address feedback from markj:

• wrap long line
• close file descriptor being leaked

feedback from jhb..use SUN_LEN()

looks good, thanks

What about receiving the file descriptors from the nvlist?

In D32886#791073, @asomers wrote:

OTOH, on my system the only collisions are caused by the ZFS sysctls marked "LEGACY", and vm.uma.tcp_log_bucket.size/vm.uma.tcp_log.bucket_size. So a less disruptive option would simply be to ignore the LEGACY sysctls, and rename vm.uma.tcp_log.bucket_size.

From the PR...it looks like ~25 metrics are broken. And compare that to the total number of exported metrics:

% prometheus_sysctl_exporter | wc -l
    13360

If the name of the metric being exported changes...how will that not break existing setups?

In D34880#790376, @kp wrote:

I'm very, very tempted to say we should just remove DIOCCHANGERULE, but at least miniupnpd actually uses it.

Taking this review in a different direction..

feedback from imp: use ${NO_WUNUSED_BUT_SET_VARIABLE}

remaining mandoc warnings left to preserve format of original text

All revisions in this stack need to be updated and rebased.

In D34584#783549, @0mp wrote:

MFC?

You have broken up https://reviews.freebsd.org/D30471 into multiple reviews without addressing the issues I raised in the original review.

I’ll commit this tomorrow.

not sure if using CSTD=c11 in usr.bin/zstd/Makefile would be acceptable/work here, to avoid touching contrib code.

I split out the bhyvectl changes into a separate review:

updated per feedback from jhb

In D34372#779008, @chuck wrote:

Perhaps I'm getting confused. Are we saying on FreeBSD mutex's are initialized automatically, or in this particular case? If so, where/how does the initialization happen?

I didn't realize that these mutex's get initialized on FreeBSD, even without your patch.

looking at the comment above the struct virtio_softc definition in usr.sbin/bhyve/virtio.h:

[ ... snipped ... ]
* Note: inside each hypervisor virtio driver, changes to these
* data structures must be locked against other threads, if any.
* Except for PCI config space register read/write, we assume each
* driver does the required locking, but we need a pointer to the
* lock (if there is one) for PCI config space read/write ops.
[ ... snipped ... ]