Hello,

I seem to have a problem where 4.4 hangs writing to swap.
I can run: stress --vm 5 --vm-bytes 5M --vmhang 5 --timeout 1m
under 4.3 but under 4.4 the machine hangs.  Here's the background.

I'm ran nothing but bind (+ cron etc.) on a 586 with 48M of RAM
(machine A, the problem machine).  After upgrading from 4.2 to 4.4
stable patch 7 (via 4.3) it began hanging, no response to keyboard or
network.  The console's monitor shows video so I could see there
were no messages on the console.  (I have reflexively pressed the
shift key before the attached crt warms up so I can't say whether
there is keyboard response that turns off any sort of video blanking.)
I tried adding another 8M of RAM for a total of 64M, and told bind to
decrease the size of it's data cache, and neither helped.

I ran 9 passes of memtest86+ and booted into knoppix and ran 2 passes
of badblocks -w and got no errors.  (I believe I have since messed
with the RAM sticks, FWIW.)  Moving the hard drive to another
(slightly slower) 586 with 64M of RAM made the problem go away.  The
two boxes are slightly different versions of the same product, old HP
Vectra desktops.

It seemed to me the problem occurred when pages were being swapped
out.  There was no thrashing, swapctl showed between 6 and 30%
swap utilization.  Sometimes writing to swap works fine because
I can watch increased swap utilization with swapctl -l.

I ran: stress --vm 5 --vm-bytes 5M --vmhang 5 --timeout 1m
after putting another drive in box A.  (I picked these
numbers to fill physical ram but not swap.)

On box A the stress command runs fine on 4.3, but if I run the same
command after upgrading to 4.4 it hangs every time.  I can run this
command on box B 4.4 (again, stable patched to patch 7) without
problem.

I've already spent more time on this problem that it's worth, but it's
driving me batty.  I've even completely disassembled and cleaned box A
down to reseating the processor.  Do I have some sort of strange
hardware problem, a software bug, or something else?

Thanks for the help.

-----------------<snip>-------------------
Box A dmesg (the box that fails, aprox 40M swap, had 90M of swap
until I replaced the drive):

OpenBSD 4.4-stable (GENERIC) #31: Fri Jan 16 00:18:55 CST 2009
    k...@forge.meme.com:/usr/src/sys/arch/i386/compile/GENERIC
cpu0: Intel Pentium (P54C) ("GenuineIntel" 586-class) 167 MHz
cpu0: FPU,V86,DE,PSE,TSC,MSR,MCE,CX8
cpu0: F00F bug workaround installed
real mem  = 66678784 (63MB)
avail mem = 55009280 (52MB)
mainbus0 at root
bios0 at mainbus0: AT/286+ BIOS, date 02/03/97, BIOS32 rev. 0 @ 0xf705f
apm0 at bios0: Power Management spec V1.2 (BIOS management disabled)
apm0: APM power management enable: unrecognized device ID (9)
apm0: AC on, battery charge unknown
pcibios0 at bios0: rev 2.1 @ 0xf6fa0/0x990
pcibios0: PCI BIOS has 7 Interrupt Routing table entries
pcibios0: PCI Exclusive IRQs: 9
pcibios0: PCI Interrupt Router at 000:15:0 ("Intel 82371SB ISA" rev
0x00)
pcibios0: PCI bus #0 is the last bus
bios0: ROM list: 0xc0000/0x8000
cpu0 at mainbus0
pci0 at mainbus0 bus 0: configuration mode 1 (bios)
pchb0 at pci0 dev 0 function 0 "Intel 82439HX" rev 0x03
rl0 at pci0 dev 7 function 0 "Realtek 8139" rev 0x10: irq 9, address
00:50:fc:27:3d:ea
rlphy0 at rl0 phy 0: RTL internal PHY
vga1 at pci0 dev 13 function 0 "S3 Trio64V2/DX" rev 0x04
wsdisplay0 at vga1 mux 1: console (80x25, vt100 emulation)
wsdisplay0: screen 1-5 added (80x25, vt100 emulation)
drm at vga1 unsupported
pcib0 at pci0 dev 15 function 0 "Intel 82371SB ISA" rev 0x01
pciide0 at pci0 dev 15 function 1 "Intel 82371SB IDE" rev 0x00: DMA,
channel 0 wired to compatibility, channel 1 wired to compatibility
wd0 at pciide0 channel 0 drive 0: <ST32132A>
wd0: 16-sector PIO, LBA, 2015MB, 4127760 sectors
wd0(pciide0:0:0): using PIO mode 4, DMA mode 2
atapiscsi0 at pciide0 channel 1 drive 0
scsibus0 at atapiscsi0: 2 targets, initiator 7
cd0 at scsibus0 targ 0 lun 0: <TEAC, CD-516E, 1.0C> ATAPI 5/cdrom
removable
cd0(pciide0:1:0): using PIO mode 4, DMA mode 2
isa0 at pcib0
isadma0 at isa0
com0 at isa0 port 0x3f8/8 irq 4: ns16550a, 16 byte fifo
com1 at isa0 port 0x2f8/8 irq 3: ns16550a, 16 byte fifo
pckbc0 at isa0 port 0x60/5
pckbd0 at pckbc0 (kbd slot)
pckbc0: using irq 1 for kbd slot
wskbd0 at pckbd0: console keyboard, using wsdisplay0
pcppi0 at isa0 port 0x61
midi0 at pcppi0: <PC speaker>
spkr0 at pcppi0
lpt0 at isa0 port 0x378/4 irq 7
npx0 at isa0 port 0xf0/16: reported by CPUID; using exception 16
fdc0 at isa0 port 0x3f0/6 irq 6 drq 2
fd0 at fdc0 drive 0: 1.44MB 80 cyl, 2 head, 18 sec
biomask fd65 netmask ff65 ttymask ffff
softraid0 at root
root on wd0a swap on wd0b dump on wd0b
WARNING: / was not properly unmounted
-----------------<snip>-------------------
Box B dmesg (aprox 90M swap):

