netdev - AX88179 interface is always NET_ADDR

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-next>] [day] [month] [year] [list]

Message-ID: <CAPY8ntBB+qDuw9M+2ZSFFuP78dFP9d5WPL93TdYAGxbdg=Msfw@mail.gmail.com>
Date: Fri, 22 Mar 2024 12:51:58 +0000
From: Dave Stevenson <dave.stevenson@...pberrypi.com>
To: jtornosm@...hat.com, weihao.bj@...system.com
Cc: netdev@...r.kernel.org, edumazet@...gle.com, pabeni@...hat.com, 
	kuba@...nel.org, davem@...emloft.net
Subject: AX88179 interface is always NET_ADDR_RANDOM after d2689b6a86b9

Hi

A few of our users have reported that in 6.6 and 6.8 an AX88179 is not
using predictable net names.

Digging into it, it's because addr_assign_type is NET_ADDR_RANDOM, and
that is caused by commit d2689b6a86b9 "net: usb: ax88179_178a: avoid
two consecutive device resets" (I haven't got that patch in my
mailbox, so I can't reply to it directly). That has also been
backported to 6.6.

Prior to that commit ax88179_reset was called as part of ax88179_bind,
and therefore set dev_addr before usbnet_probe looked at it.
That now doesn't happen, so usbnet_probe flags the interface as
NET_ADDR_RANDOM. The flag is not removed when ax88179_reset comes
along later and sets dev_addr.

I can fully understand the intention of the patch, but it seems that
dev_addr must be set as part of the device binding. I don't know this
driver well enough, so am making the report for those more familiar to
be able to evaluate it.

Many thanks
  Dave