Folks,
Perhaps someone here can help me shed some light on this mystery. At
some point in one of the recent major merges with onnv, the following
very-early-in-boot crash on x86 appeared in the clearview gate. This
does not happen on SPARC, and it does not happen when this same system is
running a stock onnv nightly build from the same day.
WARNING: kcf:compute_min_max_threads cpupart_get_cpus: failed, setting
kcf_minthreads to 1
panic[cpu0]/thread=fec1e260: BAD TRAP: type=8 (#df Double fault)
rp=fec22cbc addr=0
#df Double fault
pid=0, pc=0xfe800406, sp=0xfec36804, eflags=0x202
cr0: 80050019<pg,wp,et,ts,pe> cr4: 98<pge,pse,de>
cr2: f000ff74cr3: 2638000
gs: 1b0 fs: 0 es: 160 ds: 160
edi: 0 esi: 0 ebp: 0 esp: fec22cf4
ebx: 54 edx: d ecx: 0 eax: fec36b80
trp: 8 err: 0 eip: fe800406 cs: 158
efl: 202 usp: fec36804 ss: 160
cpu address timestamp type vc handler pc
0 fec3965c dc0aac05a trap 8 #df cmntrap+106
0 fec395b4 dc0aaac6e trap e #pf mutex_owner_running+e
tss.tss_link: 0x0
tss.tss_esp0: 0x0
tss.tss_ss0: 0x160
tss.tss_esp1: 0x0
tss.tss_ss1: 0x0
tss.tss_esp2: 0x0
tss.tss_ss2: 0x0
tss.tss_cr3: 0x2638000
tss.tss_eip: 0xfe800406
tss.tss_eflags: 0x202
tss.tss_eax: 0xfec36b80
tss.tss_ebx: 0x54
tss.tss_ecx: 0xf0000000
tss.tss_edx: 0xd
tss.tss_esp: 0xfec36804
fec22c0c unix:die+105 (8, fec22cbc, 0, 0)
fec22ca8 unix:trap+1357 (fec22cbc, 0, 0)
fec22cbc unix:cmntrap+10b (1b0, 0, 160, 160, 0)
skipping system dump - no dump device configured
rebooting...
panic[cpu0]/thread=fec1e260: BAD TRAP: type=e (#pf Page fault)
rp=fec407f8 addr=f0002707
dump aborted: please record the above information!
rebooting...
The system panics with that last page fault repeatedly thereafter trying
to reboot.
I can get the thing to break into kmdb, but I'm at a loss at this point.
The double-fault is throwing me off the scent. Any pointers out there?
-Seb
