Re: [sdcc-devel] Proposal for return-from-main and exit(), and simplifying the regression tests.

[Daniel D. <dr...@ma...>]

On Mon, 26 Jul 2004, Phuah Yee Keat wrote:

> Frieder Ferlemann wrote:
> > Phuah Yee Keat wrote:
> >> The program will be put into a loop when run in the hardware or
> >> simulator. It will be as a system-reset have happened. Previously, the
> >
> > unfortunately not - it will just continue executing at 0x0000
>
> As I understand from the mcs51 port, jumping to 0x0000 will be as good
> as a software-reset. All the SFRs will be re-initialized again, and all

Why? I don't think so. Jumping to 0 will just continue at 0, no SFR
initialisation will happen.

> [a bit off-topic]
> I am actually working on a new 8051 simulator, mainly because I found
> the current one inadequate for my own testing purposes, here are some of
> the reasons:
> 1. Its in C++, which is not one of my strongest language.
> 2. It provided too much of a debugger interface instead of a simulator
> interface. What I would like to see in simulators is that plugins can be
> written to mimic the connection to the external devices (something like
> cereal.sourceforge.net, but a lot more simpler), like 7-segment LED

uCsim simulates microcontroller and _not any external circuit_! It is
not a simulation framework like the cereal, so you can think that
uCsim is just a plugin of a big framework doing a micro. I'm trying to
keep to unix philosophy that one program should do just one thing.
uCsim will never simulate other than internal of the micro.

So if you expect a framework, you did a good decision not using uCsim.

Daniel