From: Henrik Carlqvist on
Peter Chant <REMpeteOVE(a)petezilla.co.uk> wrote:
> bash-4.1# /etc/rc.d/rc.wicd restart
> Stopping wicd daemon...
> Starting wicd daemon: /usr/sbin/wicd &

Others have guessed on old lock files being left since the unclean
shutdown and they might be right. One way to get a clue of what is really
happening might be to try:

strace -f /usr/sbin/wicd

regards Henrik
--
The address in the header is only to prevent spam. My real address is:
hc123(at)poolhem.se Examples of addresses which go to spammers:
root(a)localhost postmaster(a)localhost

From: Peter Chant on
Henrik Carlqvist wrote:


> strace -f /usr/sbin/wicd

Problem seems to lie with wicd itself. However, strace pulls up a HUGE
number of files - many missing.

--
http://www.petezilla.co.uk
From: Peter Chant on
Grant wrote:

> Also check for stale files under /var/{run,lock} and friends.
>
> Or, installpkg --reinstall \
> /path/to/slackware*_$version/slackware*/*/t*z [1]
>
> for the grossly overdone fix :(

Its reaching that point - that or reinstall.

Whilst on the subject of doing an install to a machine with no dvd drive.

Handy hint, don't format your external drives to ext4 until small distro's
such as Slax support them.

--
http://www.petezilla.co.uk
From: Henrik Carlqvist on
Peter Chant <REMpeteOVE(a)petezilla.co.uk> wrote:
> Problem seems to lie with wicd itself.

If you really think that the wicd binary is broken that could easily be
verified by running md5sum on the file and compare the result with the
result from a good system.

> However, strace pulls up a HUGE number of files - many missing.

It does not mean that something is wrong when strace shows a lot of
attempts to load shared libraries from different directories. That is how
ld.so works, it tries to load every shared library from a number of
directories until the file is found.

Look at the end of the output from strace. Do you find any clue there to
why wicd decided to stop?

regards Henrik
--
The address in the header is only to prevent spam. My real address is:
hc123(at)poolhem.se Examples of addresses which go to spammers:
root(a)localhost postmaster(a)localhost

From: hba on
On Aug 3, 5:59 am, Henrik Carlqvist <Henrik.Carlqv...(a)deadspam.com>
wrote:
> Peter Chant <REMpete...(a)petezilla.co.uk> wrote:
> > bash-4.1# /etc/rc.d/rc.wicd restart
> > Stopping wicd daemon...
> > Starting wicd daemon: /usr/sbin/wicd &
>
> Others have guessed on old lock files being left since the unclean
> shutdown and they might be right.

It's not just for an unclean shutdown but also for a change of
runlevels (3 or 4 to 1). I fixed by editing /etc/rc.d/rc.K

--- /etc/rc.d/rc.K 2008-12-02 14:31:48.000000000 -0600
+++ etc/rc.d/rc.K 2010-07-08 00:22:57.000000000 -0500
@@ -42,6 +42,11 @@
/etc/rc.d/rc.local_shutdown stop
fi

+# Stop Mysql server:
+if [ -x /etc/rc.d/rc.mysqld ]; then
+ /etc/rc.d/rc.mysqld stop
+fi
+
# Stop the Apache web server:
if [ -x /etc/rc.d/rc.httpd ]; then
/etc/rc.d/rc.httpd stop
@@ -52,6 +57,11 @@
. /etc/rc.d/rc.samba stop
fi

+# Stop wicd:
+if [ -x /etc/rc.d/rc.wicd ]; then
+ . /etc/rc.d/rc.wicd stop
+fi
+
# Shut down the NFS server:
if [ -x /etc/rc.d/rc.nfsd ]; then
/etc/rc.d/rc.nfsd stop

As i can see the "problem" with the current rc.K script is that send a
SIGHUP, SIGTERM, SIGKILL to wicd (and currently to *all* process that
are not "stop" by and /etc/rc.d/rc.* file) and wicd can't delete its
PID file by itself. This also was an error i got with rc.hdapsd file
(http://slackbuilds.org/slackbuilds/13.0/system/hdapsd/rc.hdapsd)
since it wasn't in /etc/rc.d/rc.local_shutdown.

- hba