NIC driver specific reset occurs in absence of NETDEV WATCHDOG Tx timeout
Issue
- We would expect the the following messages:
[1021832.529822] NETDEV WATCHDOG: eno1 (tg3): transmit queue 0 timed out
[1021832.536274] WARNING: CPU: 17 PID: 0 at net/sched/sch_generic.c:482 dev_watchdog+0x29a/0x2b0
.
.
[1021832.924605] tg3 0000:c1:00.0 eno1: transmit timed out, resetting
But sometimes we see only:
[1100490.705915] Injecting an artificial tx watchdog timeout for eno1
[1100490.712103] tg3 0000:c1:00.0 eno1: transmit timed out, resetting
Or for i40e:
[24566383.265023] i40e 0000:0b:00.0 ens1f0: tx_timeout: VSI_seid: 308, Q 17, NTC: 0x1d7, HWB: 0x1a5, NTU: 0x1a5, TAIL: 0x1a5, INT: 0x0
Environment
- Red Hat Enterprise Linux
- All NIC drivers
Subscriber exclusive content
A Red Hat subscription provides unlimited access to our knowledgebase, tools, and much more.