From mboxrd@z Thu Jan 1 00:00:00 1970 From: Arnd Bergmann Subject: Re: [PATCH] unicore32 fix: remove arch-specific futex support Date: Wed, 30 Mar 2011 14:17:19 +0200 Message-ID: <201103301417.20027.arnd@arndb.de> References: <01eb01cbeeb7$30390460$90ab0d20$@mprc.pku.edu.cn> Mime-Version: 1.0 Content-Type: Text/Plain; charset="iso-8859-1" Content-Transfer-Encoding: 7bit Return-path: Received: from moutng.kundenserver.de ([212.227.17.9]:62775 "EHLO moutng.kundenserver.de" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S932396Ab1C3MRc (ORCPT ); Wed, 30 Mar 2011 08:17:32 -0400 In-Reply-To: <01eb01cbeeb7$30390460$90ab0d20$@mprc.pku.edu.cn> Sender: linux-arch-owner@vger.kernel.org List-ID: To: Guan Xuetao , Paul Mundt Cc: linux-kernel@vger.kernel.org, linux-arch@vger.kernel.org On Wednesday 30 March 2011, Guan Xuetao wrote: > The futex functions in unicore32 are not used and verified, > so just replaced by asm-generic version. > > Signed-off-by: Guan Xuetao Note that the a working futex implementation would be really useful to get efficient multithreaded applications. On uniprocessor machines, this can easily be done by disabling interrupts, like you also do for spinlock and atomic. I believe we could actually copy the arch/sh/include/asm/futex-irq.h implementation to include/asm-generic and use that whenever CONFIG_SMP is disabled. That would also help a lot of other architectures. Arnd