[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Re: compiler flags



On  14 Jan, this message from Chris Anderson echoed through cyberspace:
> On Wed, 2004-01-14 at 11:22, Boehn, Gunnar von wrote:
>> Hello everybody,
>> 
>> 
>> MySQL will soon provide precompiled packages for PowerPC as well.
>> I think they will provide tgz for MySQL 4.0, 4.1 and 5.0 
>> 
>> I just spoke to the MySQl package maintainer and
>> we wondered which compiler flags will be the best to use.
>> 
>> 
>> Can someone please give us an overview of the PPC compiler options ?
>> 
>> - Do you think that altivec will give a measurable speed boost ?
>> - Can gcc create code which will use altivec for g4/g5
>>   but will run 'normal' code on non altivec PPC ?
>> - Or would you recommand to create to binaries optimized
>>   one normal and one optimised for altivec ?
> 
> It might help if someone did a comparision between a package compiled
> with altivec support and one without it. If the difference is
> justifiable then it may make sense to seperate the packages, otherwise
> it probably isn't worth the trouble.

This discussion is relevant only if your source code contains Altivec
code as an optional optimisation.

AFAIK, gcc will never produce Altivec code if it wasn't present in the
source; i.e. it will not optimise code for you. That's something you
need to do yourself.

Cheers

Michel

-------------------------------------------------------------------------
Michel Lanners                 |  " Read Philosophy.  Study Art.
23, Rue Paul Henkes            |    Ask Questions.  Make Mistakes.
L-1710 Luxembourg              |
email   mlan@cpu.lu            |
http://www.cpu.lu/~mlan        |                     Learn Always. "



Reply to: