[PATCH] drivers: acpi: fan.c move a dereference below the NULL test

[PATCH] drivers: acpi: fan.c move a dereference below the NULL test

[Darren Jenkins]

In acpi_fan_remove() device is being dereferenced before the NULL test.
This reorders the code to ensure it is checked for NULL first.

Coverity CID: 2758

Signed-off-by: Darren Jenkins <darrenrjenkins@gmail.com>
---
 drivers/acpi/fan.c |    9 +++++++--
 1 files changed, 7 insertions(+), 2 deletions(-)

diff --git a/drivers/acpi/fan.c b/drivers/acpi/fan.c
index acf2ab2..dc39640 100644
--- a/drivers/acpi/fan.c
+++ b/drivers/acpi/fan.c
@@ -298,9 +298,14 @@ static int acpi_fan_add(struct acpi_device *device)
 
 static int acpi_fan_remove(struct acpi_device *device, int type)
 {
-	struct thermal_cooling_device *cdev = acpi_driver_data(device);
+	struct thermal_cooling_device *cdev;
+
+	if (!device)
+		return -EINVAL;
+
+	cdev = acpi_driver_data(device);
 
-	if (!device || !cdev)
+	if (!cdev)
 		return -EINVAL;
 
 	acpi_fan_remove_fs(device);
-- 
1.6.3.3

[PATCH] drivers: acpi: fan.c move a dereference below the NULL test

[Darren Jenkins]

In acpi_fan_remove() device is being dereferenced before the NULL test.
This reorders the code to ensure it is checked for NULL first.

Coverity CID: 2758

Signed-off-by: Darren Jenkins <darrenrjenkins@gmail.com>
---
 drivers/acpi/fan.c |    9 +++++++--
 1 files changed, 7 insertions(+), 2 deletions(-)

diff --git a/drivers/acpi/fan.c b/drivers/acpi/fan.c
index acf2ab2..dc39640 100644
--- a/drivers/acpi/fan.c
+++ b/drivers/acpi/fan.c
@@ -298,9 +298,14 @@ static int acpi_fan_add(struct acpi_device *device)
 
 static int acpi_fan_remove(struct acpi_device *device, int type)
 {
-	struct thermal_cooling_device *cdev = acpi_driver_data(device);
+	struct thermal_cooling_device *cdev;
+
+	if (!device)
+		return -EINVAL;
+
+	cdev = acpi_driver_data(device);
 
-	if (!device || !cdev)
+	if (!cdev)
 		return -EINVAL;
 
 	acpi_fan_remove_fs(device);
-- 
1.6.3.3

Re: [PATCH] drivers: acpi: fan.c move a dereference below the NULL test

[Len Brown]

I think think this is a run-time check for a programming error,
and the current fashion is to delete the check and take a fault
if this happens so the caller can be fixed.

There are a couple of checks like this in fan.c --
perhaps Rui can clean them up when he comes back next week.

thanks,
Len Brown, Intel Open Source Technology Center

On Thu, 11 Feb 2010, Darren Jenkins wrote:

> In acpi_fan_remove() device is being dereferenced before the NULL test.
> This reorders the code to ensure it is checked for NULL first.
> 
> Coverity CID: 2758
> 
> Signed-off-by: Darren Jenkins <darrenrjenkins@gmail.com>
> ---
>  drivers/acpi/fan.c |    9 +++++++--
>  1 files changed, 7 insertions(+), 2 deletions(-)
> 
> diff --git a/drivers/acpi/fan.c b/drivers/acpi/fan.c
> index acf2ab2..dc39640 100644
> --- a/drivers/acpi/fan.c
> +++ b/drivers/acpi/fan.c
> @@ -298,9 +298,14 @@ static int acpi_fan_add(struct acpi_device *device)
>  
>  static int acpi_fan_remove(struct acpi_device *device, int type)
>  {
> -	struct thermal_cooling_device *cdev = acpi_driver_data(device);
> +	struct thermal_cooling_device *cdev;
> +
> +	if (!device)
> +		return -EINVAL;
> +
> +	cdev = acpi_driver_data(device);
>  
> -	if (!device || !cdev)
> +	if (!cdev)
>  		return -EINVAL;
>  
>  	acpi_fan_remove_fs(device);
> -- 
> 1.6.3.3
> 
> 
> 
> 
> --
> To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
> the body of a message to majordomo@vger.kernel.org
> More majordomo info at  http://vger.kernel.org/majordomo-info.html
> Please read the FAQ at  http://www.tux.org/lkml/
> 

