From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1754293AbYLIQ2y (ORCPT ); Tue, 9 Dec 2008 11:28:54 -0500 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S1753278AbYLIQ2q (ORCPT ); Tue, 9 Dec 2008 11:28:46 -0500 Received: from mx2.redhat.com ([66.187.237.31]:56258 "EHLO mx2.redhat.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1752713AbYLIQ2p (ORCPT ); Tue, 9 Dec 2008 11:28:45 -0500 Date: Tue, 9 Dec 2008 11:27:45 -0500 From: Jason Baron To: Greg KH Cc: mingo@elte.hu, linux-kernel@vger.kernel.org, rostedt@goodmis.org, fweisbec@gmail.com Subject: Re: [rfc] [patch] hook tracer to dynamic debug Message-ID: <20081209162745.GA3107@redhat.com> References: <20081205193928.GA3134@redhat.com> <20081207054840.GA20541@suse.de> MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: <20081207054840.GA20541@suse.de> User-Agent: Mutt/1.5.18 (2008-05-17) Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org On Sat, Dec 06, 2008 at 09:48:40PM -0800, Greg KH wrote: > On Fri, Dec 05, 2008 at 02:39:28PM -0500, Jason Baron wrote: > > hi, > > > > Currently, the dynamic debug framework uses 'printk' for output. Here, > > i've hooked the backend of the dynamic debug framework to the tracing > > framework. This allows the dynamic debug output to be seaprated from the > > system log, and allows for higher performance. > > Is this really an issue most of the time? > > I know logs can get full, but this seems like a bit of an overkill :) > > > I've implemented it as a compile time option, but a user might want to > > be able to toggle b/w 'printk' and the tracer at run-time...other > > ideas? Patch is against the -tip tree... > > It would have to be a run-time option if you want any distro to ever be > able to enable this in their kernels. > There haven't been any specfic complaints about this, although the all 'on' case does really spew quite a lot into the logs. Using Frederic's suggestion I'm able to implement this with only a few lines of code...so I'll keep it around if we ever need it... thanks, -Jason