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

Back to Subject List

Old thread has been locked -- no new posts accepted in this thread
???
07/04/07 09:29
Read: times


 
#141483 - RE: C interrupt overhead
Responding to: ???'s previous message
Jan Waclawek said:
The interrupt overhead in C is what most probably prevents you to reach a higher rate. Running the code in simulator and watching the timer will show you, what's going on.

Or look at the generated assembler to see what's going on.

Note: for Keil, adding a using option significantly reduces the overhead, as the compiler then won't save all registers...
http://www.keil.com/support/m...cdecls.htm

Try rewriting the ISR in inline asm.

No, don't do that!

Write it as a separate assembler module, in an assembler source file.


List of 24 messages in thread
TopicAuthorDate
How to generate 100 Khz square wave 80C51            01/01/70 00:00      
   Tips            01/01/70 00:00      
      1 KHz            01/01/70 00:00      
         that's ...            01/01/70 00:00      
         SI Units            01/01/70 00:00      
            1024 = Ki            01/01/70 00:00      
   C interrupt overhead            01/01/70 00:00      
      RE: C interrupt overhead            01/01/70 00:00      
         and listen to the guru! :-)            01/01/70 00:00      
            several reasons            01/01/70 00:00      
               Yet another reason to avoid using HLL's!            01/01/70 00:00      
               Optimizations            01/01/70 00:00      
   Use Timer0 auo reload            01/01/70 00:00      
      using            01/01/70 00:00      
      interrupt vectoring vs. latency            01/01/70 00:00      
         Then the Interupt must be in ASM            01/01/70 00:00      
            do you mean...            01/01/70 00:00      
               No            01/01/70 00:00      
               Compiler generated IRQ vectors            01/01/70 00:00      
                  OOPS we both meant LJMP REALisr            01/01/70 00:00      
                  the original problem...            01/01/70 00:00      
         Interrupt Latency + Vectoring + Service            01/01/70 00:00      
   indufficient info            01/01/70 00:00      
   use MCS52 better option            01/01/70 00:00      

Back to Subject List