Re: [PATCH v2] reset: core: honour optional resets when the GPIO fallback returns -ENOENT

From: Philipp Zabel

Date: Tue Nov 17 2026 - 05:30:40 EST


On Di, 2026-11-17 at 17:55 +0800, Yang Siyuan wrote:
> __fwnode_reset_control_get() honours RESET_CONTROL_FLAGS_BIT_OPTIONAL by
> returning NULL when the requested reset is not described in the device
> tree. The reset-gpio fallback breaks this: when
> __reset_add_reset_gpio_device() fails with -ENOENT, the error is returned
> even for an optional reset.
>
> This is reachable on Allwinner SoCs, whose pinctrl uses three-cell GPIO
> specifiers. The reset-gpio fallback only supports two cells, so it
> returns -ENOENT and a consumer asking for an optional reset fails to
> probe instead of falling back to its own GPIO handling. mmc-pwrseq-simple
> on the Anbernic RG35XX family hits this: its wifi_pwrseq node has
> reset-gpios but no resets, so the SDIO WiFi adapter never probes.
>
> Return NULL only for -ENOENT when the reset was requested as optional.
> A reset that is described but cannot be used is still an error, so
> -EPROBE_DEFER and all other failures are propagated.
>
> Fixes: c721f189e89c ("reset: Instantiate reset GPIO controller for shared reset-gpios")
> Cc: stable@xxxxxxxxxxxxxxx

This won't apply on top of c721f189e89c. The surroundings have most
recently been touched by commits 1f10008aff71 ("reset: use lock guards
in reset core") and 9035073d0ef1 ("reset: convert reset core to using
firmware nodes"). I suggest you drop the Fixes: and Cc:stable tags.
You can provide a backport if you need this on v7.0 or older.


regards
Philipp