new to list - boot problem

new to list - boot problem

[Miles Lott]

I have an Indy R5000, so says the case anyway.  I was able to setup
bootp/tftp/nfs.  I can mount the nfs dir from within IRIX.  I tried
booting from the nfs partition using bootp, and during the boot
message it seems to identify the machine and hardware ok.  Then I get
the "unable to open initial console" message( read: hang ).

I'm sure you have seen this before.  Have I buggered the unpacking
of HardHat?

TIA

Re: new to list - boot problem

[Robin Humble]

>the "unable to open initial console" message( read: hang ).

There are a few things that this can be, but the most obvious is to
check that the dev/ directory in the HardHat distribution looks like:

crw-------   1 root     root       4,   0 Sep 14 19:39 console
crw-rw-rw-   1 root     root       1,   3 Sep 14 19:43 null
brw-r-----   1 root     disk       1,   1 Sep 14 19:43 ram
crw-------   1 root     root       4,   0 Sep 14 19:44 systty
crw-------   1 root     root       4,   1 Sep 14 19:44 tty1
crw-------   1 root     root       4,   2 Sep 14 19:44 tty2
crw-------   1 root     root       4,   3 Sep 14 19:44 tty3
crw-------   1 root     root       4,   4 Sep 14 19:44 tty4
crw-------   1 root     root       4,   5 Sep 14 19:44 tty5

If not, then nuke the devices and remake them with mknod, chown, chmod.
Some tar programs (IRIX's in particular) screws these up. You can
re-make these devices under Linux/x86 or IRIX - both work fine. If
re-making from under IRIX then 'disk' gid = 6, and 'root' gid = 0 = sys
under IRIX.

Other possibilities are that you're running on hardware not well
supported by HardHat (new kernels are available) or the nfsroot
procedure isn't working well enough. I'm sure other people will have
other failure modes to tell you as well :)

cheers,
robin
+
He'd found that even the people whose job of work was, so to speak, the
Universe, didn't really believe in it and were actually quite proud of not
knowing what it really was or even if it could theoretically exist.
     Robin Humble     /      http://www.maths.monash.edu.au/~rjh/

Re: new to list - boot problem

[Miles Lott]

Thanks, the devs were in fact screwed up.  I will try again when I get
into work today.

Thanks for the quick reply :)

On Sat, 27 Feb 1999, Robin Humble wrote:

> 
> >the "unable to open initial console" message( read: hang ).
> 
> There are a few things that this can be, but the most obvious is to
> check that the dev/ directory in the HardHat distribution looks like:
> 
> crw-------   1 root     root       4,   0 Sep 14 19:39 console
> crw-rw-rw-   1 root     root       1,   3 Sep 14 19:43 null
> brw-r-----   1 root     disk       1,   1 Sep 14 19:43 ram
> crw-------   1 root     root       4,   0 Sep 14 19:44 systty
> crw-------   1 root     root       4,   1 Sep 14 19:44 tty1
> crw-------   1 root     root       4,   2 Sep 14 19:44 tty2
> crw-------   1 root     root       4,   3 Sep 14 19:44 tty3
> crw-------   1 root     root       4,   4 Sep 14 19:44 tty4
> crw-------   1 root     root       4,   5 Sep 14 19:44 tty5
> 
> If not, then nuke the devices and remake them with mknod, chown, chmod.
> Some tar programs (IRIX's in particular) screws these up. You can
> re-make these devices under Linux/x86 or IRIX - both work fine. If
> re-making from under IRIX then 'disk' gid = 6, and 'root' gid = 0 = sys
> under IRIX.
> 
> Other possibilities are that you're running on hardware not well
> supported by HardHat (new kernels are available) or the nfsroot
> procedure isn't working well enough. I'm sure other people will have
> other failure modes to tell you as well :)
> 
> cheers,
> robin
> +
> He'd found that even the people whose job of work was, so to speak, the
> Universe, didn't really believe in it and were actually quite proud of not
> knowing what it really was or even if it could theoretically exist.
>      Robin Humble     /      http://www.maths.monash.edu.au/~rjh/
> 

Re: new to list - boot problem

[Miles Lott]

OK, after fixing the devs, I no longer get the error message.
However, after loading init and ld.so.2.0.6 from the server, the Indy
hangs.

something simple must be missing... ;)

On Sat, 27 Feb 1999, Robin Humble wrote:

> 
> >the "unable to open initial console" message( read: hang ).
> 
> There are a few things that this can be, but the most obvious is to
> check that the dev/ directory in the HardHat distribution looks like:
> 
> crw-------   1 root     root       4,   0 Sep 14 19:39 console
> crw-rw-rw-   1 root     root       1,   3 Sep 14 19:43 null
> brw-r-----   1 root     disk       1,   1 Sep 14 19:43 ram
> crw-------   1 root     root       4,   0 Sep 14 19:44 systty
> crw-------   1 root     root       4,   1 Sep 14 19:44 tty1
> crw-------   1 root     root       4,   2 Sep 14 19:44 tty2
> crw-------   1 root     root       4,   3 Sep 14 19:44 tty3
> crw-------   1 root     root       4,   4 Sep 14 19:44 tty4
> crw-------   1 root     root       4,   5 Sep 14 19:44 tty5
> 
> If not, then nuke the devices and remake them with mknod, chown, chmod.
> Some tar programs (IRIX's in particular) screws these up. You can
> re-make these devices under Linux/x86 or IRIX - both work fine. If
> re-making from under IRIX then 'disk' gid = 6, and 'root' gid = 0 = sys
> under IRIX.
> 
> Other possibilities are that you're running on hardware not well
> supported by HardHat (new kernels are available) or the nfsroot
> procedure isn't working well enough. I'm sure other people will have
> other failure modes to tell you as well :)
> 
> cheers,
> robin
> +
> He'd found that even the people whose job of work was, so to speak, the
> Universe, didn't really believe in it and were actually quite proud of not
> knowing what it really was or even if it could theoretically exist.
>      Robin Humble     /      http://www.maths.monash.edu.au/~rjh/
>