------- Comment #1 from pinskia at gcc dot gnu dot org 2007-02-11 10:31 ------- Actually this has nothing to do with -fsplit-wide-types, This shows up in some cases without that, There is a reload patch floating about which fixes this. The problem is that reload does not elimate the soft frame pointer. In fact this is really a regression in 4.1.0 exposed by: 2005-06-30 Jakub Jelinek <[EMAIL PROTECTED]>
* config/rs6000/rs6000.h (FIRST_PSEUDO_REGISTER): Increment. (DWARF_FRAME_REGISTERS, DWARF_REG_TO_UNWIND_COLUMN): Adjust, so that addition of sfp doesn't change these. (FIXED_REGISTERS, CALL_USED_REGISTERS, CALL_REALLY_USED_REGISTERS, .... I actually saw it in a modified version of 4.1.1. -- pinskia at gcc dot gnu dot org changed: What |Removed |Added ---------------------------------------------------------------------------- GCC target triplet|powerpc-suse-linux |powerpc*-*-* Summary|[4.3 regression] Error: |[4.1/4.2/4.3 regression] |unsupported relocation |Error: unsupported |against sfp |relocation against sfp Target Milestone|--- |4.1.2 http://gcc.gnu.org/bugzilla/show_bug.cgi?id=30761