| Summary: | calling XScreenSaverQueryInfo crashes the server | ||||||
|---|---|---|---|---|---|---|---|
| Product: | xorg | Reporter: | Nicholas Miell <nmiell> | ||||
| Component: | Server/General | Assignee: | Xorg Project Team <xorg-team> | ||||
| Status: | RESOLVED FIXED | QA Contact: | |||||
| Severity: | normal | ||||||
| Priority: | high | CC: | mharris, roland.mainz | ||||
| Version: | 6.8.2 | ||||||
| Hardware: | x86 (IA32) | ||||||
| OS: | Linux (All) | ||||||
| Whiteboard: | |||||||
| i915 platform: | i915 features: | ||||||
| Attachments: |
|
||||||
A simple test program that does nothing XSS-related besides call XScreenSaverQueryExtension, XScreenSaverAllocInfo and XScreenSaverQueryInfo does not crash the server. beforelight does, everytime. (In reply to comment #1) > A simple test program that does nothing XSS-related besides call > XScreenSaverQueryExtension, XScreenSaverAllocInfo and XScreenSaverQueryInfo does > not crash the server. > > beforelight does, everytime. Does beforelight (I assume this is a window manager, right ?) work when you turn the screensaver extension off in the Xserver (you can do that at command line via % Xorg -extension MIT-SCREEN-SAVER #) ? beforelight is, as far as I know, the canonical example of a screensaver that uses the MIT-SCREEN-SAVER extension. It comes with X. Disabling the extension prevents beforelight from ever using it, and thus the X server does not crash. Turns out that the bug was in XScreenSaverSetAttributes, and was fixed here: http://cvs.freedesktop.org/xorg/xc/programs/Xserver/Xext/saver.c?r1=1.2&r2=1.3&only_with_tag=MAIN Needs to be merged into 6.8.2, though. Info for Fedora Core users experiencing this problem: Patch added to 6.8.2-11 which will be in all future FC3/FC4/RHEL4 builds. * Tue Mar 15 2005 Mike A. Harris <mharris@redhat.com> - Added xorg-x11-6.8.2-config-StaticNeedsPicForShared.patch for (#108026) - Removed StaticNeedsPicForShared from host.def section - Added xorg-x11-6.8.2-XScreenSaverQueryInfo-crash-fix.patch to fix (#147890) Created attachment 3526 [details] [review] Patch to saver.c as applied to CVS head Since this is fixed in CVS head, and a patch for it is in the 6.8.3 request queue, closing bug as FIXED. |
Use of freedesktop.org services, including Bugzilla, is subject to our Code of Conduct. How we collect and use information is described in our Privacy Policy.
backtrace without debugging symbols is as follows: #0 0x000000301972e4dd in *__GI_raise (sig=6700) at ../nptl/sysdeps/unix/sysv/linux/raise.c:67 #1 0x000000301972fc8e in *__GI_abort () at ../sysdeps/generic/abort.c:88 #2 0x0000003019762b91 in __libc_message (do_abort=2, fmt=0x3019810850 "*** glibc detected *** %s: 0x%s ***\n") at ../sysdeps/unix/sysv/linux/libc_fatal.c:145 #3 0x00000030197684cf in _int_free (av=0x301992e620, mem=0x1a2c) at malloc.c:5523 #4 0x0000003019768a76 in __libc_free (mem=0x6) at malloc.c:3404 #5 0x00000000005179c6 in ProcScreenSaverQueryInfo () #6 0x0000000000470f9d in Dispatch () #7 0x0000000000456de5 in main () This is with 6.8.1.904. 903 and, iirc, 902 did it also. I don't know about previous versions.