Bug#170635: libc6 2.3.1-3 to 2.3.1-5 upgrade breaks

2002-12-02 Thread Magnus Danielson
From: Anthony DeRobertis <[EMAIL PROTECTED]>
Subject: Re: Bug#170635: libc6 2.3.1-3 to 2.3.1-5 upgrade breaks
Date: 02 Dec 2002 01:17:27 -0500

> On Sun, 2002-12-01 at 21:19, Magnus Danielson wrote:
> 
> > rm /lib/ld-2.3.1.so
> > 
> > I got the "Device or resource busy" message as a reply. It's because the 
> > file
> > is in use by many tools:
> 
> That shouldn't happen. "File in use" != "name in use". rm removes names,
> not files.

Exactly! I know... it's even traditional UNIX AFAIK.

> And you using something weird as your root file system? That is
> something besides ext[23] or reiserfs?

EXT3. No strange options or anything. I can't think of any specific things
other than over-generous selection of Debian/unstable packages.

Cheers,
Magnus




Bug#170635: libc6 2.3.1-3 to 2.3.1-5 upgrade breaks

2002-12-02 Thread Magnus Danielson
From: Anthony DeRobertis <[EMAIL PROTECTED]>
Subject: Re: Bug#170635: libc6 2.3.1-3 to 2.3.1-5 upgrade breaks
Date: 02 Dec 2002 01:17:27 -0500

> On Sun, 2002-12-01 at 21:19, Magnus Danielson wrote:
> 
> > rm /lib/ld-2.3.1.so
> > 
> > I got the "Device or resource busy" message as a reply. It's because the file
> > is in use by many tools:
> 
> That shouldn't happen. "File in use" != "name in use". rm removes names,
> not files.

Exactly! I know... it's even traditional UNIX AFAIK.

> And you using something weird as your root file system? That is
> something besides ext[23] or reiserfs?

EXT3. No strange options or anything. I can't think of any specific things
other than over-generous selection of Debian/unstable packages.

Cheers,
Magnus


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




Bug#170635: libc6 2.3.1-3 to 2.3.1-5 upgrade breaks

2002-12-01 Thread Magnus Danielson
From: Jeff Bailey <[EMAIL PROTECTED]>
Subject: Re: Bug#170635: libc6 2.3.1-3 to 2.3.1-5 upgrade breaks
Date: 25 Nov 2002 06:58:16 -0500

Hi Jeff,

> On Mon, 2002-11-25 at 02:03, Magnus Danielson wrote:
> 
> > Package: libc6
> > Version: 2.3.1-5
> > Severity: critical
> > 
> > On upgrading my Debian/unstable system (oddly called heaven) from libc6 
> > 2.3.1-3
> > to 2.3.1-5 it halts the installation with this text:
> 
> I've done -3 to -5 upgrades on many i386 systems, and 2.2.5 to 2.3.1-5 
> upgrades
> hurd-i386 and sparc and never seen this message.  I need more details.

OK. This is what happend:

When the dpkg wanted to remove /lib/ld-2.3.1.so in order to insert it's own
(identical) this was not possible since the rm command got:

rm /lib/ld-2.3.1.so

I got the "Device or resource busy" message as a reply. It's because the file
is in use by many tools:

# lsof | grep ld-2.3.1.so
bash   907 magnus  memREG3,1   83504   1208304 /lib/ld-2.3.1.so
startx1005 magnus  memREG3,1   83504   1208304 /lib/ld-2.3.1.so
xinit 1016 magnus  memREG3,1   83504   1208304 /lib/ld-2.3.1.so
fvwm2 1021 magnus  memREG3,1   83504   1208304 /lib/ld-2.3.1.so
xsnow 1023 magnus  memREG3,1   83504   1208304 /lib/ld-2.3.1.so
xterm 1024 magnus  memREG3,1   83504   1208304 /lib/ld-2.3.1.so
xterm 1025 magnus  memREG3,1   83504   1208304 /lib/ld-2.3.1.so
xterm 1026 magnus  memREG3,1   83504   1208304 /lib/ld-2.3.1.so
bubblefis 1027 magnus  memREG3,1   83504   1208304 /lib/ld-2.3.1.so
bash  1028 magnus  memREG3,1   83504   1208304 /lib/ld-2.3.1.so
bash  1029 magnus  memREG3,1   83504   1208304 /lib/ld-2.3.1.so
bash  1030 magnus  memREG3,1   83504   1208304 /lib/ld-2.3.1.so
FvwmPager 1031 magnus  memREG3,1   83504   1208304 /lib/ld-2.3.1.so
emacs 3834 magnus  memREG3,1   83504   1208304 /lib/ld-2.3.1.so
lsof  6487 magnus  memREG3,1   83504   1208304 /lib/ld-2.3.1.so
grep  6488 magnus  memREG3,1   83504   1208304 /lib/ld-2.3.1.so
lsof  6489 magnus  memREG3,1   83504   1208304 /lib/ld-2.3.1.so

However, it is fishy that I should run into that problem... libc6 updates
should run smoothly and they have before.

