Linux/Mips installation

Linux/Mips installation

[Jake Griesbach]

Hi,

I'm new to this list, so please forgive me if this has already been
addressed earlier...

I'm trying to install SGI/Linux onto an Indy workstation using the
instructions found at http://www.linux.sgi.com/mips/manhattan.  I've got
the machine so that it net boots the kernel and starts the installation
program.  However, after I select the hard drive I want to install to, I
get the following error:

                                         Oops: 0000
$0 : 00000000 1000fc00 00001000 ffffffe0
$4 : 00000020 00000000 1000fc00 1000fc00
$8 : 00000010 881095c0 00000001 00000068
$12: 00000000 8df81740 8df80000 88139e80
$16: 00000000 00001000 adf49000 8df47800
$20: 8df50e70 bfbc0003 00000000 bfb90000
$24: 00000001 2abe2f30
$28: 8de6a000 8de6bb98 8df50e70 880fa8c4
epc   : 88020f80
Status: 1000fc02
Cause : 00000008
install exited abnormally -- received signal 11
sending termination signals...done
sending kill signals...done
umounting filesystems...
/proc
umount failed /tmp
you may safely reboot your system


This happens directly after I select my target hard drive from the hard
drive list.  It doesn't seem to matter whether or not I try to partition
it with the Edit option on the menu.

Do you have any ideas as to what may be wrong?  I'm using the
hardhat-sgi-5.1.tar.gz file from
ftp://ftp.linux.sgi.com:/pub/linux/mips/RedHat.

Thanks!!

Jake Griesbach
University of Colorado

Re: Linux/Mips installation

[Thomas Bogendoerfer]

On Mon, Mar 08, 1999 at 09:57:29AM -0700, Jake Griesbach wrote:
> Do you have any ideas as to what may be wrong?  I'm using the
> hardhat-sgi-5.1.tar.gz file from
> ftp://ftp.linux.sgi.com:/pub/linux/mips/RedHat.

Please try a newer kernel. Take the one from

ftp://ftp.linux.sgi.com/pub/linux/mips/test/vmlinux-indy-2.2.1-990226.tar.gz

If you have problems with this kernel, please send the output of
hinv (IRIX command).

Thomas.

-- 
   This device has completely bogus header. Compaq scores again :-|
It's a host bridge, but it should be called ghost bridge instead ;^)
                                        [Martin `MJ' Mares on linux-kernel]

Re: Linux/Mips installation

[Jake Griesbach]

Hmm.  That still didn't work, although the new kernel booted just fine.  I
still got an Oops message right after I selected the drive to install to.
Could this be a swapon problem?  I also have an internal floppy/floptical
drive (this also appears on the select drive list).  Could this be
throwing things off?  Does this work for R4600 processors?

Here's my hinv output:
CPU: MIPS R4600 Processor Chip Revision: 2.0
FPU: MIPS R4600 Floating Point Coprocessor Revision: 2.0
1 133 MHZ IP22 Processor
Main memory size: 96 Mbytes
Instruction cache size: 16 Kbytes
Data cache size: 16 Kbytes
Integral SCSI controller 0: Version WD33C93B, revision D
  Disk drive: unit 1 on SCSI controller 0
  Disk drive / removable media: unit 2 on SCSI controller 0
  Disk drive: unit 3 on SCSI controller 0
  CDROM: unit 6 on SCSI controller 0
On-board serial ports: 2
On-board bi-directional parallel port
Graphics board: Indy 8-bit
Integral Ethernet: ec0, version 1
Integral ISDN: Basic Rate Interface unit 0, revision 1.0
Iris Audio Processor: version A2 revision 4.1.0
Vino video: unit 0, revision 0, IndyCam connected

Thanks,

Jake Griesbach
University of Colorado


On Tue, 9 Mar 1999, Thomas Bogendoerfer wrote:

> On Mon, Mar 08, 1999 at 09:57:29AM -0700, Jake Griesbach wrote:
> > Do you have any ideas as to what may be wrong?  I'm using the
> > hardhat-sgi-5.1.tar.gz file from
> > ftp://ftp.linux.sgi.com:/pub/linux/mips/RedHat.
> 
> Please try a newer kernel. Take the one from
> 
> ftp://ftp.linux.sgi.com/pub/linux/mips/test/vmlinux-indy-2.2.1-990226.tar.gz
> 
> If you have problems with this kernel, please send the output of
> hinv (IRIX command).
> 
> Thomas.
> 
> -- 
>    This device has completely bogus header. Compaq scores again :-|
> It's a host bridge, but it should be called ghost bridge instead ;^)
>                                         [Martin `MJ' Mares on linux-kernel]
> 

Re: Linux/Mips installation

[Thomas Bogendoerfer]

On Tue, Mar 09, 1999 at 02:34:09PM -0700, Jake Griesbach wrote:
> Hmm.  That still didn't work, although the new kernel booted just fine.  I
> still got an Oops message right after I selected the drive to install to.

please mail me the oops, so I could look up, where is happens.

