Re: [hatari-devel] DSP "running" on non-falcon

[Nicolas Pomarède <npomarede@xxxxxxxxxxxx>]

Le 27/07/2022 à 12:05, Eero Tamminen a écrit :
>          if (dsp_core.running == 0)
>                  return;
> ----------------------------------------
>
> This is because dsp_core.running gets set only after first write to 
> Falcon DSP CPU_HOST_TRXL IO register, which obviously can happen only on 
> Falcon.

yes, it will be harmless on TT for example, but considering the 
ressources needed to run 68030+cache+mmu, all code path that can be 
skipped as soon as possible will speed up emulation a little

> > so, as soon as DSP is enabled in the GUI or with some --options, this 
> > calls DSP_Enable() in falcon/dsp.c, which sets bDspEnabled=true , even 
> > if machine is not a falcon and has no DSP at all (even an STF will 
> > report bDspEnabled=true)  :(((
>
> Yeah, DSP_Enabled() seems to be called from many places based just on 
> the DSP type.
>
> I think the reason for doing things based on specific emulation 
> settings, instead of machine type, is that machine type can change quite 
> late in emulation (re-)initialization (when TOS is loaded), based on the 
> TOS version (in tos.c).

maybe ; in such case code should be checked one day, because in all 
accuracy when changing machine type, we should first change machine 
type, then re-init all depending HW such as DSP or SCC or SCSI or whatever.

> Ok, you already pushed the change.  I tested it and it seems to work OK. 
>   I also updated the release-notes description a bit.

As it would avoid to call DSP_Run() and I didn't see problem when 
testing, I thought it could go in dev version anyway.

> (And fixed another issue AddressSanitizer noticed.)
>
>
>      - Eero