diff --git a/llvm/lib/CodeGen/TwoAddressInstructionPass.cpp b/llvm/lib/CodeGen/TwoAddressInstructionPass.cpp
index 45f61262faf93..41b888592cc18 100644
--- a/llvm/lib/CodeGen/TwoAddressInstructionPass.cpp
+++ b/llvm/lib/CodeGen/TwoAddressInstructionPass.cpp
@@ -1868,12 +1868,16 @@ bool TwoAddressInstructionPass::runOnMachineFunction(MachineFunction &Func) {
             // %reg.subidx.
             LaneBitmask LaneMask =
                 TRI->getSubRegIndexLaneMask(mi->getOperand(0).getSubReg());
-            SlotIndex Idx = LIS->getInstructionIndex(*mi);
+            SlotIndex Idx = LIS->getInstructionIndex(*mi).getRegSlot();
             for (auto &S : LI.subranges()) {
               if ((S.LaneMask & LaneMask).none()) {
-                LiveRange::iterator UseSeg = S.FindSegmentContaining(Idx);
-                LiveRange::iterator DefSeg = std::next(UseSeg);
-                S.MergeValueNumberInto(DefSeg->valno, UseSeg->valno);
+                LiveRange::iterator DefSeg = S.FindSegmentContaining(Idx);
+                if (mi->getOperand(0).isUndef()) {
+                  S.removeValNo(DefSeg->valno);
+                } else {
+                  LiveRange::iterator UseSeg = std::prev(DefSeg);
+                  S.MergeValueNumberInto(DefSeg->valno, UseSeg->valno);
+                }
               }
             }
 
diff --git a/llvm/test/CodeGen/Thumb2/mve-fmath.ll b/llvm/test/CodeGen/Thumb2/mve-fmath.ll
index 844da3baa42ba..3a477f987cee6 100644
--- a/llvm/test/CodeGen/Thumb2/mve-fmath.ll
+++ b/llvm/test/CodeGen/Thumb2/mve-fmath.ll
@@ -1,6 +1,8 @@
 ; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py
-; RUN: llc -mtriple=thumbv8.1m.main-none-none-eabi -mattr=+mve,+fullfp16 -verify-machineinstrs %s -o - | FileCheck %s --check-prefix=CHECK
-; RUN: llc -mtriple=thumbv8.1m.main-none-none-eabi -mattr=+mve.fp -verify-machineinstrs %s -o - | FileCheck %s --check-prefix=CHECK
+; RUN: llc -mtriple=thumbv8.1m.main-none-none-eabi -mattr=+mve,+fullfp16 -verify-machineinstrs %s -o - | FileCheck %s --check-prefixes=CHECK,CHECK-LV
+; RUN: llc -mtriple=thumbv8.1m.main-none-none-eabi -mattr=+mve.fp -verify-machineinstrs %s -o - | FileCheck %s --check-prefixes=CHECK,CHECK-LV
+; RUN: llc -early-live-intervals -mtriple=thumbv8.1m.main-none-none-eabi -mattr=+mve,+fullfp16 -verify-machineinstrs %s -o - | FileCheck %s --check-prefixes=CHECK,CHECK-LIS
+; RUN: llc -early-live-intervals -mtriple=thumbv8.1m.main-none-none-eabi -mattr=+mve.fp -verify-machineinstrs %s -o - | FileCheck %s --check-prefixes=CHECK,CHECK-LIS
 
 define arm_aapcs_vfpcc <4 x float> @sqrt_float32_t(<4 x float> %src) {
 ; CHECK-LABEL: sqrt_float32_t:
@@ -1085,21 +1087,37 @@ entry:
 }
 
 define arm_aapcs_vfpcc <2 x double> @copysign_float64_t(<2 x double> %src1, <2 x double> %src2) {
-; CHECK-LABEL: copysign_float64_t:
-; CHECK:       @ %bb.0: @ %entry
-; CHECK-NEXT:    .save {r7, lr}
-; CHECK-NEXT:    push {r7, lr}
-; CHECK-NEXT:    vmov r0, r1, d3
-; CHECK-NEXT:    vmov r0, lr, d2
-; CHECK-NEXT:    vmov r0, r3, d1
-; CHECK-NEXT:    vmov r12, r2, d0
-; CHECK-NEXT:    lsrs r1, r1, #31
-; CHECK-NEXT:    bfi r3, r1, #31, #1
-; CHECK-NEXT:    lsr.w r1, lr, #31
-; CHECK-NEXT:    bfi r2, r1, #31, #1
-; CHECK-NEXT:    vmov d1, r0, r3
-; CHECK-NEXT:    vmov d0, r12, r2
-; CHECK-NEXT:    pop {r7, pc}
+; CHECK-LV-LABEL: copysign_float64_t:
+; CHECK-LV:       @ %bb.0: @ %entry
+; CHECK-LV-NEXT:    .save {r7, lr}
+; CHECK-LV-NEXT:    push {r7, lr}
+; CHECK-LV-NEXT:    vmov r0, r1, d3
+; CHECK-LV-NEXT:    vmov r0, lr, d2
+; CHECK-LV-NEXT:    vmov r0, r3, d1
+; CHECK-LV-NEXT:    vmov r12, r2, d0
+; CHECK-LV-NEXT:    lsrs r1, r1, #31
+; CHECK-LV-NEXT:    bfi r3, r1, #31, #1
+; CHECK-LV-NEXT:    lsr.w r1, lr, #31
+; CHECK-LV-NEXT:    bfi r2, r1, #31, #1
+; CHECK-LV-NEXT:    vmov d1, r0, r3
+; CHECK-LV-NEXT:    vmov d0, r12, r2
+; CHECK-LV-NEXT:    pop {r7, pc}
+;
+; CHECK-LIS-LABEL: copysign_float64_t:
+; CHECK-LIS:       @ %bb.0: @ %entry
+; CHECK-LIS-NEXT:    .save {r4, lr}
+; CHECK-LIS-NEXT:    push {r4, lr}
+; CHECK-LIS-NEXT:    vmov r0, r12, d3
+; CHECK-LIS-NEXT:    vmov r0, lr, d2
+; CHECK-LIS-NEXT:    vmov r4, r3, d1
+; CHECK-LIS-NEXT:    vmov r1, r2, d0
+; CHECK-LIS-NEXT:    lsr.w r0, r12, #31
+; CHECK-LIS-NEXT:    bfi r3, r0, #31, #1
+; CHECK-LIS-NEXT:    lsr.w r0, lr, #31
+; CHECK-LIS-NEXT:    bfi r2, r0, #31, #1
+; CHECK-LIS-NEXT:    vmov d1, r4, r3
+; CHECK-LIS-NEXT:    vmov d0, r1, r2
+; CHECK-LIS-NEXT:    pop {r4, pc}
 entry:
   %0 = call fast <2 x double> @llvm.copysign.v2f64(<2 x double> %src1, <2 x double> %src2)
   ret <2 x double> %0
diff --git a/llvm/test/CodeGen/Thumb2/mve-shuffle.ll b/llvm/test/CodeGen/Thumb2/mve-shuffle.ll
index 93a058828765e..6e644c58687fa 100644
--- a/llvm/test/CodeGen/Thumb2/mve-shuffle.ll
+++ b/llvm/test/CodeGen/Thumb2/mve-shuffle.ll
@@ -1,6 +1,8 @@
 ; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py
-; RUN: llc -mtriple=thumbv8.1m.main-none-none-eabi -mattr=+mve,+fullfp16 -verify-machineinstrs %s -o - | FileCheck %s --check-prefix=CHECK
-; RUN: llc -mtriple=thumbv8.1m.main-none-none-eabi -mattr=+mve.fp -verify-machineinstrs %s -o - | FileCheck %s --check-prefixes=CHECK,CHECKFP
+; RUN: llc -mtriple=thumbv8.1m.main-none-none-eabi -mattr=+mve,+fullfp16 -verify-machineinstrs %s -o - | FileCheck %s --check-prefixes=CHECK,CHECK-LV
+; RUN: llc -mtriple=thumbv8.1m.main-none-none-eabi -mattr=+mve.fp -verify-machineinstrs %s -o - | FileCheck %s --check-prefixes=CHECK,CHECK-LV,CHECKFP
+; RUN: llc -early-live-intervals -mtriple=thumbv8.1m.main-none-none-eabi -mattr=+mve,+fullfp16 -verify-machineinstrs %s -o - | FileCheck %s --check-prefixes=CHECK,CHECK-LIS
+; RUN: llc -early-live-intervals -mtriple=thumbv8.1m.main-none-none-eabi -mattr=+mve.fp -verify-machineinstrs %s -o - | FileCheck %s --check-prefixes=CHECK,CHECK-LIS,CHECKFP
 
 define arm_aapcs_vfpcc <4 x i32> @shuffle1_i32(<4 x i32> %src) {
 ; CHECK-LABEL: shuffle1_i32:
@@ -221,18 +223,31 @@ entry:
 }
 
 define arm_aapcs_vfpcc <8 x i16> @shuffle3_i16(<8 x i16> %src) {
-; CHECK-LABEL: shuffle3_i16:
-; CHECK:       @ %bb.0: @ %entry
-; CHECK-NEXT:    vmov q1, q0
-; CHECK-NEXT:    vmovx.f16 s2, s5
-; CHECK-NEXT:    vmovx.f16 s0, s4
-; CHECK-NEXT:    vins.f16 s5, s4
-; CHECK-NEXT:    vins.f16 s2, s0
-; CHECK-NEXT:    vmov.f32 s3, s5
-; CHECK-NEXT:    vmovx.f16 s1, s7
-; CHECK-NEXT:    vmov.f32 s0, s6
-; CHECK-NEXT:    vins.f16 s1, s7
-; CHECK-NEXT:    bx lr
+; CHECK-LV-LABEL: shuffle3_i16:
+; CHECK-LV:       @ %bb.0: @ %entry
+; CHECK-LV-NEXT:    vmov q1, q0
+; CHECK-LV-NEXT:    vmovx.f16 s2, s5
+; CHECK-LV-NEXT:    vmovx.f16 s0, s4
+; CHECK-LV-NEXT:    vins.f16 s5, s4
+; CHECK-LV-NEXT:    vins.f16 s2, s0
+; CHECK-LV-NEXT:    vmov.f32 s3, s5
+; CHECK-LV-NEXT:    vmovx.f16 s1, s7
+; CHECK-LV-NEXT:    vmov.f32 s0, s6
+; CHECK-LV-NEXT:    vins.f16 s1, s7
+; CHECK-LV-NEXT:    bx lr
+;
+; CHECK-LIS-LABEL: shuffle3_i16:
+; CHECK-LIS:       @ %bb.0: @ %entry
+; CHECK-LIS-NEXT:    vmovx.f16 s5, s3
+; CHECK-LIS-NEXT:    vmovx.f16 s6, s1
+; CHECK-LIS-NEXT:    vmovx.f16 s4, s0
+; CHECK-LIS-NEXT:    vins.f16 s1, s0
+; CHECK-LIS-NEXT:    vins.f16 s6, s4
+; CHECK-LIS-NEXT:    vins.f16 s5, s3
+; CHECK-LIS-NEXT:    vmov.f32 s7, s1
+; CHECK-LIS-NEXT:    vmov.f32 s4, s2
+; CHECK-LIS-NEXT:    vmov q0, q1
+; CHECK-LIS-NEXT:    bx lr
 entry:
   %out = shufflevector <8 x i16> %src, <8 x i16> undef, <8 x i32> <i32 4, i32 5, i32 7, i32 6, i32 3, i32 1, i32 2, i32 0>
   ret <8 x i16> %out
@@ -476,42 +491,79 @@ entry:
 }
 
