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: <e0e2cc71-70b5-d3be-32c4-27d8dace00cf@gmail.com>
Date:   Tue, 25 Feb 2020 11:47:38 -0800
From:   Florian Fainelli <f.fainelli@...il.com>
To:     Sriram Chadalavada <sriram.chadalavada@...dleap.ca>,
        netdev@...r.kernel.org, Andrew Lunn <andrew@...n.ch>,
        Vivien Didelot <vivien.didelot@...il.com>
Subject: Re: Fwd: Kernel crash due to conflict between legacy and current DSA
 mv88e6xxx drivers ?

On 2/25/20 11:44 AM, Sriram Chadalavada wrote:
> The following error occurred using 4.19.101 kernel and NOT with our
> last stable 4.1.16 kernel on enabling CONFIG_NET_DSA_LEGACY.
> 
> [    1.505428] libphy: mdiobus_find: mii bus [igb_enet_mii_bus] found
> [    1.505515] sysfs: cannot create duplicate filename '/kernel/marvell/access'
> [    1.505529] CPU: 0 PID: 1 Comm: swapper/0 Not tainted 4.19.101 #0
> [    1.505534] Hardware name: Freescale i.MX6 Quad/DualLite (Device Tree)
> [    1.505539] Backtrace:
> [    1.505549] Function entered at [<80013a28>] from [<80013d60>]
> [    1.505561]  r7:00000000 r6:60000013 r5:00000000 r4:806437b4
> [    1.505565] Function entered at [<80013d48>] from [<804e94b4>]
> [    1.505570] Function entered at [<804e941c>] from [<8013fffc>]
> [    1.505579]  r7:00000000 r6:80575d54 r5:e931a000 r4:eee2d000
> [    1.505585] Function entered at [<8013ff9c>] from [<8013fcbc>]
> [    1.505593]  r7:00000000 r6:e931a000 r5:ffffffef r4:8063860c
> [    1.505599] Function entered at [<8013fb50>] from [<8013fd68>]
> [    1.505606]  r6:e932f880 r5:8063860c r4:00000000
> [    1.505612] Function entered at [<8013fcf4>] from [<802f3190>]
> [    1.505620]  r7:80673f0c r6:00000000 r5:e91df078 r4:00000000
> [    1.505626] Function entered at [<802f30f4>] from [<804df104>]
> [    1.505638]  r10:e932b940 r9:8052c8e8 r8:00000000 r7:e91df078
> r6:00000000 r5:802f30f4
> [    1.505644]  r4:80638598
> [    1.505649] Function entered at [<804deb8c>] from [<80271310>]
> [    1.505661]  r10:00000000 r9:8064366c r8:00000000 r7:80672acc
> r6:8064366c r5:eef36410
> [    1.505665]  r4:00000000
> [    1.505670] Function entered at [<802712c0>] from [<8026f7b0>]
> [    1.505679]  r7:80672acc r6:00000000 r5:80672ac8 r4:eef36410
> [    1.505684] Function entered at [<8026f65c>] from [<8026fb90>]
> 
> On disabling the option CONFIG_NET_DSA_LEGACY,  the crash goes away
> but the ethernet interfaces are NOT enumerated.
> 
> https://www.spinics.net/lists/netdev/msg556413.html seems to suggest
> that legacy DSA is still being used for Marvell 88e6xxx drivers. If
> the kernel community intends to move away from that, we would want to
> adopt that approach. Please advise on what we might have to
> investigate/change that will also enable resolving this kernel crash.
> As I understand, one of the things I would need to change is that the
> relevant device tree needs to use the "current binding" as opposed to
> the "deprecated binding" which is what is being used now.

Correct, and also it looks like you have out of tree changes applied, as
nothing in the vanialla kernel will try to create a
/kernel/marvell/access hierarchy under /sys.
-- 
Florian

Powered by blists - more mailing lists

Powered by Openwall GNU/*/Linux Powered by OpenVZ