From mboxrd@z Thu Jan 1 00:00:00 1970 From: Russell King - ARM Linux Subject: Re: [PATCH v2] coresight-stm: adding driver for CoreSight STM component Date: Thu, 26 Feb 2015 22:58:53 +0000 Message-ID: <20150226225853.GM8656@n2100.arm.linux.org.uk> References: <1424907152-18808-1-git-send-email-mathieu.poirier@linaro.org> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Return-path: Content-Disposition: inline In-Reply-To: Sender: linux-kernel-owner@vger.kernel.org To: Rob Herring Cc: Mathieu Poirier , Will Deacon , "linux-api@vger.kernel.org" , Jonathan Corbet , "linux-kernel@vger.kernel.org" , "linux-arm-kernel@lists.infradead.org" , "linux-doc@vger.kernel.org" List-Id: linux-api@vger.kernel.org On Thu, Feb 26, 2015 at 04:24:53PM -0600, Rob Herring wrote: > We really shouldn't do private implementation here. It there really > any reason not to allow readq/writeq generically for 32-bit or just > for arm32? My argument has always been that drivers should do the emulation of 64-bit accesses when there is no native support. IO registers tend to have side effects when read/written. How do we know whether the low-half or the high-half should be written first? This isn't something that an architecture can really dictate. What may be right for one hardware device may not be correct for another. -- FTTC broadband for 0.8mile line: currently at 10.5Mbps down 400kbps up according to speedtest.net.