Where Is My Backtrace?
Submitted by axel on Fri, 18/02/2011 - 09:33
In the past I got a nice stacktrace showing what called what, and that I was supposed to send the crash to myself - and now?
The temporary fix is
Hi!
I noticed that lately, on Ubuntu 10.10, I dont get any! backtrace! anymore! Nothing! I ran e.g.
root [0] TObject *o = (TObject*)8764586 root [1] o->GetName() *** Break *** segmentation violation Root > .q
Ubuntu's kernel hardening efforts are causing that:
$ cat /etc/sysctl.d/10-ptrace.conf [...] kernel.yama.ptrace_scope = 1
sudo bash -c "echo 0 > /proc/sys/kernel/yama/ptrace_scope", the proper one is to switch the flag in /etc/sysctl.d/10-ptrace.conf to 0.
Et voilà:
=========================================================== There was a crash (kSigSegmentationViolation). This is the entire stack trace of all threads: =========================================================== #0 0x00007fd0b6ef8dee in waitpid () from /lib/libc.so.6 #1 0x00007fd0b6e90a79 in ?? () from /lib/libc.so.6 #2 0x00007fd0b93ef9f0 in TUnixSystem::Exec (this=0x16dc9b0, shellcmd=0x1c024d8 "/mnt/ram/t/etc/gdb-backtrace.sh 32333 1>&2") at /build/root/srcram/core/unix/src/TUnixSystem.cxx:2010 #3 0x00007fd0b93f0387 in TUnixSystem::StackTrace (this=0x16dc9b0) at /build/root/srcram/core/unix/src/TUnixSystem.cxx:2237 #4 0x00007fd0b93edb23 in TUnixSystem::DispatchSignals (this=0x16dc9b0, sig=kSigSegmentationViolation) at /build/root/srcram/core/unix/src/TUnixSystem.cxx:1136 #5 0x00007fd0b93eb7d9 in SigHandler (sig=kSigSegmentationViolation) at /build/root/srcram/core/unix/src/TUnixSystem.cxx:357 #6 0x00007fd0b93f3aea in sighandler (sig=11) at /build/root/srcram/core/unix/src/TUnixSystem.cxx:3524 #7 <signal handler called> #8 0x00007fd0b9695aec in G__G__Base2_10_0_26 (result7=0x7fff593e15b0, funcname=0x18535e0 "", libp=0x7fff593d6b80, hash=0) at core/base/src/G__Base2.cxx:6687 [...]
Welcome back, dear stack frames!
Cheers,