 define arm_aapcs_vfpcc <16 x i8> @shuffle3_i8(<16 x i8> %src) {
-; CHECK-LABEL: shuffle3_i8:
-; CHECK:       @ %bb.0: @ %entry
-; CHECK-NEXT:    vmov q1, q0
-; CHECK-NEXT:    vmov.u8 r0, q0[4]
-; CHECK-NEXT:    vmov.8 q0[0], r0
-; CHECK-NEXT:    vmov.u8 r0, q1[5]
-; CHECK-NEXT:    vmov.8 q0[1], r0
-; CHECK-NEXT:    vmov.u8 r0, q1[15]
-; CHECK-NEXT:    vmov.8 q0[2], r0
-; CHECK-NEXT:    vmov.u8 r0, q1[7]
-; CHECK-NEXT:    vmov.8 q0[3], r0
-; CHECK-NEXT:    vmov.u8 r0, q1[14]
-; CHECK-NEXT:    vmov.8 q0[4], r0
-; CHECK-NEXT:    vmov.u8 r0, q1[9]
-; CHECK-NEXT:    vmov.8 q0[5], r0
-; CHECK-NEXT:    vmov.u8 r0, q1[6]
-; CHECK-NEXT:    vmov.8 q0[6], r0
-; CHECK-NEXT:    vmov.u8 r0, q1[3]
-; CHECK-NEXT:    vmov.8 q0[7], r0
-; CHECK-NEXT:    vmov.u8 r0, q1[10]
-; CHECK-NEXT:    vmov.8 q0[8], r0
-; CHECK-NEXT:    vmov.u8 r0, q1[12]
-; CHECK-NEXT:    vmov.8 q0[9], r0
-; CHECK-NEXT:    vmov.u8 r0, q1[1]
-; CHECK-NEXT:    vmov.8 q0[10], r0
-; CHECK-NEXT:    vmov.u8 r0, q1[13]
-; CHECK-NEXT:    vmov.8 q0[11], r0
-; CHECK-NEXT:    vmov.u8 r0, q1[2]
-; CHECK-NEXT:    vmov.8 q0[12], r0
-; CHECK-NEXT:    vmov.u8 r0, q1[8]
-; CHECK-NEXT:    vmov.8 q0[13], r0
-; CHECK-NEXT:    vmov.u8 r0, q1[0]
-; CHECK-NEXT:    vmov.8 q0[14], r0
-; CHECK-NEXT:    vmov.u8 r0, q1[11]
-; CHECK-NEXT:    vmov.8 q0[15], r0
-; CHECK-NEXT:    bx lr
+; CHECK-LV-LABEL: shuffle3_i8:
+; CHECK-LV:       @ %bb.0: @ %entry
+; CHECK-LV-NEXT:    vmov q1, q0
+; CHECK-LV-NEXT:    vmov.u8 r0, q0[4]
+; CHECK-LV-NEXT:    vmov.8 q0[0], r0
+; CHECK-LV-NEXT:    vmov.u8 r0, q1[5]
+; CHECK-LV-NEXT:    vmov.8 q0[1], r0
+; CHECK-LV-NEXT:    vmov.u8 r0, q1[15]
+; CHECK-LV-NEXT:    vmov.8 q0[2], r0
+; CHECK-LV-NEXT:    vmov.u8 r0, q1[7]
+; CHECK-LV-NEXT:    vmov.8 q0[3], r0
+; CHECK-LV-NEXT:    vmov.u8 r0, q1[14]
+; CHECK-LV-NEXT:    vmov.8 q0[4], r0
+; CHECK-LV-NEXT:    vmov.u8 r0, q1[9]
+; CHECK-LV-NEXT:    vmov.8 q0[5], r0
+; CHECK-LV-NEXT:    vmov.u8 r0, q1[6]
+; CHECK-LV-NEXT:    vmov.8 q0[6], r0
+; CHECK-LV-NEXT:    vmov.u8 r0, q1[3]
+; CHECK-LV-NEXT:    vmov.8 q0[7], r0
+; CHECK-LV-NEXT:    vmov.u8 r0, q1[10]
+; CHECK-LV-NEXT:    vmov.8 q0[8], r0
+; CHECK-LV-NEXT:    vmov.u8 r0, q1[12]
+; CHECK-LV-NEXT:    vmov.8 q0[9], r0
+; CHECK-LV-NEXT:    vmov.u8 r0, q1[1]
+; CHECK-LV-NEXT:    vmov.8 q0[10], r0
+; CHECK-LV-NEXT:    vmov.u8 r0, q1[13]
+; CHECK-LV-NEXT:    vmov.8 q0[11], r0
+; CHECK-LV-NEXT:    vmov.u8 r0, q1[2]
+; CHECK-LV-NEXT:    vmov.8 q0[12], r0
+; CHECK-LV-NEXT:    vmov.u8 r0, q1[8]
+; CHECK-LV-NEXT:    vmov.8 q0[13], r0
+; CHECK-LV-NEXT:    vmov.u8 r0, q1[0]
+; CHECK-LV-NEXT:    vmov.8 q0[14], r0
+; CHECK-LV-NEXT:    vmov.u8 r0, q1[11]
+; CHECK-LV-NEXT:    vmov.8 q0[15], r0
+; CHECK-LV-NEXT:    bx lr
+;
+; CHECK-LIS-LABEL: shuffle3_i8:
+; CHECK-LIS:       @ %bb.0: @ %entry
+; CHECK-LIS-NEXT:    vmov.u8 r0, q0[4]
+; CHECK-LIS-NEXT:    vmov.8 q1[0], r0
+; CHECK-LIS-NEXT:    vmov.u8 r0, q0[5]
+; CHECK-LIS-NEXT:    vmov.8 q1[1], r0
+; CHECK-LIS-NEXT:    vmov.u8 r0, q0[15]
+; CHECK-LIS-NEXT:    vmov.8 q1[2], r0
+; CHECK-LIS-NEXT:    vmov.u8 r0, q0[7]
+; CHECK-LIS-NEXT:    vmov.8 q1[3], r0
+; CHECK-LIS-NEXT:    vmov.u8 r0, q0[14]
+; CHECK-LIS-NEXT:    vmov.8 q1[4], r0
+; CHECK-LIS-NEXT:    vmov.u8 r0, q0[9]
+; CHECK-LIS-NEXT:    vmov.8 q1[5], r0
+; CHECK-LIS-NEXT:    vmov.u8 r0, q0[6]
+; CHECK-LIS-NEXT:    vmov.8 q1[6], r0
+; CHECK-LIS-NEXT:    vmov.u8 r0, q0[3]
+; CHECK-LIS-NEXT:    vmov.8 q1[7], r0
+; CHECK-LIS-NEXT:    vmov.u8 r0, q0[10]
+; CHECK-LIS-NEXT:    vmov.8 q1[8], r0
+; CHECK-LIS-NEXT:    vmov.u8 r0, q0[12]
+; CHECK-LIS-NEXT:    vmov.8 q1[9], r0
+; CHECK-LIS-NEXT:    vmov.u8 r0, q0[1]
+; CHECK-LIS-NEXT:    vmov.8 q1[10], r0
+; CHECK-LIS-NEXT:    vmov.u8 r0, q0[13]
+; CHECK-LIS-NEXT:    vmov.8 q1[11], r0
+; CHECK-LIS-NEXT:    vmov.u8 r0, q0[2]
+; CHECK-LIS-NEXT:    vmov.8 q1[12], r0
+; CHECK-LIS-NEXT:    vmov.u8 r0, q0[8]
+; CHECK-LIS-NEXT:    vmov.8 q1[13], r0
+; CHECK-LIS-NEXT:    vmov.u8 r0, q0[0]
+; CHECK-LIS-NEXT:    vmov.8 q1[14], r0
+; CHECK-LIS-NEXT:    vmov.u8 r0, q0[11]
+; CHECK-LIS-NEXT:    vmov.8 q1[15], r0
+; CHECK-LIS-NEXT:    vmov q0, q1
+; CHECK-LIS-NEXT:    bx lr
 entry:
   %out = shufflevector <16 x i8> %src, <16 x i8> undef, <16 x i32> <i32 4, i32 5, i32 15, i32 7, i32 14, i32 9, i32 6, i32 3, i32 10, i32 12, i32 1, i32 13, i32 2, i32 8, i32 0, i32 11>
   ret <16 x i8> %out
@@ -1143,18 +1195,31 @@ entry:
 }
 
 define arm_aapcs_vfpcc <8 x half> @shuffle3_f16(<8 x half> %src) {
-; CHECK-LABEL: shuffle3_f16:
-; CHECK:       @ %bb.0: @ %entry
-; CHECK-NEXT:    vmov q1, q0
-; CHECK-NEXT:    vmovx.f16 s2, s5
-; CHECK-NEXT:    vmovx.f16 s0, s4
-; CHECK-NEXT:    vins.f16 s5, s4
-; CHECK-NEXT:    vins.f16 s2, s0
-; CHECK-NEXT:    vmov.f32 s3, s5
-; CHECK-NEXT:    vmovx.f16 s1, s7
-; CHECK-NEXT:    vmov.f32 s0, s6
-; CHECK-NEXT:    vins.f16 s1, s7
-; CHECK-NEXT:    bx lr
+; CHECK-LV-LABEL: shuffle3_f16:
+; CHECK-LV:       @ %bb.0: @ %entry
+; CHECK-LV-NEXT:    vmov q1, q0
+; CHECK-LV-NEXT:    vmovx.f16 s2, s5
+; CHECK-LV-NEXT:    vmovx.f16 s0, s4
+; CHECK-LV-NEXT:    vins.f16 s5, s4
+; CHECK-LV-NEXT:    vins.f16 s2, s0
+; CHECK-LV-NEXT:    vmov.f32 s3, s5
+; CHECK-LV-NEXT:    vmovx.f16 s1, s7
+; CHECK-LV-NEXT:    vmov.f32 s0, s6
+; CHECK-LV-NEXT:    vins.f16 s1, s7
+; CHECK-LV-NEXT:    bx lr
+;
+; CHECK-LIS-LABEL: shuffle3_f16:
+; CHECK-LIS:       @ %bb.0: @ %entry
+; CHECK-LIS-NEXT:    vmovx.f16 s5, s3
+; CHECK-LIS-NEXT:    vmovx.f16 s6, s1
+; CHECK-LIS-NEXT:    vmovx.f16 s4, s0
+; CHECK-LIS-NEXT:    vins.f16 s1, s0
+; CHECK-LIS-NEXT:    vins.f16 s6, s4
+; CHECK-LIS-NEXT:    vins.f16 s5, s3
+; CHECK-LIS-NEXT:    vmov.f32 s7, s1
+; CHECK-LIS-NEXT:    vmov.f32 s4, s2
+; CHECK-LIS-NEXT:    vmov q0, q1
+; CHECK-LIS-NEXT:    bx lr
 entry:
   %out = shufflevector <8 x half> %src, <8 x half> undef, <8 x i32> <i32 4, i32 5, i32 7, i32 6, i32 3, i32 1, i32 2, i32 0>
   ret <8 x half> %out
diff --git a/llvm/test/CodeGen/Thumb2/mve-vld3.ll b/llvm/test/CodeGen/Thumb2/mve-vld3.ll
index 4895eabb71ec0..8a94e571e9836 100644
--- a/llvm/test/CodeGen/Thumb2/mve-vld3.ll
+++ b/llvm/test/CodeGen/Thumb2/mve-vld3.ll
@@ -1,5 +1,6 @@
 ; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py
-; RUN: llc -mtriple=thumbv8.1m.main-none-none-eabi -mattr=+mve.fp,+fp64 -verify-machineinstrs %s -o - | FileCheck %s
+; RUN: llc -mtriple=thumbv8.1m.main-none-none-eabi -mattr=+mve.fp,+fp64 -verify-machineinstrs %s -o - | FileCheck %s -check-prefixes=CHECK,CHECK-LV
+; RUN: llc -early-live-intervals -mtriple=thumbv8.1m.main-none-none-eabi -mattr=+mve.fp,+fp64 -verify-machineinstrs %s -o - | FileCheck %s -check-prefixes=CHECK,CHECK-LIS
 
