diff --git a/llvm/lib/CodeGen/GlobalISel/LegalizerHelper.cpp b/llvm/lib/CodeGen/GlobalISel/LegalizerHelper.cpp
index ed8bd25698c03..dace204b1251f 100644
--- a/llvm/lib/CodeGen/GlobalISel/LegalizerHelper.cpp
+++ b/llvm/lib/CodeGen/GlobalISel/LegalizerHelper.cpp
@@ -6139,6 +6139,7 @@ LegalizerHelper::moreElementsVector(MachineInstr &MI, unsigned TypeIdx,
   case TargetOpcode::G_FCANONICALIZE:
   case TargetOpcode::G_SEXT_INREG:
   case TargetOpcode::G_ABS:
+  case TargetOpcode::G_CTLZ:
     if (TypeIdx != 0)
       return UnableToLegalize;
     Observer.changingInstr(MI);
diff --git a/llvm/lib/Target/AArch64/GISel/AArch64LegalizerInfo.cpp b/llvm/lib/Target/AArch64/GISel/AArch64LegalizerInfo.cpp
index 0da3c73b6926d..3270a8145399b 100644
--- a/llvm/lib/Target/AArch64/GISel/AArch64LegalizerInfo.cpp
+++ b/llvm/lib/Target/AArch64/GISel/AArch64LegalizerInfo.cpp
@@ -326,12 +326,23 @@ AArch64LegalizerInfo::AArch64LegalizerInfo(const AArch64Subtarget &ST)
       .maxScalarEltSameAsIf(always, 1, 0);
 
   getActionDefinitionsBuilder(G_CTLZ)
-      .legalForCartesianProduct(
-          {s32, s64, v8s8, v16s8, v4s16, v8s16, v2s32, v4s32})
-      .scalarize(1)
+      .legalFor({{s32, s32},
+                 {s64, s64},
+                 {v8s8, v8s8},
+                 {v16s8, v16s8},
+                 {v4s16, v4s16},
+                 {v8s16, v8s16},
+                 {v2s32, v2s32},
+                 {v4s32, v4s32}})
       .widenScalarToNextPow2(1, /*Min=*/32)
       .clampScalar(1, s32, s64)
+      .clampNumElements(0, v8s8, v16s8)
+      .clampNumElements(0, v4s16, v8s16)
+      .clampNumElements(0, v2s32, v4s32)
+      .moreElementsToNextPow2(0)
+      .scalarizeIf(scalarOrEltWiderThan(0, 32), 0)
       .scalarSameSizeAs(0, 1);
+
   getActionDefinitionsBuilder(G_CTLZ_ZERO_UNDEF).lower();
 
   getActionDefinitionsBuilder(G_CTTZ)
diff --git a/llvm/test/CodeGen/AArch64/ctlz.ll b/llvm/test/CodeGen/AArch64/ctlz.ll
index 437e3d5ff75c6..59a845044e164 100644
--- a/llvm/test/CodeGen/AArch64/ctlz.ll
+++ b/llvm/test/CodeGen/AArch64/ctlz.ll
@@ -1,6 +1,6 @@
 ; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py UTC_ARGS: --version 5
 ; RUN: llc -mtriple=aarch64-none-eabi -verify-machineinstrs %s -o - | FileCheck %s --check-prefixes=CHECK,CHECK-SD
