From mboxrd@z Thu Jan 1 00:00:00 1970 From: "Paul E. McKenney" Subject: Re: Compilers and RCU readers: Once more unto the breach! Date: Fri, 22 May 2015 06:12:52 -0700 Message-ID: <20150522131252.GG6776@linux.vnet.ibm.com> References: <20150520024148.GD6776@linux.vnet.ibm.com> <20150520114745.GC11498@arm.com> <20150520121522.GH6776@linux.vnet.ibm.com> <20150520154617.GE11498@arm.com> <20150520181606.GT6776@linux.vnet.ibm.com> <20150521192422.GC19204@arm.com> <20150521200212.GW6776@linux.vnet.ibm.com> <20150522064344.GA17518@gmail.com> Reply-To: paulmck@linux.vnet.ibm.com Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Return-path: Received: from e33.co.us.ibm.com ([32.97.110.151]:59003 "EHLO e33.co.us.ibm.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1756856AbbEVNNA (ORCPT ); Fri, 22 May 2015 09:13:00 -0400 Received: from /spool/local by e33.co.us.ibm.com with IBM ESMTP SMTP Gateway: Authorized Use Only! Violators will be prosecuted for from ; Fri, 22 May 2015 07:13:00 -0600 Content-Disposition: inline In-Reply-To: <20150522064344.GA17518@gmail.com> Sender: linux-arch-owner@vger.kernel.org List-ID: To: Ingo Molnar Cc: Linus Torvalds , Will Deacon , Linux Kernel Mailing List , "c++std-parallel@accu.org" , "linux-arch@vger.kernel.org" , "gcc@gcc.gnu.org" , p796231 , "mark.batty@cl.cam.ac.uk" , Peter Zijlstra , Ramana Radhakrishnan , David Howells , Andrew Morton , "michaelw@ca.ibm.com" On Fri, May 22, 2015 at 08:43:44AM +0200, Ingo Molnar wrote: > > * Linus Torvalds wrote: > > > (a) the "official" rules are completely pointless, and make sense > > only because the standard is written for some random "abstract > > machine" that doesn't actually exist. > > Presuming the intent of the abstract machine specification is to avoid > being seen as biased towards any specific machine (politics), maybe > write this as: > > (a) the "official" rules are written for a somewhat weird and > complex "union of all known and theoretically possible CPU > architectures that exist or which might exist in the future", > which machine does not actually exist in practice, but which > allows a single abstract set of rules to apply to all machines. > These rules are complex, but if applied to a specific machine > they become considerably simpler. Here's a few examples: ... > > ? > > (Assuming it's a goal of this standard to be human parseable to more > than a few dozen people on the planet.) Should something based on Section 7.9 go in, then I would need to add a more developer-friendly explanation in Documentation/RCU, no two ways about it! ;-) Thanx, Paul