see below, please. regards,
Richard Erlacher ----- Original Message ----- From: "Jan Waclawek" <[EMAIL PROTECTED]> To: <sdcc-user@lists.sourceforge.net> Sent: Friday, August 29, 2008 3:08 PM Subject: Re: [Sdcc-user] Quickstart document,was: Virus in SDCC-2.8.0-setup.exe - MD5 etc tutorial > Richard, > > I answer the simulator-specific part separately. > >>I've looked for it, but haven't found one that operates as one might >>expect. >>Perhaps its doc's need clarification. Are there any? I recall reading >>something about a simulator and a link to a website that seemingly no >>longer >>works. >> > > C:\Program Files\SDCC\doc\usim\index.html is the starting point of the > documentation, but you don't need to download anything. > Simply run "C:\Program Files\SDCC\bin\s51.exe" and type "help". That might > give you a rough feeling of what it is all about. > >>I always found simulators quite straightforward. The difficult part is >>getting a complete characterization of whatever's to be simulated. >>Approximations don't make a simulator work well. In this context, the >>"trick" is to predefine the environment within which the simulator has to >>function. Handwaves don't help, either. Some things have to be clearly >>defined, at least in terms of inputs, outputs, and processes. > > I don't quite understand what do you mean by this. > There are times of the year when I have routinely have some time on my hands. A couple of years ago, I requested information from one of the 805x-clone makers about the specifics of their internal timing so I could implement a relatively precise, (timing-valid) simulator that allowed me to use scheduled event-relative stimuli to the MCU. though I've repeated my request, I'm still waiting. Things like the built-in timers and UART(s) that operate synchronously with the 12-ticks-per-system-clock model are easy to understand, as Intel, in this case, documented them, hence, are easy to simulate. In the case of one-clockers, particularly those that have programmable options as to how they treat relationship between system clock and that external timebase input, behave differently. If I want to obtain precise details about how the MCU interprets an event nn nanoseconds after the rising edge on X1, and when it sets this flag or that, I need detailed information from the manufacturer. Suppose I want to make the UART in MODE 0 function as a shift register, and not stop generating TxCk when a byte is assembled or disassembled. How can I do that if I don't know when it decides whether to generate a clock? How do I know when I have to read SBUF? How do I know when to clear the RI flag so it doesn't miss a clock? How do I determine whether I should even attempt do those things? So far, it's all a mystery, and any attempt at simulation would be a guess. As for the environment, the simulator can operate from a driver clock "event" or it can operate from "real" time. A schedule, in some implementations, implies "real" time, though the schedule actually is developed by a tool that interprets the schedule source and synthesizes the "real" time on which the schedule operates and from which it derives the required stimuli. This allows you to specify an oscillator as a loop, and a recurring event, similarly, as a periodic function. It also allows the insertion of random intervals in the event schedule, as you might want to do in order to "monte-carlo" your simulation in order to examine the effect of input timing variations. > > I think, a step-by-step quickstart example on a trivial "blinkey" program > would help here, too; but I am not competent enough to make one. > Right now, perhaps I'm not either. I'm not understanding on what you mean. > > Jan Waclawek > ------------------------------------------------------------------------- This SF.Net email is sponsored by the Moblin Your Move Developer's challenge Build the coolest Linux based applications with Moblin SDK & win great prizes Grand prize is a trip for two to an Open Source event anywhere in the world http://moblin-contest.org/redirect.php?banner_id=100&url=/ _______________________________________________ Sdcc-user mailing list Sdcc-user@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/sdcc-user