Re: [PATCH] drivers: acpi: fan.c move a dereference below the NULL test

[Len Brown]

I think think this is a run-time check for a programming error,
and the current fashion is to delete the check and take a fault
if this happens so the caller can be fixed.

There are a couple of checks like this in fan.c --
perhaps Rui can clean them up when he comes back next week.

thanks,
Len Brown, Intel Open Source Technology Center

On Thu, 11 Feb 2010, Darren Jenkins wrote:

> In acpi_fan_remove() device is being dereferenced before the NULL test.
> This reorders the code to ensure it is checked for NULL first.
> 
> Coverity CID: 2758
> 
> Signed-off-by: Darren Jenkins <darrenrjenkins@gmail.com>
> ---
>  drivers/acpi/fan.c |    9 +++++++--
>  1 files changed, 7 insertions(+), 2 deletions(-)
> 
> diff --git a/drivers/acpi/fan.c b/drivers/acpi/fan.c
> index acf2ab2..dc39640 100644
> --- a/drivers/acpi/fan.c
> +++ b/drivers/acpi/fan.c
> @@ -298,9 +298,14 @@ static int acpi_fan_add(struct acpi_device *device)
>  
>  static int acpi_fan_remove(struct acpi_device *device, int type)
>  {
> -	struct thermal_cooling_device *cdev = acpi_driver_data(device);
> +	struct thermal_cooling_device *cdev;
> +
> +	if (!device)
> +		return -EINVAL;
> +
> +	cdev = acpi_driver_data(device);
>  
> -	if (!device || !cdev)
> +	if (!cdev)
>  		return -EINVAL;
>  
>  	acpi_fan_remove_fs(device);
> -- 
> 1.6.3.3
> 
> 
> 
> 
> --
> To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
> the body of a message to majordomo@vger.kernel.org
> More majordomo info at  http://vger.kernel.org/majordomo-info.html
> Please read the FAQ at  http://www.tux.org/lkml/
> 

Re: [PATCH] drivers: acpi: fan.c move a dereference below the NULL test

[Thomas Renninger]

On Friday 19 February 2010 06:27:46 Len Brown wrote:
> I think think this is a run-time check for a programming error,
> and the current fashion is to delete the check and take a fault
> if this happens so the caller can be fixed.
> 
> There are a couple of checks like this in fan.c --
> perhaps Rui can clean them up when he comes back next week.
Please let us clean this up first and remove:
static inline void *acpi_driver_data(struct acpi_device *d)
{
        return d->driver_data;
}
This has nothing to do with acpica code, it's in kernel code only.
Is there a reason for above abstraction?
If not, it's error prone and really should get eliminated.

