freebsd-update bug?

I have updated some FreeBSD 15-RELEASE machines, I wanted to go to -p1, however, it seems on all systems freebsd-update did only update userland. After the reboot,
uname -a still prints
FreeBSD log 15.0-RELEASE FreeBSD 15.0-RELEASE releng/15.0-n280995-7aedc8de6446 GENERIC amd64. Also,
freebsd-version -kru prints
Code:
15.0-RELEASE
15.0-RELEASE
15.0-RELEASE-p1

I also updated my pkgbase test VM, and this machine indeed did a kernel update to FreeBSD vmfreebsd 15.0-RELEASE-p1 FreeBSD 15.0-RELEASE-p1 releng/15.0-n280999-7bceec30b351 GENERIC amd64, this is also represented in the freebsd-version output.

Shall I file a bug report?
 
Not answering your question directly, but I can see that my freebsd-update output doesn't show a new kernel file, so the kernel version won't change:
Code:
# freebsd-update fetch
src component not installed, skipped
Looking up update.FreeBSD.org mirrors... 3 mirrors found.
Fetching metadata signature for 15.0-RELEASE from update2.freebsd.org... done.
Fetching metadata index... done.
Fetching 2 metadata patches.. done.
Applying metadata patches... done.
Inspecting system... done.
Preparing to download files... done.
Fetching 6 patches.... done.
Applying patches... done.
The following files will be updated as part of updating to
15.0-RELEASE-p1:
/bin/freebsd-version
/boot/kernel/vmm.ko
/boot/kernel/zfs.ko
/rescue/[
...
So I get the same output as you, but I'd expect that because there's no new kernel file:
Code:
 # freebsd-version -ruk
15.0-RELEASE
15.0-RELEASE
15.0-RELEASE-p1

But maybe there was meant to be a new kernel file (and pkgbase installed it?)
 
Yep. It's a userland only thing. So no new kernel, thus not updated.

Only rtsold(8)/rtsol(8)/resolvconf(8) had a bug.


This one got fixed in the kernel module.

This one too, a kernel module.

Both weren't included in the kernel itself (GENERIC) but are dynamically loaded. There was no need to update the kernel itself.
 
With respect to the remote packages of the FreeBSD-base repository, my guess is that all base packages have been rebuilt, resulting in p1 "three times", just as would be the case when one would have applied these EN-s and SAs based on source.
 
There seem to be 9 base packages affected by these EN-s and SA-s:
Code:
[1-0] #  pkg que -x -e '%o~base/*' '%o %n %v' '15\.0p1'  | column -t | nl
     1  base/FreeBSD-kernel-generic      FreeBSD-kernel-generic      15.0p1
     2  base/FreeBSD-kernel-generic-dbg  FreeBSD-kernel-generic-dbg  15.0p1
     3  base/FreeBSD-rescue              FreeBSD-rescue              15.0p1
     4  base/FreeBSD-runtime             FreeBSD-runtime             15.0p1
     5  base/FreeBSD-runtime-dbg         FreeBSD-runtime-dbg         15.0p1
     6  base/FreeBSD-src                 FreeBSD-src                 15.0p1
     7  base/FreeBSD-src-sys             FreeBSD-src-sys             15.0p1
     8  base/FreeBSD-utilities           FreeBSD-utilities           15.0p1
     9  base/FreeBSD-utilities-dbg       FreeBSD-utilities-dbg       15.0p1
[2-0] #
IF one had choosen to upgrade only these specifically, then—most likely—there would have been only one p1 in the output of freebsd-version -kru


Yeah, they're built as a whole. Which kind of defeats their purpose to be honest.
That might change in the future, once all the remote "build wrinkles" have been ironed out.
I'm hopeful :)
 
Back
Top