On 11/3/2016 10:18 AM, Andrew Lunn wrote: >> There are a couple of things about this. Russel's work isn't part of >> the kernel yet so I can't make use of it. > > Well, you could guide it into the kernel. Part of it has already made > the way in. And i know of other platforms which would benefit from it. > >> Additionally, the I2C device is integrated into the IP of the >> network device with register addresses being offsets of the network >> device BAR so I'm not sure how I would go about getting it setup in >> order to use the i2c infrastructure. > > Have you looked at the core i2c stuff? All you need is an > i2c_algorithim structure: > > http://lxr.free-electrons.com/source/include/linux/i2c.h#L407 > > and an i2c_adaptor structure: > > http://lxr.free-electrons.com/source/include/linux/i2c.h#L532 > > and then you can call i2c_add_adapter() to register your i2c bus with > the i2c core. Embedded such an i2c driver inside another driver is not > a problem.
I looked into this but I need to investigate this more. The I2C bus for this device is dedicated to multiple instances of this device which requires obtaining a hardware mutex before using it. I see how the locking is done, but I might need additional mutexes to protect certain scenarios and it could affect performance if I have to obtain and release the hardware mutex more than is needed. I'd like to stay with what I have for now in order to enable the driver and then come back and look at what it would take to integrate this support into the driver knowing that userspace could be accessing the bus, also. Thanks, Tom > > Andrew >
