[...]
> Hi Alexander,
>
> The good news: I can consistently reproduce the hang problem.  But the
> bad news is that even with a WITNESS kernel and kern.witness.watch=2
> (or even 3) I don't see any message or kernel panic.
>
> Any additional suggestion in order to increase debug information or
> force the machine to go into ddb?

I tried a MP_LOCKDEBUG kernel and finally got a panic, shown below.
Sorry for not including 'show locks' and all the cpus, but my serial
console session froze before that.

Hope it helps,
--Kor

pmap_tlb_shootwait: spun out
Stopped at  db_enter+0x14:  popq    %rbp
ddb{13}>
--
ddb{13}> show panic
the kernel did not panic
--
ddb{13}> ps /o
    TID    PID    UID     PRFLAGS     PFLAGS  CPU  COMMAND
 307801  52176      0         0x2  0x4000000   10  semaphore
 410431  54690      0         0x2  0x4000000   15  semaphore
 257513  26918      0         0x2  0x4000000   11  semaphore
 138054  68532      0         0x2  0x4000000    5  semaphore
 203059   2408      0         0x2  0x4000000    9  semaphore
  89141   5998      0         0x2  0x4000000    8  semaphore
 263439  99498      0         0x2  0x4000000    7  semaphore
 477129  15941      0         0x2  0x4000000   12  semaphore
  41845  21800      0         0x3  0x4000000    2  nfdump
 371022  21800      0         0x3  0x4000000    0  nfdump
*391905  66338      0         0x3  0x4000000   13  nfdump
 367615  66338      0         0x3  0x4000000    6  nfdump
 394180  56727      0         0x3  0x4000000    3  nfdump
  46184  33300      0    0x100003          0    1K top
--
ddb{13}> trace
db_enter() at db_enter+0x14
pmap_page_remove(fffffd8184bdc408) at pmap_page_remove+0x38e
uvm_anfree_list(fffffd9c88e2be50,0) at uvm_anfree_list+0x63
amap_wipeout(fffffd9ca390fc88) at amap_wipeout+0x198
uvm_unmap_detach(ffff80005608fcc0,0) at uvm_unmap_detach+0xbe
sys_munmap(ffff800055cda540,ffff80005608fdc0,ffff80005608fd40) at
sys_munmap+0x185
syscall(ffff80005608fdc0) at syscall+0x5f9
Xsyscall() at Xsyscall+0x128
end of kernel
end trace frame: 0x123cf0e2b80, count: -8
--
ddb{13}> machine ddbcpu 1
Stopped at      x86_ipi_db+0x16:        leave

ddb{1}> trace
x86_ipi_db(ffff8000552aaff0) at x86_ipi_db+0x16
x86_ipi_handler() at x86_ipi_handler+0x80
Xresume_lapic_ipi() at Xresume_lapic_ipi+0x27
pmap_enter(fffffd9cd8988d00,7103880b6000,1c7767a000,3,33) at pmap_enter+0x703
uvm_fault_upper(ffff800055da4928,ffff800055da4960,ffff800055da4820) at
uvm_fault_upper+0x1e0
uvm_fault(fffffd9ca1ab8b58,7103880b6000,2,3) at uvm_fault+0xce
uvm_fault_wire(fffffd9ca1ab8b58,7103880b6000,7103880b7000,3) at
uvm_fault_wire+0x58
uvm_map_pageable_wire(fffffd9ca1ab8b58,fffffd9c8f522700,fffffd9ca0915ab8,0,3,0)
 at uvm_map_pageable_wire+0x19d
sysctl_vslock(7103880b6d68,18) at sysctl_vslock+0x9b
sys_sysctl(ffff800055d61c60,ffff800055da4c80,ffff800055da4c00) at
sys_sysctl+0x187
syscall(ffff800055da4c80) at syscall+0x5f9
Xsyscall() at Xsyscall+0x128
end of kernel
end trace frame: 0x7103880b6dc0, count: -12
--

[...]

Reply via email to