From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: from mx2.suse.de (cantor2.suse.de [195.135.220.15]) (using TLSv1 with cipher DHE-RSA-AES256-SHA (256/256 bits)) (Client CN "mx2.suse.de", Issuer "CAcert Class 3 Root" (not verified)) by ozlabs.org (Postfix) with ESMTPS id 62C8E2C007F for ; Tue, 3 Jul 2012 07:34:29 +1000 (EST) Subject: Re: [PATCH] PPC: use CURRENT_THREAD_INFO instead of open coded assembly Mime-Version: 1.0 (Apple Message framework v1278) Content-Type: text/plain; charset=us-ascii From: Alexander Graf In-Reply-To: <4FF21343.20408@freescale.com> Date: Mon, 2 Jul 2012 23:34:21 +0200 Message-Id: <23F3EB4C-799C-41EF-BAB8-360DF220FC62@suse.de> References: <1341262574-2841-1-git-send-email-stuart.yoder@freescale.com> <40388C8A-6D3B-409E-9C6D-45005E1F6016@suse.de> <1341264368.2588.49.camel@pasglop> <4FF21343.20408@freescale.com> To: Scott Wood Cc: linuxppc-dev@lists.ozlabs.org, Stuart Yoder List-Id: Linux on PowerPC Developers Mail List List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , On 02.07.2012, at 23:31, Scott Wood wrote: > On 07/02/2012 04:27 PM, Alexander Graf wrote: >>=20 >> On 02.07.2012, at 23:26, Benjamin Herrenschmidt wrote: >>=20 >>> On Mon, 2012-07-02 at 22:59 +0200, Alexander Graf wrote: >>>> This should go into an #ifdef __ASSEMBLY__ block, right? :) >>>=20 >>> We almost never use #ifdef __ASSEMBLY__, we use it the other way >>> around, to prevent C stuff from being included in assembly. The >>> other way around is legit since things might be used in inline asm >>> for example. >>=20 >> I'm not sure I want to see this bit of code used as is in inline asm >> :). I don't even think it's possible, since it's a full statement. >> Either way, it's safer with the guard. >=20 > Safer from what? It won't be expanded unless referenced. How is this > better than putting ifdefs on #includes, prototypes, struct = definitions, > etc.? The ifdef is just clutter. Well, it'd make it easier to read the errors resulting of it. Calling = CURRENT_THREAD_INFO from within C code would throw random compiler = errors at you that are quite unintelligible, while a missing definition = would be a reasonably obvious thing to fix, no? Either way, not married to this. I just find it cleaner to not expose = something as a define that wouldn't work in the first place. Alex