This isn't surprising. Backingstore doesn't guarantee that you won't get expose events. It merely attempts to reduce them by saving occluded bits. When the server switches away to another VT everything in the framebuffer is lost. Backing store allowed it to recover the occluded bits that it saved earlier. I believe this is the expected behavior. People didn't have VT switches back when backing store was designed. Mark. On Wed, 9 Jun 2004 davidek@xxxxxxxxxxxxxxxxxxxxxx wrote: > Hello, > I noticed a problem maybe related to backingstore option with the > release 4.4.0. My XF86Config-4 looks like: > Section "Device" > Identifier "Generic Video Card" > Driver "i810" > # BusID "PCI:0:2:0" > VideoRam 65536 > Option "BackingStore" "true" > Option "DisplayInfo" "FALSE" > EndSection > > I have the "famous" i855GM chip, so I had to introduce the displayinfo > option as already mentioned in several posts in this list. > > The backingstore itself works well, i.e. if a graphical window hides below > some other one it still keeps its content, but when switching back to > the text console (Ctrl-Alt-F1) and back, the content is lost. What's funny > is that the content is actually preserved in the part of the window which > was overlapped, the rest (that was on the top when switching there and > back) gets empty. > > Maybe the option DisplayInfo "false" has some influence ? I checked with > xdpyinfo and I got as usual: > options: backing-store YES, save-unders YES > > Thanks a lot for any hint, > > best regards > Tomas > > E-mail : davidek@xxxxxxxxxxxxxxxxxxxxxx, > Tomas.Davidek@xxxxxxx > > _______________________________________________ > XFree86 mailing list > XFree86@xxxxxxxxxxx > http://XFree86.Org/mailman/listinfo/xfree86 > _______________________________________________ XFree86 mailing list XFree86@xxxxxxxxxxx http://XFree86.Org/mailman/listinfo/xfree86