 ; i32
 
@@ -67,46 +68,87 @@ entry:
 }
 
 define void @vld3_v8i32(ptr %src, ptr %dst) {
-; CHECK-LABEL: vld3_v8i32:
-; CHECK:       @ %bb.0: @ %entry
-; CHECK-NEXT:    .vsave {d8, d9, d10, d11}
-; CHECK-NEXT:    vpush {d8, d9, d10, d11}
-; CHECK-NEXT:    vldrw.u32 q0, [r0, #64]
-; CHECK-NEXT:    vldrw.u32 q1, [r0, #48]
-; CHECK-NEXT:    vldrw.u32 q4, [r0, #80]
-; CHECK-NEXT:    vmov.f32 s10, s2
-; CHECK-NEXT:    vmov.f32 s13, s0
-; CHECK-NEXT:    vmov.f32 s14, s3
-; CHECK-NEXT:    vmov.f32 s8, s4
-; CHECK-NEXT:    vmov.f32 s9, s7
-; CHECK-NEXT:    vmov.f32 s12, s5
-; CHECK-NEXT:    vmov.f32 s15, s18
-; CHECK-NEXT:    vmov.f32 s11, s17
-; CHECK-NEXT:    vadd.i32 q2, q2, q3
-; CHECK-NEXT:    vmov.f32 s0, s6
-; CHECK-NEXT:    vmov.f32 s2, s16
-; CHECK-NEXT:    vldrw.u32 q1, [r0, #16]
-; CHECK-NEXT:    vmov.f32 s3, s19
-; CHECK-NEXT:    vldrw.u32 q3, [r0, #32]
-; CHECK-NEXT:    vadd.i32 q0, q2, q0
-; CHECK-NEXT:    vldrw.u32 q2, [r0]
-; CHECK-NEXT:    vmov.f32 s17, s4
-; CHECK-NEXT:    vstrw.32 q0, [r1, #16]
-; CHECK-NEXT:    vmov.f32 s18, s7
-; CHECK-NEXT:    vmov.f32 s22, s6
-; CHECK-NEXT:    vmov.f32 s16, s9
-; CHECK-NEXT:    vmov.f32 s19, s14
-; CHECK-NEXT:    vmov.f32 s20, s8
-; CHECK-NEXT:    vmov.f32 s21, s11
-; CHECK-NEXT:    vmov.f32 s23, s13
-; CHECK-NEXT:    vadd.i32 q4, q5, q4
-; CHECK-NEXT:    vmov.f32 s4, s10
-; CHECK-NEXT:    vmov.f32 s6, s12
-; CHECK-NEXT:    vmov.f32 s7, s15
-; CHECK-NEXT:    vadd.i32 q1, q4, q1
-; CHECK-NEXT:    vstrw.32 q1, [r1]
-; CHECK-NEXT:    vpop {d8, d9, d10, d11}
-; CHECK-NEXT:    bx lr
+; CHECK-LV-LABEL: vld3_v8i32:
+; CHECK-LV:       @ %bb.0: @ %entry
+; CHECK-LV-NEXT:    .vsave {d8, d9, d10, d11}
+; CHECK-LV-NEXT:    vpush {d8, d9, d10, d11}
+; CHECK-LV-NEXT:    vldrw.u32 q0, [r0, #64]
+; CHECK-LV-NEXT:    vldrw.u32 q1, [r0, #48]
+; CHECK-LV-NEXT:    vldrw.u32 q4, [r0, #80]
+; CHECK-LV-NEXT:    vmov.f32 s10, s2
+; CHECK-LV-NEXT:    vmov.f32 s13, s0
+; CHECK-LV-NEXT:    vmov.f32 s14, s3
+; CHECK-LV-NEXT:    vmov.f32 s8, s4
+; CHECK-LV-NEXT:    vmov.f32 s9, s7
+; CHECK-LV-NEXT:    vmov.f32 s12, s5
+; CHECK-LV-NEXT:    vmov.f32 s15, s18
+; CHECK-LV-NEXT:    vmov.f32 s11, s17
+; CHECK-LV-NEXT:    vadd.i32 q2, q2, q3
+; CHECK-LV-NEXT:    vmov.f32 s0, s6
+; CHECK-LV-NEXT:    vmov.f32 s2, s16
+; CHECK-LV-NEXT:    vldrw.u32 q1, [r0, #16]
+; CHECK-LV-NEXT:    vmov.f32 s3, s19
+; CHECK-LV-NEXT:    vldrw.u32 q3, [r0, #32]
+; CHECK-LV-NEXT:    vadd.i32 q0, q2, q0
+; CHECK-LV-NEXT:    vldrw.u32 q2, [r0]
+; CHECK-LV-NEXT:    vmov.f32 s17, s4
+; CHECK-LV-NEXT:    vstrw.32 q0, [r1, #16]
+; CHECK-LV-NEXT:    vmov.f32 s18, s7
+; CHECK-LV-NEXT:    vmov.f32 s22, s6
+; CHECK-LV-NEXT:    vmov.f32 s16, s9
+; CHECK-LV-NEXT:    vmov.f32 s19, s14
+; CHECK-LV-NEXT:    vmov.f32 s20, s8
+; CHECK-LV-NEXT:    vmov.f32 s21, s11
+; CHECK-LV-NEXT:    vmov.f32 s23, s13
+; CHECK-LV-NEXT:    vadd.i32 q4, q5, q4
+; CHECK-LV-NEXT:    vmov.f32 s4, s10
+; CHECK-LV-NEXT:    vmov.f32 s6, s12
+; CHECK-LV-NEXT:    vmov.f32 s7, s15
+; CHECK-LV-NEXT:    vadd.i32 q1, q4, q1
+; CHECK-LV-NEXT:    vstrw.32 q1, [r1]
+; CHECK-LV-NEXT:    vpop {d8, d9, d10, d11}
+; CHECK-LV-NEXT:    bx lr
+;
+; CHECK-LIS-LABEL: vld3_v8i32:
+; CHECK-LIS:       @ %bb.0: @ %entry
+; CHECK-LIS-NEXT:    .vsave {d8, d9, d10, d11}
+; CHECK-LIS-NEXT:    vpush {d8, d9, d10, d11}
+; CHECK-LIS-NEXT:    vldrw.u32 q0, [r0, #64]
+; CHECK-LIS-NEXT:    vldrw.u32 q1, [r0, #48]
+; CHECK-LIS-NEXT:    vldrw.u32 q4, [r0, #80]
+; CHECK-LIS-NEXT:    vmov.f32 s10, s2
+; CHECK-LIS-NEXT:    vmov.f32 s13, s0
+; CHECK-LIS-NEXT:    vmov.f32 s14, s3
+; CHECK-LIS-NEXT:    vmov.f32 s8, s4
+; CHECK-LIS-NEXT:    vmov.f32 s9, s7
+; CHECK-LIS-NEXT:    vmov.f32 s12, s5
+; CHECK-LIS-NEXT:    vmov.f32 s15, s18
+; CHECK-LIS-NEXT:    vmov.f32 s11, s17
+; CHECK-LIS-NEXT:    vmov.f32 s0, s6
+; CHECK-LIS-NEXT:    vadd.i32 q2, q2, q3
+; CHECK-LIS-NEXT:    vmov.f32 s2, s16
+; CHECK-LIS-NEXT:    vldrw.u32 q1, [r0, #16]
+; CHECK-LIS-NEXT:    vmov.f32 s3, s19
+; CHECK-LIS-NEXT:    vldrw.u32 q4, [r0, #32]
+; CHECK-LIS-NEXT:    vadd.i32 q0, q2, q0
+; CHECK-LIS-NEXT:    vldrw.u32 q2, [r0]
+; CHECK-LIS-NEXT:    vmov.f32 s13, s4
+; CHECK-LIS-NEXT:    vstrw.32 q0, [r1, #16]
+; CHECK-LIS-NEXT:    vmov.f32 s14, s7
+; CHECK-LIS-NEXT:    vmov.f32 s22, s6
+; CHECK-LIS-NEXT:    vmov.f32 s12, s9
+; CHECK-LIS-NEXT:    vmov.f32 s15, s18
+; CHECK-LIS-NEXT:    vmov.f32 s20, s8
+; CHECK-LIS-NEXT:    vmov.f32 s21, s11
+; CHECK-LIS-NEXT:    vmov.f32 s23, s17
+; CHECK-LIS-NEXT:    vadd.i32 q3, q5, q3
+; CHECK-LIS-NEXT:    vmov.f32 s4, s10
+; CHECK-LIS-NEXT:    vmov.f32 s6, s16
+; CHECK-LIS-NEXT:    vmov.f32 s7, s19
+; CHECK-LIS-NEXT:    vadd.i32 q1, q3, q1
+; CHECK-LIS-NEXT:    vstrw.32 q1, [r1]
+; CHECK-LIS-NEXT:    vpop {d8, d9, d10, d11}
+; CHECK-LIS-NEXT:    bx lr
 entry:
   %l1 = load <24 x i32>, ptr %src, align 4
   %s1 = shufflevector <24 x i32> %l1, <24 x i32> undef, <8 x i32> <i32 0, i32 3, i32 6, i32 9, i32 12, i32 15, i32 18, i32 21>
@@ -119,80 +161,155 @@ entry:
 }
 
