Re: halt -p/reboot - ddb (was Re: halt -p: Stopped at gettick+0xec: inb $0x40,%al)

2007-12-12 Thread Stuart Henderson 
same box ... -current MP, looks like two ddb running at the same time.

# reboot
/etc/rc.shutdown in progress...
/etc/rc.shutdown complete.
sycing disks... done
pSatnoipcp:e dm taxt_enter: lockin g  a  g ai_n_mspt_ lmoycske+l0fx
ep:a nmiocv:l mtx_enter :  lo 0cxki4n(g% eadgxa)i,n%seta xm3
sReUlNfy
Tp aLnEiAcS:T  m'ttxr_aecnet'e rA:N Dl o'cpksi'n gA NaDg aIiNnCsLtU DmEy
sOeUlTfP
Tp aWnHiEcN:  RmEtPxO_ReTnItNeGr :T HlIoSc kPiAnNgI Ca!g
   U
iDnOs tN OmTy sEeVlEfN  a
BpOaTnHiEcR:  RmEtPxO_ReTnItNeGr :T HlIoSc kWiInTgH OaUgTa iInNsCtL
UmDyIsNeGl fT
ApTa nIiNcF:O RmMtAxT_IeOnNt!e
   H
:d dlbo{c1k}in g against myself
panic: mtx_enter: locking against myself
panic: mtx_enter: locking against myself
panic: mtx_enter: locking against myself
panic: mtx_enter: locking against myselfpanic: 
cpu1: fp_save ipi didn't (P)panic: 
mtx_enter: locking against myself
panic: mtx_enter: locking against myself
panic: mtx_enter: locking against myself
panic: mtx_enter: locking against myself
panic: mtx_enter: locking against myself
panic: mtx_enter: locking against myself
[repeats]

I won't re-quote the last mail again, the other details are all
here: http://marc.info/?l=openbsd-miscm=119739509606536w=2

halt -p/reboot - ddb (was Re: halt -p: Stopped at gettick+0xec: inb $0x40,%al)

2007-12-11 Thread Stuart Henderson 
On 2007/12/10 17:32, Stuart Henderson wrote:
 On 2007/12/10 17:06, Stuart Henderson wrote:
  I've got a ServerWorks-based Fujitsu-Siemens Xeon box. At 'halt -p'
  (with or without acpi) the following happens (no panic).
 
 Ugh. 'reboot', too.

More info: with bsd.mp, reboot drops to ddb too, but differently.
(GENERIC.MP dmesg is added right at the bottom). halt -p works.

At this point 'c' allows the reboot/halt to proceed, so at least
I no longer need a power-cycle.

This works ok in 4.0, not in 4.1/newer.

sycing disks... done
Stopped at  __mp_lock+0x3e: movl0x4(%edx),%eax
ddb{1} tr
__mp_lock(d07cc144,d048767a,dac34f1c,dac34f1c) at __mp_lock+0x3e
i386_softintlock(0,d0350058,d6b50010,10,dac30010) at i386_softintlock+0x10
Xintrltimer() at Xintrltimer+0x47
--- interrupt ---
cpu_idle_cycle(d1270800) at cpu_idle_cycle+0xf
Bad frame pointer: 0xd0911e78
ddb{1} ps
   PID   PPID   PGRPUID  S   FLAGS  WAIT  COMMAND
  6898  1   6898  0  7   0x2004002reboot
15  0  0  0  30x100200  bored crypto
14  0  0  0  3   0x2100200  aiodoned  aiodoned
13  0  0  0  2   0x2100200update
12  0  0  0  3   0x2100200  cleaner   cleaner
11  0  0  0  30x100200  reaperreaper
10  0  0  0  3   0x2100200  pgdaemon  pagedaemon
 9  0  0  0  2   0x2100600pfpurge
 8  0  0  0  3   0x2100200  usbtskusbtask
 7  0  0  0  3   0x2100200  usbevtusb0
 6  0  0  0  3   0x2100200  acpi_idle acpi0
*5  0  0  0  70x100200idle1
 4  0  0  0  30x100200  bored syswq
 3  0  0  0  30x100200idle0
 2  0  0  0  3   0x2100200  kmalloc   kmthread
 1  0  1  0  3   0x2004080  wait  init
 0 -1  0  0  3   0x2080200  scheduler swapper

