netdev - Re: [PATCH] net: dev: Add API to check net

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: <CA+5LGR1KwePssqhCkZ6qT_W87fO2o1XPze53mJwjkTWtphiWrA@mail.gmail.com>
Date:   Tue, 4 Aug 2020 20:47:18 +0300
From:   Ilia Lin <ilia.lin@...il.com>
To:     David Miller <davem@...emloft.net>
Cc:     andrew@...n.ch, ilial@...eaurora.org, kuba@...nel.org,
        jiri@...lanox.com, edumazet@...gle.com, ap420073@...il.com,
        xiyou.wangcong@...il.com, maximmi@...lanox.com,
        Ilia Lin <ilia.lin@...nel.org>, netdev@...r.kernel.org,
        open list <linux-kernel@...r.kernel.org>
Subject: Re: [PATCH] net: dev: Add API to check net_dev readiness

Hi Andrew and David,

Thank you for your comments!

The client driver is still work in progress, but it can be seen here:
https://source.codeaurora.org/quic/la/kernel/msm-4.19/tree/drivers/platform/msm/ipa/ipa_api.c#n3842

For HW performance reasons, it has to be in subsys_initcall.

Here is the register_netdev call:
https://source.codeaurora.org/quic/la/kernel/msm-4.19/tree/drivers/platform/msm/ipa/ipa_v3/rmnet_ipa.c#n2497

And it is going to be in the subsys_initcall as well.

Thanks,
Ilia



On Mon, Jul 27, 2020 at 8:32 PM David Miller <davem@...emloft.net> wrote:
>
> From: Andrew Lunn <andrew@...n.ch>
> Date: Sun, 26 Jul 2020 21:45:28 +0200
>
> > I also have to wonder why a network device driver is being probed the
> > subsys_initcall.
>
> This makes me wonder how this interface could even be useful.  The
> only way to fix the problem is to change when the device is probed,
> which would mean changing which initcall it uses.  So at run time,
> this information can't do much.