Nice, I like this

Address @glebius comment

I'm happy to land this. Abdelkader Boudih <freebsd@seuros.com> as the Git author info?

The first bit of the commit message (subject) should be the general subsystem, I imagine it could be cam: here

@bnovkov do you know why the "needs close" check determines this one has already landed? (i.e. fbsd_author_needs_close.html)

Code wise looks good to me. But I'm not protocol expert.

Adjust minimum required python version to 3.11 to match pyproject.toml

• Previous revision stack was committed.
• All possible improvements that derive from this file, will be re-submitted.

Don't we need to do a broadcast invalidate on fini (or remember the allocated VMIDs from the current epoch), otherwise we'll reallocate the same VMIDs if we kldload and start VMs? Also, don't leave the epoch as is whilst clearing the bitmap and next, that doesn't make sense, if the epoch is staying then so does the allocated set, and if it's being reset then so does the allocated set (along with invalidate).

Remove "opt_route.h" header from all files except for:
sys/net/route/route_tables.c:#include "opt_route.h"
sys/net/route/fib_algo.c:#include "opt_route.h"
sys/netinet/in_fib.c:#include "opt_route.h"
sys/netinet6/in6_fib.c:#include "opt_route.h"

The build looks good on 16amd64: https://pkg.ftfl.ca/build.html?mastername=16amd64-default&build=2026-03-19_12h15m13s

This is not complete, but would be useful to get feedback on the approach

In D55831#1279764, @jhb wrote:

I can't recall if kgdb has to know the size of struct trapframe is my only concern. I think we already only use the 16 bits for cs/ss today.

I can't recall if kgdb has to know the size of struct trapframe is my only concern. I think we already only use the 16 bits for cs/ss today.

Tweak sbuf overflow case

Why not simply not allocate the new number for the new ioctl. Rename current ioctl to something like SIOCGI2C_NOBANK. Then, at the ioctl level, do two things:

• for new SIOCGI2C check that reserved bytes in the structure are zero
• for SIOCGI2C_NOBANK, zero bank/page/padding and call into SIOCGI2C

Actually, the documentation doesn't permit sleeping between the start and stop callbacks, so it will need to be a single copyout() at the end regardless.

Rewrite loop

In D55931#1279683, @jhb wrote:

Speaking of this btw, it might be nice to fix other architectures to also pull the pcb out of the kstack in general. Certainly for CHERI we prefer that arrangement.

In D55831#1279681, @jhb wrote:

FYI, kgdb will have to know about the change in layout to trapframe I think, so bump __FreeBSD_version when you land this just to be on the safe side please?

Actually, to support SEQ_SKIP (which is in the docs but we don't currently implement), I do indeed need to keep the sbuf_clear/finish in the loop. Will upload a new version in a bit.

Looks good to me.

Any chance we can make zstd a loadable module?

Two failures:

• Cross-build Kernel / aarch64 macos-latest (clang-18) Failed in 19 minutes and 24 seconds
• Cross-build Kernel / amd64 macos-latest (clang-18) Failed in 5 minutes and 15 seconds

Issue created: https://codeberg.org/tobik/portfmt/issues/34

Updating D55907: clk/broadcom: Add CPRMAN clkdev backend and peripheral clock node