Updated the diff: the previous upload accidentally included an unrelated hunk reverting the pmap_invalidate_range() operand order, an artifact of generating the diff against a base branch that already carried that separate fix. This revision now contains only the dcbz page zero/copy change.

What about powerpc?

Can you please update the docs in this commit?

In D59501#1365171, @kbowling wrote:

FWIW when we did LB we were really worried about breaking software, and in hindsight we maybe should have just aligned REUSEPORT across a major FreeBSD version with Linux to minimize third party software adaptation. Think hard before adding yet another non-portable sockopt; the LB_CPU_CURRENT behavior sounds desirable unless there is some corner you see. So I agree with Gleb.

This revision is superseded by D57633. Its build-wiring change is folded into the rebuilt D57633 because that wiring is the activation switch for the PMU implementation and does not form a useful standalone review boundary.

Update to the rebuilt PerfMonV2 grouping and multiplexing stack.