Re: [AD] Allegro's mixer

[Chris <kcat@xxxxxxxxxx>]

Robert Jr Ohannessian wrote:

> long long should be fine in interrupt context - as long as the data 
> isn't modified both inside and outside the interrupt simultaneously. 
> At this point, it would be possible for half the operation to break. 
> As long as the operation result isn't committed to memory, there 
> shouldn't be any problems though, as registers are saved across 
> interrupts.


The only issue I have against using a 64-bit buffer, then, is speed 
implications. AFAIK, DOS doesn't touch the final buffer outside of the 
timer, since it writes first to the mix_buffer, which isn't touched 
outside of the mixer functions, and then to the driver-provided buffer, 
which would more than likely be the DMA buffer under DOS.

> That said, there's no way to force the compiler to store 
> intermediaries in registers, so unless we explicitly provide locks on 
> the sample data, we should either:
> - Keep the code 386 compatible, for all 3 people on the planet who 
> still use those.
> - Make the behavior explicitly undefined.


I was under the impression that the sample data was locked? Unless you 
mean unwriteable. I was also under the impression those 64-bit warnings 
applied to the 386 and up, not just the 386. I personally would not 
object to requiring a 486 to use Allegro in interrupt driven OSs. If 
someone has a 386, I don't think they'd need anything more than Allegro 
3.11 functionality anyway.

- Kitty Cat