| Summary: | segfault in ovs thread when using more than one VM | ||||||
|---|---|---|---|---|---|---|---|
| Product: | Red Hat Enterprise Linux 7 | Reporter: | Andrew Theurer <atheurer> | ||||
| Component: | openvswitch | Assignee: | Kevin Traynor <ktraynor> | ||||
| Status: | CLOSED INSUFFICIENT_DATA | QA Contact: | ovs-qe | ||||
| Severity: | urgent | Docs Contact: | |||||
| Priority: | urgent | ||||||
| Version: | 7.3 | CC: | atheurer, atragler, fleitner, rkhan | ||||
| Target Milestone: | rc | ||||||
| Target Release: | --- | ||||||
| Hardware: | Unspecified | ||||||
| OS: | Unspecified | ||||||
| Whiteboard: | |||||||
| Fixed In Version: | Doc Type: | If docs needed, set a value | |||||
| Doc Text: | Story Points: | --- | |||||
| Clone Of: | Environment: | ||||||
| Last Closed: | 2017-02-17 18:02:48 UTC | Type: | Bug | ||||
| Regression: | --- | Mount Type: | --- | ||||
| Documentation: | --- | CRM: | |||||
| Verified Versions: | Category: | --- | |||||
| oVirt Team: | --- | RHEL 7.3 requirements from Atomic Host: | |||||
| Cloudforms Team: | --- | Target Upstream Version: | |||||
| Attachments: |
|
||||||
Kevin, This might be related to your other vhost fixes. Can you test with openvswitch-2.5.0-23.git20160727.el7fdb https://brewweb.engineering.redhat.com/brew/buildinfo?buildID=526711 |
Created attachment 1221652 [details] log file for ovs-vswitchd Description of problem: When starting testpmd in a VM, one of the threads from ovs-vswitchd segfaults with: [ 4535.253497] vhost_thread2[4613]: segfault at 18 ip 00007fe9e07a9af0 sp 00007fe9d1ee0638 error 4 in ovs-vswitchd[7fe9e06c9000+319000] Version-Release number of selected component (if applicable): RHEL7.3 openvswitch-2.5.0-14.git20160727.el7fdp How reproducible: Steps to Reproduce: Configure ovs with the following: ovsbr0: dpdk0 vhost-user-1 ovsbr1: dpdk1 vhostuser-2 vhostuser-3 Create 5 PMD threads, 1 for each port. create vm1 with 2 vhostuser ports, vhost-user1 and vhost-user-2 create vm2 with 1 vhostuser ports, vhost-user-3 start both VMs on vm2, bring up the interface with kernel driver (virtio) on vm1, run testpmd on the two vhostuser interfaces Actual results: Once vm1 starts testpmd, the segfault will happen in the host. If vm2 is not started, the segfault will not happen. Expected results: There should be no segfault Additional info: