For packed and unpacked calculations, the orthogonalization_single is done batchified. This remove the calls to batch_get_states for the CG eigensolver. Some tests are not passing at the moment and performance need to be checked. At the moment, the spinors are treated as before, as there is no clear gain.
Improve the performance of the CG eigensolver.
- I have checked that my code follows the Octopus coding standards
- I have added tests for all the new features added in this request.