Dual GPS. problem with GPS leap second definition on gps2?

I use Mission Planner 1.3.32 with a arduplane piwhawk 1.3.4.
I have 2 GPs on board + a independant datalogger for GPS2. I save at the time the RAW signal of GPS2 ( ublox M8N) on a Sdcard in in datalogger and as GPs2 in the log files. Gps1 is a standard 3DR Gps ublox 6.
When I compare , for a exact same position, the time given by the independant datalogger and the time for Gps2 with the MP log , I found a difference of 2 sec. ( GPs2 mentionned time in MP is 2sec later than the true time)
Also when I see the log browser and I check the calculation of the date and time with the given GMS and Gwk, I find a correct figure with 17 s leap second with Gps2 but a wrong 15s leap second with Gps1.( But the time mentionned by Mp for Gps1 is correct!)
please see the joined image …

I don’t see where is my mistake or if this a bug

the code that calculates the time will be based on gps1 only, as well as the timeUS field.

this is to keep all values consistant across the log, otherwise gps latency would factor in.