Hide Forgot
description of problem: domain cannot start with virtio and serial console Version-Release number of selected component (if applicable): libvirt-1.2.8-9.el7.x86_64 How reproducible: 100% Steps to Reproduce: 1.notice the sequence of two console # virsh dumpxml test3 <serial type='pty'> <target type='isa-serial' port='0'/> </serial> <console type='pty'> <target type='virtio' port='0'/> </console> <console type='pty'> <target type='serial' port='1'/> </console> 2. # virsh start test3 error: Failed to start domain test3 error: internal error: no assigned pty for device charconsole1 Actual results: Cannot start if virtio console before serial console, but can start if after it. # virsh start test3 Domain test3 started # virsh dumpxml test3 <serial type='pty'> <source path='/dev/pts/8'/> <target type='isa-serial' port='0'/> <alias name='serial0'/> </serial> <console type='pty' tty='/dev/pts/8'> <source path='/dev/pts/8'/> <target type='serial' port='0'/> <alias name='serial0'/> </console> <console type='pty'> <source path='/dev/pts/9'/> <target type='virtio' port='1'/> <alias name='console1'/> </console> Expected results: Can start or if we do not support this should give a check before call qemuProcessLookupPTYs and give a useful error. Additional info: After check the code maybe the cause is in qemuProcessFindCharDevicePTYsMonitor: /* For historical reasons, console[0] can be just an alias * for serial[0]. That's why we need to update it as well. */ if (vm->def->nconsoles) { virDomainChrDefPtr chr = vm->def->consoles[0]; if (vm->def->nserials && libvirt only check if the first console is serial, but when set virtio console be the first console libvirt do the wrong things here, but qemu can recognize this sequence, so the second console cannot get a 'entry' in qemuProcessLookupPTYs.
Fixed upstream: commit d85dcae4b2bb7277503001519e533d21133f65d4 Author: Erik Skultety <eskultet> Date: Thu Dec 11 12:28:22 2014 +0100 conf: move the check for secondary consoles of targetType serial For historical reasons, only the first <console> element might be of targetType serial, but we checked for other consoles of targetType serial in our post-parse callback if and only if we knew the first console was serial, otherwise the check was skipped. This patch moves the check one level up, so first the check for secondary console of type serial is performed and then the rest of operations continue unchanged. v1.2.11-6-gd85dcae
Verified as below: [root@localhost ~]# rpm -q libvirt qemu-kvm-rhev libvirt-1.2.15-1.el7.x86_64 qemu-kvm-rhev-2.3.0-1.el7.x86_64 Add below lines to domain xml <serial type='pty'> <target type='isa-serial' port='0'/> </serial> <console type='pty'> <target type='virtio' port='0'/> </console> <console type='pty'> <target type='serial' port='1'/> </console> Then, save the xml in "virsh edit" [root@localhost ~]# virsh edit r71 error: unsupported configuration: Only the first console can be a serial port Failed. Try again? [y,n,i,f,?]: The expected error message will be toasted.
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://rhn.redhat.com/errata/RHBA-2015-2202.html