dvhltc@linux.vnet.ibm.com wrote on 09/02/2010 01:04:28 AM:

Subject

Re: [PATCH][RFC] preempt_count corruption across H_CEDE call with
CONFIG_PREEMPT on pseries

With this in place, we no longer see the preempt_count dropping below
zero. However, if I offline/online a CPU about 246 times I hit the
opposite problem, a preempt_count() overflow. There appears to be a
missing preempt_enable() somewhere in the offline/online paths.

This (preempt_count overflow) also occurred in mainline (with
CONFIG_PREEMPT=y) in 2.6.35, but not in 2.6.36-rc3.   A
bisect seems to indicate it was fixed with
a7c2bb8279d20d853e43c34584eaf2b039de8026   "powerpc: Re-enable preemption
before cpu_die()".

Which may look familiar.  :-)

It looks like this patch went to mainline (likely via the powerpc tree),
but may
have not gotten back into the -rt branch.

-Will

--
Darren Hart
IBM Linux Technology Center
Real-Time Linux Team