Re: [PATCH v4 6/6] regulator: mt6359: Add proper ldo_vcn33_[12] regulators

[sashiko-bot@kernel.org]

Thank you for your contribution! Sashiko AI review found 1 potential issue(s) to consider:
- [High] The `mt6359_noop_ops` struct is empty, which causes the dummy regulators to be treated as always-on and prevents them from being disabled, permanently locking the shared `ldo_vcn33_[12]` supply in the enabled state.
--

commit fb19e47461ecd9ff9e40d149eebb44edec04294e
Author: Chen-Yu Tsai <wenst@chromium.org>

regulator: mt6359: Add proper ldo_vcn33_[12] regulators

This commit replaces the separate WiFi and BT regulators for ldo_vcn33_1 and
ldo_vcn33_2 with single regulators. It retains backward compatibility by
converting the existing regulators to no-op regulators that are fed from their
respective new parent regulators.

> diff --git a/drivers/regulator/mt6359-regulator.c b/drivers/regulator/mt6359-regulator.c
> index 46cafe93b24e4..af0e0339fbdd6 100644
> --- a/drivers/regulator/mt6359-regulator.c
> +++ b/drivers/regulator/mt6359-regulator.c
[ ... ]
> @@ -475,6 +489,9 @@ static const struct regulator_ops mt6359p_vemc_ops = {
>  	.get_status = mt6359_get_status,
>  };
>  
> +/* Used for backward-compatible placeholder regulators */
> +static const struct regulator_ops mt6359_noop_ops = {};

Could this empty structure cause the parent supply to remain enabled
permanently?

Without an .is_enabled callback, _regulator_is_enabled() defaults to
returning 1. During registration, set_supply() checks if the regulator is
enabled. Since it appears to be enabled, it proactively calls
regulator_enable() on the parent supply (ldo_vcn33_1 or ldo_vcn33_2),
incrementing its use_count.

Later, if a consumer calls regulator_disable() on the dummy regulator,
_regulator_do_disable() returns -EINVAL because the .disable callback is
NULL. Due to this error, _regulator_disable() aborts early without
decrementing its own use_count or calling regulator_disable() on the parent
supply.

Would it make sense to add dummy .enable, .disable, and .is_enabled callbacks
that return 0 to mt6359_noop_ops?

-- 
Sashiko AI review · https://sashiko.dev/#/patchset/20260514091520.2718987-1-wenst@chromium.org?part=6