Bug#413674: fails to attach/find old screens after update

2007-03-06 Thread Alexander Rydekull
Package: screen
Version: 4.0.3-0.3
Severity: grave
Justification: causes non-serious data loss



-- System Information:
Debian Release: 4.0
  APT prefers unstable
  APT policy: (500, 'unstable')
Architecture: i386 (i686)
Shell:  /bin/sh linked to /bin/bash
Kernel: Linux 2.6.18-3-686
Locale: LANG=en_US.UTF-8, LC_CTYPE=en_US.UTF-8 (charmap=UTF-8)

Versions of packages screen depends on:
ii  base-passwd 3.5.11   Debian base system master password
ii  debconf [debconf-2.0]   1.5.12   Debian configuration management sy
ii  libc6   2.3.6.ds1-13 GNU C Library: Shared libraries
ii  libncursesw55.5-5Shared libraries for terminal hand
ii  libpam0g0.79-4   Pluggable Authentication Modules l
ii  passwd  1:4.0.18.1-7 change and administer password and

screen recommends no packages.

-- debconf information:
  screen/old_upgrade_prompt: false



So, I tripped over this when I arrived at work, connected to my server through 
SSH and tried to get my screen up and 
met a There is no screen to be attached. sign...


Here's some information of things I tried, and various info that might be 
usefull.

I've also tried and find other people with the same problem, I've found atleast 
two where one came into #screen on 
freenode, just like me and supplied this pastebin.

http://pastebin.ca/raw/383517

What I think is the reason is the upgrade from 4.0.3-0.2 to 4.0.3-0.3 that I 
did yesterday.

In my mind, what I suspect will fail it is

start a screen (4.0.3-0.2)
upgrade to screen 4.0.3-0.3
detach
attach (fails)





[EMAIL PROTECTED]:~$ dpkg --status screen
Package: screen
Status: install ok installed
Priority: optional
Section: misc
Installed-Size: 896
Maintainer: Adam Lazur [EMAIL PROTECTED]
Architecture: i386
Version: 4.0.3-0.3
Depends: libc6 (= 2.3.6-6), libncursesw5 (= 5.4-5), libpam0g (= 0.76), 
base-passwd (= 2.0.3.4), passwd (= 
1:4.0.3-10)
Pre-Depends: debconf (= 0.5) | debconf-2.0
Conflicts: suidmanager ( 0.52)
Conffiles:
 /etc/screenrc 0f03c578a71c3af3ed4d92a5557e7a5a
 /etc/init.d/screen-cleanup c0b20a5c34424b377ed33473d6422fb5
Description: a terminal multiplexor with VT100/ANSI terminal emulation
 screen is a terminal multiplexor that runs several separate screens on a
 single physical character-based terminal.  Each virtual terminal emulates a
 DEC VT100 plus several ANSI X3.64 and ISO 2022 functions.  Screen sessions
 can be detached and resumed later on a different terminal.
 .
 Screen also supports a whole slew of other features.  Some of these are:
 configurable input and output translation, serial port support, configurable
 logging, multi-user support, and utf8 charset support.
[EMAIL PROTECTED]:~$ dpkg -l screen
Desired=Unknown/Install/Remove/Purge/Hold
| Status=Not/Installed/Config-files/Unpacked/Failed-config/Half-installed
|/ Err?=(none)/Hold/Reinst-required/X=both-problems (Status,Err: uppercase=bad)
||/ Name  Version   Description
+++-=-=-==
ii  screen4.0.3-0.3 a terminal multiplexor 
with VT100/ANSI terminal emulation
[EMAIL PROTECTED]:~$ 






[EMAIL PROTECTED]:~$ screen -rx -d
There is no screen to be attached.
[EMAIL PROTECTED]:~$ ps -ef | grep rydekull | grep SCREEN
rydekull  3045 1  0 Jan09 ?00:08:15 SCREEN
rydekull  6360  6302  0 10:18 pts/300:00:00 grep SCREEN
[EMAIL PROTECTED]:~$ ls -l /var/run/screen/S-rydekull/
total 0
prw--- 1 rydekull rydekull 0 2007-03-06 09:21 3045.pts-0.munin
[EMAIL PROTECTED]:~$ screen -r 3045.pts-0.munin
There is no screen to be resumed matching 3045.pts-0.munin.
[EMAIL PROTECTED]:~$ screen -r 3045
There is no screen to be resumed matching 3045.
[EMAIL PROTECTED]:~$ screen -d 3045.pts-0.munin
There is no screen to be detached matching 3045.pts-0.munin.
[EMAIL PROTECTED]:~$ screen -ls
No Sockets found in /var/run/screen/S-rydekull.