 define void @vld3_v16i32(ptr %src, ptr %dst) {
-; CHECK-LABEL: vld3_v16i32:
-; CHECK:       @ %bb.0: @ %entry
-; CHECK-NEXT:    .vsave {d8, d9, d10, d11, d12, d13, d14, d15}
-; CHECK-NEXT:    vpush {d8, d9, d10, d11, d12, d13, d14, d15}
-; CHECK-NEXT:    vldrw.u32 q0, [r0, #64]
-; CHECK-NEXT:    vldrw.u32 q1, [r0, #48]
-; CHECK-NEXT:    vldrw.u32 q4, [r0, #80]
-; CHECK-NEXT:    vldrw.u32 q6, [r0, #176]
-; CHECK-NEXT:    vmov.f32 s10, s2
-; CHECK-NEXT:    vmov.f32 s13, s0
-; CHECK-NEXT:    vmov.f32 s14, s3
-; CHECK-NEXT:    vmov.f32 s8, s4
-; CHECK-NEXT:    vmov.f32 s9, s7
-; CHECK-NEXT:    vmov.f32 s12, s5
-; CHECK-NEXT:    vmov.f32 s15, s18
-; CHECK-NEXT:    vmov.f32 s11, s17
-; CHECK-NEXT:    vadd.i32 q2, q2, q3
-; CHECK-NEXT:    vmov.f32 s0, s6
-; CHECK-NEXT:    vmov.f32 s2, s16
-; CHECK-NEXT:    vldrw.u32 q1, [r0, #16]
-; CHECK-NEXT:    vmov.f32 s3, s19
-; CHECK-NEXT:    vldrw.u32 q3, [r0, #32]
-; CHECK-NEXT:    vadd.i32 q0, q2, q0
-; CHECK-NEXT:    vldrw.u32 q2, [r0]
-; CHECK-NEXT:    vmov.f32 s17, s4
-; CHECK-NEXT:    vmov.f32 s18, s7
-; CHECK-NEXT:    vmov.f32 s22, s6
-; CHECK-NEXT:    vmov.f32 s16, s9
-; CHECK-NEXT:    vmov.f32 s19, s14
-; CHECK-NEXT:    vmov.f32 s20, s8
-; CHECK-NEXT:    vmov.f32 s21, s11
-; CHECK-NEXT:    vmov.f32 s23, s13
-; CHECK-NEXT:    vmov.f32 s4, s10
-; CHECK-NEXT:    vldrw.u32 q2, [r0, #160]
-; CHECK-NEXT:    vmov.f32 s6, s12
-; CHECK-NEXT:    vadd.i32 q4, q5, q4
-; CHECK-NEXT:    vmov.f32 s7, s15
-; CHECK-NEXT:    vldrw.u32 q3, [r0, #144]
-; CHECK-NEXT:    vadd.i32 q1, q4, q1
-; CHECK-NEXT:    vmov.f32 s18, s10
-; CHECK-NEXT:    vmov.f32 s21, s8
-; CHECK-NEXT:    vmov.f32 s22, s11
-; CHECK-NEXT:    vmov.f32 s16, s12
-; CHECK-NEXT:    vmov.f32 s17, s15
-; CHECK-NEXT:    vmov.f32 s20, s13
-; CHECK-NEXT:    vmov.f32 s23, s26
-; CHECK-NEXT:    vmov.f32 s19, s25
-; CHECK-NEXT:    vadd.i32 q4, q4, q5
-; CHECK-NEXT:    vmov.f32 s8, s14
-; CHECK-NEXT:    vmov.f32 s10, s24
-; CHECK-NEXT:    vldrw.u32 q3, [r0, #112]
-; CHECK-NEXT:    vmov.f32 s11, s27
-; CHECK-NEXT:    vldrw.u32 q5, [r0, #128]
-; CHECK-NEXT:    vadd.i32 q2, q4, q2
-; CHECK-NEXT:    vldrw.u32 q4, [r0, #96]
-; CHECK-NEXT:    vmov.f32 s25, s12
-; CHECK-NEXT:    vstrw.32 q2, [r1, #48]
-; CHECK-NEXT:    vmov.f32 s26, s15
-; CHECK-NEXT:    vstrw.32 q0, [r1, #16]
-; CHECK-NEXT:    vmov.f32 s30, s14
-; CHECK-NEXT:    vstrw.32 q1, [r1]
-; CHECK-NEXT:    vmov.f32 s24, s17
-; CHECK-NEXT:    vmov.f32 s27, s22
-; CHECK-NEXT:    vmov.f32 s28, s16
-; CHECK-NEXT:    vmov.f32 s29, s19
-; CHECK-NEXT:    vmov.f32 s31, s21
-; CHECK-NEXT:    vadd.i32 q6, q7, q6
-; CHECK-NEXT:    vmov.f32 s12, s18
-; CHECK-NEXT:    vmov.f32 s14, s20
-; CHECK-NEXT:    vmov.f32 s15, s23
-; CHECK-NEXT:    vadd.i32 q3, q6, q3
-; CHECK-NEXT:    vstrw.32 q3, [r1, #32]
-; CHECK-NEXT:    vpop {d8, d9, d10, d11, d12, d13, d14, d15}
-; CHECK-NEXT:    bx lr
+; CHECK-LV-LABEL: vld3_v16i32:
+; CHECK-LV:       @ %bb.0: @ %entry
+; CHECK-LV-NEXT:    .vsave {d8, d9, d10, d11, d12, d13, d14, d15}
+; CHECK-LV-NEXT:    vpush {d8, d9, d10, d11, d12, d13, d14, d15}
+; CHECK-LV-NEXT:    vldrw.u32 q0, [r0, #64]
+; CHECK-LV-NEXT:    vldrw.u32 q1, [r0, #48]
+; CHECK-LV-NEXT:    vldrw.u32 q4, [r0, #80]
+; CHECK-LV-NEXT:    vldrw.u32 q6, [r0, #176]
+; CHECK-LV-NEXT:    vmov.f32 s10, s2
+; CHECK-LV-NEXT:    vmov.f32 s13, s0
+; CHECK-LV-NEXT:    vmov.f32 s14, s3
+; CHECK-LV-NEXT:    vmov.f32 s8, s4
+; CHECK-LV-NEXT:    vmov.f32 s9, s7
+; CHECK-LV-NEXT:    vmov.f32 s12, s5
+; CHECK-LV-NEXT:    vmov.f32 s15, s18
+; CHECK-LV-NEXT:    vmov.f32 s11, s17
+; CHECK-LV-NEXT:    vadd.i32 q2, q2, q3
+; CHECK-LV-NEXT:    vmov.f32 s0, s6
+; CHECK-LV-NEXT:    vmov.f32 s2, s16
+; CHECK-LV-NEXT:    vldrw.u32 q1, [r0, #16]
+; CHECK-LV-NEXT:    vmov.f32 s3, s19
+; CHECK-LV-NEXT:    vldrw.u32 q3, [r0, #32]
+; CHECK-LV-NEXT:    vadd.i32 q0, q2, q0
+; CHECK-LV-NEXT:    vldrw.u32 q2, [r0]
+; CHECK-LV-NEXT:    vmov.f32 s17, s4
+; CHECK-LV-NEXT:    vmov.f32 s18, s7
+; CHECK-LV-NEXT:    vmov.f32 s22, s6
+; CHECK-LV-NEXT:    vmov.f32 s16, s9
+; CHECK-LV-NEXT:    vmov.f32 s19, s14
+; CHECK-LV-NEXT:    vmov.f32 s20, s8
+; CHECK-LV-NEXT:    vmov.f32 s21, s11
+; CHECK-LV-NEXT:    vmov.f32 s23, s13
+; CHECK-LV-NEXT:    vmov.f32 s4, s10
+; CHECK-LV-NEXT:    vldrw.u32 q2, [r0, #160]
+; CHECK-LV-NEXT:    vmov.f32 s6, s12
+; CHECK-LV-NEXT:    vadd.i32 q4, q5, q4
+; CHECK-LV-NEXT:    vmov.f32 s7, s15
+; CHECK-LV-NEXT:    vldrw.u32 q3, [r0, #144]
+; CHECK-LV-NEXT:    vadd.i32 q1, q4, q1
+; CHECK-LV-NEXT:    vmov.f32 s18, s10
+; CHECK-LV-NEXT:    vmov.f32 s21, s8
+; CHECK-LV-NEXT:    vmov.f32 s22, s11
+; CHECK-LV-NEXT:    vmov.f32 s16, s12
+; CHECK-LV-NEXT:    vmov.f32 s17, s15
+; CHECK-LV-NEXT:    vmov.f32 s20, s13
+; CHECK-LV-NEXT:    vmov.f32 s23, s26
+; CHECK-LV-NEXT:    vmov.f32 s19, s25
+; CHECK-LV-NEXT:    vadd.i32 q4, q4, q5
+; CHECK-LV-NEXT:    vmov.f32 s8, s14
+; CHECK-LV-NEXT:    vmov.f32 s10, s24
+; CHECK-LV-NEXT:    vldrw.u32 q3, [r0, #112]
+; CHECK-LV-NEXT:    vmov.f32 s11, s27
+; CHECK-LV-NEXT:    vldrw.u32 q5, [r0, #128]
+; CHECK-LV-NEXT:    vadd.i32 q2, q4, q2
+; CHECK-LV-NEXT:    vldrw.u32 q4, [r0, #96]
+; CHECK-LV-NEXT:    vmov.f32 s25, s12
+; CHECK-LV-NEXT:    vstrw.32 q2, [r1, #48]
+; CHECK-LV-NEXT:    vmov.f32 s26, s15
+; CHECK-LV-NEXT:    vstrw.32 q0, [r1, #16]
+; CHECK-LV-NEXT:    vmov.f32 s30, s14
+; CHECK-LV-NEXT:    vstrw.32 q1, [r1]
+; CHECK-LV-NEXT:    vmov.f32 s24, s17
+; CHECK-LV-NEXT:    vmov.f32 s27, s22
+; CHECK-LV-NEXT:    vmov.f32 s28, s16
+; CHECK-LV-NEXT:    vmov.f32 s29, s19
+; CHECK-LV-NEXT:    vmov.f32 s31, s21
+; CHECK-LV-NEXT:    vadd.i32 q6, q7, q6
+; CHECK-LV-NEXT:    vmov.f32 s12, s18
+; CHECK-LV-NEXT:    vmov.f32 s14, s20
+; CHECK-LV-NEXT:    vmov.f32 s15, s23
+; CHECK-LV-NEXT:    vadd.i32 q3, q6, q3
+; CHECK-LV-NEXT:    vstrw.32 q3, [r1, #32]
+; CHECK-LV-NEXT:    vpop {d8, d9, d10, d11, d12, d13, d14, d15}
+; CHECK-LV-NEXT:    bx lr
+;
+; CHECK-LIS-LABEL: vld3_v16i32:
+; CHECK-LIS:       @ %bb.0: @ %entry
+; CHECK-LIS-NEXT:    .vsave {d8, d9, d10, d11, d12, d13, d14, d15}
+; CHECK-LIS-NEXT:    vpush {d8, d9, d10, d11, d12, d13, d14, d15}
+; CHECK-LIS-NEXT:    vldrw.u32 q0, [r0, #64]
+; CHECK-LIS-NEXT:    vldrw.u32 q1, [r0, #48]
+; CHECK-LIS-NEXT:    vldrw.u32 q4, [r0, #80]
+; CHECK-LIS-NEXT:    vldrw.u32 q6, [r0, #176]
+; CHECK-LIS-NEXT:    vmov.f32 s10, s2
+; CHECK-LIS-NEXT:    vmov.f32 s13, s0
+; CHECK-LIS-NEXT:    vmov.f32 s14, s3
+; CHECK-LIS-NEXT:    vmov.f32 s8, s4
+; CHECK-LIS-NEXT:    vmov.f32 s9, s7
+; CHECK-LIS-NEXT:    vmov.f32 s12, s5
+; CHECK-LIS-NEXT:    vmov.f32 s15, s18
+; CHECK-LIS-NEXT:    vmov.f32 s11, s17
+; CHECK-LIS-NEXT:    vmov.f32 s0, s6
+; CHECK-LIS-NEXT:    vadd.i32 q2, q2, q3
+; CHECK-LIS-NEXT:    vmov.f32 s2, s16
+; CHECK-LIS-NEXT:    vldrw.u32 q1, [r0, #16]
+; CHECK-LIS-NEXT:    vmov.f32 s3, s19
+; CHECK-LIS-NEXT:    vldrw.u32 q4, [r0, #32]
+; CHECK-LIS-NEXT:    vadd.i32 q0, q2, q0
+; CHECK-LIS-NEXT:    vldrw.u32 q2, [r0]
+; CHECK-LIS-NEXT:    vmov.f32 s13, s4
+; CHECK-LIS-NEXT:    vmov.f32 s14, s7
+; CHECK-LIS-NEXT:    vmov.f32 s22, s6
+; CHECK-LIS-NEXT:    vmov.f32 s12, s9
+; CHECK-LIS-NEXT:    vmov.f32 s15, s18
+; CHECK-LIS-NEXT:    vmov.f32 s20, s8
+; CHECK-LIS-NEXT:    vmov.f32 s21, s11
+; CHECK-LIS-NEXT:    vmov.f32 s23, s17
+; CHECK-LIS-NEXT:    vadd.i32 q3, q5, q3
+; CHECK-LIS-NEXT:    vmov.f32 s4, s10
+; CHECK-LIS-NEXT:    vmov.f32 s7, s19
+; CHECK-LIS-NEXT:    vldrw.u32 q2, [r0, #160]
+; CHECK-LIS-NEXT:    vmov.f32 s6, s16
+; CHECK-LIS-NEXT:    vadd.i32 q1, q3, q1
+; CHECK-LIS-NEXT:    vldrw.u32 q3, [r0, #144]
+; CHECK-LIS-NEXT:    vmov.f32 s18, s10
+; CHECK-LIS-NEXT:    vmov.f32 s21, s8
+; CHECK-LIS-NEXT:    vmov.f32 s22, s11
+; CHECK-LIS-NEXT:    vmov.f32 s16, s12
+; CHECK-LIS-NEXT:    vmov.f32 s17, s15
+; CHECK-LIS-NEXT:    vmov.f32 s20, s13
+; CHECK-LIS-NEXT:    vmov.f32 s23, s26
+; CHECK-LIS-NEXT:    vmov.f32 s19, s25
+; CHECK-LIS-NEXT:    vmov.f32 s8, s14
+; CHECK-LIS-NEXT:    vadd.i32 q4, q4, q5
+; CHECK-LIS-NEXT:    vmov.f32 s10, s24
+; CHECK-LIS-NEXT:    vldrw.u32 q3, [r0, #112]
+; CHECK-LIS-NEXT:    vmov.f32 s11, s27
+; CHECK-LIS-NEXT:    vldrw.u32 q6, [r0, #128]
+; CHECK-LIS-NEXT:    vadd.i32 q2, q4, q2
+; CHECK-LIS-NEXT:    vldrw.u32 q4, [r0, #96]
+; CHECK-LIS-NEXT:    vmov.f32 s21, s12
+; CHECK-LIS-NEXT:    vstrw.32 q2, [r1, #48]
+; CHECK-LIS-NEXT:    vmov.f32 s22, s15
+; CHECK-LIS-NEXT:    vstrw.32 q0, [r1, #16]
+; CHECK-LIS-NEXT:    vmov.f32 s30, s14
+; CHECK-LIS-NEXT:    vstrw.32 q1, [r1]
+; CHECK-LIS-NEXT:    vmov.f32 s20, s17
+; CHECK-LIS-NEXT:    vmov.f32 s23, s26
+; CHECK-LIS-NEXT:    vmov.f32 s28, s16
+; CHECK-LIS-NEXT:    vmov.f32 s29, s19
+; CHECK-LIS-NEXT:    vmov.f32 s31, s25
+; CHECK-LIS-NEXT:    vadd.i32 q5, q7, q5
+; CHECK-LIS-NEXT:    vmov.f32 s12, s18
+; CHECK-LIS-NEXT:    vmov.f32 s14, s24
+; CHECK-LIS-NEXT:    vmov.f32 s15, s27
+; CHECK-LIS-NEXT:    vadd.i32 q3, q5, q3
+; CHECK-LIS-NEXT:    vstrw.32 q3, [r1, #32]
+; CHECK-LIS-NEXT:    vpop {d8, d9, d10, d11, d12, d13, d14, d15}
+; CHECK-LIS-NEXT:    bx lr
 entry:
   %l1 = load <48 x i32>, ptr %src, align 4
   %s1 = shufflevector <48 x i32> %l1, <48 x i32> undef, <16 x i32> <i32 0, i32 3, i32 6, i32 9, i32 12, i32 15, i32 18, i32 21, i32 24, i32 27, i32 30, i32 33, i32 36, i32 39, i32 42, i32 45>
@@ -247,32 +364,59 @@ entry:
 }
 
 define void @vld3_v4i16(ptr %src, ptr %dst) {
-; CHECK-LABEL: vld3_v4i16:
-; CHECK:       @ %bb.0: @ %entry
-; CHECK-NEXT:    .save {r4, r5, r6, lr}
-; CHECK-NEXT:    push {r4, r5, r6, lr}
-; CHECK-NEXT:    vldrw.u32 q0, [r0]
-; CHECK-NEXT:    vldrh.u32 q1, [r0, #16]
-; CHECK-NEXT:    vmov.u16 r5, q0[6]
-; CHECK-NEXT:    vmov.u16 r6, q0[0]
-; CHECK-NEXT:    vmov r0, r3, d2
-; CHECK-NEXT:    vmov.u16 lr, q0[2]
-; CHECK-NEXT:    vmov r2, r4, d3
-; CHECK-NEXT:    vmov q1[2], q1[0], r6, r5
-; CHECK-NEXT:    vmov.u16 r5, q0[7]
-; CHECK-NEXT:    vmov.u16 r6, q0[1]
-; CHECK-NEXT:    vmov q2[2], q2[0], r6, r5
-; CHECK-NEXT:    vmov.u16 r5, q0[3]
-; CHECK-NEXT:    vmov.u16 r6, q0[4]
-; CHECK-NEXT:    vmov q1[3], q1[1], r5, r3
-; CHECK-NEXT:    vmov q2[3], q2[1], r6, r2
-; CHECK-NEXT:    vmov.u16 r12, q0[5]
-; CHECK-NEXT:    vadd.i32 q0, q1, q2
-; CHECK-NEXT:    vmov q1[2], q1[0], lr, r0
-; CHECK-NEXT:    vmov q1[3], q1[1], r12, r4
-; CHECK-NEXT:    vadd.i32 q0, q0, q1
-; CHECK-NEXT:    vstrh.32 q0, [r1]
-; CHECK-NEXT:    pop {r4, r5, r6, pc}
+; CHECK-LV-LABEL: vld3_v4i16:
+; CHECK-LV:       @ %bb.0: @ %entry
+; CHECK-LV-NEXT:    .save {r4, r5, r6, lr}
+; CHECK-LV-NEXT:    push {r4, r5, r6, lr}
+; CHECK-LV-NEXT:    vldrw.u32 q0, [r0]
+; CHECK-LV-NEXT:    vldrh.u32 q1, [r0, #16]
+; CHECK-LV-NEXT:    vmov.u16 r5, q0[6]
+; CHECK-LV-NEXT:    vmov.u16 r6, q0[0]
+; CHECK-LV-NEXT:    vmov r0, r3, d2
+; CHECK-LV-NEXT:    vmov.u16 lr, q0[2]
+; CHECK-LV-NEXT:    vmov r2, r4, d3
+; CHECK-LV-NEXT:    vmov q1[2], q1[0], r6, r5
+; CHECK-LV-NEXT:    vmov.u16 r5, q0[7]
+; CHECK-LV-NEXT:    vmov.u16 r6, q0[1]
+; CHECK-LV-NEXT:    vmov q2[2], q2[0], r6, r5
+; CHECK-LV-NEXT:    vmov.u16 r5, q0[3]
+; CHECK-LV-NEXT:    vmov.u16 r6, q0[4]
+; CHECK-LV-NEXT:    vmov q1[3], q1[1], r5, r3
+; CHECK-LV-NEXT:    vmov q2[3], q2[1], r6, r2
+; CHECK-LV-NEXT:    vmov.u16 r12, q0[5]
+; CHECK-LV-NEXT:    vadd.i32 q0, q1, q2
+; CHECK-LV-NEXT:    vmov q1[2], q1[0], lr, r0
+; CHECK-LV-NEXT:    vmov q1[3], q1[1], r12, r4
+; CHECK-LV-NEXT:    vadd.i32 q0, q0, q1
+; CHECK-LV-NEXT:    vstrh.32 q0, [r1]
+; CHECK-LV-NEXT:    pop {r4, r5, r6, pc}
+;
+; CHECK-LIS-LABEL: vld3_v4i16:
+; CHECK-LIS:       @ %bb.0: @ %entry
+; CHECK-LIS-NEXT:    .save {r4, r5, r6, lr}
+; CHECK-LIS-NEXT:    push {r4, r5, r6, lr}
+; CHECK-LIS-NEXT:    vldrw.u32 q0, [r0]
+; CHECK-LIS-NEXT:    vldrh.u32 q1, [r0, #16]
+; CHECK-LIS-NEXT:    vmov.u16 r5, q0[6]
+; CHECK-LIS-NEXT:    vmov.u16 r6, q0[0]
+; CHECK-LIS-NEXT:    vmov r0, r2, d2
+; CHECK-LIS-NEXT:    vmov.u16 r12, q0[2]
+; CHECK-LIS-NEXT:    vmov r3, r4, d3
+; CHECK-LIS-NEXT:    vmov q1[2], q1[0], r6, r5
+; CHECK-LIS-NEXT:    vmov.u16 r5, q0[7]
+; CHECK-LIS-NEXT:    vmov.u16 r6, q0[1]
+; CHECK-LIS-NEXT:    vmov q2[2], q2[0], r6, r5
+; CHECK-LIS-NEXT:    vmov.u16 r5, q0[3]
+; CHECK-LIS-NEXT:    vmov.u16 r6, q0[4]
+; CHECK-LIS-NEXT:    vmov q1[3], q1[1], r5, r2
+; CHECK-LIS-NEXT:    vmov q2[3], q2[1], r6, r3
+; CHECK-LIS-NEXT:    vmov.u16 lr, q0[5]
+; CHECK-LIS-NEXT:    vadd.i32 q0, q1, q2
+; CHECK-LIS-NEXT:    vmov q1[2], q1[0], r12, r0
+; CHECK-LIS-NEXT:    vmov q1[3], q1[1], lr, r4
+; CHECK-LIS-NEXT:    vadd.i32 q0, q0, q1
+; CHECK-LIS-NEXT:    vstrh.32 q0, [r1]
+; CHECK-LIS-NEXT:    pop {r4, r5, r6, pc}
 entry:
   %l1 = load <12 x i16>, ptr %src, align 4
   %s1 = shufflevector <12 x i16> %l1, <12 x i16> undef, <4 x i32> <i32 0, i32 3, i32 6, i32 9>
@@ -340,86 +484,167 @@ entry:
 }
 
