[PATCH] i2c: busses: Fix error handling in synquacer_i2c_probe

[PATCH] i2c: busses: Fix error handling in synquacer_i2c_probe

[Miaoqian Lin]

In the error handling path, the clk_prepare_enable() function
call should be balanced by a corresponding 'clk_disable_unprepare()'
call, as already done in the remove function.

Fixes: 0d676a6c4390 ("i2c: add support for Socionext SynQuacer I2C controller")
Signed-off-by: Miaoqian Lin <linmq006@gmail.com>
---
 drivers/i2c/busses/i2c-synquacer.c | 24 +++++++++++++++++-------
 1 file changed, 17 insertions(+), 7 deletions(-)

diff --git a/drivers/i2c/busses/i2c-synquacer.c b/drivers/i2c/busses/i2c-synquacer.c
index e4026c5416b1..cd955224d629 100644
--- a/drivers/i2c/busses/i2c-synquacer.c
+++ b/drivers/i2c/busses/i2c-synquacer.c
@@ -569,22 +569,27 @@ static int synquacer_i2c_probe(struct platform_device *pdev)
 	    i2c->pclkrate > SYNQUACER_I2C_MAX_CLK_RATE) {
 		dev_err(&pdev->dev, "PCLK missing or out of range (%d)\n",
 			i2c->pclkrate);
-		return -EINVAL;
+		ret = -EINVAL;
+		goto err_disable_clk;
 	}
 
 	i2c->base = devm_platform_ioremap_resource(pdev, 0);
-	if (IS_ERR(i2c->base))
-		return PTR_ERR(i2c->base);
+	if (IS_ERR(i2c->base)) {
+		ret = PTR_ERR(i2c->base);
+		goto err_disable_clk;
+	}
 
 	i2c->irq = platform_get_irq(pdev, 0);
-	if (i2c->irq < 0)
-		return i2c->irq;
+	if (i2c->irq < 0) {
+		ret = i2c->irq;
+		goto err_disable_clk;
+	}
 
 	ret = devm_request_irq(&pdev->dev, i2c->irq, synquacer_i2c_isr,
 			       0, dev_name(&pdev->dev), i2c);
 	if (ret < 0) {
 		dev_err(&pdev->dev, "cannot claim IRQ %d\n", i2c->irq);
-		return ret;
+		goto err_disable_clk;
 	}
 
 	i2c->state = STATE_IDLE;
@@ -607,7 +612,7 @@ static int synquacer_i2c_probe(struct platform_device *pdev)
 	ret = i2c_add_numbered_adapter(&i2c->adapter);
 	if (ret) {
 		dev_err(&pdev->dev, "failed to add bus to i2c core\n");
-		return ret;
+		goto err_disable_clk;
 	}
 
 	platform_set_drvdata(pdev, i2c);
@@ -616,6 +621,11 @@ static int synquacer_i2c_probe(struct platform_device *pdev)
 		 dev_name(&i2c->adapter.dev));
 
 	return 0;
+
+err_disable_clk:
+	if (!IS_ERR(i2c->pclk))
+		clk_disable_unprepare(i2c->pclk);
+	return ret;
 }
 
 static int synquacer_i2c_remove(struct platform_device *pdev)
-- 
2.17.1

Re: [PATCH] i2c: busses: Fix error handling in synquacer_i2c_probe

[Miaoqian Lin]

Hi, Andy Shevchenko

On Wed, Mar 16, 2022 at 09:04:06AM +0200, Andy Shevchenko wrote:
> 
> > In the error handling path, the clk_prepare_enable() function
> > call should be balanced by a corresponding 'clk_disable_unprepare()'
> > call, as already done in the remove function.
> 
> 
> It’s not good to mix devm approach with non-devm.
> 
Thanks for your review. I'm sorry, I don't quite understand. Could you please explain more? What functions are your referring to? Thanks. 

