netdev - Re: qmi_wwan not using netif_carrier

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]

Date:   Wed, 17 Jun 2020 11:59:33 -0500
From:   Dan Williams <dcbw@...hat.com>
To:     Andrew Lunn <andrew@...n.ch>, Tanjeff-Nicolai Moos <tmoos@...ec.de>
Cc:     netdev@...r.kernel.org
Subject: Re: qmi_wwan not using netif_carrier_*()

On Wed, 2020-06-17 at 18:48 +0200, Andrew Lunn wrote:
> On Wed, Jun 17, 2020 at 03:21:53PM +0200, Tanjeff-Nicolai Moos wrote:
> > Hi netdevs,
> > 
> > Kernel version:
> > 
> >   I'm working with kernel 4.14.137 (OpenWRT project). But I looked
> > at
> >   the source of kernel 5.7 and found the same situation.
> > 
> > Problem:
> > 
> >   I'm using the qmi_wwan driver for a Sierra Wireless EM7455 LTE
> >   modem. This driver does not use
> >   netif_carrier_on()/netif_carrier_off() to update its link status.
> >   This confuses ledtrig_netdev which uses netif_carrier_ok() to
> > obtain
> >   the link status.
> > 
> > My solution:
> > 
> >   As a solution (or workaround?) I would try:
> > 
> >   1) In drivers/net/usb/qmi_wwan.c, lines 904/913: Add the flag
> >      FLAG_LINK_INTR.
> > 
> >   2) In drivers/net/usb/usbnet.c, functions usbnet_open() and
> >      usbnet_stop(): Add a call to netif_carrier_*(),
> >      but only if FLAG_LINK_INTR is set.
> > 
> > Question:
> > 
> >   Is this the intended way to use FLAG_LINK_INTR and
> > netif_carrier_*()?
> >   Or is there another recommended way to obtain the link status of
> >   network devices (I could change ledtrig_netdev)?
> 
> Hi Tanjeff
> 
> With Ethernet, having a carrier means there is a link partner, the
> layer 2 of the OSI 7 layer stack model is working. If the interface
> is
> not open()ed, it clearly should not have carrier. However, just
> because it is open, does not mean it has carrier. The cable could be
> unplugged, etc.
> 
> This is an LTE modem. What does carrier mean here? I don't know if it
> is well defined, but i would guess it is connected to a base station
> which is offering service. I'm assuming you are interested in data
> here, not wanting to make a 911/999/112/$EMERGENCY_SERVICE call which
> in theory all base stations should accept.
> 
> Is there a way to get this state information from the hardware? That
> would be the correct way to set the carrier.

There isn't. All the setup that would result in IFF_LOWER_UP (eg
ability to pass packets to the cellular network) happens over channels
*other* than the ethernet one. eg CDC-WDM, CDC-ACM, CDC-MBIM, AT
commands, QMI commands, MBIM commands, etc.

Something in userspace handles the actual IP-level connection setup and
once that's done, only then do you really have IFF_LOWER_UP. One way to
solve this could be to require userspace connection managers to manage
the carrier state of the device, which is possible for some drivers
already IIRC.

Dan