Red Hat Bugzilla – Bug 621161
qemu-kvm crashes with I/O Possible message
Last modified: 2013-01-09 17:59:04 EST
Created attachment 436510 [details]
strace of qemu-kvm
I'm facing one issue with dual head on winxp guest, please let me know if it's
related to this or if I should open separate bug for this issue.
Here are steps that I made before it happened
1) Clean install of WinXP into qemu-kvm
2) install rhev-tools into guest
3) reboot -> shutdown
4) sudo /usr/libexec/qemu-kvm -vga qxl -device qxl -m 512 -soundhw ac97 -spice
5) spicec -h localhost -p 3002
5) Right click on desktop -> Settings -> Extend display + set up both displays
(Same crash occurs on each boot of winxp guest ,removing -device qxl does not make any change)
qemu-kvm crashes with "I/O Possible" message
See attached strace output (strace of qemu-kvm with the same params as in Ad. 4) and qemu-kvm output
Version of software is:
Created attachment 436511 [details]
Please attach the crash log.
Try connecting gdb to qemu-kvm and getting a backtrace.
QEMU_AUDIO_DRV=pa (default) fails.
When using spice QEMU_AUDIO_DRV=spice should be used, otherwise forwarding audio to the spice client doesn't work. libvirt will do that for you btw.
I *think* one of the pulseaudio threads receives SIGIO (need to double-check that though). Doesn't looks spice-related at first glance. Nevertheless it triggers only with spice enabled.
Yes removing -soundhw ac97 does help
Patch posted to rhvirt-patches.
Verified on qemu-kvm-0.12.1.2-2.109.el6. Works for me now.
*** Bug 614915 has been marked as a duplicate of this bug. ***
Red Hat Enterprise Linux 6.0 is now available and should resolve
the problem described in this bug report. This report is therefore being closed
with a resolution of CURRENTRELEASE. You may reopen this bug report if the
solution does not work for you.