There is still an issue at exactly Ye=0.5, but the openmp bug is fixed. It turns out we have to include the several variable assignments before the eos call in the critical block, because those are actually helmholtz varibales and not nulib variables. If optimization is important, one could call the eos for many values at once, but I think it works fine the way it is now (as long as you stay away from ye=0.5)
There is still an issue at exactly Ye=0.5, but the openmp bug is fixed. It turns out we have to include the several variable assignments before the eos call in the critical block, because those are actually helmholtz varibales and not nulib variables. If optimization is important, one could call the eos for many values at once, but I think it works fine the way it is now (as long as you stay away from ye=0.5)