From mboxrd@z Thu Jan 1 00:00:00 1970 From: Jean Delvare Subject: Re: i2c-i801 retries on lost arbitration (resent: no gpg) Date: Fri, 30 Apr 2010 11:48:22 +0200 Message-ID: <20100430114822.1a5e0cf9@hyperion.delvare> References: <20100430094412.GC3553@swordfish.minsk.epam.com> Mime-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Return-path: In-Reply-To: <20100430094412.GC3553@swordfish.minsk.epam.com> Sender: linux-kernel-owner@vger.kernel.org To: Sergey Senozhatsky Cc: "Ben Dooks (embedded platforms)" , Wolfram Sang , linux-i2c@vger.kernel.org, linux-kernel@vger.kernel.org List-Id: linux-i2c@vger.kernel.org Hi Sergey, On Fri, 30 Apr 2010 12:44:12 +0300, Sergey Senozhatsky wrote: > Soory, forgot to turn-off gpg. gpg is fine. > Both i2c spec and ICH7 datasheet requires to restart transaction in > case of lost arbitration. > > "No information is lost during the arbitration process. A master that > loses the arbitration can generate clock pulses until the end of the > byte in which it loses the arbitration and must restart its transaction > when the bus is idle." > > Please see the following patch. Is it correct? No, it's not. As you wrote above, this is not specific to the Intel ICH but a general I2C issue. As such it must be handled in i2c-core and not by individual drivers. And as a matter of fact, it is already handled. Look at functions i2c_transfer and i2c_smbus_xfer in i2c-core.c, see the comments "Retry automatically on arbitration loss", the code is already there. It's just a matter of bus drivers properly setting adapter->retries (which i2c-i801.c does already.) So I don't know which kernel you're using, but with the latest upstream kernel, things should work just fine already. -- Jean Delvare