John, On Wed, May 7, 2014 at 11:10 AM, John Stultz <john.stu...@linaro.org> wrote: > On 05/06/2014 09:19 PM, Doug Anderson wrote: >> John, >> >> On Tue, May 6, 2014 at 5:25 PM, John Stultz <john.stu...@linaro.org> wrote: >>> Really, I'm curious about the backstory that made you generate the test? >>> I assume something bit you where udelay was way off? Or were you using >>> udelay for some sort of accuracy sensitive use? >> Several times we've seen cases where udelay() was pretty broken with >> cpufreq if you were actually implementing udelay() with >> loops_per_jiffy. I believe it may also be broken upstream on >> multicore systems, though now that ARM arch timers are there maybe we >> don't care as much? >> >> Specifically, there is a lot of confusion between the global loops per >> jiffy and the per CPU one. On ARM I think we always use the global >> one and we attempt to scale it as cpufreq changes. ...but... >> >> * cores tend scale together and there's a single global. That means >> you might have started the delay loop at one freq and ended it at >> another (if another CPU changes the freq). > > Good point. The loops based delay would clearly be broken w/ ASMP unless > we use per-cpu values that are scaled(and as you point out, we don't > scale the value mid-delay). Time based counters for udelay() - like the > arch timer you mention - are a much better way to work around this.
Locally we have this: * https://chromium-review.googlesource.com/189885 ARM: Don't ever downscale loops_per_jiffy in SMP systems I didn't think upstream would really want this given the move to arch timers, but I'm happy to post it. >> * I believe there's some strange issues in terms of how the loops per >> jiffy variable is initialized and how the "original CPU freq" is. I >> know we ran into issues on big.LITTLE where the LITTLE cores came up >> and clobbered the loops_per_jiffy variable but it was still doing math >> based on the big cores. > > Hrm. I don't have a theory on this right now, but clearly there are > issues to be resolved, so having your tests included would be a good > thing to help find these issues. Locally we added: * https://chromium-review.googlesource.com/189725 init: Don't decrease loops_per_jiffy when a CPU comes up ...and that "fixed" things for us. Specifically what happened was: - A15s boot up at 1.8GHz (though they can actually go up to 1.9/2.0). - A7s boot up at ~600MHz and clobbered loops_per_jiffy with something tiny. - In the first "cpufreq_callback" in "arch/arm/kernel/smp.c", we stored the current _A15_ frequency upon the first cpufreq transition and the current _A7_ loops per jiffy (since the A7s clobbered it). It seemed like our situation was so messed up that upstream probably wouldn't want the fix (using loops_per_jiffy in an HMP system is pretty insane), but again I'm happy to send it up. We also took out some IKS (big.LITTLE) attempt to fix up loops_per_jiffy that was massively flawed in a number of ways <https://chromium-review.googlesource.com/189728> but that's not upstream. -Doug -- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majord...@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/
