I recently came into possession of a brand new, multi-processor
machine with the specific intention of running a number of domUs on
it simultaneously. Unfortunately, the dom0 kernel locks up in
predictable fashion, with repeatable results, completely crashing the
machine.
Machine Specs:
Dual Xeon 3.2 processors
Dell Poweredge case
2 gigabit ethernet cards
from lspci:
0000:06:07.0 Ethernet controller: Intel Corp. 82541GI/PI Gigabit
Ethernet Controller (rev 05)
Subsystem: Dell: Unknown device 016d
Control: I/O+ Mem+ BusMaster+ SpecCycle- MemWINV+ VGASnoop-
ParErr- Stepping- SERR+ FastB2B-
Status: Cap+ 66MHz+ UDF- FastB2B- ParErr- DEVSEL=medium
>TAbort- <TAbort- <MAbort- >SERR- <PERR-
Latency: 32 (63750ns min), Cache Line Size: 0x10 (64 bytes)
Interrupt: pin A routed to IRQ 209
Region 0: Memory at fe6e0000 (32-bit, non-prefetchable)
[size=128K]
Region 2: I/O ports at dcc0 [size=64]
Capabilities: <available only to root>
0000:07:08.0 Ethernet controller: Intel Corp. 82541GI/PI Gigabit
Ethernet Controller (rev 05)
Subsystem: Dell: Unknown device 016d
Control: I/O+ Mem+ BusMaster+ SpecCycle- MemWINV+ VGASnoop-
ParErr- Stepping- SERR+ FastB2B-
Status: Cap+ 66MHz+ UDF- FastB2B- ParErr- DEVSEL=medium
>TAbort- <TAbort- <MAbort- >SERR- <PERR-
Latency: 32 (63750ns min), Cache Line Size: 0x10 (64 bytes)
Interrupt: pin A routed to IRQ 217
Region 0: Memory at fe4e0000 (32-bit, non-prefetchable)
[size=128K]
Region 2: I/O ports at ccc0 [size=64]
Capabilities: <available only to root>
Build process (used both the 3.0 kernel and 3.0-testing), with the
exact same results:
untar
cd into linux-2.6.12, make ARCH=xen menuconfig
changed the processor type to xeon, and enabled SMP (i've done with
without smp, with the same results)
cd ..
make world
make install
the grub entry is correct, kernel boots fine
Then, when i try to transfer a file via scp to/from the machine, the
transfer starts off fine (9.9MB/s), then stalls completely as the
dom0 kernel dies. This only seems to happen in response to network
traffic which lasts longer than a second. Unfortunately, since the
kernel dies quickly, I haven't been able to try too many other
behaviors. However, network transfer kills the kernel each time, as
long as it takes more than 15 seconds to transfer.
It's a server, and when i plug in a monitor and keyboard to check the
screen, nothing has appeared, and the screen doesn't respond to
keyboard input. Almost seems like a problem with interrupt handling,
but that's a bit of an unfounded guess.
This machine is running debian sarge. All the debian sarge kernels
work just fine (smp, non-smp, etc.). It seems to be only a Xen issue.
Any more information needed? Any suggestions?
_______________________________________________
Xen-devel mailing list
Xen-devel@xxxxxxxxxxxxxxxxxxx
http://lists.xensource.com/xen-devel
|
Home