From: jon AT heber DOT co DOT uk (Jon McGugan)
Subject: cross compiling to m68k no floating point
14 May 1997 19:06:42 -0700
Approved: cygnus DOT gnu-win32 AT cygnus DOT com
Distribution: cygnus
Message-ID: <06391549900115.cygnus.gnu-win32@DOMAIN_H>
References: Your message of "Mon, 12 May 1997 19:08:01 +0900."             <3376EC01 DOT 317B AT pi DOT titech DOT ac DOT jp>
Reply-To: jon AT heber DOT co DOT uk
Mime-Version: 1.0
Content-Type: text/plain; charset=US-ASCII
Content-Transfer-Encoding: 7BIT
Comments: Authenticated sender is <jon@[192.0.0.40]>
Original-To: gnu-win32 AT cygnus DOT com
Original-CC: jon AT heber DOT co DOT uk
In-reply-to: <9705122346.AA10143@modi.xraylith.wisc.edu>
X-mailer: Pegasus Mail for Win32 (v2.53/R1)
Original-Sender: owner-gnu-win32 AT cygnus DOT com

Am trying to cross compile the gnu gcc compiler b17 version to target 
m68k with no floating point. Gas and Ld configured.

The make in gcc fails in the libgcc2 when it tries to extract 
_floatdidf with errors of form:

Error: Invalid instruction for this architecture; needs 68040 or 6060 
-- statement `fdmove.l .....` ignored

Lots of this. It appears to be trying to use floating point opcodes 
to implement functions/macros.

Help appreciated in what I've done wrong. configured with

--target=m68k-motorola-coff --nfp

Regards

Jon
-
For help on using this list (especially unsubscribing), send a message to
"gnu-win32-request AT cygnus DOT com" with one line of text: "help".