From mboxrd@z Thu Jan 1 00:00:00 1970 From: Alexander Aring Subject: Re: [PATCH] mrf24j40: seperate mrf24j40 h/w init and add checkings Date: Wed, 23 Apr 2014 16:00:09 +0200 Message-ID: <20140423140007.GA24286@omega> References: <1398258716-18319-1-git-send-email-varkab@cdac.in> <5357C565.8080009@signal11.us> Mime-Version: 1.0 Content-Type: text/plain; charset=utf-8 Cc: Varka Bhadram , linux-zigbee-devel@lists.sourceforge.net, alex.bluesman.smirnov@gmail.com, dbaryshkov@gmail.com, netdev@vger.kernel.org, davem@davemloft.net, Varka Bhadram To: Alan Ott Return-path: Received: from mail-ee0-f41.google.com ([74.125.83.41]:46133 "EHLO mail-ee0-f41.google.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1752386AbaDWOAR (ORCPT ); Wed, 23 Apr 2014 10:00:17 -0400 Received: by mail-ee0-f41.google.com with SMTP id t10so842498eei.28 for ; Wed, 23 Apr 2014 07:00:16 -0700 (PDT) Content-Disposition: inline In-Reply-To: <5357C565.8080009@signal11.us> Sender: netdev-owner@vger.kernel.org List-ID: Hi Alan, On Wed, Apr 23, 2014 at 09:51:33AM -0400, Alan Ott wrote: > On 04/23/2014 09:11 AM, Varka Bhadram wrote: > >I followed the process that you mailed earlier, thnks for that. > > > >I am expecting the mail from Alan about the changes. > > Hi Varka, > > Is there a specific problem you're seeing? Typically in the kernel we expect > the SPI controller to succeed for a couple reasons: > 1. It's part of the basic, core functionality of a system. Checking for > errors on SPI transfers is analogous to making sure RAM you wrote actually > got written. > 2. Most of the time an SPI failure is not something we can detect anyway. > (disconnect one of the lines and see what you get). > 3. The code to check for it just adds a lot of bloat without much measurable > benefit. > > I've read the above in the comments in other drivers, but I can't remember > exactly where right now. There are plenty of examples in the kernel of SPI > being done this way, as it seems to be accepted practice in the kernel. > > If there is a specific issue that you're seeing, then let's talk about it, > otherwise I'm going to NAK this change. > if somebody hasn't a right spi configuration the probe function should fail. Assumed that spi_sync will return a errno then. - Alex