igc Adapter Resets #74
Labels
No labels
affects/hardware
affects/legal
affects/reproducibility
affects/security
affects/style
affects/usability
blocked by/release 21.05
blocked by/release 21.11
blocked by/release 22.05
blocked by/testing needed
blocked by/testing needed/fuuko
blocked by/testing needed/sayuri
blocked by/upstream
resolution
deferred
resolution
permanent workaround
resolution
upstream wontfix
resolution
wontfix
type
bug
type
chore
type
feature
type
new machine
type
question
type
regression
type
tracking
No milestone
No project
No assignees
1 participant
Notifications
Due date
No due date set.
Dependencies
No dependencies set.
Reference: simon/nixos-config#74
Loading…
Reference in a new issue
No description provided.
Delete branch "%!s()"
Deleting a branch is permanent. Although the deleted branch may continue to exist for a short time before it actually gets removed, it CANNOT be undone in most cases. Continue?
Problem
Sometimes, the NIC resets and throws the following message into dmesg (all at the same time):
(In this case, the cable was removed, so the link going down is not a problem here, but that is not always the case)
On Linux 6.4.1, I observed a kernel trace (however it isn’t printed every time the issue occurs):
Scope
This affects all devices with Intel i225-V NICs, namely shinbu (4 NICs) and fuuko (1 external NIC).
shinobu:
lspci -vnn -d 8086:15f3
fuuko:
lspci -vnn -d 8086:15f3
Firmware version
Kernel version
Sometimes, it happens just spontaneously with nothing from my end, but the chances of it happening are increased by the following factors:
Upstream issues
There seem to be problems with the specific integration of the i225-V on ASUS mainboards, which are related to power saving measures and cause the device to disappear from the PCI bus. From what I can see, this is not what happens in this case.
however, this also has a
Detected Tx Unit Hang
error message, which I have never seenI do not know if this refers to the same problem, but this (updating the firmware) might solve it.
However, there is no way laid out how to do this on Linux.
Solution
?
I reached out to Protectli, since 4 out of 5 affected ports are on their device (and they might have other customers with the same problem).