How to configure application properly?

Does Keil provide source for their __main, or is it library code? If there's no debugging symbols for __main (as would be if they did not provide source) then you can't step into it with a symbolic debugger.

The debugger should let you single-step the processor (rather than going line-by-line through code). You might try that.

code.

No, it's library code. I can't even step into __main(). The debugger does let

me single step until I get to following instruction:

bx __main()

When I try to step into __main(), the code starts running.

It could be problem with my installation of uVision version 5.4 or debugger firmware.

--------------------------------------- Posted through

Yes it is the start up library code. If you open up the disassembly window, give the disassembly window focus, then single step you can step through its assembly code - you see it doing zeroing and scatter loading, and then bx'ing to main().

As far as how to configure the application - start with an example for your chip that comes with the tools, when that works, adapt or copy it.

Regards, Richard.

• formatting link
  Designed for microcontrollers. More than 103000 downloads in 2012.

• formatting link
  Trace, safety certification, FAT FS, TCP/IP, training, and more...

Single step THE PROCESSOR. If you're "single stepping" in the symbolic debugger then the behavior you'll see is exactly what you are seeing.

It sounds like Richard has already given you guidance, but it's wise to remember that ANY debugger worth its salt (which is nearly all that I have used) will have some sort of bare-metal mode that lets you look at the disassembled code in memory, and step through individual processor instructions, even when what's there is not tied to any code that the debugger can find.

If you haven't identified and used the debugger mode that lets you single step the processor itself, then the debugging problem is somewhere between the front of the monitor and the keyboard.

_sys_exit is what gets called when your main() exits, which it should never do in any normal embedded system.

it,

going

does

debugger

minutes but haven't reached my main(). If I run, I get stuck at while(1) mentioned earlier.

I'll try to find example for this chip and see if I can use it to get to my main().

--------------------------------------- Posted through

What does your main() look like? If this is just a bare-bones main() for test/evaluation purposes, are you "falling off" the end of it? Stick in a "while (1) { Toggle_a_pin(); /* or something */ }" and see what happens.

Ten

while(1)

my

Now I just have bare-bones main() now. It has while(1), but I still can't break there. I end up in my initialization assembly startup file.

I can make it main() with an example application.

--------------------------------------- Posted through

Something is happening within the startup code that is making it decide to bomb.

Finding startup code that is correct for your chip is probably best, if such exists.

for

file.

your help!

--------------------------------------- Posted through

Something that I haven't seen anyone ask, but should have thought of myself:

Is there an evaluation board for this chip out there someplace, and does it come with a Keil tool chain? Often your best bet is to start with a demo application and go from there.