Re: [MiNT] Mintlib: VcheckMode() and Srealloc()

[Paul Wratt <paul.wratt@gmail.com>]

On Sun, Dec 13, 2009 at 5:39 AM, Peter Persson <pep.fishmoose@gmail.com> wrote:
>
> On 12 dec 2009, at 19.24, Paul Wratt wrote:
>> first, have you tried ALL the doc source, it may also be undocumented
>> somewhere, as it were
>
> I've searched the relevant headers in mintlib (/mint/falcon.h and /mint/osbind.h).
>
>> To add it, start with a call you do know, and reference that, you can
>> always do a little test prog to check it is functioning
>
> Thanks, I know how it works.
>
>> If you can find them used, then you can "devine" what they pass, and
>> what they return, which should help when referencing the known call,
>> and creating the entry..
>
> WORD VcheckMode(WORD modecode);
>  - it takes a modecode as input parameter, parses it, and returns a suitable modecode based on the current monitor type.
>
>> I think Ozk tried a couple of different VcheckMode() variations when
>> trying for Milan S3, RageII, and Falcon compatible rez changing
>
> Sounds weird. VcheckMode() is a TOS call. It's used in the k_init.c, and by the TOS VDI (I think I wrote desktop earlier, I think it's in w_opnwk() though) prior to setting the screen resolution.
> The call is useful, and I already know how it works. Since we're using it in XaAES too, I think it's right to put it in mintlib as well. The right place for this would be falcon.h, and it's not there.
>
> Srealloc() is a bit tricky, because I get the impression (based on Vsetscreen() behavior) that it re-allocates memory. But re-allocation could also be placed in the Vsetscreen() function. Again, it's not documented, but TOS uses it.
>
> -- PeP

It has just occured to me that you might have been asking "how do I
get it in CVS".. someone will have to review it, and I think, under
the circumstances, Johan should be consulted on any VDI stuff, if for
no other reason than the knowledge he has, there may be a reason it is
undocumented. Srealloc() may also have "unreliable results" if it too
is undocumented. If these are however Falcon only (Vcheckmode is not,
but may be dependant on driver also), then there should not be any
issues adding them immediately, as long as CTxx hardware can produce
consistant results, and again, other graphics card for Falcon..

maybe simple post the additions as you have them in the headers, and
possibly some test code so others can do some quick tests, that should
also allow any review to happen immediately.. (is there someone who
handles this area?)

Paul