netdev - Re: introduce function wrapper for sk_data

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]

Message-ID: <20230307005638.76597-1-kuniyu@amazon.com>
Date:   Mon, 6 Mar 2023 16:56:38 -0800
From:   Kuniyuki Iwashima <kuniyu@...zon.com>
To:     <aahringo@...hat.com>
CC:     <cluster-devel@...hat.com>, <cong.wang@...edance.com>,
        <davem@...emloft.net>, <netdev@...r.kernel.org>,
        <peilin.ye@...edance.com>, <kuniyu@...zon.com>
Subject: Re: introduce function wrapper for sk_data_ready() call?

From:   Alexander Aring <aahringo@...hat.com>
Date:   Mon, 6 Mar 2023 07:47:02 -0500
> 
> Hi,
> 
> I saw that in 6.3-rc1 the following patch introduced something in dlm
> socket application handling 40e0b0908142 ("net/sock: Introduce
> trace_sk_data_ready()"). I am asking myself if we could instead
> introduce a wrapper in net/ protocol family implementations and they
> do such trace event calls there inside the socket implementation
> instead of letting the application layer do it. It looks pretty
> generic for me and it does not trace any application specific
> information.

I think you cannot apply the same logic to some functions which call
trace_sk_data_ready() twice, e.g. subflow_data_ready, tls_data_ready().

Then, only such functions need an additional trace_sk_data_ready(),
which is not clean, I think.

Thanks,
Kuniyuki


> 
> I did something similar for sk_error_report(), see e3ae2365efc1 ("net:
> sock: introduce sk_error_report").
> 
> Thanks.
> 
> - Alex