netdev - Re: [PATCH v1 1/2] driver core: fw_devlink: Add support for FWNODE_FLAG_BROKEN

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:   Tue, 31 Aug 2021 18:46:59 -0700
From:   Saravana Kannan <saravanak@...gle.com>
To:     Andrew Lunn <andrew@...n.ch>
Cc:     Greg Kroah-Hartman <gregkh@...uxfoundation.org>,
        "Rafael J. Wysocki" <rafael@...nel.org>,
        Linus Walleij <linus.walleij@...aro.org>,
        Vivien Didelot <vivien.didelot@...il.com>,
        Florian Fainelli <f.fainelli@...il.com>,
        Vladimir Oltean <olteanv@...il.com>,
        "David S. Miller" <davem@...emloft.net>,
        Jakub Kicinski <kuba@...nel.org>, Len Brown <lenb@...nel.org>,
        Alvin Sipraga <ALSI@...g-olufsen.dk>, kernel-team@...roid.com,
        linux-kernel@...r.kernel.org, netdev@...r.kernel.org,
        linux-acpi@...r.kernel.org
Subject: Re: [PATCH v1 1/2] driver core: fw_devlink: Add support for FWNODE_FLAG_BROKEN_PARENT

On Tue, Aug 31, 2021 at 4:02 PM Andrew Lunn <andrew@...n.ch> wrote:
>
> > If the switches are broken without the phy-handle or ethernet change,
> > I'm not sure if the "BROKEN_PARENT" patch would help.
>
> > > Which is not enough to fix these Ethernet switches.
> >
> > Ok, if you can give more specifics on this, I'll look into it.
>
> The switches probe, but get the wrong PHY driver, genphy, not the
> Marvell PHY driver. And genphy is not sufficient for this hardware.
>
> I'd need:
> > 1) The DTS file that you see the issue on.
>
> I did the bisect on arch/arm/boot/dts/vf610-zii-dev-rev-c.dts but i
> also tested arch/arm/boot/dts/vf610-zii-dev-rev-b.dts.

Thanks for the detailed info Andrew. I looked at the DT files. So
yeah, this is similar to the realtek issue and my generic fix for DSA
should work for all of these unless I'm forgetting something. Please
let me know if it doesn't.

