Re: [PATCH] phy: berlin-usb: adjust USB_PHY_RX_CTRL init flags

[Antoine Tenart]

Hi Alexander,

I've added Jisheng in Cc so he can have a look at this.

Thanks!
Antoine

On Sun, Mar 04, 2018 at 09:18:07PM +0300, Alexander Monakov wrote:
> Make the value written into the USB_PHY_RX_CTRL configuration register
> match 0xAA79 value written by manufacturer-supplied kernels for Sony
> NSZ-GS7 (Berlin2 SoC), Google Chromecast and Valve Steam Link (BG2CD).
> 
> This fixes timeouts communicating to the internal hub on Steam Link.
> 
> Cc: Kishon Vijay Abraham I <kishon@xxxxxx>
> Cc: Antoine Tenart <antoine.tenart@xxxxxxxxxxx> 
> Signed-off-by: Alexander Monakov <amonakov@xxxxxxxxx>
> ---
> Hello,
> 
> I'm trying to run mainline kernel on "Steam Link" device built around
> Marvell BG2CD SoC, and in doing so I've found that USB almost always
> fails to initialize with timeout errors when connecting to the hub:
> 
> [    0.223563] usb usb1: New USB device found, idVendor=1d6b, idProduct=0002
> [    0.223580] usb usb1: New USB device strings: Mfr=3, Product=2, SerialNumber=1
> [    0.223596] usb usb1: Product: EHCI Host Controller
> [    0.223608] usb usb1: Manufacturer: Linux 4.15.0-00014-gc6f881cd62dc-dirty ehci_hcd
> [    0.223624] usb usb1: SerialNumber: ci_hdrc.0
> [    0.223985] hub 1-0:1.0: USB hub found
> [    0.224022] hub 1-0:1.0: 1 port detected
> [    0.524726] ci_hdrc ci_hdrc.0: port 1 reset error -110
> [    1.044673] ci_hdrc ci_hdrc.0: port 1 reset error -110
> [    1.399721] mwifiex_sdio mmc0:0001:1: info: FW download over, size 814048 bytes
> [    1.484666] ci_hdrc ci_hdrc.0: port 1 reset error -110
> [    2.113451] mwifiex_sdio mmc0:0001:1: WLAN FW is active
> [    2.113725] sdio platform data not available
> [    2.192570] mwifiex_sdio mmc0:0001:1: info: MWIFIEX VERSION: mwifiex 1.0 (15.68.7.p112) 
> [    2.192599] mwifiex_sdio mmc0:0001:1: driver_version = mwifiex 1.0 (15.68.7.p112) 
> [    2.213433] usb 1-1: new high-speed USB device number 2 using ci_hdrc
> [    2.294676] ci_hdrc ci_hdrc.0: port 1 reset error -110
> 
> and so on until...
> 
> [   19.254658] ci_hdrc ci_hdrc.0: port 1 reset error -110
> [   19.913433] usb usb1-port1: Cannot enable. Maybe the USB cable is bad?
> [   19.913472] usb usb1-port1: unable to enumerate USB device
> 
> I've inspected how the original kernel performs phy initialization, and
> found that the value written into the USB_PHY_RX_CTRL register differs
> in one field. Correcting the field is enough to stabilize it in my case.
> 
> I don't know where the _260 choice in mainline source comes from. I've
> checked three publicly available kernels for devices with this phy, and
> they all write the same value, 0xAA79.
> 
> FWIW, this is not the only difference I've found. Other kernels also:
> - do not touch TX config registers;
> - drive a particular GPIO low before and high after configuring the phy.
> 
> I am unsure if anything needs to or can be done about those differences.
> 
> Thanks.
> Alexander
> 
>  drivers/phy/marvell/phy-berlin-usb.c | 2 +-
>  1 file changed, 1 insertion(+), 1 deletion(-)
> 
> diff --git a/drivers/phy/marvell/phy-berlin-usb.c b/drivers/phy/marvell/phy-berlin-usb.c
> index 2017751ede26..8f2b5cae360f 100644
> --- a/drivers/phy/marvell/phy-berlin-usb.c
> +++ b/drivers/phy/marvell/phy-berlin-usb.c
> @@ -127,7 +127,7 @@ static int phy_berlin_usb_power_on(struct phy *phy)
>  	writel(V2I_VCO_RATIO(0x5) | R_ROTATE_0 | ANA_TEST_DC_CTRL(0x5),
>  	       priv->base + USB_PHY_ANALOG);
>  	writel(PHASE_FREEZE_DLY_4_CL | ACK_LENGTH_16_CL | SQ_LENGTH_12 |
> -	       DISCON_THRESHOLD_260 | SQ_THRESHOLD(0xa) | LPF_COEF(0x2) |
> +	       DISCON_THRESHOLD_270 | SQ_THRESHOLD(0xa) | LPF_COEF(0x2) |
>  	       INTPL_CUR_30, priv->base + USB_PHY_RX_CTRL);
>  
>  	writel(TX_VDD12_13 | TX_OUT_AMP(0x3), priv->base + USB_PHY_TX_CTRL1);
> -- 
> 2.11.0
> 

-- 
Antoine Ténart, Bootlin (formerly Free Electrons)
Embedded Linux and Kernel engineering
https://bootlin.com