-; RUN: llc -mtriple=aarch64-none-eabi -global-isel -global-isel-abort=2 -verify-machineinstrs %s -o - 2>&1 | FileCheck %s --check-prefixes=CHECK,CHECK-GI
+; RUN: llc -mtriple=aarch64-none-eabi -global-isel -verify-machineinstrs %s -o - | FileCheck %s --check-prefixes=CHECK,CHECK-GI
 
 
 define void @v2i8(ptr %p1) {
@@ -21,14 +21,14 @@ define void @v2i8(ptr %p1) {
 ;
 ; CHECK-GI-LABEL: v2i8:
 ; CHECK-GI:       // %bb.0: // %entry
-; CHECK-GI-NEXT:    ldrb w8, [x0]
-; CHECK-GI-NEXT:    ldrb w9, [x0, #1]
-; CHECK-GI-NEXT:    clz w8, w8
-; CHECK-GI-NEXT:    clz w9, w9
-; CHECK-GI-NEXT:    sub w8, w8, #24
-; CHECK-GI-NEXT:    sub w9, w9, #24
-; CHECK-GI-NEXT:    strb w8, [x0]
-; CHECK-GI-NEXT:    strb w9, [x0, #1]
+; CHECK-GI-NEXT:    ldr b0, [x0]
+; CHECK-GI-NEXT:    ldr b1, [x0, #1]
+; CHECK-GI-NEXT:    add x8, x0, #1
+; CHECK-GI-NEXT:    mov v0.b[0], v0.b[0]
+; CHECK-GI-NEXT:    mov v0.b[1], v1.b[0]
+; CHECK-GI-NEXT:    clz v0.8b, v0.8b
+; CHECK-GI-NEXT:    st1 { v0.b }[0], [x0]
+; CHECK-GI-NEXT:    st1 { v0.b }[1], [x8]
 ; CHECK-GI-NEXT:    ret
 entry:
   %d = load <2 x i8>, ptr %p1
@@ -59,18 +59,18 @@ define void @v3i8(ptr %p1) {
 ;
 ; CHECK-GI-LABEL: v3i8:
 ; CHECK-GI:       // %bb.0: // %entry
-; CHECK-GI-NEXT:    ldrb w8, [x0]
-; CHECK-GI-NEXT:    ldrb w9, [x0, #1]
-; CHECK-GI-NEXT:    ldrb w10, [x0, #2]
-; CHECK-GI-NEXT:    clz w8, w8
-; CHECK-GI-NEXT:    clz w9, w9
-; CHECK-GI-NEXT:    clz w10, w10
-; CHECK-GI-NEXT:    sub w8, w8, #24
-; CHECK-GI-NEXT:    sub w9, w9, #24
-; CHECK-GI-NEXT:    strb w8, [x0]
-; CHECK-GI-NEXT:    sub w8, w10, #24
-; CHECK-GI-NEXT:    strb w9, [x0, #1]
-; CHECK-GI-NEXT:    strb w8, [x0, #2]
+; CHECK-GI-NEXT:    ldr b0, [x0]
+; CHECK-GI-NEXT:    ldr b1, [x0, #1]
+; CHECK-GI-NEXT:    add x8, x0, #1
+; CHECK-GI-NEXT:    add x9, x0, #2
+; CHECK-GI-NEXT:    mov v0.b[0], v0.b[0]
+; CHECK-GI-NEXT:    mov v0.b[1], v1.b[0]
+; CHECK-GI-NEXT:    ldr b1, [x0, #2]
+; CHECK-GI-NEXT:    mov v0.b[2], v1.b[0]
+; CHECK-GI-NEXT:    clz v0.8b, v0.8b
+; CHECK-GI-NEXT:    st1 { v0.b }[0], [x0]
+; CHECK-GI-NEXT:    st1 { v0.b }[1], [x8]
+; CHECK-GI-NEXT:    st1 { v0.b }[2], [x9]
 ; CHECK-GI-NEXT:    ret
 entry:
   %d = load <3 x i8>, ptr %p1
@@ -95,29 +95,15 @@ define void @v4i8(ptr %p1) {
 ; CHECK-GI:       // %bb.0: // %entry
