interesting(?) data on network interrupt servicing

[Julian Elischer]

Paolo Pisati wrote:

>On Wed, Mar 22, 2006 at 12:29:06PM -0800, Luigi Rizzo wrote:
>  
>
>>Paolo should follow up in the next days with graphs and more data.
>>    
>>
>
>Here are graphs and data:
>
>http://mercurio.sm.dsi.unimi.it/~pisati/
>
>and here is a tarball with all images, data and patches:
>
>http://mercurio.sm.dsi.unimi.it/~pisati/interrupt.tgz
>
>  
>
Interesting that the timespent actually in the driver is similar. jus 
that there is a constant extra 10000
units of overhead for every interrupt at the lower levels before it gets 
to the driver..