Bug 965711

I experience the same and I suspect a (host) kernel bug. If I boot 3.9.2-200.fc18.x86_64 on the VM host it doesn't work but using kernel-3.8.11-200.fc18.x86_64 works.

When I start an existing CentOS 5 VM this is the last thing I see (after grub):
...
root (hd0,0)
 Filesystem type is ext2fs, partition type 0x83
kernel /vmlinux-2.6.18-348.1.1.el5 ro root=... clocksource=acpi_pm
   [Linux-bzimage, setup=0x1e00, size=0x2050fc]
initrd /initrd-2.6.18-348.1.1.el5.img
   [Linux-initrd @ 0x37c98000, 0x357ed0 bytes]

It seems to be an upstream but or at least the same problem exists on Arch Linux:
https://bugs.archlinux.org/task/35247

Still broken with kernel-3.9.4-200.fc18.x86_64.

Reassigning to kernel.

I can confirm that there is a difference between

qemu-system-x86_64 -cdrom CentOS-5.9-x86_64-netinstall.iso

and

qemu-system-x86_64 -cdrom CentOS-5.9-x86_64-netinstall.iso -enable-kvm

meaning that the latter exhibits the problem. Setup is Debian sid or testing:

ii  qemu-system-x86       1.5.0+dfsg-4    i386            QEMU full system emulation binaries (x86)
Linux tglase.lan.tarent.de 3.9-1-amd64 #1 SMP Debian 3.9.6-1 i686 GNU/Linux
Linux ldegen-wirt.lan.tarent.de 3.9-1-amd64 #1 SMP Debian 3.9.6-1 x86_64 GNU/Linux

A “ceteris paribus” reboot into the stable Debian kernel makes things work with -enable-kvm (and blazingly fast compared to without it):

Linux ldegen-wirt.lan.tarent.de 3.2.0-4-amd64 #1 SMP Debian 3.2.46-1 x86_64 GNU/Linux

I’ll also prod the Debian Kernel people. Our primary symptom was, in case someone is interested, a Windows® VM with virtio drivers no longer coming up (“crc error
 system halted” with enough debug level bumping); my co-admin referenced this bugreport.

I suspect this might also be related to the host CPU. I can run a CentOS VM (or the netinstall) just fine on a quite recent Intel machine but not on an old Intel laptop of mine. The libvirt config is identical as well as the Fedora patch level/installed rpms.

working machine:
  Intel(R) Core(TM) i5-2400 CPU @ 3.10GHz
  4 cores
  cpu family 6
  model 42

not working:
  Intel(R) Core(TM)2 Duo CPU     T7250  @ 2.00GHz
  2 cores
  cpu family 6
  model 15

Probably time for some kernel bisecting?

Not quite…

Not working #1:

processor       : 7
vendor_id       : GenuineIntel
cpu family      : 6
model           : 26
model name      : Intel(R) Core(TM) i7 CPU         950  @ 3.07GHz
stepping        : 5
microcode       : 0x16
cpu MHz         : 1596.000
cache size      : 8192 KB
physical id     : 0
siblings        : 8
core id         : 3
cpu cores       : 4
apicid          : 7
initial apicid  : 7
fpu             : yes
fpu_exception   : yes
cpuid level     : 11
wp              : yes
flags           : fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov pat pse36 clflush dts acpi mmx fxsr sse sse2 ss ht tm pbe syscall nx rdtscp lm constant_tsc arch_perfmon pebs bts rep_good nopl xtopology nonstop_tsc aperfmperf pni dtes64 monitor ds_cpl vmx est tm2 ssse3 cx16 xtpr pdcm sse4_1 sse4_2 popcnt lahf_lm ida dtherm tpr_shadow vnmi flexpriority ept vpid
bogomips        : 6133.32
clflush size    : 64
cache_alignment : 64
address sizes   : 36 bits physical, 48 bits virtual
power management:


Not working #2:

processor       : 7
vendor_id       : GenuineIntel
cpu family      : 6
model           : 26
model name      : Intel(R) Core(TM) i7 CPU         950  @ 3.07GHz
stepping        : 5
microcode       : 0x11
cpu MHz         : 3060.000
cache size      : 8192 KB
physical id     : 0
siblings        : 8
core id         : 3
cpu cores       : 4
apicid          : 7
initial apicid  : 7
fpu             : yes
fpu_exception   : yes
cpuid level     : 11
wp              : yes
flags           : fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov pat pse36 clflush dts acpi mmx fxsr sse sse2 ss ht tm pbe syscall nx rdtscp lm constant_tsc arch_perfmon pebs bts rep_good nopl xtopology nonstop_tsc aperfmperf pni dtes64 monitor ds_cpl vmx est tm2 ssse3 cx16 xtpr pdcm sse4_1 sse4_2 popcnt lahf_lm ida dtherm tpr_shadow vnmi flexpriority ept vpid
bogomips        : 6128.71
clflush size    : 64
cache_alignment : 64
address sizes   : 36 bits physical, 48 bits virtual
power management:

Pretty sure this is a dup, please reopen if I'm wrong.

*** This bug has been marked as a duplicate of bug 967652 ***