This is the mail archive of the
gcc-patches@gcc.gnu.org
mailing list for the GCC project.
Default to software floating point for xscale-elf port
- From: Nick Clifton <nickc at cambridge dot redhat dot com>
- To: gcc-patches at gcc dot gnu dot org
- Date: 19 Dec 2001 15:22:09 +0000
- Subject: Default to software floating point for xscale-elf port
Hi Guys,
The XScale does not have a hardware floating point coprocessor but
the binaries being generated for the xscale-elf toolchain were not
passing the -mno-fpu switch on to the assembler. Fixewd by applying
the following patch.
Cheers
Nick
2001-12-19 Nick Clifton <nickc@cambridge.redhat.com>
* config/arm/xscale-elf.h (SUBTARGET_EXTRA_ASM_SPEC): Pass
-mno-fpu onto gas unless -mhard-float is specified.
(MULTILIB_DEFAULTS): Include msoft-float.
Index: xscale-elf.h
===================================================================
RCS file: /cvs/gcc/gcc/gcc/config/arm/xscale-elf.h,v
retrieving revision 1.2
diff -p -c -r1.2 xscale-elf.h
*** xscale-elf.h 2001/09/14 10:19:30 1.2
--- xscale-elf.h 2001/12/19 15:23:53
*************** Boston, MA 02111-1307, USA. */
*** 28,36 ****
#define SUBTARGET_CPU_DEFAULT TARGET_CPU_xscale
#endif
! #define SUBTARGET_EXTRA_ASM_SPEC "%{!mcpu=*:-mxscale}"
#ifndef MULTILIB_DEFAULTS
#define MULTILIB_DEFAULTS \
! { "mlittle-endian", "mno-thumb-interwork", "marm" }
#endif
--- 28,36 ----
#define SUBTARGET_CPU_DEFAULT TARGET_CPU_xscale
#endif
! #define SUBTARGET_EXTRA_ASM_SPEC "%{!mcpu=*:-mxscale} %{!mhard-float:-mno-fpu}"
#ifndef MULTILIB_DEFAULTS
#define MULTILIB_DEFAULTS \
! { "mlittle-endian", "mno-thumb-interwork", "marm", "msoft-float" }
#endif