Do you think it might help with debugging if we consulted the
zfs_no_trimtuneable directly, rather than having it affect the
If we don't do this here, but rather consult the tuneable directly in the other places (e.g.,
vdev_autotrim_should_stop()) we can still see whether the TRIM bits would be active for a vdev, and we can tune it on without reimporting the pool. This might also help out when testing a potential blacklist feature?