Trying to boot an Indy

Trying to boot an Indy

[Dave Gilbert]

Hi,
  I'm trying to boot my Indy into Linux for the first time and not having
any luck.

1) I tried bootp - bootp()vmlinux - it says 'no server for vmlinux'.  The
bootp server is a Linux/Alpha box running 2.4.0-ac9 - I've already done
the trick with no_pmtu.  tcpdump shows bootp sending a packet with
apparently the correct mac address.

2) So I ftp'd the file over under Irix, gzip -d'd it and then rebooted and
from sash did boot indyvmlinux - this immediatly after starting gave
'Exception: <vector=UTLB Miss>' plus details (available on request).
(This is kernel vmlinux=001027-test9-r4x00.gz off
download.ichilton.co.uk/pub/ichilton/linux-mips/kernels).

This is an Indy R4600PC with 64MB RAM.

Ideas appreciated.

Dave

P.S. I've got a lead wired here to do the 13w3 but it doesn't seem to be
spot on - I don't get any green, but it is sync'd - this is going to an
Iiyama 8617 (original 17") monitor. Ideas on that would also be
appreciated.

-- 
 ---------------- Have a happy GNU millennium! ----------------------   
/ Dr. David Alan Gilbert      | Running GNU/Linux on m68k, |  Happy  \ 
\   gro.gilbert @ treblig.org |  Alpha, x86, ARM and SPARC |  In Hex /
 \ ___________________________|___ http://www.treblig.org  |________/

Re: Trying to boot an Indy

[Ralf Baechle]

On Sun, Jan 21, 2001 at 02:01:15AM +0000, Dave Gilbert wrote:

> 2) So I ftp'd the file over under Irix, gzip -d'd it and then rebooted and
> from sash did boot indyvmlinux - this immediatly after starting gave
> 'Exception: <vector=UTLB Miss>' plus details (available on request).
> (This is kernel vmlinux=001027-test9-r4x00.gz off
> download.ichilton.co.uk/pub/ichilton/linux-mips/kernels).
> 
> This is an Indy R4600PC with 64MB RAM.

You may have become a victim of a sash bug which makes the firmware report
used memory as free memory and in the end results in the kernel overwriting
itself.  The solution is easy; don't use sash.  This means you either
have to boot the kernel via ``bootp() -f ...'' from the network or you
use IRIX's dvhtool to write it into the volume header, then boot it from
the kernel prompt by just entering the kernel name.

  Ralf

Re: Trying to boot an Indy

[Dave Gilbert]

On Sun, 21 Jan 2001, Ralf Baechle wrote:

> On Sun, Jan 21, 2001 at 02:01:15AM +0000, Dave Gilbert wrote:
> 
> > 2) So I ftp'd the file over under Irix, gzip -d'd it and then rebooted and
> > from sash did boot indyvmlinux - this immediatly after starting gave
> > 'Exception: <vector=UTLB Miss>' plus details (available on request).
> > (This is kernel vmlinux=001027-test9-r4x00.gz off
> > download.ichilton.co.uk/pub/ichilton/linux-mips/kernels).
> > 
> > This is an Indy R4600PC with 64MB RAM.
> 
> You may have become a victim of a sash bug which makes the firmware report
> used memory as free memory and in the end results in the kernel overwriting
> itself.  The solution is easy; don't use sash.  This means you either
> have to boot the kernel via ``bootp() -f ...'' from the network or you
> use IRIX's dvhtool to write it into the volume header, then boot it from
> the kernel prompt by just entering the kernel name.

Well the bootp() stuff still won't respond so that is out, and dvhtool is
saying there isn't enough room for my linux image. Hmph.

Running bootpc under Irix seems to give sensible results.

Is there any other mechanism for loading other than bootp ?

Dave

-- 
 ---------------- Have a happy GNU millennium! ----------------------   
/ Dr. David Alan Gilbert      | Running GNU/Linux on m68k, |  Happy  \ 
\   gro.gilbert @ treblig.org |  Alpha, x86, ARM and SPARC |  In Hex /
 \ ___________________________|___ http://www.treblig.org  |________/

Re: Trying to boot an Indy

[Christoph Martin]

Dave Gilbert <gilbertd@treblig.org> writes:

> 
> Hi,
>   I'm trying to boot my Indy into Linux for the first time and not having
> any luck.
> 
> 1) I tried bootp - bootp()vmlinux - it says 'no server for vmlinux'.  The
> bootp server is a Linux/Alpha box running 2.4.0-ac9 - I've already done
> the trick with no_pmtu.  tcpdump shows bootp sending a packet with
> apparently the correct mac address.
> 

I have the same problem serving bootp from my i386 2.4.0 box. bootp
with kernel 2.2.x on the same box works. And it is only the bootp from
the command console that is failing. the bootp part later on in the
kernel is working from the 2.4.0 box.

Weird.

