This is the mail archive of the
gcc-patches@gcc.gnu.org
mailing list for the GCC project.
Re: Simple commit_edge_insertions fix
> On Sun, May 14, 2000 at 10:26:03PM +0200, Jan Hubicka wrote:
> > In that case the sibling call to exit edge outght to have abnormall flag
> > or not?
>
> It should, yes.
>
>
> r~
OK? (once bootstrap finishes :)
BTW whats about non-simple jumps? The clobbers may be dangerous as well.
Mon May 8 20:25:57 CEST 2000 Jan Hubicka <jh@suse.cz>
* flow.c (make_edge): Edges from sibbling calls to exit are abnormal.
*** flow.c.old Mon May 8 20:25:18 2000
--- flow.c Mon May 8 20:25:37 2000
*************** make_edges (label_value_list)
*** 1007,1013 ****
wouldn't have created the sibling call in the first place. */
if (code == CALL_INSN && SIBLING_CALL_P (insn))
! make_edge (edge_cache, bb, EXIT_BLOCK_PTR, 0);
else
/* If this is a CALL_INSN, then mark it as reaching the active EH
--- 1007,1013 ----
wouldn't have created the sibling call in the first place. */
if (code == CALL_INSN && SIBLING_CALL_P (insn))
! make_edge (edge_cache, bb, EXIT_BLOCK_PTR, EDGE_ABNORMAL);
else
/* If this is a CALL_INSN, then mark it as reaching the active EH