Re: OpenBSD seems to fail on Primepower 250

Hello Ben,

OpenBSD has been running on my PrimePower 250 (2 x 1.1GHz, 2 GB of
RAM, no PCI cards) for well over a year now, and although some
programs crash consistently (Epiphany comes to mind), the system
itself has never let me down. True, I rarely push the disks or CPUs to
their limits for a prolonged period of time. But I do compile and link
large amounts of C++ code, and so far never had problems. If there is
some specific output you're interested in (simple tests only! nothing
that will crash/corrupt my disks!), let me know.

Cheers,
Tijs

On Tue, Sep 17, 2013 at 3:17 PM, Benjamin Bochmann <[email protected]> wrote:
> Hello,
>
> I observed failures of serveral OpenBSD versions on Primepower 250
> (Sparc64 V) servers which are described below. Because of the hardware
> similarity of the other Primepower servers this problem could be also
> present on Primepower 450, 650 and 850.
> Are there any chances to run OpenBSD on Primepower?
>
> Thanks in advance,
> Ben
>
> 1. Fujitsu Primepower 250, 2x SPARC64 V, (1,1 GHz), 2 GB of RAM
> 2. Fujitsu Primepower 250, 2x SPARC64 V, (1,28 GHz), 8 GB of RAM
> (both supported by OpenBSD)
>
> - good hardware condition, no errors reported by OpenBoot hardware tests
> - both machines run stable on FreeBSD 8.x for 1 year and FreeBSD 9.1
> - Machine 1 has an additional PCI-SCSI contoller (QLA10162 Dual
>   U3-SCSI) with no disks connected (disks are connected to built-in
>   SCSI controller), no graphics card => almost standard configuration
> - Machine 2 has an unused Radeon graphics card (PCI) by Fujitsu, no
>   additional PCI cards => standard configuration
> - both machines run stable on FreeBSD 8.3 for some weeks, also under
>   high CPU and disk load
> - clean power supply lines (double converter USV)
>
> Symptoms
>
> - machine hangs suddenly by high disk _and_ CPU load after processing
>   a certain amount of data, symptoms:
>   - serial console shows no reaction
>   - ssh server is dead
>   - but systems seems to be alive due to reply to ethernet ping commands
>   - after restart: damaged disk partition which was written, can not be
>   restored, no syslog entries
> => probably disk subsystem is dead
>
> following command kills the machine within 1...5 MB of written data:
> dd if=/dev/random of=testfile.dat bs=1024 count=80000000
>
> following command does not kill the machine (=> very low CPU usage):
> dd if=/dev/null of=testfile.dat bs=1024 count=80000000
>
> following command does not kill the machine (=> no disk usage):
> dd if=/dev/random of=/dev/null bs=1024 count=80000000
>
> - standard installation from CD freezes rarely the machines (media and
>   drives are ok)
>
> marks:
>   - same behavior on both machines (hardware damage very unlikely)
>   - symptoms observed on OpenBSD 4.9, 5.0, 5.1, 5.2 and 5.3 current,
> minimal MP installation without X.Org, 1GBit ethernet controller
> configured, no modifications
>   - independend on disk type, but the crash probability decreases with
>     faster disks (e.g. 15k rpm disks)
>
> Debug-console:
>
>
> ddb{0}> trace
>
> intr_list_handler(40001f4f500, 0, e0017ec8, 106020103fe, 1461480,
> 106020103fd) a
>
> t intr_list_handler+0x2c
>
> sparc_interrupt(1889038, 0, e0017ec8, 40025575cf0, 1461480, 180e268) at
> sparc_i
>
> nterrupt+0x2a8
>
> sparc_interrupt(188ccb0, 1c18000, 1547390, 18883e0, 0, 1) at
> sparc_interrupt+0x
>
> 2d4
>
> sched_idle(1c18000, 4000c0164e0, 18, f009e5bc, f009e5bc, 1801010) at
> sched_idle
>
> +0x180
>
