I'm still looking into this. If you happen to get another core in the meantime:
1) you don't need to do the backtrace anymore. I can read the stacks just fine when I don't have incorrect paths in my gdb debugging setup.
2) Check the logs with journalctl -xe, and see if there are any suspicious messages related to vs_exe before the core dump message. Probably a long shot, but I wanted to cover that base just in case.
Sorry for the delay. Usually, even if the coredump doesn't show me exactly what happened, it lets me narrow things down into a range I can probe for problems with valgrind or other tools. But there are some things I'm still figuring out with these dumps, so the search area is still a little large.