[EMAIL PROTECTED]:~$ ls -l /var/run/ | grep screen
drwxrwxr-x 14 rootutmp 4096 2006-08-27 16:29 screen
[EMAIL PROTECTED]:~$ ps -ef | grep rydekull
rydekull  3045 1  0 Jan09 ?00:08:15 SCREEN
rydekull 30964  3045  0 Feb13 pts/000:00:00 /bin/bash
rydekull  9395 30964  1 Mar01 pts/001:22:46 centericq
[EMAIL PROTECTED]:~$ kill -9 30964
[EMAIL PROTECTED]:~$ ps -ef | grep 30964
rydekull  7898  7615  0 10:48 pts/27   00:00:00 grep 30964
[EMAIL PROTECTED]:~$ screen -rx -d
There is no screen to be attached.
[EMAIL PROTECTED]:~$ 





Perhaps it's just me, but I think this should work quite alright?

Anyone got a clue?



found something funny aswell

[EMAIL PROTECTED]:~$ strace -f -o strace-screen.txt screen -rx -d
Directory '/var/run/screen' must have mode 777.
[EMAIL PROTECTED]:~$ su -
Password: 
[EMAIL PROTECTED]:~# chmod 777 /var/run/screen/
[EMAIL PROTECTED]:~# logout
[EMAIL PROTECTED]:~$ strace 

Bug#413674: fails to attach/find old screens after update

2007-03-06 Thread Christian Perrier
Quoting Alexander Rydekull ([EMAIL PROTECTED]):
 Package: screen
 Version: 4.0.3-0.3
 Severity: grave
 Justification: causes non-serious data loss
 


I got the following provate mail from another user:

==
I'm writing to you about today's (well, Monday's) NMU on the 'screen'
package, version 4.0.3-0.3. It seems to me that on my several systems
(i386 running debian sid), this new version switches from the use of
FIFOs to local sockets in /var/run/screen/S-user. Thus, after the
upgrade, already ready running screens are not recognized anymore
(and cannot be reattached). What a big deal! I want my screen back! :)

Well, according to the changelog, being a l10n NMU, I'm not sure this
waes really intented, was it?
==

Actually, I fail to see why this could have been broken by a l10n NMU
but this could maybe be a side effect of the triggered recompile
(which would mean there's a bad underlying problem).

The build I did was done in a supposedly clean chroot, with
pbuilderwhat I ignore is how was the previous builds done.




signature.asc
Description: Digital signature


Bug#413674: fails to attach/find old screens after update

2007-03-06 Thread Alexander Rydekull

Additional information, something that my friend noticed.

The file is

prw--- 1 rydekull rydekull 0 2007-03-06 09:21 3045.pts-0.munin

while it should be

srw--- 1 rydekull rydekull 0 2007-03-06 09:21 3045.pts-0.munin

In other words, it should be a socket, but it isnt. Why isnt it?

/Alexander Rydekull


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



Processed: Re: Bug#413674: fails to attach/find old screens after update

2007-03-06 Thread Debian Bug Tracking System
Processing commands for [EMAIL PROTECTED]:

 severity 413674 important
Bug#413674: fails to attach/find old screens after update
Severity set to `important' from `grave'

 thanks
Stopping processing here.

Please contact me if you need assistance.

Debian bug tracking system administrator
(administrator, Debian Bugs database)


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



Bug#413674: fails to attach/find old screens after update

2007-03-06 Thread Steve Langasek
severity 413674 important
thanks

On Tue, Mar 06, 2007 at 03:21:57PM +0100, Alexander Rydekull wrote:
 So, I tripped over this when I arrived at work, connected to my server 
 through SSH and tried to get my screen up and 
 met a There is no screen to be attached. sign...

 Here's some information of things I tried, and various info that might be 
 usefull.

 I've also tried and find other people with the same problem, I've found 
 atleast two where one came into #screen on 
 freenode, just like me and supplied this pastebin.

 http://pastebin.ca/raw/383517

 What I think is the reason is the upgrade from 4.0.3-0.2 to 4.0.3-0.3 that I 
 did yesterday.

 In my mind, what I suspect will fail it is

 start a screen (4.0.3-0.2)
 upgrade to screen 4.0.3-0.3
 detach
 attach (fails)

I can confirm this bug; I can further confirm that the problem is specific
to the i386 build of 4.0.3-0.3, and that a rebuild of the package in my
environment fixes the problem.

I don't know what's particular about Christian's build environment that
would cause this problem, but I've triggered a binNMU on i386 so that we can
get this fixed in the near term.

There is still a bug here certainly, because misbuilds of this sort should
be protected against; but there's no longer any reason we should need to
treat this as release-critical.

Thanks,
-- 
Steve Langasek   Give me a lever long enough and a Free OS
Debian Developer   to set it on, and I can move the world.
[EMAIL PROTECTED]   http://www.debian.org/


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