[PATCH] dmaengine: fsl-edma: disable clks in case of failure in probe

[PATCH] dmaengine: fsl-edma: disable clks in case of failure in probe

[Alexey Khoroshilov]

fsl_edma_probe() does not disable already enabled clocks
in case of failure. The patch fixes that.

Found by Linux Driver Verification project (linuxtesting.org).

Signed-off-by: Alexey Khoroshilov <khoroshilov@ispras.ru>
---
 drivers/dma/fsl-edma.c | 27 +++++++++++++++++++--------
 1 file changed, 19 insertions(+), 8 deletions(-)

diff --git a/drivers/dma/fsl-edma.c b/drivers/dma/fsl-edma.c
index be2e62b87948..58271302e43d 100644
--- a/drivers/dma/fsl-edma.c
+++ b/drivers/dma/fsl-edma.c
@@ -885,20 +885,23 @@ static int fsl_edma_probe(struct platform_device *pdev)
 
 		res = platform_get_resource(pdev, IORESOURCE_MEM, 1 + i);
 		fsl_edma->muxbase[i] = devm_ioremap_resource(&pdev->dev, res);
-		if (IS_ERR(fsl_edma->muxbase[i]))
-			return PTR_ERR(fsl_edma->muxbase[i]);
+		if (IS_ERR(fsl_edma->muxbase[i])) {
+			ret = PTR_ERR(fsl_edma->muxbase[i]);
+			goto err_clk;
+		}
 
 		sprintf(clkname, "dmamux%d", i);
 		fsl_edma->muxclk[i] = devm_clk_get(&pdev->dev, clkname);
 		if (IS_ERR(fsl_edma->muxclk[i])) {
 			dev_err(&pdev->dev, "Missing DMAMUX block clock.\n");
-			return PTR_ERR(fsl_edma->muxclk[i]);
+			ret = PTR_ERR(fsl_edma->muxclk[i]);
+			goto err_clk;
 		}
 
 		ret = clk_prepare_enable(fsl_edma->muxclk[i]);
 		if (ret) {
 			dev_err(&pdev->dev, "DMAMUX clk block failed.\n");
-			return ret;
+			goto err_clk;
 		}
 
 	}
@@ -923,7 +926,7 @@ static int fsl_edma_probe(struct platform_device *pdev)
 	edma_writel(fsl_edma, ~0, fsl_edma->membase + EDMA_INTR);
 	ret = fsl_edma_irq_init(pdev, fsl_edma);
 	if (ret)
-		return ret;
+		goto err_allclk;
 
 	dma_cap_set(DMA_PRIVATE, fsl_edma->dma_dev.cap_mask);
 	dma_cap_set(DMA_SLAVE, fsl_edma->dma_dev.cap_mask);
@@ -952,20 +955,28 @@ static int fsl_edma_probe(struct platform_device *pdev)
 	ret = dma_async_device_register(&fsl_edma->dma_dev);
 	if (ret) {
 		dev_err(&pdev->dev, "Can't register Freescale eDMA engine.\n");
-		return ret;
+		goto err_allclk;
 	}
 
 	ret = of_dma_controller_register(np, fsl_edma_xlate, fsl_edma);
 	if (ret) {
 		dev_err(&pdev->dev, "Can't register Freescale eDMA of_dma.\n");
-		dma_async_device_unregister(&fsl_edma->dma_dev);
-		return ret;
+		goto err_async;
 	}
 
 	/* enable round robin arbitration */
 	edma_writel(fsl_edma, EDMA_CR_ERGA | EDMA_CR_ERCA, fsl_edma->membase + EDMA_CR);
 
 	return 0;
+
+err_async:
+	dma_async_device_unregister(&fsl_edma->dma_dev);
+err_allclk:
+	i = DMAMUX_NR;
+err_clk:
+	while (--i >= 0)
+		clk_disable_unprepare(fsl_edma->muxclk[i]);
+	return ret;
 }
 
 static int fsl_edma_remove(struct platform_device *pdev)
-- 
1.9.1

Re: [PATCH] dmaengine: fsl-edma: disable clks in case of failure in probe

[Vinod Koul]

On Sat, Mar 12, 2016 at 12:28:39AM +0300, Alexey Khoroshilov wrote:
> fsl_edma_probe() does not disable already enabled clocks
> in case of failure. The patch fixes that.
> 
> Found by Linux Driver Verification project (linuxtesting.org).

