From mboxrd@z Thu Jan 1 00:00:00 1970 From: David Brownell Subject: Re: Revisiting "problem with converting from at91_spi to atmel_spi (AT91RM9200)" Date: Wed, 03 Oct 2007 09:33:58 -0700 Message-ID: <20071003163358.6C79A235028@adsl-69-226-248-13.dsl.pltn13.pacbell.net> References: <37ED64EE21A398409D609EAF7B0525A02164EB@hw-mail.hannover.huwag.net> Mime-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 7bit Cc: andrew-eS41wJS13H5l57MIdRCFDg@public.gmane.org To: spi-devel-general-5NWGOfrQmneRv+LV9MX5uipxlwaOVQ5f@public.gmane.org, LTI-mKfYHFOTJHu+onKQ3Rj6Dg@public.gmane.org Return-path: In-Reply-To: <37ED64EE21A398409D609EAF7B0525A02164EB-4fFKm6SzF271qYPpFx2fzhn/E98GtLqgrE5yTffgRl4@public.gmane.org> List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Sender: spi-devel-general-bounces-5NWGOfrQmneRv+LV9MX5uipxlwaOVQ5f@public.gmane.org Errors-To: spi-devel-general-bounces-5NWGOfrQmneRv+LV9MX5uipxlwaOVQ5f@public.gmane.org List-Id: linux-spi.vger.kernel.org > I was wondering if the problem described in the thread "problem with > converting from at91_spi to atmel_spi (AT91RM9200)" from last May > was ever solved, as I'm currently running into the same trouble. > at91_spi works nicely with an at25-driven EEPROM on CS0, but atmel_spi > doesn't - it just blocks whatever thread is using it. Andrew Victor had partial success after updating the platform setup code appropriately. I don't know what the status of that patch is. > While that's no problem in the EEPROM case, a customer wants to use > SPI basically as a very fast UART between two AT91RM9200 based boards > now, on which only CS0 is available. Is there no other chance than > converting the layout to use another CS instead? Unfortunately CS0 on rm9200 seems especially full of trouble; from what I've seen and heard, using another chipselect would be good just on the basis of the number of CS0-specific errata. Of course, if you just want UART-ish behavior, you might also be able to use one of the SSC channels. The SPI stack aims to handle master/slave style communication, not peer-to-peer. When you're modeling a bidirectional byte stream, that's the wrong model. Plus, you'd have fewer chip errata to cope with. - Dave ------------------------------------------------------------------------- This SF.net email is sponsored by: Splunk Inc. Still grepping through log files to find problems? Stop. Now Search log events and configuration files using AJAX and a browser. Download your FREE copy of Splunk now >> http://get.splunk.com/