[PATCH 0/3] ARM: dts: omap: workarounds for Ethernet PHY strap on DRA71 and DRA72 EVMs

[PATCH 0/3] ARM: dts: omap: workarounds for Ethernet PHY strap on DRA71 and DRA72 EVMs

[Sekhar Nori]

Hi Tony,

Here are some patches to overcome incorrect Ethernet PHY strap settings
on DRA72x EVM Rev C and DRA71 EVM. Both of these boards use TI's DP83867
PHY.

The incorrect PHY strap settings cause a voilation of PHY specification
and need to be worked around in software as documented in the commit
messages. But there is no known issue that these patches fix.

I did verify that there are no regressions with NFS boot on these boards
after the patches are applied on v4.12-rc4.

The series itself has been rebased on your omap-for-v4.13/dt branch since
it has some conflicts with patches you have queued.

Sekhar Nori (3):
  ARM: dts: dra71-evm: workaround incorrect DP83867 RX_CTRL pin strap
  ARM: dts: dra72-evm-revc: workaround incorrect DP83867 RX_CTRL pin
    strap
  ARM: omap2plus_defconfig: enable DP83867 phy driver

 arch/arm/boot/dts/dra71-evm.dts      | 2 ++
 arch/arm/boot/dts/dra72-evm-revc.dts | 2 ++
 arch/arm/configs/omap2plus_defconfig | 1 +
 3 files changed, 5 insertions(+)

-- 
2.9.0

[PATCH 1/3] ARM: dts: dra71-evm: workaround incorrect DP83867 RX_CTRL pin strap

[Sekhar Nori]

The DRA71 EVM straps the DP83867 GigaBit Ethernet phy's RX_DV/RX_CTRL pin
in mode 1. Unfortunately, the phy data manual disallows this.

Add "ti,dp83867-rxctrl-strap-quirk" property to the phy's device-tree node
to allow kernel to enable software workaround for this incorrect strap
setting. This is as suggested by the phy's datamanual and ensures proper
operation of this PHY.

This needs to be done for both instances of this PHY present on the board.

Signed-off-by: Sekhar Nori <nsekhar@ti.com>
---
 arch/arm/boot/dts/dra71-evm.dts | 2 ++
 1 file changed, 2 insertions(+)

diff --git a/arch/arm/boot/dts/dra71-evm.dts b/arch/arm/boot/dts/dra71-evm.dts
index 570f06e4f71f..c1a7ee09218f 100644
--- a/arch/arm/boot/dts/dra71-evm.dts
+++ b/arch/arm/boot/dts/dra71-evm.dts
@@ -205,6 +205,7 @@
 		ti,tx-internal-delay = <DP83867_RGMIIDCTL_250_PS>;
 		ti,fifo-depth = <DP83867_PHYCR_FIFO_DEPTH_8_B_NIB>;
 		ti,min-output-impedance;
+		ti,dp83867-rxctrl-strap-quirk;
 	};
 
 	dp83867_1: ethernet-phy at 3 {
@@ -213,6 +214,7 @@
 		ti,tx-internal-delay = <DP83867_RGMIIDCTL_250_PS>;
 		ti,fifo-depth = <DP83867_PHYCR_FIFO_DEPTH_8_B_NIB>;
 		ti,min-output-impedance;
+		ti,dp83867-rxctrl-strap-quirk;
 	};
 };
 
-- 
2.9.0

[PATCH 2/3] ARM: dts: dra72-evm-revc: workaround incorrect DP83867 RX_CTRL pin strap

[Sekhar Nori]

The DRA72 EVM Rev C straps the DP83867 GigaBit Ethernet phy's RX_DV/RX_CTRL
pin in mode 1. Unfortunately, the phy data manual disallows this.

Add "ti,dp83867-rxctrl-strap-quirk" property to the phy's device-tree node
to allow kernel to enable software workaround for this incorrect strap
setting. This is as suggested by the phy's datamanual and ensures proper
operation of this PHY.

This needs to be done for both instances of this PHY present on the board.

