Hide Forgot
Description of problem: Boot rhel7.0 guest with a VF assigned, then fail to shutdown the guest. Version-Release number of selected component (if applicable): kernel: 3.10.0-57.el7.x86_64 qemu-kvm: qemu-kvm-rhev-1.5.3-20.el7.x86_64 How reproducible: 100% Steps to Reproduce: 1. Bring up VFs and boot a guest /usr/libexec/qemu-kvm -M q35 -cpu SandyBridge -enable-kvm -m 4G \ -smp 1,sockets=1,cores=2,threads=1 -name test \ -rtc base=utc,clock=host,driftfix=slew -k en-us -boot menu=on -vga qxl \ -spice disable-ticketing,port=5931 -monitor stdio -qmp tcp:0:4444,server,nowait \ -device ioh3420,bus=pcie.0,id=root.0 -device x3130-upstream,bus=root.0,id=upstream -device xio3130-downstream,bus=upstream,id=downstream0,chassis=1 \ -drive file=/home/RHEL-Server-7.0-64.raw,if=none,id=drive-system-disk,media=disk,format=raw,aio=native,werror=stop,rerror=stop -device virtio-blk-pci,bus=downstream0,drive=drive-system-disk,id=system-disk,bootindex=1 \ -device xio3130-downstream,bus=upstream,id=downstream2,chassis=2 -net none -device vfio-pci,host=09:04.3,id=hostnet_VF3 \ -serial unix:/tmp/ttyS0,server,nowait 2. Shutdown the guest: Hang at "Actual results" 3. Actual results: Fail to shutdown the guest: [ 54.560539] Bad IO access at port 0xac (return inl(port)) [ 56.561120] Bad IO access at port 0xac (return inl(port)) [ 57.662415] pciehp 0000:00:02.0:pcie04: Device 0000:01:00.0 already exists at 0000:01:00, cannot hot-add [ 57.662417] pciehp 0000:00:02.0:pcie04: Cannot add device at 0000:01:00 [ 58.562020] Bad IO access at port 0xac (return inl(port)) [ 60.563077] Bad IO access at port 0xac (return inl(port)) [ 62.564146] Bad IO access at port 0xac (return inl(port)) [ 62.767556] pciehp 0000:02:00.0:pcie24: Device 0000:03:00.0 already exists at 0000:03:00, cannot hot-add [ 62.767557] pciehp 0000:02:00.0:pcie24: Cannot add device at 0000:03:00 [ 64.565160] Bad IO access at port 0xac (return inl(port)) [ 66.566148] Bad IO access at port 0xac (return inl(port)) [ 68.567039] Bad IO access at port 0xac (return inl(port)) [ 70.568050] Bad IO access at port 0xac (return inl(port)) [ 72.569040] Bad IO access at port 0xac (return inl(port)) [ 114.590184] be2net 0000:00:03.0: POST timeout; stage=0xffff [ 114.590190] dpm_run_callback(): pci_pm_resume+0x0/0xb0 returns -1 [ 114.590194] PM: Device 0000:00:03.0 failed to resume async: error -1 Cannot finalize remaining file systems and devices, giving up. dracut Warning: Killing all remaining processes Powering off. [ 160.068240] be2net 0000:00:03.0: FW not responding [ 160.072997] Power down. Expected results: Succeed to shutdown the guest. Additional info: Succeed to shutdown windows guest.
What is the device being assigned? Does the shutdown problem only happen with that VF or is it reproducible with other VFs (82576/82599)? Does this problem only happen on q35 guests?
