Am 06.02.2016 um 17:31 schrieb Michael Hennebry:
It seems to me that the way to handle a specific invalid instruction,
0xFFFF or anything else, is to add it to the instruction set.
If desired, startup code can read a configuration file.
Yes, implement all the options we discussed today ends up in configuring
the simulavr with an additional config file. But this also opens a door
to make the instruction set specific to the core itself. Some
instructions are not valid on all cpu cores.
My idea is quite different from all that:
simulavr is a tool from programmers for programmers. If someone need for
a debug session a different handling of an opcode, it is quite simple to
implement the behavior in three lines of c code. I have no idea how
flexible a config file should be. This ends in writing a simulator
configuration language :-) At this point it looks simpler for all people
to change 3 lines of simulavr source code instead of reading the
simulavr manual for the "how to config the decoder for special
instructions".
Is there really someone who needs a additional config file to get all
the options in? I compile the tool in 3 or 4 seconds and change the
behavior of an instruction in less then a minute. So the solution we
discuss did not give more flexibility then the open source code itself.
But if someone wants to implement a simulavr configuration language...
:-) I will not working on this topic. There are a lot more jobs todo. If
some one can spend some time on twi? for mega128 and others? ;)
Regards
Klaus
_______________________________________________
Simulavr-devel mailing list
Simulavr-devel@nongnu.org
https://lists.nongnu.org/mailman/listinfo/simulavr-devel