 define void @vld3_v16i16(ptr %src, ptr %dst) {
-; CHECK-LABEL: vld3_v16i16:
-; CHECK:       @ %bb.0: @ %entry
-; CHECK-NEXT:    .vsave {d8, d9, d10, d11}
-; CHECK-NEXT:    vpush {d8, d9, d10, d11}
-; CHECK-NEXT:    vldrw.u32 q0, [r0, #48]
-; CHECK-NEXT:    vldrw.u32 q2, [r0, #64]
-; CHECK-NEXT:    vldrw.u32 q3, [r0, #80]
-; CHECK-NEXT:    vmovx.f16 s6, s2
-; CHECK-NEXT:    vmov.f32 s4, s1
-; CHECK-NEXT:    vins.f16 s4, s6
-; CHECK-NEXT:    vmovx.f16 s6, s9
-; CHECK-NEXT:    vmov.f32 s5, s8
-; CHECK-NEXT:    vmovx.f16 s7, s12
-; CHECK-NEXT:    vins.f16 s5, s6
-; CHECK-NEXT:    vmov.f32 s6, s11
-; CHECK-NEXT:    vins.f16 s6, s7
-; CHECK-NEXT:    vmovx.f16 s16, s15
-; CHECK-NEXT:    vmov.f32 s7, s14
-; CHECK-NEXT:    vmovx.f16 s17, s3
-; CHECK-NEXT:    vins.f16 s7, s16
-; CHECK-NEXT:    vmovx.f16 s16, s0
-; CHECK-NEXT:    vins.f16 s16, s2
-; CHECK-NEXT:    vmovx.f16 s2, s1
-; CHECK-NEXT:    vins.f16 s0, s2
-; CHECK-NEXT:    vmovx.f16 s2, s8
-; CHECK-NEXT:    vins.f16 s3, s2
-; CHECK-NEXT:    vmovx.f16 s2, s11
-; CHECK-NEXT:    vmovx.f16 s8, s14
-; CHECK-NEXT:    vmovx.f16 s18, s10
-; CHECK-NEXT:    vmovx.f16 s19, s13
-; CHECK-NEXT:    vins.f16 s10, s2
-; CHECK-NEXT:    vins.f16 s13, s8
-; CHECK-NEXT:    vmov.f32 s1, s3
-; CHECK-NEXT:    vins.f16 s18, s12
-; CHECK-NEXT:    vins.f16 s19, s15
-; CHECK-NEXT:    vmov.f32 s3, s13
-; CHECK-NEXT:    vldrw.u32 q3, [r0, #16]
-; CHECK-NEXT:    vins.f16 s17, s9
-; CHECK-NEXT:    vmov.f32 s2, s10
-; CHECK-NEXT:    vadd.i16 q0, q0, q4
-; CHECK-NEXT:    vldrw.u32 q2, [r0, #32]
-; CHECK-NEXT:    vadd.i16 q0, q0, q1
-; CHECK-NEXT:    vmovx.f16 s6, s14
-; CHECK-NEXT:    vldrw.u32 q4, [r0]
-; CHECK-NEXT:    vins.f16 s6, s8
-; CHECK-NEXT:    vmov.f32 s22, s15
-; CHECK-NEXT:    vmovx.f16 s8, s8
-; CHECK-NEXT:    vins.f16 s22, s8
-; CHECK-NEXT:    vmovx.f16 s8, s11
-; CHECK-NEXT:    vmov.f32 s23, s10
-; CHECK-NEXT:    vmovx.f16 s4, s16
-; CHECK-NEXT:    vins.f16 s23, s8
-; CHECK-NEXT:    vmovx.f16 s8, s17
-; CHECK-NEXT:    vins.f16 s16, s8
-; CHECK-NEXT:    vmovx.f16 s8, s12
-; CHECK-NEXT:    vmovx.f16 s5, s19
-; CHECK-NEXT:    vins.f16 s19, s8
-; CHECK-NEXT:    vmovx.f16 s8, s15
-; CHECK-NEXT:    vmovx.f16 s7, s9
-; CHECK-NEXT:    vins.f16 s14, s8
-; CHECK-NEXT:    vmovx.f16 s8, s10
-; CHECK-NEXT:    vins.f16 s4, s18
-; CHECK-NEXT:    vmov.f32 s20, s17
-; CHECK-NEXT:    vmovx.f16 s18, s18
-; CHECK-NEXT:    vins.f16 s9, s8
-; CHECK-NEXT:    vins.f16 s5, s13
-; CHECK-NEXT:    vins.f16 s20, s18
-; CHECK-NEXT:    vmov.f32 s17, s19
-; CHECK-NEXT:    vins.f16 s7, s11
-; CHECK-NEXT:    vmovx.f16 s13, s13
-; CHECK-NEXT:    vmov.f32 s21, s12
-; CHECK-NEXT:    vmov.f32 s18, s14
-; CHECK-NEXT:    vins.f16 s21, s13
-; CHECK-NEXT:    vmov.f32 s19, s9
-; CHECK-NEXT:    vstrw.32 q0, [r1, #16]
-; CHECK-NEXT:    vadd.i16 q1, q4, q1
-; CHECK-NEXT:    vadd.i16 q1, q1, q5
-; CHECK-NEXT:    vstrw.32 q1, [r1]
-; CHECK-NEXT:    vpop {d8, d9, d10, d11}
-; CHECK-NEXT:    bx lr
+; CHECK-LV-LABEL: vld3_v16i16:
+; CHECK-LV:       @ %bb.0: @ %entry
+; CHECK-LV-NEXT:    .vsave {d8, d9, d10, d11}
+; CHECK-LV-NEXT:    vpush {d8, d9, d10, d11}
+; CHECK-LV-NEXT:    vldrw.u32 q0, [r0, #48]
+; CHECK-LV-NEXT:    vldrw.u32 q2, [r0, #64]
+; CHECK-LV-NEXT:    vldrw.u32 q3, [r0, #80]
+; CHECK-LV-NEXT:    vmovx.f16 s6, s2
+; CHECK-LV-NEXT:    vmov.f32 s4, s1
+; CHECK-LV-NEXT:    vins.f16 s4, s6
+; CHECK-LV-NEXT:    vmovx.f16 s6, s9
+; CHECK-LV-NEXT:    vmov.f32 s5, s8
+; CHECK-LV-NEXT:    vmovx.f16 s7, s12
+; CHECK-LV-NEXT:    vins.f16 s5, s6
+; CHECK-LV-NEXT:    vmov.f32 s6, s11
+; CHECK-LV-NEXT:    vins.f16 s6, s7
+; CHECK-LV-NEXT:    vmovx.f16 s16, s15
+; CHECK-LV-NEXT:    vmov.f32 s7, s14
+; CHECK-LV-NEXT:    vmovx.f16 s17, s3
+; CHECK-LV-NEXT:    vins.f16 s7, s16
+; CHECK-LV-NEXT:    vmovx.f16 s16, s0
+; CHECK-LV-NEXT:    vins.f16 s16, s2
+; CHECK-LV-NEXT:    vmovx.f16 s2, s1
+; CHECK-LV-NEXT:    vins.f16 s0, s2
+; CHECK-LV-NEXT:    vmovx.f16 s2, s8
+; CHECK-LV-NEXT:    vins.f16 s3, s2
+; CHECK-LV-NEXT:    vmovx.f16 s2, s11
+; CHECK-LV-NEXT:    vmovx.f16 s8, s14
+; CHECK-LV-NEXT:    vmovx.f16 s18, s10
+; CHECK-LV-NEXT:    vmovx.f16 s19, s13
+; CHECK-LV-NEXT:    vins.f16 s10, s2
+; CHECK-LV-NEXT:    vins.f16 s13, s8
+; CHECK-LV-NEXT:    vmov.f32 s1, s3
+; CHECK-LV-NEXT:    vins.f16 s18, s12
+; CHECK-LV-NEXT:    vins.f16 s19, s15
+; CHECK-LV-NEXT:    vmov.f32 s3, s13
+; CHECK-LV-NEXT:    vldrw.u32 q3, [r0, #16]
+; CHECK-LV-NEXT:    vins.f16 s17, s9
+; CHECK-LV-NEXT:    vmov.f32 s2, s10
+; CHECK-LV-NEXT:    vadd.i16 q0, q0, q4
+; CHECK-LV-NEXT:    vldrw.u32 q2, [r0, #32]
+; CHECK-LV-NEXT:    vadd.i16 q0, q0, q1
+; CHECK-LV-NEXT:    vmovx.f16 s6, s14
+; CHECK-LV-NEXT:    vldrw.u32 q4, [r0]
+; CHECK-LV-NEXT:    vins.f16 s6, s8
+; CHECK-LV-NEXT:    vmov.f32 s22, s15
+; CHECK-LV-NEXT:    vmovx.f16 s8, s8
+; CHECK-LV-NEXT:    vins.f16 s22, s8
+; CHECK-LV-NEXT:    vmovx.f16 s8, s11
+; CHECK-LV-NEXT:    vmov.f32 s23, s10
+; CHECK-LV-NEXT:    vmovx.f16 s4, s16
+; CHECK-LV-NEXT:    vins.f16 s23, s8
+; CHECK-LV-NEXT:    vmovx.f16 s8, s17
+; CHECK-LV-NEXT:    vins.f16 s16, s8
+; CHECK-LV-NEXT:    vmovx.f16 s8, s12
+; CHECK-LV-NEXT:    vmovx.f16 s5, s19
+; CHECK-LV-NEXT:    vins.f16 s19, s8
+; CHECK-LV-NEXT:    vmovx.f16 s8, s15
+; CHECK-LV-NEXT:    vmovx.f16 s7, s9
+; CHECK-LV-NEXT:    vins.f16 s14, s8
+; CHECK-LV-NEXT:    vmovx.f16 s8, s10
+; CHECK-LV-NEXT:    vins.f16 s4, s18
+; CHECK-LV-NEXT:    vmov.f32 s20, s17
+; CHECK-LV-NEXT:    vmovx.f16 s18, s18
+; CHECK-LV-NEXT:    vins.f16 s9, s8
+; CHECK-LV-NEXT:    vins.f16 s5, s13
+; CHECK-LV-NEXT:    vins.f16 s20, s18
+; CHECK-LV-NEXT:    vmov.f32 s17, s19
+; CHECK-LV-NEXT:    vins.f16 s7, s11
+; CHECK-LV-NEXT:    vmovx.f16 s13, s13
+; CHECK-LV-NEXT:    vmov.f32 s21, s12
+; CHECK-LV-NEXT:    vmov.f32 s18, s14
+; CHECK-LV-NEXT:    vins.f16 s21, s13
+; CHECK-LV-NEXT:    vmov.f32 s19, s9
+; CHECK-LV-NEXT:    vstrw.32 q0, [r1, #16]
+; CHECK-LV-NEXT:    vadd.i16 q1, q4, q1
+; CHECK-LV-NEXT:    vadd.i16 q1, q1, q5
+; CHECK-LV-NEXT:    vstrw.32 q1, [r1]
+; CHECK-LV-NEXT:    vpop {d8, d9, d10, d11}
+; CHECK-LV-NEXT:    bx lr
+;
+; CHECK-LIS-LABEL: vld3_v16i16:
+; CHECK-LIS:       @ %bb.0: @ %entry
+; CHECK-LIS-NEXT:    .vsave {d8, d9, d10, d11}
+; CHECK-LIS-NEXT:    vpush {d8, d9, d10, d11}
+; CHECK-LIS-NEXT:    vldrw.u32 q0, [r0, #48]
+; CHECK-LIS-NEXT:    vldrw.u32 q2, [r0, #64]
+; CHECK-LIS-NEXT:    vldrw.u32 q3, [r0, #80]
+; CHECK-LIS-NEXT:    vmovx.f16 s6, s2
+; CHECK-LIS-NEXT:    vmov.f32 s4, s1
+; CHECK-LIS-NEXT:    vins.f16 s4, s6
+; CHECK-LIS-NEXT:    vmovx.f16 s6, s9
+; CHECK-LIS-NEXT:    vmov.f32 s5, s8
+; CHECK-LIS-NEXT:    vmovx.f16 s7, s12
+; CHECK-LIS-NEXT:    vins.f16 s5, s6
+; CHECK-LIS-NEXT:    vmov.f32 s6, s11
+; CHECK-LIS-NEXT:    vins.f16 s6, s7
+; CHECK-LIS-NEXT:    vmovx.f16 s16, s15
+; CHECK-LIS-NEXT:    vmov.f32 s7, s14
+; CHECK-LIS-NEXT:    vmovx.f16 s17, s3
+; CHECK-LIS-NEXT:    vins.f16 s7, s16
+; CHECK-LIS-NEXT:    vmovx.f16 s16, s0
+; CHECK-LIS-NEXT:    vins.f16 s16, s2
+; CHECK-LIS-NEXT:    vmovx.f16 s2, s1
+; CHECK-LIS-NEXT:    vins.f16 s0, s2
+; CHECK-LIS-NEXT:    vmovx.f16 s2, s8
+; CHECK-LIS-NEXT:    vins.f16 s3, s2
+; CHECK-LIS-NEXT:    vmovx.f16 s2, s11
+; CHECK-LIS-NEXT:    vmovx.f16 s8, s14
+; CHECK-LIS-NEXT:    vmovx.f16 s18, s10
+; CHECK-LIS-NEXT:    vmovx.f16 s19, s13
+; CHECK-LIS-NEXT:    vins.f16 s10, s2
+; CHECK-LIS-NEXT:    vins.f16 s13, s8
+; CHECK-LIS-NEXT:    vmov.f32 s1, s3
+; CHECK-LIS-NEXT:    vins.f16 s18, s12
+; CHECK-LIS-NEXT:    vins.f16 s19, s15
+; CHECK-LIS-NEXT:    vmov.f32 s3, s13
+; CHECK-LIS-NEXT:    vins.f16 s17, s9
+; CHECK-LIS-NEXT:    vmov.f32 s2, s10
+; CHECK-LIS-NEXT:    vldrw.u32 q3, [r0, #16]
+; CHECK-LIS-NEXT:    vadd.i16 q0, q0, q4
+; CHECK-LIS-NEXT:    vldrw.u32 q4, [r0]
+; CHECK-LIS-NEXT:    vadd.i16 q0, q0, q1
+; CHECK-LIS-NEXT:    vldrw.u32 q1, [r0, #32]
+; CHECK-LIS-NEXT:    vmovx.f16 s10, s14
+; CHECK-LIS-NEXT:    vmov.f32 s22, s15
+; CHECK-LIS-NEXT:    vins.f16 s10, s4
+; CHECK-LIS-NEXT:    vmovx.f16 s4, s4
+; CHECK-LIS-NEXT:    vins.f16 s22, s4
+; CHECK-LIS-NEXT:    vmovx.f16 s4, s7
+; CHECK-LIS-NEXT:    vmov.f32 s23, s6
+; CHECK-LIS-NEXT:    vmovx.f16 s8, s16
+; CHECK-LIS-NEXT:    vins.f16 s23, s4
+; CHECK-LIS-NEXT:    vmovx.f16 s4, s17
+; CHECK-LIS-NEXT:    vins.f16 s16, s4
+; CHECK-LIS-NEXT:    vmovx.f16 s4, s12
+; CHECK-LIS-NEXT:    vmovx.f16 s9, s19
+; CHECK-LIS-NEXT:    vins.f16 s19, s4
+; CHECK-LIS-NEXT:    vmovx.f16 s4, s15
+; CHECK-LIS-NEXT:    vmovx.f16 s11, s5
+; CHECK-LIS-NEXT:    vins.f16 s14, s4
+; CHECK-LIS-NEXT:    vmovx.f16 s4, s6
+; CHECK-LIS-NEXT:    vins.f16 s8, s18
+; CHECK-LIS-NEXT:    vmov.f32 s20, s17
+; CHECK-LIS-NEXT:    vmovx.f16 s18, s18
+; CHECK-LIS-NEXT:    vins.f16 s5, s4
+; CHECK-LIS-NEXT:    vins.f16 s9, s13
+; CHECK-LIS-NEXT:    vins.f16 s20, s18
+; CHECK-LIS-NEXT:    vmov.f32 s17, s19
+; CHECK-LIS-NEXT:    vins.f16 s11, s7
+; CHECK-LIS-NEXT:    vmovx.f16 s13, s13
+; CHECK-LIS-NEXT:    vmov.f32 s21, s12
+; CHECK-LIS-NEXT:    vmov.f32 s18, s14
+; CHECK-LIS-NEXT:    vins.f16 s21, s13
+; CHECK-LIS-NEXT:    vmov.f32 s19, s5
+; CHECK-LIS-NEXT:    vstrw.32 q0, [r1, #16]
+; CHECK-LIS-NEXT:    vadd.i16 q1, q4, q2
+; CHECK-LIS-NEXT:    vadd.i16 q1, q1, q5
+; CHECK-LIS-NEXT:    vstrw.32 q1, [r1]
+; CHECK-LIS-NEXT:    vpop {d8, d9, d10, d11}
+; CHECK-LIS-NEXT:    bx lr
 entry:
   %l1 = load <48 x i16>, ptr %src, align 4
   %s1 = shufflevector <48 x i16> %l1, <48 x i16> undef, <16 x i32> <i32 0, i32 3, i32 6, i32 9, i32 12, i32 15, i32 18, i32 21, i32 24, i32 27, i32 30, i32 33, i32 36, i32 39, i32 42, i32 45>
@@ -710,35 +935,65 @@ entry:
 ; i64
 
