This is the mail archive of the
gcc@gcc.gnu.org
mailing list for the GCC project.
Re: Anyone knows about a testcase which was fixed by Wed Dec 24 18:05:13 1997 SPARC change?
On Fri, Oct 06, 2000 at 01:07:26AM -0700, Richard Henderson wrote:
> On Fri, Oct 06, 2000 at 04:03:02AM -0400, Jakub Jelinek wrote:
> > ... so all functions which it might get through should be within one
> > translation unit ...
>
> But that patch was written before builtin_setjmp_receiver was
> created. It seems likely that nonlocal_goto_receiver should
> be renamed, and triggered only for -mflat -fpic.
Oh yes, you're right. I'll kill nonlocal_goto_receiver then and add
exception_receiver and builtin_setjmp_receiver for -mflat -fpic, which will
unconditionally call newly created load_pic_register. I don't think it is
worth optimizing -mflat -fpic so that the pic register would be in such
places reloaded only if it is needed and it will simplify the code a lot
(i386 does output it unconditionally in receivers as well).
Jakub