Signed-off-by: Sekhar Nori <nsekhar@ti.com>
---
 arch/arm/boot/dts/dra72-evm-revc.dts | 2 ++
 1 file changed, 2 insertions(+)

diff --git a/arch/arm/boot/dts/dra72-evm-revc.dts b/arch/arm/boot/dts/dra72-evm-revc.dts
index 5532309971ae..9b439851ff79 100644
--- a/arch/arm/boot/dts/dra72-evm-revc.dts
+++ b/arch/arm/boot/dts/dra72-evm-revc.dts
@@ -80,6 +80,7 @@
 		ti,min-output-impedance;
 		interrupt-parent = <&gpio6>;
 		interrupts = <16 IRQ_TYPE_EDGE_FALLING>;
+		ti,dp83867-rxctrl-strap-quirk;
 	};
 
 	dp83867_1: ethernet-phy at 3 {
@@ -90,6 +91,7 @@
 		ti,min-output-impedance;
 		interrupt-parent = <&gpio6>;
 		interrupts = <16 IRQ_TYPE_EDGE_FALLING>;
+		ti,dp83867-rxctrl-strap-quirk;
 	};
 };
 
-- 
2.9.0

[PATCH 3/3] ARM: omap2plus_defconfig: enable DP83867 phy driver

[Sekhar Nori]

TI's DP83867 phy is used on DRA72x EVM rev C and DRA71x
EVMs. Enable support for it in omap2plus_defconfig.

The driver is built into the kernel to help NFS booting.

Signed-off-by: Sekhar Nori <nsekhar@ti.com>
---
 arch/arm/configs/omap2plus_defconfig | 1 +
 1 file changed, 1 insertion(+)

diff --git a/arch/arm/configs/omap2plus_defconfig b/arch/arm/configs/omap2plus_defconfig
index a120ae816260..d4afa2ca2537 100644
--- a/arch/arm/configs/omap2plus_defconfig
+++ b/arch/arm/configs/omap2plus_defconfig
@@ -170,6 +170,7 @@ CONFIG_TI_CPTS=y
 # CONFIG_NET_VENDOR_WIZNET is not set
 CONFIG_AT803X_PHY=y
 CONFIG_DP83848_PHY=y
+CONFIG_DP83867_PHY=y
 CONFIG_MICREL_PHY=y
 CONFIG_SMSC_PHY=y
 CONFIG_PPP=m
-- 
2.9.0

[PATCH 3/3] ARM: omap2plus_defconfig: enable DP83867 phy driver

[Grygorii Strashko]

On 06/27/2017 10:06 AM, Sekhar Nori wrote:
> TI's DP83867 phy is used on DRA72x EVM rev C and DRA71x
> EVMs. Enable support for it in omap2plus_defconfig.
> 
> The driver is built into the kernel to help NFS booting.
> 
> Signed-off-by: Sekhar Nori <nsekhar@ti.com>

Reviewed-by: Grygorii Strashko <grygorii.strashko@ti.com>

this can go alone I think.
> ---
>   arch/arm/configs/omap2plus_defconfig | 1 +
>   1 file changed, 1 insertion(+)
> 
> diff --git a/arch/arm/configs/omap2plus_defconfig b/arch/arm/configs/omap2plus_defconfig
> index a120ae816260..d4afa2ca2537 100644
> --- a/arch/arm/configs/omap2plus_defconfig
> +++ b/arch/arm/configs/omap2plus_defconfig
> @@ -170,6 +170,7 @@ CONFIG_TI_CPTS=y
>   # CONFIG_NET_VENDOR_WIZNET is not set
>   CONFIG_AT803X_PHY=y
>   CONFIG_DP83848_PHY=y
> +CONFIG_DP83867_PHY=y
>   CONFIG_MICREL_PHY=y
>   CONFIG_SMSC_PHY=y
>   CONFIG_PPP=m
> 

-- 
regards,
-grygorii

[PATCH 0/3] ARM: dts: omap: workarounds for Ethernet PHY strap on DRA71 and DRA72 EVMs

[Sekhar Nori]

Hi Tony,

On Tuesday 27 June 2017 08:36 PM, Sekhar Nori wrote:
> Hi Tony,
> 
> Here are some patches to overcome incorrect Ethernet PHY strap settings
> on DRA72x EVM Rev C and DRA71 EVM. Both of these boards use TI's DP83867
> PHY.
> 
> The incorrect PHY strap settings cause a voilation of PHY specification
> and need to be worked around in software as documented in the commit
> messages. But there is no known issue that these patches fix.
> 
> I did verify that there are no regressions with NFS boot on these boards
> after the patches are applied on v4.12-rc4.

Please ignore this series. It is prematurely posted. The DT property I
use is not upstream yet. I relied on some recollection of it being
upstream. I should have grepped instead. Since there was no regression
and no change in behavior could also be explained, I did not catch it in
my testing. Thanks Grygorii for pointing out.

Regards,
Sekhar