PATCH for alias sets/stack slots, again

Richard Henderson rth@cygnus.com
Fri Feb 19 17:26:00 GMT 1999


On Fri, Feb 19, 1999 at 12:37:59AM -0700, Jeffrey A Law wrote:
> If the old slot has alias set zero and the desired alias set is zero, then we
> can reuse the slot.  Right?

No, Mark is right.  The problem is that we have "false zeros", in that
we lose alias set information on slots when we regenerate the slots 
during inline function expansion.  If it weren't for that, we could
reuse them.

I think the only proper solution to this problem is to do inline expansion
as trees before any stack anythings exist.  Then we'll never lose alias
set information, and this particular problem won't exist.

Your points about possibly not reusing memory for different sizes etc
is probably something that we should look at anyway.


r~


More information about the Gcc-patches mailing list