Re: [beagleboard] Re: PRU0 base address

2014-11-11 Thread Karl Karpfen
OK, solved, using this initialisation PRU is at least initialised: HWREG(SOC_PRM_PER_REGS)|=0x0002; HWREG(SOC_PRM_PER_REGS)|=0xFFFD; HWREG(SOC_CM_PER_REGS+CM_PER_PRU_ICSS_CLKCTRL)=0x0002;

[beagleboard] Re: PRU0 base address

2014-11-10 Thread Karl Karpfen
OK, seems I have to clarify this a bit: in my environment no Linux is involved, I'm digging within the hardware directly. What I found meanwhile: some clocks have to be enabled. But setting CM_PER_PRU_ICSS_CLKCTRL to 0x0002 and CM_PER_PRU_ICSS_CLKSTCTRL to 0x0010

[beagleboard] Re: PRU0 base address

2014-11-06 Thread ericggregory
The PRU does need to be enabled, typically by a Device Tree entry similar to this: fragment@4 { target = pruss; __overlay__ { status = okay; }; }; If using a cape of some kind, this would normally be in the device tree setup for that cape. If it's your own

[beagleboard] Re: PRU0 base address

2014-11-06 Thread josmfernandez
Hey man, You should read this PDF file. It helps a lot. http://mythopoeic.org/BBB-PRU/am335xPruReferenceGuide.pdf From there most of your doubts will surely be solved. If not, keep posting. Best of luck, Jose On Tuesday, November 4, 2014 10:54:03 AM UTC-6, Karl Karpfen wrote: OK, I'm sure

[beagleboard] Re: PRU0 base address

2014-11-05 Thread Karl Karpfen
OK, the base address seems to be correct. Nevertheless CTRL-register isn ot readable, so it seems some important PRU-initialisations are missing. So is there any clock or power that has to be turned on for PRU? Am Dienstag, 4. November 2014 17:54:03 UTC+1 schrieb Karl Karpfen: OK, I'm sure