Re: [PATCH v3 2/4] KVM: use simple waitqueue for vcpu->wq

[Paolo Bonzini]

On 20/10/2015 09:28, Daniel Wagner wrote:
> 	min              max          mean            std
> count   382.000000       382.000000    382.000000     382.000000
> mean   6068.552356    269502.528796   8056.016198    3912.128273
> std     707.404966    848866.474783   1062.472704    9835.891707
> min    2335.000000     29828.000000   7337.426000     445.738750
> 25%    6004.500000     44237.500000   7471.094250    1078.834837
> 50%    6372.000000     64175.000000   7663.133700    1783.172446
> 75%    6465.500000    150384.500000   8210.771900    2759.734524
> max    6886.000000  10188451.000000  15466.434000  120469.205668
> 
> After
> 	min             max          mean           std
> count   300.000000      300.000000    300.000000    300.000000
> mean   5618.380000   217464.786667   7745.545114   3258.483272
> std     824.719741   516371.888369    847.391685   5632.943904
> min    3494.000000    31410.000000   7083.574800    438.445477
> 25%    4937.000000    45446.000000   7214.102850   1045.536261
> 50%    6118.000000    67023.000000   7417.330800   1699.574075
> 75%    6224.000000   134191.500000   7871.625600   2809.536185
> max    6654.000000  4570896.000000  13528.788600  52206.226799

Anything above ~10000 cycles means that the host went to C1 or
lower---the number means more or less nothing in that case.

The mean shows an improvement indeed.

Paolo
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/