Re: Testing the effects of the low latency patch

[J Sloan]

It was during gameplay, no dbench running, which
as it turns out is a workload of interest for me -

Joe

Andrew Morton wrote:

>J Sloan wrote:
>
>>>>2.4.18-pre6+tux+nfs-fixes
>>>>-------------
>>>>...
>>>>7.6 1
>>>>7.8 1
>>>>21.1 1
>>>>
>>>This is the stock kernel.  In twenty minutes you suffered
>>>precisely *one* scheduling overrun which is perceptible
>>>by a human.  The rest are much shorter than your monitor's
>>>refresh interval.  Interesting, yes?
>>>
>>Yes, the stock kernel is much improved from
>>say 6 months ago. I will take a look at the
>>kernel that shipped with my distro just for
>>giggles as well...
>>
>
>Was you histogram generated during a game session, or during
>dbench?  write()-intensive workloads are the main common
>offender.
>
>-
>