[hatari-devel] Twelve demo bombs with fast-forward and freezes on Falcon?

[Laurent Sallafranque]

Sorry for the typo in my precedent mail : you must read :

I've tested with IoMem_WriteByte : voice.prg works now (with a little 
other problem, sound play more than it should !).

Regards

Laurent



Le 01/06/2011 00:50, Laurent Sallafranque a écrit :
> I've tested with IoMem_WriteByte : voice.prg works no (with a little 
> other problem).
>
> The game conquest of elysium works too (no more freeze).
> Should I commit this ?
>
> May this corrects the problem you encounter in monochrome mode with 
> twelve demo ?
>
> Regards
>
> Laurent
>
>
>
>
> Le 01/06/2011 00:41, Laurent Sallafranque a écrit :
>> I've found something, but I need your advice :
>>
>> in crossbar.c, (function Crossbar_Process_DMAPlay_Transfer), there's 
>> the following code :
>>
>>     /* Check if end-of-frame has been reached and raise interrupts if 
>> needed. */
>>     if (dmaPlay.frameCounter >= dmaPlay.frameLen)
>>     {
>>         /* Update sound */
>>         //Sound_Update(false);
>>
>>         /* Send a MFP15_Int (I7) at end of replay buffer if enabled */
>>         if (dmaPlay.mfp15_int) {
>>             MFP_InputOnChannel(MFP_TIMER_GPIP7_BIT, MFP_IERA, 
>> &MFP_IPRA);
>>             LOG_TRACE(TRACE_CROSSBAR, "Crossbar : MFP15 (IT7) 
>> interrupt from DMA play\n");
>>         }
>>
>>         /* Send a TimerA_Int at end of replay buffer if enabled */
>>         if (dmaPlay.timerA_int) {
>>             if (MFP_TACR == 0x08) {       /* Is timer A in Event 
>> Count mode? */
>>                 MFP_TimerA_EventCount_Interrupt();
>>                 LOG_TRACE(TRACE_CROSSBAR, "Crossbar : MFP Timer A 
>> interrupt from DMA play\n");
>>             }
>>         }
>>
>>         if (dmaPlay.loopMode) {
>>             Crossbar_setDmaPlay_Settings();
>>         }
>>         else {
>>             dmaCtrlReg = IoMem_ReadByte(0xff8901) & 0xfe;
>>             //IoMem_bput(0xff8901, dmaCtrlReg);
>>         }
>>     }
>> }
>>
>>
>> I've commented the IoMem_bput line, and I don't get anymore the 
>> M68000 Bus Error writing at address $ff8901.
>> (tested with voice.prg). Of course, without this line, it doesn't 
>> work well, as sound continues to play.
>>
>>
>> Should I use something else than IoMem_bput here (IoMem_WriteByte for 
>> example) ?
>>
>> What's the difference between these to memory acces ?
>>
>> Regards
>>
>> Laurent
>>
>>
>>
>>
>> Le 31/05/2011 23:36, Laurent Sallafranque a écrit :
>>> I'd enjoy to reply, but I don't know how to break when the 68030 
>>> reads or write from/to $ff8901
>>>
>>> Any tips would be appreciate
>>>
>>> Laurent
>>>
>>>
>>> Le 31/05/2011 23:14, Eero Tamminen a écrit :
>>>> Hi,
>>>>
>>>> On keskiviikko 01 kesäkuu 2011, Laurent Sallafranque wrote:
>>>>> Here are some examples taken from old threads :
>>>>>
>>>>> https://lists.berlios.de/pipermail/hatari-devel/2010q3/002307.html
>>>>>
>>>>> https://lists.berlios.de/pipermail/hatari-devel/2010q1/001232.html
>>>>>
>>>>> There's also voice.prg
>>>>>     -->  M68000 Bus Error writing at address $ff8901.
>>>> Are they doing byte, word or long access?
>>>>
>>>>
>>>>> I remember there were 2 or 3 more programs with the same problem, 
>>>>> but I
>>>>> can't find them.
>>>>
>>>>     - Eero
>>>> _______________________________________________
>>>> hatari-devel mailing list
>>>> hatari-devel at lists.berlios.de
>>>> https://lists.berlios.de/mailman/listinfo/hatari-devel
>>>>
>>>
>>> _______________________________________________
>>> hatari-devel mailing list
>>> hatari-devel at lists.berlios.de
>>> https://lists.berlios.de/mailman/listinfo/hatari-devel
>>>
>>
>> _______________________________________________
>> hatari-devel mailing list
>> hatari-devel at lists.berlios.de
>> https://lists.berlios.de/mailman/listinfo/hatari-devel
>>
>
> _______________________________________________
> hatari-devel mailing list
> hatari-devel at lists.berlios.de
> https://lists.berlios.de/mailman/listinfo/hatari-devel
>