Jason Thorpe <thorpej@wasabisystems.com> writes:
> On Sunday, September 21, 2003, at 03:41  PM, Perry E. Metzger wrote:
> 
> > Well, we did switch from gcc 2.x to gcc 3.x. Could you check if that
> > had some impact?
> 
> Wasn't openssl itself also upgraded during that timeframe?

Indeed it was. So we have a number of variables that might be coming
in to play. It might be that we forgot to turn on some sort of
platform optimizations, or it might be other openssl slowdowns, or it
might be gcc, or...

I would suggest that Bill might want to poke around a bit and figure
out the exact source of the slowdown.

FYI, this sort of thing is one reason why I wish we had a machine
running benchmarks automatically every day or so -- so we might notice
changes with unintended effects on performance. Sadly, I don't know
that anyone has the time to set it up...


-- 
Perry E. Metzger		perry@piermont.com