[v1] USB: pxa27x_udc: check return value of clk_enable

[PATCH] USB: pxa27x_udc: check return value of clk_enable

Posted by Zhaoyang Yu 1 month, 2 weeks ago

clk_enable() may fail according to the API contract.
Previously, udc_enable() ignored its return value.
This patch checks the return value and logs an error
without continuing initialization if clk_enable fails.

Signed-off-by: Zhaoyang Yu <2426767509@qq.com>
---
 drivers/usb/gadget/udc/pxa27x_udc.c | 8 +++++++-
 1 file changed, 7 insertions(+), 1 deletion(-)

diff --git a/drivers/usb/gadget/udc/pxa27x_udc.c b/drivers/usb/gadget/udc/pxa27x_udc.c
index 897f53601b5b..69fcecaf6061 100644
--- a/drivers/usb/gadget/udc/pxa27x_udc.c
+++ b/drivers/usb/gadget/udc/pxa27x_udc.c
@@ -1696,7 +1696,13 @@ static void udc_enable(struct pxa_udc *udc)
 	if (udc->enabled)
 		return;
 
-	clk_enable(udc->clk);
+	int ret;
+
+	ret = clk_enable(udc->clk);
+	if (ret) {
+		dev_err(udc->dev, "clk_enable failed: %d\n", ret);
+		return;
+	}
 	udc_writel(udc, UDCICR0, 0);
 	udc_writel(udc, UDCICR1, 0);
 	udc_clear_mask_UDCCR(udc, UDCCR_UDE);
-- 
2.34.1

Re: [PATCH] USB: pxa27x_udc: check return value of clk_enable

Posted by Sergey Shtylyov 1 month, 2 weeks ago

On 3/1/26 7:23 PM, Zhaoyang Yu wrote:

> clk_enable() may fail according to the API contract.
> Previously, udc_enable() ignored its return value.
> This patch checks the return value and logs an error
> without continuing initialization if clk_enable fails.
> 
> Signed-off-by: Zhaoyang Yu <2426767509@qq.com>
> ---
>  drivers/usb/gadget/udc/pxa27x_udc.c | 8 +++++++-
>  1 file changed, 7 insertions(+), 1 deletion(-)
> 
> diff --git a/drivers/usb/gadget/udc/pxa27x_udc.c b/drivers/usb/gadget/udc/pxa27x_udc.c
> index 897f53601b5b..69fcecaf6061 100644
> --- a/drivers/usb/gadget/udc/pxa27x_udc.c
> +++ b/drivers/usb/gadget/udc/pxa27x_udc.c
> @@ -1696,7 +1696,13 @@ static void udc_enable(struct pxa_udc *udc)
>  	if (udc->enabled)
>  		return;
>  
> -	clk_enable(udc->clk);
> +	int ret;

   Please don't declare variables amidst of a function. IIRC, the Linux coding
style doesn't allow for this C++-ism...

[...]

MBR, Sergey