netdev - Re: [PATCH net-next] net: phy: fix autoneg mismatch case in genphy_read

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: <c5421a7b-dcb5-cfb1-1123-90a785a951e6@gmail.com>
Date:   Fri, 5 Apr 2019 19:40:51 +0200
From:   Heiner Kallweit <hkallweit1@...il.com>
To:     Simon Horman <horms@...ge.net.au>
Cc:     Andrew Lunn <andrew@...n.ch>,
        Florian Fainelli <f.fainelli@...il.com>,
        David Miller <davem@...emloft.net>, netdev@...r.kernel.org,
        linux-renesas-soc@...r.kernel.org
Subject: Re: [PATCH net-next] net: phy: fix autoneg mismatch case in
 genphy_read_status

On 05.04.2019 18:16, Simon Horman wrote:
> On Tue, Apr 02, 2019 at 08:43:30PM +0200, Heiner Kallweit wrote:
>> The original patch didn't consider the case that autoneg process
>> finishes successfully but both link partners have no mode in common.
>> In this case there's no link, nevertheless we may be interested in
>> what the link partner advertised.
>>
>> Like phydev->link we set phydev->autoneg_complete in
>> genphy_update_link() and use the stored value in genphy_read_status().
>> This way we don't have to read register BMSR again.
>>
>> Fixes: b6163f194c69 ("net: phy: improve genphy_read_status")
>> Signed-off-by: Heiner Kallweit <hkallweit1@...il.com>
> 
> Hi,
> 
> I have observed a regression with this patch as present as
> 4950c2ba49cc ("net: phy: fix autoneg mismatch case in genphy_read_status")
> in net-next.
> 
> The platform is the Renesas R-Car E3 (r8a77990) based Ebisu-4D board.
> 
> 
> Without this patch (commit before 4950c2ba49cc) the link is negotiated
> as follows:
> 
> [    3.257699] libphy: ravb_mii: probed
> [    3.266498] ravb e6800000.ethernet eth0: Base address at 0xe6800000, 2e:09:0a:03:85:38, IRQ 104.
> [    3.537082] Micrel KSZ9031 Gigabit PHY e6800000.ethernet-ffffffff:00: attached PHY driver [Micrel KSZ9031 Gigabit PHY] (mii_bus:phy_addr=e6800000.ethernet-ffffffff:00, irq=165)
> [    9.667161] ravb e6800000.ethernet eth0: Link is Up - 1Gbps/Full - flow control off
> 
> # ethtool --version
> ethtool version 3.16
> # ethtool eth0
> Settings for eth0:
>         Supported ports: [ TP MII ]
>         Supported link modes:   100baseT/Full 
>                                 1000baseT/Full 
>         Supported pause frame use: No
>         Supports auto-negotiation: Yes
>         Advertised link modes:  100baseT/Full 
>                                 1000baseT/Full 
>         Advertised pause frame use: No
>         Advertised auto-negotiation: Yes
>         Link partner advertised link modes:  10baseT/Half 10baseT/Full 
>                                              100baseT/Half 100baseT/Full 
>                                              1000baseT/Full 
>         Link partner advertised pause frame use: No
>         Link partner advertised auto-negotiation: Yes
>         Speed: 1000Mb/s
>         Duplex: Full
>         Port: MII
>         PHYAD: 0
>         Transceiver: external
>         Auto-negotiation: on
>         Supports Wake-on: g
>         Wake-on: d
>         Current message level: 0x000000cc (204)
>                                link timer rx_err tx_err
>         Link detected: yes
> 
> 
> With this patch the link is negotiated as follows, note the "Unknown":
> 
> [    3.268609] libphy: ravb_mii: probed
> [    3.277402] ravb e6800000.ethernet eth0: Base address at 0xe6800000, 2e:09:0a:03:85:38, IRQ 104.
> [    3.565057] Micrel KSZ9031 Gigabit PHY e6800000.ethernet-ffffffff:00: attached PHY driver [Micrel KSZ9031 Gigabit PHY] (mii_bus:phy_addr=e6800000.ethernet-ffffffff:00, irq=165)
> [    9.804423] ravb e6800000.ethernet eth0: Link is Up - Unknown/Unknown - flow control off
> 
Looks like the PHY doesn't set the "aneg complete" bit. But then, how
can the link be up? Could you please add a debug output in
genphy_update_link() printing the value of register BMSR?
I wonder whether your case may be caused by a chip erratum. Item 5 in
the errata documentation
(http://ww1.microchip.com/downloads/en/DeviceDoc/80000692D.pdf)
may be a candidate. Could you please check (as explained in the
errata document) and report the exact PHY version?

