m68k clock problem

[Nicolas PIGEON]

Hi Victor,


>Nicolas, 

>Disclaimer: I have no previous experience with 68302. 

>I guess you are using debugger to trace the interrupt handler. Sometimes 
>it is possible. But have in mind that timer continues to tick while you 
>trace the interrupt handler. When you leave the interrupt handler, next 
>timer interrupt ready to be served. It is not happend before because 
>processor 
>level is higher than interrupt level while interrupt handler executed. 
>RTS instruction restore previous processor level, and next interrupt 
>occurs immediately after RTS instruction. 
>You may try to patch gdb stub to reset timer when trace interrupt handler 
>leaved. 


I don't use a debugger but a HP logic analyzer which monitors various signals around the 68302 (Address bus, data bus, chip select...). The system always runs as it would in a real environment.


>If this is not a case, the situation you describe looks like 
>you miss to reset interrupt condition in timer block when interrupt handler 
>executed. Looking through ckinit.c, I see that this is done. 
>Nevertheless check 
>twice all definition against documentation. 


Indeed, the interrupt condition seems to be reset inside the interrupt handler (according to the code), but I never jump inside that routine, not even once !!


>Another possibility timer ticks too frequent due to wrong divider. Check 
>this too. 


I tried to set a Prescaler value of 16 in the TMR, but the problem occurs exactly at the same place...


>Victor 

Thanks a lot for your ideas, it helped me check some points!