[PATCH] mxc-nand: no need to check for validity of platform drive data

[PATCH] mxc-nand: no need to check for validity of platform drive data

[Uwe Kleine-König]

The probe function calls platform_set_drvdata with a valid pointer when
the probe is successful.  As mxcnd_suspend and mxcnd_resume are only
called on bound devices, platform_get_drvdata always returns non-NULL.

This fix isn't critical as the pointer is always valid so it doesn't
matter if the compiler generated code for it or not.

Signed-off-by: Uwe Kleine-König <u.kleine-koenig@pengutronix.de>
Reported-by: David Binderman <dcb314@hotmail.com>
Cc: David Woodhouse <dwmw2@infradead.org>
Cc: Sascha Hauer <s.hauer@pengutronix.de>
Cc: Artem Bityutskiy <Artem.Bityutskiy@nokia.com>
Cc: Vladimir Barinov <vova.barinov@gmail.com>
---
 drivers/mtd/nand/mxc_nand.c |   17 +++++++----------
 1 files changed, 7 insertions(+), 10 deletions(-)

diff --git a/drivers/mtd/nand/mxc_nand.c b/drivers/mtd/nand/mxc_nand.c
index 45dec57..84f3635 100644
--- a/drivers/mtd/nand/mxc_nand.c
+++ b/drivers/mtd/nand/mxc_nand.c
@@ -886,11 +886,10 @@ static int mxcnd_suspend(struct platform_device *pdev, pm_message_t state)
 	int ret = 0;
 
 	DEBUG(MTD_DEBUG_LEVEL0, "MXC_ND : NAND suspend\n");
-	if (mtd) {
-		ret = mtd->suspend(mtd);
-		/* Disable the NFC clock */
-		clk_disable(host->clk);
-	}
+
+	ret = mtd->suspend(mtd);
+	/* Disable the NFC clock */
+	clk_disable(host->clk);
 
 	return ret;
 }
@@ -904,11 +903,9 @@ static int mxcnd_resume(struct platform_device *pdev)
 
 	DEBUG(MTD_DEBUG_LEVEL0, "MXC_ND : NAND resume\n");
 
-	if (mtd) {
-		/* Enable the NFC clock */
-		clk_enable(host->clk);
-		mtd->resume(mtd);
-	}
+	/* Enable the NFC clock */
+	clk_enable(host->clk);
+	mtd->resume(mtd);
 
 	return ret;
 }
-- 
1.6.5.2

[PATCH] mxc-nand: don't disable clock in mxcnd-suspend

[Uwe Kleine-König]

The clock must already be off after mtd->suspend.  Disabling it again
results in an negative overflow of the clock usage count.  This didn't
hurt as mxcnd_resume undid it after wake up.

Signed-off-by: Uwe Kleine-König <u.kleine-koenig@pengutronix.de>
Cc: David Woodhouse <dwmw2@infradead.org>
Cc: Sascha Hauer <s.hauer@pengutronix.de>
Cc: Artem Bityutskiy <Artem.Bityutskiy@nokia.com>
Cc: Vladimir Barinov <vova.barinov@gmail.com>
---
 drivers/mtd/nand/mxc_nand.c |   10 ++++++----
 1 files changed, 6 insertions(+), 4 deletions(-)

diff --git a/drivers/mtd/nand/mxc_nand.c b/drivers/mtd/nand/mxc_nand.c
index 84f3635..970ce6b 100644
--- a/drivers/mtd/nand/mxc_nand.c
+++ b/drivers/mtd/nand/mxc_nand.c
@@ -888,8 +888,12 @@ static int mxcnd_suspend(struct platform_device *pdev, pm_message_t state)
 	DEBUG(MTD_DEBUG_LEVEL0, "MXC_ND : NAND suspend\n");
 
 	ret = mtd->suspend(mtd);
-	/* Disable the NFC clock */
-	clk_disable(host->clk);
+
+	/*
+	 * nand_suspend locks the device for exclusive access, so
+	 * the clock must already be off.
+	 */
+	BUG_ON(!ret && host->clk_act);
 
 	return ret;
 }
@@ -903,8 +907,6 @@ static int mxcnd_resume(struct platform_device *pdev)
 
 	DEBUG(MTD_DEBUG_LEVEL0, "MXC_ND : NAND resume\n");
 
-	/* Enable the NFC clock */
-	clk_enable(host->clk);
 	mtd->resume(mtd);
 
 	return ret;
-- 
1.6.5.2

Re: [PATCH] mxc-nand: don't disable clock in mxcnd-suspend

[Sascha Hauer]