OpenBSD 4.4-stable (GENERIC) #27: Wed Nov 19 23:38:45 CST 2008
    k...@forge.meme.com:/usr/src/sys/arch/i386/compile/GENERIC
cpu0: Intel Pentium (P54C) ("GenuineIntel" 586-class) 100 MHz
cpu0: FPU,V86,DE,PSE,TSC,MSR,MCE,CX8
cpu0: F00F bug workaround installed
real mem  = 66678784 (63MB)
avail mem = 55009280 (52MB)
mainbus0 at root
bios0 at mainbus0: AT/286+ BIOS, date 06/27/97, BIOS32 rev. 0 @ 0xf849d
apm0 at bios0: Power Management spec V1.1 (BIOS management disabled)
apm0: APM power management enable: unrecognized device ID (9)
apm0: AC on, battery charge unknown
pcibios0 at bios0: rev 2.1 @ 0xf83b0/0x920
pcibios0: PCI BIOS has 5 Interrupt Routing table entries
pcibios0: PCI Exclusive IRQs: 9
pcibios0: PCI Interrupt Router at 000:15:0 ("Intel 82371FB ISA" rev
0x00)
pcibios0: PCI bus #0 is the last bus
bios0: ROM list: 0xc0000/0x8000
cpu0 at mainbus0
pci0 at mainbus0 bus 0: configuration mode 1 (bios)
pchb0 at pci0 dev 0 function 0 "Intel 82437FX" rev 0x02
rl0 at pci0 dev 7 function 0 "Realtek 8139" rev 0x10: irq 9, address
00:50:fc:4e:9b:5b
rlphy0 at rl0 phy 0: RTL internal PHY
vga1 at pci0 dev 13 function 0 "S3 Trio32/64" rev 0x00
wsdisplay0 at vga1 mux 1: console (80x25, vt100 emulation)
wsdisplay0: screen 1-5 added (80x25, vt100 emulation)
drm at vga1 unsupported
pcib0 at pci0 dev 15 function 0 "Intel 82371FB ISA" rev 0x02
pciide0 at pci0 dev 15 function 1 "Intel 82371FB IDE" rev 0x02: DMA,
channel 0 wired to compatibility, channel 1 wired to compatibility
wd0 at pciide0 channel 0 drive 0: <QUANTUM FIREBALL_TM1280A>
wd0: 16-sector PIO, LBA, 1222MB, 2503872 sectors
wd0(pciide0:0:0): using PIO mode 4, DMA mode 2
atapiscsi0 at pciide0 channel 1 drive 0
scsibus0 at atapiscsi0: 2 targets, initiator 7
wdc_atapi_intr: warning: reading only 14 of 18 bytes
wdc_atapi_intr: warning: reading only 0 of 36 bytes
wdc_atapi_intr: warning: reading only 14 of 18 bytes
cd0 at scsibus0 targ 0 lun 0: <NEC, CD-ROM DRIVE:273, 4.26> ATAPI
5/cdrom removable
cd0(pciide0:1:0): using PIO mode 3
isa0 at pcib0
isadma0 at isa0
com0 at isa0 port 0x3f8/8 irq 4: ns16550a, 16 byte fifo
com1 at isa0 port 0x2f8/8 irq 3: ns16550a, 16 byte fifo
pckbc0 at isa0 port 0x60/5
pckbd0 at pckbc0 (kbd slot)
pckbc0: using irq 1 for kbd slot
wskbd0 at pckbd0: console keyboard, using wsdisplay0
pcppi0 at isa0 port 0x61
midi0 at pcppi0: <PC speaker>
spkr0 at pcppi0
lpt0 at isa0 port 0x378/4 irq 7
npx0 at isa0 port 0xf0/16: reported by CPUID; using exception 16
fdc0 at isa0 port 0x3f0/6 irq 6 drq 2
fd0 at fdc0 drive 0: 1.44MB 80 cyl, 2 head, 18 sec
biomask fd65 netmask ff65 ttymask ffff
softraid0 at root
root on wd0a swap on wd0b dump on wd0b
WARNING: / was not properly unmounted
-----------------<snip>-------------------

Karl <k...@meme.com>
Free Software:  "You don't pay back, you pay forward."
                 -- Robert A. Heinlein

Reply via email to