Windows 10 does not recognise ethernet over USB for STM32MP157C-DK2

[CRong]

I connected the board to the OTG port. In the DFU mode, the USB port can be recognised correctly. When switching to the normal mode, Windows 10 cannot recognise the device. I tried to force to install the device as "Remote NDIS Compatible Device". Windows shows device status as "This device cannot start (Code 10)".

I have stuck on this issue for more than 2 weeks. Please help!

[Olivier GALLIEN]

Hi @CRong​ 

Could you please confirm you are following the procedure to setup ethernet over USB present in the help of SW4STM32 ?

STM32 MPU Family - MCU Support -> Tips and Tricks -> Ethernet-over-USB Windows 10 Set-up

Thx

Olivier

[CRong]

Hi Oliver @Community member​ 

Thanks for your reply.

Yes, I did follow that procedure. The USB0 interface is up. However, in device manager, the OTG usb device is not recognised. It shows as "Unknown device". When using "Let me pick from a list of device drivers on my computer", it does not show "Remote NDIS compatible device" as in the procedure.

Thanks

Chris

[robin_elecgator]

I had the same problem on my new PC. But connecting the STM32MP via a USB2.0 HUB makes it work. I think it has something to do with OTG driver and new motherboards which supports usb C.

[CRong]

Thanks, Robin

Just ordered a USB hub. Will give a try ....

[CRong]

Hi, @robin_elecgator​ 

It works now with USB hub ... Thanks a lot.

[robin_elecgator]

Good, I'm just wondering on which PC didn't it work? I have it tested only on Dell computers with a 8th gen intel processor.

I had it on both the devkit and a custom board that we designed.

Which type of PC do you use?

I currently have the following 3 options:

A) There is a problem with the drivers of intel

B) There is a problem with the USB gadget driver in the kernel

C) There is a problem in the hw usb driver

But I don't have really the time to pin point it.

[Olivier GALLIEN]

Hi @robin_elecgator​  and @CRong​ 

We are interesting to pin point the root cause of this.

Could you please both confirm that on target side during boot you have following traces :

[ 442.850073] dwc2 49000000.usb-otg: new device is high-speed

[ 442.885715] dwc2 49000000.usb-otg: new address 8

[ 442.934950] configfs-gadget gadget: high-speed config #1: c

[ 442.939395] IPv6: ADDRCONF(NETDEV_CHANGE): usb0: link 

Then can you please share as many details as possible on your PC and USB controller :

model, USB2 and USB3 controller name and type etc...

Thanks a lot,

Olivier

[Olivier GALLIEN]

Fact that USB Hub is a workaround can point a problem of power compatibility with PC USB controller.

Could you please try following change for a test :

On target :

root@stm32mp1:~# /sbin/stm32_usbotg_eth_config.sh stop # stop the USB gadget

root@stm32mp1:~# vi /sbin/stm32_usbotg_eth_config.sh #edit the script

replace echo 250 > "${d}/configs/${c}/MaxPower"

with echo 100 > "${d}/configs/${c}/MaxPower"

Save and quit

root@stm32mp1:~# /sbin/stm32_usbotg_eth_config.sh start # restart the Gadget

Check if RNDIS is now up on PC side.

Thx

Olivier

[robin_elecgator]

Hi @Community member​ 

I connect my DK2 to my pcb with a Type C to Type C cable. I get the following:

[ 16.947132] dwc2 49000000.usb-otg: new device is high-speed
[ 16.981167] dwc2 49000000.usb-otg: new address 4
[ 17.027155] dwc2 49000000.usb-otg: Failed to exit L1 sleep state in 200us.
[ 17.032624] configfs-gadget gadget: high-speed config #1: c
[ 17.038284] dwc2 49000000.usb-otg: dwc2_hsotg_enqueue_setup: failed queue (-11)
[ 17.045822] dwc2 49000000.usb-otg: Failed to exit L1 sleep state in 200us.
[ 17.065150] IPv6: ADDRCONF(NETDEV_UP): usb0: link is not ready
[ 17.765371] IPv6: ADDRCONF(NETDEV_CHANGE): usb0: link becomes ready
[ 22.028030] dwc2 49000000.usb-otg: Failed to exit L1 sleep state in 200us.

The PC is a Dell Precision 3630, with a i7-8700k running on Intel C246 (Cannon lake chipset).

