RE: Busybox of Hardhat preview 2.1 for sandpoint 74xx ?

I rebuilt the zImage and this time everything is ok.

Thanks.

Bo Yu
Motorola GTSS
1-817-245-6725


-----Original Message-----
From: Mark Hatle [mailto:fray@mvista.com]
Sent: Tuesday, May 06, 2003 7:45 AM
To: Matt Porter
Cc: Dale Farnsworth; Yu Bo-BOYU1; linuxppc-embedded@lists.linuxppc.org
Subject: Re: Busybox of Hardhat preview 2.1 for sandpoint 74xx ?


Matt Porter wrote:

On Mon, May 05, 2003 at 08:15:44PM -0700, Dale Farnsworth wrote:

quoted

On Mon, May 05, 2003 at 08:31:34PM +0000, Yu Bo-BOYU1 wrote:

quoted

Did anyone know if busybox of hardhat preview 2.1 for sandpoint 74xx work

or

quoted

quoted

not ?
I got error "Bad trap at PC: 300162fc, SR: f032, vector=f00    Not

tainted"

quoted

quoted

no matter
the root is mounted on NFS or ramfs.

It's been a while since I last saw it, but this reminds me of what
happens when running the 74xx userspace on a kernel that doesn't have
altivec support enabled.

Just to confirm, with the MSR VEC off on a 74xx core, it will definitely
generate that bad trap (UnknownException) on mvista 74xx userspace.  I've
seen it on 7xx's that were unintentionally run on the ppc_74xx userspace,
for example.

That would definatly explain it then.  MontaVista ppc_74xx userspace has
altivec
enabled in glibc.  So the userspace DOES use altivec calls.  (Altivec
registers
are saved on signals and longjmp... this allows you to use altivec
instructions
safely in complex programs.)

--Mark

** Sent via the linuxppc-embedded mail list. See http://lists.linuxppc.org/