Hi Simon, On 2026-09-10 Simon Glass wrote: > Pin a10 is already used by hog_input_active_low further down in this > file.
I picked a10 after grepping for '&gpio_a 10' and the hogs address their pins by bare number, so I missed it. Bank a has no free offset at all: all 25 are taken between the phandle users and the hogs on 10 to 13, and a15 to a18 are in use as well. Rather than grow the bank (three gpio tests assert its size), v2 moves the regulator to c8 in the pinmux-gpios bank, whose two cells also let the node carry the flags the way a real regulator-fixed does. > The node name regulator-fixed matches the compatible Will rename it fixed-gpio-reg and look it up under that name. > Patch 2 moves the same claim-in-probe pattern into gpio-regulator and > fixed-clock, but only fixed-regulator is covered here. fixed-clock drops out of patch 2 (see that thread). For gpio-regulator v2 adds a node with two states on c9 and a test that mirrors this one: unclaimed after of_to_plat(), output after probe(), voltage switching the pin. > gpio_get_function() returns GPIOF_UNUSED whenever the pin is not > claimed via the DM name table, regardless of the sandbox pad state. Good point. v2 also asserts sandbox_gpio_get_direction() is still input after of_to_plat() and becomes output only after probe(), so the test watches the pad and not just the name table. Regards, Mehmet
