Error in posit subtraction

[Mark1626]

There is an issue in posit subtraction that causes the following to returns an incorrect result. Issue is most likely due to incorrect regime bit being created

nbits - 32
es - 2
7.000091552734375 - 7.00

Reproducing

This can be reproduced by the adding the following unit test to PositAddSpec.scala

it should "return posit" in {
    posit_add_test(32, 2, 1442843648, 1442840576, 2101346304, sub=true)
}