agent(rhel8): skip tests in nspawn on RHEL 8.6 and older

As we're missing [0] in these versions which is not trivial to backport
into such old systemd versions, and it breaks the nspawn runs with:

Spawning container TEST-01-BASIC--1 on /var/tmp/systemd-test.0kzKzL/nspawn-root.
Press ^] three times within 1s to kill container.
systemd 239 running in system mode. (+PAM +AUDIT +SELINUX +IMA -APPARMOR +SMACK +SYSVINIT +UTMP +LIBCRYPTSETUP +GCRYPT +GNUTLS +ACL +XZ +LZ4 +SECCOMP
+BLKID +ELFUTILS +KMOD +IDN2 -IDN +PCRE2 default-hierarchy=legacy)
Detected virtualization systemd-nspawn.
Detected architecture x86-64.

Welcome to CentOS Stream 9!

Set hostname to <systemd-testsuite>.
Initializing machine ID from container UUID.
Failed to add address 127.0.0.1 to loopback interface: Operation not permitted
Failed to add address ::1 to loopback interface: Operation not permitted
Failed to bring loopback interface up: Operation not permitted
Found cgroup on /sys/fs/cgroup/systemd, legacy hierarchy
Using cgroup controller name=systemd. File system hierarchy is at /sys/fs/cgroup/systemd.
Failed to install release agent, ignoring: No such file or directory
Failed to create /init.scope control group: Operation not permitted
Failed to allocate manager object: Operation not permitted
[!!!!!!] Failed to allocate manager object, freezing.
Freezing execution.
Trying to halt container. Send SIGTERM again to trigger immediate termination.
Container TEST-01-BASIC--1 terminated by signal KILL.

[0] redhat-plumbers/systemd-rhel8#286

Author: mrc0mmand

agent/testsuite-rhel8.sh

@@ -155,6 +155,19 @@ for t in test/TEST-??-*; do
     # about them, so let's tell it to whitelist all known syscalls
     export NSPAWN_ARGUMENTS="--system-call-filter=@known"
 
+    if grep 'seccomp_init_for_arch(.*SCMP_ACT_ERRNO(EPERM)' src/nspawn/nspawn-seccomp.c; then
+        # In very old RHEL 8 systemd versions (8.6.0 and older) we're missing patch series [0] that configure
+        # nspawn's seccomp filters to return ENOSYS instead of EPERM for unknown syscalls. And given we have
+        # to run the CI for RHEL 8 systemd on C9S, we get a bunch of unknown syscalls when running nspawn which
+        # breaks the nspawn runs (the workaround above with $NSPAWN_ARGUMENTS doesn't help here).
+        #
+        # Given the series we're missing is not trivial to backport, especially into such old systemd versions,
+        # let's just skip the nspawn runs there to have at least some test coverage.
+        #
+        # [0] https://github.com/redhat-plumbers/systemd-rhel8/pull/286
+        export TEST_NO_NSPAWN=1
+    fi
+
     # Suffix the $TESTDIR of each retry with an index to tell them apart
     export MANGLE_TESTDIR=1
     # FIXME: retry each task again if it fails (i.e. run each task twice at most)