Version-Release number of selected component: PackageKit-0.8.13-1.fc21 Additional info: reporter: libreport-2.1.9 backtrace_rating: 4 cmdline: /usr/libexec/packagekitd crash_function: g_string_chunk_free executable: /usr/libexec/packagekitd kernel: 3.12.1-2.fc21.x86_64 runlevel: N 5 type: CCpp uid: 0 Truncated backtrace: Thread no. 1 (10 frames) #6 g_string_chunk_free at gstringchunk.c:154 #7 lr_yum_repomdrecord_free at /usr/src/debug/librepo/py3/librepo/repomd.c:52 #8 g_slist_foreach at gslist.c:896 #9 g_slist_free_full at gslist.c:179 #10 lr_yum_repomd_free at /usr/src/debug/librepo/py3/librepo/repomd.c:69 #11 lr_result_clear at /usr/src/debug/librepo/py3/librepo/result.c:45 #12 lr_result_free at /usr/src/debug/librepo/py3/librepo/result.c:55 #13 hif_source_free at hif-source.c:65 #14 g_ptr_array_foreach at garray.c:1460 #15 ptr_array_free at garray.c:1085
Created attachment 828879 [details] File: backtrace
Created attachment 828880 [details] File: cgroup
Created attachment 828881 [details] File: core_backtrace
Created attachment 828882 [details] File: dso_list
Created attachment 828883 [details] File: environ
Created attachment 828884 [details] File: limits
Created attachment 828885 [details] File: maps
Created attachment 828886 [details] File: open_fds
Created attachment 828887 [details] File: proc_pid_status
Created attachment 828888 [details] File: var_log_messages
Reassigning to see if this is interesting.
Cannot reproduce, Valgrind shows everything is ok. I've inspect the code for possibility of double free, but found nothing. SIGABRT during glibc free() could be heap corruption as well, but as I wrote above, Valrgind doesn't show any error and all my tests pass. For this moment no fix is available. Please let me know if this bug appear again.
Cannot reproduce with the current (1.7.10) version. Feel free to re-open if reappears.