> Could this be a swapon problem?  I also have an internal floppy/floptical

maybe. Did you try to create swap ? AFAIK this doesn't work with HardHat.

> drive (this also appears on the select drive list).  Could this be
> throwing things off?  Does this work for R4600 processors?

it works perfect here.

Thomas.

-- 
   This device has completely bogus header. Compaq scores again :-|
It's a host bridge, but it should be called ghost bridge instead ;^)
                                        [Martin `MJ' Mares on linux-kernel]

Re: Linux/Mips installation

[Jake Griesbach]

On Tue, 9 Mar 1999, Thomas Bogendoerfer wrote:
> please mail me the oops, so I could look up, where is happens.

Here it is:
                            Oops: 0000
$0 : 00000000 1000fc00 00001000 ffffffe0
$4 : 00000020 00000000 1000fc00 8df1fcc0
$8 : 00000010 880ff520 00000001 88152b28
$12: 00000001 00000001 00000001 fffffffc
$16: 00000000 00001000 a83b6000 883b3800
$20: 00000002 bfbc0003 00000000 bfb90000
$24: 1000fc01 80000000
$28: 8875e000 8875fb78 883b7e78 880f167c
epc   : 8801f3f8
Status: 1000fc02
Cause : 00000000
Process runinstall2 (pid: 7, stackpage=8875e000)
Stack: 03200801 00802421 808abd88 808abddd 883b3800 bfbc0003 880f167c 
88101b38
       8df1f4a0 00000074 883bc700 8875fc30 00000078 00000001 883b3800 
bfbc0003
       883b7e78 883b50e0 00000002 8875fcc0 00000000 00000000 00000005
880f1df4
       8df220a8 88086778 8df21074 8807e11c 00000000 880b3b60 883b3800
1000fc00
       883b7e78 880f1a6c 00000000 1000fc01 00000001 1000fc01 883b7e00
883b3800
       08000000 ...
Call Trace: [<880f167c>] [<88101b38>] [<880f1df4>] [<88086778>] (more ...)
Code: 00041823 02032824 00431024 <bcb50000> 14a2fffe  00a42821 40866000
00000000 00000000
install exited abnormally -- recieved signal 11
sending termination signals...done

This happens whether or not I use the disk partitioner via the Edit button
or not.  I'm choosing the third hard disk on my menu which corresponds to
the external drive I have after choosing the install button on the first
menu.

> 
> > Could this be a swapon problem?  I also have an internal floppy/floptical
> 
> maybe. Did you try to create swap ? AFAIK this doesn't work with HardHat.

I'm not using swapon explicitly.  I was just wondering if maybe the
installation program was trying to swapon the raw partition that IRIX
defines on a "root" type partition table (using fx).

Jake Griesbach
University of Colorado

Re: Linux/Mips installation

[Thomas Bogendoerfer]

On Tue, Mar 09, 1999 at 03:43:23PM -0700, Jake Griesbach wrote:
> On Tue, 9 Mar 1999, Thomas Bogendoerfer wrote:
> > please mail me the oops, so I could look up, where is happens.
> 
> Here it is:

thanks. The oops is caused by something going wrong in the SCSI driver.

I saw in your hinv, that there are different devices on your scsi bus.
Could you please try to remove as much as possible from the bus and
try it again ? What devices do have on the bus ?

Thomas.

-- 
   This device has completely bogus header. Compaq scores again :-|
It's a host bridge, but it should be called ghost bridge instead ;^)
                                        [Martin `MJ' Mares on linux-kernel]

Re: Linux/Mips installation

[Jake Griesbach]

That was it!  I removed the CDROM and internal floptical drive, and the
installation proceeded.  However, now I get the prom error when rebooting
that is caused by elf kernels on old proms:

Cannot load bootp()/vmlinux
Illegal f_magic number 0x7f45, expected MIPSELMAGIC or MIPSEBMAGIC.

I tried downloading the ecoff kernel you posted in the test directory, but
I still get the error.  Maybe I haven't installed it correctly, or maybe I
am not typing the correct boot command.

I have unzipped the ecoff kernel and placed it in the installation (remote
computer's) root directory.  (It does boot over NFS into the installation
script.)  I have verified the kernel installed on the local drive after
installation is the same size.  (I simply copied it over to the local
drive.)  Do I have to run something equivalent to lilo??

When rebooting I'm typing:  boot dksc(0,3)

3 is the SCSI id for the local installed drive.

Thanks!!

