linux-kernel - Re: [PATCH] net: optimize cmpxchg in ip_idents

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Message-ID: <5377f988-9855-8322-a459-26376f50ee34@hisilicon.com>
Date:   Tue, 21 Jan 2020 10:40:11 +0800
From:   Shaokun Zhang <zhangshaokun@...ilicon.com>
To:     Eric Dumazet <edumazet@...gle.com>
CC:     Peter Zijlstra <peterz@...radead.org>,
        Eric Dumazet <eric.dumazet@...il.com>,
        David Miller <davem@...emloft.net>,
        netdev <netdev@...r.kernel.org>,
        LKML <linux-kernel@...r.kernel.org>, <jinyuqi@...wei.com>,
        Alexey Kuznetsov <kuznet@....inr.ac.ru>,
        "Hideaki YOSHIFUJI" <yoshfuji@...ux-ipv6.org>,
        <guoyang2@...wei.com>, Will Deacon <will@...nel.org>
Subject: Re: [PATCH] net: optimize cmpxchg in ip_idents_reserve

Hi Eric,

On 2020/1/19 12:12, Eric Dumazet wrote:
> On Sat, Jan 18, 2020 at 7:47 PM Shaokun Zhang
> <zhangshaokun@...ilicon.com> wrote:
>>
> 
>> We have used the atomic_add_return[1], but it makes the UBSAN unhappy followed
>> by the comment.
>> It seems that Eric also agreed to do it if some comments are added. I will do
>> it later.
>>
>> Thanks,
>> Shaokun
>>
>> [1] https://lkml.org/lkml/2019/7/26/217
>>
> 
> In case you have missed it, we needed a proper analysis.
> My feedback was quite simple :
> 
> <quote>
> Have you first checked that current UBSAN versions will not complain anymore ?
> </quote>
> 
> You never did this work, never replied to my question, and months

Yeah, I'm not sure how to deal with the UBSAN issue and you said that some of
you would work this.

> later you come back
> with a convoluted patch while we simply can proceed with a revert now

After several months, we thought that we can do it like refcount_add_not_zero,
so we submit this patch.

> we are sure that linux kernels are compiled with the proper option.
> 
> As mentioned yesterday, no need for a comment.
> Instead the changelog should be explaining why the revert is now safe.
> 

Ok, it is really needed to consider this.

Thanks,
Shaokun

> .
>