netdev - Re: [PATCH/RFC 1/5] dt-bindings: net: renesas,ravb: Document internal clock delay properties

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]

Date:   Sat, 20 Jun 2020 07:47:16 +0200
From:   Oleksij Rempel <linux@...pel-privat.de>
To:     Geert Uytterhoeven <geert+renesas@...der.be>,
        Sergei Shtylyov <sergei.shtylyov@...entembedded.com>,
        "David S . Miller" <davem@...emloft.net>,
        Jakub Kicinski <kuba@...nel.org>,
        Rob Herring <robh+dt@...nel.org>
Cc:     Andrew Lunn <andrew@...n.ch>,
        Philippe Schenker <philippe.schenker@...adex.com>,
        Florian Fainelli <f.fainelli@...il.com>,
        Heiner Kallweit <hkallweit1@...il.com>,
        Kazuya Mizuguchi <kazuya.mizuguchi.ks@...esas.com>,
        Wolfram Sang <wsa+renesas@...g-engineering.com>,
        netdev@...r.kernel.org, devicetree@...r.kernel.org,
        linux-renesas-soc@...r.kernel.org
Subject: Re: [PATCH/RFC 1/5] dt-bindings: net: renesas,ravb: Document internal
 clock delay properties

Hi Geert,

Am 19.06.20 um 21:15 schrieb Geert Uytterhoeven:
> Some EtherAVB variants support internal clock delay configuration, which
> can add larger delays than the delays that are typically supported by
> the PHY (using an "rgmii-*id" PHY mode, and/or "[rt]xc-skew-ps"
> properties).
>
> Add properties for configuring the internal MAC delays.
> These properties are mandatory, even when specified as zero, to
> distinguish between old and new DTBs.
>
> Update the example accordingly.
>
> Signed-off-by: Geert Uytterhoeven <geert+renesas@...der.be>
> ---
>  .../devicetree/bindings/net/renesas,ravb.txt  | 29 ++++++++++---------
>  1 file changed, 16 insertions(+), 13 deletions(-)
>
> diff --git a/Documentation/devicetree/bindings/net/renesas,ravb.txt b/Documentation/devicetree/bindings/net/renesas,ravb.txt
> index 032b76f14f4fdb38..488ada78b6169b8e 100644
> --- a/Documentation/devicetree/bindings/net/renesas,ravb.txt
> +++ b/Documentation/devicetree/bindings/net/renesas,ravb.txt
> @@ -64,6 +64,18 @@ Optional properties:
>  			 AVB_LINK signal.
>  - renesas,ether-link-active-low: boolean, specify when the AVB_LINK signal is
>  				 active-low instead of normal active-high.
> +- renesas,rxc-delay-ps: Internal RX clock delay.

may be it make sense to add a generic delay property for MACs and PHYs?

> +			This property is mandatory and valid only on R-Car Gen3
> +			and RZ/G2 SoCs.
> +			Valid values are 0 and 1800.
> +			A non-zero value is allowed only if phy-mode = "rgmii".
> +			Zero is not supported on R-Car D3.
> +- renesas,txc-delay-ps: Internal TX clock delay.
> +			This property is mandatory and valid only on R-Car H3,
> +			M3-W, M3-W+, M3-N, V3M, and V3H, and RZ/G2M and RZ/G2N.
> +			Valid values are 0 and 2000.

In the driver i didn't found sanity check for valid values.

> +			A non-zero value is allowed only if phy-mode = "rgmii".
> +			Zero is not supported on R-Car V3H.>  Example:
>
> @@ -105,8 +117,10 @@ Example:
>  				  "ch24";
>  		clocks = <&cpg CPG_MOD 812>;
>  		power-domains = <&cpg>;
> -		phy-mode = "rgmii-id";
> +		phy-mode = "rgmii";
>  		phy-handle = <&phy0>;
> +		renesas,rxc-delay-ps = <0>;
> +		renesas,txc-delay-ps = <2000>;
>
>  		pinctrl-0 = <&ether_pins>;
>  		pinctrl-names = "default";
> @@ -115,18 +129,7 @@ Example:
>  		#size-cells = <0>;
>
>  		phy0: ethernet-phy@0 {
> -			rxc-skew-ps = <900>;
> -			rxdv-skew-ps = <0>;
> -			rxd0-skew-ps = <0>;
> -			rxd1-skew-ps = <0>;
> -			rxd2-skew-ps = <0>;
> -			rxd3-skew-ps = <0>;
> -			txc-skew-ps = <900>;
> -			txen-skew-ps = <0>;
> -			txd0-skew-ps = <0>;
> -			txd1-skew-ps = <0>;
> -			txd2-skew-ps = <0>;
> -			txd3-skew-ps = <0>;
> +			rxc-skew-ps = <1500>;


I'm curios, how this numbers ware taken?
Old configurations was:
TX delay:
(txd*-skew-ps = 0) == -420ns
(txc-skew-ps = 900) == 0ns
resulting delays 0.420ns

RX delay:
(rxd*-skew-ps = 0) == -420ns
(rxc-skew-ps = 900) == 0ns
internal delay 1.2 + 420ns will be 1.62ns

I was not able to find actual delay values provided by the MAC. But from your patches it looks like 2ns.

So, end result will be:
TXID = 2.4ns
RXID = 3.62ns

Both values seems to be a bit out of spec. New values are:
TXID = PHY 0ns + MAC 2.0ns
RXID =
(rxd*-skew-ps = no change) == 0ns
(rxc-skew-ps = 1500) == +600ns
internal delay 1.2 + 600ns = 1.8ns

From the PHY point of view, it is RGMII_RXID mode. Are you sure, RGMII_ID is not working for you?
Till now the numbers was not looking as a fine tuning.
I assume, it is better to use proper phy-mode instead of skew-ps values. I feel like no one had time
to understand real configured values in this PHY.

>  			reg = <0>;
>  			interrupt-parent = <&gpio2>;
>  			interrupts = <11 IRQ_TYPE_LEVEL_LOW>;
>


--
Regards,
Oleksij