[PATCH] net: trivial: fix fsl fec_main mdio write return value

[PATCH] net: trivial: fix fsl fec_main mdio write return value

[Albert ARIBAUD (3ADEV)]

If the MDIO enf-of-transfer wait loop does not time out,
0 should be returned to the caller through ret, but ret
was not reset to 0 after receiving the return value of
pm_runtime_get_sync(). Reset ret when no time out occurred.

Signed-off-by: Albert ARIBAUD (3ADEV) <albert.aribaud@3adev.fr>
---
 drivers/net/ethernet/freescale/fec_main.c | 2 ++
 1 file changed, 2 insertions(+)

diff --git a/drivers/net/ethernet/freescale/fec_main.c b/drivers/net/ethernet/freescale/fec_main.c
index 01f7e81..0413c05 100644
--- a/drivers/net/ethernet/freescale/fec_main.c
+++ b/drivers/net/ethernet/freescale/fec_main.c
@@ -1818,6 +1818,8 @@ static int fec_enet_mdio_write(struct mii_bus *bus, int mii_id, int regnum,
 		netdev_err(fep->netdev, "MDIO write timeout\n");
 		ret  = -ETIMEDOUT;
 	}
+	else
+		ret = 0;
 
 	pm_runtime_mark_last_busy(dev);
 	pm_runtime_put_autosuspend(dev);
-- 
2.9.3

Re: [PATCH] net: trivial: fix fsl fec_main mdio write return value

[Albert ARIBAUD]

... and of course I only see the typo /now/...

Le Mon, 26 Sep 2016 21:11:40 +0200, "Albert ARIBAUD (3ADEV)"
<albert.aribaud@3adev.fr> a écrit :

> If the MDIO enf-of-transfer wait loop does not time out,
              ~~~~~~~~~~~~~~~

Make that "end-of-transfer".

Should I post a v2?

Cordialement,
Albert ARIBAUD
3ADEV

Re: [PATCH] net: trivial: fix fsl fec_main mdio write return value

[Fabio Estevam]

Hi Albert,

On Mon, Sep 26, 2016 at 4:11 PM, Albert ARIBAUD (3ADEV)
<albert.aribaud@3adev.fr> wrote:
> If the MDIO enf-of-transfer wait loop does not time out,
> 0 should be returned to the caller through ret, but ret
> was not reset to 0 after receiving the return value of
> pm_runtime_get_sync(). Reset ret when no time out occurred.Is this really needed?

Commit 42ea4457 ("net: fec: normalize return value of
pm_runtime_get_sync() in MDIO write") does the following:

diff --git a/drivers/net/ethernet/freescale/fec_main.c
b/drivers/net/ethernet/freescale/fec_main.c
index 91925e3..6cc3340 100644
--- a/drivers/net/ethernet/freescale/fec_main.c
+++ b/drivers/net/ethernet/freescale/fec_main.c
@@ -1816,11 +1816,13 @@ static int fec_enet_mdio_write(struct mii_bus
*bus, int mii_id, int regnum,
        struct fec_enet_private *fep = bus->priv;
        struct device *dev = &fep->pdev->dev;
        unsigned long time_left;
-       int ret = 0;
+       int ret;

        ret = pm_runtime_get_sync(dev);
        if (ret < 0)
                return ret;
+       else
+               ret = 0;

> diff --git a/drivers/net/ethernet/freescale/fec_main.c b/drivers/net/ethernet/freescale/fec_main.c
> index 01f7e81..0413c05 100644
> --- a/drivers/net/ethernet/freescale/fec_main.c
> +++ b/drivers/net/ethernet/freescale/fec_main.c
> @@ -1818,6 +1818,8 @@ static int fec_enet_mdio_write(struct mii_bus *bus, int mii_id, int regnum,
>                 netdev_err(fep->netdev, "MDIO write timeout\n");
>                 ret  = -ETIMEDOUT;
>         }
> +       else
> +               ret = 0;

, so is this really needed? ret should be 0 at this point.

