Hide Forgot
Description of problem: When selinux is enforcing, restore the saved domain will cause avc denial, so fail to restore. Version-Release number of selected component (if applicable): libvirt-0.8.1-2.el6.x86_64 kernel-2.6.32-24.el6.x86_64 qemu-kvm-0.12.1.2-2.51.el6.x86_64 # rpm -qa|grep selinux libselinux-2.0.90-3.el6.x86_64 libselinux-python-2.0.90-3.el6.x86_64 selinux-policy-3.7.19-12.el6.noarch libselinux-utils-2.0.90-3.el6.x86_64 selinux-policy-targeted-3.7.19-12.el6.noarch How reproducible: Always Steps to Reproduce: 1. Make sure selinux is enforcing # getenforce Enforcing 2. Save a running domain # virsh save demo /tmp/demo.save Domain demo saved to /tmp/demo.save # ll -Z /tmp/demo.save -rw-------. root root system_u:object_r:svirt_image_t:s0:c165,c811 /tmp/demo.save 3. Restore the saved domain # virsh restore /tmp/demo.save Domain restored from /tmp/demo.save # ps -efZ |grep kvm system_u:system_r:svirt_t:s0:c207,c818 qemu 1922 1 16 01:06 ? 00:00:33 /usr/libexec/qemu-kvm -S -M rhel6.0.0 -enable-kvm -m 1024 -smp 1,sockets=1,cores=1,threads=1 -name demo -uuid 7f6b41f0-2dea-51ae-9162-0b757ad010e6 -nodefaults -chardev socket,id=monitor,path=/var/lib/libvirt/qemu/demo.monitor,server,nowait -mon chardev=monitor,mode=control -rtc base=utc -boot c -drive file=/var/lib/libvirt/images/demo.img,if=none,id=drive-ide0-0-0,boot=on -device ide-drive,bus=ide.0,unit=0,drive=drive-ide0-0-0,id=ide0-0-0 -netdev tap,fd=25,id=hostnet0 -device rtl8139,netdev=hostnet0,id=net0,mac=52:54:00:0d:a5:09,bus=pci.0,addr=0x4 -chardev pty,id=serial0 -device isa-serial,chardev=serial0 -usb -vnc 127.0.0.1:0 -k en-us -vga cirrus -incoming exec:cat -device virtio-balloon-pci,id=balloon0,bus=pci.0,addr=0x3 Actual results: After virsh restore command, the domain will re-start automatically, not go back to the saved point. Fail to restore. In log file, the following avc denial is seen: type=AVC msg=audit(1273554370.058:70861): avc: denied { read } for pid=1922 comm="qemu-kvm" path="/tmp/demo.save" dev=sda6 ino=983110 scontext=system_u:system_r:svirt_t:s0:c207,c818 tcontext=system_u:object_r:svirt_image_t:s0:c165,c811 tclass=file type=SYSCALL msg=audit(1273554370.058:70861): arch=c000003e syscall=59 success=yes exit=0 a0=7f727046dd20 a1=7f7270470730 a2=7f727046dc30 a3=7f7284d25ef0 items=0 ppid=1 pid=1922 auid=500 uid=107 gid=107 euid=107 suid=107 fsuid=107 egid=107 sgid=107 fsgid=107 tty=(none) ses=1 comm="qemu-kvm" exe="/usr/libexec/qemu-kvm" subj=system_u:system_r:svirt_t:s0:c207,c818 key=(null) Expected results: After virsh restore command, the domain should go back to the saved point directly, restore successfully. Additional info:
This request was evaluated by Red Hat Product Management for inclusion in a Red Hat Enterprise Linux major release. Product Management has requested further review of this request by Red Hat Engineering, for potential inclusion in a Red Hat Enterprise Linux Major release. This request is not yet committed for inclusion.
