With both yesterday's and today's snapshots for sparc64, my LDOMs panic
if they run pfctl. If I get in single user mode and chmod 000 /sbin/pfctl,
then the machine finishes boot into multi-user.
Here's what I got from ddb
panic: kernel data fault: pc=101356c addr=0
Stopped at db_enter+0x8: nop
TID PID UID PRFLAGS PFLAGS CPU COMMAND
*515341 61157 0 0x3 0 1K pfctl
424322 92540 0 0x14000 0x200 5 zerothread
data_access_fault(400b76cd390, 31, 101356c, 0, 0, 1) at data_access_fault+0x30c
sun4v_datatrap(0, 10040200000001, 4149580000000000, 0, 40011bfbdd0, 1) at sun4v
_datatrap+0x210
pool_get(0, 40011cf3e00, 70, 4000, 40, 0) at pool_get+0xcc
pf_osfp_add(40011cf3e00, 0, 0, 198ac98, 40011cf3e50, 1c00) at pf_osfp_add+0x378
pfioctl(4900, c088444f, 40011cf3e00, 3, 40011c7b7b0, 0) at pfioctl+0x1e8c
spec_ioctl(400b76cd9f8, 40291ff8000, 17838f8, a717794028, a717794000, 1) at spe
c_ioctl+0xb8
VOP_IOCTL(40011c17470, c088444f, 40011cf3e00, 198ac98, 40011ca3ec0, 40011c7b7b0
) at VOP_IOCTL+0x6c
vn_ioctl(40011c53800, c088444f, 40011cf3e00, 40011c7b7b0, 1012f50, 40011cf3e00)
at vn_ioctl+0x64
sys_ioctl(0, 400b76cddb0, 400b76cddf0, 17e83f8, 15ec3a0, 1e) at sys_ioctl+0x234
syscall(400b76cded0, 436, a46e82e068, a46e82e06c, 0, 1) at syscall+0x390
syscall_setup(3, c088444f, fffffffffffddd40, fffffffffffdd468, 0, 2) at syscall
_setup+0x134
https://www.openbsd.org/ddb.html describes the minimum info required in bug
reports. Insufficient info makes it difficult to find and fix bugs.
ddb{1}> trace
data_access_fault(400b76cd390, 31, 101356c, 0, 0, 1) at data_access_fault+0x30c
sun4v_datatrap(0, 10040200000001, 4149580000000000, 0, 40011bfbdd0, 1) at sun4v
_datatrap+0x210
pool_get(0, 40011cf3e00, 70, 4000, 40, 0) at pool_get+0xcc
pf_osfp_add(40011cf3e00, 0, 0, 198ac98, 40011cf3e50, 1c00) at pf_osfp_add+0x378
pfioctl(4900, c088444f, 40011cf3e00, 3, 40011c7b7b0, 0) at pfioctl+0x1e8c
spec_ioctl(400b76cd9f8, 40291ff8000, 17838f8, a717794028, a717794000, 1) at spe
c_ioctl+0xb8
VOP_IOCTL(40011c17470, c088444f, 40011cf3e00, 198ac98, 40011ca3ec0, 40011c7b7b0
) at VOP_IOCTL+0x6c
vn_ioctl(40011c53800, c088444f, 40011cf3e00, 40011c7b7b0, 1012f50, 40011cf3e00)
at vn_ioctl+0x64
sys_ioctl(0, 400b76cddb0, 400b76cddf0, 17e83f8, 15ec3a0, 1e) at sys_ioctl+0x234
syscall(400b76cded0, 436, a46e82e068, a46e82e06c, 0, 1) at syscall+0x390
syscall_setup(3, c088444f, fffffffffffddd40, fffffffffffdd468, 0, 2) at syscall
_setup+0x134
ddb{1}> ps
PID TID PPID UID S FLAGS WAIT COMMAND
*61157 515341 16571 0 7 0x3 pfctl
16571 426607 1 0 3 0x10008b pause sh
70519 276043 0 0 3 0x14200 bored smr
92540 424322 0 0 7 0x14200 zerothread
87004 32019 0 0 3 0x14200 aiodoned aiodoned
5172 399130 0 0 3 0x14200 syncer update
89163 105042 0 0 3 0x14200 cleaner cleaner
43692 475871 0 0 3 0x14200 reaper reaper
31065 289077 0 0 3 0x14200 pgdaemon pagedaemon
88791 267552 0 0 3 0x14200 bored crynlk
68417 328642 0 0 3 0x14200 bored crypto
88795 453910 0 0 3 0x40014200 idle5
7467 171346 0 0 7 0x40014200 idle4
70440 423883 0 0 7 0x40014200 idle3
4602 474640 0 0 7 0x40014200 idle2
33593 101368 0 0 3 0x40014200 idle1
50435 169391 0 0 3 0x14200 bored softnet
67100 501094 0 0 3 0x14200 bored systqmp
39979 405988 0 0 3 0x14200 bored systq
91889 106504 0 0 3 0x40014200 bored softclock
46855 75622 0 0 7 0x40014200 idle0
80081 331540 0 0 3 0x14200 kmalloc kmthread
1 273837 0 0 3 0x82 wait init
0 0 -1 0 3 0x10200 scheduler swapper
Attempting to use mach ddbcpu 0 just hung forever.
--Kurt
