On 06/17/2013 08:20 PM, Paul Turner wrote: > On Fri, Jun 7, 2013 at 12:20 AM, Alex Shi <alex....@intel.com> wrote: >> > blocked_load_avg sometime is too heavy and far bigger than runnable load >> > avg, that make balance make wrong decision. So remove it. > Ok so this is going to have terrible effects on the correctness of > shares distribution; I'm fairly opposed to it in its present form. > > So let's see, what could be happening.. > > In "sched: compute runnable load avg in cpu_load and > cpu_avg_load_per_task" you already update the load average weights > solely based on current runnable load. While this is generally poor > for stability (and I suspect the benefit is coming largely from > weighted_cpuload() where you do want to use runnable_load_avg and not > get_rq_runnable_load() where I suspect including blocked_load_avg() is > correct in the longer term).
If the 'poor stability' means your previous example of 2 40% busy task and one 90% busy task. It occasionally happens. but at least in all testing, kbuild, aim7, tbench, oltp, hackbench, ltp etc. involve blocked_load_avg is just worse, guess due to above reason. > > Ah so.. I have an inkling: > Inside weighted_cpuload() where you're trying to use only > runnable_load_avg; this is in-fact still including blocked_load_avg > for a cgroup since in the cgroup case a group entities' contribution > is a function of both runnable and blocked load. with this patch tg will not include blocked_load_avg. Honestly, blocked_load_avg should has its meaning, like in your scenario. but just now, we only can see it bring more harm without any help on all we tested benchmarks. I can't find a reason to enable sth that hurt performance. > > Having weighted_cpuload() pull rq->load (possibly moderated by > rq->avg) would reasonably avoid this since issued shares are > calculated using instantaneous weights, without breaking the actual > model for how much load overall that we believe the group has. > I considered to use rq->avg in weighted_cpuload, but when we do move_tasks to balance load between cpu, we just consider the cfs tasks not rt task, consider rq->load/avg will involved a unnecessary rt interference. So I changed to cfs load only. -- Thanks Alex -- 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/