[PATCH] [staging] iio, adc: Do not leak memory in ad7280_event_handler()

[PATCH] [staging] iio, adc: Do not leak memory in ad7280_event_handler()

[Jesper Juhl]

If ad7280_read_all_channels() returns <0 then we'll leak the memory
allocated to 'channels' when we return and that variable goes out of
scope.
This patch fixes the leak.

Signed-off-by: Jesper Juhl <jj@chaosbits.net>
---
 drivers/staging/iio/adc/ad7280a.c |    4 +++-
 1 files changed, 3 insertions(+), 1 deletions(-)

 note: I don't have the hardware, so patch is compile tested only.

diff --git a/drivers/staging/iio/adc/ad7280a.c b/drivers/staging/iio/adc/ad7280a.c
index 372d059..dddc03c 100644
--- a/drivers/staging/iio/adc/ad7280a.c
+++ b/drivers/staging/iio/adc/ad7280a.c
@@ -687,8 +687,10 @@ static irqreturn_t ad7280_event_handler(int irq, void *private)
 		return IRQ_HANDLED;
 
 	ret = ad7280_read_all_channels(st, st->scan_cnt, channels);
-	if (ret < 0)
+	if (ret < 0) {
+		kfree(channels);
 		return IRQ_HANDLED;
+	}
 
 	for (i = 0; i < st->scan_cnt; i++) {
 		if (((channels[i] >> 23) & 0xF) <= AD7280A_CELL_VOLTAGE_6) {
-- 
1.7.7.2


-- 
Jesper Juhl <jj@chaosbits.net>       http://www.chaosbits.net/
Don't top-post http://www.catb.org/jargon/html/T/top-post.html
Plain text mails only, please.

Re: [PATCH] [staging] iio, adc: Do not leak memory in ad7280_event_handler()

[Jonathan Cameron]

On 11/06/2011 09:49 PM, Jesper Juhl wrote:
> If ad7280_read_all_channels() returns <0 then we'll leak the memory
> allocated to 'channels' when we return and that variable goes out of
> scope.
> This patch fixes the leak.
> 
Looks right to me - good spot. Only choice is whether a single exit
point makes sense rather than undwinding it here? Michael?
I'm happy with either solution, hence the ack.

> Signed-off-by: Jesper Juhl <jj@chaosbits.net>
Acked-by: Jonathan Cameron <jic23@cam.ac.uk>
> ---
>  drivers/staging/iio/adc/ad7280a.c |    4 +++-
>  1 files changed, 3 insertions(+), 1 deletions(-)
> 
>  note: I don't have the hardware, so patch is compile tested only.
> 
> diff --git a/drivers/staging/iio/adc/ad7280a.c b/drivers/staging/iio/adc/ad7280a.c
> index 372d059..dddc03c 100644
> --- a/drivers/staging/iio/adc/ad7280a.c
> +++ b/drivers/staging/iio/adc/ad7280a.c
> @@ -687,8 +687,10 @@ static irqreturn_t ad7280_event_handler(int irq, void *private)
>  		return IRQ_HANDLED;
>  
>  	ret = ad7280_read_all_channels(st, st->scan_cnt, channels);
> -	if (ret < 0)
> +	if (ret < 0) {
> +		kfree(channels);
>  		return IRQ_HANDLED;
> +	}
>  
>  	for (i = 0; i < st->scan_cnt; i++) {
>  		if (((channels[i] >> 23) & 0xF) <= AD7280A_CELL_VOLTAGE_6) {

Re: [PATCH] [staging] iio, adc: Do not leak memory in ad7280_event_handler()

[Lars-Peter Clausen]

On 11/06/2011 10:49 PM, Jonathan Cameron wrote:
> On 11/06/2011 09:49 PM, Jesper Juhl wrote:
>> If ad7280_read_all_channels() returns <0 then we'll leak the memory
>> allocated to 'channels' when we return and that variable goes out of
>> scope.
>> This patch fixes the leak.
>>
> Looks right to me - good spot. Only choice is whether a single exit
> point makes sense rather than undwinding it here? Michael?
> I'm happy with either solution, hence the ack.

A similar patch has already been sent last week and should be in Gregs queue.
See: http://comments.gmane.org/gmane.linux.kernel.iio/2433

- Lars

Re: [PATCH] [staging] iio, adc: Do not leak memory in ad7280_event_handler()

[Greg KH]

On Sun, Nov 06, 2011 at 10:49:22PM +0100, Jesper Juhl wrote:
> If ad7280_read_all_channels() returns <0 then we'll leak the memory
> allocated to 'channels' when we return and that variable goes out of
> scope.
> This patch fixes the leak.

This patch no longer applies to my tree, can you redo it against the
next linux-next release and resend it?

thanks,

greg k-h

Re: [PATCH] [staging] iio, adc: Do not leak memory in ad7280_event_handler()

[Jesper Juhl]

On Sat, 26 Nov 2011, Greg KH wrote:

> On Sun, Nov 06, 2011 at 10:49:22PM +0100, Jesper Juhl wrote:
> > If ad7280_read_all_channels() returns <0 then we'll leak the memory
> > allocated to 'channels' when we return and that variable goes out of
> > scope.
> > This patch fixes the leak.
> 
> This patch no longer applies to my tree, can you redo it against the
> next linux-next release and resend it?
> 
> thanks,
> 
> greg k-h

Sure. I'll take a look at it this weekend.


-- 
Jesper Juhl <jj@chaosbits.net>       http://www.chaosbits.net/
Don't top-post http://www.catb.org/jargon/html/T/top-post.html
Plain text mails only, please.