-Saravana

> Rev B is interesting because switch0 and switch1 got genphy, while
> switch2 got the correct Marvell PHY driver. switch2 PHYs don't have
> interrupt properties, so don't loop back to their parent device.
>
> Here is Rev B. I trimmed out other devices probing in parallel:
>
> [    1.029100] fec 400d1000.ethernet: Invalid MAC address: 00:00:00:00:00:00
> [    1.034735] fec 400d1000.ethernet: Using random MAC address: 42:f2:14:33:78:f5
> [    1.042272] libphy: fec_enet_mii_bus: probed
> [    1.455932] libphy: mdio_mux: probed
> [    1.459432] mv88e6085 0.1:00: switch 0x3520 detected: Marvell 88E6352, revision 1
> [    1.494076] libphy: mdio: probed
> [    1.518958] libphy: mdio_mux: probed
> [    1.522553] mv88e6085 0.2:00: switch 0x3520 detected: Marvell 88E6352, revision 1
> [    1.537295] libphy: mdio: probed
> [    1.556571] libphy: mdio_mux: probed
> [    1.559719] mv88e6085 0.4:00: switch 0x1a70 detected: Marvell 88E6185, revision 2
> [    1.574614] libphy: mdio: probed
> [    1.733104] mv88e6085 0.1:00 lan0 (uninitialized): PHY [!mdio-mux!mdio@...witch@...dio:00] driver [Generic PHY] (irq=POLL)
> [    1.750737] mv88e6085 0.1:00 lan1 (uninitialized): PHY [!mdio-mux!mdio@...witch@...dio:01] driver [Generic PHY] (irq=POLL)
> [    1.768273] mv88e6085 0.1:00 lan2 (uninitialized): PHY [!mdio-mux!mdio@...witch@...dio:02] driver [Generic PHY] (irq=POLL)
> [    1.806561] mv88e6085 0.2:00 lan3 (uninitialized): PHY [!mdio-mux!mdio@...witch@...dio:00] driver [Generic PHY] (irq=POLL)
> [    1.824033] mv88e6085 0.2:00 lan4 (uninitialized): PHY [!mdio-mux!mdio@...witch@...dio:01] driver [Generic PHY] (irq=POLL)
> [    1.841496] mv88e6085 0.2:00 lan5 (uninitialized): PHY [!mdio-mux!mdio@...witch@...dio:02] driver [Generic PHY] (irq=POLL)
> [    1.943535] mv88e6085 0.4:00 lan6 (uninitialized): PHY [!mdio-mux!mdio@...witch@...dio:00] driver [Marvell 88E1545] (irq=POLL)
> [    2.003529] mv88e6085 0.4:00 lan7 (uninitialized): PHY [!mdio-mux!mdio@...witch@...dio:01] driver [Marvell 88E1545] (irq=POLL)
> [    2.063535] mv88e6085 0.4:00 lan8 (uninitialized): PHY [!mdio-mux!mdio@...witch@...dio:02] driver [Marvell 88E1545] (irq=POLL)
> [    2.084768] DSA: tree 0 setup
> [    2.087791] libphy: mdio_mux: probed
> [    2.265477] Micrel KSZ8041 400d0000.ethernet-1:00: attached PHY driver (mii_bus:phy_addr=400d0000.ethernet-1:00, irq=POLL)
>
> root@...-devel-b:~# cat /sys/kernel/debug/devices_deferred
> root@...-devel-b:~#
>
> For Rev C we see:
>
> [    1.244417] fec 400d1000.ethernet: Invalid MAC address: 00:00:00:00:00:00
> [    1.250081] fec 400d1000.ethernet: Using random MAC address: c6:42:89:ed:5f:dd
> [    1.257507] libphy: fec_enet_mii_bus: probed
> [    1.570725] libphy: mdio_mux: probed
> [    1.574208] mv88e6085 0.1:00: switch 0xa10 detected: Marvell 88E6390X, revision 1
> [    1.590272] libphy: mdio: probed
> [    1.627721] libphy: mdio_mux: probed
> [    1.631222] mv88e6085 0.2:00: switch 0xa10 detected: Marvell 88E6390X, revision 1
> [    1.659643] libphy: mdio: probed
> [    1.811665] mv88e6085 0.1:00 lan1 (uninitialized): PHY [!mdio-mux!mdio@...witch@...dio:01] driver [Generic PHY] (irq=POLL)
> [    1.829230] mv88e6085 0.1:00 lan2 (uninitialized): PHY [!mdio-mux!mdio@...witch@...dio:02] driver [Generic PHY] (irq=POLL)
> [    1.845884] mv88e6085 0.1:00 lan3 (uninitialized): PHY [!mdio-mux!mdio@...witch@...dio:03] driver [Generic PHY] (irq=POLL)
> [    1.863237] mv88e6085 0.1:00 lan4 (uninitialized): PHY [!mdio-mux!mdio@...witch@...dio:04] driver [Generic PHY] (irq=POLL)
> [    1.884078] mv88e6085 0.2:00 lan5 (uninitialized): PHY [!mdio-mux!mdio@...witch@...dio:01] driver [Generic PHY] (irq=POLL)
> [    1.901630] mv88e6085 0.2:00 lan6 (uninitialized): PHY [!mdio-mux!mdio@...witch@...dio:02] driver [Generic PHY] (irq=POLL)
> [    1.918287] mv88e6085 0.2:00 lan7 (uninitialized): PHY [!mdio-mux!mdio@...witch@...dio:03] driver [Generic PHY] (irq=POLL)
> [    1.933721] mv88e6085 0.2:00 lan8 (uninitialized): PHY [!mdio-mux!mdio@...witch@...dio:04] driver [Generic PHY] (irq=POLL)
> [    1.948722] DSA: tree 0 setup
> [    1.951599] libphy: mdio_mux: probed
>
> [   21.565550] Micrel KSZ8041 400d0000.ethernet-1:00: attached PHY driver (mii_bus:phy_addr=400d0000.ethernet-1:00, irq=48)
>
> I have Rev B using NFS root, so the interfaces are configured up by
> the kernel during boot. Rev C has a local root filesystem, so user
> space brings the interfaces up, and it is only when the FEC is opened
> does it attach to the Micrel PHY. That explains the difference between
> 2.265 and 21.565 seconds for the last line.
>
> Again, nothing deferred.
>
>        Andrew