On Thu, Aug 20, 2015 at 06:38:51PM +0200, Andrew Lunn wrote: > > It's true that this is something that we might have overlooked. Is it > > expected to maintain that compatibility when moving a driver from one > > framework to another (and this is a real question, not a troll)? > > Yes. There will be user space applications reading from the eeprom > file in /sys. In fact, until the NVMEM framework arrived, it was not > easy to access the eeprom from kernel space, meaning the majority of > users must of been user space... Ack. > > If so, we might provide a compatibility layer to add the former file > > too, protected by a kconfig option maybe ? > > There is one other detail you might of missed. Both AT24 and AT25 do > have an in kernel API. In the at24_platform_data you can have a > callback function "setup" which gets called when the device is > probed. setup() is called with a struct memory_accessor which contains > function pointers for reading and writing to the EEPROM. A few > platforms use these for getting the MAC address out of the EEPROM. > And these platforms are old style, not DT. Actually, we took it into account. The in-kernel API is even a big chunk of the framework. The only thing we still need to figure out is what interface we need to register cells statically. AT25's memory accessor can be removed, there's no users for it. The only user of the AT24 is some omap l138 boards is mach-davinci. Maxime -- Maxime Ripard, Free Electrons Embedded Linux, Kernel and Android engineering http://free-electrons.com
Attachment:
signature.asc
Description: Digital signature
