Summary: | xcb backend: XNextEvent() segfault when XlibDisplayIOError is true | ||
---|---|---|---|
Product: | xorg | Reporter: | Fumihito YOSHIDA <fumihito.yoshida> |
Component: | Lib/Xlib | Assignee: | Jamey Sharp <jamey> |
Status: | RESOLVED NOTOURBUG | QA Contact: | Xorg Project Team <xorg-team> |
Severity: | critical | ||
Priority: | high | CC: | ThJaeger |
Version: | git | ||
Hardware: | All | ||
OS: | All | ||
Whiteboard: | 2011BRB_Reviewed | ||
i915 platform: | i915 features: |
Description
Fumihito YOSHIDA
2008-10-06 04:08:24 UTC
Does your application use XTest? If so, can you check if this patch fixes the issue? http://cgit.freedesktop.org/xorg/xserver/commit/?id=b2756a71a432f7cf7c870a48676c98625512558d This is a bit old. Is it still an issue? I'm not sure how this bug ever could have triggered. It requires that you enter XNextEvent after _XIOError was called, but Xlib calls exit() on IO error. So either you longjmp'd out of the IO error handler, in which case you're not allowed to touch the Display further, or you're trying to clean up the Display in an atexit handler, where it seems like a bad plan to process events. I'm also not seeing anything sane that Xlib can do if this happens. The caller has done something wrong. We could replace the check with an assertion, since the program is going to crash one way or another. I'd happily take a patch that replaces the early-return with an assert, but barring that or evidence that this is actually a libX11 bug, I'm closing this. |
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.