> proc_trampoline(0, 0, 0, 0, 0, 0) at proc_trampoline+0x14
>
> ddb{0}>
>
> ddb{0}> ps
>
>    PID   PPID   PGRP    UID  S       FLAGS  WAIT          COMMAND
>
>  16116   3107  16116   1000  3           0  bqwait        dd
>
>   3107  11564   3107   1000  3        0x88  pause         ksh
>
>  11564  20267  20267   1000  3        0x80  select        sshd
>
>  20267  31522  20267      0  3        0x80  poll          sshd
>
>   3730      1   3730      0  3           0  biowait       getty
>
>  18681      1  18681      0  3        0x80  select        cron
>
>  10611      1  10611     99  3        0x80  poll          sndiod
>
>   6256      1   6256      0  3        0x80  select        inetd
>
>  14775      1  14775      0  3        0x80  select        sendmail
>
>  31522      1  31522      0  3        0x80  select        sshd
>
>  27248   5823   5823     74  3        0x80  bpf           pflogd
>
>   5823      1   5823      0  3        0x80  netio         pflogd
>
>   9515  11997  11997     73  2        0x80                syslogd
>
>  11997      1  11997      0  3        0x80  netio         syslogd
>
>     14      0      0      0  3    0x100200  aiodoned      aiodoned
>
>     13      0      0      0  3    0x100200  bqwait        update
>
>     12      0      0      0  3    0x100200  cleaner       cleaner
>
>     11      0      0      0  3    0x100200  biowait       reaper
>
>     10      0      0      0  3    0x100200  pgdaemon      pagedaemon
>
>      9      0      0      0  3    0x100200  bored         crypto
>
>      8      0      0      0  3    0x100200  pftm          pfpurge
>
>      7      0      0      0  3    0x100200  usbtsk        usbtask
>
>      6      0      0      0  3    0x100200  usbatsk       usbatsk
>
>      5      0      0      0  7  0x40100200                idle1
>
>      4      0      0      0  3    0x100200  bored         syswq
>
> *    3      0      0      0  7  0x40100200                idle0
>
>      2      0      0      0  3    0x100200  kmalloc       kmthread
>
>      1      0      1      0  3        0x80  wait          init
>
>      0     -1      0      0  3       0x200  scheduler     swapper
>
>   6389   3107   6389   1000  6      0x2000                top
>
> ddb{0}>
>
> ddb{0}> trace
>
> intr_list_handler(40001f4f500, 0, e0017ec8, 106020103fe, 1461480,
> 106020103fd) a
>
> t intr_list_handler+0x2c
>
> sparc_interrupt(1889038, 0, e0017ec8, 40025575cf0, 1461480, 180e268) at
> sparc_i
>
> nterrupt+0x2a8
>
> sparc_interrupt(188ccb0, 1c18000, 1547390, 18883e0, 0, 1) at
> sparc_interrupt+0x
>
> 2d4
>
> sched_idle(1c18000, 4000c0164e0, 18, f009e5bc, f009e5bc, 1801010) at
> sched_idle
>
> +0x180
>
> proc_trampoline(0, 0, 0, 0, 0, 0) at proc_trampoline+0x14
>
> ddb{0}>
>
> ddb{0}> show panic
>
> the kernel did not panic
>
> ddb{0}>
>
>
>
> < Power-off and on again >
>
>
>
> Mar 14 16:06:20 MET 2013 XSCF:I::[02020100] Start power off sequence
> Mar 14 16:06:23 MET 2013 XSCF:I::[02020300] Power off
> Mar 14 16:06:32 MET 2013 XSCF:I::[02010100] Power-on request from Power
> Switch Mar 14 16:06:34 MET 2013 XSCF:I::[02020400] Complete power off
> sequence. Mar 14 16:06:34 MET 2013 XSCF:I::[02010300] Start power on
> sequence Mar 14 16:06:41 MET 2013 XSCF:I::[04010E00] I/Os power on ...
> complete
>
> Mar 14 16:06:41 MET 2013 XSCF:I::[02010600] Initialize hardware:phase 1
> Mar 14 16:06:41 MET 2013 XSCF:I::[02010B00] Initialize hardware:phase 2
> Mar 14 16:06:41 MET 2013 XSCF:I::[02010C00] Initialize hardware:phase 3
> Mar 14 16:06:42 MET 2013 XSCF:I::[02010D00] Reset released...