 ; CHECK-GI-NEXT:    ldr w8, [x0]
 ; CHECK-GI-NEXT:    fmov s0, w8
-; CHECK-GI-NEXT:    uxtb w8, w8
-; CHECK-GI-NEXT:    clz w8, w8
 ; CHECK-GI-NEXT:    mov b1, v0.b[1]
-; CHECK-GI-NEXT:    mov b2, v0.b[2]
-; CHECK-GI-NEXT:    sub w8, w8, #24
+; CHECK-GI-NEXT:    mov v2.b[0], v0.b[0]
+; CHECK-GI-NEXT:    mov b3, v0.b[2]
 ; CHECK-GI-NEXT:    mov b0, v0.b[3]
-; CHECK-GI-NEXT:    fmov w9, s1
-; CHECK-GI-NEXT:    fmov w10, s2
-; CHECK-GI-NEXT:    fmov s1, w8
-; CHECK-GI-NEXT:    uxtb w9, w9
-; CHECK-GI-NEXT:    uxtb w8, w10
-; CHECK-GI-NEXT:    clz w9, w9
-; CHECK-GI-NEXT:    clz w8, w8
-; CHECK-GI-NEXT:    sub w9, w9, #24
-; CHECK-GI-NEXT:    sub w8, w8, #24
-; CHECK-GI-NEXT:    mov v1.b[1], w9
-; CHECK-GI-NEXT:    fmov w9, s0
-; CHECK-GI-NEXT:    uxtb w9, w9
-; CHECK-GI-NEXT:    mov v1.b[2], w8
-; CHECK-GI-NEXT:    clz w8, w9
-; CHECK-GI-NEXT:    sub w8, w8, #24
-; CHECK-GI-NEXT:    mov v1.b[3], w8
-; CHECK-GI-NEXT:    fmov w8, s1
+; CHECK-GI-NEXT:    mov v2.b[1], v1.b[0]
+; CHECK-GI-NEXT:    mov v2.b[2], v3.b[0]
+; CHECK-GI-NEXT:    mov v2.b[3], v0.b[0]
+; CHECK-GI-NEXT:    clz v0.8b, v2.8b
+; CHECK-GI-NEXT:    fmov w8, s0
 ; CHECK-GI-NEXT:    str w8, [x0]
 ; CHECK-GI-NEXT:    ret
 entry:
@@ -148,145 +134,11 @@ entry:
 }
 
 define <32 x i8> @v32i8(<32 x i8> %d) {
-; CHECK-SD-LABEL: v32i8:
-; CHECK-SD:       // %bb.0: // %entry
-; CHECK-SD-NEXT:    clz v0.16b, v0.16b
-; CHECK-SD-NEXT:    clz v1.16b, v1.16b
-; CHECK-SD-NEXT:    ret
-;
-; CHECK-GI-LABEL: v32i8:
-; CHECK-GI:       // %bb.0: // %entry
-; CHECK-GI-NEXT:    umov w9, v0.b[0]
-; CHECK-GI-NEXT:    umov w11, v1.b[0]
-; CHECK-GI-NEXT:    umov w10, v0.b[1]
-; CHECK-GI-NEXT:    umov w13, v1.b[1]
-; CHECK-GI-NEXT:    umov w8, v0.b[2]
-; CHECK-GI-NEXT:    clz w9, w9
-; CHECK-GI-NEXT:    clz w11, w11
-; CHECK-GI-NEXT:    clz w10, w10
-; CHECK-GI-NEXT:    sub w14, w9, #24
-; CHECK-GI-NEXT:    sub w12, w11, #24
-; CHECK-GI-NEXT:    clz w11, w13
-; CHECK-GI-NEXT:    fmov s2, w14
-; CHECK-GI-NEXT:    fmov s3, w12
-; CHECK-GI-NEXT:    umov w9, v1.b[2]
-; CHECK-GI-NEXT:    sub w10, w10, #24
-; CHECK-GI-NEXT:    sub w11, w11, #24
-; CHECK-GI-NEXT:    clz w8, w8
-; CHECK-GI-NEXT:    sub w8, w8, #24
-; CHECK-GI-NEXT:    mov v2.b[1], w10
-; CHECK-GI-NEXT:    mov v3.b[1], w11
