Return-Path: <cyrus@holtmann.org>
Message-ID: <5434F191.3080403@xsilon.com>
Date: Wed, 08 Oct 2014 09:10:57 +0100
From: Martin Townsend <martin.townsend@xsilon.com>
MIME-Version: 1.0
To: Alexander Aring <alex.aring@gmail.com>,
 Martin Townsend <mtownsend1973@gmail.com>
CC: linux-bluetooth@vger.kernel.org, linux-wpan@vger.kernel.org,
 marcel@holtmann.org, jukka.rissanen@linux.intel.com
Subject: Re: [PATCH v3 bluetooth-next] 6lowpan: Use pskb_expand_head in IPHC
 decompression.
References: <1412753745-6469-1-git-send-email-martin.townsend@xsilon.com> <1412753745-6469-2-git-send-email-martin.townsend@xsilon.com> <20141008075458.GA32554@omega> <20141008080618.GB32554@omega>
In-Reply-To: <20141008080618.GB32554@omega>
Content-Type: text/plain; charset=utf-8
List-ID: <linux-bluetooth.vger.kernel.org>

Hi Alex,

On 08/10/14 09:06, Alexander Aring wrote:
> Hi Martin,
>
> On Wed, Oct 08, 2014 at 09:54:58AM +0200, Alexander Aring wrote:
> ...
>> here we lost the skb reference from parameter if failed and I don't see
>> that skb_unshare free it on failing (maybe I am wrong here).
>>
> I will do this step by step.
>
> 1. call skb_unshare
>
> 2. check if cloned
>
> 3. call skb_copy
>
> 4. call __alloc_skb
>
> 5. __alloc_skb return NULL
>
> 6. assign to temp nskb;
>
> 7. parameter skb will be freed. kfree_skb(skb).
>
> 8. skb = nskb;
>
> 9. return skb;
>
> So I see that kfree_skb(skb) is always called also on failure.
Yes this caught me out too :)
> So your code seems to be correct. This is a lack of documentation.
>
> - Alex
So is v3 patch good to go?

- Martin.