RFR: 8371450: AES performance improvements for key schedule generation [v3]

[Ferenc Rakoczi]

On Mon, 10 Nov 2025 02:13:53 GMT, Shawn M Emery <duke at openjdk.org> wrote:

>> The int[256] SBOX solution would increase performance by 2.64% for the AESReinit benchmark.
>
> Switching from a byte[16][16] to a byte[256] SBOX would still increase performance by 1.24% for the AESReinit benchmark w/o sacrificing memory.

> Yes, this would definitely speed up key schedule generation, but at the cost of additional memory; +768 bytes per instance, where the total AES cipher would be 1,268 bytes more per instance than the original Cryptix implementation.

I don't think that when even a Raspberry Pi comes with gigabytes of memory, a 1k increase in constant data size should be an issue to consider. On the other hand, a 1-2% speed increase in a rarely called function is also insignificant, so if you want to keep things unchanged here, I won't object to it.

-------------

PR Review Comment: https://git.openjdk.org/jdk/pull/28188#discussion_r2510666809