[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [MiNT] XaAES video init implementation (Falcon)



Send a patch. I'll get it committed.

Alan.

On 03/29/13 12:17, Peter Persson wrote:
Anyone? Basically I think the correct thing would be to remove the hard-coded modecode filtering completely, since that's not a job for the AES imo. Any input on this, or can I just create a patch which someone hopefully will commit?


2013/3/22 Peter Persson <pep.fishmoose@gmail.com <mailto:pep.fishmoose@gmail.com>>

    Hi,

    I found the following code in k_init.c

    nvmode = vcheckmode(cfg.videomode);
    if ((nvmode & (1 << 4)) &&/* VGA_FALCON */
        (nvmode & 7) == 4)/* is 16bit */
    {
    nvmode &= ~(1 << 3);/* Set 320 pixels */
    }

    So first it uses vcheckmode() to filter the modecode to something
    which works on the current monitor (isn't this undocumented, btw?
    I see nothing wrong with this though, it is after all a useful
    function for the AES). It then makes assumption as to what modes
    are actually valid. I question this; is this really something that
    the AES should interfere with?

    Basically, what happens on the SuperVidel is that the resolution
    has to be set as a fake-RGB resolution (VGA bit not set),
    otherwise XaAES enforces doubled pixels horizontally. The
    alternative is to not set the mode using xaaes.cnf, but fall back
    to a default mode specified in the SuperVidel configuration file
    (this requires NVDI however).

    Thoughts?

    -- PeP