[openib-general] Re: [PATCH] RFC Verbs: add support for transport specific verbs

[Grant Grundler]

On Wed, Mar 01, 2006 at 10:11:10AM -0800, Fabian Tillier wrote:
...
> > And this still means we add complexity (need to verify device type)
> > for calling transport-specific functions, which also has performance
> > implications.
> 
> Can you quantify the performance penalty?  I would guess that for the
> device specific operations, beside not being speed-path operations,
> the overhead of checking the device type is likely lost in the noise
> compared to the time it takes to actually perform the operation.

It's not realistic to quantify the performance impact of a single change.
The "roar of the crowd" (many small changes) can be measured.
Just don't add tests or indirect function calls that aren't absolutely
needed. Or attempt to move those into an initialization path.

grant