Email: Password: Remember Me | Create Account (Free)

Back to Subject List

Old thread has been locked -- no new posts accepted in this thread
???
01/01/05 22:18
Read: times


 
#84212 - Please use Correct Terms
Responding to: ???'s previous message
Shahzad said:
this is possible that the code which is compiled correctly using Keil and also run as expected in keil debugger will not run on 8052 because the execution environment for debugger (PC hardware) and actual 8052 hardware are different.

The above, as stated, is not true[1]: the Keil Debugger does run the code in the actual 8052.

However, I am now fairly certain that you must be talking about the Keil simulator - not the debugger.
In this case, it is true that you are not running on an actual 8052 in your actual target hardware - you a running a simulation using the PC.

If that is true then how can i code some thing which i am sure that it will run the same on PC and 8052.

You must understand the limitations of simulation - as I explained before:
I said:
You must mean the Simulator?

Yes, it is possible - because it is only a Simulation.
If there is something - anything - about your hardware that does not exactly match what the simulator simulates, then that obvioulsy invalidates at least that part of the simulation!

http://www.keil.com/discuss/docs/thread5315.htm
http://www.keil.com/discuss/docs/thread4173.htm
http://www.keil.com/discuss/docs/thread4173.htm




[1] Actually, all testing & measurement involves some disturbance of the system under test:
  • When you take a voltage reading, you load the circuit - which affects the voltage;
  • When you take a current reading, you add extra series resistance - which affects the current
  • etc, etc,...
    See Heisenberg's Uncertainty Principle.

    Similarly, any form of debugging has some impact on the software under tes:
  • Using LEDs, you have to add code to drive the LEDs;
  • Adding debug trace with printf uses memory space & processor time;
  • A monitor uses processor resources;
  • Even an ICE can add loading on the processor pins, and introduce all sorts of problems with actually connecting it to the target system

    Thus there is no such thing as entirely non-intrusive debugging - no matter what tools you have.

  • List of 25 messages in thread
    TopicAuthorDate
    Keil C Code not working            01/01/70 00:00      
       Desktop turbo c++ ?            01/01/70 00:00      
          Slightly??!!            01/01/70 00:00      
          agreed            01/01/70 00:00      
             i got the math to do div / multiply            01/01/70 00:00      
             Debugger? Simulator? Maths?            01/01/70 00:00      
                Question...            01/01/70 00:00      
                   Please use Correct Terms            01/01/70 00:00      
                Re Andy Debugger, Simulator? Math            01/01/70 00:00      
             *unnecessarily* complex!            01/01/70 00:00      
                Re andy            01/01/70 00:00      
                   Divide and Conquer!            01/01/70 00:00      
             Bad Burner?            01/01/70 00:00      
                Re Andy Bad Burner            01/01/70 00:00      
             Project Configuration?            01/01/70 00:00      
       debugging?            01/01/70 00:00      
          thanks for the link            01/01/70 00:00      
       Data Sizes - Lesson not yet learned?            01/01/70 00:00      
          Re Andy            01/01/70 00:00      
       brackets ?            01/01/70 00:00      
          (no) brackets            01/01/70 00:00      
       WriteCommandLCD ?            01/01/70 00:00      
          Re Andy            01/01/70 00:00      
             Problem not fixed yet?            01/01/70 00:00      
                Problem not fixed yet?            01/01/70 00:00      

    Back to Subject List