Am Wed, 1 Jun 2011 05:25:13 +0800 schrieb Greg KH <gregkh@suse.de>:
On Tue, May 31, 2011 at 09:57:12AM +0200, Stefan Seyfried wrote:
I never expected to see such pathological workload in the wild, but it exists ;)
Have you reported this bug to the kernel developers?
This is no bug. There are pathological workloads. Pseudo code: for(;;) { receive(&data); /* blocks until data is received */ compute_with_high_cpuload_but finish_quickly(&data); send(&data); } If the computation is finished in << 100ms, the cpufreq will not be switched up. Sleep time between data blocks is large (> 100ms) Now imagine three such processes on three servers, handing data around in turns. Switching to "performance" governor did increase the end-user performance by 30%. The power consumption of those systems did not change in a measurable way. Yes, the workload is pathological. Yes, the software is stupid. Yes, people are paying big $$$ for this software. No comment. Tuning the ondemand governor was also something I tried, and performance did not suffer anymore after that. But then it was so triggerhappy that even on an idle system, it was switching to turbo mode all the time. The performance governor at least saves the kernel from doing this work ;) -- Stefan Seyfried "Dispatch war rocket Ajax to bring back his body!" -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org For additional commands, e-mail: opensuse-factory+help@opensuse.org