Re: [PATCH 2/2] hwmon: (w83627ehf) Fix memory leak in probe function

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

 



On Mon, 2012-03-12 at 13:58 -0400, Jean Delvare wrote:
> On Mon, 12 Mar 2012 08:43:25 -0700, Guenter Roeck wrote:
> > The driver probe function leaked memory if creating the cpu0_vid attribute file
> > failed. Fix by converting the driver to use devm_kzalloc.
> > 
> > Signed-off-by: Guenter Roeck <linux@xxxxxxxxxxxx>
> > ---
> >  drivers/hwmon/w83627ehf.c |    7 +++----
> >  1 files changed, 3 insertions(+), 4 deletions(-)
> > 
> > diff --git a/drivers/hwmon/w83627ehf.c b/drivers/hwmon/w83627ehf.c
> > index 378142c..60e929c 100644
> > --- a/drivers/hwmon/w83627ehf.c
> > +++ b/drivers/hwmon/w83627ehf.c
> > @@ -2021,7 +2021,8 @@ static int __devinit w83627ehf_probe(struct platform_device *pdev)
> >  		goto exit;
> >  	}
> >  
> > -	data = kzalloc(sizeof(struct w83627ehf_data), GFP_KERNEL);
> > +	data = devm_kzalloc(&pdev->dev, sizeof(struct w83627ehf_data),
> > +			    GFP_KERNEL);
> >  	if (!data) {
> >  		err = -ENOMEM;
> >  		goto exit_release;
> > @@ -2514,9 +2515,8 @@ static int __devinit w83627ehf_probe(struct platform_device *pdev)
> >  
> >  exit_remove:
> >  	w83627ehf_device_remove_files(dev);
> > -	kfree(data);
> > -	platform_set_drvdata(pdev, NULL);
> >  exit_release:
> > +	platform_set_drvdata(pdev, NULL);
> >  	release_region(res->start, IOREGION_LENGTH);
> >  exit:
> >  	return err;
> > @@ -2530,7 +2530,6 @@ static int __devexit w83627ehf_remove(struct platform_device *pdev)
> >  	w83627ehf_device_remove_files(&pdev->dev);
> >  	release_region(data->addr, IOREGION_LENGTH);
> >  	platform_set_drvdata(pdev, NULL);
> > -	kfree(data);
> >  
> >  	return 0;
> >  }
> 
> I'm not familiar with devm but this looks correct.

Hi Jean,

I just noticed it recently, as more and mode drivers are converted to
use it. It auto-frees memory when the device is released. It typically
reduces driver size by some 80+ bytes, and it ensures that there are no
memory leaks. Quite convenient.

The only thing I don't know for sure is if platform_set_drvdata(pdev,
NULL); is still needed. It doesn't hurt, though, so I don't remove it.

Guenter



_______________________________________________
lm-sensors mailing list
lm-sensors@xxxxxxxxxxxxxx
http://lists.lm-sensors.org/mailman/listinfo/lm-sensors


[Index of Archives]     [Linux Kernel]     [Linux Hardware Monitoring]     [Linux USB Devel]     [Linux Audio Users]     [Linux Kernel]     [Linux SCSI]     [Yosemite Backpacking]

  Powered by Linux