Path: utzoo!attcan!uunet!lll-winken!csd4.milw.wisc.edu!leah!rpi!rpi.edu!hiebeler From: hiebeler@turing.cs.rpi.edu (Dave Hiebeler) Newsgroups: comp.windows.x Subject: fix2 patch to window.c destroys backing_store Message-ID: Date: 13 Apr 89 18:26:10 GMT Sender: usenet@rpi.edu Distribution: na Organization: RPI CS Dept Lines: 36 Well, after asking about backing_store, and playing around with it on various machines, I have finally found this out: in the "fix2" patches, the following 2 lines are added to server/dix/window.c: if (!pWin->firstChild) return; This is supposed to fix "core dump on no-op UnmapSubwindows"; these 2 lines are located at the beginning of UnmapSubwindows(). Now, I have tested my program which uses backing_store on a Sun-3/50, using a version of the X11R3 server with those 2 lines, and without them. With them, backing_store does NOT work. Without them, it DOES work fine. 2 questions: - Under what circumstances does this core dump occurr? I don't know what a "no-op UnmapSubwindows" is. I.e. is it "dangerous" to run the server without that patch? I've never gotten a core-dump while running a server without the patch. - Is this patch really supposed to disable backing_store? I am not extremely familiar with X, but those 2 lines seem to say "if you're not the first child, don't do anything", not giving the server a chance to do backing_store. The window I'm trying to use backing_store on is a box widget, which is the child of a viewport widget, which is the child of a top-level shell widget. This seems to be hitting that return. Again, much thanks for any help... -- Dave Hiebeler Internet: hiebeler@cs.rpi.edu (preferred address) Computer Science Dept. hiebeler@itsgw.rpi.edu Amos Eaton Bldg. Bitnet: userfrzk@rpitsmts.bitnet Rensselaer Polytechnic Institute / Troy, NY 12180-3590