It seems like this is not the case.
One of the USB sockets on the Ixora carrier board does not work with the V1.2 device tree (X7) and without doing any electrical testing, it would appear that because the USBH_EN pin (84) is now an I2C_SDA pin then it seems not to be powering the USB port.
No, that is actually as designed as none of them Apalis TK1 modules, regardless of their version, ever had any functional USBH3. As a matter of fact, USBH3 is routed to the on-module companion K20 MCU, so one could implement custom USB functionality e.g. in the FreeRTOS running on there.
Please note that while on Ixora V1.0A USBH2 was routed to X7 this got changed for Ixora V1.1A as outlined in its revision history:
What I have done so far to get things working:
Take the Apalis TK1 LXDE 2.8.6 Image
Use the scripts from within the apalis-tk1_bin folder from a TK1 2.7.3 image but modify them for ext3 file system, zImage (and other bits I can’t remember at the moment)
Nowadays, one should really use ext4 in every and all cases.
Include the tegra124-apalis-v1.2-eval.dtb device-tree (it is definitely being used)
Write to microSD card and upload to the TK1
Rather than the legacy update procedure, we really strongly recommend using the Toradex Easy Installer, which, BTW, does even come pre-installed with any and all V1.2A module.
Module boots through to Angstrom and everything works except USB (looks like the EN signals aren’t quite correct).
I really don;'t think so.
I fully admit that this is a complete hash and that there will be issues with what I’m doing but I am under significant pressure to get this up and running to meet deadlines. What with having no base-line to work with (images onwards from 2.7b3 just DO NOT work - tried on multiple TK1 modules, all fail to boot) then I’ve had to try everything which my limited Linux knowledge will allow.
The only thing I need to figure out is why the USB host port is not working on our hardware and its probably V1.2 device-tree related. The I2C on the DDC port is working fine so something tells me that parts of the V1.2 device-tree are working.
No, sorry, but this is really as per design. Nothing wrong really.