On Mon, Jan 11, 2010 at 05:53:16PM +0100, Uwe Kleine-König wrote:
> The clock must already be off after mtd->suspend.  Disabling it again
> results in an negative overflow of the clock usage count.  This didn't
> hurt as mxcnd_resume undid it after wake up.
> 
> Signed-off-by: Uwe Kleine-König <u.kleine-koenig@pengutronix.de>
> Cc: David Woodhouse <dwmw2@infradead.org>
> Cc: Sascha Hauer <s.hauer@pengutronix.de>
> Cc: Artem Bityutskiy <Artem.Bityutskiy@nokia.com>
> Cc: Vladimir Barinov <vova.barinov@gmail.com>

Acked-by: Sascha Hauer <s.hauer@pengutronix.de>

> ---
>  drivers/mtd/nand/mxc_nand.c |   10 ++++++----
>  1 files changed, 6 insertions(+), 4 deletions(-)
> 
> diff --git a/drivers/mtd/nand/mxc_nand.c b/drivers/mtd/nand/mxc_nand.c
> index 84f3635..970ce6b 100644
> --- a/drivers/mtd/nand/mxc_nand.c
> +++ b/drivers/mtd/nand/mxc_nand.c
> @@ -888,8 +888,12 @@ static int mxcnd_suspend(struct platform_device *pdev, pm_message_t state)
>  	DEBUG(MTD_DEBUG_LEVEL0, "MXC_ND : NAND suspend\n");
>  
>  	ret = mtd->suspend(mtd);
> -	/* Disable the NFC clock */
> -	clk_disable(host->clk);
> +
> +	/*
> +	 * nand_suspend locks the device for exclusive access, so
> +	 * the clock must already be off.
> +	 */
> +	BUG_ON(!ret && host->clk_act);
>  
>  	return ret;
>  }
> @@ -903,8 +907,6 @@ static int mxcnd_resume(struct platform_device *pdev)
>  
>  	DEBUG(MTD_DEBUG_LEVEL0, "MXC_ND : NAND resume\n");
>  
> -	/* Enable the NFC clock */
> -	clk_enable(host->clk);
>  	mtd->resume(mtd);
>  
>  	return ret;
> -- 
> 1.6.5.2
> 
> 

-- 
Pengutronix e.K.                           |                             |
Industrial Linux Solutions                 | http://www.pengutronix.de/  |
Peiner Str. 6-8, 31137 Hildesheim, Germany | Phone: +49-5121-206917-0    |
Amtsgericht Hildesheim, HRA 2686           | Fax:   +49-5121-206917-5555 |

Re: [PATCH] mxc-nand: no need to check for validity of platform drive data

[Sascha Hauer]

On Mon, Jan 11, 2010 at 03:05:35PM +0100, Uwe Kleine-König wrote:
> The probe function calls platform_set_drvdata with a valid pointer when
> the probe is successful.  As mxcnd_suspend and mxcnd_resume are only
> called on bound devices, platform_get_drvdata always returns non-NULL.
> 
> This fix isn't critical as the pointer is always valid so it doesn't
> matter if the compiler generated code for it or not.
> 
> Signed-off-by: Uwe Kleine-König <u.kleine-koenig@pengutronix.de>
> Reported-by: David Binderman <dcb314@hotmail.com>
> Cc: David Woodhouse <dwmw2@infradead.org>
> Cc: Sascha Hauer <s.hauer@pengutronix.de>
> Cc: Artem Bityutskiy <Artem.Bityutskiy@nokia.com>
> Cc: Vladimir Barinov <vova.barinov@gmail.com>

