This is the mail archive of the
newlib@sources.redhat.com
mailing list for the newlib project.
RE: arm and -msoft-float
- To: <newlib at sources dot redhat dot com>
- Subject: RE: arm and -msoft-float
- From: Ole Sather <ole dot saether at nvlsi dot no>
- Date: Tue, 30 Oct 2001 07:20:18 +0100
- Reply-To: <ole dot saether at nvlsi dot no>
Try the "-nfp" option to configure (gcc and newlib)
Ole
-----Original Message-----
From: newlib-owner@sources.redhat.com
[mailto:newlib-owner@sources.redhat.com]On Behalf Of Michael Butts
Sent: 29. oktober 2001 23:44
To: newlib@sources.redhat.com
Subject: arm and -msoft-float
I am trying to build an arm-elf toolset using gcc
2.95.3, binutils 2.11.2 and newlib 1.9.0 on a FreeBSD
4.4 system.
I am able to build everything successfully, but cannot
get soft-float support.
binutils was configured using:
--prefix=/usr/local/cross_dev/arm --target=arm-elf
In the gcc-2.95.3 dir, I created a link to
../newlib-1.9.0/newlib and configured as:
--prefix=/usr/local/cross_dev/arm --target=arm-elf
--enable-multilib --with-newlib
Before making gcc, I went into the gcc dir and
uncommented the lines in the Makefile:
MULTILIB_OPTIONS = mlittle-endian/mbig-endian
mhard-float/msoft-float mapcs-32/mapcs-26
fno-leading-underscore/fleading-underscore
MULITLIB_DIRNAMES = le be fpu soft 32bit 26bit elf
under
After make/install, I end up with the following
directories in the arm/arm-elf/lib dir:
be fpu 26bit elf under
No soft!
When building my source code:
/usr/local/cross_dev/arm/bin/arm-elf-ld: Error:
memcpy.o uses hard floating point, whereas hotrod.elf
uses soft floating point
File in wrong format: failed to merge target specific
data of file
/usr/local/cross_dev/arm/arm-elf/lib/be/soft/libc.a(memcpy.o)
Any ideas on how to get soft-float support?
Thanks,
Mike
__________________________________________________
Do You Yahoo!?
Make a great connection at Yahoo! Personals.
http://personals.yahoo.com