Re: [PATCH] net: trivial: fix fsl fec_main mdio write return value

[Albert ARIBAUD]

Hi Fabio,

You are correct -- I'd rebased my patch as part of a series, and I did
not notice that the fix was already there. Apologies and thanks for
pointing it out.

Le Mon, 26 Sep 2016 18:00:32 -0300, Fabio Estevam <festevam@gmail.com>
a écrit :

> Hi Albert,
> 
> On Mon, Sep 26, 2016 at 4:11 PM, Albert ARIBAUD (3ADEV)
> <albert.aribaud@3adev.fr> wrote:
> > If the MDIO enf-of-transfer wait loop does not time out,
> > 0 should be returned to the caller through ret, but ret
> > was not reset to 0 after receiving the return value of
> > pm_runtime_get_sync(). Reset ret when no time out occurred.Is this really needed?  
> 
> Commit 42ea4457 ("net: fec: normalize return value of
> pm_runtime_get_sync() in MDIO write") does the following:
> 
> diff --git a/drivers/net/ethernet/freescale/fec_main.c
> b/drivers/net/ethernet/freescale/fec_main.c
> index 91925e3..6cc3340 100644
> --- a/drivers/net/ethernet/freescale/fec_main.c
> +++ b/drivers/net/ethernet/freescale/fec_main.c
> @@ -1816,11 +1816,13 @@ static int fec_enet_mdio_write(struct mii_bus
> *bus, int mii_id, int regnum,
>         struct fec_enet_private *fep = bus->priv;
>         struct device *dev = &fep->pdev->dev;
>         unsigned long time_left;
> -       int ret = 0;
> +       int ret;
> 
>         ret = pm_runtime_get_sync(dev);
>         if (ret < 0)
>                 return ret;
> +       else
> +               ret = 0;
> 
> > diff --git a/drivers/net/ethernet/freescale/fec_main.c
> > b/drivers/net/ethernet/freescale/fec_main.c index 01f7e81..0413c05
> > 100644 --- a/drivers/net/ethernet/freescale/fec_main.c
> > +++ b/drivers/net/ethernet/freescale/fec_main.c
> > @@ -1818,6 +1818,8 @@ static int fec_enet_mdio_write(struct mii_bus
> > *bus, int mii_id, int regnum, netdev_err(fep->netdev, "MDIO write
> > timeout\n"); ret  = -ETIMEDOUT;
> >         }
> > +       else
> > +               ret = 0;  
> 
> , so is this really needed? ret should be 0 at this point.



Cordialement,
Albert ARIBAUD
3ADEV

Re: [PATCH] net: trivial: fix fsl fec_main mdio write return value

[Sergei Shtylyov]

Hello.

On 09/26/2016 10:11 PM, Albert ARIBAUD (3ADEV) wrote:

> If the MDIO enf-of-transfer wait loop does not time out,
> 0 should be returned to the caller through ret, but ret
> was not reset to 0 after receiving the return value of
> pm_runtime_get_sync(). Reset ret when no time out occurred.
>
> Signed-off-by: Albert ARIBAUD (3ADEV) <albert.aribaud@3adev.fr>
> ---
>  drivers/net/ethernet/freescale/fec_main.c | 2 ++
>  1 file changed, 2 insertions(+)
>
> diff --git a/drivers/net/ethernet/freescale/fec_main.c b/drivers/net/ethernet/freescale/fec_main.c
> index 01f7e81..0413c05 100644
> --- a/drivers/net/ethernet/freescale/fec_main.c
> +++ b/drivers/net/ethernet/freescale/fec_main.c
> @@ -1818,6 +1818,8 @@ static int fec_enet_mdio_write(struct mii_bus *bus, int mii_id, int regnum,
>  		netdev_err(fep->netdev, "MDIO write timeout\n");
>  		ret  = -ETIMEDOUT;
>  	}
> +	else

	} else {

> +		ret = 0;

	}

[...]

MBR, Sergei