[U-Boot] [PATCH 2/2] armv8/ls1046a: RGMII PHY requires internal delay on Tx

[Madalin-cristian Bucur]

> -----Original Message-----
> From: York Sun
> Sent: Wednesday, August 02, 2017 11:36 PM
> To: Madalin-cristian Bucur <madalin.bucur at nxp.com>; u-boot at lists.denx.de;
> joe.hershberger at ni.com
> Cc: Mingkai Hu <mingkai.hu at nxp.com>
> Subject: Re: [PATCH 2/2] armv8/ls1046a: RGMII PHY requires internal delay
> on Tx
> 
> On 04/04/2017 04:44 AM, Madalin Bucur wrote:
> > Signed-off-by: Madalin Bucur <madalin.bucur at nxp.com>
> > ---
> >   drivers/net/fm/ls1046.c | 4 ++--
> >   1 file changed, 2 insertions(+), 2 deletions(-)
> >
> > diff --git a/drivers/net/fm/ls1046.c b/drivers/net/fm/ls1046.c
> > index bf55554..6c91fb9 100644
> > --- a/drivers/net/fm/ls1046.c
> > +++ b/drivers/net/fm/ls1046.c
> > @@ -72,12 +72,12 @@ phy_interface_t fman_port_enet_if(enum fm_port port)
> >   	if (port == FM1_DTSEC3)
> >   		if ((rcwsr13 & FSL_CHASSIS2_RCWSR13_EC1) ==
> >   				FSL_CHASSIS2_RCWSR13_EC1_DTSEC3_RGMII)
> > -			return PHY_INTERFACE_MODE_RGMII;
> > +			return PHY_INTERFACE_MODE_RGMII_TXID;
> >
> >   	if (port == FM1_DTSEC4)
> >   		if ((rcwsr13 & FSL_CHASSIS2_RCWSR13_EC2) ==
> >   				FSL_CHASSIS2_RCWSR13_EC2_DTSEC4_RGMII)
> > -			return PHY_INTERFACE_MODE_RGMII;
> > +			return PHY_INTERFACE_MODE_RGMII_TXID;
> >
> >   	/* handle SGMII, only MAC 2/5/6/9/10 available */
> >   	switch (port) {
> >
> 
> I don't know what test you have run. This change breaks network on
> LS1043ARDB FM1 at DTSEC3.
> 
> York

Hi York,

You're likely using an old kernel, if you move to one that has removed the
hardcoded internal delay for RGMII [1] you'll notice that RGMII no longer
works with the old u-boot and you need this patch. We're in this situation
because someone decided to hardcode internal delay for RGMII mode [2] in the
Linux RTL8211F PHY driver. Someone else hardcoded the PHY fixup in u-boot to
RGMII instead of RGMII_TXID [3] to align to that initial mistake. The issue
with this u-boot fixup is that it overwrites device tree information, so one
cannot correct this from the dts.

Regards,
Madalin

[1] Linux kernel commit that removes hardcoded Tx internal delay for RTL8211F

commit e3230494b57ece68750e3e32d3e53d6b00917058
Author: Martin Blumenstingl <martin.blumenstingl at googlemail.com>
Date:   Fri Nov 25 14:12:01 2016 +0100

    net: phy: realtek: fix enabling of the TX-delay for RTL8211F

    The old logic always enabled the TX-delay when the phy-mode was set to
    PHY_INTERFACE_MODE_RGMII. There are dedicated phy-modes which tell the
    PHY driver to enable the RX and/or TX delays:
    - PHY_INTERFACE_MODE_RGMII should disable the RX and TX delay in the
      PHY (if required, the MAC should add the delays in this case)
    - PHY_INTERFACE_MODE_RGMII_ID should enable RX and TX delay in the PHY
    - PHY_INTERFACE_MODE_RGMII_TXID should enable the TX delay in the PHY
    - PHY_INTERFACE_MODE_RGMII_RXID should enable the RX delay in the PHY
      (currently not supported by RTL8211F)

    With this patch we enable the TX delay for PHY_INTERFACE_MODE_RGMII_ID
    and PHY_INTERFACE_MODE_RGMII_TXID.
    Additionally we now explicity disable the TX-delay, which seems to be
    enabled automatically after a hard-reset of the PHY (by triggering it's
    reset pin) to get a consistent state (as defined by the phy-mode).

    This fixes a compatibility problem with some SoCs where the TX-delay was
    also added by the MAC. With the TX-delay being applied twice the TX
    clock was off and TX traffic was broken or very slow (<10Mbit/s) on
    1000Mbit/s links.

    Signed-off-by: Martin Blumenstingl <martin.blumenstingl at googlemail.com>
    Reviewed-by: Florian Fainelli <f.fainelli at gmail.com>
    Signed-off-by: David S. Miller <davem at davemloft.net>

[2] Linux patch that hardcoded Tx ID for RGMII on RTL8211F

commit 3447cf2e9a11fd672de8b3cc613eabd97fe82053
Author: Shengzhou Liu <Shengzhou.Liu at freescale.com>
Date:   Thu Jun 18 16:42:47 2015 +0800

    net/phy: Add support for Realtek RTL8211F

    RTL8211F has different register definitions from RTL8211E.
    Specially it needs to enable TXDLY in case of RGMII.

    Signed-off-by: Shengzhou Liu <Shengzhou.Liu at freescale.com>
    Signed-off-by: David S. Miller <davem at davemloft.net>

[3] u-boot commit that hardcoded RGMII instead of the required RGMII_TXID

commit 9d3b8bd166e6c23ab83d0dcd49d81c6f0bf39e17
Author: Mingkai Hu <mingkai.hu at nxp.com>
Date:   Tue Jul 5 16:01:56 2016 +0800

    drivers: net/fm: Add Fman support for LS1046A

    The Fman module on LS1046A is similiar with that on LS1043A but
    LS1046A has one more XFI (10GbE) interface.

    Signed-off-by: Shaohui Xie <Shaohui.Xie at nxp.com>
    Signed-off-by: Mingkai Hu <mingkai.hu at nxp.com>
    Signed-off-by: Gong Qianyu <Qianyu.Gong at nxp.com>
    Reviewed-by: York Sun <york.sun at nxp.com>