linux-kernel.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: Vladimir Oltean <vladimir.oltean@nxp.com>
To: Michael Sit Wei Hong <michael.wei.hong.sit@intel.com>
Cc: "Jose.Abreu@synopsys.com" <Jose.Abreu@synopsys.com>,
	"andrew@lunn.ch" <andrew@lunn.ch>,
	"hkallweit1@gmail.com" <hkallweit1@gmail.com>,
	"linux@armlinux.org.uk" <linux@armlinux.org.uk>,
	"kuba@kernel.org" <kuba@kernel.org>,
	"netdev@vger.kernel.org" <netdev@vger.kernel.org>,
	"peppe.cavallaro@st.com" <peppe.cavallaro@st.com>,
	"alexandre.torgue@foss.st.com" <alexandre.torgue@foss.st.com>,
	"davem@davemloft.net" <davem@davemloft.net>,
	"mcoquelin.stm32@gmail.com" <mcoquelin.stm32@gmail.com>,
	"weifeng.voon@intel.com" <weifeng.voon@intel.com>,
	"boon.leong.ong@intel.com" <boon.leong.ong@intel.com>,
	"tee.min.tan@intel.com" <tee.min.tan@intel.com>,
	"vee.khee.wong@linux.intel.com" <vee.khee.wong@linux.intel.com>,
	"vee.khee.wong@intel.com" <vee.khee.wong@intel.com>,
	"linux-stm32@st-md-mailman.stormreply.com" 
	<linux-stm32@st-md-mailman.stormreply.com>,
	"linux-arm-kernel@lists.infradead.org" 
	<linux-arm-kernel@lists.infradead.org>,
	"linux-kernel@vger.kernel.org" <linux-kernel@vger.kernel.org>
Subject: Re: [RESEND PATCH net-next v5 2/3] net: pcs: add 2500BASEX support for Intel mGbE controller
Date: Fri, 4 Jun 2021 12:00:18 +0000	[thread overview]
Message-ID: <20210604120017.gbpwm27jjwoi2b32@skbuf> (raw)
In-Reply-To: <20210604105733.31092-3-michael.wei.hong.sit@intel.com>

On Fri, Jun 04, 2021 at 06:57:32PM +0800, Michael Sit Wei Hong wrote:
> From: Voon Weifeng <weifeng.voon@intel.com>
> 
> XPCS IP supports 2500BASEX as PHY interface. It is configured as
> autonegotiation disable to cater for PHYs that does not supports 2500BASEX
> autonegotiation.
> 
> v2: Add supported link speed masking.
> v3: Restructure to introduce xpcs_config_2500basex() used to configure the
>     xpcs for 2.5G speeds. Added 2500BASEX specific information for
>     configuration.
> v4: Fix indentation error
> 
> Signed-off-by: Voon Weifeng <weifeng.voon@intel.com>
> Signed-off-by: Michael Sit Wei Hong <michael.wei.hong.sit@intel.com>
> ---
>  drivers/net/pcs/pcs-xpcs.c   | 56 ++++++++++++++++++++++++++++++++++++
>  include/linux/pcs/pcs-xpcs.h |  1 +
>  2 files changed, 57 insertions(+)
> 
> diff --git a/drivers/net/pcs/pcs-xpcs.c b/drivers/net/pcs/pcs-xpcs.c
> index 34164437c135..98c4a3973402 100644
> --- a/drivers/net/pcs/pcs-xpcs.c
> +++ b/drivers/net/pcs/pcs-xpcs.c
> @@ -57,9 +57,12 @@
>  
>  /* Clause 37 Defines */
>  /* VR MII MMD registers offsets */
> +#define DW_VR_MII_MMD_CTRL		0x0000
>  #define DW_VR_MII_DIG_CTRL1		0x8000
>  #define DW_VR_MII_AN_CTRL		0x8001
>  #define DW_VR_MII_AN_INTR_STS		0x8002
> +/* Enable 2.5G Mode */
> +#define DW_VR_MII_DIG_CTRL1_2G5_EN	BIT(2)
>  /* EEE Mode Control Register */
>  #define DW_VR_MII_EEE_MCTRL0		0x8006
>  #define DW_VR_MII_EEE_MCTRL1		0x800b
> @@ -86,6 +89,11 @@
>  #define DW_VR_MII_C37_ANSGM_SP_1000		0x2
>  #define DW_VR_MII_C37_ANSGM_SP_LNKSTS		BIT(4)
>  
> +/* SR MII MMD Control defines */
> +#define AN_CL37_EN		BIT(12)	/* Enable Clause 37 auto-nego */
> +#define SGMII_SPEED_SS13	BIT(13)	/* SGMII speed along with SS6 */
> +#define SGMII_SPEED_SS6		BIT(6)	/* SGMII speed along with SS13 */
> +
>  /* VR MII EEE Control 0 defines */
>  #define DW_VR_MII_EEE_LTX_EN		BIT(0)  /* LPI Tx Enable */
>  #define DW_VR_MII_EEE_LRX_EN		BIT(1)  /* LPI Rx Enable */
> @@ -161,6 +169,14 @@ static const int xpcs_sgmii_features[] = {
>  	__ETHTOOL_LINK_MODE_MASK_NBITS,
>  };
>  
> +static const int xpcs_2500basex_features[] = {
> +	ETHTOOL_LINK_MODE_Asym_Pause_BIT,
> +	ETHTOOL_LINK_MODE_Autoneg_BIT,
> +	ETHTOOL_LINK_MODE_2500baseX_Full_BIT,
> +	ETHTOOL_LINK_MODE_2500baseT_Full_BIT,
> +	__ETHTOOL_LINK_MODE_MASK_NBITS,
> +};
> +

