Red Hat Bugzilla – Bug 1425765
The guest failed to start with ich6 sound when machine type is rhel6.*.0
Last modified: 2017-08-01 23:37:31 EDT
Description of problem: The guest failed to start with ich6 sound when machine type is rhel6.*.0 Version-Release number of selected component (if applicable): qemu-kvm-rhev-2.8.0-4.el7.x86_64 How reproducible: 100% Steps to Reproduce: 1. on rhel7 host, define a rhel6 guest rhel6.6.0 machine type: <domain type='kvm'> <name>rhel6.8</name> ... <os> <type arch='x86_64' machine='rhel6.6.0'>hvm</type> <boot dev='hd'/> </os> ... </domain> 2. start the guest. # virsh start rhel6.8 Domain rhel6.8 started 3. Edit guest xml and add ich6 sound model into it: <sound model='ich6'> </sound> 4. start the guest again： # virsh start rhel6.8 error: Failed to start domain rhel6.8 error: internal error: qemu unexpectedly closed the monitor: Unexpected error in input_type_enum() at qapi/qapi-visit-core.c:364: 2017-02-22T13:36:20.207356Z qemu-kvm: -device intel-hda,id=sound0,bus=pci.0,addr=0x6: can't apply global intel-hda.msi=0: Invalid parameter '0' Actual results: As step 4, The guest failed to start with ich6 sound when machine type is rhel6.*.0. Expected results: should start the guest successfully. Additional info: 1.When machine type is 7.3.0, the guest with ich6 sound can be started. # virsh dumpxml V7.3-full ... <type arch='x86_64' machine='pc-i440fx-rhel7.3.0'>hvm</type> ... <sound model='ich6'> <address type='pci' domain='0x0000' bus='0x00' slot='0x07' function='0x0'/> </sound> ... # virsh start V7.3-full Domain V7.3-full started 2.Can not reproduce on rhel7.3.z host: # rpm -q qemu-kvm-rhev qemu-kvm-rhev-2.6.0-28.el7_3.6.x86_64 # virsh dumpxml rhel6.8 ... <type arch='x86_64' machine='rhel6.6.0'>hvm</type> ... <sound model='ich6'> <address type='pci' domain='0x0000' bus='0x00' slot='0x06' function='0x0'/> </sound> ... # virsh start rhel6.8 Domain rhel6.8 started
can also reproduce on qemu-kvm-rhev-2.8.0-5.el7.x86_64.
test build https://brewweb.engineering.redhat.com/brew/taskinfo?taskID=12614634
patch posted.
Test on qemu-kvm-rhev-2.8.0-5.el7.bz1425765.2.x86_64， the result is as expected: steps: 1.define a guest with following xml: # virsh dumpxml rhel6.8 <domain type='kvm'> <name>rhel6.8</name> ... <type arch='x86_64' machine='rhel6.6.0'>hvm</type> ... <sound model='ich6'> <address type='pci' domain='0x0000' bus='0x00' slot='0x06' function='0x0'/> </sound> ... 2.try to start the guest: # virsh start rhel6.8 Domain rhel6.8 started
Fix included in qemu-kvm-rhev-2.8.0-6.el7
Reproduce this issue against qemu-kvm-rhev-2.8.0-3.el7.x86_64 # /usr/libexec/qemu-kvm -machine rhel6.6.0 -device intel-hda Unexpected error in input_type_enum() at qapi/qapi-visit-core.c:364: qemu-kvm: -device intel-hda: can't apply global intel-hda.msi=0: Invalid parameter '0' Aborted (core dumped) Verify this issue agasint qemu-kvm-rhev-2.8.0-6.el7.x86_64, no qemu error happen when using same qemu cli
Verified per comment 9
Since the problem described in this bug report should be resolved in a recent advisory, it has been closed with a resolution of ERRATA. For information on the advisory, and where to find the updated files, follow the link below. If the solution does not work for you, open a new bug report. https://access.redhat.com/errata/RHSA-2017:2392