KVM is running: scontext=system_u:system_r:svirt_t:s0:c207,c818 While the disk is labelled: tcontext=system_u:object_r:svirt_image_t:s0:c165,c811 So for some reason libvirt is failing to set the correct label at restore
On libvirt-0.8.1-7.el6 and the guest is rhel5.4 ,'save' and 'restore' successfully! My enviroment: [root@dhcp-65-73 /]# rpm -qa |grep libvirt libvirt-client-0.8.1-7.el6.x86_64 libvirt-python-0.8.1-7.el6.x86_64 libvirt-0.8.1-7.el6.x86_64 [root@dhcp-65-73 /]# rpm -qa |grep qemu qemu-img-0.12.1.2-2.68.el6.x86_64 gpxe-roms-qemu-0.9.7-6.3.el6.noarch qemu-kvm-0.12.1.2-2.68.el6.x86_64 [root@dhcp-65-73 /]# rpm -q kernel kernel-2.6.32-25.el6.x86_64 kernel-2.6.32-28.el6.x86_64 kernel-2.6.32-30.el6.x86_64 My procedure: 1. Start a VM # virsh start rhel5.4_64 2. In guest, issue some command, such as: ls, or ifconfig 3. Save the running vm, and check the context of the save file. # virsh save rhel5.4_64 /tmp/rhel5.4_64.save 5. Restore the vm # virsh restore /tmp/rhel5.4_64.save Then,"vm is restored successfully, and the output of the step 2 still is seen."
(In reply to comment #4) > On libvirt-0.8.1-7.el6 and the guest is rhel5.4 ,'save' and 'restore' > successfully! > My enviroment: > [root@dhcp-65-73 /]# rpm -qa |grep libvirt > libvirt-client-0.8.1-7.el6.x86_64 > libvirt-python-0.8.1-7.el6.x86_64 > libvirt-0.8.1-7.el6.x86_64 > [root@dhcp-65-73 /]# rpm -qa |grep qemu > qemu-img-0.12.1.2-2.68.el6.x86_64 > gpxe-roms-qemu-0.9.7-6.3.el6.noarch > qemu-kvm-0.12.1.2-2.68.el6.x86_64 > [root@dhcp-65-73 /]# rpm -q kernel > kernel-2.6.32-25.el6.x86_64 > kernel-2.6.32-28.el6.x86_64 > kernel-2.6.32-30.el6.x86_64 > > My procedure: > 1. Start a VM > > # virsh start rhel5.4_64 > > 2. In guest, issue some command, such as: ls, or ifconfig > > 3. Save the running vm, and check the context of the save file. > > # virsh save rhel5.4_64 /tmp/rhel5.4_64.save > > 5. Restore the vm > > # virsh restore /tmp/rhel5.4_64.save > > Then,"vm is restored successfully, and the output of the step 2 still is seen." Sorry! My below description is not clearly! If saved directory is my $HOME,'save ' procedure will failed! If saved directory is '/tmp','restore' will failed ,and start the guest normally,not my saved time action on my guest when I saved!
Although the example showed saving/restoring in /tmp, the problem existed for any directory. Patches verified to fix the problem have been submitted upstream. Waiting for ACK: https://www.redhat.com/archives/libvir-list/2010-June/msg00661.html
libvirt-0_8_1-11_el6 has been built in RHEL-6-candidate with the fix. Dave
Verified this defect with libvirt-0_8_1-11_el6, and PASSED. # virsh save win2008-virtio /tmp/my1.save Domain win2008-virtio saved to /tmp/my1.save # virsh restore /tmp/my1.save Domain restored from /tmp/my1.save Guest is running happily. # ll -Z /tmp/my1.save -rw-------. root root system_u:object_r:virt_content_t:s0 /tmp/my1.save
Verified this bug with RHEL6 RC build and it passed: libvirt-0.8.1-27.el6.x86_64 qemu-kvm-0.12.1.2-2.113.el6.x86_64 kernel-2.6.32-70.el6.x86_64
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.