♟ jhb

ctl: Use CAM_PRIORITY_NORMAL for queued CCBs

ctld: Only check physical port linking in a single configuration context

ctld: Refactor ioctl port handling

ctld: Don't ignore pp/vp values of kernel ports with pp == 0 but vp != 0

I was able to effectively test this today using pre-existing ioctl ports.

I was able to test this directly today playing around with ioctl ports.

Permit using ioctl/0/0

riscv64: Move pcb out of kstack into a new UMA zone

arm64: Move pcb out of kstack into a new UMA zone

sys: Consistently use ptoa() with td_kstack_pages

sys: Add td_kstack_top inline helper function

arm64 locore: Fix a comment typo: Of -> If

amd64: Inline set_top_of_stack_td into cpu_thread_new_kstack

arm64: Adjust the kernel stack pointer at the end of fork_trampoline

cpu_thread_new_kstack: Introduce new MD callback

amd64/i386: Remove dead code to clear XSAVE header

cpu_fork: Remove redundant assignments to td_pcb and td_frame

amd64: Don't set an initial FPU state save size

arm64: Remove unneeded declaration from <machine/pcpu.h>

amd64: Drop segment descriptor details from trap messsages

RELNOTES: Add entries for various commits through early March 2026

While the previous patch in this stack fixes a reproducer from Robert, this followup patch is intended as additional robustness. With both of these I was able to still mount a FreeBSD install CD and all of the attributes reported by find /mnt -ls were the same before and after the patches.

powerpc: Remove stale include line from MPC85XX

arm64/iommu: Add a missing close parenthesis

vnic: Add missing #include

ctld: Mark a few more isns_* methods in the conf class private

ctld: Add a dedicated conf method for shutting down

ctld: Simplify pidfile rename handling in conf::apply

ctld: Remove redundant call to conf::isns_schedule_update

sys: Use is_pci_device instead of checking device or devclass names

sys: Use is_pci_device instead of direct comparisons to devclasses

pci: Add is_pci_device helper function

LinuxKPI: Check the devclass of the parent device in dev_is_pci

In D56996#1305937, @bz wrote:

I was contemplating the name but I couldn't come up with anything nice that started with pci_ ..

In D57066#1307689, @bz wrote:

We assume that parent it is not a drmn device. Given I added this, it was for mt76 back then likely, so this was clearly a bug.

Rebase on top of LinuxKPI fixup

Restore seg reg values

In D56989#1305907, @kib wrote:

For me, the segment registers values from the trap frame appeared to be very useful and sometimes even instrumental, e.g. during the recent FRED work. If the kernel idea of the trap frame address is misaligned with the location where the hardware put it, segment values are canaries that allow to see that immediately. Also, bugs with segments handling in doreti or fred_gouser are relatively obvious then. I am asking to keep that in the frame dump.

In terms of accessors in general: I recently added a new "has" wrapper method to __BUS_ACCESSOR, but it only works if your IVAR is using a "global" number and not the private range. All of the PCI IVARs are in the private range.

clock_gettime.2: Fix a couple of typos

arm64/vmm: Enforce a data barrier before tlbi for non-VHE

mandoc: Updating FreeBSD thead library descriptions

mandoc: Add an entry for libsys to lib.in

In D56250#1296293, @minsoochoo0122_proton.me wrote:

In D56250#1295915, @jhb wrote:

Hmmm, cross debugging of cross-architectures does work on a FreeBSD host?

Oh, I meant debugging across different operating systems (e.g. FreeBSD arm64 dump on Linux amd64 LLDB). This was something we wanted since the foundation sponsored LLDB's kernel debugging project, but the approach at that time was to port libkvm to other OSes (libfbsdvmcore). The end result was portable libkvm, but the issue was that most distros didn't ship LLDB with that third-party library, hence debugging FreeBSD crash dump was impossible without building LLDB from source with that library included. Also libfbsdvmcore had to catch up with libkvm, but it has been unmaintained for years.

kobj: Use M_WAITOK in kobj_init

kobj: Assert class passed to kobj_init_static is compiled

jhb (John Baldwin)
User

Projects (10)
View All

User Details

Recent Activity
View All

Thu, May 28

Wed, May 27

Tue, May 26

Wed, May 20

Tue, May 19

Mon, May 18

Thu, May 14

Wed, May 13

Mon, May 11

Fri, May 8

Thu, May 7

Wed, May 6

May 2 2026

jhb (John Baldwin)User

Projects (10)View All