![]() | |||||||||||||
Messages in this thread |
On Sun, 9 Mar 2003, Robert Love wrote: > On Sun, 2003-03-09 at 18:44, Zwane Mwaikambo wrote: > > > --- linux-2.5.64-unwashed/include/linux/brlock.h 5 Mar 2003 05:07:54 -0000 1.1.1.1 > > +++ linux-2.5.64-unwashed/include/linux/brlock.h 9 Mar 2003 23:42:26 -0000 > > @@ -85,8 +85,7 @@ > > if (idx >= __BR_END) > > __br_lock_usage_bug(); > > > > - preempt_disable(); > > - _raw_read_lock(&__brlock_array[smp_processor_id()][idx]); > > + read_lock(&__brlock_array[smp_processor_id()][idx]); > > } > > This is wrong. I was waiting for this ;) > We have to disable preemption prior to reading smp_processor_id(). > Otherwise we may lock/unlock the wrong processor's brlock. The above as > you changed it is equivalent to: > > cpu = smp_processor_id(); > /* do not want to preempt here, but we can! */ > preempt_disable(); > _raw_read_lock(&__brlock_array[cpu][idx]); How are we able to preempt there? Timer tick? Thanks, Zwane -- function.linuxpower.ca - To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/ | ||||||||||||
| Last update: 2005-03-22 13:33 [from the cache] ©2003-2008 | |||||||||||||