Re: [RESEND PATCH 2/2] sched/fair: Optimize __update_sched_avg()

From: Paul Turner
Date: Thu Mar 30 2017 - 18:03:23 EST


On Thu, Mar 30, 2017 at 7:14 AM, Peter Zijlstra <peterz@xxxxxxxxxxxxx> wrote:
> On Thu, Mar 30, 2017 at 02:16:58PM +0200, Peter Zijlstra wrote:
>> On Thu, Mar 30, 2017 at 04:21:08AM -0700, Paul Turner wrote:
>
>> > > +
>> > > + if (unlikely(periods >= LOAD_AVG_MAX_N))
>> > > return LOAD_AVG_MAX;
>
>> >
>> > Is this correct in the iterated periods > LOAD_AVG_MAX_N case?
>> > I don't think the decay above is guaranteed to return these to zero.
>>
>> Ah!
>>
>> Indeed, so decay_load() needs LOAD_AVG_PERIOD * 63 before it truncates
>> to 0, because every LOAD_AVG_PERIOD we half the value; loose 1 bit; so
>> 63 of those and we're 0.
>>
>> But __accumulate_sum() OTOH returns LOAD_AVG_MAX after only
>> LOAD_AVG_MAX_N, which < LOAD_AVG_PERIOD * 63.
>>
>> So yes, combined we exceed LOAD_AVG_MAX, which is bad. Let me think what
>> to do about that.
>
>
> So at the very least it should be decay_load(LOAD_AVG_MAX, 1) (aka
> LOAD_AVG_MAX - 1024), but that still doesn't account for the !0
> decay_load() of the first segment.
>
> I'm thinking that we can compute the middle segment, by taking the max
> value and chopping off the ends, like:
>
>
> p
> c2 = 1024 \Sum y^n
> n=1
>
> inf inf
> = 1024 ( \Sum y^n - \Sum y^n - y^0 )
> n=0 n=p
>
>

So this is endemic to what I think is a deeper problem:

The previous rounds of optimization folded weight into load_sum. I
think this introduced a number of correctness problems:

a) the load_avg is no longer independent of weight; a high weight
entity can linger for eons. [63 LOAD_AVG_PERIODS]
b) it breaks the dynamic response of load_avg on a weight change.
While nice is not common, there's a case that this is really important
for which is cgroups with a low number of threads running. E.g. When we
transition from 1->2 threads we immediately halve the weight, but
because of the folding it takes a very large time to be numerically
correct again.
c) It doesn't work with scale_load_down and fractional shares below
SCHED_LOAD_SCALE [we multiply in a zero -> zero rq load]
d) It makes doing stability/clipping above a nightmare.

I think it's actually *costing* us cycles, since we end up multiplying
in the weight every partial [load_sum] update, but we only actually
need to compute it when updating load_avg [which is per period
overflow].

> Which gives something like the below.. Or am I completely off my rocker?
>
> ---
> kernel/sched/fair.c | 70 ++++++++++++++---------------------------------------
> 1 file changed, 18 insertions(+), 52 deletions(-)
>
> diff --git a/kernel/sched/fair.c b/kernel/sched/fair.c
> index 76f67b3e34d6..4f17ec0a378a 100644
> --- a/kernel/sched/fair.c
> +++ b/kernel/sched/fair.c
> @@ -2744,26 +2744,6 @@ static const u32 runnable_avg_yN_inv[] = {
> };
>
> /*
> - * Precomputed \Sum y^k { 1<=k<=n }. These are floor(true_value) to prevent
> - * over-estimates when re-combining.
> - */
> -static const u32 runnable_avg_yN_sum[] = {
> - 0, 1002, 1982, 2941, 3880, 4798, 5697, 6576, 7437, 8279, 9103,
> - 9909,10698,11470,12226,12966,13690,14398,15091,15769,16433,17082,
> - 17718,18340,18949,19545,20128,20698,21256,21802,22336,22859,23371,
> -};
> -
> -/*
> - * Precomputed \Sum y^k { 1<=k<=n, where n%32=0). Values are rolled down to
> - * lower integers. See Documentation/scheduler/sched-avg.txt how these
> - * were generated:
> - */
> -static const u32 __accumulated_sum_N32[] = {
> - 0, 23371, 35056, 40899, 43820, 45281,
> - 46011, 46376, 46559, 46650, 46696, 46719,
> -};
> -
> -/*
> * Approximate:
> * val * y^n, where y^32 ~= 0.5 (~1 scheduling period)
> */
> @@ -2795,40 +2775,25 @@ static u64 decay_load(u64 val, u64 n)
> return val;
> }
>
> -static u32 __accumulate_sum(u64 periods, u32 period_contrib, u32 remainder)
> +static u32 __accumulate_pelt_segments(u64 periods, u32 d1, u32 d3)
> {
> - u32 c1, c2, c3 = remainder; /* y^0 == 1 */
> -
> - if (!periods)
> - return remainder - period_contrib;
> -
> - if (unlikely(periods >= LOAD_AVG_MAX_N))
> - return LOAD_AVG_MAX;
> + u32 c1, c2, c3 = d3; /* y^0 == 1 */
>
> /*
> * c1 = d1 y^(p+1)
> */
> - c1 = decay_load((u64)(1024 - period_contrib), periods);
> + c1 = decay_load((u64)d1, periods);
>
> - periods -= 1;
> /*
> - * For updates fully spanning n periods, the contribution to runnable
> - * average will be:
> + * p
> + * c2 = 1024 \Sum y^n
> + * n=1
> *
> - * c2 = 1024 \Sum y^n
> - *
> - * We can compute this reasonably efficiently by combining:
> - *
> - * y^PERIOD = 1/2 with precomputed 1024 \Sum y^n {for: n < PERIOD}
> + * inf inf
> + * = 1024 ( \Sum y^n - \Sum y^n - y^0 )
> + * n=0 n=p+1
> */
> - if (likely(periods <= LOAD_AVG_PERIOD)) {
> - c2 = runnable_avg_yN_sum[periods];
> - } else {
> - c2 = __accumulated_sum_N32[periods/LOAD_AVG_PERIOD];
> - periods %= LOAD_AVG_PERIOD;
> - c2 = decay_load(c2, periods);
> - c2 += runnable_avg_yN_sum[periods];
> - }
> + c2 = LOAD_AVG_MAX - decay_load(LOAD_AVG_MAX, periods) - 1024;
>
> return c1 + c2 + c3;
> }
> @@ -2861,8 +2826,8 @@ accumulate_sum(u64 delta, int cpu, struct sched_avg *sa,
> unsigned long weight, int running, struct cfs_rq *cfs_rq)
> {
> unsigned long scale_freq, scale_cpu;
> + u32 contrib = delta;
> u64 periods;
> - u32 contrib;
>
> scale_freq = arch_scale_freq_capacity(NULL, cpu);
> scale_cpu = arch_scale_cpu_capacity(NULL, cpu);
> @@ -2880,13 +2845,14 @@ accumulate_sum(u64 delta, int cpu, struct sched_avg *sa,
> decay_load(cfs_rq->runnable_load_sum, periods);
> }
> sa->util_sum = decay_load((u64)(sa->util_sum), periods);
> - }
>
> - /*
> - * Step 2
> - */
> - delta %= 1024;
> - contrib = __accumulate_sum(periods, sa->period_contrib, delta);
> + /*
> + * Step 2
> + */
> + delta %= 1024;
> + contrib = __accumulate_pelt_segments(periods,
> + 1024 - sa->period_contrib, delta);
> + }
> sa->period_contrib = delta;
>
> contrib = cap_scale(contrib, scale_freq);