This is the mail archive of the
gcc@gcc.gnu.org
mailing list for the GCC project.
Re: reload buglet
- To: Joern Rennecke <amylaar at cygnus dot co dot uk>
- Subject: Re: reload buglet
- From: Bernd Schmidt <bernds at pathia dot cygnus dot co dot uk>
- Date: Sat, 27 Nov 1999 13:22:11 +0000 (GMT)
- cc: gcc at gcc dot gnu dot org, gcc-patches at gcc dot gnu dot org
On Thu, 25 Nov 1999, Joern Rennecke wrote:
> Just found this in reloads_conflict:
>
> case RELOAD_FOR_OUTPUT:
> return (r2_type == RELOAD_FOR_INSN || r2_type == RELOAD_FOR_OUTPUT
> || ((r2_type == RELOAD_FOR_OUTPUT_ADDRESS
> || r2_type == RELOAD_FOR_OUTADDR_ADDRESS)
> && r2_opnum >= r1_opnum));
>
> That should be r2_opnum <= r1_opnum .
That seems correct to me. Could someone approve this change?
Bernd
diff -c -p -r1.182 reload1.c
*** reload1.c 1999/11/09 16:52:19 1.182
--- reload1.c 1999/11/27 13:14:25
*************** reloads_conflict (r1, r2)
*** 5232,5238 ****
return (r2_type == RELOAD_FOR_INSN || r2_type == RELOAD_FOR_OUTPUT
|| ((r2_type == RELOAD_FOR_OUTPUT_ADDRESS
|| r2_type == RELOAD_FOR_OUTADDR_ADDRESS)
! && r2_opnum >= r1_opnum));
case RELOAD_FOR_INSN:
return (r2_type == RELOAD_FOR_INPUT || r2_type == RELOAD_FOR_OUTPUT
--- 5232,5238 ----
return (r2_type == RELOAD_FOR_INSN || r2_type == RELOAD_FOR_OUTPUT
|| ((r2_type == RELOAD_FOR_OUTPUT_ADDRESS
|| r2_type == RELOAD_FOR_OUTADDR_ADDRESS)
! && r2_opnum <= r1_opnum));
case RELOAD_FOR_INSN:
return (r2_type == RELOAD_FOR_INPUT || r2_type == RELOAD_FOR_OUTPUT