On 03/31/2016 05:21 PM, Bin Liu wrote:

quoted

-	glue = kzalloc(sizeof(*glue), GFP_KERNEL);
+	glue = devm_kzalloc(&pdev->dev, sizeof(*glue), GFP_KERNEL);
  	if (!glue) {
  		dev_err(&pdev->dev, "failed to allocate glue context\n");
-		goto err0;
+		return -ENOMEM;
  	}

-	clk = clk_get(&pdev->dev, "usb20");
+	clk = devm_clk_get(&pdev->dev, "usb20");
  	if (IS_ERR(clk)) {
  		dev_err(&pdev->dev, "failed to get clock\n");
-		ret = PTR_ERR(clk);
-		goto err3;
+		return PTR_ERR(clk);

memory leak due to not kfree(glue).

It is my understanding that since glue is allocated with devm_kzalloc(), 
that if the probe function returns and error, glue and everything else 
allocated with devm_* will be automatically freed.

If this is not the case, wouldn't devm_kfree() be the appropriate 
function instead?

quoted

@@ -576,8 +569,6 @@ static int da8xx_remove(struct platform_device *pdev)
  	platform_device_unregister(glue->musb);
  	usb_phy_generic_unregister(glue->phy);
  	clk_disable(glue->clk);
-	clk_put(glue->clk);
-	kfree(glue);

Doesn't match with $subject, I'd put them into a seperate patch.

I disagree. Since these are now automatically freed because of changes 
in the probe function, these changes belong in the same patch.