-; CHECK-GI-NEXT:    umov w10, v0.b[3]
-; CHECK-GI-NEXT:    clz w9, w9
-; CHECK-GI-NEXT:    umov w11, v1.b[3]
-; CHECK-GI-NEXT:    sub w9, w9, #24
-; CHECK-GI-NEXT:    mov v2.b[2], w8
-; CHECK-GI-NEXT:    mov v3.b[2], w9
-; CHECK-GI-NEXT:    clz w8, w10
-; CHECK-GI-NEXT:    umov w9, v0.b[4]
-; CHECK-GI-NEXT:    clz w10, w11
-; CHECK-GI-NEXT:    umov w11, v1.b[4]
-; CHECK-GI-NEXT:    sub w8, w8, #24
-; CHECK-GI-NEXT:    sub w10, w10, #24
-; CHECK-GI-NEXT:    mov v2.b[3], w8
-; CHECK-GI-NEXT:    mov v3.b[3], w10
-; CHECK-GI-NEXT:    umov w8, v0.b[5]
-; CHECK-GI-NEXT:    clz w9, w9
-; CHECK-GI-NEXT:    clz w10, w11
-; CHECK-GI-NEXT:    umov w11, v1.b[5]
-; CHECK-GI-NEXT:    sub w9, w9, #24
-; CHECK-GI-NEXT:    sub w10, w10, #24
-; CHECK-GI-NEXT:    mov v2.b[4], w9
-; CHECK-GI-NEXT:    mov v3.b[4], w10
-; CHECK-GI-NEXT:    clz w8, w8
-; CHECK-GI-NEXT:    umov w9, v0.b[6]
-; CHECK-GI-NEXT:    clz w10, w11
-; CHECK-GI-NEXT:    umov w11, v1.b[6]
-; CHECK-GI-NEXT:    sub w8, w8, #24
-; CHECK-GI-NEXT:    sub w10, w10, #24
-; CHECK-GI-NEXT:    mov v2.b[5], w8
-; CHECK-GI-NEXT:    mov v3.b[5], w10
-; CHECK-GI-NEXT:    umov w8, v0.b[7]
-; CHECK-GI-NEXT:    clz w9, w9
-; CHECK-GI-NEXT:    clz w10, w11
-; CHECK-GI-NEXT:    umov w11, v1.b[7]
-; CHECK-GI-NEXT:    sub w9, w9, #24
-; CHECK-GI-NEXT:    sub w10, w10, #24
-; CHECK-GI-NEXT:    mov v2.b[6], w9
-; CHECK-GI-NEXT:    mov v3.b[6], w10
-; CHECK-GI-NEXT:    clz w8, w8
-; CHECK-GI-NEXT:    umov w9, v0.b[8]
-; CHECK-GI-NEXT:    clz w10, w11
-; CHECK-GI-NEXT:    umov w11, v1.b[8]
-; CHECK-GI-NEXT:    sub w8, w8, #24
-; CHECK-GI-NEXT:    sub w10, w10, #24
-; CHECK-GI-NEXT:    mov v2.b[7], w8
-; CHECK-GI-NEXT:    mov v3.b[7], w10
-; CHECK-GI-NEXT:    umov w8, v0.b[9]
-; CHECK-GI-NEXT:    clz w9, w9
-; CHECK-GI-NEXT:    clz w10, w11
-; CHECK-GI-NEXT:    umov w11, v1.b[9]
-; CHECK-GI-NEXT:    sub w9, w9, #24
-; CHECK-GI-NEXT:    sub w10, w10, #24
-; CHECK-GI-NEXT:    mov v2.b[8], w9
-; CHECK-GI-NEXT:    mov v3.b[8], w10
-; CHECK-GI-NEXT:    clz w8, w8
-; CHECK-GI-NEXT:    umov w9, v0.b[10]
-; CHECK-GI-NEXT:    clz w10, w11
-; CHECK-GI-NEXT:    umov w11, v1.b[10]
-; CHECK-GI-NEXT:    sub w8, w8, #24
-; CHECK-GI-NEXT:    sub w10, w10, #24
-; CHECK-GI-NEXT:    mov v2.b[9], w8
-; CHECK-GI-NEXT:    mov v3.b[9], w10
