TI_NN_NodeEnumInit

[parisseb]

The declaration of TI_NN_NodeEnumInit does not use the same argument type for the calc and for the device: for the calc it's nn_ch_t, for the PC it's nn_oh_t, for more details cf.
https://tiplanet.org/forum/viewtopic.php?f=20&t=23919&view=unread#p253664

[Vogtinator]

Were you able to verify that it works as expected after changing to nn_oh_t?

[parisseb]

I tried to force with
int ans=TI_NN_NodeEnumInit((nn_ch_t) oh);
then it compiles, but I do not have a working example of code to try, and I abandonned because I could not get my code slice working on the nspire. I do not need USB anymore, I'm going to read port 0x900b0028 to check if the calc is connected to a PC (this is the real reason I tried USB).

[Vogtinator]

I know that calc <-> calc communication works currently (there are some threads on tiplanet) and calc <-> PC communication was also verified on earlier versions of the SDK: https://github.com/ndless-nspire/nsptools-history/blob/master/Ndless/trunk/arm/tests/navnet/ndless_navnet_tests.c

Maybe you didn't enable interrupts necessary for USB traffic?

I'm going to read port 0x900b0028 to check if the calc is connected to a PC (this is the real reason I tried USB).

That's not documented on hackspire, how did you discover this?

I can already tell that it won't work on CX II as it uses a different PMU and 0x900b00XX does not exist anymore.

[parisseb]

I made tests inside the KhiCAS shell, read32(0x9000b0028) returns 0x11110114 if the calc is plugged and 0x11070114 if not. I discovered that by checking the differences of all values in the power management port range. I guess the CX ii has a similar port range.