Re: [PATCH] tracing: add trace console

From: Johannes Berg <johannes@sipsolutions.net>
To: Peter Zijlstra <peterz@infradead.org>
Cc: Steven Rostedt <rostedt@goodmis.org>,
	LKML <linux-kernel@vger.kernel.org>,
	Frederic Weisbecker <fweisbec@gmail.com>,
	Ingo Molnar <mingo@redhat.com>,
	Thomas Gleixner <tglx@linutronix.de>,
	Andrew Morton <akpm@linux-foundation.org>
Subject: Re: [PATCH] tracing: add trace console
Date: Wed, 16 Nov 2011 15:30:46 +0100	[thread overview]
Message-ID: <1321453846.4502.0.camel@jlt3.sipsolutions.net> (raw)
In-Reply-To: <1321453537.27735.0.camel@twins>

On Wed, 2011-11-16 at 15:25 +0100, Peter Zijlstra wrote:

> > >  kernel/trace/Kconfig         |   13 ++++++++++
> > >  kernel/trace/Makefile        |    2 +
> > >  kernel/trace/trace-console.c |   55 +++++++++++++++++++++++++++++++++++++++++++
> > >  3 files changed, 70 insertions(+)
> > > 
> > > --- a/kernel/trace/Kconfig    2011-11-16 11:11:27.000000000 +0100
> > > +++ b/kernel/trace/Kconfig    2011-11-16 11:11:35.000000000 +0100
> > > @@ -368,6 +368,19 @@ config BLK_DEV_IO_TRACE
> > >  
> > >         If unsure, say N.
> > >  
> > > +config TRACE_CONSOLE
> > > +     tristate "Support for a tracing console"
> > > +     help
> > > +       Say M (or Y if you must, but this is not recommended) 
> 
> Why not?

It'll always be registered and you have no way to turn it off then, so
you can't trace without getting every printk. I didn't want to add extra
API to toggle it, but building it in might be a bit odd that way.

johannes