cvs commit: src/share/mk bsd.cpu.mk bsd.dep.mk bsd.lib.mk bsd.sys.mk src/sys/conf files kern.mk kern.pre.mk kmod.mk src/sys/dev/aic7xxx/aicasm Makefile

[Tom Rhodes]

On Thu, 18 Mar 2004 14:05:56 -0500
Ender <ender at tog.net> wrote:

> Tom Rhodes wrote:
> 
> >On Thu, 18 Mar 2004 09:28:27 -0800
> >"David O'Brien" <obrien at freebsd.org> wrote:
> >
> >  
> >
> >>Just to be clear -- I'm not against supporting a 2nd compiler in /usr/src
> >>    
> >>
> >
> >Then why voice any opinion at all?
> >
> >  
> >
> >>at all.  Just don't think Intel is the most gracious compiler vendor.
> >>    
> >>
> >
> >Oh, because you don't like Intel?  I agree that they're greedy, but
> >that shouldn't really ...
> >
> >  
> >
> >>I'll also strongly push back on ever using 'icc' as part of the release
> >>build.
> >>    
> >>
> >
> >push back a release build.  Are you worried that the 'rumors?' of
> >better, more optimized binaries going to hurt the AMD effort?
> >
> >  
> >
> Your comments are a lot more biased than the original comments. It has 
> already been shown that the intel compiler "advantages" run exactly the 
> same on a AMD cpu. (after forcing the changes to ignore the cpu type) It 
> has also been stated that a company is in full right to apply whatever 
> advantages to there product they see fit, and i agree. I also believe 
> that if we are going to support a change that clearly gives advantages 
> to one cpu type, we should spend a equal amount of time supporting 
> advantages to all specific cpu types. ( the EFER.NXE in the 
> FreeBSD/amd64 loader for example)
> 

Umm, I'm really wondering if you're talking to David or myself.

-- 
Tom Rhodes