Jake Griesbach


On Wed, 10 Mar 1999, Thomas Bogendoerfer wrote:

> On Tue, Mar 09, 1999 at 03:43:23PM -0700, Jake Griesbach wrote:
> > On Tue, 9 Mar 1999, Thomas Bogendoerfer wrote:
> > > please mail me the oops, so I could look up, where is happens.
> > 
> > Here it is:
> 
> thanks. The oops is caused by something going wrong in the SCSI driver.
> 
> I saw in your hinv, that there are different devices on your scsi bus.
> Could you please try to remove as much as possible from the bus and
> try it again ? What devices do have on the bus ?
> 
> Thomas.
> 
> -- 
>    This device has completely bogus header. Compaq scores again :-|
> It's a host bridge, but it should be called ghost bridge instead ;^)
>                                         [Martin `MJ' Mares on linux-kernel]
> 

Re: Linux/Mips installation

[Thomas Bogendoerfer]

On Wed, Mar 10, 1999 at 04:02:08PM -0700, Jake Griesbach wrote:
> That was it!  I removed the CDROM and internal floptical drive, and the
> installation proceeded.  However, now I get the prom error when rebooting
> that is caused by elf kernels on old proms:
> 
> Cannot load bootp()/vmlinux
> Illegal f_magic number 0x7f45, expected MIPSELMAGIC or MIPSEBMAGIC.

The kernel worked before the installation and doesn't afterwards ? Strange.

> I tried downloading the ecoff kernel you posted in the test directory, but
> I still get the error.  Maybe I haven't installed it correctly, or maybe I
> am not typing the correct boot command.

Really the same error ? The 0x7f45 is part of the elf magic, which the
ECOFF doesnt contain for sure (at least not at the beginning of the file).

> I have unzipped the ecoff kernel and placed it in the installation (remote
> computer's) root directory.  (It does boot over NFS into the installation

If you boot via tftp, you need to put new kernel in your /tftp directory 
(like you did it with the other kernel).

> script.)  I have verified the kernel installed on the local drive after
> installation is the same size.  (I simply copied it over to the local
> drive.)  Do I have to run something equivalent to lilo??

Right now, you have two choices for booting the kernel. Either use
bootp/tftp or put it on your Irix root partition and boot with 
"boot vmlinux root=/dev/sdc1" (or the approriate /dev/sdxx).

Thomas.

-- 
   This device has completely bogus header. Compaq scores again :-|
It's a host bridge, but it should be called ghost bridge instead ;^)
                                        [Martin `MJ' Mares on linux-kernel]

Re: Linux/Mips installation

[Jake Griesbach]

Thanks, that helped.  I didn't have the kernel installed correctly.  Now
the system boots by itself and runs linux!

I've noticed that the web pages haven't been updated for a while.  Is X
available?  I get the following errors when I try to run X (with startx):

_X11TransSocketUNIXConnect: Can't connect: errno = 146

This repeats until I kill startx.  Also, I'm not familiar with the Red-Hat
linux flavor.  How do I start the package manager to get the latest set of
packages?

Thanks!

Jake Griesbach
University of Colorado




On Thu, 11 Mar 1999, Thomas Bogendoerfer wrote:

> On Wed, Mar 10, 1999 at 04:02:08PM -0700, Jake Griesbach wrote:
> > That was it!  I removed the CDROM and internal floptical drive, and the
> > installation proceeded.  However, now I get the prom error when rebooting
> > that is caused by elf kernels on old proms:
> > 
> > Cannot load bootp()/vmlinux
> > Illegal f_magic number 0x7f45, expected MIPSELMAGIC or MIPSEBMAGIC.
> 
> The kernel worked before the installation and doesn't afterwards ? Strange.
> 
> > I tried downloading the ecoff kernel you posted in the test directory, but
> > I still get the error.  Maybe I haven't installed it correctly, or maybe I
> > am not typing the correct boot command.
> 
> Really the same error ? The 0x7f45 is part of the elf magic, which the
> ECOFF doesnt contain for sure (at least not at the beginning of the file).
> 
> > I have unzipped the ecoff kernel and placed it in the installation (remote
> > computer's) root directory.  (It does boot over NFS into the installation
> 
> If you boot via tftp, you need to put new kernel in your /tftp directory 
> (like you did it with the other kernel).
> 
> > script.)  I have verified the kernel installed on the local drive after
> > installation is the same size.  (I simply copied it over to the local
> > drive.)  Do I have to run something equivalent to lilo??
> 
> Right now, you have two choices for booting the kernel. Either use
> bootp/tftp or put it on your Irix root partition and boot with 
> "boot vmlinux root=/dev/sdc1" (or the approriate /dev/sdxx).
> 
> Thomas.
> 
> -- 
>    This device has completely bogus header. Compaq scores again :-|
> It's a host bridge, but it should be called ghost bridge instead ;^)
>                                         [Martin `MJ' Mares on linux-kernel]
> 

Re: Linux/Mips installation

[ralf]

On Thu, Mar 11, 1999 at 10:44:15AM -0700, Jake Griesbach wrote:

> Thanks, that helped.  I didn't have the kernel installed correctly.  Now
> the system boots by itself and runs linux!
> 
> I've noticed that the web pages haven't been updated for a while.  Is X
> available?  I get the following errors when I try to run X (with startx):
> 
> _X11TransSocketUNIXConnect: Can't connect: errno = 146
> 
> This repeats until I kill startx.  Also, I'm not familiar with the Red-Hat
> linux flavor.  How do I start the package manager to get the latest set of
> packages?

Well, that's ok because it's what is supposed to happen where you're trying
to run X without a server.  But we've got a full set of clients.

  Ralf