Michael S. Tsirkin wrote:
So, what about an actual benchmark for the two approaches that were proposed
(perfquery / sysfs)?

A benchmark would be nice to see if sysfs could be used, but I'm not sure that it's needed in order to compare the performance between the two. How would sysfs counters get populated if not via a PM MAD?

- Sean

_______________________________________________
openib-general mailing list
[email protected]
http://openib.org/mailman/listinfo/openib-general

To unsubscribe, please visit http://openib.org/mailman/listinfo/openib-general

Reply via email to