Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 2620:137:e000::1:20 as permitted sender) client-ip=2620:137:e000::1:20;
Date:   Sun, 21 May 2023 15:35:12 +0300
From:   Vladimir Oltean <olteanv@gmail.com>
To:     David Epping <david.epping@missinglinkelectronics.com>
Cc:     Andrew Lunn <andrew@lunn.ch>,
        Heiner Kallweit <hkallweit1@gmail.com>,
        Russell King <linux@armlinux.org.uk>,
        "David S . Miller" <davem@davemloft.net>,
        Eric Dumazet <edumazet@google.com>,
        Jakub Kicinski <kuba@kernel.org>,
        Paolo Abeni <pabeni@redhat.com>, netdev@vger.kernel.org,
        linux-kernel@vger.kernel.org, UNGLinuxDriver@microchip.com
Subject: Re: [PATCH net 3/3] net: phy: mscc: enable VSC8501/2 RGMII RX clock
Message-ID: <20230521123512.3kpy66sjnzj2chie@skbuf>
References: <20230520160603.32458-1-david.epping@missinglinkelectronics.com>
 <20230520160603.32458-4-david.epping@missinglinkelectronics.com>
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
In-Reply-To: <20230520160603.32458-4-david.epping@missinglinkelectronics.com>
Precedence: bulk

On Sat, May 20, 2023 at 06:06:03PM +0200, David Epping wrote:
> By default the VSC8501 and VSC8502 RGMII RX clock output is disabled.
> To allow packet forwarding towards the MAC it needs to be enabled.
> The same may be necessary for GMII and MII modes, but that's currently
> unclear.
> 
> For VSC853x and VSC854x the respective disable bit is reserved and the
> clock output is enabled by default.
> 
> Signed-off-by: David Epping <david.epping@missinglinkelectronics.com>
> ---
>  drivers/net/phy/mscc/mscc.h      |  1 +
>  drivers/net/phy/mscc/mscc_main.c | 24 ++++++++++++++++++++++++
>  2 files changed, 25 insertions(+)
> 
> diff --git a/drivers/net/phy/mscc/mscc.h b/drivers/net/phy/mscc/mscc.h
> index 79cbb2418664..defe5cc6d4fc 100644
> --- a/drivers/net/phy/mscc/mscc.h
> +++ b/drivers/net/phy/mscc/mscc.h
> @@ -179,6 +179,7 @@ enum rgmii_clock_delay {
>  #define VSC8502_RGMII_CNTL		  20
>  #define VSC8502_RGMII_RX_DELAY_MASK	  0x0070
>  #define VSC8502_RGMII_TX_DELAY_MASK	  0x0007
> +#define VSC8502_RGMII_RX_CLK_DISABLE	  0x0800
>  
>  #define MSCC_PHY_WOL_LOWER_MAC_ADDR	  21
>  #define MSCC_PHY_WOL_MID_MAC_ADDR	  22
> diff --git a/drivers/net/phy/mscc/mscc_main.c b/drivers/net/phy/mscc/mscc_main.c
> index 29fc27a16805..c7a8f5561c66 100644
> --- a/drivers/net/phy/mscc/mscc_main.c
> +++ b/drivers/net/phy/mscc/mscc_main.c
> @@ -547,6 +547,26 @@ static int vsc85xx_rgmii_set_skews(struct phy_device *phydev, u32 rgmii_cntl,
>  	return rc;
>  }
>  
> +/* For VSC8501 and VSC8502 the RGMII RX clock output is disabled by default. */

This statement is not exactly true, proven by my board where I've just
printed these values:

[    6.454638] Microsemi GE VSC8502 SyncE 0000:00:00.3:03: vsc85xx_rgmii_enable_rx_clk: RGMII_CNTL 0x44, RX_CLK_DISABLE 0x0
[    6.544652] sja1105 spi2.2 sw2p0 (uninitialized): PHY [0000:00:00.3:03] driver [Microsemi GE VSC8502 SyncE] (irq=POLL)
[    6.630864] Microsemi GE VSC8502 SyncE 0000:00:00.3:02: vsc85xx_rgmii_enable_rx_clk: RGMII_CNTL 0x44, RX_CLK_DISABLE 0x0
[    6.720218] sja1105 spi2.2 sw2p1 (uninitialized): PHY [0000:00:00.3:02] driver [Microsemi GE VSC8502 SyncE] (irq=POLL)
[    6.806876] Microsemi GE VSC8502 SyncE 0000:00:00.3:11: vsc85xx_rgmii_enable_rx_clk: RGMII_CNTL 0x44, RX_CLK_DISABLE 0x0
[    6.896185] sja1105 spi2.2 sw2p2 (uninitialized): PHY [0000:00:00.3:11] driver [Microsemi GE VSC8502 SyncE] (irq=POLL)
[    6.982775] Microsemi GE VSC8502 SyncE 0000:00:00.3:10: vsc85xx_rgmii_enable_rx_clk: RGMII_CNTL 0x44, RX_CLK_DISABLE 0x0
[    7.071988] sja1105 spi2.2 sw2p3 (uninitialized): PHY [0000:00:00.3:10] driver [Microsemi GE VSC8502 SyncE] (irq=POLL)

Let's resolve that difference before the patches are merged, and write
some correct comments.

I agree that the datasheet is not clear, but I think that the RX_CLK
output is enabled or not based on the strapping of the RCVRDCLK1 and
RCVRDCLK2 pins. Coincidentally, these are also muxed with PHYADD1 and
PHYADD2, so the default value of RX_CLK_DISABLE might depend on the
PHY address (?!).

What is your PHY address? Mine are 0x10 and 0x11 for the VSC8502 on my
board.

Not saying that the patch is wrong or that the resolution should be any
different than it is. Just that it's clear we can't both be right, and
my PHYs clearly work (re-tested just now).

--
pw-bot: changes-requested