-
Notifications
You must be signed in to change notification settings - Fork 2.5k
Description
With the various gstreamer targets, we have setup a custom message handler to raise abort() on critical messages (which are essentially codepaths that shouldn't be reached).
The problem is that we end up with stack traces like this:
#0 0x7f15a7835427 in gsignal /build/glibc-9tT8Do/glibc-2.23/sysdeps/unix/sysv/linux/raise.c:54
#1 0x7f15a7837029 in abort /build/glibc-9tT8Do/glibc-2.23/stdlib/abort.c:89
#2 0x42fa49 in custom_logger /src/gst-ci/fuzzing/gst-discoverer.c:79:5
#3 0x7264cc in g_logv /work/glib-2.54.2/glib/gmessages.c:1341:11
#4 0x726117 in g_log /work/glib-2.54.2/glib/gmessages.c:1403:3
#5 0x5a0a50 in What actually triggered the critical
...
It seems only the 3 functions before abort() are the ones being used to detect identical issues (which will always be the same), whereas we would want to add at least a function or two before that to detect similarities.
Is there a way to customize the depth for similar testcase detection ?