Hello Stephen, So, you should have received the latest revision [0]. It takes the approach you advised: - One driver (clk) is platform. Others (reset, pinctrl) are auxiliary drivers. - The clk driver spawns auxiliary devices, passing to it the iomem pointer using ->platform_data. - The auxdevs spawned are based on compatible match data. We don't need any info to spawn them except their name, so match data only has an optional string. No array needed even, just two pointers: plain, simple. - This means the iomem register logic is split across each driver. [0]: https://lore.kernel.org/lkml/20240620-mbly-olb-v3-0-5f29f8ca289c@xxxxxxxxxxx/ On Tue May 7, 2024 at 11:48 PM CEST, Stephen Boyd wrote: > I could see the MFD subsystem gaining support for creating child > auxiliary devices for some compatible string node, and passing those > devices a regmap. Maybe that would be preferable to having to pick a > driver subsystem to put the platform driver in. Outside of making a > general purpose framework, you could put the platform driver in > drivers/mfd and have that populate the child devices like clk, reset, > pinctrl, etc. Having one of the driver be platform and spawn others reduces the amount of boilerplate (no driver that only creates sub devices). That sounds like a nice advantage; to be contrasted with having unrelated code in subsystems (eg auxdev spawning code in drivers/clk/). Thanks for your pieces of advice Stephen, -- Théo Lebrun, Bootlin Embedded Linux and Kernel engineering https://bootlin.com
