Re: [hatari-devel] question about natfeats' code

[Nicolas Pomarède <npomarede@xxxxxxxxxxxx>]

Le 26/10/2014 23:17, Douglas Little a écrit :
>     Douglas, have you seen anything funny related to SR,
>     when using NatFeats?
>
>
> Not that I have noticed - although SR accidents with the low byte would
> mainly affect condition codes, which are pretty context sensitive and
> 68k sets these continuously anyway.
>
> If the upper byte is wrong though it would probably show up pretty quickly.
>
> D

hi

Unless in very special case, I don't see why the C code for a natfeat 
command would require to change SR (or any other registers in a fact), 
except maybe to indicate a specific return status in CCR or in a register.

So, I think it's up to each natfeat command to do what it wants with the 
registers, even if the default behaviour will certainly be to not do 
anything, but only call debugger/change emulation speed/...

I will commmit a patch to remove SR save/restore, let me know if it 
breaks anything, but I don't think it will.

Nicolas