Messages in this thread | | | Subject | RE: [patch] Reorganize the cpufreq cpu hotplug locking to not be totally bizare | Date | Thu, 27 Jul 2006 14:29:01 -0500 | From | "Langsdorf, Mark" <> |
| |
> before introducing the ugly recursion we did try the > preempt_disable() for cpufreq, and it worked for most all > governers with preempt_disable(), but powernowk8 called > set_cpus_allowed() in the callback path that threw out a > scheduling while in atomic BUG(). > > http://lkml.org/lkml/2005/10/31/239
Is there some other preferred call we could be making instead of set_cpus_allowed() ? We need to be able to program the MSRs on that specific core, and as far as I know, the only way to do that is to guarantee that we are scheduled on that particular core.
If there's a better way to hop to a specific core, I'll gladly rewrite the code in question.
-Mark Langsdorf AMD, Inc.
- 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/
| |