 define void @vld3_v2i64(ptr %src, ptr %dst) {
-; CHECK-LABEL: vld3_v2i64:
-; CHECK:       @ %bb.0: @ %entry
-; CHECK-NEXT:    .save {r4, r5, r6, r7, r8, lr}
-; CHECK-NEXT:    push.w {r4, r5, r6, r7, r8, lr}
-; CHECK-NEXT:    vldrw.u32 q0, [r0]
-; CHECK-NEXT:    vldrw.u32 q1, [r0, #32]
-; CHECK-NEXT:    vldrw.u32 q2, [r0, #16]
-; CHECK-NEXT:    vmov.f32 s12, s2
-; CHECK-NEXT:    vmov.f32 s13, s3
-; CHECK-NEXT:    vmov.f32 s2, s4
-; CHECK-NEXT:    vmov.f32 s3, s5
-; CHECK-NEXT:    vmov r0, r3, d5
-; CHECK-NEXT:    vmov r2, r4, d3
-; CHECK-NEXT:    vmov r6, r7, d0
-; CHECK-NEXT:    vmov r5, r8, d6
-; CHECK-NEXT:    vmov lr, r12, d1
-; CHECK-NEXT:    adds.w r0, r0, lr
-; CHECK-NEXT:    adc.w r3, r3, r12
-; CHECK-NEXT:    adds r0, r0, r2
-; CHECK-NEXT:    adc.w r2, r3, r4
-; CHECK-NEXT:    vmov r3, r4, d4
-; CHECK-NEXT:    adds r6, r6, r5
-; CHECK-NEXT:    adc.w r7, r7, r8
-; CHECK-NEXT:    adds r3, r3, r6
-; CHECK-NEXT:    adcs r7, r4
-; CHECK-NEXT:    vmov q0[2], q0[0], r3, r0
-; CHECK-NEXT:    vmov q0[3], q0[1], r7, r2
-; CHECK-NEXT:    vstrw.32 q0, [r1]
-; CHECK-NEXT:    pop.w {r4, r5, r6, r7, r8, pc}
+; CHECK-LV-LABEL: vld3_v2i64:
+; CHECK-LV:       @ %bb.0: @ %entry
+; CHECK-LV-NEXT:    .save {r4, r5, r6, r7, r8, lr}
+; CHECK-LV-NEXT:    push.w {r4, r5, r6, r7, r8, lr}
+; CHECK-LV-NEXT:    vldrw.u32 q0, [r0]
+; CHECK-LV-NEXT:    vldrw.u32 q1, [r0, #32]
+; CHECK-LV-NEXT:    vldrw.u32 q2, [r0, #16]
+; CHECK-LV-NEXT:    vmov.f32 s12, s2
+; CHECK-LV-NEXT:    vmov.f32 s13, s3
+; CHECK-LV-NEXT:    vmov.f32 s2, s4
+; CHECK-LV-NEXT:    vmov.f32 s3, s5
+; CHECK-LV-NEXT:    vmov r0, r3, d5
+; CHECK-LV-NEXT:    vmov r2, r4, d3
+; CHECK-LV-NEXT:    vmov r6, r7, d0
+; CHECK-LV-NEXT:    vmov r5, r8, d6
+; CHECK-LV-NEXT:    vmov lr, r12, d1
+; CHECK-LV-NEXT:    adds.w r0, r0, lr
+; CHECK-LV-NEXT:    adc.w r3, r3, r12
+; CHECK-LV-NEXT:    adds r0, r0, r2
+; CHECK-LV-NEXT:    adc.w r2, r3, r4
+; CHECK-LV-NEXT:    vmov r3, r4, d4
+; CHECK-LV-NEXT:    adds r6, r6, r5
+; CHECK-LV-NEXT:    adc.w r7, r7, r8
+; CHECK-LV-NEXT:    adds r3, r3, r6
+; CHECK-LV-NEXT:    adcs r7, r4
+; CHECK-LV-NEXT:    vmov q0[2], q0[0], r3, r0
+; CHECK-LV-NEXT:    vmov q0[3], q0[1], r7, r2
+; CHECK-LV-NEXT:    vstrw.32 q0, [r1]
+; CHECK-LV-NEXT:    pop.w {r4, r5, r6, r7, r8, pc}
+;
+; CHECK-LIS-LABEL: vld3_v2i64:
+; CHECK-LIS:       @ %bb.0: @ %entry
+; CHECK-LIS-NEXT:    .save {r4, r5, r6, r7, r8, lr}
+; CHECK-LIS-NEXT:    push.w {r4, r5, r6, r7, r8, lr}
+; CHECK-LIS-NEXT:    vldrw.u32 q0, [r0]
+; CHECK-LIS-NEXT:    vldrw.u32 q2, [r0, #32]
+; CHECK-LIS-NEXT:    vldrw.u32 q1, [r0, #16]
+; CHECK-LIS-NEXT:    vmov.f32 s12, s2
+; CHECK-LIS-NEXT:    vmov.f32 s13, s3
+; CHECK-LIS-NEXT:    vmov.f32 s2, s8
+; CHECK-LIS-NEXT:    vmov.f32 s3, s9
+; CHECK-LIS-NEXT:    vmov r0, r3, d3
+; CHECK-LIS-NEXT:    vmov r2, r4, d5
+; CHECK-LIS-NEXT:    vmov r6, r7, d0
+; CHECK-LIS-NEXT:    vmov r5, r8, d6
+; CHECK-LIS-NEXT:    vmov lr, r12, d1
+; CHECK-LIS-NEXT:    adds.w r0, r0, lr
+; CHECK-LIS-NEXT:    adc.w r3, r3, r12
+; CHECK-LIS-NEXT:    adds r0, r0, r2
+; CHECK-LIS-NEXT:    adc.w r2, r3, r4
+; CHECK-LIS-NEXT:    vmov r3, r4, d2
+; CHECK-LIS-NEXT:    adds r6, r6, r5
+; CHECK-LIS-NEXT:    adc.w r7, r7, r8
+; CHECK-LIS-NEXT:    adds r3, r3, r6
+; CHECK-LIS-NEXT:    adcs r7, r4
+; CHECK-LIS-NEXT:    vmov q0[2], q0[0], r3, r0
+; CHECK-LIS-NEXT:    vmov q0[3], q0[1], r7, r2
+; CHECK-LIS-NEXT:    vstrw.32 q0, [r1]
+; CHECK-LIS-NEXT:    pop.w {r4, r5, r6, r7, r8, pc}
 entry:
   %l1 = load <6 x i64>, ptr %src, align 4
   %s1 = shufflevector <6 x i64> %l1, <6 x i64> undef, <2 x i32> <i32 0, i32 3>
@@ -751,62 +1006,120 @@ entry:
 }
 