Now, the trick around this problem is to do

cp /lib/ld-2.3.1.so /lib/ld-2.3.1.so.foo
rm /lib/ld-linux.so.2
ldconfig

Since the copied file which is more recent and ldconfig will fetch the
later. However, the bindings to the old one remains, so a reboot is the only
cure, but then the bindings will point to the .foo variant at next time, so
a new attempt to install will now work. What ever caused this problem, I don't
know. I've run pretty tight updates to unstable on this machine and this is one
of the few hickups I've had.

Now I have put these troubles behind me, but someone should figure out how to
handle this problem, both on how it could occur (I haven't fiddle with that
file prior to this!) and when it occurs. If you fuck up the libc6 installation
a lot of other stuff falls right off.

Cheers,
Magnus




Bug#170635: libc6 2.3.1-3 to 2.3.1-5 upgrade breaks

2002-12-01 Thread Magnus Danielson
From: Jeff Bailey <[EMAIL PROTECTED]>
Subject: Re: Bug#170635: libc6 2.3.1-3 to 2.3.1-5 upgrade breaks
Date: 25 Nov 2002 06:58:16 -0500

Hi Jeff,

> On Mon, 2002-11-25 at 02:03, Magnus Danielson wrote:
> 
> > Package: libc6
> > Version: 2.3.1-5
> > Severity: critical
> > 
> > On upgrading my Debian/unstable system (oddly called heaven) from libc6 2.3.1-3
> > to 2.3.1-5 it halts the installation with this text:
> 
> I've done -3 to -5 upgrades on many i386 systems, and 2.2.5 to 2.3.1-5 upgrades
> hurd-i386 and sparc and never seen this message.  I need more details.

OK. This is what happend:

When the dpkg wanted to remove /lib/ld-2.3.1.so in order to insert it's own
(identical) this was not possible since the rm command got:

rm /lib/ld-2.3.1.so

I got the "Device or resource busy" message as a reply. It's because the file
is in use by many tools:

# lsof | grep ld-2.3.1.so
bash   907 magnus  memREG3,1   83504   1208304 /lib/ld-2.3.1.so
startx1005 magnus  memREG3,1   83504   1208304 /lib/ld-2.3.1.so
xinit 1016 magnus  memREG3,1   83504   1208304 /lib/ld-2.3.1.so
fvwm2 1021 magnus  memREG3,1   83504   1208304 /lib/ld-2.3.1.so
xsnow 1023 magnus  memREG3,1   83504   1208304 /lib/ld-2.3.1.so
xterm 1024 magnus  memREG3,1   83504   1208304 /lib/ld-2.3.1.so
xterm 1025 magnus  memREG3,1   83504   1208304 /lib/ld-2.3.1.so
xterm 1026 magnus  memREG3,1   83504   1208304 /lib/ld-2.3.1.so
bubblefis 1027 magnus  memREG3,1   83504   1208304 /lib/ld-2.3.1.so
bash  1028 magnus  memREG3,1   83504   1208304 /lib/ld-2.3.1.so
bash  1029 magnus  memREG3,1   83504   1208304 /lib/ld-2.3.1.so
bash  1030 magnus  memREG3,1   83504   1208304 /lib/ld-2.3.1.so
FvwmPager 1031 magnus  memREG3,1   83504   1208304 /lib/ld-2.3.1.so
emacs 3834 magnus  memREG3,1   83504   1208304 /lib/ld-2.3.1.so
lsof  6487 magnus  memREG3,1   83504   1208304 /lib/ld-2.3.1.so
grep  6488 magnus  memREG3,1   83504   1208304 /lib/ld-2.3.1.so
lsof  6489 magnus  memREG3,1   83504   1208304 /lib/ld-2.3.1.so

However, it is fishy that I should run into that problem... libc6 updates
should run smoothly and they have before.

Now, the trick around this problem is to do

cp /lib/ld-2.3.1.so /lib/ld-2.3.1.so.foo
rm /lib/ld-linux.so.2
ldconfig

Since the copied file which is more recent and ldconfig will fetch the
later. However, the bindings to the old one remains, so a reboot is the only
cure, but then the bindings will point to the .foo variant at next time, so
a new attempt to install will now work. What ever caused this problem, I don't
know. I've run pretty tight updates to unstable on this machine and this is one
of the few hickups I've had.

Now I have put these troubles behind me, but someone should figure out how to
handle this problem, both on how it could occur (I haven't fiddle with that
file prior to this!) and when it occurs. If you fuck up the libc6 installation
a lot of other stuff falls right off.

