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;
Message-ID: <529eadf1-5c7a-2068-2932-8e75ae02e405@linux.alibaba.com>
Date:   Thu, 30 Jun 2022 22:15:14 +0800
MIME-Version: 1.0
User-Agent: Mozilla/5.0 (Macintosh; Intel Mac OS X 10.15; rv:91.0)
 Gecko/20100101 Thunderbird/91.9.1
Subject: Re: [PATCH] net: hinic: avoid kernel hung in hinic_get_stats64()
To:     Eric Dumazet <edumazet@google.com>
Cc:     David Miller <davem@davemloft.net>,
        Jakub Kicinski <kuba@kernel.org>,
        Paolo Abeni <pabeni@redhat.com>, gustavoars@kernel.org,
        cai.huoqing@linux.dev, Aviad Krawczyk <aviad.krawczyk@huawei.com>,
        zhaochen6@huawei.com, netdev <netdev@vger.kernel.org>,
        LKML <linux-kernel@vger.kernel.org>
References: <07736c2b7019b6883076a06129e06e8f7c5f7154.1656487154.git.mqaio@linux.alibaba.com>
 <CANn89iLpW4zFf2ABADbMNERPFr=OrAXEMm6ZgCxYA5VpcDpYTw@mail.gmail.com>
 <8b012bbd-a175-5699-1f26-108dd52fc5b7@linux.alibaba.com>
 <CANn89iJ-2BTR1SfFBbNG3jSgHK-TuRE_J-Khbbednu=pWnFtmw@mail.gmail.com>
From:   maqiao <mqaio@linux.alibaba.com>
In-Reply-To: <CANn89iJ-2BTR1SfFBbNG3jSgHK-TuRE_J-Khbbednu=pWnFtmw@mail.gmail.com>
Content-Type: text/plain; charset=UTF-8; format=flowed
Content-Transfer-Encoding: 8bit
Precedence: bulk


在 2022/6/30 下午9:59, Eric Dumazet 写道:
> On Thu, Jun 30, 2022 at 3:57 PM maqiao <mqaio@linux.alibaba.com> wrote:
>>
>>
>>
>> 在 2022/6/30 下午6:23, Eric Dumazet 写道:
> 
>>> Note: The following is racy, because multiple threads can call
>>> hinic_get_stats64() at the same time.
>>> It needs a loop, see include/linux/u64_stats_sync.h for detail.
>> Thanks for reminding, and I noticed that nic_tx_stats/nic_rx_stats has
>> been protected by u64_stats_sync in update_t/rx_stats(), it seems that
>> it's unnecessary to use spinlock in update_nic_stats().
> 
> It is necessary to use the spinlock to protect writers among themselves.
Ohhh, sorry, I was wrong.
I did not realize that seqlock cannot prevent mutil writers enter 
critical section...
> 
> 
>>
>> I will send v2 as soon as possible, thanks.