Received-SPF: pass (google.com: best guess record for domain of linux-kernel-owner@vger.kernel.org designates 209.132.180.67 as permitted sender) client-ip=209.132.180.67;
Subject: Re: [PATCH bpf v2] xdp: fix hang while unregistering device bound
 to xdp socket
From:   Ilya Maximets <i.maximets@samsung.com>
To:     Jakub Kicinski <jakub.kicinski@netronome.com>
Cc:     netdev@vger.kernel.org, linux-kernel@vger.kernel.org,
        bpf@vger.kernel.org, xdp-newbies@vger.kernel.org,
        "David S. Miller" <davem@davemloft.net>,
        =?UTF-8?B?QmrDtnJuIFTDtnBlbA==?= <bjorn.topel@intel.com>,
        Magnus Karlsson <magnus.karlsson@intel.com>,
        Jonathan Lemon <jonathan.lemon@gmail.com>
Message-ID: <06eee1e3-283a-d665-904b-f0bc89b73232@samsung.com>
Date:   Mon, 10 Jun 2019 19:17:37 +0300
User-Agent: Mozilla/5.0 (X11; Linux x86_64; rv:60.0) Gecko/20100101
        Thunderbird/60.7.0
MIME-Version: 1.0
In-Reply-To: <d8b2bc92-3b7e-dfe7-35ee-61a68d46ff02@samsung.com>
Content-Language: en-GB
Content-Transfer-Encoding: 7bit
Content-Type: text/plain; charset="utf-8"
CMS-TYPE: 201P
References: <CGME20190607173149eucas1p1d2ebedcab469ebd66acfe7c7dcd18d7e@eucas1p1.samsung.com>
        <20190607173143.4919-1-i.maximets@samsung.com>
        <20190607163156.12cd3418@cakuba.netronome.com>
        <d8b2bc92-3b7e-dfe7-35ee-61a68d46ff02@samsung.com>
Sender: linux-kernel-owner@vger.kernel.org
Precedence: bulk

On 10.06.2019 11:05, Ilya Maximets wrote:
> On 08.06.2019 2:31, Jakub Kicinski wrote:
>> On Fri,  7 Jun 2019 20:31:43 +0300, Ilya Maximets wrote:
>>> +static int xsk_notifier(struct notifier_block *this,
>>> +			unsigned long msg, void *ptr)
>>> +{
>>> +	struct sock *sk;
>>> +	struct net_device *dev = netdev_notifier_info_to_dev(ptr);
>>> +	struct net *net = dev_net(dev);
>>> +	int i, unregister_count = 0;
>>
>> Please order the var declaration lines longest to shortest.
>> (reverse christmas tree)
> 
> Hi.
> I'm not a fan of mixing 'struct's with bare types in the declarations.
> Moving the 'sk' to the third place will make a hole like this:
> 
> 	struct net_device *dev = netdev_notifier_info_to_dev(ptr);
> 	struct net *net = dev_net(dev);
> 	struct sock *sk;
> 	int i, unregister_count = 0;
> 
> Which is not looking good.
> Moving to the 4th place:
> 
> 	struct net_device *dev = netdev_notifier_info_to_dev(ptr);
> 	struct net *net = dev_net(dev);
> 	int i, unregister_count = 0;
> 	struct sock *sk;

I've sent v3 with this variant and with moved msg check to the top level.

> 
> This variant doesn't look good for me because of mixing 'struct's with
> bare integers.
> 
> Do you think I need to use one of above variants?
> 
>>
>>> +	mutex_lock(&net->xdp.lock);
>>> +	sk_for_each(sk, &net->xdp.list) {
>>> +		struct xdp_sock *xs = xdp_sk(sk);
>>> +
>>> +		mutex_lock(&xs->mutex);
>>> +		switch (msg) {
>>> +		case NETDEV_UNREGISTER:
>>
>> You should probably check the msg type earlier and not take all the
>> locks and iterate for other types..
> 
> Yeah. I thought about it too. Will fix in the next version.
> 
> Best regards, Ilya Maximets.
>