• Remove unneeded intr_disable() call.

In D16688#399572, @raichoo_googlemail.com wrote:

I have already made such a change by importing documentation from the Handbook into the manpage. I've talked to @bcr asking the same question at the Essen Hackathon and he told me it would be okay. I personally would like to have something like this in the manpage since I use dtrace quite a bit and I'm missing proper documentation that is being shipped with FreeBSD that I can consume on the terminal. The manpage seems to be the obvious place for this since it's where people look first.

In D18728#399576, @greg_unrelenting.technology wrote:

I think that's the right place in sctp. I'm not sure how to test it though, or the sdp thing…

In D16688#399496, @raichoo_googlemail.com wrote:

Bumping this since it's been laying dormant for quite a while. Does this still need work to get merged? I've had people asking me things at 35c3 that have been documented here.

Looks fine to me.

This seems fine to me. sdp_sosend() needs the same treatment. sctp_sosend() probably does as well, but I don't see where. @tuexen might know.

• Don't bother zeroing the top-level page.

In D18721#399440, @kib wrote:

I must admit that the use of allpmaps in riscv pmap is somewhat strange. Can we preallocate kernel l1 tables and do not touch them later at all, only copying the kernel part into each new user pmap ?

• Fix the bug that kib pointed out.

A couple of minor changes:

• ensure the bucket list is initialized
• when reclaiming, adjust imin and imax at the same time as nitems

The main portion of the change looks fine to me, I did not look very closely at the mlx5 change though.