On Tue, Feb 05, 2002 at 05:54:32PM +0100, Christoph Hellwig wrote:
> In article <20020205211826.B32506@in.ibm.com> you wrote:
> > 3. A per-cpu timer support ? - This will allow us to get rid of the krcud
> > stuff and make RCU even simpler.
>
> Something like http://people.redhat.com/mingo/scalable-timers-patches/smptimers-2.4.16-A0?
Almost. IIUC, there is still the possibility that a timer queued
in one CPU may get executed in another. While this by itself
doesn't cause a problem for this RCU implementation (we end up
checking some other CPU's queue), if one CPU is starved of timers,
it could be problematic.
Since we have many per-cpu data structures, it would be a useful thing
to have a per-cpu mechanism to manipulate these data structures
in a cache-sensitive way.
>
> Ingo, Linus: Any chance to see that in 2.5 soon?
IIRC, timerlist_lock used to show up in lockmetering in some workloads
we were using some time ago. It would be a good thing to get rid
of this global lock soon.
Thanks
Dipankar
-- Dipankar Sarma <dipankar@in.ibm.com> http://lse.sourceforge.net Linux Technology Center, IBM Software Lab, Bangalore, India. - 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/
This archive was generated by hypermail 2b29 : Thu Feb 07 2002 - 21:00:43 EST