[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <CAFd5g449iu6DGvMuP8-n8C6GBbr62G6_ZYS6Lh9038EB0XO9YA@mail.gmail.com>
Date: Tue, 27 Jun 2017 16:47:45 -0700
From: Brendan Higgins <brendanhiggins@...gle.com>
To: Wolfram Sang <wsa@...-dreams.de>
Cc: Rob Herring <robh+dt@...nel.org>,
Mark Rutland <mark.rutland@....com>,
Greg KH <gregkh@...uxfoundation.org>,
"David S . Miller" <davem@...emloft.net>, mchehab@...nel.org,
Joel Stanley <joel@....id.au>, martin.petersen@...cle.com,
Benjamin Herrenschmidt <benh@...nel.crashing.org>,
linux-i2c@...r.kernel.org, devicetree <devicetree@...r.kernel.org>,
Linux Kernel Mailing List <linux-kernel@...r.kernel.org>,
OpenBMC Maillist <openbmc@...ts.ozlabs.org>
Subject: Re: [PATCH v11 3/4] i2c: aspeed: added driver for Aspeed I2C
On Tue, Jun 27, 2017 at 2:00 AM, Wolfram Sang <wsa@...-dreams.de> wrote:
>
>> > If SCL is stuck low, how do you want to send a STOP?
>> >
>>
>> Fair point. I should probably drop that in the future and just do a
>> reset, and even then, doing a
>> reset is probably just wishful thinking. If a slave is holding down
>> SCL, we are pretty screwed.
>
> Yes, you'd need to reset the client device. And that has to be done in
> the client driver. I wonder if i2c_transfer and friends should return a
> specific error value in that case... need to think about it.
>
Well, a good first step would be to make my recovery code conform to
the struct i2c_bus_recovery_info. Is i2c_generic_scl_recovery supposed
to be part of the user interface, or is it just intended to help put the
main recovery function together?
Powered by blists - more mailing lists