Crash oomd on startup if `--cgroup-fs` path is not cgroup2

[danobi]

Letting oomd be silently run against a cgroup1 hierarchy could lead to subtle issues down the line.

[danobi]

Reported in #61

[hakavlad]

@danobi

You can find actual cgroup2 root location in /proc/mounts:

$ cat /proc/mounts | grep cgroup2
cgroup /sys/fs/cgroup/unified cgroup2 rw,nosuid,nodev,noexec,relatime,nsdelegate 0 0

Perhaps it is worth taking the default cgroup2 root from /proc/mounts instead of using /sys/fs/cgroup by default.

[danobi]

Perhaps it is worth taking the default cgroup2 root from /proc/mounts instead of using /sys/fs/cgroup by default.

Oh yeah I like this idea. I'll file a separate issue for this.