Improving dm-crypt performance by making it able to use more than one core

Issue

• Currently "dm-crypt" does all encryption work per dmcrypt mapping in a single workqueue. This does not scale well when multiple cores are submitting IO at a high rate. The single core running the single thread cannot keep up with the encryption and encrypted IO performance tasks, which causes a bottleneck on the IO worker. How to make dm-crypt able to use more than one core.