This is a general design comment, perhaps you could address this later,
but do keep it in mind:

I don't think the PCS has anything to do with whether the link will
support flow control.
Similarly, Aquantia (now Marvell) PHYs operating in 2500base-x mode are
capable of negotiating the copper-side link to 10/100/1000/2500 even if
the system-side link is fixed at 2500. The way this is achieved is by
the PHY emitting PAUSE frames towards the MAC in order to achieve rate
adaptation with the external link speed.
This is not completely standardized in phylink at the moment, but we
have systems where this works. My point is that maybe the PCS driver
isn't the most appropriate place to implement the phylink_validate
method - the MAC driver is almost always in the position to know better.

If you could move the xpcs validation inside stmmac I think that would
be an improvement. For example with the NXP SJA1105 patches that I am
going to send out for review soon, I am not calling xpcs_validate() at
all.

>  static const phy_interface_t xpcs_usxgmii_interfaces[] = {
>  	PHY_INTERFACE_MODE_USXGMII,
>  };
> @@ -177,11 +193,17 @@ static const phy_interface_t xpcs_sgmii_interfaces[] = {
>  	PHY_INTERFACE_MODE_SGMII,
>  };
>  
> +static const phy_interface_t xpcs_2500basex_interfaces[] = {
> +	PHY_INTERFACE_MODE_2500BASEX,
> +	PHY_INTERFACE_MODE_MAX,
> +};
> +
>  enum {
>  	DW_XPCS_USXGMII,
>  	DW_XPCS_10GKR,
>  	DW_XPCS_XLGMII,
>  	DW_XPCS_SGMII,
> +	DW_XPCS_2500BASEX,
>  	DW_XPCS_INTERFACE_MAX,
>  };
>  
> @@ -306,6 +328,7 @@ static int xpcs_soft_reset(struct mdio_xpcs_args *xpcs,
>  		dev = MDIO_MMD_PCS;
>  		break;
>  	case DW_AN_C37_SGMII:
> +	case DW_2500BASEX:
>  		dev = MDIO_MMD_VEND2;
>  		break;
>  	default:
> @@ -804,6 +827,28 @@ static int xpcs_config_aneg_c37_sgmii(struct mdio_xpcs_args *xpcs)
>  	return xpcs_write(xpcs, MDIO_MMD_VEND2, DW_VR_MII_DIG_CTRL1, ret);
>  }
>  
> +static int xpcs_config_2500basex(struct mdio_xpcs_args *xpcs)
> +{
> +	int ret;
> +
> +	ret = xpcs_read(xpcs, MDIO_MMD_VEND2, DW_VR_MII_DIG_CTRL1);
> +	if (ret < 0)
> +		return ret;
> +	ret |= DW_VR_MII_DIG_CTRL1_2G5_EN;
> +	ret &= ~DW_VR_MII_DIG_CTRL1_MAC_AUTO_SW;
> +	ret = xpcs_write(xpcs, MDIO_MMD_VEND2, DW_VR_MII_DIG_CTRL1, ret);
> +	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;
> +	ret |= SGMII_SPEED_SS6;
> +	ret &= ~SGMII_SPEED_SS13;
> +	return xpcs_write(xpcs, MDIO_MMD_VEND2, DW_VR_MII_MMD_CTRL, ret);
> +}
> +
>  static int xpcs_do_config(struct mdio_xpcs_args *xpcs,
>  			  phy_interface_t interface, unsigned int mode)
>  {
> @@ -827,6 +872,11 @@ static int xpcs_do_config(struct mdio_xpcs_args *xpcs,
>  		if (ret)
