mudflap versus cgraph

[Jan Hubicka]

> On Tue, 22 Jun 2004 10:31:31 -0700
> Richard Henderson <rth@redhat.com> wrote:
> 
> > On Mon, Jun 21, 2004 at 09:14:04PM -0700, Zack Weinberg wrote:
> ...
> > > provided that we could figure out how to construct a function in a
> > > completely language-independent manner (which we need anyway).
> > 
> > We have gimple/generic.  We *can* create completely languages
> > independent functions now.
> 
> What should it look like?  All existing examples of synthetic function
> creation involve calling a few language-dependent hooks (if nothing
> else, the start_function and finish_function equivalents).

I have little time to walk down the thread, but just small warning.  At
the moment there are quite few categories of the functinos - mudflap,
C++ constructors, objc interfaces, gcov constructors and probably few
others.  So this is definitly cool thing to do, but it is going to take
considerable effort.

I am having problems to think of what possibly broke for you with
mudflap, perhaps there is easier sollution to break the dependency in
between this and IMA?

Honza
> 
> > >         .section .rodata.mf.statics, "a"
> > >         .long x
> > >         .long 4
> > 
> > This is pic-safe if you put it in a writeable section.
> 
> Good to know, but this change is more invasive than I want to get into
> right now.
> 
> zw