[PATCH -next] mmc: aspeed: Fix return value check in aspeed_sdc_probe()

[PATCH -next] mmc: aspeed: Fix return value check in aspeed_sdc_probe()

[Wei Yongjun]

In case of error, the function of_platform_device_create() returns
NULL pointer not ERR_PTR(). The IS_ERR() test in the return value
check should be replaced with NULL test.

Fixes: 09eed7fffd33 ("mmc: Add support for the ASPEED SD controller")
Signed-off-by: Wei Yongjun <weiyongjun1@huawei.com>
---
 drivers/mmc/host/sdhci-of-aspeed.c | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/drivers/mmc/host/sdhci-of-aspeed.c b/drivers/mmc/host/sdhci-of-aspeed.c
index 8bb095ca2fa9..d5acb5afc50f 100644
--- a/drivers/mmc/host/sdhci-of-aspeed.c
+++ b/drivers/mmc/host/sdhci-of-aspeed.c
@@ -261,9 +261,9 @@ static int aspeed_sdc_probe(struct platform_device *pdev)
 		struct platform_device *cpdev;
 
 		cpdev = of_platform_device_create(child, NULL, &pdev->dev);
-		if (IS_ERR(cpdev)) {
+		if (!cpdev) {
 			of_node_put(child);
-			ret = PTR_ERR(cpdev);
+			ret = -ENODEV;
 			goto err_clk;
 		}
 	}

[PATCH -next] mmc: aspeed: Fix return value check in aspeed_sdc_probe()

[Dan Carpenter]

> Fixes: 09eed7fffd33 ("mmc: Add support for the ASPEED SD controller")
                        ^^^^
When we're adding new files, could we use the prefix for the new driver
instead of just the subsystem?  "mmc: aspeed: Add new driver"?
Otherwise it's tricky to know what people want for the driver.

I just wrote this same patch and I swear I would have sent my patch
earlier but I spent hours thinking about the patch prefix and then the
census people came to the house and delayed me even more.

regards,
dan carpenter

Re: [PATCH -next] mmc: aspeed: Fix return value check in aspeed_sdc_probe()

[Andrew Jeffery]

On Mon, 26 Aug 2019, at 22:34, Dan Carpenter wrote:
> > Fixes: 09eed7fffd33 ("mmc: Add support for the ASPEED SD controller")
>                         ^^^^
> When we're adding new files, could we use the prefix for the new driver
> instead of just the subsystem?  "mmc: aspeed: Add new driver"?
> Otherwise it's tricky to know what people want for the driver.

I don't have any issue with the request, but I don't understand this last
bit. What do you mean by "it's tricky to know what people want for the
driver"?

Andrew

Re: [PATCH -next] mmc: aspeed: Fix return value check in aspeed_sdc_probe()

[Andrew Jeffery]

On Mon, 26 Aug 2019, at 21:27, Wei Yongjun wrote:
> In case of error, the function of_platform_device_create() returns
> NULL pointer not ERR_PTR(). The IS_ERR() test in the return value
> check should be replaced with NULL test.
> 
> Fixes: 09eed7fffd33 ("mmc: Add support for the ASPEED SD controller")
> Signed-off-by: Wei Yongjun <weiyongjun1@huawei.com>
> ---
>  drivers/mmc/host/sdhci-of-aspeed.c | 4 ++--
>  1 file changed, 2 insertions(+), 2 deletions(-)
> 
> diff --git a/drivers/mmc/host/sdhci-of-aspeed.c 
> b/drivers/mmc/host/sdhci-of-aspeed.c
> index 8bb095ca2fa9..d5acb5afc50f 100644
> --- a/drivers/mmc/host/sdhci-of-aspeed.c
> +++ b/drivers/mmc/host/sdhci-of-aspeed.c
> @@ -261,9 +261,9 @@ static int aspeed_sdc_probe(struct platform_device 
> *pdev)
>  		struct platform_device *cpdev;
>  
>  		cpdev = of_platform_device_create(child, NULL, &pdev->dev);
> -		if (IS_ERR(cpdev)) {
> +		if (!cpdev) {
>  			of_node_put(child);
> -			ret = PTR_ERR(cpdev);
> +			ret = -ENODEV;
>  			goto err_clk;
>  		}
>  	}

I ... have no idea why I wrote it that way. I must have just assumed it returned
an ERR_PTR(). Thanks for finding/fixing that.

Reviewed-by: Andrew Jeffery <andrew@aj.id.au>

[PATCH -next] mmc: aspeed: Fix return value check in aspeed_sdc_probe()

[Julia Lawall]

On Tue, 27 Aug 2019, Andrew Jeffery wrote:

>
>
> On Mon, 26 Aug 2019, at 22:34, Dan Carpenter wrote:
> > > Fixes: 09eed7fffd33 ("mmc: Add support for the ASPEED SD controller")
> >                         ^^^^
> > When we're adding new files, could we use the prefix for the new driver
> > instead of just the subsystem?  "mmc: aspeed: Add new driver"?
> > Otherwise it's tricky to know what people want for the driver.
>
> I don't have any issue with the request, but I don't understand this last
> bit. What do you mean by "it's tricky to know what people want for the
> driver"?

There is no obvious algorithm that tells how to go from a file name to an
appropriate subject line prefix.

julia

[PATCH -next] mmc: aspeed: Fix return value check in aspeed_sdc_probe()

[Ulf Hansson]

On Tue, 27 Aug 2019 at 02:47, Julia Lawall <julia.lawall@lip6.fr> wrote:
>
>
>
> On Tue, 27 Aug 2019, Andrew Jeffery wrote:
>
> >
> >
> > On Mon, 26 Aug 2019, at 22:34, Dan Carpenter wrote:
> > > > Fixes: 09eed7fffd33 ("mmc: Add support for the ASPEED SD controller")
> > >                         ^^^^
> > > When we're adding new files, could we use the prefix for the new driver
> > > instead of just the subsystem?  "mmc: aspeed: Add new driver"?
> > > Otherwise it's tricky to know what people want for the driver.
> >
> > I don't have any issue with the request, but I don't understand this last
> > bit. What do you mean by "it's tricky to know what people want for the
> > driver"?
>
> There is no obvious algorithm that tells how to go from a file name to an
> appropriate subject line prefix.

For MMC we normally use the name of the host driver file (excluding
".c") as part of the prefix.

For this case that means I amended the header into: mmc:
sdhci-of-aspeed: Fix return value check in aspeed_sdc_probe() and
applied it for next.

I also took the liberty to change this for the other related patches
for the "aspeed" driver to follow the same pattern.

Kind regards
Uffe