Re: [PATCH] pinctrl: mcp23s08: Propagate reset GPIO errors

[sashiko-bot@kernel.org]

Thank you for your contribution! Sashiko AI review found 1 potential issue(s) to consider:
- [High] pinctrl: uninitialized variable ret used in dev_err_probe() on error path
--

commit 102caea02ef327a69988ba143522defe7312fabf
Author: bui duc phuc <phucduc.bui@gmail.com>

pinctrl: mcp23s08: Propagate reset GPIO errors

This commit updates mcp23s08_probe_one() to catch and propagate errors
encountered when requesting the reset GPIO. It adds an error check to avoid
silently ignoring issues like deferred probing.

> diff --git a/drivers/pinctrl/pinctrl-mcp23s08.c b/drivers/pinctrl/pinctrl-mcp23s08.c
> index cc668f84bf8bc..fab6d19a559e3 100644
> --- a/drivers/pinctrl/pinctrl-mcp23s08.c
> +++ b/drivers/pinctrl/pinctrl-mcp23s08.c
> @@ -613,6 +613,8 @@ int mcp23s08_probe_one(struct mcp23s08 *mcp, struct device *dev,
>  	mcp->chip.owner = THIS_MODULE;
>  
>  	mcp->reset_gpio = devm_gpiod_get_optional(dev, "reset", GPIOD_OUT_LOW);
> +	if (IS_ERR(mcp->reset_gpio))
> +		return dev_err_probe(dev, ret, "failed to get reset GPIO\n");

[Severity: High]
Is ret uninitialized when used here? Looking at mcp23s08_probe_one() in
drivers/pinctrl/pinctrl-mcp23s08.c, ret is declared at the top of the
function but does not appear to be assigned a value before this point.

If devm_gpiod_get_optional() fails, will this return a garbage value
from the stack instead of the actual error code? Would passing
PTR_ERR(mcp->reset_gpio) to dev_err_probe() instead of ret correctly
propagate the error?

>  
>  	/* verify MCP_IOCON.SEQOP = 0, so sequential reads work,
>  	 * and MCP_IOCON.HAEN = 1, so we work with all chips.

-- 
Sashiko AI review · https://sashiko.dev/#/patchset/20261009112634.33436-1-phucduc.bui@gmail.com?part=1