Dumping update here from some findings in MacOSx86.it discord server.

[    2.209904]: igb: start()
[    2.210042]: igb: initEventSources() ===>
[    2.210133]: igb: MSI interrupt could not be enabled.
[    2.210135]: igb: Error initializing event sources.
[    2.210136]: igb: Failed to initEventSources()
[    2.211652]: igb: start()
[    2.211685]: igb: initEventSources() ===>
[    2.211996]: igb: MSI interrupt could not be enabled.
[    2.211998]: igb: Error initializing event sources.
[    2.211999]: igb: Failed to initEventSources()
[   15.147184]: igb: start()
[   15.180226]: igb: start()

My reply there:

`So, for as far as I could get, the driver is calling standard macos function to register interruptHandler https://developer.apple.com/documentation/kernel/iointerrupteventsource/1812661-interrupteventsource
with default intIndex=0. It seems the hardware expects further indexes (not sure whether it's because it has 2 ethernet ports) as, for example implemented in IntelMausi https://github.com/acidanthera/IntelMausi/blob/f0e95ef36b5073fc80172474fd8e1a6f3bd14bea/IntelMausiEthernet/IntelMausiSetup.cpp#L299
but that would most probably mean that full-featured MSI interrupts should be enabled and glued to proper parts in Intel IGB source code in the driver which seems a really big piece of work. As now it's mainly change-reboot-and-test kind of loop unless the whole driver is implemented with the new driver kit (which is even more ambitious project) promoted by Apple for future release (this will help in debugging dramatically).

So no low hanging fruit here, even though such major advancement will benefit everyone, incl B450. I won't have enough time for this up to, at least, the New Year. But would be happy to assist and help anyone eager to contribute. Moreover Shaneee merged all my changes recently, he might have someone or ideas as well

Shaneee#2

i211 TRX40 (no ethernet detected) #4

Description

Metadata

Metadata

Assignees

Labels

Projects

Milestone

Relationships

Development

Issue actions