>
> Mar 14 15:06:45 GMT 2013 PowerOn SelfTest start
> POST:Testing Flash/SRAM
> POST:Testing SC
> POST:Testing XSCF
> POST:Banner
> POST:FATAL check
> POST:Testing Timer1
> POST:Testing Tick
> POST:Testing MMU
> POST:Testing CPU Type
> POST:Testing DTAG
> POST:Memory Probe
> POST:Testing Memory
> POST:Testing Softint
> POST:Testing U2P
> POST:Testing Slave Device
> POST:Testing Master Device
> POST:System Configure
> POST:OBP Start
> screen:r1024x768x75 not found.
> keyboard not found.
>
>
> Fujitsu Siemens PRIMEPOWER250 2x SPARC64 V, No Keyboard
> OpenBoot 3.18.1-1, 2048 MB memory installed
> Ethernet address xx:xx:xx:xx:xx:xx, Host ID: xxxxxxxx.
> XSCF Version: 4.10.1
>
>
>
> Boot device: /pci@83,4000/FJSV,ulsa@2,1/disk@0,0  File and args:
> OpenBSD IEEE 1275 Bootblock 1.3
> ..>> OpenBSD BOOT 1.4
> Trying bsd...
> | / - \ | Booting /pci@83,4000/FJSV,ulsa@2,1/disk@0,0:a/bsd
> 6632144@0x1000000/ - \ | / - \ | / - \ | / - \ | / - \ | / - \ | / - \ | / - \ | / - \ | / - \ | / - \ | / - \ | / - \ | / - \ | / - \ | / - \ | / - \ | / - \ | / - \ | / - \ | / - \ | / - \ | / - \ | / - \ | / - \ | / - \ | / - \ | / - \ | / - \ | / - \ | / - \ | / - \ | / - \ | / - \ | / - \ | / - \ | / - \ | / - \ | / - \ | / - \ | / - \ | / - \ | / - \ | / - \ | / - \ | / - \ | / - \ | / - \ | / - \ | / - \ | / - \ | / - \ | / - \ | / - \ | / - \ | / - \ | / - \ | / - \ | / - \ | / - \ | / - \ | / - \ | / - \ | / - \ | / - \ | / - \ | / - \ | / - \ | / - \ | / - \ | / - \ | / - \ | / - \ | / - \ | / - \ | / - \ | / - \ | / - \ | / - \ | / - \ | / - \ | / - \ | / - \ | / - \ | / - \ | / - \ | / - \ | / - \ | / - \ | / - \ | / - \ | / - \ | / - \ | / - \ | / - \ | / - \ | / - \ | / - \
  | / - \ | / - \ | / - \ | / - +3376@0x16532d0\ +174928@0x1800000| / - \ | / - \ | / - \ +4019376@0x182ab50
> symbols @ 0xfea3c3c0
> 156+408840| / - \ | / - \ | / - \ | / - \ | / - \ | / - \ | +258160/ - \ | / - \ | / - \ | / - \ |
> start=0x1000000 [ using 668120 bytes of bsd ELF symbol table ]
>
> console is /pci@83,4000/isa@7/su@0,3f8
>
> Copyright (c) 1982, 1986, 1989, 1991, 1993
>
>        The Regents of the University of California.  All rights
>        reserved.
>
> Copyright (c) 1995-2013 OpenBSD. All rights reserved.
