diff --git a/src/cmd/asm/internal/asm/testdata/arm64.s b/src/cmd/asm/internal/asm/testdata/arm64.s
index 0860cbbf22f..2eeaa971f51 100644
--- a/src/cmd/asm/internal/asm/testdata/arm64.s
+++ b/src/cmd/asm/internal/asm/testdata/arm64.s
@@ -107,6 +107,34 @@ TEXT	foo(SB), DUPOK|NOSPLIT, $-8
 	MOVD	(R2)(R6<<3), R4                 // 447866f8
 	MOVD	(R3)(R7.SXTX<<3), R8            // 68f867f8
 	MOVWU	(R5)(R4.UXTW), R10              // aa4864b8
+	MOVBU	(R3)(R9.UXTW), R8               // 68486938
+	MOVBU	(R5)(R8), R10                   // MOVBU	(R5)(R8*1), R10         // aa686838
+	MOVHU	(R2)(R7.SXTW<<1), R11           // 4bd86778
+	MOVHU	(R1)(R2<<1), R5                 // 25786278
+	MOVB	(R9)(R3.UXTW), R6               // 2649a338
+	MOVB	(R10)(R6), R15                  // MOVB	(R10)(R6*1), R15                // 4f69a638
+	MOVH	(R5)(R7.SXTX<<1), R18           // b2f8a778
+	MOVH	(R8)(R4<<1), R10                // 0a79a478
+	MOVW	(R9)(R8.SXTW<<2), R19           // 33d9a8b8
+	MOVW	(R1)(R4.SXTX), R11              // 2be8a4b8
+	MOVW	(R1)(R4.SXTX), ZR               // 3fe8a4b8
+	MOVW	(R2)(R5), R12                   // MOVW	(R2)(R5*1), R12                  // 4c68a5b8
+	MOVD	R5, (R2)(R6<<3)                 // 457826f8
+	MOVD	R9, (R6)(R7.SXTX<<3)            // c9f827f8
+	MOVD	ZR, (R6)(R7.SXTX<<3)            // dff827f8
+	MOVW	R8, (R2)(R3.UXTW<<2)            // 485823b8
+	MOVW	R7, (R3)(R4.SXTW)               // 67c824b8
+	MOVB	R4, (R2)(R6.SXTX)               // 44e82638
+	MOVB	R8, (R3)(R9.UXTW)               // 68482938
+	MOVB	R10, (R5)(R8)                   // MOVB	R10, (R5)(R8*1)                  // aa682838
+	MOVH	R11, (R2)(R7.SXTW<<1)           // 4bd82778
+	MOVH	R5, (R1)(R2<<1)                 // 25782278
+	MOVH	R7, (R2)(R5.SXTX<<1)            // 47f82578
+	MOVH	R8, (R3)(R6.UXTW)               // 68482678
+	MOVB	(R29)(R30<<0), R14              // ae7bbe38
+	MOVB	(R29)(R30), R14                 // MOVB	(R29)(R30*1), R14                // ae6bbe38
+	MOVB	R4, (R2)(R6.SXTX)               // 44e82638
+
 //	LTYPE1 imsr ',' spreg ','
 //	{
 //		outcode($1, &$2, $4, &nullgen);
diff --git a/src/cmd/asm/internal/asm/testdata/arm64error.s b/src/cmd/asm/internal/asm/testdata/arm64error.s
index 4a1142e8a8c..6f27af2f899 100644
--- a/src/cmd/asm/internal/asm/testdata/arm64error.s
+++ b/src/cmd/asm/internal/asm/testdata/arm64error.s
@@ -14,6 +14,8 @@ TEXT errors(SB),$0
 	MOVD	(R3)(R7.SXTX<<2), R8                             // ERROR "invalid index shift amount"
 	MOVWU	(R5)(R4.UXTW<<3), R10                            // ERROR "invalid index shift amount"
 	MOVWU	(R5)(R4<<1), R10                                 // ERROR "invalid index shift amount"
+	MOVB	(R5)(R4.SXTW<<5), R10                            // ERROR "invalid index shift amount"
+	MOVH	R5, (R6)(R2<<3)                                  // ERROR "invalid index shift amount"
 	VLD1	(R8)(R13), [V2.B16]                              // ERROR "illegal combination"
 	VLD1	8(R9), [V2.B16]                                  // ERROR "illegal combination"
 	VST1	[V1.B16], (R8)(R13)                              // ERROR "illegal combination"
diff --git a/src/cmd/internal/obj/arm64/asm7.go b/src/cmd/internal/obj/arm64/asm7.go
index 85eb5fe5323..b8def184de5 100644
--- a/src/cmd/internal/obj/arm64/asm7.go
+++ b/src/cmd/internal/obj/arm64/asm7.go
@@ -448,6 +448,15 @@ var optab = []Optab{
 	/* load with shifted or extended register offset */
 	{AMOVD, C_ROFF, C_NONE, C_REG, 98, 4, 0, 0, 0},
 	{AMOVW, C_ROFF, C_NONE, C_REG, 98, 4, 0, 0, 0},
+	{AMOVH, C_ROFF, C_NONE, C_REG, 98, 4, 0, 0, 0},
+	{AMOVB, C_ROFF, C_NONE, C_REG, 98, 4, 0, 0, 0},
+	{AMOVBU, C_ROFF, C_NONE, C_REG, 98, 4, 0, 0, 0},
+
+	/* store with extended register offset */
+	{AMOVD, C_REG, C_NONE, C_ROFF, 99, 4, 0, 0, 0},
+	{AMOVW, C_REG, C_NONE, C_ROFF, 99, 4, 0, 0, 0},
+	{AMOVH, C_REG, C_NONE, C_ROFF, 99, 4, 0, 0, 0},
+	{AMOVB, C_REG, C_NONE, C_ROFF, 99, 4, 0, 0, 0},
 