-- 
============================================================================
Christoph Martin, Uni-Mainz, Germany
 Internet-Mail:  Christoph.Martin@Uni-Mainz.DE
--------------export-a-crypto-system-sig -RSA-3-lines-PERL------------------
#!/usr/bin/perl -sp0777i<X+d*lMLa^*lN%0]dsXx++lMlN/dsM0<j]dsj
$/=unpack('H*',$_);$_=`echo 16dio\U$k"SK$/SM$n\EsN0p[lN*1
lK[d2%Sa2/d0$^Ixp"|dc`;s/\W//g;$_=pack('H*',/((..)*)$/)
#what's this? see http://www.dcs.ex.ac.uk/~aba/rsa/

Re: Trying to boot an Indy

[Dr. David Gilbert]

On 23 Jan 2001, Christoph Martin wrote:

> Dave Gilbert <gilbertd@treblig.org> writes:

> > 1) I tried bootp - bootp()vmlinux - it says 'no server for vmlinux'.  The
> > bootp server is a Linux/Alpha box running 2.4.0-ac9 - I've already done
> > the trick with no_pmtu.  tcpdump shows bootp sending a packet with
> > apparently the correct mac address.
> >
>
> I have the same problem serving bootp from my i386 2.4.0 box. bootp
> with kernel 2.2.x on the same box works. And it is only the bootp from
> the command console that is failing. the bootp part later on in the
> kernel is working from the 2.4.0 box.

This sprang into life when I added the 'sa' entry in the bootp file - that
is the server address.

Dave

-- 
/------------------------------------------------------------------\
| Dr. David Alan Gilbert | Work:dg@px.uk.com +44-161-286-2000 Ex258|
| -------- G7FHJ --------|---------------------------------------- |
| Home: dave@treblig.org            http://www.treblig.org         |
\------------------------------------------------------------------/

Re: Trying to boot an Indy

[Dr. David Gilbert]

On 23 Jan 2001, Christoph Martin wrote:

> Dave Gilbert <gilbertd@treblig.org> writes:

> > 1) I tried bootp - bootp()vmlinux - it says 'no server for vmlinux'.  The
> > bootp server is a Linux/Alpha box running 2.4.0-ac9 - I've already done
> > the trick with no_pmtu.  tcpdump shows bootp sending a packet with
> > apparently the correct mac address.
> >
>
> I have the same problem serving bootp from my i386 2.4.0 box. bootp
> with kernel 2.2.x on the same box works. And it is only the bootp from
> the command console that is failing. the bootp part later on in the
> kernel is working from the 2.4.0 box.

This sprang into life when I added the 'sa' entry in the bootp file - that
is the server address.

Dave

-- 
/------------------------------------------------------------------\
| Dr. David Alan Gilbert | Work:dg@px.uk.com +44-161-286-2000 Ex258|
| -------- G7FHJ --------|---------------------------------------- |
| Home: dave@treblig.org            http://www.treblig.org         |
\------------------------------------------------------------------/

Re: Trying to boot an Indy

[Wills]

  I am also suddenly without success:

> > > 1) I tried bootp - bootp()vmlinux - it says 'no server for vmlinux'.  The
> > > bootp server is a Linux/Alpha box running 2.4.0-ac9 - I've already done
> > > the trick with no_pmtu.  tcpdump shows bootp sending a packet with
> > > apparently the correct mac address.
> >
> > I have the same problem serving bootp from my i386 2.4.0 box. bootp
> > with kernel 2.2.x on the same box works. And it is only the bootp from
> > the command console that is failing. the bootp part later on in the
> > kernel is working from the 2.4.0 box.
> 
> This sprang into life when I added the 'sa' entry in the bootp file - that
> is the server address.

  I have a new set of IP addresses, new (2.2.18) kernel, new nfs
server utilities (all together, unfortunately) since my Indy last
spoke sanely to my intel linux box; neither the 'sa' bootptab entry
nor /proc/sys/net/ipv4/ip_no_pmtu_disc content tips have any evident
effect - ie. I will see "Setting $netaddr...", but the "obtaining
[...]" ticker no longer appears. I am eventually told, variously,
"netaddr may be set incorrectly or network too busy" or "unable to
execute bootp():vmlinux"



:(
	Wills.

Re: Patch for Philips Nino...

[Steven J. Hill]

"Steven J. Hill" wrote:
> 
> I would like to submit a patch for the Philips Nino. It can be
> gotten from here (ftp://ftp.cotw.com/pub/nino/mips-tools/).
> I didn't want to waste bandwidth by attaching it. It overlaps
> some of Florian's stuff. Criticisms very welcome. Thanks.
> 
And then he realizes that in his sleepiness, he attached it
anyway. Sorry everyone.

-Steve

-- 
 Steven J. Hill - Embedded SW Engineer
 Public Key: 'finger sjhill@mail.cotw.com'
 FPR1: E124 6E1C AF8E 7802 A815
 FPR2: 7D72 829C 3386 4C4A E17D

Re: Trying to boot an Indy

[Ralf Baechle]

On Tue, Jan 23, 2001 at 05:11:44PM +0100, Christoph Martin wrote:

> > 1) I tried bootp - bootp()vmlinux - it says 'no server for vmlinux'.  The
> > bootp server is a Linux/Alpha box running 2.4.0-ac9 - I've already done
> > the trick with no_pmtu.  tcpdump shows bootp sending a packet with
> > apparently the correct mac address.
> > 
> 
> I have the same problem serving bootp from my i386 2.4.0 box. bootp
> with kernel 2.2.x on the same box works. And it is only the bootp from
> the command console that is failing. the bootp part later on in the
> kernel is working from the 2.4.0 box.
> 
> Weird.

Not weired at all.  The firmware bootp()... does bootp and then uses tftp
to download the kernel; the kernel then only uses bootp to figure out it's
network configuration.  So the two are not only doing different things,
they're also two are not only two independant and completly different
implementations.

  Ralf