    Thomas

> thanks,
> Len Brown, Intel Open Source Technology Center
> 
> On Thu, 11 Feb 2010, Darren Jenkins wrote:
> 
> > In acpi_fan_remove() device is being dereferenced before the NULL 
test.
> > This reorders the code to ensure it is checked for NULL first.
> > 
> > Coverity CID: 2758
> > 
> > Signed-off-by: Darren Jenkins <darrenrjenkins@gmail.com>
> > ---
> >  drivers/acpi/fan.c |    9 +++++++--
> >  1 files changed, 7 insertions(+), 2 deletions(-)
> > 
> > diff --git a/drivers/acpi/fan.c b/drivers/acpi/fan.c
> > index acf2ab2..dc39640 100644
> > --- a/drivers/acpi/fan.c
> > +++ b/drivers/acpi/fan.c
> > @@ -298,9 +298,14 @@ static int acpi_fan_add(struct acpi_device 
*device)
> >  
> >  static int acpi_fan_remove(struct acpi_device *device, int type)
> >  {
> > -	struct thermal_cooling_device *cdev = acpi_driver_data(device);
> > +	struct thermal_cooling_device *cdev;
> > +
> > +	if (!device)
> > +		return -EINVAL;
> > +
> > +	cdev = acpi_driver_data(device);
> >  
> > -	if (!device || !cdev)
> > +	if (!cdev)
> >  		return -EINVAL;
> >  
> >  	acpi_fan_remove_fs(device);
> > -- 
> > 1.6.3.3
> > 
> > 
> > 
> > 
> > --
> > To unsubscribe from this list: send the line "unsubscribe linux-
kernel" in
> > the body of a message to majordomo@vger.kernel.org
> > More majordomo info at  http://vger.kernel.org/majordomo-info.html
> > Please read the FAQ at  http://www.tux.org/lkml/
> > 
> 

Re: [PATCH] drivers: acpi: fan.c move a dereference below the NULL test

[Thomas Renninger]

On Friday 19 February 2010 06:27:46 Len Brown wrote:
> I think think this is a run-time check for a programming error,
> and the current fashion is to delete the check and take a fault
> if this happens so the caller can be fixed.
> 
> There are a couple of checks like this in fan.c --
> perhaps Rui can clean them up when he comes back next week.
Please let us clean this up first and remove:
static inline void *acpi_driver_data(struct acpi_device *d)
{
        return d->driver_data;
}
This has nothing to do with acpica code, it's in kernel code only.
Is there a reason for above abstraction?
If not, it's error prone and really should get eliminated.

    Thomas

> thanks,
> Len Brown, Intel Open Source Technology Center
> 
> On Thu, 11 Feb 2010, Darren Jenkins wrote:
> 
> > In acpi_fan_remove() device is being dereferenced before the NULL 
test.
> > This reorders the code to ensure it is checked for NULL first.
> > 
> > Coverity CID: 2758
> > 
> > Signed-off-by: Darren Jenkins <darrenrjenkins@gmail.com>
> > ---
> >  drivers/acpi/fan.c |    9 +++++++--
> >  1 files changed, 7 insertions(+), 2 deletions(-)
> > 
> > diff --git a/drivers/acpi/fan.c b/drivers/acpi/fan.c
> > index acf2ab2..dc39640 100644
> > --- a/drivers/acpi/fan.c
> > +++ b/drivers/acpi/fan.c
> > @@ -298,9 +298,14 @@ static int acpi_fan_add(struct acpi_device 
*device)
> >  
> >  static int acpi_fan_remove(struct acpi_device *device, int type)
> >  {
> > -	struct thermal_cooling_device *cdev = acpi_driver_data(device);
> > +	struct thermal_cooling_device *cdev;
> > +
> > +	if (!device)
> > +		return -EINVAL;
> > +
> > +	cdev = acpi_driver_data(device);
> >  
> > -	if (!device || !cdev)
> > +	if (!cdev)
> >  		return -EINVAL;
> >  
> >  	acpi_fan_remove_fs(device);
> > -- 
> > 1.6.3.3
> > 
> > 
> > 
> > 
> > --
> > To unsubscribe from this list: send the line "unsubscribe linux-
kernel" in
> > the body of a message to majordomo@vger.kernel.org
> > More majordomo info at  http://vger.kernel.org/majordomo-info.html
> > Please read the FAQ at  http://www.tux.org/lkml/
> > 
> 

Re: [PATCH] drivers: acpi: fan.c move a dereference below the NULL

[Zhang Rui]

On Fri, 2010-02-19 at 13:27 +0800, Len Brown wrote:
> I think think this is a run-time check for a programming error,
> and the current fashion is to delete the check and take a fault
> if this happens so the caller can be fixed.
> 
I agree.