 	/* pre/post-indexed/signed-offset load/store register pair
 	   (unscaled, signed 10-bit quad-aligned and long offset) */
@@ -2367,10 +2376,19 @@ func (c *ctxt7) checkoffset(p *obj.Prog, as obj.As) {
 
 /* checkShiftAmount checks whether the index shift amount is valid */
 /* for load with register offset instructions */
-func (c *ctxt7) checkShiftAmount(p *obj.Prog, as obj.As) {
-	amount := (p.From.Index >> 5) & 7
-	switch as {
-	case AMOVWU:
+func (c *ctxt7) checkShiftAmount(p *obj.Prog, a *obj.Addr) {
+	var amount int16
+	amount = (a.Index >> 5) & 7
+	switch p.As {
+	case AMOVB, AMOVBU:
+		if amount != 0 {
+			c.ctxt.Diag("invalid index shift amount: %v", p)
+		}
+	case AMOVH, AMOVHU:
+		if amount != 1 && amount != 0 {
+			c.ctxt.Diag("invalid index shift amount: %v", p)
+		}
+	case AMOVW, AMOVWU:
 		if amount != 2 && amount != 0 {
 			c.ctxt.Diag("invalid index shift amount: %v", p)
 		}
@@ -2914,7 +2932,7 @@ func (c *ctxt7) asmout(p *obj.Prog, o *Optab, out []uint32) {
 			c.ctxt.Diag("REGTMP used in large offset store: %v", p)
 		}
 		o1 = c.omovlit(AMOVD, p, &p.To, REGTMP)
-		o2 = c.olsxrr(p, int32(c.opstrr(p, p.As)), int(p.From.Reg), r, REGTMP)
+		o2 = c.olsxrr(p, int32(c.opstrr(p, p.As, false)), int(p.From.Reg), r, REGTMP)
 
 	case 31: /* movT L(R), R -> ldrT */
 		// if offset L can be split into hi+lo, and both fit into instructions, do
@@ -4293,7 +4311,7 @@ func (c *ctxt7) asmout(p *obj.Prog, o *Optab, out []uint32) {
 	case 98: /* MOVD (Rn)(Rm.SXTW[<<amount]),Rd */
 		if p.From.Offset != 0 {
 			// extended or shifted offset register.
-			c.checkShiftAmount(p, p.As)
+			c.checkShiftAmount(p, &p.From)
 			o1 = c.opldrr(p, p.As, true)
 			o1 |= uint32(p.From.Offset) /* includes reg, op, etc */
 		} else {
@@ -4303,10 +4321,23 @@ func (c *ctxt7) asmout(p *obj.Prog, o *Optab, out []uint32) {
 		}
 		o1 |= uint32(p.From.Reg&31) << 5
 		rt := int(p.To.Reg)
-		if p.To.Type == obj.TYPE_NONE {
-			rt = REGZERO
-		}
 		o1 |= uint32(rt & 31)
+
+	case 99: /* MOVD Rt, (Rn)(Rm.SXTW[<<amount]) */
+		if p.To.Offset != 0 {
+			// extended or shifted offset register.
+			c.checkShiftAmount(p, &p.To)
+			o1 = c.opstrr(p, p.As, true)
+			o1 |= uint32(p.To.Offset) /* includes reg, op, etc */
+		} else {
+			// (Rn)(Rm), no extension or shift.
+			o1 = c.opstrr(p, p.As, false)
+			o1 |= uint32(p.To.Index&31) << 16
+		}
+		o1 |= uint32(p.To.Reg&31) << 5
+		rf := int(p.From.Reg)
+		o1 |= uint32(rf & 31)
+
 	}
 	out[0] = o1
 	out[1] = o2
@@ -5668,20 +5699,25 @@ func (c *ctxt7) opldrr(p *obj.Prog, a obj.As, extension bool) uint32 {
 
