mm driver message handling in DefDriverProc

[Saulius Krasuckas]

* 6 years, 7 months ago Eric Pouech did a change [*]:
> Started calling mixer as a 32 bit low level driver (no longer 16 bit)
> Added support for yield proc in mci struct
> Added mmTaskXXX16 and mmThreadXXX16 functions
> Refreshed implementation of 16 bit native & asynchronous MCI driver command (through mmtask.tsk).

...
> @@ -4193,13 +4273,13 @@ LRESULT WINAPI DefDriverProc(DWORD dwDri
>  {
>      switch (Msg) {
>      case DRV_LOAD:
> - case DRV_FREE:
> - case DRV_ENABLE:
> case DRV_DISABLE:
> -
>    return 1;
> case DRV_INSTALL:
> -
>    return 0;
> - case DRV_ENABLE:
> - case DRV_FREE:
> case DRV_REMOVE:
> -
>    return 1;
> -
>    return DRV_SUCCESS;
> default:
>        return 0;
>    }

I'm do not follow a logic, as nowadays PSDK defines DRV_SUCCESS as 0x0001.  
Eric, can't both cases of 1 and DRV_SUCCESS be merged into a single case?  
Maybe this change address some 16-bit issues.  No?

[*] http://cvs.winehq.org/cvsweb/wine/multimedia/Attic/mmsystem.c.diff?r1=1.24&a...

[Eric Pouech]

Saulius Krasuckas wrote:
> - 6 years, 7 months ago Eric Pouech did a change [*]:
> > Started calling mixer as a 32 bit low level driver (no longer 16 bit)
> > Added support for yield proc in mci struct
> > Added mmTaskXXX16 and mmThreadXXX16 functions
> > Refreshed implementation of 16 bit native & asynchronous MCI driver command (through mmtask.tsk).
>
> I'm do not follow a logic, as nowadays PSDK defines DRV_SUCCESS as 0x0001.  
> Eric, can't both cases of 1 and DRV_SUCCESS be merged into a single case?  
> Maybe this change address some 16-bit issues.  No?

trying to get my mind 6 years & 7 months ago. As it was a Monday, it 
makes things even harder ;-)
It seems that the main change in that area was to change all return 
values, but IIRC some were documented (with DRV_SUCCESS) some other 
weren't (or were documented as boolean).
Setting everything as DRV_SUCCESS should be fine.

But why are you looking at this patch ?
A+

[Saulius Krasuckas]

* On Sat, 12 Nov 2005, Eric Pouech wrote:
> - Saulius Krasuckas wrote:
> > Eric, can't both cases of 1 and DRV_SUCCESS be merged into a single 
> > case? Maybe this change address some 16-bit issues.  No?
>
> Setting everything as DRV_SUCCESS should be fine.
>
> But why are you looking at this patch ?

Thanks.  I was trying to see what happens if I set JACK audio driver via 
winecfg.  After I inserted thee TRACEs I saw they likely can be reduced 
down to two ones.

But as that is spread through all the current wine-audio-drivers, I didn't 
feel having right to change them all.  And I am not sure I will. :-)