-; CHECK-GI-NEXT:    umov w8, v0.b[11]
-; CHECK-GI-NEXT:    clz w9, w9
-; CHECK-GI-NEXT:    clz w10, w11
-; CHECK-GI-NEXT:    umov w11, v1.b[11]
-; CHECK-GI-NEXT:    sub w9, w9, #24
-; CHECK-GI-NEXT:    sub w10, w10, #24
-; CHECK-GI-NEXT:    mov v2.b[10], w9
-; CHECK-GI-NEXT:    mov v3.b[10], w10
-; CHECK-GI-NEXT:    clz w8, w8
-; CHECK-GI-NEXT:    umov w9, v0.b[12]
-; CHECK-GI-NEXT:    clz w10, w11
-; CHECK-GI-NEXT:    umov w11, v1.b[12]
-; CHECK-GI-NEXT:    sub w8, w8, #24
-; CHECK-GI-NEXT:    sub w10, w10, #24
-; CHECK-GI-NEXT:    mov v2.b[11], w8
-; CHECK-GI-NEXT:    mov v3.b[11], w10
-; CHECK-GI-NEXT:    umov w8, v0.b[13]
-; CHECK-GI-NEXT:    clz w9, w9
-; CHECK-GI-NEXT:    clz w10, w11
-; CHECK-GI-NEXT:    umov w11, v1.b[13]
-; CHECK-GI-NEXT:    sub w9, w9, #24
-; CHECK-GI-NEXT:    sub w10, w10, #24
-; CHECK-GI-NEXT:    mov v2.b[12], w9
-; CHECK-GI-NEXT:    mov v3.b[12], w10
-; CHECK-GI-NEXT:    clz w8, w8
-; CHECK-GI-NEXT:    umov w9, v0.b[14]
-; CHECK-GI-NEXT:    clz w10, w11
-; CHECK-GI-NEXT:    umov w11, v1.b[14]
-; CHECK-GI-NEXT:    sub w8, w8, #24
-; CHECK-GI-NEXT:    sub w10, w10, #24
-; CHECK-GI-NEXT:    mov v2.b[13], w8
-; CHECK-GI-NEXT:    mov v3.b[13], w10
-; CHECK-GI-NEXT:    umov w8, v0.b[15]
-; CHECK-GI-NEXT:    umov w10, v1.b[15]
-; CHECK-GI-NEXT:    clz w9, w9
-; CHECK-GI-NEXT:    clz w11, w11
-; CHECK-GI-NEXT:    sub w9, w9, #24
-; CHECK-GI-NEXT:    sub w11, w11, #24
-; CHECK-GI-NEXT:    mov v2.b[14], w9
-; CHECK-GI-NEXT:    mov v3.b[14], w11
-; CHECK-GI-NEXT:    clz w8, w8
-; CHECK-GI-NEXT:    clz w9, w10
-; CHECK-GI-NEXT:    sub w8, w8, #24
-; CHECK-GI-NEXT:    sub w9, w9, #24
-; CHECK-GI-NEXT:    mov v2.b[15], w8
-; CHECK-GI-NEXT:    mov v3.b[15], w9
-; CHECK-GI-NEXT:    mov v0.16b, v2.16b
-; CHECK-GI-NEXT:    mov v1.16b, v3.16b
-; CHECK-GI-NEXT:    ret
+; CHECK-LABEL: v32i8:
+; CHECK:       // %bb.0: // %entry
+; CHECK-NEXT:    clz v0.16b, v0.16b
+; CHECK-NEXT:    clz v1.16b, v1.16b
+; CHECK-NEXT:    ret
 entry:
   %s = call <32 x i8> @llvm.ctlz(<32 x i8> %d, i1 false)
   ret <32 x i8> %s
@@ -310,14 +162,12 @@ define void @v2i16(ptr %p1) {
 ;
 ; CHECK-GI-LABEL: v2i16:
 ; CHECK-GI:       // %bb.0: // %entry
-; CHECK-GI-NEXT:    ldrh w8, [x0]
-; CHECK-GI-NEXT:    ldrh w9, [x0, #2]
-; CHECK-GI-NEXT:    clz w8, w8