 define void @vld3_v4i64(ptr %src, ptr %dst) {
-; CHECK-LABEL: vld3_v4i64:
-; CHECK:       @ %bb.0: @ %entry
-; CHECK-NEXT:    .save {r4, r5, r6, r7, r8, lr}
-; CHECK-NEXT:    push.w {r4, r5, r6, r7, r8, lr}
-; CHECK-NEXT:    .vsave {d8, d9, d10, d11, d12}
-; CHECK-NEXT:    vpush {d8, d9, d10, d11, d12}
-; CHECK-NEXT:    vldrw.u32 q0, [r0]
-; CHECK-NEXT:    vldrw.u32 q3, [r0, #32]
-; CHECK-NEXT:    vldrw.u32 q2, [r0, #16]
-; CHECK-NEXT:    vldrw.u32 q5, [r0, #48]
-; CHECK-NEXT:    vmov.f32 s4, s2
-; CHECK-NEXT:    vldrw.u32 q4, [r0, #64]
-; CHECK-NEXT:    vmov.f32 s5, s3
-; CHECK-NEXT:    vmov.f32 s2, s12
-; CHECK-NEXT:    vmov.f32 s3, s13
-; CHECK-NEXT:    vmov r5, r4, d5
-; CHECK-NEXT:    vmov r3, r8, d7
-; CHECK-NEXT:    vldrw.u32 q3, [r0, #80]
-; CHECK-NEXT:    vmov.f32 s24, s22
-; CHECK-NEXT:    vmov.f32 s25, s23
-; CHECK-NEXT:    vmov lr, r12, d1
-; CHECK-NEXT:    vmov.f32 s2, s12
-; CHECK-NEXT:    vmov.f32 s3, s13
-; CHECK-NEXT:    vmov r6, r7, d12
-; CHECK-NEXT:    adds.w r0, r5, lr
-; CHECK-NEXT:    adc.w r5, r4, r12
-; CHECK-NEXT:    adds.w lr, r0, r3
-; CHECK-NEXT:    vmov r4, r2, d10
-; CHECK-NEXT:    adc.w r12, r5, r8
-; CHECK-NEXT:    vmov r5, r0, d8
-; CHECK-NEXT:    adds r6, r6, r4
-; CHECK-NEXT:    adcs r2, r7
-; CHECK-NEXT:    adds r6, r6, r5
-; CHECK-NEXT:    adc.w r8, r2, r0
-; CHECK-NEXT:    vmov r7, r4, d1
-; CHECK-NEXT:    vmov r2, r5, d9
-; CHECK-NEXT:    vmov r3, r0, d0
-; CHECK-NEXT:    adds r2, r2, r7
-; CHECK-NEXT:    adc.w r7, r5, r4
-; CHECK-NEXT:    vmov r5, r4, d7
-; CHECK-NEXT:    adds r2, r2, r5
-; CHECK-NEXT:    adcs r7, r4
-; CHECK-NEXT:    vmov r5, r4, d2
-; CHECK-NEXT:    vmov q1[2], q1[0], r6, r2
-; CHECK-NEXT:    vmov q1[3], q1[1], r8, r7
-; CHECK-NEXT:    vstrw.32 q1, [r1, #16]
-; CHECK-NEXT:    adds r3, r3, r5
-; CHECK-NEXT:    adcs r0, r4
-; CHECK-NEXT:    vmov r4, r5, d4
-; CHECK-NEXT:    adds r3, r3, r4
-; CHECK-NEXT:    vmov q0[2], q0[0], r3, lr
-; CHECK-NEXT:    adcs r0, r5
-; CHECK-NEXT:    vmov q0[3], q0[1], r0, r12
-; CHECK-NEXT:    vstrw.32 q0, [r1]
-; CHECK-NEXT:    vpop {d8, d9, d10, d11, d12}
-; CHECK-NEXT:    pop.w {r4, r5, r6, r7, r8, pc}
+; CHECK-LV-LABEL: vld3_v4i64:
+; CHECK-LV:       @ %bb.0: @ %entry
+; CHECK-LV-NEXT:    .save {r4, r5, r6, r7, r8, lr}
+; CHECK-LV-NEXT:    push.w {r4, r5, r6, r7, r8, lr}
+; CHECK-LV-NEXT:    .vsave {d8, d9, d10, d11, d12}
+; CHECK-LV-NEXT:    vpush {d8, d9, d10, d11, d12}
+; CHECK-LV-NEXT:    vldrw.u32 q0, [r0]
+; CHECK-LV-NEXT:    vldrw.u32 q3, [r0, #32]
+; CHECK-LV-NEXT:    vldrw.u32 q2, [r0, #16]
+; CHECK-LV-NEXT:    vldrw.u32 q5, [r0, #48]
+; CHECK-LV-NEXT:    vmov.f32 s4, s2
+; CHECK-LV-NEXT:    vldrw.u32 q4, [r0, #64]
+; CHECK-LV-NEXT:    vmov.f32 s5, s3
+; CHECK-LV-NEXT:    vmov.f32 s2, s12
+; CHECK-LV-NEXT:    vmov.f32 s3, s13
+; CHECK-LV-NEXT:    vmov r5, r4, d5
+; CHECK-LV-NEXT:    vmov r3, r8, d7
+; CHECK-LV-NEXT:    vldrw.u32 q3, [r0, #80]
+; CHECK-LV-NEXT:    vmov.f32 s24, s22
+; CHECK-LV-NEXT:    vmov.f32 s25, s23
+; CHECK-LV-NEXT:    vmov lr, r12, d1
+; CHECK-LV-NEXT:    vmov.f32 s2, s12
+; CHECK-LV-NEXT:    vmov.f32 s3, s13
+; CHECK-LV-NEXT:    vmov r6, r7, d12
+; CHECK-LV-NEXT:    adds.w r0, r5, lr
+; CHECK-LV-NEXT:    adc.w r5, r4, r12
+; CHECK-LV-NEXT:    adds.w lr, r0, r3
+; CHECK-LV-NEXT:    vmov r4, r2, d10
+; CHECK-LV-NEXT:    adc.w r12, r5, r8
+; CHECK-LV-NEXT:    vmov r5, r0, d8
+; CHECK-LV-NEXT:    adds r6, r6, r4
+; CHECK-LV-NEXT:    adcs r2, r7
+; CHECK-LV-NEXT:    adds r6, r6, r5
+; CHECK-LV-NEXT:    adc.w r8, r2, r0
+; CHECK-LV-NEXT:    vmov r7, r4, d1
+; CHECK-LV-NEXT:    vmov r2, r5, d9
+; CHECK-LV-NEXT:    vmov r3, r0, d0
+; CHECK-LV-NEXT:    adds r2, r2, r7
+; CHECK-LV-NEXT:    adc.w r7, r5, r4
+; CHECK-LV-NEXT:    vmov r5, r4, d7
+; CHECK-LV-NEXT:    adds r2, r2, r5
+; CHECK-LV-NEXT:    adcs r7, r4
+; CHECK-LV-NEXT:    vmov r5, r4, d2
+; CHECK-LV-NEXT:    vmov q1[2], q1[0], r6, r2
+; CHECK-LV-NEXT:    vmov q1[3], q1[1], r8, r7
+; CHECK-LV-NEXT:    vstrw.32 q1, [r1, #16]
+; CHECK-LV-NEXT:    adds r3, r3, r5
+; CHECK-LV-NEXT:    adcs r0, r4
+; CHECK-LV-NEXT:    vmov r4, r5, d4
+; CHECK-LV-NEXT:    adds r3, r3, r4
+; CHECK-LV-NEXT:    vmov q0[2], q0[0], r3, lr
+; CHECK-LV-NEXT:    adcs r0, r5
+; CHECK-LV-NEXT:    vmov q0[3], q0[1], r0, r12
+; CHECK-LV-NEXT:    vstrw.32 q0, [r1]
+; CHECK-LV-NEXT:    vpop {d8, d9, d10, d11, d12}
+; CHECK-LV-NEXT:    pop.w {r4, r5, r6, r7, r8, pc}
+;
+; CHECK-LIS-LABEL: vld3_v4i64:
+; CHECK-LIS:       @ %bb.0: @ %entry
+; CHECK-LIS-NEXT:    .save {r4, r5, r6, r7, r8, lr}
+; CHECK-LIS-NEXT:    push.w {r4, r5, r6, r7, r8, lr}
+; CHECK-LIS-NEXT:    .vsave {d8, d9, d10, d11, d12}
+; CHECK-LIS-NEXT:    vpush {d8, d9, d10, d11, d12}
+; CHECK-LIS-NEXT:    vldrw.u32 q0, [r0]
+; CHECK-LIS-NEXT:    vldrw.u32 q3, [r0, #32]
+; CHECK-LIS-NEXT:    vldrw.u32 q1, [r0, #16]
+; CHECK-LIS-NEXT:    vldrw.u32 q5, [r0, #48]
+; CHECK-LIS-NEXT:    vmov.f32 s8, s2
+; CHECK-LIS-NEXT:    vldrw.u32 q4, [r0, #64]
+; CHECK-LIS-NEXT:    vmov.f32 s9, s3
+; CHECK-LIS-NEXT:    vmov.f32 s2, s12
+; CHECK-LIS-NEXT:    vmov.f32 s3, s13
+; CHECK-LIS-NEXT:    vmov r2, r3, d3
+; CHECK-LIS-NEXT:    vmov r4, r8, d7
+; CHECK-LIS-NEXT:    vldrw.u32 q3, [r0, #80]
+; CHECK-LIS-NEXT:    vmov.f32 s24, s22
+; CHECK-LIS-NEXT:    vmov.f32 s25, s23
+; CHECK-LIS-NEXT:    vmov.f32 s7, s19
+; CHECK-LIS-NEXT:    vmov lr, r12, d1
+; CHECK-LIS-NEXT:    vmov.f32 s2, s12
+; CHECK-LIS-NEXT:    vmov.f32 s3, s13
+; CHECK-LIS-NEXT:    vmov r6, r7, d12
+; CHECK-LIS-NEXT:    adds.w r0, r2, lr
+; CHECK-LIS-NEXT:    adc.w r2, r3, r12
+; CHECK-LIS-NEXT:    adds.w lr, r0, r4
+; CHECK-LIS-NEXT:    vmov r3, r5, d10
+; CHECK-LIS-NEXT:    adc.w r12, r2, r8
+; CHECK-LIS-NEXT:    vmov r2, r0, d8
+; CHECK-LIS-NEXT:    adds r3, r3, r6
+; CHECK-LIS-NEXT:    adcs r7, r5
+; CHECK-LIS-NEXT:    adds r2, r2, r3
+; CHECK-LIS-NEXT:    adc.w r8, r7, r0
+; CHECK-LIS-NEXT:    vmov r6, r5, d1
+; CHECK-LIS-NEXT:    vmov r3, r7, d3
+; CHECK-LIS-NEXT:    vmov r4, r0, d0
+; CHECK-LIS-NEXT:    adds r3, r3, r6
+; CHECK-LIS-NEXT:    adcs r7, r5
+; CHECK-LIS-NEXT:    vmov r6, r5, d7
+; CHECK-LIS-NEXT:    adds r3, r3, r6
+; CHECK-LIS-NEXT:    adcs r7, r5
+; CHECK-LIS-NEXT:    vmov r6, r5, d4
+; CHECK-LIS-NEXT:    adds r4, r4, r6
+; CHECK-LIS-NEXT:    adcs r0, r5
+; CHECK-LIS-NEXT:    vmov r5, r6, d2
+; CHECK-LIS-NEXT:    vmov q1[2], q1[0], r2, r3
+; CHECK-LIS-NEXT:    vmov q1[3], q1[1], r8, r7
+; CHECK-LIS-NEXT:    vstrw.32 q1, [r1, #16]
+; CHECK-LIS-NEXT:    adds r4, r4, r5
+; CHECK-LIS-NEXT:    vmov q0[2], q0[0], r4, lr
+; CHECK-LIS-NEXT:    adcs r0, r6
+; CHECK-LIS-NEXT:    vmov q0[3], q0[1], r0, r12
+; CHECK-LIS-NEXT:    vstrw.32 q0, [r1]
+; CHECK-LIS-NEXT:    vpop {d8, d9, d10, d11, d12}
+; CHECK-LIS-NEXT:    pop.w {r4, r5, r6, r7, r8, pc}
 entry:
   %l1 = load <12 x i64>, ptr %src, align 4
   %s1 = shufflevector <12 x i64> %l1, <12 x i64> undef, <4 x i32> <i32 0, i32 3, i32 6, i32 9>
@@ -881,46 +1194,87 @@ entry:
 }
 
