[A89] Re: Traps...
[Prev][Next][Index][Thread]
[A89] Re: Traps...
At 22:53 2001-07-23, you wrote:
>Olle, I don't need to worry about setting the trace bit again from within the
>trace function... rte does it for me. The supervisor stack will contain the
>previous sr and pc so that rte can return like nothing happened. In fact,
>rte is
>the only instruction that I have executed at every trace interrupt... so I
>should get something around a 50% slow down...
Oh, yeah, didn't think about that.
The slowdown should be much more then 50% though, it doesn't take 0 time to
save all those registers on supervisor stack, fetch the exception vector,
and jump to it.
if it doesn't look all that slowed down, it is probably because interrupts
are still run at full speed as far as I can see in my 68k docs.
>I don't plan on 'releasing' this program, it was more of an experiment than
>anything serious... although I might post the source on my webpage. BTW:
>it does
>not work on queue... :( Queue disables interrupts apperantly... darn
>rowread :)
ofcourse. sounds like 99% of my "programming" :)
///Olle