From mboxrd@z Thu Jan 1 00:00:00 1970 From: Tony Lindgren Subject: Re: [PATCH v2 15/31] arm64: SMP support Date: Fri, 17 Aug 2012 02:39:12 -0700 Message-ID: <20120817093912.GW11011@atomide.com> References: <1344966752-16102-1-git-send-email-catalin.marinas@arm.com> <1344966752-16102-16-git-send-email-catalin.marinas@arm.com> <20120817092133.GR11011@atomide.com> <20120817093250.GB24389@arm.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Return-path: Received: from mho-01-ewr.mailhop.org ([204.13.248.71]:58395 "EHLO mho-01-ewr.mailhop.org" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S964833Ab2HQJjP (ORCPT ); Fri, 17 Aug 2012 05:39:15 -0400 Content-Disposition: inline In-Reply-To: <20120817093250.GB24389@arm.com> Sender: linux-arch-owner@vger.kernel.org List-ID: To: Catalin Marinas Cc: "linux-arch@vger.kernel.org" , "linux-arm-kernel@lists.infradead.org" , "linux-kernel@vger.kernel.org" , Arnd Bergmann , Will Deacon , Marc Zyngier * Catalin Marinas [120817 02:33]: > On Fri, Aug 17, 2012 at 10:21:33AM +0100, Tony Lindgren wrote: > > * Catalin Marinas [120814 11:05]: > > > This patch adds SMP initialisation and spinlocks implementation for > > > AArch64. The spinlock support uses the new load-acquire/store-release > > > instructions to avoid explicit barriers. The architecture also specifies > > > that an event is automatically generated when clearing the exclusive > > > monitor state to wake up processors in WFE, so there is no need for an > > > explicit DSB/SEV instruction sequence. The SEVL instruction is used to > > > set the exclusive monitor locally as there is no conditional WFE and a > > > branch is more expensive. > > > > Do we always have SMP hardware on arm64? Or are we going to need to > > again add smp_on_up support later on? > > There isn't anything in the architecture specs that mandates multiple > cores but given the current trend it's very likely that we'll always > have MP. > > An improvement in AArch64 is that we can use the SMP cache/TLB ops (the > inner shareable variants) even on a UP system so there is no need for > run-time code patching for correct execution. That's good to hear! Tony