V7.D2, V6, i3 \ VADD i3.D2, i1.D2, i4.D2 \ SHA512H2 i0.D2, i1, i3 // func blockSHA512(dig *Digest, p []byte) TEXT ·blockSHA512(SB),NOSPLIT,$0 MOVD dig+0(FP), R0 MOVD p_base+8(FP), R1 MOVD p_len+16(FP), R2 MOVD $·_K+0(SB), R3 // long enough to prefetch PRFM (R3), PLDL3KEEP // load digest VLD1 (R0), [V8.D2, V9.D2, V10.D2, V11.D2] loop: // load digest in V0-V3 keeping original in V8-V11 VMOV V8.B16, V0.B16 VMOV V9.B16, V1.B16 VMOV V10.B16, V2.B16 VMOV V11.B16, V3.B16 // load message data in V12-V19 VLD1.P...