Comment 6 for bug 1935617

As root:

$ mkdir /tmp/debugtest
$ export ARTIFACT_DIRECTORY=/tmp/debugtest
$ export TEST_SHOW_JOURNAL=warning
$ export TEST_REQUIRE_INSTALL_TESTS=0
$ export TEST_PREFER_NSPAWN=1
$ export NO_BUILD=1
$ export SELECTED_TESTS="TEST-08-*"
./test/run-integration-tests.sh

Reproduces the issue, it will keep the journal in /tmp/debugtest and run in emulation.

The crash looks like:
[ 15.988730] kernel: ext4 filesystem being mounted at /root supports timestamps until 2038 (0x7fffffff)
[ 18.235447] systemd[1]: Failed to find module 'autofs4'
[ 21.938779] kernel: ext4 filesystem being remounted at / supports timestamps until 2038 (0x7fffffff)
[ 25.404984] systemd-coredump[371]: [🡕] Process 369 (rm) of user 0 dumped core.

                                      Stack trace of thread 369:
                                      #0 0x00000b1dcc8c61d0 n/a (rm + 0x61d0)
                                      #1 0x00000b1dcc8c8f64 n/a (rm + 0x8f64)
                                      #2 0x00000b1dcc8c3538 n/a (rm + 0x3538)
                                      #3 0x00000b1dcc8c4424 n/a (rm + 0x4424)
                                      #4 0x00000b1dcc8c2b70 n/a (rm + 0x2b70)
                                      #5 0x00007ee0760aa0b4 n/a (libc.so.6 + 0x2a0b4)
                                      #6 0x00007ee0760aa2a0 __libc_start_main (libc.so.6 + 0x2a2a0)
[ 25.480260] systemd[1]: testsuite-08.service: Control process exited, code=dumped, status=11/SEGV

It seems always "rm" that breaks, here from a fill log:

$ grep -e "of user 0 dumped core" systemd-ppc-qemu6.0.txt
[ 47.691335] systemd-coredump[286]: Process 279 (rm) of user 0 dumped core.
[ 44.911447] systemd-coredump[284]: Process 279 (rm) of user 0 dumped core.
[ 45.089287] systemd-coredump[284]: Process 282 (rm) of user 0 dumped core.
[ 44.799519] systemd-coredump[284]: Process 278 (rm) of user 0 dumped core.
[ 44.666997] systemd-coredump[285]: Process 283 (rm) of user 0 dumped core.
[ 46.306449] systemd-coredump[282]: Process 279 (rm) of user 0 dumped core.
[ 43.898582] systemd-coredump[283]: Process 280 (rm) of user 0 dumped core.
[ 53.298322] systemd-coredump[338]: Process 335 (rm) of user 0 dumped core.
[ 45.731700] systemd-coredump[284]: Process 282 (rm) of user 0 dumped core.
[ 45.354492] systemd-coredump[283]: Process 281 (rm) of user 0 dumped core.
[ 43.579587] systemd-coredump[281]: Process 278 (rm) of user 0 dumped core.
[ 45.047475] systemd-coredump[284]: Process 278 (rm) of user 0 dumped core.
[ 50.662839] systemd-coredump[281]: Process 278 (rm) of user 0 dumped core.
[ 45.634806] systemd-coredump[283]: Process 278 (rm) of user 0 dumped core.

It is interesting to see such a crash as all other services report a valid green "ok".

If I start the very same guest, but without the pre-defined systemd.wants=end.service (as it is meant to shut it down again) I can log in as root without a PW.

$ /bin/qemu-system-ppc64 -smp 4 -net none -m 512M -nographic -vga none -kernel /boot/vmlinux-5.11.0-22-generic -drive format=raw,cache=unsafe,file=/var/tmp/systemd-test.zx7kOu/test08.img -initrd /boot/initrd.img-5.11.0-22-generic -append ' root=/dev/sda1 rw raid=noautodetect rd.luks=0 loglevel=2 init=/lib/systemd/systemd console=hvc0 selinux=0 SYSTEMD_UNIT_PATH=/usr/lib/systemd/tests/testdata/testsuite-08.units:/usr/lib/systemd/tests/testdata/units: systemd.unit=testsuite.target systemd.wants=testsuite-08.service

In this boot mode the console is far from perfect, but I can find testsuite-08.service to be failed on start with that SEGFAULT ... trying to isolate the failing command.