The typical method to give this credit is Reported-by: ...

In this case you may use the mail id which sent the error report to you, see
examples of kbuild:

Reported-by: kbuild test robot <fengguang.wu@intel.com>

> 
> Signed-off-by: Alexey Khoroshilov <khoroshilov@ispras.ru>
> ---
>  drivers/dma/fsl-edma.c | 27 +++++++++++++++++++--------
>  1 file changed, 19 insertions(+), 8 deletions(-)
> 
> diff --git a/drivers/dma/fsl-edma.c b/drivers/dma/fsl-edma.c
> index be2e62b87948..58271302e43d 100644
> --- a/drivers/dma/fsl-edma.c
> +++ b/drivers/dma/fsl-edma.c
> @@ -885,20 +885,23 @@ static int fsl_edma_probe(struct platform_device *pdev)
>  
>  		res = platform_get_resource(pdev, IORESOURCE_MEM, 1 + i);
>  		fsl_edma->muxbase[i] = devm_ioremap_resource(&pdev->dev, res);
> -		if (IS_ERR(fsl_edma->muxbase[i]))
> -			return PTR_ERR(fsl_edma->muxbase[i]);
> +		if (IS_ERR(fsl_edma->muxbase[i])) {
> +			ret = PTR_ERR(fsl_edma->muxbase[i]);
> +			goto err_clk;
> +		}
>  
>  		sprintf(clkname, "dmamux%d", i);
>  		fsl_edma->muxclk[i] = devm_clk_get(&pdev->dev, clkname);
>  		if (IS_ERR(fsl_edma->muxclk[i])) {
>  			dev_err(&pdev->dev, "Missing DMAMUX block clock.\n");
> -			return PTR_ERR(fsl_edma->muxclk[i]);
> +			ret = PTR_ERR(fsl_edma->muxclk[i]);
> +			goto err_clk;
>  		}
>  
>  		ret = clk_prepare_enable(fsl_edma->muxclk[i]);
>  		if (ret) {
>  			dev_err(&pdev->dev, "DMAMUX clk block failed.\n");
> -			return ret;
> +			goto err_clk;
>  		}
>  
>  	}
> @@ -923,7 +926,7 @@ static int fsl_edma_probe(struct platform_device *pdev)
>  	edma_writel(fsl_edma, ~0, fsl_edma->membase + EDMA_INTR);
>  	ret = fsl_edma_irq_init(pdev, fsl_edma);
>  	if (ret)
> -		return ret;
> +		goto err_allclk;
>  
>  	dma_cap_set(DMA_PRIVATE, fsl_edma->dma_dev.cap_mask);
>  	dma_cap_set(DMA_SLAVE, fsl_edma->dma_dev.cap_mask);
> @@ -952,20 +955,28 @@ static int fsl_edma_probe(struct platform_device *pdev)
>  	ret = dma_async_device_register(&fsl_edma->dma_dev);
>  	if (ret) {
>  		dev_err(&pdev->dev, "Can't register Freescale eDMA engine.\n");
> -		return ret;
> +		goto err_allclk;
>  	}
>  
>  	ret = of_dma_controller_register(np, fsl_edma_xlate, fsl_edma);
>  	if (ret) {
>  		dev_err(&pdev->dev, "Can't register Freescale eDMA of_dma.\n");
> -		dma_async_device_unregister(&fsl_edma->dma_dev);
> -		return ret;
> +		goto err_async;
>  	}
>  
>  	/* enable round robin arbitration */
>  	edma_writel(fsl_edma, EDMA_CR_ERGA | EDMA_CR_ERCA, fsl_edma->membase + EDMA_CR);
>  
>  	return 0;
> +
> +err_async:
> +	dma_async_device_unregister(&fsl_edma->dma_dev);
> +err_allclk:
> +	i = DMAMUX_NR;
> +err_clk:
> +	while (--i >= 0)
> +		clk_disable_unprepare(fsl_edma->muxclk[i]);
> +	return ret;
>  }
>  
>  static int fsl_edma_remove(struct platform_device *pdev)
> -- 
> 1.9.1
> 
> --
> To unsubscribe from this list: send the line "unsubscribe dmaengine" in
> the body of a message to majordomo@vger.kernel.org
> More majordomo info at  http://vger.kernel.org/majordomo-info.html

-- 
~Vinod