mass contribution to dH/dlambda has wrong sign
Commit b8440aeb incorrectly change the sign of the kinetic contribution to dH/dl. The commit message says:
in update.c, in calc_ke_part_normal and calc_ke_part_visc
- sign error in mass change; if mass B is greater than mass A, then the
change in free energy is positive, not negative.
This is incorrect, if massB is greater the free-energy change is negative. The kinetic contribution is obtained from the derivative of p^2/(2*mass(lambda) which has the opposite sign.
The fix is only a sign change (plus a comment explaining this tricky sign issue).
Since mass contributions vanish is a thermodynamic cycle, I don't know if we should fix this for 4.6.6 or for 5.0.