Could you please also test with another link partner?

> And ethtool reports:
> 
> # ethtool eth0
> Settings for eth0:
>         Supported ports: [ TP MII ]
>         Supported link modes:   100baseT/Full 
>                                 1000baseT/Full 
>         Supported pause frame use: No
>         Supports auto-negotiation: Yes
>         Advertised link modes:  100baseT/Full 
>                                 1000baseT/Full 
>         Advertised pause frame use: No
>         Advertised auto-negotiation: Yes
>         Speed: Unknown!
>         Duplex: Unknown! (255)
>         Port: MII
>         PHYAD: 0
>         Transceiver: external
>         Auto-negotiation: on
>         Supports Wake-on: g
>         Wake-on: d
>         Current message level: 0x000000cc (204)
>                                link timer rx_err tx_err
>         Link detected: yes
> 
> 
> 
> I noticed this when preparing a patch limit the maximum speed the device to
> 100Mbit/s as follows:
> 
> --- a/arch/arm64/boot/dts/renesas/r8a77990-ebisu.dts
> +++ b/arch/arm64/boot/dts/renesas/r8a77990-ebisu.dts
> @@ -272,6 +272,7 @@
>                 interrupt-parent = <&gpio2>;
>                 interrupts = <21 IRQ_TYPE_LEVEL_LOW>;
>                 reset-gpios = <&gpio1 20 GPIO_ACTIVE_LOW>;
> +               max-speed = <100>;
>         };
>  };
>  
> While the 100Mbit/s limit applied on top of the commit before 4950c2ba49cc
> things work as expected:
> 
> [    3.258347] libphy: ravb_mii: probed
> [    3.266739] ravb e6800000.ethernet eth0: Base address at 0xe6800000, 2e:09:0a:03:85:38, IRQ 104.
> [    3.553642] Micrel KSZ9031 Gigabit PHY e6800000.ethernet-ffffffff:00: attached PHY driver [Micrel KSZ9031 Gigabit PHY] (mii_bus:phy_addr=e6800000.ethernet-ffffffff:00, irq=165)
> [    5.585027] ravb e6800000.ethernet eth0: Link is Up - 100Mbps/Full - flow control off
> 
> # ethtool eth0
> Settings for eth0:
>         Supported ports: [ TP MII ]
>         Supported link modes:   100baseT/Full 
>         Supported pause frame use: No
>         Supports auto-negotiation: Yes
>         Advertised link modes:  100baseT/Full 
>         Advertised pause frame use: No
>         Advertised auto-negotiation: Yes
>         Link partner advertised link modes:  10baseT/Half 10baseT/Full 
>                                              100baseT/Half 100baseT/Full 
I wondered why the link partner suddenly doesn't report 1Gbps support.
Seems to be a small flaw in phylib, if locally 1Gbps isn't supported
(or support is disabled) then we don't check for the link partners
1Gbps capability.

>         Link partner advertised pause frame use: No
>         Link partner advertised auto-negotiation: Yes
>         Speed: 100Mb/s
>         Duplex: Full
>         Port: MII
>         PHYAD: 0
>         Transceiver: external
>         Auto-negotiation: on
>         Supports Wake-on: g
>         Wake-on: d
>         Current message level: 0x000000cc (204)
>                                link timer rx_err tx_err
>         Link detected: yes
> 
> 
> However, with the 100Mbit/s limit applied on top of 4950c2ba49cc the boot hangs
> while (or just after?) negotiating the link:
> 
> [    3.541549] Micrel KSZ9031 Gigabit PHY e6800000.ethernet-ffffffff:00: attached PHY driver [Micrel KSZ9031 Gigabit PHY] (mii_bus:phy_addr=e6800000.ethernet-ffffffff:00, irq=165)
> [    5.536389] ravb e6800000.ethernet eth0: Link is Up - Unknown/Unknown - flow control off
> 
>