To implement AES-GCM bulk function to improve AES-GCM performance using Power Assembly.
The implementation is for Power 10 and above. The improvement is as follow for AES128,
GCM enc | 0.240 ns/B 3969 MiB/s - c/B GCM dec | 0.240 ns/B 3966 MiB/s - c/B
GCM enc | 0.192 ns/B 4971 MiB/s - c/B GCM dec | 0.193 ns/B 4951 MiB/s - c/B
The benchmark was run on P10 with 3.447 GHz.
Also, This is the first time I am using the platform, I would need advise how to submit the patch. Thanks.
Files added and changed.
- cipher/rijndael-p10le.c - new
- cipher/rijndael-gcm-p10le.s - new