-; CHECK-GI-NEXT:    clz w9, w9
-; CHECK-GI-NEXT:    sub w8, w8, #16
-; CHECK-GI-NEXT:    sub w9, w9, #16
-; CHECK-GI-NEXT:    strh w8, [x0]
-; CHECK-GI-NEXT:    strh w9, [x0, #2]
+; CHECK-GI-NEXT:    ldr h0, [x0]
+; CHECK-GI-NEXT:    add x8, x0, #2
+; CHECK-GI-NEXT:    ld1 { v0.h }[1], [x8]
+; CHECK-GI-NEXT:    clz v0.4h, v0.4h
+; CHECK-GI-NEXT:    str h0, [x0]
+; CHECK-GI-NEXT:    st1 { v0.h }[1], [x8]
 ; CHECK-GI-NEXT:    ret
 entry:
   %d = load <2 x i16>, ptr %p1
@@ -338,18 +188,15 @@ define void @v3i16(ptr %p1) {
 ;
 ; CHECK-GI-LABEL: v3i16:
 ; CHECK-GI:       // %bb.0: // %entry
-; CHECK-GI-NEXT:    ldrh w8, [x0]
-; CHECK-GI-NEXT:    ldrh w9, [x0, #2]
-; CHECK-GI-NEXT:    ldrh w10, [x0, #4]
-; CHECK-GI-NEXT:    clz w8, w8
-; CHECK-GI-NEXT:    clz w9, w9
-; CHECK-GI-NEXT:    clz w10, w10
-; CHECK-GI-NEXT:    sub w8, w8, #16
-; CHECK-GI-NEXT:    sub w9, w9, #16
-; CHECK-GI-NEXT:    strh w8, [x0]
-; CHECK-GI-NEXT:    sub w8, w10, #16
-; CHECK-GI-NEXT:    strh w9, [x0, #2]
-; CHECK-GI-NEXT:    strh w8, [x0, #4]
+; CHECK-GI-NEXT:    ldr h0, [x0]
+; CHECK-GI-NEXT:    add x8, x0, #2
+; CHECK-GI-NEXT:    add x9, x0, #4
+; CHECK-GI-NEXT:    ld1 { v0.h }[1], [x8]
+; CHECK-GI-NEXT:    ld1 { v0.h }[2], [x9]
+; CHECK-GI-NEXT:    clz v0.4h, v0.4h
+; CHECK-GI-NEXT:    str h0, [x0]
+; CHECK-GI-NEXT:    st1 { v0.h }[1], [x8]
+; CHECK-GI-NEXT:    st1 { v0.h }[2], [x9]
 ; CHECK-GI-NEXT:    ret
 entry:
   %d = load <3 x i16>, ptr %p1
@@ -379,81 +226,11 @@ entry:
 }
 
 define <16 x i16> @v16i16(<16 x i16> %d) {
-; CHECK-SD-LABEL: v16i16:
-; CHECK-SD:       // %bb.0: // %entry
-; CHECK-SD-NEXT:    clz v0.8h, v0.8h
-; CHECK-SD-NEXT:    clz v1.8h, v1.8h
-; CHECK-SD-NEXT:    ret
-;
-; CHECK-GI-LABEL: v16i16:
-; CHECK-GI:       // %bb.0: // %entry
-; CHECK-GI-NEXT:    umov w8, v0.h[0]
-; CHECK-GI-NEXT:    umov w10, v1.h[0]
-; CHECK-GI-NEXT:    umov w9, v0.h[1]
-; CHECK-GI-NEXT:    umov w11, v1.h[1]
-; CHECK-GI-NEXT:    umov w12, v0.h[2]
-; CHECK-GI-NEXT:    umov w13, v1.h[2]
-; CHECK-GI-NEXT:    clz w8, w8
-; CHECK-GI-NEXT:    clz w10, w10
-; CHECK-GI-NEXT:    clz w9, w9
-; CHECK-GI-NEXT:    sub w8, w8, #16
-; CHECK-GI-NEXT:    sub w10, w10, #16
-; CHECK-GI-NEXT:    clz w11, w11
-; CHECK-GI-NEXT:    fmov s2, w8
-; CHECK-GI-NEXT:    fmov s3, w10
