This adds the code described in the paper Memory-Efficient High-Speed Implementation of
Kyber on Cortex-M4. It should also be available on eprint later this week.

For more information see https://github.com/mupq/nttm4

Note that there is a performance difference in terms of speed and stack usage between the numbers reported here and those in the paper or on nttm4. This is due to -flto being present in nttm4 but not in pqm4. -flto improves speed but also increases stack space by a bit.