Message-ID: <93bc1f0a-d1ed-e0cf-30ec-a3b59a1ec678@datenfreihafen.org>
Date:   Thu, 16 Mar 2023 17:33:15 +0100
MIME-Version: 1.0
User-Agent: Mozilla/5.0 (X11; Linux x86_64; rv:102.0) Gecko/20100101
 Thunderbird/102.8.0
Subject: Re: [PATCH next] ca8210: Fix unsigned mac_len comparison with zero in
 ca8210_skb_tx()
Content-Language: en-US
To:     Simon Horman <simon.horman@corigine.com>,
        Harshit Mogalapalli <harshit.m.mogalapalli@oracle.com>
Cc:     Dan Carpenter <error27@gmail.com>,
        Alexander Aring <alex.aring@gmail.com>,
        Miquel Raynal <miquel.raynal@bootlin.com>,
        "David S. Miller" <davem@davemloft.net>,
        Eric Dumazet <edumazet@google.com>,
        Jakub Kicinski <kuba@kernel.org>,
        Paolo Abeni <pabeni@redhat.com>,
        Marcel Holtmann <marcel@holtmann.org>,
        Harry Morris <harrymorris12@gmail.com>,
        linux-wpan@vger.kernel.org, netdev@vger.kernel.org,
        linux-kernel@vger.kernel.org
References: <20230306191824.4115839-1-harshit.m.mogalapalli@oracle.com>
 <ZAb5BlS+OgFfJM6t@corigine.com>
From:   Stefan Schmidt <stefan@datenfreihafen.org>
In-Reply-To: <ZAb5BlS+OgFfJM6t@corigine.com>
Content-Type: text/plain; charset=UTF-8; format=flowed
Content-Transfer-Encoding: 7bit
Precedence: bulk

Hello Simon.

On 07.03.23 09:42, Simon Horman wrote:
> On Mon, Mar 06, 2023 at 11:18:24AM -0800, Harshit Mogalapalli wrote:
>> mac_len is of type unsigned, which can never be less than zero.
>>
>> 	mac_len = ieee802154_hdr_peek_addrs(skb, &header);
>> 	if (mac_len < 0)
>> 		return mac_len;
>>
>> Change this to type int as ieee802154_hdr_peek_addrs() can return negative
>> integers, this is found by static analysis with smatch.
>>
>> Fixes: ded845a781a5 ("ieee802154: Add CA8210 IEEE 802.15.4 device driver")
>> Signed-off-by: Harshit Mogalapalli <harshit.m.mogalapalli@oracle.com>
> 
> I discussed this briefly with Harshit offline.
> 
> The commit referenced above tag does add the call to
> ieee802154_hdr_peek_addrs(), an there is a sign miss match between
> the return value and the variable.
> 
> The code to check the mac_len was added more recently, by the following
> commit. However the fixes tag is probably fine as-is, because it's fixing
> error handling of a call made in that commit.
> 
> 6c993779ea1d ("ca8210: fix mac_len negative array access")
> 
> Reviewed-by: Simon Horman <simon.horman@corigine.com>

I agree that the commit above is the better Fixes tag as it makes clear 
it only comes after this change. I amended the commit message 
accordingly when applying this to wpan.

regards
Stefan Schmidt