cvs commit: src/share/mk bsd.cpu.mk bsd.dep.mk bsd.lib.mkbsd.sys.mksrc/sys/dev/aic7xxx/aicasm Makefile

[Ender]

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)