processor type.

[Wilko Bulte]

On Sat, Jan 15, 2005 at 10:48:33AM +0100, Wilko Bulte wrote..
> On Sat, Jan 15, 2005 at 10:06:05AM +0100, Bernd Walter wrote..
> > On Sat, Jan 15, 2005 at 09:40:53AM +0100, Wilko Bulte wrote:
> > > On Fri, Jan 14, 2005 at 11:45:23PM +0100, Wilko Bulte wrote..
> > > > NO_CPU_COPTFLAGS=true   # Don't add -march=<cpu> to COPTFLAGS automatically
> > > > 
> > > > fixes it on RELENG_4
> > > 
> > > Well, the elapsed wall clock time for a buildworld running an ev5 world
> > > versus running an ev6 world is only roughly 2 or 3 minutes in favor of a ev6
> > > world on a elapsed time of roughly 75 minutes. This is on a DS10 600MHz 
> > > 
> > > Not too interesting in my book
> > 
> > I wonder if the build environment was compiled using those flags too or
> > just the resulting binaries.
> 
> To clarify:  my tests were 'make buildworlds'.  The first number was while
> building running a EV5 optimised world to do so, the second used the EV6
> world.
> 
> I have not tried to build an EV5 buildworld using the EV6 world
> (so in other words an empty /etc/make.conf).  The machine is currently
> churning on RC3 for re@ so I cannot go try it.

More numbers, this time for a buildrelease of 4.11-RC

while running an 'ev5 world'    32086.29 real     23579.73 user      3221.72 sys
while running an 'ev6 world'    30268.58 real     22634.77 user      3285.35 sys

-- 
Wilko Bulte				wilko at FreeBSD.org