Quoting Duje Mihanović (2024-04-20 06:32:56) > On 4/20/24 00:24, Stephen Boyd wrote: > > Quoting Duje Mihanović (2024-04-19 07:31:14) > >> On Friday, April 12, 2024 4:57:09 AM GMT+2 Stephen Boyd wrote: > >>> Quoting Duje Mihanović (2024-04-11 03:15:34) > >>> > >>>> On 4/11/2024 10:00 AM, Stephen Boyd wrote: > >>>>> Is there a reason this file can't be a platform driver? > >>>> > >>>> Not that I know of, I did it like this only because the other in-tree > >>>> MMP clk drivers do so. I guess the initialization should look like any > >>>> of the qcom GCC drivers then? > >>> > >>> Yes. > >> > >> With the entire clock driver code in one file this is quite messy as I also > >> needed to add module_init and module_exit functions to (un)register each > >> platform driver, presumably because the module_platform_driver macro doesn't > >> work with multiple platform drivers in one module. If I split up the driver > >> code for each clock controller block into its own file (such as clk-of- > >> pxa1908-apbc.c) as I believe is the best option, should the commits be split > >> up accordingly as well? > > > > Sure. Why is 'of' in the name? Maybe that is unnecessary? > > That seems to be a historical leftover from when Marvell was just adding > DT support to the ARM32 MMP SoCs which Rob followed along with in the > PXA1928 clk driver and so have I. Should I drop it then as Marvell has > in the PXA1908 vendor kernel? > Sounds good to me.
