From mboxrd@z Thu Jan 1 00:00:00 1970 From: Arnd Bergmann Subject: Re: [RFC] asm-generic: default BUG_ON(x) to "if(x) BUG()" Date: Mon, 23 Nov 2015 17:52:38 +0100 Message-ID: <4150271.zrLzq08T60@wuerfel> References: <5868782.RxZY0W5S4d@wuerfel> <20151123163359.GC768@1wt.eu> <20151123163750.GT8644@n2100.arm.linux.org.uk> Mime-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 7Bit Return-path: Received: from mout.kundenserver.de ([212.227.126.131]:52436 "EHLO mout.kundenserver.de" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1752053AbbKWQx1 (ORCPT ); Mon, 23 Nov 2015 11:53:27 -0500 In-Reply-To: <20151123163750.GT8644@n2100.arm.linux.org.uk> Sender: linux-arch-owner@vger.kernel.org List-ID: To: Russell King - ARM Linux Cc: Willy Tarreau , linux-arch@vger.kernel.org, linux-kernel@vger.kernel.org, linux-arm-kernel@lists.infradead.org, josh@joshtriplett.org On Monday 23 November 2015 16:37:50 Russell King - ARM Linux wrote: > On Mon, Nov 23, 2015 at 05:33:59PM +0100, Willy Tarreau wrote: > > I think you could do better by simply calling panic("BUG!") instead as > > BUG() does. It will avoid the printk() call and pushing the file/line > > number onto the stack. It will also probably not inflate the rodata this > > way. > > Does that not depend on the architectures BUG() implementation? If an > architecture implements it as a signalling illegal instruction and a > lookup table, changing it to be a panic() would probably be more code. Correct, overall, we are down to a 1.40% size reduction compared to 1.70% without my patch and 1.49% with my version: section CONFIG_BUG=y CONFIG_BUG=n CONFIG_BUG=n+patch panic("BUG!") .text 8320248 | 8180944 | 8207688 | 8221848 .rodata 3633720 | 3567144 | 3570648 | 3567344 __bug_table 32508 | --- | --- | --- __modver 692 | 1584 | 2176 | 1384 .init.text 558132 | 548300 | 550088 | 550592 .exit.text 12380 | 12256 | 12380 | 12448 .data 1016672 | 1016064 | 1016128 | 1016064 Total 14622556 | 14374510 | 14407326 | 14417898 Arnd