How do you debug embedded programs?
In this article, we will discuss alternative ways to correct embedded programs?!!
Some manufacturers – Intel as an example – show special debugging probes for embedded processors. In essence, this is frequently a Flexi-cable probe with a ‘head’ that completely matches the target processor kind and size and clamps into the processor socket instead of the processor. The cable ends up in a processor soul box (separately powered), that really replaces the processor and would possibly run at something from an outsized fraction of the traditional processor speed up to full speed, depending. This soul box, in turn, hooks up to a laptop and provides a full symbolic computer program the same as the computer program in Visual Studio.
In the picture below is an associate example debugging probe set connected to associate embedded board (credit Intel).
For sure, this suggests that you just square measure using a compiler/linker that outputs symbolic information supported by the manufacturer, so inside the case of Intel this can be one issue like PL/M as an example.
Below is a helping example symbolic debugging window (credit Intel).
Effectively, then, you have a very embedded application that couldn’t even have a computer screen of any kind, however, you are able to correct nearly as if you are testing a desktop application.
Embedded systems include a large shape of environments that have little in common.
Some of the debugging methods I have employed in my embedded career:
- Attach the Wyse terminal to an interface, burgled kernel program, examine and modify professional the kernel data structures.
- Plug the interface analyzer into the interface line between the box and additionally the peripheral, repeat the action at law, watch the interface protocol exchange.
- Telnet into the rectify TCP port that I coded up to support ME read/write access to the components of the system. I had the telnet open on the laptop as I walked through sensors, triggering them under control sequences and debugging the responses.
- On one system that had a touchscreen thereon, I had a secret keyboard input, thus I might open up the box, pull a cable out, insert my special cable instead, connect the keyboard to its cable, and kill the screen to reveal the console.
- Hook up a piece of honest electronic equipment to record the input going into my box and the output going out additionally, and run a check state of affairs on the machine (which jointly drove the trigger on the scope) million times for each set.
- Using a sample of actual inputs, pre-programmed things, and also irregular” fuzz”, record elaborate event logs from every inputs and output (events were logged by inflicting them over the network to a work server), run the inputs through a software system machine and examine each discrepancy between outputs.
- Writing (and overwriting) a small, however terribly special event log to NVRAM just in the case of the watchdog kills the system (the first factor I did on startup was to scan that log).
- In notably luxurious things, remote gdb.
- And that’s only a little sample of the techniques embedded programmers which are used.
To read more articles like this visit Eduengteam.com
You can find more useful tutorials at Click Here