I am trying to understand the IEEE 802.3 standard specifically for the MAC address tables. According to Clauses 2.3 and 31.3 The MAC Client must provision the SA (source address) parameter within the MA_DATA.request primitive prior to its transit to the MAC sublayer (through the optional MAC Control sublayer). I have three questions about this


1) Where is the burned-in address configured/located?

2) How is this requested or cited by MA_DATA.request?

3) How does the operating system through a simple command write to the firmware in Logical Link Control sublayer if the layer is specified to be independent from all above layers?

The MAC address is part of the interface (software construct) while the burned-in address is part of the NIC (hardware). A given NIC may have multiple interfaces, so source address on xmit packets is supplied by software. The driver usually reads the burned-in address through non-standard means and may load it into the card registers to drop non-matching dest addresses on rcv packets. In the case where you have a bridge configured, there is no filtering and the dest matching is done in software.

No idea what firmware has to do with LLC or why it is a problem that the layers are independent. Maybe you could give an example of a command and why you think it is a problem.