>  			return ret;
>  		break;
> +	case DW_2500BASEX:
> +		ret = xpcs_config_2500basex(xpcs);
> +		if (ret)
> +			return ret;
> +		break;
>  	default:
>  		return -1;
>  	}
> @@ -1023,6 +1073,12 @@ static const struct xpcs_compat synopsys_xpcs_compat[DW_XPCS_INTERFACE_MAX] = {
>  		.num_interfaces = ARRAY_SIZE(xpcs_sgmii_interfaces),
>  		.an_mode = DW_AN_C37_SGMII,
>  	},
> +	[DW_XPCS_2500BASEX] = {
> +		.supported = xpcs_2500basex_features,
> +		.interface = xpcs_2500basex_interfaces,
> +		.num_interfaces = ARRAY_SIZE(xpcs_2500basex_features),
> +		.an_mode = DW_2500BASEX,
> +	},
>  };
>  
>  static const struct xpcs_id xpcs_id_list[] = {
> diff --git a/include/linux/pcs/pcs-xpcs.h b/include/linux/pcs/pcs-xpcs.h
> index 0860a5b59f10..4d815f03b4b2 100644
> --- a/include/linux/pcs/pcs-xpcs.h
> +++ b/include/linux/pcs/pcs-xpcs.h
> @@ -13,6 +13,7 @@
>  /* AN mode */
>  #define DW_AN_C73			1
>  #define DW_AN_C37_SGMII			2
> +#define DW_2500BASEX			3
>  
>  struct xpcs_id;
>  
> -- 
> 2.17.1
> 

  reply	other threads:[~2021-06-04 12:00 UTC|newest]

Thread overview: 8+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2021-06-04 10:57 [RESEND PATCH net-next v5 0/3] Enable 2.5Gbps speed for stmmac Michael Sit Wei Hong
2021-06-04 10:57 ` [RESEND PATCH net-next v5 1/3] net: stmmac: split xPCS setup from mdio register Michael Sit Wei Hong
2021-06-04 11:51   ` Vladimir Oltean
2021-06-04 13:08     ` Vladimir Oltean
2021-06-04 10:57 ` [RESEND PATCH net-next v5 2/3] net: pcs: add 2500BASEX support for Intel mGbE controller Michael Sit Wei Hong
2021-06-04 12:00   ` Vladimir Oltean [this message]
2021-06-04 10:57 ` [RESEND PATCH net-next v5 3/3] net: stmmac: enable Intel mGbE 2.5Gbps link speed Michael Sit Wei Hong
2021-06-04 12:02   ` Vladimir Oltean

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=20210604120017.gbpwm27jjwoi2b32@skbuf \
    --to=vladimir.oltean@nxp.com \
    --cc=Jose.Abreu@synopsys.com \
    --cc=alexandre.torgue@foss.st.com \
    --cc=andrew@lunn.ch \
    --cc=boon.leong.ong@intel.com \
    --cc=davem@davemloft.net \
    --cc=hkallweit1@gmail.com \
    --cc=kuba@kernel.org \
    --cc=linux-arm-kernel@lists.infradead.org \
    --cc=linux-kernel@vger.kernel.org \
    --cc=linux-stm32@st-md-mailman.stormreply.com \
    --cc=linux@armlinux.org.uk \
    --cc=mcoquelin.stm32@gmail.com \
    --cc=michael.wei.hong.sit@intel.com \
    --cc=netdev@vger.kernel.org \
    --cc=peppe.cavallaro@st.com \
    --cc=tee.min.tan@intel.com \
    --cc=vee.khee.wong@intel.com \
    --cc=vee.khee.wong@linux.intel.com \
    --cc=weifeng.voon@intel.com \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).