linux-mips-fnet
[Top] [All Lists]

Re: M700-10

To: linux-mips@fnet.fr
Subject: Re: M700-10
From: Systemkennung Linux <linux@mailhost.uni-koblenz.de>
Date: Mon, 11 Mar 1996 15:42:02 +0100 (MET)
In-reply-to: <199603102318.AAA17030@pallas.spacetec.no> from "Tor Arntsen" at Mar 11, 96 00:18:04 am
Hi,

> Ok, that explains why I couldn't read it.. I've got only 1.44 MB drives here.
> This floppy says 'ED' instead of 'HD' as the others.  I'll just try to 
> borrow a 2.88MB drive in case I need that utility.  

If everything fails I could send you an image of the 2.88mb disk.

> Now the first problem..
> 
> I have set up my cross-development environment on an SGI Challenge.
> I'm using gcc 2.7.2, binutils 2.6, all set up for mipsel-linuxelf, 
> corresponding patches applied.
> I have also grabbed and installed the mipsel-linuxelf version of libc-951212,
> and linux-1.3.62.tar.gz (all from ftp.fnet.fr).
> This went mostly without problems, although IRIX has some annoying limits
> w.r.t. the size of commands arg list etc.  Had to modify the .hdepend target 
> in the main Makefile because of this.
> 
> The problem I ran into during the kernel build is that gas chokes with the 
> following message:
> 
> magnum4000.S: Assembler messages:
> magnum4000.S:30: Error: Can not represent relocation in this object file 
> format
> 
> The line in question is
>                 /*
>                  * Get pending interrupts
>                  */
>                 mfc0    t0,CP0_CAUSE            # get pending interrupts
>                 mfc0    t1,CP0_STATUS           # get enabled interrupts
>                 and     t0,t1                   # isolate allowed ones
>                 andi    t0,0xff00               # isolate pending bits
> --->            beqz    t0,spurious_interrupt
>                 sll     t0,16                   # delay slot

This is a branch to an external target address.  Unfortunately this can
not be represented in ELF.

The switch of Linux/MIPS to ELF isn't complete yet.  You still need the
a.out compiler to compile the kernel; *all* user code is ELF.  Though
I promise this alredy for some time:  I'll fix the ELF problems with the
kernel as soon as I can and will then eleminate the a.out support in
compiler/linker.

  Ralf

<Prev in Thread] Current Thread [Next in Thread>