From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id ; Fri, 24 May 2002 03:16:14 -0400 Received: (majordomo@vger.kernel.org) by vger.kernel.org id ; Fri, 24 May 2002 03:16:13 -0400 Received: from ausmtp02.au.ibm.COM ([202.135.136.105]:8628 "EHLO ausmtp02.au.ibm.com") by vger.kernel.org with ESMTP id ; Fri, 24 May 2002 03:16:12 -0400 From: Rusty Russell To: Benjamin LaHaise Cc: torvalds@transmeta.com, linux-kernel@vger.kernel.org, akpm@zip.com.au Subject: Re: [PATCH] XBUG(comment) BUG enhancement In-Reply-To: Your message of "Fri, 24 May 2002 02:49:06 -0400." <20020524024906.A1547@redhat.com> Date: Fri, 24 May 2002 17:19:51 +1000 Message-Id: Sender: linux-kernel-owner@vger.kernel.org X-Mailing-List: linux-kernel@vger.kernel.org In message <20020524024906.A1547@redhat.com> you write: > On Fri, May 24, 2002 at 03:24:30PM +1000, Rusty Russell wrote: > > __asm__ __volatile__( "ud2\n" \ > > - "\t.word %c0\n" \ > > + "\t.long %c0\n" \ > > "\t.long %c1\n" \ > > - : : "i" (__LINE__), "i" (__FILE__)) > > + : : "i" (__stringify(__LINE__)), "i" (__FILE__)) > > This part I can't agree with: changing the line number to a string > results in excess pollution of the data segment with useless strings > that are frequently duplicates. Why not leave it as a number? To unify the trap handler to handle both cases. If you really think this is unacceptable bloat, please measure the difference, then use line number zero or something for XBUG and place the comment string at in a third .long. But I was lazy, and this was tested... Rusty. -- Anyone who quotes me in their sig is an idiot. -- Rusty Russell.