I'm trying to track down the source of an exception in a C674x dsp core (OMAP-L137). The easiest way seems to be to use the ROV LOG_system record that shows the trace information, but for some reason, I'm getting "null value" in the log entries 0 through 5 rather than the trace messages.
I've been able to manually locate the exception by setting a break point at hwi1 and reading out the B3, NRP, EFR and IERR registers, but this requires running with the debugger attached. Ideally, I'd be able to read back the ROV log after a runtime event and figure out what happened.
The rest of the ROV data looks correct, and I can step through the code and see where the calls to _LogEvent are happening during EXC_exceptionhandler and EXC_internal. Prior to the exception there are no entries in LOG_system, and after there are 6 entries each with "(null format)"
System: CCS v4.2.4, DSP/BIOS 5.41.09
This maybe relevant: we're building the system under linux, starting the processor, and then attaching the jtag debugger (XDS100v2) with a minimal gel file under CCS for debugging. We load symbols from the .out file to get support for the ROV.
I appreciate any suggestions. Thanks!
