Mail Archives: djgpp/1997/05/20/11:34:31
Andrew Crabtree wrote:
>
> > I've downloaded binaries of PGCC for DOS... and I'm completely
> > disappointment. The code that it has produced from my program (last
> > address main minus first addres of my first function) is about 20%
> > larger due to incredible amounts on nops inside the code. The
>
> My experience has been that pgcc tends NOT to add no-ops into
> the code, this was something 2.7.2 was guilty of. If you have code
> that demonstrates otherwise feel free to post it (instead of
> just complaining). It would also help is you posted the
> assembly output of the two compilers, but I can redo that myself.
>
What you said IS TRUE, but the nops are on the end of all functions. The
amount of nops is so big thet it makes my code 20% bigger (don't know if
only this), I've heard that PGCC produces smaller code, becouse it
doesn't use nops. Me expirience is different, it uses nops but in
different way.
> > instructions are unscheduled (just swaping some pairs could halp) and
> > have a lot of AGI stalls. I didn't notice any speed differences
> > comparing to 2.7.2 (-02). I've used -O6 option and even some swiches
>
> How were you testing and measuring speed?
>
Just simple fps counting. I'm shure there is some difference (maybe
small) but I didn't noticed any with this method (I've disabled all my
assembler code in triangle rendering functions).
> You might want to consider sending posts like this to the pgcc mailing list
> if you can provide sufficient details.
>
My intention was not to complain, but to come to know what maight
happend. Meybe there's a swich that I don't know abour or somethink like
this? Or meybe my code is so specyfic that the compiler doesn't do its
work on it (I don't think so; unscheduled instructions...)?.
> Andrew
- Raw text -