From mboxrd@z Thu Jan 1 00:00:00 1970 From: Arnd Bergmann Subject: Re: [PATCH 1/1] mmc:Extension of MMC Block IOCTL Command support for testing of non read/write Commands Date: Thu, 17 Nov 2011 14:41:48 +0000 Message-ID: <201111171441.48504.arnd@arndb.de> References: <1321441734-24191-1-git-send-email-shashidharh@vayavyalabs.com> Mime-Version: 1.0 Content-Type: Text/Plain; charset="iso-8859-15" Content-Transfer-Encoding: 7bit Return-path: Received: from moutng.kundenserver.de ([212.227.17.8]:53200 "EHLO moutng.kundenserver.de" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1753900Ab1KQOmD (ORCPT ); Thu, 17 Nov 2011 09:42:03 -0500 In-Reply-To: <1321441734-24191-1-git-send-email-shashidharh@vayavyalabs.com> Sender: linux-mmc-owner@vger.kernel.org List-Id: linux-mmc@vger.kernel.org To: Shashidhar Hiremath Cc: Venkatraman S , Linus Walleij , Ulf Hansson , Jaehoon Chung , Philip Rakity , Girish K S , Lucas De Marchi , Aries Lee , Arindam Nath , Zhangfei Gao , Kyungmin Park , Andrei Warkentin , Chris Ball , linux-mmc@vger.kernel.org, pk , Sandeep , Rayagond On Wednesday 16 November 2011, Shashidhar Hiremath wrote: > The patch provides An infrastructure to test commands other than Read/Write commands using the IOCTL interface.The Patch can be useful incase of validating the device to verify whether device support a given command or not. The result of the sent command will be written to the result element of the mmc_ioc_cmd structure passed through IOCTL interface. > > Signed-off-by: Shashidhar Hiremath Hi Shashidhar, Unfortunately, I have to tell you that an implementation like this is inappropriate for a number of reasons. My feeling is that most importantly you should instead have the test logic entirely in user space and only call the existing ioctl in the kernel. This would make the entire patch obsolete, so I'm not commenting on the other problems. What has lead you to implementing the individual test cases in the kernel? Arnd