Acked-by: Sascha Hauer <s.hauer@pengutronix.de>

> ---
>  drivers/mtd/nand/mxc_nand.c |   17 +++++++----------
>  1 files changed, 7 insertions(+), 10 deletions(-)
> 
> diff --git a/drivers/mtd/nand/mxc_nand.c b/drivers/mtd/nand/mxc_nand.c
> index 45dec57..84f3635 100644
> --- a/drivers/mtd/nand/mxc_nand.c
> +++ b/drivers/mtd/nand/mxc_nand.c
> @@ -886,11 +886,10 @@ static int mxcnd_suspend(struct platform_device *pdev, pm_message_t state)
>  	int ret = 0;
>  
>  	DEBUG(MTD_DEBUG_LEVEL0, "MXC_ND : NAND suspend\n");
> -	if (mtd) {
> -		ret = mtd->suspend(mtd);
> -		/* Disable the NFC clock */
> -		clk_disable(host->clk);
> -	}
> +
> +	ret = mtd->suspend(mtd);
> +	/* Disable the NFC clock */
> +	clk_disable(host->clk);
>  
>  	return ret;
>  }
> @@ -904,11 +903,9 @@ static int mxcnd_resume(struct platform_device *pdev)
>  
>  	DEBUG(MTD_DEBUG_LEVEL0, "MXC_ND : NAND resume\n");
>  
> -	if (mtd) {
> -		/* Enable the NFC clock */
> -		clk_enable(host->clk);
> -		mtd->resume(mtd);
> -	}
> +	/* Enable the NFC clock */
> +	clk_enable(host->clk);
> +	mtd->resume(mtd);
>  
>  	return ret;
>  }
> -- 
> 1.6.5.2
> 
> 

-- 
Pengutronix e.K.                           |                             |
Industrial Linux Solutions                 | http://www.pengutronix.de/  |
Peiner Str. 6-8, 31137 Hildesheim, Germany | Phone: +49-5121-206917-0    |
Amtsgericht Hildesheim, HRA 2686           | Fax:   +49-5121-206917-5555 |

Re: [PATCH] mxc-nand: no need to check for validity of platform drive data

[Artem Bityutskiy]

On Mon, 2010-01-11 at 15:05 +0100, Uwe Kleine-König wrote:
> The probe function calls platform_set_drvdata with a valid pointer when
> the probe is successful.  As mxcnd_suspend and mxcnd_resume are only
> called on bound devices, platform_get_drvdata always returns non-NULL.
> 
> This fix isn't critical as the pointer is always valid so it doesn't
> matter if the compiler generated code for it or not.

I've put these patches to my l2-mtd-2.6.git/master

-- 
Best Regards,
Artem Bityutskiy (Артём Битюцкий)

Re: [PATCH] mxc-nand: no need to check for validity of platform drive data

[Uwe Kleine-König]

Hello Artem,

On Sun, Jan 17, 2010 at 10:25:54AM +0200, Artem Bityutskiy wrote:
> On Mon, 2010-01-11 at 15:05 +0100, Uwe Kleine-König wrote:
> > The probe function calls platform_set_drvdata with a valid pointer when
> > the probe is successful.  As mxcnd_suspend and mxcnd_resume are only
> > called on bound devices, platform_get_drvdata always returns non-NULL.
> > 
> > This fix isn't critical as the pointer is always valid so it doesn't
> > matter if the compiler generated code for it or not.
> 
> I've put these patches to my l2-mtd-2.6.git/master
Up to now this patch didn't hit next.  Maybe you want to ask Steven
Rothwell to merge your tree, too?

Best regards
Uwe

-- 
Pengutronix e.K.                              | Uwe Kleine-König            |
Industrial Linux Solutions                    | http://www.pengutronix.de/  |

Re: [PATCH] mxc-nand: no need to check for validity of platform drive data

[Artem Bityutskiy]

Hi,

On Fri, 2010-02-05 at 14:47 +0100, ext Uwe Kleine-König wrote:
> Hello Artem,
> 
> On Sun, Jan 17, 2010 at 10:25:54AM +0200, Artem Bityutskiy wrote:
> > On Mon, 2010-01-11 at 15:05 +0100, Uwe Kleine-König wrote:
> > > The probe function calls platform_set_drvdata with a valid pointer when
> > > the probe is successful.  As mxcnd_suspend and mxcnd_resume are only
> > > called on bound devices, platform_get_drvdata always returns non-NULL.
> > > 
> > > This fix isn't critical as the pointer is always valid so it doesn't
> > > matter if the compiler generated code for it or not.
> > 
> > I've put these patches to my l2-mtd-2.6.git/master
> Up to now this patch didn't hit next.  Maybe you want to ask Steven
> Rothwell to merge your tree, too?

This is up to dwmw2, who is the MTD maintainer. All I do is I am helping
dwmw2 by providing him a git tree with all MTD patches from the mailing
list. So he does not have to go through whole mailing list. I put stuff
I reviewed to "master", and stuff I did/could not review to "dunno"

Hence, my tree is "level 2".

To get into the "next" tree, you should contact dwmw2.

-- 
Best Regards,
Artem Bityutskiy (Артём Битюцкий)