Summary: |
networkd fails assertion on tap device going away |
Product: |
systemd
|
Reporter: |
Dave Reisner <d> |
Component: |
general | Assignee: |
systemd-bugs |
Status: |
RESOLVED
FIXED
|
QA Contact: |
systemd-bugs |
Severity: |
normal
|
|
|
Priority: |
medium
|
|
|
Version: |
unspecified | |
|
Hardware: |
x86-64 (AMD64) | |
|
OS: |
Linux (All) | |
|
Whiteboard: |
|
i915 platform:
|
|
i915 features:
|
|
Attachments: |
bt full
|
Use of freedesktop.org services, including Bugzilla, is subject to our Code of Conduct. How we collect and use information is described in our Privacy Policy.
Created attachment 101800 [details] bt full I can semi-reliably hit a failed assertion when shutting down a VM: tap0 : link 41 added tap0 : udev initializing link... tap0 : flags change: +MULTICAST +BROADCAST tap0 : MAC address: fe:8d:25:75:f8:a1 br-qemu : MAC address: fe:8d:25:75:f8:a1 tap0 : flags change: +UP +LOWER_UP +RUNNING tap0 : gained carrier br-qemu : flags change: +LOWER_UP +RUNNING br-qemu : gained carrier sd-rtnl: discarding 20 bytes of incoming message tap0 : added address: fe80::fc8d:25ff:fe75:f8a1/64 tap0 : udev initialized link tap0 : link state is up-to-date tap0 : unmanaged tap0 : flags change: -UP -LOWER_UP -RUNNING tap0 : removed address: fe80::fc8d:25ff:fe75:f8a1/64 tap0 : link removed tap0 : link 41 added tap0 : could not find udev device could not add new link br-qemu : MAC address: 00:00:00:00:00:00 br-qemu : flags change: -LOWER_UP tap0 : link removed br-qemu : flags change: -RUNNING br-qemu : lost carrier sd-rtnl: discarding 20 bytes of incoming message Assertion 'm' failed at src/network/networkd-link.c:2029, function link_rtnl_process_address(). Aborting. The KVM is given a bridge device: # /etc/systemd/network/br-qemu.netdev [NetDev] Name=br-qemu Kind=bridge # /etc/systemd/network/br-qemu.network [Match] Name=br-qemu [Network] Description=Qemu vlan Address=10.0.2.1/24 Address=2001:470:736c:107::1/64 And is started with args: -netdev bridge,br=br-qemu,id=mynet0 -device virtio-net,netdev=mynet0,mac=de:ad:be:ef:00:07 I've bisected this breakage to fbbeb65a93e2f90f6576001b. Backtrace attached. It's likely a race condition somewhere -- the SIGABRT sometimes turns into a SIGSEGV when I set SYSTEMD_LOG_LEVEL=debug.