FW: potato - telnetd can't get pty?

2000-06-14 Thread Schulz, Martin

... there are no stupid questions:
strace of in.telnetd showed it accesses /dev/ptmx, and my SuSE i386
Linux in.telnetd doesn't
access /dev/ptmx.

crw-r-   1 root tty5,   2 Jan 19  1999 /dev/ptmx

That doesn't work with telnetd.telnetd access rights!

What's wrong here, in.telnetd or the device's initial access?
(everything's rocking with 777 for /dev/ptmx)

Thnx.

> Maybe this is a stupid question, but what's wrong with the following:
> 
> bash-2.03# telnet osolemio
> Trying 192.168.100.104...
> Connected to osolemio.
> Escape character is '^]'.
> telnetd: getpty: Permission denied
> .
> Connection closed by foreign host.
> 
> (I've also tried running manually /etc/init.d/devpts.sh)
> 
> This is a brand new potato install.
> 
> --
> Martin Schulz



Re: Sparc ELC BWTWO device

2000-06-14 Thread Schulz, Martin
...hmm, I still think I need to sort out /dev/fb#, but what's bwtwo to
look like?

/var/log/dmesg looks acceptable:
 Console: switching to mono frame buffer device 142x54
 fb0: bwtwo at 3.fe00
the Debian logo is still garbled.

For what it's worth, XsunMono dies after a read() as follows:
Program received signal SIGSEGV, Segmentation fault.
0xd2f64 in miIntersect ()
(gdb) backtrace
#0  0xd2f64 in miIntersect ()
#1  0xa96e4 in mfbCopyArea ()
#2  0xafa90 in mfbPutImage ()
#3  0xd992c in miDCInitialize ()
#4  0xd9c58 in miDCInitialize ()
#5  0xe6a18 in miSpriteInitialize ()
#6  0xe68cc in miSpriteInitialize ()
#7  0xd92a4 in miPointerUpdate ()
#8  0xd8f54 in miPointerInitialize ()
#9  0x3dd6c in DefineInitialRootWindow ()
#10 0x4573c in main ()
#11 0xe00bc46c in __libc_start_main () from /lib/libc.so.6  

Anybody seen something like that?

Martin

> what's the incantation for the BW framebuffer of the ELC?
> 
> I'm getting a garbled 'logo' on bootup. The Debian logo was fine when
tftp'ing the 2.2.1 slink kernel.
> 
> Also X dies, but the screen I'm seeing seems also to be garbled, like it
was grayscale instead of
> BW.
> 
> I'm using  a newly installed potato with the 2.2.15 kernel .  The adequate
X server is XsunMono
> I hope.
> 
> Is there anything going awry with the device detection on bootup?
> How can I check?
> 
> Martin
>


--  
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact
[EMAIL PROTECTED]



potato - telnetd can't get pty?

2000-06-12 Thread Schulz, Martin

Maybe this is a stupid question, but what's wrong with the following:

bash-2.03# telnet osolemio
Trying 192.168.100.104...
Connected to osolemio.
Escape character is '^]'.
telnetd: getpty: Permission denied
.
Connection closed by foreign host. 

(I've also tried running manually /etc/init.d/devpts.sh)

This is a brand new potato install.

-- 
Martin Schulz



Re: Sparc ELC w limited resources - trouble installing

2000-06-08 Thread Schulz, Martin
OK,

I got it going with the install root image written to the swap partition.
Booted nicely over tftp.
Then switching to the (preinstalled) base partition and going through the
movements.
No problems, even got the kernel/drivers from the net, using the IP
addresses.
The penguin is garbled in my BW FB.

Prereqs: bootp server, tftp server, rarp entry.

Version installed: May 28
(http://auric.debian.org/~bcollins/disks-sparc/current/)

I'll go through the dselect process next.

I couldn't get the installation going over NFS, though.  I've got knfsd up
and running and
can mount between intel Linux machines without a hitch.  When the kernel
tries to boot
it doesn't get a reply from the portmapper and falls back to the default
port (which is?),
Failing finally with '-5'.

I can't see why my (intel linux) portmapper wouldn't answer, or why the
whole process aborts
at the end.  It is, however, clear indication that the kernels I'd tried
before were not NFS
enabled.

Any comments on what the problem might be with the NFS boot?  (Not that I
need that now,
but to satisfy my curiosity!).  Once installed, I'll also try to NFS mount
the same drive normally.

Martin



Sparc ELC w limited resources - trouble installing

2000-06-07 Thread Schulz, Martin
Hi,

 I have a working ELC, after replacing the NVRAM... it did boot SunOS 4 last
time I tried.
I'm having some trouble figuring out the (best) way to install Linux on this
machine.

I've studied the available install doc from Debian, UltraSparc and RedHat to
no total enlightenment.
I've tried several methods to not much success so far.

Configuration:
12 MB RAM
Ethernet and a fully functional 2.2.15 PII connected (rarp and tftp set up
and working)
650 MB Hard disk I can either use on the PII Linux box or the ELC.

I thought that should do it - but not much luck.

My equipment precludes me from using either CD or Floppy, apparently the
preferred install
media. That together with little RAM seems to block the more treaden paths
(rescue disks etc.).

TFTP:
I know my memory is a little tight and none of the tftp boots have ever
worked. Should they have?
One thing I noticed that my machine always seems to hang after the load
unless I pad
the file to a multiple of a block size.  That might be worth noting or even
add to the scripts for producing these thingies we tftp over.
The doc on 'low memory TFTP install' is a complete mystery to me - see NFS
questions below.

NFS:
I also tried to boot with an NFS root, but the the kernel never managed to
mount it:  
"cant mount root on 00:00\nPress L1 A to reboot".
The docs are not clear what the correct syntax is for an NFS mounted root...
that shouldn't be hard to add.  I tried root=/dev/nfs
nfsroot=192.168.100.100:/the/root/path
Obviously that depends on the kernel to have support built in for it, but
how can I tell if the .config
isn't provided with the kernel? (Debian linux-2.2.1.a.out or so)

HD:
I managed to partition and setup the HD with the base2_1.tar.gz content.
That booted great,
but simply told me to stick in the boot floppy, then confiure the system.
Stuck.
Why is it so hard to set up a minimal system from Linux box of a different
architecture I don't understand.

If left to my own resources,  I'll try the same stuff again with a newer
potato install kit I've learned
about on the list today.  I'll also try to abuse my swap partition on the HD
as a 'root'.
That should be feasible for Debian, but RedHat wanted all their RPMS in
there, taking too much
space.

Am I missing something important?

Bonus question:
Also, I can't claim I've fully understood the full detailed mechanics of an
installation. Wha'ts the
main thing to kick off after a minimal RAM or disk based system is up and
running and why can't
it be on the same partition as the system being installed?

Your help is very much appreciated...

Martin