Re: [PATCH 2/2] phy: ti-pipe3: Fix EPROBE_DEFER handling for clock resources

[Vladimir Oltean]

On Fri, May 15, 2026 at 12:07:37PM +0800, Hongling Zeng wrote:
> ti_pipe3_get_clk() has two issues with -EPROBE_DEFER error handling:
> 
> 1. When devm_clk_get() for sysclk fails, the function returns -EINVAL
>    instead of propagating the actual error code. This masks -EPROBE_DEFER
>    to -EINVAL, breaking the probe deferral mechanism and causing permanent
>    driver initialization failure on systems with non-deterministic probe
>    ordering.
> 
> 2. For SATA PHY refclk, the function ignores all errors to support older
>    DTBs missing the refclk property. However, this incorrectly ignores
>    -EPROBE_DEFER as well, causing the driver to proceed without waiting
>    for the clock provider to become available.
> 
> Fix both issues:
> - Return PTR_ERR(phy->sys_clk) instead of -EINVAL to propagate all
>   error codes including -EPROBE_DEFER
> - For SATA refclk, only ignore -ENOENT (clock not found in old DTBs)
>   but propagate other errors like -EPROBE_DEFER
> 
> Fixes: a70143bbef6b ("drivers: phy: usb3/pipe3: Adapt pipe3 driver to Generic PHY Framework")
> Fixes: 7f33912d2978 ("phy: ti-pipe3: Fix SATA across suspend/resume")
> Signed-off-by: Hongling Zeng <zenghongling@xxxxxxxxxx>
> ---
>  drivers/phy/ti/phy-ti-pipe3.c | 17 ++++++++++++-----
>  1 file changed, 12 insertions(+), 5 deletions(-)
> 
> diff --git a/drivers/phy/ti/phy-ti-pipe3.c b/drivers/phy/ti/phy-ti-pipe3.c
> index 58fbc3b27813..2bfcd0c70abd 100644
> --- a/drivers/phy/ti/phy-ti-pipe3.c
> +++ b/drivers/phy/ti/phy-ti-pipe3.c
> @@ -604,15 +604,22 @@ static int ti_pipe3_get_clk(struct ti_pipe3 *phy)
>  {
>  	struct clk *clk;
>  	struct device *dev = phy->dev;
> +	int ret;
>  
>  	phy->refclk = devm_clk_get(dev, "refclk");

When phy->mode == PIPE3_MODE_SATA, I think it would be a good idea to
call devm_clk_get_optional() which converts ENOENT to 0 for you.
Otherwise call devm_clk_get(). In both cases, you can propagate the
returned error code without special-casing anything.

>  	if (IS_ERR(phy->refclk)) {
> -		dev_err(dev, "unable to get refclk\n");
> +		ret = PTR_ERR(phy->refclk);
>  		/* older DTBs have missing refclk in SATA PHY
> -		 * so don't bail out in case of SATA PHY.
> +		 * so don't bail out for -ENOENT, but still defer
> +		 * probe for other errors like -EPROBE_DEFER.
>  		 */
> -		if (phy->mode != PIPE3_MODE_SATA)
> -			return PTR_ERR(phy->refclk);
> +		if (ret == -ENOENT) {
> +			if (phy->mode != PIPE3_MODE_SATA)
> +				return ret;
> +		} else {
> +			dev_err(dev, "unable to get refclk\n");
> +			return ret;
> +		}
>  	}
>  
>  	if (phy->mode != PIPE3_MODE_SATA) {
> @@ -629,7 +636,7 @@ static int ti_pipe3_get_clk(struct ti_pipe3 *phy)
>  		phy->sys_clk = devm_clk_get(dev, "sysclk");
>  		if (IS_ERR(phy->sys_clk)) {
>  			dev_err(dev, "unable to get sysclk\n");
> -			return -EINVAL;
> +			return PTR_ERR(phy->sys_clk);
>  		}
>  	}
>  
> -- 
> 2.25.1
> 
> 

Because of the broken threading this patch can't be applied anyway, so

pw-bot: cr