 // opstrr returns the ARM64 opcode encoding corresponding to the obj.As opcode
 // for store instruction with register offset.
-func (c *ctxt7) opstrr(p *obj.Prog, a obj.As) uint32 {
+// The offset register can be (Rn)(Rm.UXTW<<2) or (Rn)(Rm<<2) or (Rn)(Rm).
+func (c *ctxt7) opstrr(p *obj.Prog, a obj.As, extension bool) uint32 {
+	OptionS := uint32(0x1a)
+	if extension {
+		OptionS = uint32(0) // option value and S value have been encoded into p.To.Offset.
+	}
 	switch a {
 	case AMOVD:
-		return 0x1a<<10 | 0x1<<21 | 0x1f<<27
+		return OptionS<<10 | 0x1<<21 | 0x1f<<27
 	case AMOVW, AMOVWU:
-		return 0x1a<<10 | 0x1<<21 | 0x17<<27
+		return OptionS<<10 | 0x1<<21 | 0x17<<27
 	case AMOVH, AMOVHU:
-		return 0x1a<<10 | 0x1<<21 | 0x0f<<27
+		return OptionS<<10 | 0x1<<21 | 0x0f<<27
 	case AMOVB, AMOVBU:
-		return 0x1a<<10 | 0x1<<21 | 0x07<<27
+		return OptionS<<10 | 0x1<<21 | 0x07<<27
 	case AFMOVS:
-		return 0x1a<<10 | 0x1<<21 | 0x17<<27 | 1<<26
+		return OptionS<<10 | 0x1<<21 | 0x17<<27 | 1<<26
 	case AFMOVD:
-		return 0x1a<<10 | 0x1<<21 | 0x1f<<27 | 1<<26
+		return OptionS<<10 | 0x1<<21 | 0x1f<<27 | 1<<26
 	}
 	c.ctxt.Diag("bad opstrr %v\n%v", a, p)
 	return 0
diff --git a/src/cmd/internal/obj/arm64/doc.go b/src/cmd/internal/obj/arm64/doc.go
index 4e7cb0177e4..0a7700f8ac4 100644
--- a/src/cmd/internal/obj/arm64/doc.go
+++ b/src/cmd/internal/obj/arm64/doc.go
@@ -35,10 +35,17 @@ Go Assembly for ARM64 Reference Manual
       LDXPW	(<Rn>), (<Rt1>, <Rt2>)
         Loads two 32-bit words from memory, and writes them to Rt1 and Rt2.
 
-    MOVD|MOVW: Load Register (register offset)
+    MOVD|MOVW|MOVH|MOVHU|MOVB|MOVBU: Load Register (register offset)
       MOVD	(Rn)(Rm.UXTW<<3), Rt
       MOVD	(Rn)(Rm.SXTX), Rt
+      MOVD	(Rn)(Rm<<3), Rt
       MOVD	(Rn)(Rm), Rt
+      MOVB|MOVBU	(Rn)(Rm.UXTW), Rt
+
+    MOVD|MOVW|MOVH|MOVB: Stote Register (register offset)
+      MOVD	Rt, (Rn)(Rm.UXTW<<3)
+      MOVD	Rt, (Rn)(Rm.SXTX)
+      MOVD	Rt, (Rn)(Rm)
 
     PRFM: Prefetch Memory (immediate)
       PRFM	imm(Rn), <prfop>