[beagleboard] Re: gdbserver shared library PRU loader

2016-10-13 Thread TJF
Am Donnerstag, 13. Oktober 2016 10:42:13 UTC+2 schrieb Jelle Spijker: > > The PRU program is genereted in C with the CCS IDE, the compiler is TI > V2.1.2 en the linker command file is AM335x_PRU.cmd > That doesn't work. You'd have to extract the .text section from that file and load only that d

[beagleboard] Re: gdbserver shared library PRU loader

2016-10-13 Thread Jelle Spijker
The PRU program is genereted in C with the CCS IDE, the compiler is TI V2.1.2 en the linker command file is AM335x_PRU.cmd // /* AM335x_PRU.cmd */ /* Copyright (c

[beagleboard] Re: gdbserver shared library PRU loader

2016-10-13 Thread TJF
Am Donnerstag, 13. Oktober 2016 09:16:31 UTC+2 schrieb Jelle Spijker: > > When I run the test as a normal user and output the return values to the > cout it returns > prussdrv_init returns: 0 > prussdrv_open returns: -1 > Segmentation fault (core dumped) > > I'm not sure ATM. It may also need wr

[beagleboard] Re: gdbserver shared library PRU loader

2016-10-13 Thread Jelle Spijker
Hi TJF, Thnx for your reply. I tried the following steps: I added the /dev/uio[0..7] to the group bbb (of which my user is a member) and gave them read right for a group. I changed the typo prussdrv_open(PRU_EVTOUT_0); When I run the test as a normal user and output the return values to the co

[beagleboard] Re: gdbserver shared library PRU loader

2016-10-12 Thread TJF
Just found another issue: Am Mittwoch, 12. Oktober 2016 15:05:12 UTC+2 schrieb Jelle Spijker: > > Fragment of the Shared library: > > void BBB::loadPRUfirmware(const unsigned int &number, const std::string > &firmwareLocation) { > //Initialize structure used by prussdrv_pruintc_intc > tpr

[beagleboard] Re: gdbserver shared library PRU loader

2016-10-12 Thread TJF
Hi! Am Mittwoch, 12. Oktober 2016 15:05:12 UTC+2 schrieb Jelle Spijker: > > Hi All, > > I'm running Arch Linux on my BBB, so no remoteproc for me :-( when working > with the PRU. > > You're lucky, you didn't waste time for testing remoteproc. > Because I work with prussdrv the program should b