2019-12-04 14:42:12

by Alexandre Torgue

Subject: [PATCH] pinctrl: pinmux: fix a possible null pointer in pinmux_can_be_used_for_gpio

This commit adds a check on ops pointer to avoid a kernel panic when
ops->strict is used. Indeed, on some pinctrl driver (at least for
pinctrl-stmfx) the pinmux ops is not implemented. Let's assume than gpio
can be used in this case.

Fixes: 472a61e777fe ("pinctrl/gpio: Take MUX usage into account")

Signed-off-by: Alexandre Torgue <[email protected]>

diff --git a/drivers/pinctrl/pinmux.c b/drivers/pinctrl/pinmux.c
index e914f6efd39e..9503ddf2edc7 100644
--- a/drivers/pinctrl/pinmux.c
+++ b/drivers/pinctrl/pinmux.c
@@ -85,7 +85,7 @@ bool pinmux_can_be_used_for_gpio(struct pinctrl_dev *pctldev, unsigned pin)
const struct pinmux_ops *ops = pctldev->desc->pmxops;

/* Can't inspect pin, assume it can be used */
- if (!desc)
+ if (!desc || !ops)
return true;

if (ops->strict && desc->mux_usecount)
--
2.17.1

2019-12-13 10:00:09

by Linus Walleij

Subject: Re: [PATCH] pinctrl: pinmux: fix a possible null pointer in pinmux_can_be_used_for_gpio

On Wed, Dec 4, 2019 at 3:41 PM Alexandre Torgue <[email protected]> wrote:

> This commit adds a check on ops pointer to avoid a kernel panic when
> ops->strict is used. Indeed, on some pinctrl driver (at least for
> pinctrl-stmfx) the pinmux ops is not implemented. Let's assume than gpio
> can be used in this case.
>
> Fixes: 472a61e777fe ("pinctrl/gpio: Take MUX usage into account")
> Signed-off-by: Alexandre Torgue <[email protected]>

Patch applied for fixes.

Yours,
Linus Walleij