[Qemu-devel] kqemu in x86_64: (host) exception 0x0d in monitor space

[J M Cerqueira Esteves]

Greetings

Summary: qemu-system-x86_64 with kqemu (running under Ubuntu on a Athlon
64) crashes while installing a guest Debian amd64 testing (etch) system,
with the host reporting (in kernel logs):
  kqemu: aborting: Unexpected exception 0x0d in monitor space


Host CPU: AMD Athlon 64 3500+ (machine: HP dx5150 MT)
Host operating system: Ubuntu 6.06 LTS
Host kernel: one of the Ubuntu pre-packaged ones,
             2.6.15-26-amd64-k8 (SMP PREEMPT)

VDE: 'backported' (just rebuilding the package)
     from Debian testing's vde 1.5.11-1.
QEMU: 0.8.2, configured with -cc=gcc-3.4 --enable-alsa
kqemu: 1.3.0pre9

I tried to install Debian amd64 testing (etch) from a
snapshot netinst iso image downloaded yesterday, invoking

    vdeq qemu-system-x86_64 \
        -pidfile /srv/qemu/nisaba.pid \
        -m 160 \
        -net nic,vlan=0,model=rtl8139,macaddr=4A:4D:23:00:00:01 \
        -net vde,vlan=0,sock=/var/run/vde/tap-vde-1.ctl \
        -hda /srv/qemu/$NAME.qcow \
        -cdrom /srv/ark/cd/debian-testing-amd64-netinst-20060810.iso \
        -boot d

Booted in expert mode, chose language, keyboard layout, country, locale
parameters, and just after I chose "detect and mount cdrom" qemu crashed
(apparently immediately after (very briefly) showing a progress bar with
"detecting hardware to find cd-rom drives"), with the (host-side) output

ES =0000 0000000000000000 00000000 00000000
CS =0033 0000000000000000 ffffffff 00affa00
SS =002b 0000000000000000 ffffffff 00cff200
DS =0000 0000000000000000 00000000 00000000
FS =0000 0000000000000000 00000000 00000000
GS =0000 0000000000000000 00000000 00000000
LDT=0000 0000000000000000 00000000 00008000
TR =0040 ffffffff8030e000 0000206f 80008930
GDT=     ffffffff8030c000 00000080
IDT=     ffffffff8030d000 00001000
CR0=8005003b CR2=00002b599766f800 CR3=00000000074c4000 CR4=000006e0
Unsupported return value: 0xffffffff

In a second attempt I got

RAX=00002b80af1d7d20 RBX=00002b80af1d49e8 RCX=0000000000000008
RDX=0000000000000008
RSI=00002b80af393800 RDI=000000000053f478 RBP=00007fffff9fa2c0
RSP=00007fffff9fa1d8
R8 =00002b80af393800 R9 =0000000000000000 R10=000000000053f478
R11=0000000000000002
R12=0000000000000000 R13=0000000000000005 R14=00002b80af0d54b0
R15=0000000000402a18
RIP=00002b80af0ce390 RFL=00010287 [--S--PC] CPL=3 II=0 A20=1 HLT=0
ES =0000 0000000000000000 00000000 00000000
CS =0033 0000000000000000 ffffffff 00affa00
SS =002b 0000000000000000 ffffffff 00cff200
DS =0000 0000000000000000 00000000 00000000
FS =0000 0000000000000000 00000000 00000000
GS =0000 0000000000000000 00000000 00000000
LDT=0000 0000000000000000 00000000 00008000
TR =0040 ffffffff8030e000 0000206f 80008930
GDT=     ffffffff8030c000 00000080
IDT=     ffffffff8030d000 00001000
CR0=8005003b CR2=00002b80af393800 CR3=0000000007b48000 CR4=000006e0
Unsupported return value: 0xffffffff


For every such test, the host's dmesg and kernel logs reported the
following:

kqemu: aborting: Unexpected exception 0x0d in monitor space
err=0000 CS:EIP=f180:00000000f0002806 SS:SP=0000:00000000f00c7e00


The above crash does not happen when qemu-system-x86_64 is invoked with
the additional option "-no-kqemu".

In case this issue is already known: is there any way to avoid this
crash (maybe some boot time parameter for the Debian guest kernel?)
without disabling kqemu?

Any suggestions for additional information gathering here which could
help solve this issue?


Best regards (and *many* thanks for QEMU)

                        J Esteves


-- 
+351 939838775   Skype:jmcerqueira   http://del.icio.us/jmce