Description of problem: I was deleting messages. Version-Release number of selected component: trojita-0.7-12.fc28 Additional info: reporter: libreport-2.9.5 backtrace_rating: 4 cmdline: /usr/bin/trojita crash_function: gpgme_key_unref executable: /usr/bin/trojita journald_cursor: s=c517a593197e413c8314732fdc844b35;i=1c1169;b=15a442c7dd9b486d8716b40f64216309;m=2d540faee;t=5700547cd87ce;x=9c3367ed12b9acbd kernel: 4.17.3-200.fc28.x86_64 rootdir: / runlevel: N 5 type: CCpp uid: 1000 Truncated backtrace: Thread no. 1 (6 frames) #0 gpgme_key_unref at key.c:333 #1 boost::detail::sp_counted_base::release at /usr/include/boost/smart_ptr/detail/sp_counted_base_std_atomic.hpp:110 #2 boost::detail::shared_count::~shared_count at /usr/include/boost/smart_ptr/detail/shared_count.hpp:426 #3 boost::shared_ptr<_gpgme_key>::~shared_ptr at /usr/include/boost/smart_ptr/shared_ptr.hpp:341 #4 GpgME::Key::~Key at /usr/include/KF5/gpgme++/key.h:55 #5 Cryptography::GpgMePart::extractSignatureStatus(std::shared_ptr<GpgME::Context>, GpgME::Signature const&, std::vector<std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> >, std::allocator<std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> > > >, bool, bool, bool&, bool&, bool&, QString&, QString&, QString&, QString&, QDateTime&) at /usr/src/debug/trojita-0.7-12.fc28.x86_64/src/Cryptography/GpgMe++.cpp:324
Created attachment 1455970 [details] File: backtrace
Created attachment 1455971 [details] File: cgroup
Created attachment 1455972 [details] File: core_backtrace
Created attachment 1455973 [details] File: cpuinfo
Created attachment 1455974 [details] File: dso_list
Created attachment 1455975 [details] File: environ
Created attachment 1455976 [details] File: exploitable
Created attachment 1455977 [details] File: limits
Created attachment 1455978 [details] File: maps
Created attachment 1455979 [details] File: mountinfo
Created attachment 1455980 [details] File: open_fds
Created attachment 1455981 [details] File: proc_pid_status
Reassigning according to provided backtrace.
I'm not so into C++, but this seems like issue in trojita which is not holding some reference long enough which causes crash.
(Trojita upstream here.) We're using the C++ bindings for gpgme. That thing is expected to provide a refcounted copy of everything to the callers. I tried to track this down a long time ago, but even with ASAN and stuff, I just wasn't able to pinpoint the real cause. Sorry, I cannot reproduce that on my system :(.
Jan, thanks for your statement. Then, I tend to close here with the resolution WORKSFORME. But let's wait at least for the EOL notification of the F28 branch that should arrive shortly with upcoming freeze for F29. @Reporter Is this issue still reproducible for you?
No reply any more and so far.
(In reply to Raphael Groner from comment #16) > @Reporter Is this issue still reproducible for you? I don't remember seeing _any_ trojitá crash for a few months (since upgrade to F29?)