Posted on May 17, 2011 at 12:24Hi Andy,
I am a bit confused about what exactly happened. Normally bus fault caused by debug accesses should be not affect program flow, because the error response should be received by the debugger, not by the processor core, so it should not affect your application code. Unless you are talking about using debug monitor exception code, or the debugger actually load some code in the chip and execute it. Maybe you can tell us more about the debug configuration you are using?
For normal applications, if it have a bus fault exception, usually we can just report the fault and then wait for the system to be reset.
Or if the design is running an OS, the task that generate the fault might get killed by the OS. Once a task has crashed, trying to continue from the same point may result in more errors anyway.
But if you are determined to continue the program execution, if the bus fault is imprecise (see Bus Fault Status Register in document), you can try clear the bus fault status register, and then do an exception return. If the bus fault is precise, you will have to clear the bus fault status register, increment the stacked program counter, and then do an exception return. But if the program has gone wrong, continue the execution doesn't guarantee the rest of the program can continue correctly.
Finally, I think you should try to identify the source of the bus fault exception first. You can try locate the fault by looking at the stacked PC and various fault status register to identify the problem.
