netdev - Re: [PATCH net-next 1/2] mlxsw: core: Add ethtool support for QSFP-DD transceivers

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: <CAL_jBfTKW_T-Pf2_shLm7N-ve_eg3G=nTD+6Fc3ZN4aHncm9YQ@mail.gmail.com>
Date:   Sat, 27 Jun 2020 21:42:10 +0100
From:   Adrian Pop <popadrian1996@...il.com>
To:     Ido Schimmel <idosch@...sch.org>
Cc:     Andrew Lunn <andrew@...n.ch>, netdev@...r.kernel.org,
        davem@...emloft.net, kuba@...nel.org, jiri@...lanox.com,
        vadimp@...lanox.com, mlxsw@...lanox.com,
        Ido Schimmel <idosch@...lanox.com>
Subject: Re: [PATCH net-next 1/2] mlxsw: core: Add ethtool support for QSFP-DD transceivers

>
> Hi Adrian, Andrew,
>
> Not sure I understand... You want the kernel to always pass page 03h to
> user space (potentially zeroed)? Page 03h is not mandatory according to
> the standard and page 01h contains information if page 03h is present or

Hi Ido!

Andrew was thinking of having 03h after 02h (potentially zeroed) just
for the purpose of having a similar layout for QSFP-DD the same way we
do for QSFP. But as you said, it is not mandatory according to the
standard and I also don't know the entire codebase for ethtool and
where it might be actually needed. I think Andrew can argue for its
presence better than me.

> not. So user space has the information it needs to determine if after
> page 02h we have page 03h or page 10h. Why always pass page 03h then?
>

If we decide to add 03h but only sometimes, I think we will add an
extra layer of complexity. Sometimes after 02h we would have 03h and
sometimes 10h. In qsfp-dd.h (following the convention from qsfp.h) in
my patch there are a lot of different constants defined with respect
to the offset of the parent page in the memory layout and "dynamic
offsets" don't sound very good, at least for me. So even if there's a
way of checking in the user space which page is after 02h, a more
stable memory layout works better on the long run.

Adrian