> http://www.OpenBSD.org
>
>
> OpenBSD 5.3 (GENERIC.MP) #46: Tue Mar 12 21:31:32 MDT 2013
>
>     [email protected]:/usr/src/sys/arch/sparc64/compile/GENERIC.MP
>
> real mem = 2147483648 (2048MB)
>
> avail mem = 2097233920 (2000MB)
>
> mainbus0 at root: Fujitsu Siemens PRIMEPOWER250 2x SPARC64 V
>
> cpu0 at mainbus0: FJSV,SPARC64-V (rev 0.8) @ 1100 MHz
>
> cpu0: physical 128K instruction (64 b/l), 128K data (64 b/l), 1024K
> external (64 b/l)
>
> cpu1 at mainbus0: FJSV,SPARC64-V (rev 0.8) @ 1100 MHz
>
> cpu1: physical 128K instruction (64 b/l), 128K data (64 b/l), 1024K
> external (64 b/l)
>
> psycho0 at mainbus0 addr 0xfffb2000: SUNW,psycho, impl 0, version 4,
> ign c0
>
> psycho0: bus range 0-0, PCI bus 0
>
> psycho0: dvma map fe000000-ffffffff, STC0 enabled
>
> pci0 at psycho0
>
> ebus0 at pci0 dev 1 function 0 "Sun PCIO EBus2" rev 0x01
>
> "FJSV,scfc" at ebus0 addr 210000-210085, 220000-220031, 260000-260001,
> 270000-28ffff ivec 0x23 not configured
>
> "FJSV,flashprom" at ebus0 addr 0-3fffff not configured
>
> clock1 at ebus0 addr 250000-251fff: mk48t59
>
> "FJSV,panel" at ebus0 addr 210011-210011 ivec 0x25 not configured
>
> ebus1 at pci0 dev 7 function 0 "Acer Labs M1533 ISA" rev 0x00
>
> com0 at ebus1 addr 3f8-3ff ivec 0x2b: ns16550a, 16 byte fifo
>
> com0: console
>
> com1 at ebus1 addr 2e8-2ef ivec 0x2b: ns16550a, 16 byte fifo
>
> hme0 at pci0 dev 1 function 1 "Sun HME" rev 0x01: ivec 0xe1, address
> xx:xx:xx:xx:xx:xx
>
> nsphyter0 at hme0 phy 1: DP83843 10/100 PHY, rev. 0
>
> mpi0 at pci0 dev 2 function 1 "Symbios Logic 53c1030" rev 0x07: ivec
> 0xe0
>
> scsibus0 at mpi0: 16 targets, initiator 7
>
> sd0 at scsibus0 targ 0 lun 0: <FUJITSU, MAT3147NC, 0108> SCSI3 0/direct
> fixed serial.FUJITSU_MAT3147NC_xxxxxxxxxx
>
> sd0: 140272MB, 512 bytes/sector, 287277984 sectors
>
> mpi0: target 0 Sync at 160MHz width 16bit offset 127 QAS 1 DT 1 IU 1
>
> pciide0 at pci0 dev 13 function 0 "Acer Labs M5229 UDMA IDE" rev 0xc4:
> DMA, channel 0 configured to native-PCI, channel 1 configured to
> native-PCI
>
> pciide0: using ivec 0xe4 for native-PCI interrupt
>
> atapiscsi0 at pciide0 channel 0 drive 0
>
> scsibus1 at atapiscsi0: 2 targets
>
> cd0 at scsibus1 targ 0 lun 0: <MATSHITA, DVD-ROM SR-8178, PY28> ATAPI
> 5/cdrom removable
>
> cd0(pciide0:0:0): using PIO mode 4, Ultra-DMA mode 2
>
> pciide0: channel 1 disabled (no drives)
>
> ohci0 at pci0 dev 10 function 0 "Acer Labs M5237 USB" rev 0x03: ivec
> 0xe9, version 1.0, legacy support
>
> usb0 at ohci0: USB revision 1.0
>
> uhub0 at usb0 "Acer Labs OHCI root hub" rev 1.00/1.00 addr 1
>
> psycho1 at mainbus0 addr 0xfff9e000: SUNW,psycho, impl 0, version 4,
> ign c0
>
> psycho1: bus range 128-129, PCI bus 128
>
> psycho1: dvma map fe000000-ffffffff, STC0 enabled, STC1 enabled
>
> pci1 at psycho1
>
> ppb0 at pci1 dev 2 function 0 "Intel 21154AE/BE PCI-PCI" rev 0x00
>
> pci2 at ppb0 bus 129
>
> isp0 at pci2 dev 4 function 0 "QLogic ISP10160" rev 0x06: ivec 0xc4
>
> isp0: board type 10160 rev 0x6, loaded firmware rev 10.4.41
>
> isp0: invalid NVRAM header
>
> scsibus2 at isp0: 16 targets, initiator 7
>
> isp1 at pci2 dev 5 function 0 "QLogic ISP10160" rev 0x06: ivec 0xc5
>
> isp1: board type 10160 rev 0x6, loaded firmware rev 10.4.41
>
> isp1: invalid NVRAM header
>
> scsibus3 at isp1: 16 targets, initiator 7
>
> bge0 at pci1 dev 1 function 0 "Fujitsu PRIMEPOWER250/450 LAN" rev 0x02,
> BCM5702/5703 A2 (0x1002): ivec 0xc0, address xx:xx:xx:xx:xx:xx
>
> brgphy0 at bge0 phy 1: BCM5703 10/100/1000baseT PHY, rev. 2
>
> "counter-timer" at mainbus0 addr 0xfff8bc00 not configured
>
> vscsi0 at root
>
> scsibus4 at vscsi0: 256 targets
>
> softraid0 at root
>
> scsibus5 at softraid0: 256 targets
>
> bootpath: /pci@83,4000/FJSV,ulsa@2,1/disk@0,0
>
> root on sd0a (4173aa211b13c328.a) swap on sd0b dump on sd0b
>
> WARNING: / was not properly unmounted
>
> Automatic boot in progress: starting file system checks.
