[PATCH] hid: hidraw_connect memleak fix

[PATCH] hid: hidraw_connect memleak fix

[Mariusz Kozlowski]

It looks like hidraw_connect() is leaking memory in case of failure.
Also it should return -ENOMEM when kzalloc fails.

Signed-off-by: Mariusz Kozlowski <m.kozlowski@tuxland.pl>

 drivers/hid/hidraw.c |   14 ++++++++++----
 1 file changed, 10 insertions(+), 4 deletions(-)

--- linux-2.6.23-rc8-mm1-a/drivers/hid/hidraw.c	2007-09-26 12:07:37.000000000 +0200
+++ linux-2.6.23-rc8-mm1-b/drivers/hid/hidraw.c	2007-09-26 14:36:08.000000000 +0200
@@ -282,7 +282,7 @@ EXPORT_SYMBOL_GPL(hidraw_report_event);

 int hidraw_connect(struct hid_device *hid)
 {
-	int minor, result = -EINVAL;
+	int minor, result;
 	struct hidraw *dev;

 	/* TODO currently we accept any HID device. This should later
@@ -290,8 +290,11 @@ int hidraw_connect(struct hid_device *hi
 	 * non-input applications
 	 */

-	if (!(dev = kzalloc(sizeof(struct hidraw), GFP_KERNEL)))
-		return -1;
+	dev = kzalloc(sizeof(struct hidraw), GFP_KERNEL);
+	if (!dev)
+		return -ENOMEM;
+
+	result = -EINVAL;

 	spin_lock(&minors_lock);

@@ -305,8 +308,10 @@ int hidraw_connect(struct hid_device *hi

 	spin_unlock(&minors_lock);

-	if (result)
+	if (result) {
+		kfree(dev);
 		goto out;
+	}

 	dev->dev = device_create(hidraw_class, NULL, MKDEV(hidraw_major, minor),
 				"%s%d", "hidraw", minor);
@@ -316,6 +321,7 @@ int hidraw_connect(struct hid_device *hi
 		hidraw_table[minor] = NULL;
 		spin_unlock(&minors_lock);
 		result = PTR_ERR(dev->dev);
+		kfree(dev);
 		goto out;
 	}

Re: [PATCH] hid: hidraw_connect memleak fix

[Jiri Kosina]

On Thu, 27 Sep 2007, Mariusz Kozlowski wrote:

> It looks like hidraw_connect() is leaking memory in case of failure. 
> Also it should return -ENOMEM when kzalloc fails.

Mariusz,

good catch, will apply to me tree.

Thanks,

-- 
Jiri Kosina
SUSE Labs