X only as root

2000-12-08 Thread Astrogeek
I have been using XF86v4 for a while with no trouble.  Recently however, after
exiting an X session as a user (working fine) and trying to restart it as the
same user I get:

> X: user not authorized to run the X server, aborting.

All other users except root cannot access X.  Root gets in fine.
/etc/X11/Xserver is set like this:

> /usr/bin/X11/XFree86
> Console

I also tried changing Console to Anybody, with no change.  Before exiting X I
had synced Woody with apt-get dist-upgrade, but I don't think it's an upgrade
problem.  I think I screwed it up.  I did an xhost +hostname while in X earlier
as a user trying to figure out why a program that was suid wouldn't run, but
that wouldn't do this would it?  What else can I check?

Andrew W.
-- 
[EMAIL PROTECTED] -- http://home.cwru.edu/~agw4/



PPPoE misbehaving

2001-01-15 Thread astrogeek
I recently had ADSL service installed, and I have had some strange problems 
with it connecting to certain WWW sites.  For example: http://tempo.gsu.edu, 
http://www.mbnanetaccess.com, http://www.aol.com, http://http.us.debian.org(!). 
 Sending email has also stopped working.

Almost everything else works fine.  What happens is the host is looked up, 
contacted, I receive a reply, then it stops at "transferring data".  The email 
messages all time out.  I can receive email directly, but not send it using 
exim now.  

I'm using the pppoe package from Potato.  The machine in question does ipmasq 
for a couple of Linux machines, these sites (and a few others) fail on every 
single machine including the one connected directly to the internet.  What I 
find to be really strange is that these sites work when I fall back to using 
the old dialup on the same machine, including on the ipmasq'ed machines.  So I 
figure I must have something set up wrong, but I'm at a loss as to what it 
might be.  I tried setting the MTU and MRU to 1500 with the same results (and a 
slight drop in transfer speed).

Here's what the net looks like from the ipmasqing machine (lo left out):

***
eth0  Link encap:Ethernet  HWaddr 00:50:BA:DE:3D:3D  
  UP BROADCAST RUNNING MULTICAST  MTU:1500  Metric:1
  RX packets:6659235 errors:0 dropped:133 overruns:0 frame:18
  TX packets:3829180 errors:0 dropped:0 overruns:0 carrier:0
  collisions:759 txqueuelen:100 
  Interrupt:10 Base address:0xfee0 

eth1  Link encap:Ethernet  HWaddr 00:A0:CC:D2:2D:B3  
  inet addr:192.168.1.1  Bcast:192.168.1.255  Mask:255.255.255.0
  UP BROADCAST RUNNING MULTICAST  MTU:1500  Metric:1
  RX packets:3818513 errors:0 dropped:0 overruns:0 frame:181
  TX packets:6638415 errors:145 dropped:0 overruns:4 carrier:141
  collisions:430178 txqueuelen:100 
  Interrupt:9 Base address:0xff00 

ppp0  Link encap:Point-to-Point Protocol  
  inet addr:remote.ip  P-t-P:207.69.62.1 Mask:255.255.255.255
  UP POINTOPOINT RUNNING NOARP MULTICAST  MTU:1500  Metric:1
  RX packets:2206279 errors:0 dropped:0 overruns:0 frame:0
  TX packets:1346162 errors:0 dropped:0 overruns:0 carrier:0
  collisions:0 txqueuelen:10 
***

Thanks for any help, this has had me confused for days.  I'd like to try 2.4's 
kernel-based PPPoE with the hope that it's the software doing this, but not 
being familiar with netfilter is stopping me at the moment.

Andrew W.



re: PPPoE misbehaving

2001-01-15 Thread astrogeek
Thanks list, the problem is solved.  I built the newest rp-pppoe and this made 
everything work for me again, so I can let the 2.4.0 kernel wait until I'm 
comfortable with netfilter.

Andrew W.



Can only open one xterm

2001-06-20 Thread Astrogeek
I have a problem which is probably easy to solve, but I'm having a hard time
with it.  The problem is that I can only open one xterm/aterm/etc at a time
under X.  If I try to open any more as a user (in group tty) it reports that
I can't get permission to do that.  My permissions seem to be set ok, and even
setting the relevant /dev entries 777 doesn't let it work.  Here's some text
I shunted into a file from my single term window:

**BEGIN**
crw-rw-rw-1 root tty5,   2 Jun 20 19:16 ptmx

/dev/pts/*
crw--w1 dude dude 136,   0 Jun 20 19:17 0

[EMAIL PROTECTED]:~$ aterm
aterm: can't open slave tty /dev/pts/1
[EMAIL PROTECTED]:~$ xterm
xterm: Error 1, errno 13: Permission denied
**END**

The system is an unstable Debian system using 2.4.5-ac16 and devfs/devfsd 
without the compatability names enabled.  I can open one terminal window, but
no more.  splitvt also fails, I tried that   If I delete .Xauthority and
su to root, root can open however many xterms as desired.  This is a new 
installation (wanted to repartition the old).  Restarting devfsd 
w/compatability names produces no change.

