// Copyright 2026 The Go Authors. All rights reserved. // Use of this source code is governed by a BSD-style // license that can be found in the LICENSE file. //go:build !purego #include "textflag.h" // func encryptBlockAsm(nr int, xk *uint32, dst, src *byte) TEXT ·encryptBlockAsm(SB),NOSPLIT,$0 MOV nr+0(FP), X10 MOV xk+8(FP), X11 MOV dst+16(FP), X12 MOV src+24(FP), X13 VSETIVLI $4, E32, M1, TA, MA, X0 VLE32V (X13), V2 // V2 = src // Round 0 VLE32V (X11), V10 VAESZVS V10, V2 ADD $16, X11 // Round middles 1 .. nr - 2 VLE32V (X11), V11 VAESEMVS V11, V2 ADD $16, X11 VLE32V (X11), V12 VAESEMVS V12, V2 ADD $16, X11 VLE32V (X11), V13 VAESEMVS V13, V2 ADD $16, X11 VLE32V (X11), V14 VAESEMVS V14, V2 ADD $16, X11 VLE32V (X11), V15 VAESEMVS V15, V2 ADD $16, X11 VLE32V (X11), V16 VAESEMVS V16, V2 ADD $16, X11 VLE32V (X11), V17 VAESEMVS V17, V2 ADD $16, X11 VLE32V (X11), V18 VAESEMVS V18, V2 ADD $16, X11 VLE32V (X11), V19 VAESEMVS V19, V2 ADD $16, X11 SUB $12, X10, X20 BLTZ X20, encFinalRound BEQZ X20, encLastThreeRounds VLE32V (X11), V11 VAESEMVS V11, V2 ADD $16, X11 VLE32V (X11), V12 VAESEMVS V12, V2 ADD $16, X11 encLastThreeRounds: VLE32V (X11), V13 VAESEMVS V13, V2 ADD $16, X11 VLE32V (X11), V14 VAESEMVS V14, V2 ADD $16, X11 encFinalRound: VLE32V (X11), V10 VAESEFVS V10, V2 VSE32V V2, (X12) RET // func decryptBlockAsm(nr int, xk *uint32, dst, src *byte) TEXT ·decryptBlockAsm(SB),NOSPLIT,$0 MOV nr+0(FP), X10 MOV xk+8(FP), X11 MOV dst+16(FP), X12 MOV src+24(FP), X13 VSETIVLI $4, E32, M1, TA, MA, X0 VLE32V (X13), V2 // V2 = src // Round 0 VLE32V (X11), V10 VAESZVS V10, V2 ADD $16, X11 // Round middles 1 .. nr - 2 VLE32V (X11), V11 VAESDMVS V11, V2 ADD $16, X11 VLE32V (X11), V12 VAESDMVS V12, V2 ADD $16, X11 VLE32V (X11), V13 VAESDMVS V13, V2 ADD $16, X11 VLE32V (X11), V14 VAESDMVS V14, V2 ADD $16, X11 VLE32V (X11), V15 VAESDMVS V15, V2 ADD $16, X11 VLE32V (X11), V16 VAESDMVS V16, V2 ADD $16, X11 VLE32V (X11), V17 VAESDMVS V17, V2 ADD $16, X11 VLE32V (X11), V18 VAESDMVS V18, V2 ADD $16, X11 VLE32V (X11), V19 VAESDMVS V19, V2 ADD $16, X11 SUB $12, X10, X20 BLTZ X20, decLastRound BEQZ X20, decLastThreeRounds VLE32V (X11), V11 VAESDMVS V11, V2 ADD $16, X11 VLE32V (X11), V12 VAESDMVS V12, V2 ADD $16, X11 decLastThreeRounds: VLE32V (X11), V13 VAESDMVS V13, V2 ADD $16, X11 VLE32V (X11), V14 VAESDMVS V14, V2 ADD $16, X11 decLastRound: VLE32V (X11), V10 VAESDFVS V10, V2 VSE32V V2, (X12) RET #define KF1M(ROUND, CUR, NEXT) \ VAESKF1VI ROUND, CUR, NEXT; \ VSE32V NEXT, (X12); \ ADD $16, X12; \ VSE32V NEXT, (X13); \ SUB $16, X13; \ #define KF2M(ROUND, CUR, NEXT) \ VAESKF2VI ROUND, CUR, NEXT; \ VSE32V NEXT, (X12); \ ADD $16, X12; \ VSE32V NEXT, (X13); \ SUB $16, X13; \ // func expandKeyAsm(nr int, key *byte, enc *uint32, dec *uint32) TEXT ·expandKeyAsm(SB),NOSPLIT,$0 MOV nr+0(FP), X10 MOV key+8(FP), X11 MOV enc+16(FP), X12 MOV dec+24(FP), X13 VSETIVLI $4, E32, M1, TA, MA, X0 MOV $14, X20 BEQ X20, X10, expand256 expand128: ADD $160, X13 VLE32V (X11), V2 VSE32V V2, (X12) ADD $16, X12 VSE32V V2, (X13) SUB $16, X13 KF1M($1, V2, V3) KF1M($2, V3, V4) KF1M($3, V4, V5) KF1M($4, V5, V6) KF1M($5, V6, V7) KF1M($6, V7, V8) KF1M($7, V8, V9) KF1M($8, V9, V10) KF1M($9, V10, V11) VAESKF1VI $10, V11, V12 VSE32V V12, (X12) VSE32V V12, (X13) RET expand256: ADD $224, X13 VLE32V (X11), V2 ADD $16, X11 VLE32V (X11), V3 VSE32V V2, (X12) ADD $16, X12 VSE32V V2, (X13) SUB $16, X13 VSE32V V3, (X12) ADD $16, X12 VSE32V V3, (X13) SUB $16, X13 KF2M($2, V3, V2) KF2M($3, V2, V3) KF2M($4, V3, V2) KF2M($5, V2, V3) KF2M($6, V3, V2) KF2M($7, V2, V3) KF2M($8, V3, V2) KF2M($9, V2, V3) KF2M($10, V3, V2) KF2M($11, V2, V3) KF2M($12, V3, V2) KF2M($13, V2, V3) VAESKF2VI $14, V3, V2 VSE32V V2, (X12) VSE32V V2, (X13) RET