-; CHECK-GI-NEXT:    sub w9, w9, #16
-; CHECK-GI-NEXT:    sub w11, w11, #16
-; CHECK-GI-NEXT:    umov w8, v0.h[3]
-; CHECK-GI-NEXT:    clz w10, w13
-; CHECK-GI-NEXT:    sub w10, w10, #16
-; CHECK-GI-NEXT:    mov v2.h[1], w9
-; CHECK-GI-NEXT:    mov v3.h[1], w11
-; CHECK-GI-NEXT:    clz w9, w12
-; CHECK-GI-NEXT:    umov w11, v1.h[3]
-; CHECK-GI-NEXT:    sub w9, w9, #16
-; CHECK-GI-NEXT:    clz w8, w8
-; CHECK-GI-NEXT:    sub w8, w8, #16
-; CHECK-GI-NEXT:    mov v2.h[2], w9
-; CHECK-GI-NEXT:    mov v3.h[2], w10
-; CHECK-GI-NEXT:    umov w9, v0.h[4]
-; CHECK-GI-NEXT:    clz w10, w11
-; CHECK-GI-NEXT:    umov w11, v1.h[4]
-; CHECK-GI-NEXT:    sub w10, w10, #16
-; CHECK-GI-NEXT:    mov v2.h[3], w8
-; CHECK-GI-NEXT:    mov v3.h[3], w10
-; CHECK-GI-NEXT:    umov w8, v0.h[5]
-; CHECK-GI-NEXT:    clz w9, w9
-; CHECK-GI-NEXT:    clz w10, w11
-; CHECK-GI-NEXT:    umov w11, v1.h[5]
-; CHECK-GI-NEXT:    sub w9, w9, #16
-; CHECK-GI-NEXT:    sub w10, w10, #16
-; CHECK-GI-NEXT:    mov v2.h[4], w9
-; CHECK-GI-NEXT:    mov v3.h[4], w10
-; CHECK-GI-NEXT:    clz w8, w8
-; CHECK-GI-NEXT:    umov w9, v0.h[6]
-; CHECK-GI-NEXT:    clz w10, w11
-; CHECK-GI-NEXT:    umov w11, v1.h[6]
-; CHECK-GI-NEXT:    sub w8, w8, #16
-; CHECK-GI-NEXT:    sub w10, w10, #16
-; CHECK-GI-NEXT:    mov v2.h[5], w8
-; CHECK-GI-NEXT:    mov v3.h[5], w10
-; CHECK-GI-NEXT:    umov w8, v0.h[7]
-; CHECK-GI-NEXT:    umov w10, v1.h[7]
-; CHECK-GI-NEXT:    clz w9, w9
-; CHECK-GI-NEXT:    clz w11, w11
-; CHECK-GI-NEXT:    sub w9, w9, #16
-; CHECK-GI-NEXT:    sub w11, w11, #16
-; CHECK-GI-NEXT:    mov v2.h[6], w9
-; CHECK-GI-NEXT:    mov v3.h[6], w11
-; CHECK-GI-NEXT:    clz w8, w8
-; CHECK-GI-NEXT:    clz w9, w10
-; CHECK-GI-NEXT:    sub w8, w8, #16
-; CHECK-GI-NEXT:    sub w9, w9, #16
-; CHECK-GI-NEXT:    mov v2.h[7], w8
-; CHECK-GI-NEXT:    mov v3.h[7], w9
-; CHECK-GI-NEXT:    mov v0.16b, v2.16b
-; CHECK-GI-NEXT:    mov v1.16b, v3.16b
-; CHECK-GI-NEXT:    ret
+; CHECK-LABEL: v16i16:
+; CHECK:       // %bb.0: // %entry
+; CHECK-NEXT:    clz v0.8h, v0.8h
+; CHECK-NEXT:    clz v1.8h, v1.8h
+; CHECK-NEXT:    ret
 entry:
   %s = call <16 x i16> @llvm.ctlz(<16 x i16> %d, i1 false)
   ret <16 x i16> %s
@@ -470,24 +247,10 @@ entry:
 }
 
 define <3 x i32> @v3i32(<3 x i32> %d) {
