From mboxrd@z Thu Jan 1 00:00:00 1970 From: Christian Borntraeger Subject: Re: + separate-atomic_t-declaration-from-asm-atomich-into-asm-atomic_defh.patch added to -mm tree Date: Mon, 1 Sep 2008 11:21:40 +0200 Message-ID: <200809011121.40686.borntraeger@de.ibm.com> References: <200808222056.m7MKucHA016131@imap1.linux-foundation.org> <20080822213624.GX8318@parisc-linux.org> Mime-Version: 1.0 Content-Type: text/plain; charset="iso-8859-1" Content-Transfer-Encoding: 7bit Return-path: Received: from mtagate5.uk.ibm.com ([195.212.29.138]:47126 "EHLO mtagate5.uk.ibm.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1755964AbYIAJYO (ORCPT ); Mon, 1 Sep 2008 05:24:14 -0400 In-Reply-To: <20080822213624.GX8318@parisc-linux.org> Content-Disposition: inline Sender: linux-arch-owner@vger.kernel.org List-ID: To: Matthew Wilcox Cc: akpm@linux-foundation.org, mm-commits@vger.kernel.org, ying.huang@intel.com, linux-arch@vger.kernel.org, linux-s390@vger.kernel.org, Martin Schwidefsky Am Freitag, 22. August 2008 schrieb Matthew Wilcox: > > S390 has an __aligned__((4)) on theirs -- is this really necessary? > Doesn't s390 align ints to 4 bytes automatically? If it doesn't, it Yes, the s390 abi requires a 4 byte alignment for int. The __aligned__((4)) looks unnecessary. > +++ include/linux/atomic_type.h > +typedef struct { volatile int counter; } atomic_t; > +#ifdef CONFIG_64BIT > +typedef struct { volatile long counter; } atomic64_t; > +#endif Hmm, the current s390 version has no volatile. I can recall some very long mail threads about this. What was the outcome? Christian