PATCH: Fix 20000724-1.c

[Jan Hubicka]

> > Or, I see that Jakub refers to a patch from Jan that does something
> > different, in that it puts the combinations in conservative locations?
> 
> I do seem to remember something like that.  But I thought
> it had already gone in.  Jan, am I mistaken?
No, it wasn't merged yet.  I've asked for that few times in
"unreviewed patches" mail, but didn't had luck to get it
reviewed yet.
I am just digging out the archive to find relevant patch and
send pointer shortly.
> 
> > Might something else might later move them around again?
> 
> The only other thing would be the scheduler, but volatile
> asms are scheduling barriers.
I think there is bug in scheduler anyway.
Imagine that you take address of stack slot into register,
then read using that register and then deallocate.  I don't
see any way scheduler will add dependancy between deallocate
and read.
> 
> Worse come to worse, it might make sense to not do stack
> comhining across volatile asms either.
Why exactly?

Honza
> 
> 
> r~