[macstl-dev] Re: Question

[Glen Low]

On 02/06/2005, at 9:03 PM, Michael LeBlanc wrote:

> On Jun 1, 2005, at 8:11 PM, Ilya Lipovsky wrote:
>
>
>> I cannot answer your gcc 3.4 related questions, as I am not  
>> working with it.
>>
>
> I went back through my messages and found ...
>
> On Mar 28, 2005, at 10:13 AM, Michael LeBlanc wrote:
>
>
>> This is a problem in GCC < 3.4.  In that version, GCC treats  
>> vector, bool and pixel as Apple does.
>>
>
> ... and this morning crawled through my source trees.  I obviously  
> have too many trees because I should have said "GCC < 4.0".   
> There's a somewhat red face here at the other side of the planet.
>
> Anyhow, with respect to MacSTL, I do not believe there is  
> significant difference between 3.3 and 3.4.

Hmm?

It seems from some third party testing (I asked someone who had 3.4  
etc.) and from gcc's own docs, that 3.4 uses bool as Apple does,  
while 3.3 doesn't.

See http://gcc.gnu.org/onlinedocs/gcc-3.3.5/gcc/PowerPC-AltiVec- 
Built_002din-Functions.html#PowerPC-AltiVec-Built_002din-Functions
which says "The vector bool type is deprecated and will be  
discontinued in further revisions. Use vector signed instead.", and  
there's no evidence of vector bool in the function list.

However by the time we have gcc 3.4:

http://gcc.gnu.org/onlinedocs/gcc-3.4.4/gcc/PowerPC-AltiVec- 
Built_002din-Functions.html#PowerPC-AltiVec-Built_002din-Functions
They now talk about vector bool and the functions all have vector  
bool overloads where appropriate.

Thus I submit mainline gcc 3.4 *does* have vector bool xxx as a  
distinct type, same as Apple gcc 3.3 and different from mainline gcc  
3.3.


Cheers, Glen Low


---
pixelglow software | simply brilliant stuff
www.pixelglow.com
aim: pixglen

-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://www.pixelglow.com/lists/archive/macstl-dev/attachments/20050603/e3eb1b4d/attachment.html