Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 2620:137:e000::1:20 as permitted sender) client-ip=2620:137:e000::1:20;
Subject: Re: [PATCH] net: sched: fix a error path in fw_change()
To:     Cong Wang <xiyou.wangcong@gmail.com>
Cc:     Jamal Hadi Salim <jhs@mojatatu.com>, Jiri Pirko <jiri@resnulli.us>,
        "David S . Miller" <davem@davemloft.net>,
        Eric Dumazet <edumazet@google.com>,
        Jakub Kicinski <kuba@kernel.org>,
        Paolo Abeni <pabeni@redhat.com>, netdev@vger.kernel.org,
        linux-kernel@vger.kernel.org, Yu Zhe <yuzhe@nfschina.com>
References: <20221201151532.25433-1-liqiong@nfschina.com>
 <Y4unik6y8a0MuoFt@pop-os.localdomain>
From:   liqiong <liqiong@nfschina.com>
Message-ID: <6a8a4c34-83bf-9418-987e-56ac4b44ccc6@nfschina.com>
Date:   Mon, 5 Dec 2022 10:36:00 +0800
User-Agent: Mozilla/5.0 (X11; Linux x86_64; rv:52.0) Gecko/20100101
 Thunderbird/52.2.1
MIME-Version: 1.0
In-Reply-To: <Y4unik6y8a0MuoFt@pop-os.localdomain>
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit
Content-Language: en-US
Precedence: bulk


在 2022年12月04日 03:46, Cong Wang 写道:
> On Thu, Dec 01, 2022 at 11:15:32PM +0800, Li Qiong wrote:
>> The 'pfp' pointer could be null if can't find the target filter.
>> Check 'pfp' pointer and fix this error path.
> Did you see any actual kernel crash? And do you have a reproducer too?
> Please include them if you do.
Found this by  'smatch' tool,  I check and find it may be a real problem at the risk
of NULL pointer. Like 'fw_delete()', It checks  'pfp' and 'pfp == f' too.
 

>
>> Signed-off-by: Li Qiong <liqiong@nfschina.com>
>> ---
>>  net/sched/cls_fw.c | 6 ++++++
>>  1 file changed, 6 insertions(+)
>>
>> diff --git a/net/sched/cls_fw.c b/net/sched/cls_fw.c
>> index a32351da968c..b898e4a81146 100644
>> --- a/net/sched/cls_fw.c
>> +++ b/net/sched/cls_fw.c
>> @@ -289,6 +289,12 @@ static int fw_change(struct net *net, struct sk_buff *in_skb,
>>  			if (pfp == f)
>>  				break;
>>  
>> +		if (!pfp) {
>> +			tcf_exts_destroy(&fnew->exts);
>> +			kfree(fnew);
>> +			return err;
>
> BTW, err is 0 here, you have to set some error here.
You are right,  It should return  '-EINVAL'  -- can't get the target filter.


>
> Thanks.