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: [patches] Re: flow.c/dwarf2out.c and multiple entry points


> On Fri, Feb 23, 2001 at 11:09:21AM +0000, Jason Merrill wrote:
> > Thinking about this, I wonder if it makes more sense to build up an internal
> > representation of the full unwind table while compiling and only break it
> > down into dwarf operations when we need to write it out.  This would seem
> > to offer more control over optimization of the representation.
> 
> I think this would be ideal.
Sadly, when I was writting the patch I believed that code does exactly that...
(the hunk is actually copied from code just above that I believed to do the
initialization).
> 
> The optimization I'd most like to see is to delay mentioning a
> save until the original register actually gets clobbered.  This
> allows us to cluster saves, which would reduce the number of
> advances in the info.
I don't think I am ready to implement such think - as you've noticed my knowledge
of dwarf2 is truly weak.
Sadly this patch is only that blocks the multiple entry points from working.
Do you have any sugestion how to get this working easilly, at least as temporary
solution?

Honza
> 
> 
> r~


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