Hi,
Adding Rob and HNS to Cc too for commit cf081d009c44 ("usb: musb: Set the
DT node on the child device").
* Sicelo <absicsz@xxxxxxxxx> [221010 09:00]:
> I have tried to bisect, but due to inexperience and lack of proper
> tooling (e.g. no serial console), I have not been able to pinpoint the
> cause of the error so far. However, on linux 6.0 with cf081d009c447647
> reverted, there is a more detailed trace in dmesg - the same trace
> appeared in other kernel versions while bisecting:
To me it seems that we now somehow have a probe issue for musb depending
on how it gets probed depending on the following line:
device_set_of_node_from_dev(&musb->dev, &pdev->dev);
I think commit 239071064732 ("partially Revert "usb: musb: Set the DT node
on the child device"") fixed the issue for omap3 that still uses hwmod
by removing the device_set_of_node_from_dev(). However, it somehow broke
SoCs using ti-sysc like omap4.
My guess is commit 239071064732 will fix the issue for you?
I'm not sure what the proper fix here might be to have both working
though :)
FYI, I have a pending patch set to convert omap3 also to probe with
ti-sysc for musb, but seems like there's some other issue here that
should be fixed first.
Regards,
Tony
