On Tue, Nov 9, 2021 at 11:21 PM Krzysztof Kozlowski <krzk@xxxxxxxxxx> wrote:
> On 18/08/2021 23:09, Palmer Dabbelt wrote:
> > On Wed, 18 Aug 2021 07:11:02 PDT (-0700), daire.mcnamara@xxxxxxxxxxxxx wrote:
> >> --- /dev/null
> >> +++ b/drivers/clk/microchip/Kconfig
> >> @@ -0,0 +1,7 @@
> >> +# SPDX-License-Identifier: GPL-2.0
> >> +
> >> +config MCHP_CLK_MPFS
> >> + bool "Clk driver for PolarFire SoC"
> >> + depends on (RISCV && SOC_MICROCHIP_POLARFIRE) || COMPILE_TEST
> >
> > This shouldn't depend on the SOC config. Those were meant to just
> > enable a set of drivers, not restrict what can be enabled.
>
> The kernel config is overwhelmed with amount of choices of drivers, so
> it is quite common to restrict the choices of subarch-specific drivers
> to these sub-architectures. That's how many, if not most, of ARM v7 and
> ARMv8 drivers are restricted. Some other RISC-V drivers follow this
> concept (git grep SOC_CANAAN, git grep SOC_SIFIVE) because it is
> friendly for people configuring kernels. Why making configuration choice
> more difficult and show the SoC-specific clock driver to every other
> platform (when this SoC is not used)?
+1
Gr{oetje,eeting}s,
Geert
--
Geert Uytterhoeven -- There's lots of Linux beyond ia32 -- geert@xxxxxxxxxxxxxx
In personal conversations with technical people, I call myself a hacker. But
when I'm talking to journalists I just say "programmer" or something like that.
-- Linus Torvalds
