Re: [hatari-devel] Reeking rumber analysis

[Eero Tamminen <oak@xxxxxxxxxxxxxx>]

Hi,

On maanantai 09 syyskuu 2013, Laurent Sallafranque wrote:
> The part of the code that calls the faulty function :
> 
> $06f942 : 207c 0009 3850                       movea.l   #$93850,a0
> $06f948 : 4240                                 clr.w     d0
> $06f94a : 323c 0100                            move.w    #$100,d1
> $06f94e : 7401                                 moveq     #1,d2
> $06f950 : 6100 0dd6                            bsr       $70728
> 
> 
> The part of the code that loops forever :
....
> $07075c : 7a00                                 moveq     #0,d5
....
> $070794 : 2248                                 movea.l   a0,a1
....
> $0707cc : b699                                 cmp.l     (a1)+,d3
> $0707ce : 6702                                 beq.s     $707d2
> $0707d0 : 7a01                                 moveq     #1,d5
> $0707d2 : 24c3                                 move.l    d3,(a2)+
> $0707d4 : 51ce ffca                            dbra      d6,$707a0
> 
> $0707d8 : 4a85                                 tst.l     d5
>              <------- Here is the the test that never goes throw with

What's at #$93850 and at surrounding addresses?


I would also recommend using cpu profiling and reading through
"profile save <output file>" disassembly.  It gives you an overview
& hints on what the code before the problematic point was doing,
where certain registers were modified and what other parts of code
one might need to investigate closer...


	- Eero