I'd love to get some help fixing this as I do most of my work in xterms.  TIA 
for any help!  This one has me puzzled (and using the console ).
-- 
Andrew W. <[EMAIL PROTECTED]> -- http://failsure.net/
http://home.cwru.edu/~agw4/ -- Debian GNU/Linux
Georgia State U. CS/Networking UG -- VW bus driver



Re: NVidia RivaTNT2 M64

2001-06-21 Thread Astrogeek
On Wed, Jun 20, 2001 at 09:06:58PM -0700, Reza wrote:
> Hi there.. 
> I just wondering if I was right or wrong, I'm using
> xserver-svga for my XWindows on my NVidia Riva TNT2

I'm about positive that you must be using xserver-xfree86 
v4.0.x for the driver to work.  I don't think it works with
xserver-svga, at least not the GL-accelerated drivers that you
are probably referring to from Nvidia.
-- 
Andrew W. <[EMAIL PROTECTED]> -- http://failsure.net/
http://home.cwru.edu/~agw4/ -- Debian GNU/Linux
Georgia State U. CS/Networking UG -- VW bus driver



Mount is halfway broken

2001-05-10 Thread Astrogeek
On my unstable box, "mount -a" segfaults.  It gives the following error (lots
of stuff snipped):

<1>Unable to handle kernel NULL pointer dereference at virtual address[...]
[...]
pmd entry cbb49000: 
... pmd not present!
[...]

I woke up this morning and everything was unmounted.  I actually found that 
kind of odd :-)  Windows instincts somehow kicked in and I rebooted.  And mount
segfaulted giving me only the NFS mounts, /, /proc and /dev.  I have to
manually mount several partitions.  dmesg also segfaults, and I can't see if
anything else is going wrong as (since /var is on a separate partition) I have
no kernel log!  I've been messing with it for a couple hours now, I have:

1) Installed mount/util-linux from testing and stable and compiled from source
2) Downgraded libc6 to testing and back again, no change
3) Removed and tested each stick of RAM in the machine, and some add-on cards
4) Memtest came up with no errors on five runs
5) Tried different kernel versions (all 2.4.x, running 2.4.4-ac6)
6) strace mount -a doesn't show anything obvious (to me) except at the end
+++ killed by SIGSEGV +++

Mounting each partition individually, as stated above, works for some reason.

Before I reinstall, as this is really difficult for me to troubleshoot, anybody
more experienced have any suggestions?  Nothing except mount and dmesg appear
to be having trouble.

The machines running stable are all yawning ;-)  I hope it's not bad hardware.
I be broke.
-- 
Andrew W. <[EMAIL PROTECTED]> -- http://failsure.net/
http://home.cwru.edu/~agw4/ -- Debian GNU/Linux
Georgia State U. CS/Networking UG -- VW bus driver



Halfway broken mount halfway solved

2001-05-14 Thread Astrogeek
First, a great big "Thanks!" to Karsten (who has helped me before under a
different identity long ago) for getting the information about the floppy
onto the list.  I was going to skip removing the floppy from fstab since
I figured it shouldn't make a difference, seeing how it was hard-disk 
partitions failing to mount.  But that fixed it!  Oddly, the following also
produces a segfault:

ls /dev/fd*

For the curious, I have migrated to devfs, and the user with the same problem
(sorry I lost the message) seems to have also been using devfs.  I use floppies
rarely enough that it is a minor nuisance commenting them out.  Any attempt
to use the floppy also segfaults, so it could still be hardware I guess.  It'd
be slightly a PITA to try without devfs as I've modified several files to
migrate more or less completely over to it.  I've been using devfs for a while
and this is the first problem I've had with it.

I don't think I should post my fstab or error logs since this seems a devfs or
hardware problem, but I'd be happy to share via p-mail.

Basically just wanted to say thanks, it's pretty much back to normal :-)
-- 
Andrew W. <[EMAIL PROTECTED]> -- http://failsure.net/
http://home.cwru.edu/~agw4/ -- Debian GNU/Linux
Georgia State U. CS/Networking UG -- VW bus driver