Local Fault debugging on 10G/25G Ethernet Subsystem (xxv_ethernet:v3.1)

[gianLuilui8]

Hello,

I am using the Xilinx IP 10G/25G Ethernet Subsystem. I have it configured as the Ethernet MAC + PCS/PMS 64 bit variant with 4 cores connected to 4 SFP ports. The core is running on a 156.25 MHz clock. My system works great for ~24 to 48 hours and then my application stops processing packets. I have done some digging and I found when I read register 0x404 (RX status register) I get a value of 0x0C0 and occasionally 0x0D0, though not regularly as far as I can tell - still accumulating more information about the frequency of the 0x0D0. 

I read some forums like this https://forums.xilinx.com/t5/Ethernet/10GbE-Local-Fault-detected-management-clock-dclk/m-p/713630 and others but I think most of these don't deal with systems that run for a long time and then fail after 24 hours. Has anyone experienced this? Can someone be more specific about why and what a local fault is caused by? I was thinking maybe it's the quality of my SFP adpaters? Any help would be much appreciated.

EDIT: Could this be a licensing issue? How can I make sure that I have the right license (I just assumed that I was given a license that could run in perpetuity but now I'm having doubts...)

[gianLuilui8]

Solved, I was using a Hardware Evaluation license.

[nanz · Accepted solution]

Hi @gianLuilui8 ,

RX local fault indicates that the receiver is not up and operational.  Most likely causes for an rx local fault are:

1. transceiver has not locked or the receiver is being reset
2. At least one of the lanes is not synchronized
3. The lanes are not properly aligned

A tx local fault will be indicated when:

1. Transceiver transmitter is in reset or
2. has not yet completed any other initialization or synchronization procedures needed.

Unless you have a HW evaluation license, it should continue working after 24/48hours.