On Tue, Dec 03, 2024 at 04:29:02PM +0100, Enric Balletbo i Serra wrote:

Hi,

On Thu, Nov 28, 2024 at 10:58 AM s-vadapalli [off-list ref] wrote:

quoted

On Thu, Nov 28, 2024 at 10:47:42AM +0100, Enric Balletbo i Serra wrote:

quoted

Hi,

[...]

quoted

So I changed the dr_mode to otg instead of host and tried to configure
a usb mass storage gadget but unfortunately didn't work, but this
could be a driver problem, I got the following error

  UDC core: g1: couldn't find an available UDC

As the devicetree should describe the hardware, and as far as I can
see it should support the type-c port act as a gadget, I'm fine with
changing the dr_mode, unless anyone have more information about this,
the thing that makes me think a bit more is that, in the TI kernel
this is set to host, so I'm wondering if I'm missing something or is
just that was never tested.

Are all interfaces (Type-A and Type-C) functional as Host when the
dr_mode is set to "otg"? (Do USB devices connected to the interfaces
enumerate on AM69-SK?) If yes, then it could be a DIP Switch setting
that is related to OTG mode of operation or a USB-C Mux that needs to be
configured.

Yes, all interfaces (Type-A and Type-C) are functional as host when
the dr_mode is set to "otg". Looking at SK-AM69 Processor Start Kit
User's Guide (Rev. A) [1] I don't see any DIP Switch related to the
OTG  mode. Looks like the type-c connector connects directly to the
SoC through a USB HUB (TUSB4041).

[1] https://www.ti.com/lit/ug/spruj70a/spruj70a.pdf?ts=1733215039014

In that case, please post the patch setting "dr_mode" to "otg". As to
why "peripheral"/"gadget" operation doesn't work could be addressed
separately and seems to be independent of the device-tree. The reason
for suggesting "otg" instead of "host" is to ensure that the device-tree
describes the Hardware. Since "host" mode is still functional with
"dr_mode" set to "otg", at-least for the use-cases that you have in mind,
I believe that "otg" should be acceptable.

Regards,
Siddharth.