Li Laurent
This looks like the code from SV version. It was replaced by version v1.3 (http://files.dhs.nu/files_demo/Lockup_v13.zip). This version has some fixes overflow in audio decoder being one of them.
This looks like the timer interrupt yes.
The "formula" is a bit tricky because I need to "lock" the timer interrupt with the SSI generated interrupts:
move r7,a ; SSI interrupt sample pointer
move #>DECODED_BUFFER_SIZE,x0 (buffer size)
move #>decoded_buffer,y0
sub y0,a
cmp x0,a #>decoded_buffer+DECODED_BUFFER_SIZE,r0
.IF <GT>
sub x0,a #>decoded_buffer,r0
.ENDI
; calculate new timer value to sync-lock with DMA
lsr a
lsr a #>640|(1<<13),b
sub a,b
MOVEP b,X:<<M_SCCR ;Set the interrupt rate.
With this timer stays in sync with the replay buffer. I don't remember how I came up with 640. This is probably related to the half of duration of the playback buffer.
Hope that helps
Regards
Mariusz
On 12/5/2020 2:08:18, Laurent Sallafranque <laurent.sallafranque@xxxxxxx> wrote:
Hi Mariusz,
Can you confirm me that the SCI timer interrupt runs the following code
each time the Timer triggers ?
p:002c 6e1100 (02 cyc) move r6,y:$0011
p:002d 6e9300 (02 cyc) move y:$0013,r6
p:002e 0aa524 (04 cyc) bset #4,x:$ffe5
p:002f 205e00 (02 cyc) move (r6)+,r6
p:0030 0bd669 (04 cyc) btst #9,r6
p:0031 205600 (02 cyc) move (r6)-,r6
p:0032 0e0038 (04 cyc) jcc p:$0038
p:0033 0aa504 (04 cyc) bclr #4,x:$ffe5
p:0034 085eaf (04 cyc) movep x:$ffef,x:(r6)+
p:0035 6e1300 (02 cyc) move r6,y:$0013
p:0036 6e9100 (02 cyc) move y:$0011,r6
p:0037 000004 (04 cyc) rti
How do you compute the $fff2 value ? ($a) (What is the Fosc value in
the formula ?)
Regards
Laurent
Le 04/12/2020 à 22:03, Mariusz Buras a écrit :
> That's exciting!!
>
> On Fri, 4 Dec 2020 at 20:55, Laurent Sallafranque
> wrote:
>> Mariusz, no need to send me a test case for the host interface
>> interrupts, your demo is working now ;)
>>
>> Laurent
>>
>>
>>
>> Le 04/12/2020 à 21:51, Laurent Sallafranque a écrit :
>>> Hi,
>>>
>>>
>>> My version of the demo was corrupted (2 bombs).
>>>
>>> I've downloaded another one.
>>>
>>>
>>> I've fixed the DSP Host interrupts and now, I have the demo with the
>>> tridi.
>>>
>>> It was present but had never been tested before.
>>>
>>> The demo runs super well ;)
>>>
>>>
>>> Still have to give a try to the timer now to have the music.
>>>
>>>
>>> Eero, can I comit my changes into the DSP interrupts ?
>>>
>>>
>>> Regards
>>>
>>> Laurent
>>>
>>>
>>>
>>>
>>
>