From mboxrd@z Thu Jan 1 00:00:00 1970 From: Josh Triplett Subject: Re: rcu: Make rcu_assign_pointer's assignment volatile and type-safe Date: Sun, 1 Sep 2013 17:06:26 -0700 Message-ID: <20130902000626.GA22111@leaf> References: <5223d2b1.W0Uf1KjUIaBOcCMH%wangshilong1991@gmail.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Return-path: Content-Disposition: inline In-Reply-To: <5223d2b1.W0Uf1KjUIaBOcCMH%wangshilong1991@gmail.com> Sender: linux-kernel-owner@vger.kernel.org To: Wang Shilong Cc: linux-kernel@vger.kernel.org, linux-sparse@vger.kernel.org, sparse@chrisli.org, lttng-dev@lists.lttng.org, Stephen Hemminger , "Paul E. McKenney" , Mathieu Desnoyers , Andy Whitcroft , Joe Perches List-Id: linux-sparse@vger.kernel.org On Mon, Sep 02, 2013 at 07:50:09AM +0800, Wang Shilong wrote: > Hello, Using checkpatch.pl, i get the following warnings(errors): > WARNING: line over 80 characters > #57: FILE: include/linux/rcupdate.h:518: > + extern void __rcu_assign_pointer_typecheck(int, typeof(*(v)) __kernel *); \ It'd get much uglier if wrapped, and the function name needs to stay unique to avoid conflicts with the macro context. I don't plan to change this. > ERROR: need consistent spacing around '*' (ctx:WxB) > #72: FILE: include/linux/rcupdate.h:568: > + ACCESS_ONCE(p) = (typeof(*(v)) __force space *)(v); \ False positive; checkpatch.pl seems to parse this as multiplication rather than a cast. - Josh Triplett