In article <p05101006b81de195aaac@[192.168.0.20]> [EMAIL PROTECTED]
writes:
>>    I'm assuming they're not using Xlib to render and this is root-mode
>>DGA where they are touching the framebuffer.  I've attached one of
>>my original DGA 2.0 test apps.  If it works fine for you then it's
>>probably the case that there's nothing wrong with the "mga" driver's
>>DGA support.  This app will pretty much test every mode advertised.
>>
>>                              Mark.
>
>Ok, your test program has some pretty odd behavior on my machine. 

Try not using the mga_hal_drv.o module from Matrox (assuming you're using
it).  Every version of it I've been able to try breaks DGA2 nicely on my
G400.  In single headed mode, the driver still offers all of the normal DGA2
modes, but no matter which the client requests, it always switches to the
same width, height, and pitch as the root window without any feedback to the
client that its DGA request was only half honoured.  In dualhead mode, it
also silently throws away all accelerated DGA fills, even though the flags
for both the requested and acutally used DGA2 mode indicate support.  I
poked around Matrox's web site for a while trying to find a useful way to
report my problems and offer a simple test program, but eventually gave up.

Yours,
Jeremy.
_______________________________________________
Xpert mailing list
[EMAIL PROTECTED]
http://XFree86.Org/mailman/listinfo/xpert

Reply via email to