> /dev/sd0a (4173aa211b13c328.a): 1823 files, 23825 used, 491206 free (86
> frags, 61390 blocks, 0.0% fragmentation) /dev/sd0a
> (4173aa211b13c328.a): MARKING FILE SYSTEM CLEAN /dev/sd0k
> (4173aa211b13c328.k): FREE BLK COUNT(S) WRONG IN SUPERBLK
> (SALVAGED) /dev/sd0k (4173aa211b13c328.k): SUMMARY INFORMATION BAD
> (SALVAGED) /dev/sd0k (4173aa211b13c328.k): BLK(S) MISSING IN BIT MAPS
> (SALVAGED) /dev/sd0k (4173aa211b13c328.k): 10 files, 1218616 used,
> 52819539 free (11 frags, 6602441 blocks, 0.0% fragmentation) /dev/sd0k
> (4173aa211b13c328.k): MARKING FILE SYSTEM CLEAN /dev/sd0d
> (4173aa211b13c328.d): 5 files, 4 used, 2063244 free (36 frags, 257901
> blocks, 0.0% fragmentation) /dev/sd0d (4173aa211b13c328.d): MARKING
> FILE SYSTEM CLEAN /dev/sd0f (4173aa211b13c328.f): 13397 files, 210008
> used, 820791 free (207 frags, 102573 blocks, 0.0%
> fragmentation) /dev/sd0f (4173aa211b13c328.f): MARKING FILE SYSTEM
> CLEAN /dev/sd0g (4173aa211b13c328.g): 8987 files, 71136 used, 443895
> free (1839 frags, 55257 blocks, 0.4% fragmentation) /dev/sd0g
> (4173aa211b13c328.g): MARKING FILE SYSTEM CLEAN /dev/sd0h
> (4173aa211b13c328.h): 108 files, 107 used, 5159576 free (72 frags,
> 644938 blocks, 0.0% fragmentation) /dev/sd0h (4173aa211b13c328.h):
> MARKING FILE SYSTEM CLEAN /dev/sd0j (4173aa211b13c328.j): 1 files, 1
> used, 1030798 free (14 frags, 128848 blocks, 0.0%
> fragmentation) /dev/sd0j (4173aa211b13c328.j): MARKING FILE SYSTEM
> CLEAN /dev/sd0i (4173aa211b13c328.i): 1 files, 1 used, 1030798 free (14
> frags, 128848 blocks, 0.0% fragmentation) /dev/sd0i
> (4173aa211b13c328.i): MARKING FILE SYSTEM CLEAN /dev/sd0e
> (4173aa211b13c328.e): 408 files, 2600 used, 4124634 free (106 frags,
> 515566 blocks, 0.0% fragmentation) /dev/sd0e (4173aa211b13c328.e):
> MARKING FILE SYSTEM CLEAN setting tty flags pf enabled ddb.panic: 1 ->
> 1 ddb.console: 0 -> 1 starting network starting early daemons: syslogd
> pflogd. starting RPC daemons:. savecore: no core dump checking quotas:
> done. clearing /tmp
> starting pre-securelevel daemons:.
> setting kernel security level: kern.securelevel: 0 -> 1
> creating runtime link editor directory cache.
> preserving editor files.
> starting network daemons: sshd sendmail inetd sndiod.
> starting local daemons: cron.
> Thu Mar 14 16:10:21 CET 2013
>
> OpenBSD/sparc64 (testsystem.my.domain) (console)
>
> login: