This is the mail archive of the
gcc-patches@gcc.gnu.org
mailing list for the GCC project.
Re: Bug in reload_cse_move2add()
On Jan 9, 2001, Alexandre Oliva <aoliva@redhat.com> wrote:
> Here's a slightly modified version of Joern's patch, so that it
> doesn't GET_CODE of a reg_offset before checking it's non-NULL.
> Tested with the new port I'm working on. Bootstrapping on
> i686-pc-linux-gnu right now. Ok to install?
Nope. It fails to build stage3, complaining that the floating
constant in toplev.c:1716 is out-of-range. I'll have to investigate
it. Meanwhile, may I install the simpler, safer patch suggested by
Joern:
Index: reload1.c
===================================================================
RCS file: /cvs/gcc/egcs/gcc/reload1.c,v
retrieving revision 1.248
diff -p -r1.248 reload1.c
*** reload1.c 2001/01/02 00:59:52 1.248
--- reload1.c 2001/01/03 18:54:58
*************** move2add_note_store (dst, set, data)
*** 9149,9159 ****
{
if (REGNO (src0) != regno
|| reg_offset[regno] != const0_rtx)
! {
! reg_base_reg[regno] = REGNO (src0);
! reg_set_luid[regno] = move2add_luid;
! }
reg_offset[regno] = XEXP (src, 1);
break;
}
--- 9149,9157 ----
{
if (REGNO (src0) != regno
|| reg_offset[regno] != const0_rtx)
! reg_base_reg[regno] = REGNO (src0);
+ reg_set_luid[regno] = move2add_luid;
reg_offset[regno] = XEXP (src, 1);
break;
}
--
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