Hi Lars,

On Wed, Feb 10, 2021 at 3:24 PM Lars Povlsen [off-list ref] wrote:

Geert Uytterhoeven writes:

quoted

On Wed, Feb 10, 2021 at 3:17 PM Alexandre Belloni
[off-list ref] wrote:

quoted

On 10/02/2021 14:53:01+0100, Geert Uytterhoeven wrote:

quoted

On Wed, Feb 10, 2021 at 2:45 PM Lars Povlsen [off-list ref] wrote:

quoted

Geert Uytterhoeven writes:

quoted

the Microsemi/Microchip Serial GPIO device is present only Microsemi
VCore III and Microchip Sparx5 SoCs.  Hence add a dependency on
ARCH_SPARX5 || SOC_VCOREIII, to prevent asking the user about this
driver when configuring a kernel without support for these SoCs.

Fixes: 7e5ea974e61c8dd0 ("pinctrl: pinctrl-microchip-sgpio: Add pinctrl driver for Microsemi Serial GPIO")
Signed-off-by: Geert Uytterhoeven <geert+renesas@glider.be>
---
 drivers/pinctrl/Kconfig | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/drivers/pinctrl/Kconfig b/drivers/pinctrl/Kconfig
index 113073d5f89bbf70..3b75b1d7d3d1f1b0 100644
--- a/drivers/pinctrl/Kconfig
+++ b/drivers/pinctrl/Kconfig

@@ -353,8 +353,8 @@ config PINCTRL_OCELOT

 config PINCTRL_MICROCHIP_SGPIO
        bool "Pinctrl driver for Microsemi/Microchip Serial GPIO"
-       depends on OF
-       depends on HAS_IOMEM
+       depends on OF && HAS_IOMEM
+       depends on ARCH_SPARX5 || SOC_VCOREIII || COMPILE_TEST
        select GPIOLIB
        select GPIOLIB_IRQCHIP
        select GENERIC_PINCONF

Thank you for your patch. Unfortunately, it makes it impossible to use
the driver across PCIe - which is a specifically desired configuration.

Could you add CONFIG_PCI to the || chain?

Sure.

Is PCIe the only other transport over which the register can be accessed?
Or can this also be done over e.g. SPI, like on Ocelot[1]?

[1] https://lore.kernel.org/linux-gpio/20200511145329.GV34497@piout.net/ (local)

Yes, this driver IP is also available on Ocelot (this is SOC_VCOREIII)
so this is also available over SPI.

Hence would you consider

    depends on ARCH_SPARX5 || SOC_VCOREIII || PCI || SPI || COMPILE_TEST

acceptable?  Or would that be futile, as must systems have PCI and/or
SPI enabled anyway?

Gr{oetje,eeting}s,

                        Geert

Yes, that would be fine, but as you say - not have a lot of impact.

Up to you...

OK, I'll drop this.

Gr{oetje,eeting}s,

                        Geert

-- 
Geert Uytterhoeven -- There's lots of Linux beyond ia32 -- geert@linux-m68k.org

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