From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S934034AbXC0Rk7 (ORCPT ); Tue, 27 Mar 2007 13:40:59 -0400 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S934015AbXC0Rk7 (ORCPT ); Tue, 27 Mar 2007 13:40:59 -0400 Received: from cantor2.suse.de ([195.135.220.15]:45152 "EHLO mx2.suse.de" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S934034AbXC0Rk6 (ORCPT ); Tue, 27 Mar 2007 13:40:58 -0400 To: Tom Horsley Cc: linux-kernel@vger.kernel.org, bugsy Subject: Re: TLS info and core files? References: <20070327092041.40947947@tweety> From: Andi Kleen Date: 27 Mar 2007 20:39:38 +0200 In-Reply-To: <20070327092041.40947947@tweety> Message-ID: User-Agent: Gnus/5.09 (Gnus v5.9.0) Emacs/21.3 MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Sender: linux-kernel-owner@vger.kernel.org X-Mailing-List: linux-kernel@vger.kernel.org Tom Horsley writes: > > Should core files have NOTE entries added for some sort of > thread local storage info to be recorded? (Whatever the > kernel happens to know that won't be accessible anymore > after the process is gone, which would probably vary > from architecture to architecture.) I think that would be a good idea. Shouldn't be very hard to implement. Please send a patch. We also miss some other information in core files I found useful in the past, like the CR3 and some other registers. The printks on x86-64 actually give more information than a core dump in some cases. -Andi