Re: Fwd: [PATCH v2] drivers/media/platform/s5p-tv/sdo_drv.c: fix errorreturn code

[Sylwester Nawrocki]

On 09/23/2012 09:30 PM, Mauro Carvalho Chehab wrote:
> Sylwester,
> 
> Please review.
> 
> Regards,
> Mauro
> 
> -------- Mensagem original --------
> Assunto: [PATCH v2] drivers/media/platform/s5p-tv/sdo_drv.c: fix error return code
> Data: Thu,  6 Sep 2012 10:38:29 +0200
> De: Peter Senna Tschudin <peter.senna@xxxxxxxxx>
> Para: peter.senna@xxxxxxxxx, Mauro Carvalho Chehab <mchehab@xxxxxxxxxxxxx>
> CC: linux-media@xxxxxxxxxxxxxxx, linux-kernel@xxxxxxxxxxxxxxx
> 
> From: Peter Senna Tschudin <peter.senna@xxxxxxxxx>
> 
> Convert a nonnegative error return code to a negative one, as returned
> elsewhere in the function.
> 
> A simplified version of the semantic match that finds this problem is as
> follows: (http://coccinelle.lip6.fr/)
> 
> // <smpl>
> (
> if@p1 (\(ret < 0\|ret != 0\))
>  { ... return ret; }
> |
> ret@p1 = 0
> )
> ... when != ret = e1
>     when != &ret
> *if(...)
> {
>   ... when != ret = e2
>       when forall
>  return ret;
> }
> 
> // </smpl>
> 
> Signed-off-by: Peter Senna Tschudin <peter.senna@xxxxxxxxx>
> 
> ---
>  drivers/media/platform/s5p-tv/sdo_drv.c |    3 +++
>  1 file changed, 3 insertions(+)
> 
> diff --git a/drivers/media/platform/s5p-tv/sdo_drv.c b/drivers/media/platform/s5p-tv/sdo_drv.c
> index ad68bbe..58cf56d 100644
> --- a/drivers/media/platform/s5p-tv/sdo_drv.c
> +++ b/drivers/media/platform/s5p-tv/sdo_drv.c
> @@ -369,6 +369,7 @@ static int __devinit sdo_probe(struct platform_device *pdev)
>  	sdev->fout_vpll = clk_get(dev, "fout_vpll");
>  	if (IS_ERR_OR_NULL(sdev->fout_vpll)) {
>  		dev_err(dev, "failed to get clock 'fout_vpll'\n");
> +		ret = -ENXIO;

This patch improves the situation but it doesn't look like a proper fix
to me. First of all clk_get() return code should be checked only with
IS_ERR() and the error code should be propagated up to the caller, rather
than overwriting it with -ENXIO. It's important especially in cases where,
e.g we are getting EPROBE_DEFER instead of a valid resource.

>  		goto fail_dacphy;
>  	}
>  	dev_info(dev, "fout_vpll.rate = %lu\n", clk_get_rate(sclk_vpll));
> @@ -377,11 +378,13 @@ static int __devinit sdo_probe(struct platform_device *pdev)
>  	sdev->vdac = devm_regulator_get(dev, "vdd33a_dac");
>  	if (IS_ERR_OR_NULL(sdev->vdac)) {
>  		dev_err(dev, "failed to get regulator 'vdac'\n");
> +		ret = -ENXIO;

Same situation here..

>  		goto fail_fout_vpll;
>  	}
>  	sdev->vdet = devm_regulator_get(dev, "vdet");
>  	if (IS_ERR_OR_NULL(sdev->vdet)) {
>  		dev_err(dev, "failed to get regulator 'vdet'\n");
> +		ret = -ENXIO;
>  		goto fail_fout_vpll;
>  	}

end here.

I think something more like this is needed (I'm going to queue this
patch for 3.7):

8<------------------------------------------------------------------