+Lee Jones On Wed, 29 Sept 2021 at 20:30, Mark Brown <broonie@xxxxxxxxxx> wrote: > > On Wed, Sep 29, 2021 at 04:20:37PM +0800, Chunyan Zhang wrote: > > On Tue, 28 Sept 2021 at 19:31, Mark Brown <broonie@xxxxxxxxxx> wrote: > > > > Since this is a part of a MFD I'd not expect it to have a compatible > > > string? > > > Since we switched to use devm_of_platform_populate() [1] to register > > MFD subdevices, compatible is required, IIUC. > > I'm not sure that's a good fit for these regulators, we don't gain any > extra information from the compatible here. Humm... regulators cannot be probed without compatible for this case, except we have a list of mfd_cells which include compatible and driver name in sc27xx_spi (mfd) driver. On different platforms, sc27xx_spi consists of different sub-devices with different compatible string, to avoid adding mfd_cells to sc27xx_spi driver each time we add a new platform support, we changed to use devm_of_platform_populate() which can automatically register sun-devices listed in devicetree. The above is my understand, please correct me if I'm missing something @Lee Jones Thanks, Chunyan [1] https://lkml.org/lkml/2020/6/19/207