I also your edit. This gives me the following:

root@stm32mp1-disco:~# vi /sbin/stm32_usbotg_eth_config.sh stop
root@stm32mp1-disco:~# /sbin/stm32_usbotg_eth_config.sh start
[ 1002.974225] using random self ethernet address
[ 1002.977321] using random host ethernet address
[ 1003.018066] usb0: HOST MAC 6e:3e:60:48:19:09
[ 1003.022272] usb0: MAC ca:05:50:a6:2d:81
[ 1003.028168] dwc2 49000000.usb-otg: bound driver configfs-gadget
[ 1003.191413] dwc2 49000000.usb-otg: new device is high-speed
[ 1003.224592] dwc2 49000000.usb-otg: new address 8
[ 1003.252019] dwc2 49000000.usb-otg: Failed to exit L1 sleep state in 200us.
[ 1003.257496] configfs-gadget gadget: high-speed config #1: c
[ 1003.259257] IPv6: ADDRCONF(NETDEV_UP): usb0: link is not ready
[ 1003.263135] dwc2 49000000.usb-otg: dwc2_hsotg_enqueue_setup: failed queue (-11)
[ 1003.276536] dwc2 49000000.usb-otg: Failed to exit L1 sleep state in 200us.
[ 1003.283200] IPv6: ADDRCONF(NETDEV_CHANGE): usb0: link becomes ready
ifconfig: SIOCSIFADDR: Invalid argument
root@stm32mp1-disco:~# [ 1008.253305] dwc2 49000000.usb-otg: Failed to exit L1 sleep state in 200us.
[ 1008.259049] dwc2 49000000.usb-otg: Failed to exit L1 sleep state in 200us.

[Olivier GALLIEN]

Thanks @robin_elecgator​ 

Sorry but trace from the edit trial looks strange

root@stm32mp1-disco:~# vi /sbin/stm32_usbotg_eth_config.sh stop ?

Please confirm you properly edit the file with echo 100.

Thx

Olivier

[robin_elecgator]

Hi @Community member​ 

The stop is a small mistake (vi did want to create the "stop" file) It is the second file vi wants to edit, but the changes did apply to the script

This is in my config.sh file.

 # Config
 mkdir -p "${d}/configs/${c}"
 mkdir -p "${d}/configs/${c}/strings/0x409"
 echo "Config 1: RNDIS" > "${d}/configs/${c}/strings/0x409/configuration"
 echo 100 > "${d}/configs/${c}/MaxPower"
 echo 0xC0 > "${d}/configs/${c}/bmAttributes" # self powered device
 
 mkdir -p "${d}/functions/${func_eth}"
- /sbin/stm32_usbotg_eth_config.sh 43/109 39%

Robin

[robin_elecgator]

Hi @Community member​ 

Do you need more information? did you guys found the cause of the problem?

Best regards,

Robin

[AntonioST]

Still unable to replicate it.

To understand if we are looking in the right direction, would be possible for you to test this patch in the kernel of STM32MP1 ?

Of course, it would require you to recompile the kernel; no way to apply such modification at runtime in sysfs.

diff --git a/drivers/usb/gadget/composite.c b/drivers/usb/gadget/composite.c
index b8a1584..afbcbb7 100644
--- a/drivers/usb/gadget/composite.c
+++ b/drivers/usb/gadget/composite.c
@@ -629,7 +629,7 @@ static int bos_desc(struct usb_composite_dev *cdev)
 usb_ext->bLength = USB_DT_USB_EXT_CAP_SIZE;
 usb_ext->bDescriptorType = USB_DT_DEVICE_CAPABILITY;
 usb_ext->bDevCapabilityType = USB_CAP_TYPE_EXT;
- usb_ext->bmAttributes = cpu_to_le32(USB_LPM_SUPPORT | USB_BESL_SUPPORT);
+ usb_ext->bmAttributes = cpu_to_le32(USB_LPM_SUPPORT | USB_BESL_SUPPORT | 0x4118);
 
 /*
 * The Superspeed USB Capability descriptor shall be implemented by all

[Bernard PUEL]

Any update Robin ?

[NGajic]

Hi,

I had this problem recently on my laptop with 4 3.1 USB ports. Two of them worked and two doesn't. I had same message - dwc2 49000000.usb-otg: Failed to exit L1 sleep state in 200us. I solved problem by applying this patch https://www.spinics.net/lists/linux-usb/msg205440.html on ecosystem v2.0 kernel. Now I managed to make RNDIS working and USB mass storage also.

Best regards,

Nenad

[Dat Tran]

I have same issue, when usb hub then work, I could ping from PC.

But plugin to PC directly then usb0 shown up but can not ping. Under Device Manager PC side, unknow device shown as well.