netdev - Re: [PATCH net-next 2/2] net: pcs: xpcs: fix incorrect CL37 AN sequence

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: <20210928102730.72vays22sulixodb@skbuf>
Date:   Tue, 28 Sep 2021 10:27:31 +0000
From:   Vladimir Oltean <vladimir.oltean@....com>
To:     Wong Vee Khee <vee.khee.wong@...ux.intel.com>
CC:     Andrew Lunn <andrew@...n.ch>, Jose Abreu <Jose.Abreu@...opsys.com>,
        Heiner Kallweit <hkallweit1@...il.com>,
        Russell King <linux@...linux.org.uk>,
        "David S . Miller" <davem@...emloft.net>,
        Jakub Kicinski <kuba@...nel.org>,
        "netdev@...r.kernel.org" <netdev@...r.kernel.org>,
        "linux-kernel@...r.kernel.org" <linux-kernel@...r.kernel.org>,
        Wong Vee Khee <veekhee@...il.com>
Subject: Re: [PATCH net-next 2/2] net: pcs: xpcs: fix incorrect CL37 AN
 sequence

On Tue, Sep 28, 2021 at 12:19:38PM +0800, Wong Vee Khee wrote:
> According to Synopsys DesignWare Cores Ethernet PCS databook, it is
> required to disable Clause 37 auto-negotiation by programming bit-12
> (AN_ENABLE) to 0 if it is already enabled, before programming various
> fields of VR_MII_AN_CTRL registers.
> 
> After all these programming are done, it is then required to enable
> Clause 37 auto-negotiation by programming bit-12 (AN_ENABLE) to 1.
> 
> Cc: Vladimir Oltean <vladimir.oltean@....com>
> Signed-off-by: Wong Vee Khee <vee.khee.wong@...ux.intel.com>
> ---



>  drivers/net/pcs/pcs-xpcs.c | 24 ++++++++++++++++++++----
>  1 file changed, 20 insertions(+), 4 deletions(-)
> 
> diff --git a/drivers/net/pcs/pcs-xpcs.c b/drivers/net/pcs/pcs-xpcs.c
> index da8c81d25edd..eacfb32bb229 100644
> --- a/drivers/net/pcs/pcs-xpcs.c
> +++ b/drivers/net/pcs/pcs-xpcs.c
> @@ -709,11 +709,14 @@ static int xpcs_config_aneg_c37_sgmii(struct dw_xpcs *xpcs, unsigned int mode)
>  	int ret;
>  
>  	/* For AN for C37 SGMII mode, the settings are :-
> -	 * 1) VR_MII_AN_CTRL Bit(2:1)[PCS_MODE] = 10b (SGMII AN)
> -	 * 2) VR_MII_AN_CTRL Bit(3) [TX_CONFIG] = 0b (MAC side SGMII)
> +	 * 1) VR_MII_MMD_CTRL Bit(12) [AN_ENABLE] = 0b (Disable SGMII AN in case
> +	      it is already enabled)
> +	 * 2) VR_MII_AN_CTRL Bit(2:1)[PCS_MODE] = 10b (SGMII AN)
> +	 * 3) VR_MII_AN_CTRL Bit(3) [TX_CONFIG] = 0b (MAC side SGMII)
>  	 *    DW xPCS used with DW EQoS MAC is always MAC side SGMII.
> -	 * 3) VR_MII_DIG_CTRL1 Bit(9) [MAC_AUTO_SW] = 1b (Automatic
> +	 * 4) VR_MII_DIG_CTRL1 Bit(9) [MAC_AUTO_SW] = 1b (Automatic
>  	 *    speed/duplex mode change by HW after SGMII AN complete)
> +	 * 5) VR_MII_MMD_CTRL Bit(12) [AN_ENABLE] = 1b (Enable SGMII AN)
>  	 *
>  	 * Note: Since it is MAC side SGMII, there is no need to set
>  	 *	 SR_MII_AN_ADV. MAC side SGMII receives AN Tx Config from
> @@ -721,6 +724,11 @@ static int xpcs_config_aneg_c37_sgmii(struct dw_xpcs *xpcs, unsigned int mode)
>  	 *	 between PHY and Link Partner. There is also no need to
>  	 *	 trigger AN restart for MAC-side SGMII.
>  	 */
> +	ret = xpcs_modify(xpcs, MDIO_MMD_VEND2, DW_VR_MII_MMD_CTRL, AN_CL37_EN,

Would you mind using "MDIO_CTRL1" instead of "DW_VR_MII_MMD_CTRL", and
"MDIO_AN_CTRL1_ENABLE" instead of "AN_CL37_EN"? After all, the
"SR_MII_CTRL" naming from the data book comes from "Standard Register".

> +			  0);
> +	if (ret < 0)
> +		return ret;
> +
>  	ret = xpcs_read(xpcs, MDIO_MMD_VEND2, DW_VR_MII_AN_CTRL);
>  	if (ret < 0)
>  		return ret;
> @@ -745,7 +753,15 @@ static int xpcs_config_aneg_c37_sgmii(struct dw_xpcs *xpcs, unsigned int mode)
>  	else
>  		ret &= ~DW_VR_MII_DIG_CTRL1_MAC_AUTO_SW;
>  
> -	return xpcs_write(xpcs, MDIO_MMD_VEND2, DW_VR_MII_DIG_CTRL1, ret);
> +	ret = xpcs_write(xpcs, MDIO_MMD_VEND2, DW_VR_MII_DIG_CTRL1, ret);

There is an interesting note in the data book about the MAC_AUTO_SW bit:

| This mode is valid only when DWC_xpcs is configured as MAC-side
| SGMII/USXGMII/QSGMII and should be set only when Auto-negotiation is
| enabled (AN_ENABLE bit is set to 1).

With your patch, it will be set while the AN_ENABLE bit is zero.
I wonder whether that's a poor choice of words, because it does
contradict with the sequence described in section "7.12.1 SGMII
Auto-Negotiation", and if it just means "only set MAC_AUTO_SW if you
intend to eventually set AN_ENABLE", not "you must set it only _while_
AN_ENABLE is set".

> +	if (ret < 0)
> +		return ret;
> +
> +	ret = xpcs_read(xpcs, MDIO_MMD_VEND2, DW_VR_MII_MMD_CTRL);
> +	if (ret < 0)
> +		return ret;
> +	ret |= AN_CL37_EN;

So this is the part that really confuses me.
The xpcs driver did not use to do any configuration at all for this bit.
Am I right in assuming that we were all operating with the default value?
Because in your case, the default value is set by your bootloader
(AN_CL37_EN is set), and in my case, the default is the hardware reset
(AN_CL37_EN is not set). I think (haven't tested) that this change would
actually break my setups where phylink_autoneg_inband(mode) returns false.

So could you please add this conditional?

	if (phylink_autoneg_inband(mode))
		ret |= MDIO_AN_CTRL1_ENABLE;
	else
		ret &= ~MDIO_AN_CTRL1_ENABLE;

> +	return xpcs_write(xpcs, MDIO_MMD_VEND2, DW_VR_MII_MMD_CTRL, ret);
>  }
>  
>  static int xpcs_config_2500basex(struct dw_xpcs *xpcs)
> -- 
> 2.25.1
>