[PATCH] hwmon: Don't leak info in adm1275_probe

[PATCH] hwmon: Don't leak info in adm1275_probe

[Jesper Juhl]

In drivers/hwmon/adm1275.c:adm1275_probe() we'll leak the memory
allocated to 'info' if the call to i2c_smbus_read_byte_data() returns
less than zero.

This patch fixes the leak by simply calling kfree(info) when needed.
I also considered just moving the 'info' memory allocation below the
call to i2c_smbus_read_byte_data(), but I was not sure if it mattered
to get the -ENOMEM first in that case or if it would be bad to first
have a successful i2c_smbus_read_byte_data() call and then fail the
mem alloc, so I stuck with the safe option.

Signed-off-by: Jesper Juhl <jj@chaosbits.net>
---
 drivers/hwmon/adm1275.c |    4 +++-
 1 files changed, 3 insertions(+), 1 deletions(-)

  Compile tested only.

diff --git a/drivers/hwmon/adm1275.c b/drivers/hwmon/adm1275.c
index c2ee204..2ebf255 100644
--- a/drivers/hwmon/adm1275.c
+++ b/drivers/hwmon/adm1275.c
@@ -43,8 +43,10 @@ static int adm1275_probe(struct i2c_client *client,
 		return -ENOMEM;
 
 	config = i2c_smbus_read_byte_data(client, ADM1275_PMON_CONFIG);
-	if (config < 0)
+	if (config < 0) {
+		kfree(info);
 		return config;
+	}
 
 	info->pages = 1;
 	info->direct[PSC_VOLTAGE_IN] = true;
-- 
1.7.6


-- 
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] hwmon: Don't leak info in adm1275_probe

[Guenter Roeck]

On Sat, Jul 02, 2011 at 04:04:29PM -0400, Jesper Juhl wrote:
> In drivers/hwmon/adm1275.c:adm1275_probe() we'll leak the memory
> allocated to 'info' if the call to i2c_smbus_read_byte_data() returns
> less than zero.
> 
> This patch fixes the leak by simply calling kfree(info) when needed.
> I also considered just moving the 'info' memory allocation below the
> call to i2c_smbus_read_byte_data(), but I was not sure if it mattered
> to get the -ENOMEM first in that case or if it would be bad to first
> have a successful i2c_smbus_read_byte_data() call and then fail the
> mem alloc, so I stuck with the safe option.
> 
> Signed-off-by: Jesper Juhl <jj@chaosbits.net>

Hi Jesper,

the fix for this problem is already enqueued as 
"hwmon: (adm1275) Free allocated memory if probe function fails".

There is also a second error condition, which occurs if pmbus_do_probe() 
returns an error.

Thanks,
Guenter

Re: [PATCH] hwmon: Don't leak info in adm1275_probe

[Jesper Juhl]

On Sat, 2 Jul 2011, Guenter Roeck wrote:

> On Sat, Jul 02, 2011 at 04:04:29PM -0400, Jesper Juhl wrote:
> > In drivers/hwmon/adm1275.c:adm1275_probe() we'll leak the memory
> > allocated to 'info' if the call to i2c_smbus_read_byte_data() returns
> > less than zero.
> > 
> > This patch fixes the leak by simply calling kfree(info) when needed.
> > I also considered just moving the 'info' memory allocation below the
> > call to i2c_smbus_read_byte_data(), but I was not sure if it mattered
> > to get the -ENOMEM first in that case or if it would be bad to first
> > have a successful i2c_smbus_read_byte_data() call and then fail the
> > mem alloc, so I stuck with the safe option.
> > 
> > Signed-off-by: Jesper Juhl <jj@chaosbits.net>
> 
> Hi Jesper,
> 
> the fix for this problem is already enqueued as 
> "hwmon: (adm1275) Free allocated memory if probe function fails".
> 
> There is also a second error condition, which occurs if pmbus_do_probe() 
> returns an error.
> 

Ok, cool, then that's taken care of. I'll just drop the patch from my 
personal queue. Thanks.

-- 
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.