Panic after kldload i915kms at main-n280058-e7a04a110724

From: David Wolfskill <david_at_catwhisker.org>
Date: Thu, 04 Sep 2025 12:20:09 UTC
Yesterday's (source-based) update was to main-n280031-11f1dd193af5,
and I saw no issues.  Today's was to main-n280058-e7a04a110724.

My (headless) build machine (which does not attempt to either build
or load the module), and uses the GENERIC kernel, had no issues.

I note that the machines that load the modules also build them
whenever the kernel is built.  All machines mentioned here are
amd64.

The ports tree is at main-n717402-cd0318d52939.  (Yesterday, it was at
main-n717309-326381de144d.)  Installed ports are updated daily.

On one of the affected laptops, following the panic, I booted to
single-user mode & ran:

	swapon -a && fsck -p && mount -a

That done, I (manually) issued

	kldload i915kms

and (hand-transcribed -- phone is updating):

root@:/ # kldload i915kms

Fatal tyrap 12: page fault while in kernel mode
cpuid = 1; apic id = 01
fault bvirtual address	= 0x68
fault code		= supervisor read data, page not present
instruction pointer	= 0x20:0xffffffff80a92491
stack pointer		= 0x28:0xfffffe108ecea5e0
frame pointer		= 0x28:0xfffffe108ecea630
code segment		= base 0x0, limit 0xfffff, type 0x1b
			= DPL 0, pres 1, long 1, def32 0, gran 1
processor eflags	= interrupt enabled, resume, IOPL = 0
current process		= 37 (kldload)
...[sorry; I'm getting a bit impatient....]
KDB: stack backtrace:
db_trace_self_wrapper() at ...
vanic() at ...
panic() at ...
trap_pfault() at ...
calltrap() at ...
--- trap 0xc, rip = 0xffffffff80a92491, rsp = 0xfffffe108ecea5e0, rpb = 0xfffffe108ecea630 ---
pfs_create_dir() at ...
debugfs_create_dir() at ...
drm_core_init() at ...
linker_load_module() at ...
linker_load_dependencies() at ...
link_elf_laod_file() at ...
linker_load_module() at ...
kern_kldload() at ...
sys_kldload() at ...
fast_syscall_common() at ...
--- syscall (304, FreeBSD ELF64, kldload), rip = 0x1fbdd4a14e2a, rsp = 0x1fbdd22cde08, rpb = 0x1fbdd22ce380 ---
KDB: enter: panic
[ thread pid 37 tid 100152 ]
Stopped at	0xffffffff80bce523 = kdb_enter+0x33:	movq	$0,0x1230ee2(%rip)
db> 

I will get a photo up on my Web server once the phone update is done.

Peace,
david
-- 
David H. Wolfskill                              david@catwhisker.org
Of course firing the statistician will force the statistics to conform!

See https://www.catwhisker.org/~david/publickey.gpg for my public key.