[PD] array wierdness (x86_64)

Mathieu Bouchard matju at artengine.ca
Fri May 5 17:52:20 CEST 2006

On Wed, 3 May 2006, august wrote:

> I searched for x86_64 online and see that this is a known issue, and
> that the fix is not very trivial. Anyone know what the fix would entail?  
> Where to start?  Is it just with the tables?
> Maybe I can get to it, since I will be needing it eventually.

At the very least you have to edit d_soundfile.c; in five minutes you can
find two obvious problems (int vs long) but it may take a long time to 
take the remaining bugs.

Good luck reviewing d_soundfile.c, and one last tip, if you see a function
called "pute", don't mind, Miller thought the name was funny because it
means "whore" in French, but he defined it as just a conditional post()
for debugging purposes. Don't expect the name to mean anything relevant
like s_inter.c's "sys_defaultfontshit".

I was going to fix the 64-bit mode but I figured that 64-bit mode is still
mostly useless because many other apps still have serious 64-bit mode bugs
(so I can't really use a 64-bit mode OS) and because people don't (yet)
have a need to put more than 4 gigs in the same process and because the
speedup due to increased register-set isn't that great (which compiler
and/or options of compiler do i have to use to see a real difference?) and
also because no-one did put money on the bugs.

 _ _ __ ___ _____ ________ _____________ _____________________ ...
| Mathieu Bouchard - tél:+1.514.383.3801 - http://artengine.ca/matju
| Freelance Digital Arts Engineer, Montréal QC Canada

More information about the Pd-list mailing list