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

Back to Subject List

Old thread has been locked -- no new posts accepted in this thread
Jim Granville
04/05/16 14:08
Read: 416 times

#190651 - Cycle variations
Responding to: Erik Malund's previous message
Erik Malund said:
You are probably better to use a more modern Simulator, like TSim51, that shows Cycles

these thing get more and more irrelevant/imprecise since nowadays virtually every derivative has it's own cycle/instruction table. Also for some cache misses play a role.

Hehe, well yes, this is something of a moving target.

However, Zoltán has added a while back variant-cycle support for SiLabs and others, but the cache parts are a challenge.

I am testing the cache style rules for EFM8LB1, and some code is certainly 'context dependent' - that said, it is predictable given a complex enough rule set, and mostly needs ALIGN to minimize time.

Test vehicle is a VGA display, as that easily shows any errant jitter.
I can get predictive and stable results, with some care, @ 72MHz

The new STC8F changes the timing & coding rules yet again, but hopefully simpler.

List of 8 messages in thread
AT89LP Developer Studio      Kerim Fahme      04/03/16 01:52      
   AT89LP Developer Studio reports Cycles      Jim Granville      04/04/16 22:17      
      (this and?) other chips      Erik Malund      04/05/16 08:29      
         Thank you, Erik,      Kerim Fahme      04/05/16 11:15      
         Cycle variations      Jim Granville      04/05/16 14:08      
      Thank you, Jim      Kerim Fahme      04/05/16 11:04      
         TSim51 Windows/Linux       Jim Granville      04/05/16 13:58      
            Thank you again, Jim.      Kerim Fahme      04/06/16 01:03      

Back to Subject List