netdev - Re: [PATCH net-next v4 4/4] net: tun: track dropped skb via kfree_skb

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] [day] [month] [year] [list]

Date:   Wed, 2 Mar 2022 21:55:09 -0800
From:   Dongli Zhang <dongli.zhang@...cle.com>
To:     Jakub Kicinski <kuba@...nel.org>
Cc:     dsahern@...il.com, netdev@...r.kernel.org, bpf@...r.kernel.org,
        linux-kernel@...r.kernel.org, davem@...emloft.net,
        rostedt@...dmis.org, mingo@...hat.com, ast@...nel.org,
        daniel@...earbox.net, andrii@...nel.org, imagedong@...cent.com,
        joao.m.martins@...cle.com, joe.jin@...cle.com, edumazet@...gle.com
Subject: Re: [PATCH net-next v4 4/4] net: tun: track dropped skb via
 kfree_skb_reason()

Hi Jakub,

On 3/2/22 9:21 PM, Jakub Kicinski wrote:
> On Wed, 2 Mar 2022 14:21:31 -0800 Dongli Zhang wrote:
>>> because of OOM" is what should be reported. What we were trying to
>>> allocate is not very relevant (and can be gotten from the stack trace 
>>> if needed).  
>>
>> I think OOM is not enough. Although it may not be the case in this patchset,
>> sometimes the allocation is failed because we are allocating a large chunk of
>> physically continuous pages (kmalloc vs. vmalloc) while there is still plenty of
>> memory pages available.
>>
>> As a kernel developer, it is very significant for me to identify the specific
>> line/function and specific data structure that cause the error. E.g, the bug
>> filer may be chasing which line is making trouble.
>>
>> It is less likely to SKB_TRIM more than once in a driver function, compared to
>> ENOMEM.
> 
> Nack, trim is meaningless.
> 

I will use SKB_DROP_REASON_NOMEM.

Thank you very much!

Dongli Zhang