Q35: Retest this issue with 3.10.0-54.el7.x86_64 kernel guest, After I finished to install the guest without PF assigned, First Inside guest with power can succeed to shutdown the guest. But second in qmp monitor，After system_powerdown command, the guest stop at login screen, Inside guest with poweroff, failed to shutdown the guest. Maybe regardless of PF device assigned. the command line: /usr/libexec/qemu-kvm -M q35 -cpu SandyBridge -enable-kvm -m 4G \ -smp 1,sockets=1,cores=2,threads=1 -name test -rtc base=utc,clock=host,driftfix=slew -k en-us -boot menu=on -vga qxl \ -spice disable-ticketing,port=5931 -monitor stdio -qmp tcp:0:4444,server,nowait -device ioh3420,bus=pcie.0,id=root.0 \ -device x3130-upstream,bus=root.0,id=upstream -device xio3130-downstream,bus=upstream,id=downstream0,chassis=1 -drive file=/home/RHEL7.0-64.raw,if=none,id=drive-system-disk,media=disk,format=raw,aio=native,werror=stop,rerror=stop \ -device virtio-blk-pci,bus=downstream0,drive=drive-system-disk,id=system-disk,bootindex=1 -device xio3130-downstream,bus=upstream,id=downstream2,chassis=2 \ -device virtio-net-pci,netdev=hostnet0,id=net0,bus=downstream2,mac=52:54:00:13:10:20 -netdev tap,id=hostnet0,vhost=on,script=/etc/qemu-ifup \ -serial unix:/tmp/ttyS0,server,nowait Actual Result: Hang at this stage: [root@unused ~]# [ 165.847127] pciehp 0000:00:02.0:pcie04: Device 0000:01:00.0 already exists at 0000:01:00, cannot hot-add [ 165.847129] pciehp 0000:00:02.0:pcie04: Cannot add device at 0000:01:00 [ 170.949180] pciehp 0000:02:00.0:pcie24: Device 0000:03:00.0 already exists at 0000:03:00, cannot hot-add [ 170.949182] pciehp 0000:02:00.0:pcie24: Cannot add device at 0000:03:00 [ 170.949226] pciehp 0000:02:01.0:pcie24: Device 0000:04:00.0 already exists at 0000:04:00, cannot hot-add [ 170.949227] pciehp 0000:02:01.0:pcie24: Cannot add device at 0000:04:00 [ OK ] Stopped LSB: Start the ipr init daemon. ................................. Cannot finalize remaining file systems and devices, giving up. dracut Warning: Killing all remaining processes Powering off. [ 208.919331] Power down. pc-i440fx-rhel7.0.0: Hit the same result, but hang without any information in nc command monitor.
(In reply to Alex Williamson from comment #2) > What is the device being assigned? > > Does the shutdown problem only happen with that VF or is it reproducible > with other VFs (82576/82599)?
without VF can not shutdown the guest.
(In reply to zhonglinzhang from comment #5) > without VF can not shutdown the guest. Are you saying this happens even without device assignment of any kind?
(In reply to Alex Williamson from comment #6) > (In reply to zhonglinzhang from comment #5) > > without VF can not shutdown the guest. > > Are you saying this happens even without device assignment of any kind? Yes, system_powerdown in qmp monitor, stop at login screen. Inside guest power off after login in guest, then hang.
Changing title and re-assigning back to default assignee since this has nothing to do with device assignment.
(In reply to Alex Williamson from comment #8) > Changing title and re-assigning back to default assignee since this has > nothing to do with device assignment. This bug maybe duplicate to bug 1040307 - Qemu monitor "system_powerdown" command fail to power down guest Best Regards, sluo
Hi Sluo, Can you do some investigation and test and update comment13? Best Regards, Junyi
(In reply to juzhang from comment #14) > Hi Sluo, > > Can you do some investigation and test and update comment13? > > Best Regards, > Junyi Retried this issue with the same command as comment #0 and comment #3 which just same results with bug 1040307. I think we can close it duplicate to bug 1040307. host info: # uname -r && rpm -q qemu-kvm 3.10.0-123.el7.x86_64 qemu-kvm-1.5.3-60.el7.x86_64 guest info: # uname -r 3.10.0-123.el7.x86_64 Best Regards, sluo
*** This bug has been marked as a duplicate of bug 1040307 ***