ifdeffery to be completely Kconfig based.

The result is slightly smaller in code lines, but adds a Kconfig file which contains all the rules of inlining decisions. Architecturs can select each locking function for inlining in the arch Kconfig instead of defining __always_inline__*.

Thanks,

tglx