ocata: I was able to reproduce the regression by rebooting the guest - I could not reproduce it any other way, by attaching/removing interface(s), setting their link up/down from the host side or from the guest side, adding/removing them to a bridge on the host, etc. On reboot, with qemu 1:2.8+dfsg-3ubuntu2.9~cloud5.1, after 2 reboots I was able to reproduce the guest being unable to use the interface. I stopped the guest, upgraded to 1:2.8+dfsg-3ubuntu2.9~cloud6, and repeated the test, and the guest had no problems even after 10 reboots.
** Description changed: [impact] this bug is to track re-uploading the fix for bug 1823458 plus a patch to fix a regression it introduced. instead of copying the details from bug 1823458, please see that bug for impact and testcase. [test case] - see bug 1823458 + see bug 1823458 for the original bug test case + + for the regression test case from bug 1829245, rebooting the guest will + result in the guest being unable to use the interface (possibly a couple + reboots are needed to reproduce if the interface still works after the + first reboot). [regression potential] see bug 1823458 additionally, the regression from the last fix is in bug 1829245. the change is in code where this still has the regression potential of causing guest networking to fail. ** Tags removed: verification-ocata-needed ** Tags added: verification-ocata-done -- You received this bug notification because you are a member of Ubuntu Bugs, which is subscribed to Ubuntu. https://bugs.launchpad.net/bugs/1829380 Title: race condition between vhost_net_stop and CHR_EVENT_CLOSED on shutdown crashes qemu (fix regression) To manage notifications about this bug go to: https://bugs.launchpad.net/cloud-archive/+bug/1829380/+subscriptions -- ubuntu-bugs mailing list [email protected] https://lists.ubuntu.com/mailman/listinfo/ubuntu-bugs