-; CHECK-SD-LABEL: v3i32:
-; CHECK-SD:       // %bb.0: // %entry
-; CHECK-SD-NEXT:    clz v0.4s, v0.4s
-; CHECK-SD-NEXT:    ret
-;
-; CHECK-GI-LABEL: v3i32:
-; CHECK-GI:       // %bb.0: // %entry
-; CHECK-GI-NEXT:    fmov w8, s0
-; CHECK-GI-NEXT:    mov w9, v0.s[1]
-; CHECK-GI-NEXT:    clz w8, w8
-; CHECK-GI-NEXT:    mov v1.s[0], w8
-; CHECK-GI-NEXT:    mov w8, v0.s[2]
-; CHECK-GI-NEXT:    clz w9, w9
-; CHECK-GI-NEXT:    mov v1.s[1], w9
-; CHECK-GI-NEXT:    clz w8, w8
-; CHECK-GI-NEXT:    mov v1.s[2], w8
-; CHECK-GI-NEXT:    mov v0.16b, v1.16b
-; CHECK-GI-NEXT:    ret
+; CHECK-LABEL: v3i32:
+; CHECK:       // %bb.0: // %entry
+; CHECK-NEXT:    clz v0.4s, v0.4s
+; CHECK-NEXT:    ret
 entry:
   %s = call <3 x i32> @llvm.ctlz(<3 x i32> %d, i1 false)
   ret <3 x i32> %s
@@ -504,41 +267,11 @@ entry:
 }
 
 define <8 x i32> @v8i32(<8 x i32> %d) {
-; CHECK-SD-LABEL: v8i32:
-; CHECK-SD:       // %bb.0: // %entry
-; CHECK-SD-NEXT:    clz v0.4s, v0.4s
-; CHECK-SD-NEXT:    clz v1.4s, v1.4s
-; CHECK-SD-NEXT:    ret
-;
-; CHECK-GI-LABEL: v8i32:
-; CHECK-GI:       // %bb.0: // %entry
-; CHECK-GI-NEXT:    fmov w9, s0
-; CHECK-GI-NEXT:    fmov w10, s1
-; CHECK-GI-NEXT:    mov w8, v0.s[1]
-; CHECK-GI-NEXT:    mov w11, v1.s[1]
-; CHECK-GI-NEXT:    clz w9, w9
-; CHECK-GI-NEXT:    clz w10, w10
-; CHECK-GI-NEXT:    mov v2.s[0], w9
-; CHECK-GI-NEXT:    mov v3.s[0], w10
-; CHECK-GI-NEXT:    mov w9, v0.s[2]
-; CHECK-GI-NEXT:    mov w10, v1.s[2]
-; CHECK-GI-NEXT:    clz w8, w8
-; CHECK-GI-NEXT:    clz w11, w11
-; CHECK-GI-NEXT:    mov v2.s[1], w8
-; CHECK-GI-NEXT:    mov v3.s[1], w11
-; CHECK-GI-NEXT:    mov w8, v0.s[3]
-; CHECK-GI-NEXT:    mov w11, v1.s[3]
-; CHECK-GI-NEXT:    clz w9, w9
-; CHECK-GI-NEXT:    clz w10, w10
-; CHECK-GI-NEXT:    mov v2.s[2], w9
-; CHECK-GI-NEXT:    mov v3.s[2], w10
-; CHECK-GI-NEXT:    clz w8, w8
-; CHECK-GI-NEXT:    clz w9, w11
-; CHECK-GI-NEXT:    mov v2.s[3], w8
-; CHECK-GI-NEXT:    mov v3.s[3], w9
-; CHECK-GI-NEXT:    mov v0.16b, v2.16b
-; CHECK-GI-NEXT:    mov v1.16b, v3.16b
-; CHECK-GI-NEXT:    ret
+; CHECK-LABEL: v8i32:
+; CHECK:       // %bb.0: // %entry
+; CHECK-NEXT:    clz v0.4s, v0.4s
+; CHECK-NEXT:    clz v1.4s, v1.4s
+; CHECK-NEXT:    ret
 entry:
   %s = call <8 x i32> @llvm.ctlz(<8 x i32> %d, i1 false)
   ret <8 x i32> %s