netdev - Re: [PATCH v3 net-next] rndis_host: Flag RNDIS modems as WWAN devices

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 for Android: free password hash cracker in your pocket

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

Message-ID: <20250412004203.099e482a@foxbook>
Date: Sat, 12 Apr 2025 00:42:03 +0200
From: Michał Pecio <michal.pecio@...il.com>
To: Lubomir Rintel <lkundrak@...sk>
Cc: linux-usb@...r.kernel.org, Paolo Abeni <pabeni@...hat.com>,
 linux-kernel@...r.kernel.org, netdev@...r.kernel.org, Jakub Kicinski
 <kuba@...nel.org>, Eric Dumazet <edumazet@...gle.com>, "David S. Miller"
 <davem@...emloft.net>, Andrew Lunn <andrew+netdev@...n.ch>
Subject: Re: [PATCH v3 net-next] rndis_host: Flag RNDIS modems as WWAN
 devices

On Tue, 25 Mar 2025 10:58:41 +0100, Lubomir Rintel wrote:
> Set FLAG_WWAN instead of FLAG_ETHERNET for RNDIS interfaces on Mobile
> Broadband Modems, as opposed to regular Ethernet adapters.
> 
> Otherwise NetworkManager gets confused, misjudges the device type,
> and wouldn't know it should connect a modem to get the device to work.
> What would be the result depends on ModemManager version -- older
> ModemManager would end up disconnecting a device after an unsuccessful
> probe attempt (if it connected without needing to unlock a SIM), while
> a newer one might spawn a separate PPP connection over a tty interface
> instead, resulting in a general confusion and no end of chaos.
> 
> The only way to get this work reliably is to fix the device type
> and have good enough version ModemManager (or equivalent).
> 
> Signed-off-by: Lubomir Rintel <lkundrak@...sk>
> Fixes: 63ba395cd7a5 ("rndis_host: support Novatel Verizon USB730L")

Hi,

This patch appears to have caused a regression for some users,
who opened a bug against the USB subsystem here:

https://bugzilla.kernel.org/show_bug.cgi?id=220002

Regards,
Michal