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]
Other format: [Raw text]

Re: Fix support for gigantic stack frames on x86-64


On Fri, Jun 06, 2003 at 09:15:58AM +0200, Jan Hubicka wrote:
> It appears to parse insns locally so it aborts seeing unknown register.
> Shall I add the code to look backward for load of the constant?

No, I think I'll take care of this soon.  I think this bit of
code is in need of some serious re-design, not more hackery.

> > Not that we don't need to do that anyway so that we can properly
> > annotate epilogues...
> Hmm, what do you mean?

Well, if you deallocate the stack frame and return in the next
instruction, there is a 1 insn window where a signal could be
delivered.  With -fasynchronous-unwind-tables, there are supposed
to be no such holes.

This would be an issue if you really did use the amd64 tables
e.g. for garbage collection.  And according to Geert, it *will*
be an issue for Ada.


r~


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