Cheers,
Magnus


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




Bug#170635: libc6 2.3.1-3 to 2.3.1-5 upgrade breaks

2002-11-25 Thread Magnus Danielson
Package: libc6
Version: 2.3.1-5
Severity: critical

On upgrading my Debian/unstable system (oddly called heaven) from libc6 2.3.1-3
to 2.3.1-5 it halts the installation with this text:

Preparing to replace libc6 2.3.1-3 (using .../libc6_2.3.1-5_i386.deb) ...
Unpacking replacement libc6 ...
dpkg: error processing /var/cache/apt/archives/libc6_2.3.1-5_i386.deb 
(--unpack):
 unable to install new version of `./lib/ld-2.3.1.so': Device or resource busy
dpkg-deb: subprocess paste killed by signal (Broken pipe)
Errors were encountered while processing:
 /var/cache/apt/archives/libc6_2.3.1-5_i386.deb

When the file is to be replaced, it is being used by a heap of different tools.
"lsof | grep ld-" gives a pretty good idea where the trouble is.

This have worked before, new upgrade routines in the libc6 package?

Cheers,
Magnus




Bug#170635: libc6 2.3.1-3 to 2.3.1-5 upgrade breaks

2002-11-24 Thread Magnus Danielson
Package: libc6
Version: 2.3.1-5
Severity: critical

On upgrading my Debian/unstable system (oddly called heaven) from libc6 2.3.1-3
to 2.3.1-5 it halts the installation with this text:

Preparing to replace libc6 2.3.1-3 (using .../libc6_2.3.1-5_i386.deb) ...
Unpacking replacement libc6 ...
dpkg: error processing /var/cache/apt/archives/libc6_2.3.1-5_i386.deb (--unpack):
 unable to install new version of `./lib/ld-2.3.1.so': Device or resource busy
dpkg-deb: subprocess paste killed by signal (Broken pipe)
Errors were encountered while processing:
 /var/cache/apt/archives/libc6_2.3.1-5_i386.deb

When the file is to be replaced, it is being used by a heap of different tools.
"lsof | grep ld-" gives a pretty good idea where the trouble is.

This have worked before, new upgrade routines in the libc6 package?

Cheers,
Magnus


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




Re: Ping

2001-08-20 Thread Magnus Danielson
From: Wilfried Goesgens <[EMAIL PROTECTED]>
Subject: Ping
Date: Mon, 20 Aug 2001 09:53:43 +0200

> is anybody out there?

No, you are all alone!

Cheers,
Magnus




Re: Ping

2001-08-20 Thread Magnus Danielson

From: Wilfried Goesgens <[EMAIL PROTECTED]>
Subject: Ping
Date: Mon, 20 Aug 2001 09:53:43 +0200

> is anybody out there?

No, you are all alone!

Cheers,
Magnus


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




Trouble with getopt_long_only

1999-09-26 Thread Magnus Danielson
Hi!

I run fairly bleeding edge potato box with libc6 2.1.2-2 installed.
Both my Acrobat reader (acroread 4.0-1) and Netscape (netscape 4.61-21)
will fail. I did a quick gdb session on both just to find a common denominator:

For Netscape:
heaven:~> gdb netscape 
GNU gdb 4.18
Copyright 1998 Free Software Foundation, Inc.
GDB is free software, covered by the GNU General Public License, and you are
welcome to change it and/or distribute copies of it under certain conditions.
Type "show copying" to see the conditions.
There is absolutely no warranty for GDB.  Type "show warranty" for details.
This GDB was configured as "i586-pc-linux-gnu"...
(no debugging symbols found)...
(gdb) run
Starting program: /usr/local/netscape/netscape 
(no debugging symbols found)...
...
Program received signal SIGSEGV, Segmentation fault.
0x402fc4e4 in getopt_long_only () from /lib/libc.so.6
(gdb) backtrace
#0  0x402fc4e4 in getopt_long_only () from /lib/libc.so.6
#1  0x4028c115 in signgaml () from /lib/libc.so.6
#2  0x4028c184 in signgaml () from /lib/libc.so.6
#3  0x40002205 in _dl_boot () from /lib/ld-linux.so.1
(gdb)

For Acroread:
bash-2.02$ gdb $ACRO_EXEC_CMD
GNU gdb 4.18
Copyright 1998 Free Software Foundation, Inc.
GDB is free software, covered by the GNU General Public License, and you are
welcome to change it and/or distribute copies of it under certain conditions.
Type "show copying" to see the conditions.
There is absolutely no warranty for GDB.  Type "show warranty" for details.
This GDB was configured as "i586-pc-linux-gnu"...
(no debugging symbols found)...
(gdb) run
Starting program: /usr/lib/Acrobat4/Reader/intellinux/bin/acroread 
(no debugging symbols found)...
...
Program received signal SIGSEGV, Segmentation fault.
0x408254e4 in getopt_long_only () from /lib/libc.so.6
(gdb) backtrace
#0  0x408254e4 in getopt_long_only () from /lib/libc.so.6
#1  0x407b5115 in _IcePaAuthDataEntries () from /lib/libc.so.6
#2  0x407b5184 in _IcePaAuthDataEntries () from /lib/libc.so.6
#3  0x40002205 in _dl_boot () from /lib/ld-linux.so.1
(gdb) 

They both get there in diffrent ways as you can see, but they fail at
presicely the same place (they are just loaded into diffrent offsets).

Anyway, is this a know bug?

How does we solve this?

Cheers,
Magnus