I finally tracked down the last remaining major bug in the mach64 Mesa 4.x
branch.  The secondary texture filter wasn't being set in the correct
register (was being set in scale_3d_cntl not tex_cntl), so it ended up
enabling fog instead, which disables alpha blending!  This was actually in
the last branch too, so I'm not sure why this hadn't cropped up before
(maybe I changed my quake settings and didn't remeber?).  At any rate, the
branch should now be at about the same state the previous branch was in.  
There are still some bugs remaining, and I plan on running glean and maybe
some other test suites on the new branch as well.  This would be a good
time for testers to do a fresh checkout and look for bugs.  When the trunk 
has stabilized, we can merge in any Mesa bugfixes too.

So I think we can try to focus on the DMA work now.  Frank, can you check
in whatever you've got into the -dma branch?  It doesn't matter if it's
not done, I think it would be helpful for us to see where we stand now.
One thing I want to try is using NOTEX or TINY vertices as well as
textured verticies.  It seems that mach64 ought to be able to handle
those since spec_argb, z, argb, and x_y are sequential registers, and it 
could pump up our gears numbers a bit. ;)  I'll see if I can get it to 
work with MMIO and if so, we could try those vertex formats with DMA as 
well.

-- 
Leif Delgass 
http://www.retinalburn.net


_______________________________________________
Dri-devel mailing list
[EMAIL PROTECTED]
https://lists.sourceforge.net/lists/listinfo/dri-devel

Reply via email to