diff --git a/sm4/asm_arm64.s b/sm4/asm_arm64.s
index a405c1b..baa97ec 100644
--- a/sm4/asm_arm64.s
+++ b/sm4/asm_arm64.s
@@ -170,6 +170,21 @@ TEXT ·expandKeyAsm(SB),NOSPLIT,$0
   ADD $124, R11
   VEOR ZERO.B16, ZERO.B16, ZERO.B16
 
+  MOVW.P 4(R9), R19
+  VMOV R19, x.S[0]
+  VEOR t1.B16, x.B16, x.B16
+  VEOR t2.B16, x.B16, x.B16
+  VEOR t3.B16, x.B16, x.B16
+  VAND x.B16, NIBBLE_MASK.B16, XTMP7.B16
+  VTBL XTMP7.B16, [M1L.B16], y.B16
+  VUSHR $4, x.D2, x.D2
+  VAND x.B16, NIBBLE_MASK.B16, XTMP7.B16
+  VTBL XTMP7.B16, [M1H.B16], XTMP7.B16
+  VEOR y.B16, XTMP7.B16, x.B16
+  VTBL INVERSE_SHIFT_ROWS.B16, [x.B16], x.B16
+
+  VEOR x.B16, t0.B16, t0.B16
+  
   VMOV t0.S[0], R2
   MOVW.P R2, 4(R10)
   VMOV t1.S[0], R2