Michal, does the init file look okay upstream? http://cgit.freedesktop.org/hal/tree/hald/haldaemon.in

Whoa. Making comment #0 private because of links to intranet.corp.

Also, ack.

(In reply to comment #2)
> Michal, does the init file look okay upstream?
> http://cgit.freedesktop.org/hal/tree/hald/haldaemon.in  

No. From the upstream script:

# Sanity checks.
[ -x @sbindir@/hald ] || exit 0

is the problem.

(In reply to comment #4)
> (In reply to comment #2)
> > Michal, does the init file look okay upstream?
> > http://cgit.freedesktop.org/hal/tree/hald/haldaemon.in  
> 
> No. From the upstream script:
> 
> # Sanity checks.
> [ -x @sbindir@/hald ] || exit 0
> 
> is the problem.  

Can I just change this to || exit 1, or should the line be removed altogether? I don't see how the init.d script could be installed, but not /usr/sbin/hald, as they are in the same package. It seems an odd check.

I can keep this for 5.4, or punt it to 5.5 as it's not ever triggered unless you've got a very broken system.

Quite a few of the other init scripts have the same sanity check as hal, exiting with status zero.

Changing it to exit 1 would be fine. I agree, this is a pedantic bug report. It'd be a good idea to fix the rest of the init scripts, as exiting 0 is the wrong thing to do, but... as you said, I'm not sure that if you managed to hit this bug, the exit code will be anywhere near the top of your list of things to care about.

An advisory has been issued which should help the problem
described in this bug report. This report is therefore being
closed with a resolution of ERRATA. For more information
on therefore solution and/or where to find the updated files,
please follow the link below. You may reopen this bug report
if the solution does not work for you.

http://rhn.redhat.com/errata/RHBA-2010-0256.html