Bug 2231983

Description of problem:
Starting a container via the socket causes all socket calls to hang after that. This happens specifically with systemd 252-16 and podman 4.6.0-1

Version-Release number of selected component (if applicable):


How reproducible:


Steps to Reproduce:

The environment is a CentOS Stream 9 VM (e.g. [qcow2](https://cloud.centos.org/centos/9-stream/x86_64/images/CentOS-Stream-GenericCloud-9-20220829.0.x86_64.qcow2)). On a non-root user.

1. Install specifically this version (the version 4.6.0-3 has other issue https://bugzilla.redhat.com/show_bug.cgi?id=2231975): `sudo dnf install podman-2:4.6.0-1.el9`
2. Ensure systemd 252-15 is the one installed: `sudo dnf install systemd-252-15.el9`
3. Enable socket `systemctl --user enable --now podman.socket`
4. Start nginx `podman --url unix://run/user/$(id -u)/podman/podman.sock run --name nginx-test -p 8080:80 -d docker.io/nginx`
5. List `podman --url unix://run/user/$(id -u)/podman/podman.sock ps`. Run it multiple times. They all work.
6. Remove test container: `podman container rm -f nginx-test`
7. Update systemd to 252-16: `sudo dnf install systemd-252-16.el9`
8. Reboot
9. Start nginx `podman --url unix://run/user/$(id -u)/podman/podman.sock run --name nginx-test -p 8080:80 -d docker.io/nginx`
10. List `podman --url unix://run/user/$(id -u)/podman/podman.sock ps`. This one hangs.

Sometimes, the step 10 has to be run multiple times before it hangs for good.

What I'm seeing is that the processes started in step 9 are included in the `podman.service` cgroup, and even though podman itself already exited, the `podman.service` appears as "active (running)", which seems to prevent systemd from starting podman to attend the subsequent socket requests.

```
podman.service - Podman API Service
     Loaded: loaded (/usr/lib/systemd/user/podman.service; disabled; preset: disabled)
     Active: active (running) since Mon 2023-08-14 18:03:31 EDT; 9min ago
TriggeredBy: ● podman.socket
       Docs: man:podman-system-service(1)
    Process: 1516 ExecStart=/usr/bin/podman $LOGGING system service (code=exited, status=0/SUCCESS)
   Main PID: 1516 (code=exited, status=0/SUCCESS)
      Tasks: 13 (limit: 10774)
     Memory: 30.2M
        CPU: 75ms
     CGroup: /user.slice/user-1000.slice/user/app.slice/podman.service
             ├─1526 /usr/bin/slirp4netns --disable-host-loopback --mtu=65520 --enable-sandbox --enable-seccomp --enable-ipv6 -c -r 3 -e 4 --netns-type=path /run/user/1000/n>
             ├─1529 rootlessport
             └─1538 rootlessport-child

```

Actual results:
It hangs when listing containers using the socket

Expected results:
It lists containers.

Additional info:
I submitted the ticket upstream as well: https://github.com/containers/podman/issues/19625