This is the mail archive of the gcc-patches@gcc.gnu.org mailing list for the GCC project.


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]

Re: First part of local spilling patches


> 
> So what hunk shall we go after now? :-)  Maybe the changes to make stupid
> use the reload chains?  Or should we get more ambitious and go after all
> code to use the reload chains, but implement local spilling?
> 
> Your thoughts would be greatly appreciated.

I'm not sure.  The remaining parts of the patch depend on each other rather
tightly, so I don't know how much I can still break out.  While the patch
is still large, a lot of it is of the form
  s/counted_for_groups/chain->counted_for_groups/
and similar.
I'll send a patch first that adds the maybe_fix_stack_asms function which
is pretty self-contained.  Maybe I'll also try to separate the change that
creates and deletes caller-save insns on every iteration.

Bernd



Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]