This is the mail archive of the
gcc-patches@gcc.gnu.org
mailing list for the GCC project.
Re: PATCH: if -fpic then don't consider PIC_OFFSET_TABLE_REGNUM updates dead
On Mar 15, 2001, Richard Henderson <rth@redhat.com> wrote:
> On Thu, Mar 15, 2001 at 07:29:58PM -0300, Alexandre Oliva wrote:
>> I still think marking it as MAYBE_DEAD is the right way to fix the problem.
> Certainly better than hacking flow. If Toshi has come up
> with a case in which the last GOT reference is removed by
> flow2, then do go ahead and use MAYBE_DEAD.
I'm checking this patch in mainline and branch.
Index: gcc/ChangeLog
from Alexandre Oliva <aoliva@redhat.com>
* config/sh/sh.c (sh_expand_prologue): Insns that set up the PIC
register may be dead.
Index: gcc/config/sh/sh.c
===================================================================
RCS file: /cvs/gcc/egcs/gcc/config/sh/sh.c,v
retrieving revision 1.89
diff -u -p -r1.89 sh.c
--- gcc/config/sh/sh.c 2001/02/02 18:20:29 1.89
+++ gcc/config/sh/sh.c 2001/03/16 12:30:44
@@ -3987,7 +3987,23 @@ sh_expand_prologue ()
push_regs (live_regs_mask, live_regs_mask2);
if (flag_pic && regs_ever_live[PIC_OFFSET_TABLE_REGNUM])
- emit_insn (gen_GOTaddr2picreg ());
+ {
+ rtx insn = get_last_insn ();
+ rtx last = emit_insn (gen_GOTaddr2picreg ());
+
+ /* Mark these insns as possibly dead. Sometimes, flow2 may
+ delete all uses of the PIC register. In this case, let it
+ delete the initialization too. */
+ do
+ {
+ insn = NEXT_INSN (insn);
+
+ REG_NOTES (insn) = gen_rtx_EXPR_LIST (REG_MAYBE_DEAD,
+ const0_rtx,
+ REG_NOTES (insn));
+ }
+ while (insn != last);
+ }
if (target_flags != save_flags)
{
--
Alexandre Oliva Enjoy Guarana', see http://www.ic.unicamp.br/~oliva/
Red Hat GCC Developer aoliva@{cygnus.com, redhat.com}
CS PhD student at IC-Unicamp oliva@{lsd.ic.unicamp.br, gnu.org}
Free Software Evangelist *Please* write to mailing lists, not to me