| Summary: | gnome-control-center crashes | ||
|---|---|---|---|
| Product: | Mesa | Reporter: | Alen Skondro <askondro> |
| Component: | Mesa core | Assignee: | mesa-dev |
| Status: | RESOLVED FIXED | QA Contact: | |
| Severity: | normal | ||
| Priority: | high | CC: | artafinde, biru.ionut, coder, czarkoff, gm5729, nobled, obi12341, pascal.schott, raphael.rochet, weegee |
| Version: | 8.0 | ||
| Hardware: | x86-64 (AMD64) | ||
| OS: | Linux (All) | ||
| Whiteboard: | |||
| i915 platform: | i915 features: | ||
| Attachments: |
gdb backtrace
Host.inc from llvm-3.0 GDB output valgrind log |
||
|
Description
Alen Skondro
2012-04-08 10:31:47 UTC
We have users that told us that is not related only to nouveau. Ati is involved as well and the main developer of clutter said that it might be a bug in llvmpipe or its usage. llvm::sys::getHostTriple() is kind of a weird place to crash. What version of LLVM is it, and can you build it with debug symbols before linking mesa against it? Which line in LLVM's source is calling std::string::append()? Created attachment 60489 [details]
Host.inc from llvm-3.0
It's version 3.0. I already built a debug version of LLVM and linked mesa against it but could not really see what exactly the error is since some important variables were optimized and I couldn't read their values. But the string::append call (via overloaded operator I guess) happens in llvm/lib/Support/Unix/Host.inc, line 50. You can compile llvm and mesa with -g -O0 Created attachment 60495 [details]
GDB output
This is what I got so far.
Even though I compiled with -g -O0 parameters, there are still variables marked as "optimized out".
*** Bug 49083 has been marked as a duplicate of this bug. *** We have couples of valgrid logs https://bugs.archlinux.org/task/29202?getfile=8642 FWIW, the environment variable DRAW_USE_LLVM=0 should work around the crash at a potential performance penalty. Stange enough, I experience this crash with ati driver. And there's no crash if using clutter-gtk-1.0.4-1. That's not a fix of course ... just wanted to note it here in case it matters. (In reply to comment #9) > FWIW, the environment variable DRAW_USE_LLVM=0 should work around the crash at > a potential performance penalty. Tried to run "DRAW_USE_LLVM=0 gnome-control-center" but still crashing. try MALLOC_CHECK_=1 gnome-control-center (In reply to comment #12) > try MALLOC_CHECK_=1 gnome-control-center Damned. Worked, got 10+ errors like this : *** glibc detected *** gnome-control-center: free(): invalid pointer: 0x00007f36260d74d8 *** Created attachment 60515 [details]
valgrind log
Same behaviour as #13 here, using radeon driver on 9600 card. using nouveau-dri 8.0.2-1 xf86-video-nouveau 0.0.16_git20120210-1 on archlinux 32 bits DRAW_USE_LLVM=0 gnome-control-center crashes too but MALLOC_CHECK_=1 gnome-control-center works *** Bug 49212 has been marked as a duplicate of this bug. *** I 've been using Arch Linux and I had this issue when i moved to Gnome 3.4. with control center at 3.4.1.1. Unfortunately I had to reinstall my system Yesterday and Now gnome-control-center works fine without a glitch. My original Arch linux post. https://bbs.archlinux.org/viewtopic.php?id=140131 It is fixed for me in last update. ( ArchLinux 64bits / gnome-control-center 3.4.1-2 ) (In reply to comment #19) > It is fixed for me in last update. > ( ArchLinux 64bits / gnome-control-center 3.4.1-2 ) 32bits also worked fine. It wasn't fixed, it was worked around: https://plus.google.com/104877287288155269055/posts/grHDgRNc1gi *** Bug 49039 has been marked as a duplicate of this bug. *** |
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.