> There are a couple of checks like this in fan.c --
> perhaps Rui can clean them up when he comes back next week.
> 
I'll send the cleanup patch later.

thanks,
rui

> thanks,
> Len Brown, Intel Open Source Technology Center
> 
> On Thu, 11 Feb 2010, Darren Jenkins wrote:
> 
> > In acpi_fan_remove() device is being dereferenced before the NULL test.
> > This reorders the code to ensure it is checked for NULL first.
> > 
> > Coverity CID: 2758
> > 
> > Signed-off-by: Darren Jenkins <darrenrjenkins@gmail.com>
> > ---
> >  drivers/acpi/fan.c |    9 +++++++--
> >  1 files changed, 7 insertions(+), 2 deletions(-)
> > 
> > diff --git a/drivers/acpi/fan.c b/drivers/acpi/fan.c
> > index acf2ab2..dc39640 100644
> > --- a/drivers/acpi/fan.c
> > +++ b/drivers/acpi/fan.c
> > @@ -298,9 +298,14 @@ static int acpi_fan_add(struct acpi_device *device)
> >  
> >  static int acpi_fan_remove(struct acpi_device *device, int type)
> >  {
> > -	struct thermal_cooling_device *cdev = acpi_driver_data(device);
> > +	struct thermal_cooling_device *cdev;
> > +
> > +	if (!device)
> > +		return -EINVAL;
> > +
> > +	cdev = acpi_driver_data(device);
> >  
> > -	if (!device || !cdev)
> > +	if (!cdev)
> >  		return -EINVAL;
> >  
> >  	acpi_fan_remove_fs(device);
> > -- 
> > 1.6.3.3
> > 
> > 
> > 
> > 
> > --
> > To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
> > the body of a message to majordomo@vger.kernel.org
> > More majordomo info at  http://vger.kernel.org/majordomo-info.html
> > Please read the FAQ at  http://www.tux.org/lkml/
> > 

Re: [PATCH] drivers: acpi: fan.c move a dereference below the NULL test

[Zhang Rui]

On Fri, 2010-02-19 at 13:27 +0800, Len Brown wrote:
> I think think this is a run-time check for a programming error,
> and the current fashion is to delete the check and take a fault
> if this happens so the caller can be fixed.
> 
I agree.

> There are a couple of checks like this in fan.c --
> perhaps Rui can clean them up when he comes back next week.
> 
I'll send the cleanup patch later.

thanks,
rui

> thanks,
> Len Brown, Intel Open Source Technology Center
> 
> On Thu, 11 Feb 2010, Darren Jenkins wrote:
> 
> > In acpi_fan_remove() device is being dereferenced before the NULL test.
> > This reorders the code to ensure it is checked for NULL first.
> > 
> > Coverity CID: 2758
> > 
> > Signed-off-by: Darren Jenkins <darrenrjenkins@gmail.com>
> > ---
> >  drivers/acpi/fan.c |    9 +++++++--
> >  1 files changed, 7 insertions(+), 2 deletions(-)
> > 
> > diff --git a/drivers/acpi/fan.c b/drivers/acpi/fan.c
> > index acf2ab2..dc39640 100644
> > --- a/drivers/acpi/fan.c
> > +++ b/drivers/acpi/fan.c
> > @@ -298,9 +298,14 @@ static int acpi_fan_add(struct acpi_device *device)
> >  
> >  static int acpi_fan_remove(struct acpi_device *device, int type)
> >  {
> > -	struct thermal_cooling_device *cdev = acpi_driver_data(device);
> > +	struct thermal_cooling_device *cdev;
> > +
> > +	if (!device)
> > +		return -EINVAL;
> > +
> > +	cdev = acpi_driver_data(device);
> >  
> > -	if (!device || !cdev)
> > +	if (!cdev)
> >  		return -EINVAL;
> >  
> >  	acpi_fan_remove_fs(device);
> > -- 
> > 1.6.3.3
> > 
> > 
> > 
> > 
> > --
> > To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
> > the body of a message to majordomo@vger.kernel.org
> > More majordomo info at  http://vger.kernel.org/majordomo-info.html
> > Please read the FAQ at  http://www.tux.org/lkml/
> >