Re: HEADS-UP: IFLIB implementations of sys/dev/e1000 em, lem, igb pending

From: Matthew Macy <mmacy_at_nextbsd.org>
Date: Wed, 11 Jan 2017 16:54:43 -0800
 >  A flame graph for the core cycle count and a flame graph with cache miss stats from pmc would be a great start.
 >  
 > 
 > ​I didn't know the exact event name to use for cache miss stats, but here are the flame graphs for CPU_CLK_UNHALTED_CORE:
 > http://dev.bsdrp.net/netgate.r311848.CPU_CLK_UNHALTED_CORE.svg
 > http://dev.bsdrp.net/netgate.r311849.CPU_CLK_UNHALTED_CORE.svg

Thanks. Having twice as many txqs would definitely help. It's also clear that there may be some sort of peformance issue in iflib_txq_drain. Although it could just be non-stop cache misses on the packet headers.

-M
Received on Wed Jan 11 2017 - 23:54:52 UTC

This archive was generated by hypermail 2.4.0 : Wed May 19 2021 - 11:41:09 UTC