Description of problem: Executing virsh nodedev-list will cause libvirt to go into an infinite loop. Version-Release number of selected component (if applicable): libvirt-0.6.3-13.el5 RHEL5.4-Server/Client-KVM RHEL5.4-Server-Xen How reproducible: 100% Steps to Reproduce: execute: # virsh nodedev-list computer Actual results: hang Expected results: List out all node devices Additional info:
Update Steps to Reproduce: execute: # virsh nodedev-list --tree
I really can't reproduce this: [root@test2 ~]# virsh nodedev-list --tree computer | +-pci_8086_104c | | | +-net_00_19_d1_2d_c0_45 | +-pci_8086_244e +-pci_8086_2812 +-pci_8086_2822 | | | +-pci_8086_2822_scsi_host | +-pci_8086_2822_scsi_host_0 | +-pci_8086_2822_scsi_host_1 | +-pci_8086_2822_scsi_host_2 | +-pci_8086_2822_scsi_host_3 | | | ... +-pci_8086_29a2 +-pci_8086_29a3 [root@test2 ~]# rpm -q libvirt libvirt-0.6.3-13.el5 [root@test2 ~]# There is something missing here... note that virsh nodedev-dumpxml computer hangs on that box, it's bug https://bugzilla.redhat.com/show_bug.cgi?id=509261 for which we have a fix in the pipeline, but that one I can't reproduce ! Can you make sure you didn't mix the 2 issues, if no can you run libvirtd under valgrind or gdb and then rerun the command, then report your findings, thanks Daniel
*** Bug 509952 has been marked as a duplicate of this bug. ***
Only run command # virsh nodedev-list doesn't have this issue If run command # virsh nodedev-list after run command # virsh nodedev-dumpxml computer will cause this issue.
Okay then it's a dup of 509261 which is fixed in libvirt-0.6.3-14.el5 and should be on QA at this point. Just make sure you can't reproduce the issue with that version, thanks, Daniel *** This bug has been marked as a duplicate of bug 509261 ***