Jens Owen wrote: > > Keith Whitwell wrote: > > > > > > The private-backbuffer case really has > > > > to be the backbuffer as treating the frontbuffer as private would break the > > > > blitting technique that we use to keep the (real) backbuffer uptodate... To > > > > have a private backbuffer & ignore cliprects, it really does have to be the > > > > backbuffer. > > > > > > I thought we only supported private back buffers when we had a single 3D > > > window, so wouldn't back buffer functionality be excluded when we > > > support page flipping? > > > > Yes, but we mentioned early on that when there are many cliprects it might be > > better to revert to swapping and a private backbuffer... > > Doh! Of course. > > Is the blitting technique flexible...i.e. can the new backbuffer become > the new "real" backbuffer even though it may have started out as the > front buffer?
Not at present. The normal DDX drawing operations are unchanged by this patch - I just track the dirty regions & blit to the backbuffer. > I think I may have already seen a bug related to this. I ran tuxracer, > but the desktop failed to redraw after I exited. xrefresh didn't fix > it, and running again didn't reproduce it. Maybe I should try a few > more times. If the server is confused about which is back and which is > front, I should be able to reproduce at least 50% of the time :-) The X server notices if you exit on the wrong side & should keep on blitting. There is code there for it to fix it, but it seems we aren't holding the lock or something when it is activated. Keith _______________________________________________________________ Don't miss the 2002 Sprint PCS Application Developer's Conference August 25-28 in Las Vegas -- http://devcon.sprintpcs.com/adp/index.cfm _______________________________________________ Dri-devel mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/dri-devel
