From mboxrd@z Thu Jan 1 00:00:00 1970 From: Jean Delvare Subject: Re: [PATCH v5] i2c: Adding support for Intel iSMT SMBus 2.0 host controller Date: Wed, 30 Jan 2013 22:55:57 +0100 Message-ID: <20130130225557.55b8eeee@endymion.delvare> References: <20121218150337.5b861ae3@endymion.delvare> <1359402232-21369-1-git-send-email-nhorman@tuxdriver.com> <20130129150551.498b2a9b@endymion.delvare> <20130129153253.GA14044@hmsreliant.think-freely.org> <20130129231028.41b04fc9@endymion.delvare> <20130130141504.GA2968@neilslaptop.think-freely.org> Mime-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Return-path: In-Reply-To: <20130130141504.GA2968-0o1r3XBGOEbbgkc5XkKeNuvMHUBZFtU3YPYVAmT7z5s@public.gmane.org> Sender: linux-i2c-owner-u79uwXL29TY76Z2rM5mHXA@public.gmane.org To: Neil Horman Cc: linux-i2c-u79uwXL29TY76Z2rM5mHXA@public.gmane.org, Bill Brown , Seth Heasley List-Id: linux-i2c@vger.kernel.org On Wed, 30 Jan 2013 09:15:04 -0500, Neil Horman wrote: > On Tue, Jan 29, 2013 at 11:10:28PM +0100, Jean Delvare wrote: > > This won't fix it, unless the hardware happens to update dma_size when > > receiving a block from a slave. If not then dma_size will always be > > I2C_SMBUS_BLOCK_MAX, while what we need is the number of bytes actually > > received. > > I'm sorry, I misspoke. Its the rxbytes field in the descriptor that we need to use. I > believe that field gets updated by the firmware on dmas from the device to the > cpu. I couldn't find a description of the hardware descriptor in the datasheets, so I'm afraid I can't help here. If it is really there then I'd be grateful if you can point me to the right file and page. > (...) > I'll have some tomorrow. I've got all your trivial changes fixed up, and will > hopefully be able to sort out the convoluted stuff with you and Seth in the next > few days Great. I'll be happy to test whatever you have then. -- Jean Delvare