crypto: x86/aes-gcm - optimize AVX512 precomputation of H^2 from H^1

crypto: x86/aes-gcm - optimize AVX512 precomputation of H^2 from H^1

Squaring in GF(2^128) requires fewer instructions than a generic
multiplication in GF(2^128).  Take advantage of this when computing H^2
from H^1 in aes_gcm_precompute_vaes_avx512().

Note that aes_gcm_precompute_vaes_avx2() already uses this optimization.

Acked-by: Ard Biesheuvel <ardb@kernel.org>
Tested-by: Ard Biesheuvel <ardb@kernel.org>
Link: https://lore.kernel.org/r/20251002023117.37504-8-ebiggers@kernel.org
Signed-off-by: Eric Biggers <ebiggers@kernel.org>