bestpartition in mppini no more working

[smasson]

Context

Since [13286] and [13291] bestpartition is no more finding the "best partition"

Analysis

I thought that the choice of the best partition was independent of nn_hls and that it could be done with nn_hls = 0.
This is wrong!
There is a quick example corresponding to the 1st zoom of AGRIF_DEMO) with Ni0glo = 48 and Nj0glo = 50 using 16 cores:

jpni = 16, jpnj = 1
nn_hls = 0 -> domain size:  3*50 = 150 => best
nn_hls = 1 -> domain size:  5*52 = 260
nn_hls = 2 -> domain size:  7*54 = 378

jpni = 8, jpnj = 2
nn_hls = 0 -> domain size:  6*25 = 150 => best
nn_hls = 1 -> domain size:  8*27 = 216
nn_hls = 2 -> domain size: 10*29 = 290

jpni = 4, jpnj = 4
nn_hls = 0 -> domain size: 12*13 = 156
nn_hls = 1 -> domain size: 14*15 = 210 => best
nn_hls = 2 -> domain size: 16*17 = 272 => best

Fix

modify mppini to use nn_hls when looking for the best partition.

[smasson]

In 13490:

Error: Failed to load processor CommitTicketReference

No macro or processor named 'CommitTicketReference' found