Re: Linux... (or anyone) and timeloss in joystick driver.

Linus Torvalds (
Wed, 28 Jun 1995 08:44:46 +0300

G.W. Wettstein: "Re: Linux... (or anyone) and timeloss in joystick driver." (Jun 27, 12:37):
> > From: Marek Michalkiewicz <>
> > Date: Mon, 26 Jun 1995 17:25:10 +0200 (MET DST)
> > Subject: Re: Linus... (or anyone) and timeloss in joystick driver.
> >
> > BTW, I have some ideas about the register dump feature. Since it is
> > explicitly requested by the user, why not always print it on the console
> > (like all the Oops messages) even if klogd is running? If something bad
> > happened, it may be difficult to log in and kill klogd...
> There is a command line option (-c) in klogd which defines the
> priority level of kernel messages which are to be directed to the
> console. I believe that by default anything of lower priority
> (higher number) than 1 (KERN_ALERT) gets forwarded by klogd into the
> syslogd facility.

This is one of my major gripes with klogd, actually. This "feature"
means that people who don't know about the kernel logs will never even
notice kernel messages, so they'll never know about strangenesses
occurring. It's good that they go to a log-file, but they should _also_
go to the screen. I'd much rather see the default be "-c 7", which will
print out all the informational messages (the "default" kernel message
level), and only the debugging messages would never show up, only get

(If there are messages that people don't like seeing, like the "VFS:
disk change detected", we can change those particular messages to
debugging messages, and everybody will be happy).

> By default printk issues all its output with a priority value of

I think the default is actually 6, despite the DEFAULT_MESSAGE_LOGLEVEL
being defined as 7 at the top of that file. Illogically, the place that
uses the DEFAULT_MESSAGE_LOGLEVEL will do a "-1+'0'" to it.. Strange.