This is the mail archive of the
libc-alpha@sources.redhat.com
mailing list for the glibc project.
Re: Questions regarding m68k dynamic/shared code
- From: Peter Barada <pbarada at mail dot wm dot sps dot mot dot com>
- To: schwab at suse dot de
- Cc: Peter dot Barada at motorola dot com, binutils at sources dot redhat dot com, gcc at gcc dot gnu dot org, libc-alpha at sources dot redhat dot com
- Date: Thu, 9 Jan 2003 11:36:14 -0500
- Subject: Re: Questions regarding m68k dynamic/shared code
- References: <200301081803.h08I3vV19219@hyper.wm.sps.mot.com> <jewuleo4rq.fsf@sykes.suse.de>
>|> I'm trying to build a Linux system to run on a ColdFire v4e system,
>
>Note that Linux/m68k is assuming mc68020up, and this assumption is
>hardcoded in may places.
I'm assuming that you mean the assembler, compiler and glibc. If so,
then yes, I know.... So far I've (rough outline):
1) Modified gas to add v4e instructions, -mcfv4e commanline switch.
2) Modify bfd/elf32-m68k.c to produce PLT code that is valid for ColdFire.
3) Modified the default gcc/config/m68k/linux.h to allow for commandline
selections of -m5200, -mcfv4e, as well as multilib libgcc so linking
will pick the correct library.
4) Tweaked GCC to produce -FPIC code that can assemble on ColdFire.
5) Added a --target=coldfire-linux to glibc-2.2.5(w/o FPU)
to allow me to build glibc for ColdFire.
6) Modified the 2.4.19 linux kernel to support the v4e MMU.
7) Modified gdb-5.2.1 to support FPU register size of 64 bits.
So far the kernel starts up and launches init, and dies. I'm in the
midst of trying to figure it why/where in init it dies by using a
bdi-2000 to single step the user application and looking at the
registers.
Is there anything really obvious I've overlooked?
--
Peter Barada Peter.Barada@motorola.com
Wizard 781-852-2768 (direct)
WaveMark Solutions(wholly owned by Motorola) 781-270-0193 (fax)