Re: [MiNT] CPU hogs, continued

[Michael Schwingen <rincewind@a-tuin.dascon.de>]

On Sun, Feb 10, 2002 at 01:00:42PM +0100, Konrad M. Kokoszkiewicz wrote:
> 
> I can imagine, that if timer C is emulated on Milan, the saved IPL may 
> be rather high (because it is the priority of the interrupt which 
> emulates timer C, not "real" one).

Timer C works normally on the Milan - however, there is no VBL interrupt,
and it is emulated from within the ROM's 200-hz timer by calling the VBL
handler every 4th time:

<end of 200hz timer handler in ROM>
	movem.l	(sp)+,d0-d7/a0-a6  ; restore saved regs
	btst	#1,tc_rot.w
	beq.s	1f

	move.b	#%11011111,isrb+mfp.w
	move.l	$70.w,-(sp)        ; call VBL vector
	rts

1:	move.b	#%11011111,isrb+mfp.w
	rte

So the VBL handler is called with the original stack frame from the Timer C
interrupt.

cu
Michael