PRU phandle is changing

[kaigeis...@googlemail.com]

I'm writing a kernel driver that needs to control the PRUs' state using rproc_boot(...)/rproc_shutdown(...) from linux/remoteproc.h. The functions take a `struct rproc *` as argument. I tried using `rproc_get_by_phandle(...)` to get the reference to the PRUs. The problem is that the phandles are actually different from one installation to the other. E.g. after flashing a fresh image, the PRU phandles could be 0xF4 and 0xF5. The next time, after flashing the image, the phandles are 0xF6 and 0xF7. They seem, however, to remain the same, if I don't flash a new image.

- Why are the phandles different between one image and the other?

- Is there a way to fix them, such that I reliable get pointers to the PRUs with `rproc_get_by_phandle()`?

- Is there another, better way to reliably get `struct rproc *`s to the PRUs?

Thanks,

Kai

[Igor Jędrzejczak]

Hi! I'm facing the same issue... Have you found a way to get pruss running from kernel module?

Best regards

Igor

[Dimitar Dimitrov]

Which kernel version are you using? Can you post your DTS changes?

Yes, it is expected phandle to have different value for different DTBs. See https://elinux.org/Device_Tree_Mysteries#Phandle 

Personally I found it much easier to use rpmsg, which sits on top of remoteproc. The mainline example is short and easy to follow: https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/tree/samples/rpmsg/rpmsg_client_sample.c 

Yes, RPMSG it's a bit heavy, but you do get a nice RPC. Also, any necessary DTS changes are part of recent beagleboard debian images. I've published my simple driver: https://github.com/dinuxbg/beaglemic/blob/master/driver/beaglemic.c

Regards,

Dimitar