LWN.net Logo

Re: Regression - /proc/kmsg does not (always) block for 1-byte reads

Re: Regression - /proc/kmsg does not (always) block for 1-byte reads

Posted Jul 13, 2012 1:23 UTC (Fri) by hmh (subscriber, #3838)
Parent article: Re: Regression - /proc/kmsg does not (always) block for 1-byte reads

Meh, the "daemonized dd bs=1" is weirdness that came from Ubuntu in a package Debian should, IMHO, have got rid of a long time ago (klogd and syslogd).

Standard Debian installs will use rsyslog, and not install klogd (nor its companion syslogd).

Relevant data:
http://qa.debian.org/popcon.php?package=sysklogd
http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=35325

Maybe we can simply get rid of it in Debian. And I didn't check who wrote the dd hack for Ubuntu in the first place, IMO some things are best left unknown. Besides, Debian bug #35325 linked above clearly shows how crap can easily get accepted into effectively unmaintained packages, even if someone raised a (granted, very mild) objection to it.

We need to get better at ditching unmaintained crap in Debian. In this particular case, I am told the issue is the lack of a good automated transition of craplogd^Wsysklogd config to rsyslog.


(Log in to post comments)

Re: Regression - /proc/kmsg does not (always) block for 1-byte reads

Posted Jul 13, 2012 1:26 UTC (Fri) by dlang (✭ supporter ✭, #313) [Link]

given that you can just include a sysklog config in your rsyslog.conf file with no modification, there isn't any translation needed, so any transition should be trivial.

Debian already uses the include directive to include files from /etc/rsyslog.d so this isn't even something new.

Copyright © 2013, Eklektix, Inc.
Comments and public postings are copyrighted by their creators.
Linux is a registered trademark of Linus Torvalds