SFP DDM stack-overflow (8.16 dBm)

A couple of years ago one of the networks i admin got a batch of defective plugs where the DDM didn't seem to work correctly. Otherwise they function perfectly.

When you try to poll the DDM from them, both Tx- and Rx-power always reported 8.16 dBm. Seeing as they worked, and the next batch didn't have the problem we acknowledged the problem with the vendor and then promptly forgot about the problem.

Today while discussing some troubleshooting steps with a couple of new hires this problem was mentioned by them. I casually dismissed the problem with "it's a hardware bug, don't worry about it! It's probably a translation-issue inside of the plug".

After this I went over a conversion-table from mW to dBm, and noticed this:

6 mW = 7.7815 dBm 7 mW = 8.4510 dBm 

I then wondered what 8.16 dBm maps against in mW:

8.16 dBm = 6.5463 mW 

That's awfully close to 2¹⁶ so i translated that into dBm:

6.5536 mW = 8.1647 dBm 

From what I can gather from the above information, the mW-counter seems to be 16-bits long, and seems to either be stuck to 0xffff or there is a stack-overflow happening.

I just wanted to share this with you guys, and ask if you've encountered something similar to this before? If so, please share! I found this issue extremely interesting!