Re: [BUG REPORT] usb: dwc3: Failed to link at superspeed in device mode

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

 



Hi Felipe

I received a Rock960 board. It has a rk3399 rockchip armv8 processor
with 2 DWC3 controllers. And guess what: it works perfectly fine. No
trickery to get super speed connection in device mode.

So there is something wrong with the up-board, not with the DWC3 driver.

Thanks for you support
Le mer. 10 oct. 2018 à 15:01, Felipe Balbi
<felipe.balbi@xxxxxxxxxxxxxxx> a écrit :
>
>
> Hi,
>
> Thomas JOURDAN <thomas.jourdan@xxxxxxxxx> writes:
>
> >> Ok, so HW itself negotiated HS instead of SS. Note that below ep0 max
> >> packet size is set ot 64.
> >>
> >>      irq/13-dwc3-894   [001] d...   250.891132: dwc3_event: event (00000101): Reset [U0]
> >>      irq/13-dwc3-894   [001] d...   250.941068: dwc3_event: event (00000201): Connection Done [U0]
> >>      irq/13-dwc3-894   [001] d...   250.941090: dwc3_gadget_ep_cmd: ep0out: cmd 'Set Endpoint Configuration' [1025] params 80000200 00000500 00000000 --> status: Successful
> >>      irq/13-dwc3-894   [001] d...   250.941095: dwc3_gadget_ep_enable: ep0out: mps 64/512 streams 0 burst 1 ring 0/0 flags E:swBp:e:>
> >>      irq/13-dwc3-894   [001] d...   250.941103: dwc3_gadget_ep_cmd: ep0in: cmd 'Set Endpoint Configuration' [1025] params 80000200 02000500 00000000 --> status: Successful
> >>      irq/13-dwc3-894   [001] d...   250.941106: dwc3_gadget_ep_enable: ep0in: mps 64/512 streams 0 burst 1 ring 0/0 flags E:swbp:e:<
> >
> > I am new to USB protocol. Can this packet size be an issue ?
>
> no, no. The device definitely negotiated HS :-)
>
> >> Would you have a usb sniffer to see if the link even attempts to run Rx
> >> termination detection?
> >
> > Unfortunately no but I can do eye diagram. Do you know if there is a
> > test mode I can activate for USB3.0 (like tests J/K in USB2.0) ?
>
> If you have a scope with USB3 test tools, device enters compliance mode
> when the link walks a specific pattern of link states. That's encoded on
> some scopes sporting the USB3 certification tools.
>
> >> Another thing to try is setting both of the quirks below:
> >>
> >>         snps,dis_u3_susphy_quirk
> >>         snps,dis_u2_susphy_quirk
> >
> > I attached traces with those quirks enabled. Unfortunately it didn't
> > improved the situation.
>
> okay
>
> > Do you think the issue can be software related ? IMHO it is more
> > hardware related, probably something wrong or borderline on the UP2
> > board. I will investigate the hardware side, but the schematics of the
> > board aren't opensourced, so I might take some time.
>
> right, it's unlikely to be SW. Based on registers, device is configured
> to support SuperSpeed.
>
> > I also ordered a Rock960 board. This is an ARMv8 platforrm with the
> > same DWC3 controller. I hope this one works and if it does it will be
> > a good reference for further investigation. Do you know a SBC board,
> > with the DWC3 controller, and a working USB3.0 OTG port ?
>
> TI's beagle x15 is one of them. There's also some of the 96board
> devices... I don't know of upstream support for them though.
>
> --
> balbi




[Index of Archives]     [Linux Media]     [Linux Input]     [Linux Audio Users]     [Yosemite News]     [Linux Kernel]     [Linux SCSI]     [Old Linux USB Devel Archive]

  Powered by Linux