> 
> >
> > Fixes: 0d676a6c4390 ("i2c: add support for Socionext SynQuacer I2C
> > controller")
> > Signed-off-by: Miaoqian Lin <linmq006@gmail.com>
> > ---
> >  drivers/i2c/busses/i2c-synquacer.c | 24 +++++++++++++++++-------
> >  1 file changed, 17 insertions(+), 7 deletions(-)
> >
> > diff --git a/drivers/i2c/busses/i2c-synquacer.c b/drivers/i2c/busses/i2c-
> > synquacer.c
> > index e4026c5416b1..cd955224d629 100644
> > --- a/drivers/i2c/busses/i2c-synquacer.c
> > +++ b/drivers/i2c/busses/i2c-synquacer.c
> > @@ -569,22 +569,27 @@ static int synquacer_i2c_probe(struct
> > platform_device *pdev)
> >             i2c->pclkrate > SYNQUACER_I2C_MAX_CLK_RATE) {
> >                 dev_err(&pdev->dev, "PCLK missing or out of range (%d)\n",
> >                         i2c->pclkrate);
> > -               return -EINVAL;
> > +               ret = -EINVAL;
> > +               goto err_disable_clk;
> >         }
> >
> >         i2c->base = devm_platform_ioremap_resource(pdev, 0);
> > -       if (IS_ERR(i2c->base))
> > -               return PTR_ERR(i2c->base);
> > +       if (IS_ERR(i2c->base)) {
> > +               ret = PTR_ERR(i2c->base);
> > +               goto err_disable_clk;
> > +       }
> >
> >         i2c->irq = platform_get_irq(pdev, 0);
> > -       if (i2c->irq < 0)
> > -               return i2c->irq;
> > +       if (i2c->irq < 0) {
> > +               ret = i2c->irq;
> > +               goto err_disable_clk;
> > +       }
> >
> >         ret = devm_request_irq(&pdev->dev, i2c->irq, synquacer_i2c_isr,
> >                                0, dev_name(&pdev->dev), i2c);
> >         if (ret < 0) {
> >                 dev_err(&pdev->dev, "cannot claim IRQ %d\n", i2c->irq);
> > -               return ret;
> > +               goto err_disable_clk;
> >         }
> >
> >         i2c->state = STATE_IDLE;
> > @@ -607,7 +612,7 @@ static int synquacer_i2c_probe(struct platform_device
> > *pdev)
> >         ret = i2c_add_numbered_adapter(&i2c->adapter);
> >         if (ret) {
> >                 dev_err(&pdev->dev, "failed to add bus to i2c core\n");
> > -               return ret;
> > +               goto err_disable_clk;
> >         }
> >
> >         platform_set_drvdata(pdev, i2c);
> > @@ -616,6 +621,11 @@ static int synquacer_i2c_probe(struct platform_device
> > *pdev)
> >                  dev_name(&i2c->adapter.dev));
> >
> >         return 0;
> > +
> > +err_disable_clk:
> > +       if (!IS_ERR(i2c->pclk))
> > +               clk_disable_unprepare(i2c->pclk);
> > +       return ret;
> >  }
> >
> >  static int synquacer_i2c_remove(struct platform_device *pdev)
> > --
> > 2.17.1
> >
> >
> 
> -- 
> With Best Regards,
> Andy Shevchenko

Re: [PATCH] i2c: busses: Fix error handling in synquacer_i2c_probe

[Andy Shevchenko]

On Wed, Mar 16, 2022 at 9:18 AM Miaoqian Lin <linmq006@gmail.com> wrote:
>
> Hi, Andy Shevchenko
>
> On Wed, Mar 16, 2022 at 09:04:06AM +0200, Andy Shevchenko wrote:
> >
> > > In the error handling path, the clk_prepare_enable() function
> > > call should be balanced by a corresponding 'clk_disable_unprepare()'
> > > call, as already done in the remove function.
> >
> >
> > It’s not good to mix devm approach with non-devm.
> >
> Thanks for your review. I'm sorry, I don't quite understand. Could you please explain more? What functions are your referring to? Thanks.

Your entire patch is broken. The error handling in ->probe() and
complete ->remove() now have asymmetrical ordering.

On top of that the CCF is NULL aware (and error pointer IIRC), so no
need to have additional checks.

-- 
With Best Regards,
Andy Shevchenko