??? 06/05/06 20:00 Read: times |
#117790 - Yes, that's part of it ... Responding to: ???'s previous message |
SFR definitions, and making them easy and transparent is a part of it. The simulator is part of it, too. The simulator should simulate what the part is doing, not just what the putative core SHOULD be doing. In the course of that, it should know and, if desired, report the state of every flipflop and register in the part that's in any way associated with what the programmer might want to know.
True, no simulator is perfect, but some are VERY close. The tools I was using under UNIX back 25 years ago certainly didn't simulate processors and memory and PCB's as black boxes, they attempted to assemble precise timing models and simulated with them. I could design a circuit with a 68008 and a number of MCU's, each with its own dedicated task, and it (HILO) would, as faithfully simulate the electrical behavior of all the components including the PCB traces, passives, etc, and provide a view of the simulated output waveforms. It would also tell me how long it took each device to perform each operation. A pretty slick editor/assembler/simulator I got for the 8048 some 12 years or so ago keeps pretty good track of what the 8048, its timers, etc, are doing. It can be done, but someone's got to do it. Keil's folks just didn't think it worth the effort, assuming the "dummies" out there would "buy it anyway." They were right, of course. RE |