From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: with ECARTIS (v1.0.0; list xfs); Fri, 22 Feb 2008 16:16:57 -0800 (PST) Received: from cuda.sgi.com (cuda2.sgi.com [192.48.168.29]) by oss.sgi.com (8.12.11.20060308/8.12.11/SuSE Linux 0.7) with ESMTP id m1N0GpmX031532 for ; Fri, 22 Feb 2008 16:16:53 -0800 Date: Fri, 22 Feb 2008 19:17:15 -0500 From: Christoph Hellwig Subject: Re: [patch, debug, 2/2] Use power-of-2 size ktrace buffers Message-ID: <20080223001715.GA22614@infradead.org> References: <20080218230112.GU155407@sgi.com> <20080222040806.GB16993@infradead.org> <20080222043020.GN155407@sgi.com> MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: <20080222043020.GN155407@sgi.com> Sender: xfs-bounce@oss.sgi.com Errors-to: xfs-bounce@oss.sgi.com List-Id: xfs To: David Chinner Cc: Christoph Hellwig , xfs-dev , xfs-oss On Fri, Feb 22, 2008 at 03:30:20PM +1100, David Chinner wrote: > Ok, I'll cook up another patch for that and send it out. Yeah, making that a separate patch is fine with me. It's just a logical next step. > > a second step ktrace should be replaced with something based on the > > various trace thingies floating around allowing to read out the trace > > buffer from userspace instead of having to rely on kdb. > > Yeah, that make sense for tracing from userspace. But most of the > time I find a need for the tracing is when the machine has crashed > or assert failed. Hence I don't see that we can really remove the > kdb side of things. Perhaps two different tracing modules could > be done..... It should still be one tracing buffer, with both a userspace space interface and one from the various kernel debuggers. Adding the trace buffers to core dumps in addition would be even better.