Another time,

syncing disks... done
Stopped at  lapic_delay+0x3a:   cmpl%esi,%edi
ddb{0} tr
lapic_delay(3e8,0,4,febf) at lapic_delay+0x3a
ahd_reset(d1278000,0,804010,dac4aeb0,d02032c9) at ahd_reset+0xae
ahd_shutdown(d1278000,1fca5097,2,145886) at ahd_shutdown+0x2e
dohooks(d078bf00,1,dac4af00,d047c915) at dohooks+0x6b
boot(0,0,dac4af40,0,d078af34) at boot+0x63
sys_reboot(d6a47568,dac4af68,dac4af58,,2a) at sys_reboot+0x26
syscall() at syscall+0x27e
--- syscall (number 55) ---
0x1c0009c1:
ddb{0} ps
   PID   PPID   PGRPUID  S   FLAGS  WAIT  COMMAND
*10478  1  10478  0  7   0x2805002reboot
15  0  0  0  30x100200  bored crypto
14  0  0  0  3   0x2100200  aiodoned  aiodoned
13  0  0  0  2   0x2100200update
12  0  0  0  3   0x2100200  cleaner   cleaner
11  0  0  0  30x100200  reaperreaper
10  0  0  0  3   0x2100200  pgdaemon  pagedaemon
 9  0  0  0  2   0x2100600pfpurge
 8  0  0  0  3   0x2100200  usbtskusbtask
 7  0  0  0  3   0x2100200  usbevtusb0
 6  0  0  0  3   0x2100200  acpi_idle acpi0
 5  0  0  0  70x100200idle1
 4  0  0  0  30x100200  bored syswq
 3  0  0  0  30x100200idle0
 2  0  0  0  3   0x2100200  kmalloc   kmthread
 1  0  1  0  3   0x2004080  wait  init
 0 -1  0  0  2   0x2080200swapper

On 2007/12/10 17:32, Stuart Henderson wrote:
 On 2007/12/10 17:06, Stuart Henderson wrote:
  I've got a ServerWorks-based Fujitsu-Siemens Xeon box. At 'halt -p'
  (with or without acpi) the following happens (no panic).
 
 Ugh. 'reboot', too.
 
  Any suggestions?
  
  # halt -p
  /etc/rc.shutdown in progress...
  /etc/rc.shutdown complete.
  sycing disks... done
  Stopped at  gettick+0xec:   inb $0x40,%al
  ddb tr
  gettick(d116d000,4,186a0,3e8,d116d000) at gettick+0xec
  i8254_delay(3e8,0,4,febf) at i8254_delay+0x11
  ahd_reset(d116d000,0,804010,dab27eb0,d0202251) at ahd_reset+0xae
  ahd_shutdown(d116d000,2e9b,dab27ee0,d05caedd) at ahd_shutdown+0x2e
  dohooks(d0782000,1,dab27f00,d047a654) at dohooks+0x6b
  boot(1008,0,dab27f40,0,d0781034) at boot+0x63
  sys_reboot(d693dc20,dab27f68,dab27f58,,30) at sys_reboot+0x26
  syscall() at syscall+0x24e
  --- syscall (number 55) ---
  0x1c0009c1:
  ddb ps
 PID   PPID   PGRPUID  S   FLAGS  WAIT  COMMAND
  *21404  1  21404  0  7  0x4002halt
  14  0  0  0  30x100200  bored crypto
  13  0  0  0  30x100200  aiodoned  aiodoned
  12  0  0  0  2