Hi,
I noticed that the usb ethernet gadget driver is broken since v5.5-rc2
and 5.4.4, but only for our OMAP5 board. Our OMAP3 boards work with
the same tree, kernel binary and user-space (Debian Stretch).
The symptom is that I can see the interface on the host PC being
enumerated and IP addresses etc. are set up. But a ping in either
direction fails/times out.
After inspecting diffs I found some changes in dwc3:gadget
and indeed omap3/twl4030 uses musb and omap5 uses dwc3.
Reverting
a7f7e61270f1 ("usb: dwc3: gadget: Fix logical condition");
on v5.5-rc makes it work again.
BTW: v4.19.90 works although it includes this logical condition fix.
So the real difference between v4.19 and v5.5 may be elsewhere in
newer kernels and only be revealed by the patch.
If important: my setup is running with USB2 cable and speed
only.
So please check this and other recent dwc3 patches for introducing
a stall of communication.
BR and thanks,
Nikolaus Schaller
Hi,
On 1/8/2020 2:29 PM, H. Nikolaus Schaller wrote:
> Hi,
> I noticed that the usb ethernet gadget driver is broken since v5.5-rc2
> and 5.4.4, but only for our OMAP5 board. Our OMAP3 boards work with
> the same tree, kernel binary and user-space (Debian Stretch).
>
> The symptom is that I can see the interface on the host PC being
> enumerated and IP addresses etc. are set up. But a ping in either
> direction fails/times out.
>
> After inspecting diffs I found some changes in dwc3:gadget
> and indeed omap3/twl4030 uses musb and omap5 uses dwc3.
>
> Reverting
>
> a7f7e61270f1 ("usb: dwc3: gadget: Fix logical condition");
>
> on v5.5-rc makes it work again.
>
> BTW: v4.19.90 works although it includes this logical condition fix.
> So the real difference between v4.19 and v5.5 may be elsewhere in
> newer kernels and only be revealed by the patch.
>
> If important: my setup is running with USB2 cable and speed
> only.
>
> So please check this and other recent dwc3 patches for introducing
> a stall of communication.
>
Another gadget driver issue was reported by Thinh after my fix, and he has submitted a patch for the same. You can refer the discussion https://patchwork.kernel.org/patch/11292087/.
Can you check if that patch works for you ?
> BR and thanks,
> Nikolaus Schaller
>
>
Thanks & Regards,
Tejas Joglekar
Hi,
> Am 08.01.2020 um 10:06 schrieb Tejas Joglekar <[email protected]>:
>
> Hi,
> On 1/8/2020 2:29 PM, H. Nikolaus Schaller wrote:
>> Hi,
>> I noticed that the usb ethernet gadget driver is broken since v5.5-rc2
>> and 5.4.4, but only for our OMAP5 board. Our OMAP3 boards work with
>> the same tree, kernel binary and user-space (Debian Stretch).
>>
>> The symptom is that I can see the interface on the host PC being
>> enumerated and IP addresses etc. are set up. But a ping in either
>> direction fails/times out.
>>
>> After inspecting diffs I found some changes in dwc3:gadget
>> and indeed omap3/twl4030 uses musb and omap5 uses dwc3.
>>
>> Reverting
>>
>> a7f7e61270f1 ("usb: dwc3: gadget: Fix logical condition");
>>
>> on v5.5-rc makes it work again.
>>
>> BTW: v4.19.90 works although it includes this logical condition fix.
>> So the real difference between v4.19 and v5.5 may be elsewhere in
>> newer kernels and only be revealed by the patch.
>>
>> If important: my setup is running with USB2 cable and speed
>> only.
>>
>> So please check this and other recent dwc3 patches for introducing
>> a stall of communication.
>>
> Another gadget driver issue was reported by Thinh after my fix, and he has submitted a patch for the same. You can refer the discussion https://patchwork.kernel.org/patch/11292087/.
> Can you check if that patch works for you ?
Yes, it does!
So you can add my
Tested-By: H. Nikolaus Schaller <[email protected]>
if possible (I can't reply to the original discussion).
Would be nice if you can get it into 5.5-rc series and 5.4.y asap.
BR and thanks for the quick response and solution,
Nikolaus Schaller