Looks fine to me, but I no longer have the hardware to actually test the changes, iXsystems may be able to help with that.

I'm mostly in favor of the change (smaller kernel size, etc.), but since we are removing driver from GENERIC, it's better if we can provide a mechanism to load the driver on demand, or a binary update can easily brick the system...

Why are we making this change? It was somewhat intentional to not enable SASL by default (to reduce attack surface, e.g. CVE-2019-13565), is this functionality being used by a lot of consumers?

LGTM.

Landed as rS351480.

Would like a different @releng member to take a look if possible, but the change looks reasonable to me.

Remove unused field.

Fix accounting for NumFree.

Fixed a couple of bugs.

Remove D21408 portion of change.

@markj -- do you have some suggestion on how should we test subr_compressor.c changes?

PTAL.

Fix subr_compressor.c.

In D21186#463637, @ota_j.email.ne.jp wrote:

The change looks reasonable based on PPP protocol and comparison to existing zlib code.

I didn't know ng_deflate was for ppp.
I was curious how others implemented and found that NetBSD, OpenBSD, and Linux uses Z_PACKET_FLUSH as well, interestingly.
I never looked at ppp implementation and didn't know all others have same base code along with modified zlib.
I am no longer sure if Z_PACKET_FLUSH was FreeBSD original implementation.

Assert that the empty block actually exists.

Undo unrelated space change.

Remove bogus inflateInit().

Address reviewer comments.

(Hopefully have fixed all issues I have found so far).

Revamp.

I have tested with 'dtrace -lv' which will exercise this codepath and confirmed that it worked fine after change.

Refresh patch.

Whoops, this should be committed already according to https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=21519#c18 .

Properly initialize zstream structure.