Version-Release number of selected component: tracker-1.4.0-1.fc22 Additional info: reporter: libreport-2.5.1 backtrace_rating: 4 cmdline: /usr/libexec/tracker-extract crash_function: gst_memory_unmap executable: /usr/libexec/tracker-extract global_pid: 5863 kernel: 4.0.1-300.fc22.x86_64 runlevel: N 5 type: CCpp uid: 1000 var_log_messages: [System Logs]:\n-- Logs begin at Wed 2015-01-14 03:01:01 PST, end at Wed 2015-05-06 19:39:36 PDT. -- Truncated backtrace: Thread no. 1 (10 frames) #0 gst_memory_unmap at gstmemory.c:339 #1 gst_buffer_unmap at gstbuffer.c:1622 #2 gst_faad_set_format at gstfaad.c:326 #3 gst_audio_decoder_do_caps at gstaudiodecoder.c:866 #5 gst_audio_decoder_chain at gstaudiodecoder.c:1756 #6 gst_pad_push_data at gstpad.c:3830 #8 gst_pad_push at gstpad.c:4174 #9 gst_base_parse_push_frame at gstbaseparse.c:2304 #10 gst_base_parse_chain at gstbaseparse.c:2824 #11 gst_pad_push_data at gstpad.c:3830
Created attachment 1022876 [details] File: backtrace
Created attachment 1022877 [details] File: cgroup
Created attachment 1022878 [details] File: core_backtrace
Created attachment 1022879 [details] File: dso_list
Created attachment 1022880 [details] File: environ
Created attachment 1022881 [details] File: limits
Created attachment 1022882 [details] File: maps
Created attachment 1022883 [details] File: mountinfo
Created attachment 1022884 [details] File: namespaces
Created attachment 1022885 [details] File: open_fds
Created attachment 1022886 [details] File: proc_pid_status
This bug happened to me when tracker tried to index a bunch of mp4 audio files.
I got the same backtrace when trying to play a movie using Totem, so it's probably a problem in gstreamer.
And I've been hitting that in firefox when trying to play videos in youtube. However, the backtrace contains #2 gst_faad_set_format at gstfaad.c:326 which is from gstreamer1-plugins-bad-freeworld-1.4.3-1.fc22.x86_64 which comes from rpmfusion...
Switching from gstreamer1-plugins-bad-freeworld-1.4.3-1.fc22.x86_64 to gstreamer1-plugins-bad-freeworld-1.4.3-1.fc21.x86_64 avoids this issue (in my specific situation, removing this package is fine too).
See https://bugzilla.rpmfusion.org/show_bug.cgi?id=3645
Upstream bug report is https://bugzilla.gnome.org/show_bug.cgi?id=748571
*** Bug 1220742 has been marked as a duplicate of this bug. ***
Build problem, faacDecInit2() expects an unsigned long * for the samplerate but we pass it a guint32 *, overwriting part of the stack with 0 and messing up the nearby stack allocated GstMapInfo, causing unmap to fail.
gst-plugins-bad has trouble detecting the version of faad2 in configure: checking Checking for FAAD >= 2.0... checking Checking FAAD2 version in neaacdec.h... no idea This causes it to compile the guint32 * version.
cpp output changed causing the regexp to fail: test.c: #include <neaacdec.h> GST_CHECK_FAAD_VERSION FAAD2_VERSION f21: cpp test.c # 2 "test.c" 2 GST_CHECK_FAAD_VERSION "2.7" f22: cpp test.c # 2 "test.c" GST_CHECK_FAAD_VERSION # 2 "test.c" 3 4 "2.7"
Fixed upstream https://bugzilla.gnome.org/show_bug.cgi?id=748571
*** Bug 1221734 has been marked as a duplicate of this bug. ***
*** Bug 1222350 has been marked as a duplicate of this bug. ***
This is very old bug, if you still can reproduce it - please reopen.