USB bind error about libnx

[Gillou68310]

Same issue here, usbds initialization fails at _usbDsBindDevice while testing usbds example.
I first thought this could be caused by usb sysmodule modifications on newer sysver (I'm on 5.0.2), but since it also happens to @RKX1209 on 3.0.0 there' s probably something else going on here!

[yellows8]

So you try to run the original nxdbg sysmodule and it throws a fatal-error?(Completely unrelated to client when it's during init)

What sysver?

[RKX1209]

Yes.
My client-side env.

$ lsb_release -a
No LSB modules are available.
Distributor ID: Ubuntu
Description:    Ubuntu 17.10
Release:        17.10
Codename:       artful

And stub app runs on Nintendo switch 3.0.0.

I also tried to run the original nxdbg client but, it cause some errors.
Here is the log.

$ sudo python MainWindow.py --pid 100
Traceback (most recent call last):                                              
  File "MainWindow.py", line 198, in <module>                                   
    sys.exit(main(sys.argv))                                                    
  File "MainWindow.py", line 161, in main                                       
    usb = UsbConnection()                                                       
  File "nxdbg/debugger/UsbConnection.py", line 180, in
 __init__                                                                       
    self.dev.set_configuration()                                                
  File "/usr/local/lib/python2.7/dist-packages/usb/core.py", line 869, in set_co
nfiguration                                                                     
    self._ctx.managed_set_configuration(self, configuration)                    
  File "/usr/local/lib/python2.7/dist-packages/usb/core.py", line 102, in wrappe
r                                                                               
    return f(self, *args, **kwargs)                                             
  File "/usr/local/lib/python2.7/dist-packages/usb/core.py", line 148, in managed_set_configuration 
    self.backend.set_configuration(self.handle, cfg.bConfigurationValue)          File "/usr/local/lib/python2.7/dist-packages/usb/backend/libusb1.py", line 79$, in set_configuration                  
    _check(self.lib.libusb_set_configuration(dev_handle.handle, config_value))    File "/usr/local/lib/python2.7/dist-packages/usb/backend/libusb1.py", line 59$, in _check         
    raise USBError(_strerror(ret), ret, _libusb_errno[ret])                     
usb.core.USBError: [Errno None] Other error

And original stub app failed too, which is same situation as my case.
(Same error code 0xf601 at the point)

I'd like to confirm, does usb communication capability requires kernelhax?
If so, my switch has no kernel access yet, and it can cause this problem.

But I guess it can be possible to use usb:ds service from hb apps without kernel privilege.

[yellows8]

Client should be irrelevant - 0xF601 happens without any client running right?

[RKX1209]

Yes. It happens regardless of client status, running or not.
So it can be stub side problem.

[yellows8]

usbds doesn't need kernelhax. How are you running HBL?

[RKX1209]

Along with this instruction
So backend may be pegaswitch.(not fusee launcher)

[RKX1209]

Hi. Any progress?

[yellows8]

Dunno what's wrong.

[sphaerophoria]

My switch also crashes using the usbds example inside the usbDsInitialize call somewhere. Launching on firmware 5.1.0 through SXOS 1.2 (error code 2001-0123 which should be the same as noted 0xf601 in parent)

[sphaerophoria]

I dug into my error a little further yesterday. It looks like it's actually coming from _usbDsSetVidPidBcd. I can dodge this call with a NULL deviceinfo passed into the function, but I then get it a little further up on _usbDsGetSession.

I tried ignoring all 0xf601's due to the comment in http://switchbrew.org/index.php?title=SVC#svcReplyAndReceive . Although it's not usb related I thought they may mean that another session has closed. I think it's unlikely this is the case.

I also started digging around the ptm service. Thought process was that maybe someone is holding the usb port in a certain mode for charging so we can't change it. On second thought it's probably more likely to be the HID module since I'm seeing it show up on my linux box as an hid device

└> dmesg | grep hid
[33734.203715] hid-generic 0003:057E:2000.003B: hiddev0,hidraw0: USB HID v1.11 Device [Nintendo Nintendo Switch] on usb-0000:00:1d.0-1.2/input0

[CordY5]

Any plans to fix this issue ?
Need this error to be fixed if I want to finish my app.

[misson20000]

On my 3.0.0 console, USB initialization blocks somewhere pretty early until you unplug the USB cable and plug it back in again. Try that?

[CordY5]

@misson2000 doesn't work

[JustForNothing]

It seems the USB sysmodule is completely different in 5.X
But for the debugger I found a workaround: just use TCP instead.

[yellows8]

On <5.0 with the usbds example, does it happen with both the original example and with the second param here changed to NULL? https://github.com/switchbrew/switch-examples/blob/master/usb/usbds/source/main.c#L129

[Monotek18]

@yellows8 on 3.0.0, it doesn't happen with the original example + with the second param here changed to NULL. But when you change the second param to NULL, I get 2140-0305 whenever you unplug the USB while running the homebrew.

[yellows8]

@Monotek18 Were you reading from the serial device host-side?

EDIT: Sounds like usbds is fine as long as deviceinfo is set.

[yellows8]

Can someone on 5.x also try this? #109 (comment)

[Monotek18]

@yellows8 On my 5.1.0, I still have 2001-0123 even with the 2nd param changed to NULL & with the original example.

[WinterMute]

On 5.1.0 I get the same 2001-0123 error.

On 3.0.0 the nxdbg stub app inits fine and the usbds example works fine. With the 2nd parameter set to NULL the Switch reads VID/PID as 057e/3000

[yellows8]

@RKX1209 Since it works fine for everyone else on pre-5.0.0, can you try again with RCM + latest libnx/nx-hbloader?

(5.0.0+ support can't be fixed until the new commands are documented)

[RKX1209]

On my 5.1.0, same error 2001-0123 happen even if with latest hbloader, fusee launcher;)

[SciresM]

Should be closed by PR #181

[RKX1209]

Thank you all you guys!