 define void @vld3_v8f32(ptr %src, ptr %dst) {
-; CHECK-LABEL: vld3_v8f32:
-; CHECK:       @ %bb.0: @ %entry
-; CHECK-NEXT:    .vsave {d8, d9, d10, d11}
-; CHECK-NEXT:    vpush {d8, d9, d10, d11}
-; CHECK-NEXT:    vldrw.u32 q0, [r0, #64]
-; CHECK-NEXT:    vldrw.u32 q1, [r0, #48]
-; CHECK-NEXT:    vldrw.u32 q4, [r0, #80]
-; CHECK-NEXT:    vmov.f32 s10, s2
-; CHECK-NEXT:    vmov.f32 s13, s0
-; CHECK-NEXT:    vmov.f32 s14, s3
-; CHECK-NEXT:    vmov.f32 s8, s4
-; CHECK-NEXT:    vmov.f32 s9, s7
-; CHECK-NEXT:    vmov.f32 s12, s5
-; CHECK-NEXT:    vmov.f32 s15, s18
-; CHECK-NEXT:    vmov.f32 s11, s17
-; CHECK-NEXT:    vadd.f32 q2, q2, q3
-; CHECK-NEXT:    vmov.f32 s0, s6
-; CHECK-NEXT:    vmov.f32 s2, s16
-; CHECK-NEXT:    vldrw.u32 q1, [r0, #16]
-; CHECK-NEXT:    vmov.f32 s3, s19
-; CHECK-NEXT:    vldrw.u32 q3, [r0, #32]
-; CHECK-NEXT:    vadd.f32 q0, q2, q0
-; CHECK-NEXT:    vldrw.u32 q2, [r0]
-; CHECK-NEXT:    vmov.f32 s17, s4
-; CHECK-NEXT:    vstrw.32 q0, [r1, #16]
-; CHECK-NEXT:    vmov.f32 s18, s7
-; CHECK-NEXT:    vmov.f32 s22, s6
-; CHECK-NEXT:    vmov.f32 s16, s9
-; CHECK-NEXT:    vmov.f32 s19, s14
-; CHECK-NEXT:    vmov.f32 s20, s8
-; CHECK-NEXT:    vmov.f32 s21, s11
-; CHECK-NEXT:    vmov.f32 s23, s13
-; CHECK-NEXT:    vadd.f32 q4, q5, q4
-; CHECK-NEXT:    vmov.f32 s4, s10
-; CHECK-NEXT:    vmov.f32 s6, s12
-; CHECK-NEXT:    vmov.f32 s7, s15
-; CHECK-NEXT:    vadd.f32 q1, q4, q1
-; CHECK-NEXT:    vstrw.32 q1, [r1]
-; CHECK-NEXT:    vpop {d8, d9, d10, d11}
-; CHECK-NEXT:    bx lr
+; CHECK-LV-LABEL: vld3_v8f32:
+; CHECK-LV:       @ %bb.0: @ %entry
+; CHECK-LV-NEXT:    .vsave {d8, d9, d10, d11}
+; CHECK-LV-NEXT:    vpush {d8, d9, d10, d11}
+; CHECK-LV-NEXT:    vldrw.u32 q0, [r0, #64]
+; CHECK-LV-NEXT:    vldrw.u32 q1, [r0, #48]
+; CHECK-LV-NEXT:    vldrw.u32 q4, [r0, #80]
+; CHECK-LV-NEXT:    vmov.f32 s10, s2
+; CHECK-LV-NEXT:    vmov.f32 s13, s0
+; CHECK-LV-NEXT:    vmov.f32 s14, s3
+; CHECK-LV-NEXT:    vmov.f32 s8, s4
+; CHECK-LV-NEXT:    vmov.f32 s9, s7
+; CHECK-LV-NEXT:    vmov.f32 s12, s5
+; CHECK-LV-NEXT:    vmov.f32 s15, s18
+; CHECK-LV-NEXT:    vmov.f32 s11, s17
+; CHECK-LV-NEXT:    vadd.f32 q2, q2, q3
+; CHECK-LV-NEXT:    vmov.f32 s0, s6
+; CHECK-LV-NEXT:    vmov.f32 s2, s16
+; CHECK-LV-NEXT:    vldrw.u32 q1, [r0, #16]
+; CHECK-LV-NEXT:    vmov.f32 s3, s19
+; CHECK-LV-NEXT:    vldrw.u32 q3, [r0, #32]
+; CHECK-LV-NEXT:    vadd.f32 q0, q2, q0
+; CHECK-LV-NEXT:    vldrw.u32 q2, [r0]
+; CHECK-LV-NEXT:    vmov.f32 s17, s4
+; CHECK-LV-NEXT:    vstrw.32 q0, [r1, #16]
+; CHECK-LV-NEXT:    vmov.f32 s18, s7
+; CHECK-LV-NEXT:    vmov.f32 s22, s6
+; CHECK-LV-NEXT:    vmov.f32 s16, s9
+; CHECK-LV-NEXT:    vmov.f32 s19, s14
+; CHECK-LV-NEXT:    vmov.f32 s20, s8
+; CHECK-LV-NEXT:    vmov.f32 s21, s11
+; CHECK-LV-NEXT:    vmov.f32 s23, s13
+; CHECK-LV-NEXT:    vadd.f32 q4, q5, q4
+; CHECK-LV-NEXT:    vmov.f32 s4, s10
+; CHECK-LV-NEXT:    vmov.f32 s6, s12
+; CHECK-LV-NEXT:    vmov.f32 s7, s15
+; CHECK-LV-NEXT:    vadd.f32 q1, q4, q1
+; CHECK-LV-NEXT:    vstrw.32 q1, [r1]
+; CHECK-LV-NEXT:    vpop {d8, d9, d10, d11}
+; CHECK-LV-NEXT:    bx lr
+;
+; CHECK-LIS-LABEL: vld3_v8f32:
+; CHECK-LIS:       @ %bb.0: @ %entry
+; CHECK-LIS-NEXT:    .vsave {d8, d9, d10, d11}
+; CHECK-LIS-NEXT:    vpush {d8, d9, d10, d11}
+; CHECK-LIS-NEXT:    vldrw.u32 q0, [r0, #64]
+; CHECK-LIS-NEXT:    vldrw.u32 q1, [r0, #48]
+; CHECK-LIS-NEXT:    vldrw.u32 q4, [r0, #80]
+; CHECK-LIS-NEXT:    vmov.f32 s10, s2
+; CHECK-LIS-NEXT:    vmov.f32 s13, s0
+; CHECK-LIS-NEXT:    vmov.f32 s14, s3
+; CHECK-LIS-NEXT:    vmov.f32 s8, s4
+; CHECK-LIS-NEXT:    vmov.f32 s9, s7
+; CHECK-LIS-NEXT:    vmov.f32 s12, s5
+; CHECK-LIS-NEXT:    vmov.f32 s15, s18
+; CHECK-LIS-NEXT:    vmov.f32 s11, s17
+; CHECK-LIS-NEXT:    vmov.f32 s0, s6
+; CHECK-LIS-NEXT:    vadd.f32 q2, q2, q3
+; CHECK-LIS-NEXT:    vmov.f32 s2, s16
+; CHECK-LIS-NEXT:    vldrw.u32 q1, [r0, #16]
+; CHECK-LIS-NEXT:    vmov.f32 s3, s19
+; CHECK-LIS-NEXT:    vldrw.u32 q4, [r0, #32]
+; CHECK-LIS-NEXT:    vadd.f32 q0, q2, q0
+; CHECK-LIS-NEXT:    vldrw.u32 q2, [r0]
+; CHECK-LIS-NEXT:    vmov.f32 s13, s4
+; CHECK-LIS-NEXT:    vstrw.32 q0, [r1, #16]
+; CHECK-LIS-NEXT:    vmov.f32 s14, s7
+; CHECK-LIS-NEXT:    vmov.f32 s22, s6
+; CHECK-LIS-NEXT:    vmov.f32 s12, s9
+; CHECK-LIS-NEXT:    vmov.f32 s15, s18
+; CHECK-LIS-NEXT:    vmov.f32 s20, s8
+; CHECK-LIS-NEXT:    vmov.f32 s21, s11
+; CHECK-LIS-NEXT:    vmov.f32 s23, s17
+; CHECK-LIS-NEXT:    vadd.f32 q3, q5, q3
+; CHECK-LIS-NEXT:    vmov.f32 s4, s10
+; CHECK-LIS-NEXT:    vmov.f32 s6, s16
+; CHECK-LIS-NEXT:    vmov.f32 s7, s19
+; CHECK-LIS-NEXT:    vadd.f32 q1, q3, q1
+; CHECK-LIS-NEXT:    vstrw.32 q1, [r1]
+; CHECK-LIS-NEXT:    vpop {d8, d9, d10, d11}
+; CHECK-LIS-NEXT:    bx lr
 entry:
   %l1 = load <24 x float>, ptr %src, align 4
   %s1 = shufflevector <24 x float> %l1, <24 x float> undef, <8 x i32> <i32 0, i32 3, i32 6, i32 9, i32 12, i32 15, i32 18, i32 21>
@@ -933,80 +1287,155 @@ entry:
 }
 
 define void @vld3_v16f32(ptr %src, ptr %dst) {
-; CHECK-LABEL: vld3_v16f32:
-; CHECK:       @ %bb.0: @ %entry
-; CHECK-NEXT:    .vsave {d8, d9, d10, d11, d12, d13, d14, d15}
-; CHECK-NEXT:    vpush {d8, d9, d10, d11, d12, d13, d14, d15}
-; CHECK-NEXT:    vldrw.u32 q0, [r0, #64]
-; CHECK-NEXT:    vldrw.u32 q1, [r0, #48]
-; CHECK-NEXT:    vldrw.u32 q4, [r0, #80]
-; CHECK-NEXT:    vldrw.u32 q6, [r0, #176]
-; CHECK-NEXT:    vmov.f32 s10, s2
-; CHECK-NEXT:    vmov.f32 s13, s0
-; CHECK-NEXT:    vmov.f32 s14, s3
-; CHECK-NEXT:    vmov.f32 s8, s4
-; CHECK-NEXT:    vmov.f32 s9, s7
-; CHECK-NEXT:    vmov.f32 s12, s5
-; CHECK-NEXT:    vmov.f32 s15, s18
-; CHECK-NEXT:    vmov.f32 s11, s17
-; CHECK-NEXT:    vadd.f32 q2, q2, q3
-; CHECK-NEXT:    vmov.f32 s0, s6
-; CHECK-NEXT:    vmov.f32 s2, s16
-; CHECK-NEXT:    vldrw.u32 q1, [r0, #16]
-; CHECK-NEXT:    vmov.f32 s3, s19
-; CHECK-NEXT:    vldrw.u32 q3, [r0, #32]
-; CHECK-NEXT:    vadd.f32 q0, q2, q0
-; CHECK-NEXT:    vldrw.u32 q2, [r0]
-; CHECK-NEXT:    vmov.f32 s17, s4
-; CHECK-NEXT:    vmov.f32 s18, s7
-; CHECK-NEXT:    vmov.f32 s22, s6
-; CHECK-NEXT:    vmov.f32 s16, s9
-; CHECK-NEXT:    vmov.f32 s19, s14
-; CHECK-NEXT:    vmov.f32 s20, s8
-; CHECK-NEXT:    vmov.f32 s21, s11
-; CHECK-NEXT:    vmov.f32 s23, s13
-; CHECK-NEXT:    vmov.f32 s4, s10
-; CHECK-NEXT:    vldrw.u32 q2, [r0, #160]
-; CHECK-NEXT:    vmov.f32 s6, s12
-; CHECK-NEXT:    vadd.f32 q4, q5, q4
-; CHECK-NEXT:    vmov.f32 s7, s15
-; CHECK-NEXT:    vldrw.u32 q3, [r0, #144]
-; CHECK-NEXT:    vadd.f32 q1, q4, q1
-; CHECK-NEXT:    vmov.f32 s18, s10
-; CHECK-NEXT:    vmov.f32 s21, s8
-; CHECK-NEXT:    vmov.f32 s22, s11
-; CHECK-NEXT:    vmov.f32 s16, s12
-; CHECK-NEXT:    vmov.f32 s17, s15
-; CHECK-NEXT:    vmov.f32 s20, s13
-; CHECK-NEXT:    vmov.f32 s23, s26
-; CHECK-NEXT:    vmov.f32 s19, s25
-; CHECK-NEXT:    vadd.f32 q4, q4, q5
-; CHECK-NEXT:    vmov.f32 s8, s14
-; CHECK-NEXT:    vmov.f32 s10, s24
-; CHECK-NEXT:    vldrw.u32 q3, [r0, #112]
-; CHECK-NEXT:    vmov.f32 s11, s27
-; CHECK-NEXT:    vldrw.u32 q5, [r0, #128]
-; CHECK-NEXT:    vadd.f32 q2, q4, q2
-; CHECK-NEXT:    vldrw.u32 q4, [r0, #96]
-; CHECK-NEXT:    vmov.f32 s25, s12
-; CHECK-NEXT:    vstrw.32 q2, [r1, #48]
-; CHECK-NEXT:    vmov.f32 s26, s15
-; CHECK-NEXT:    vstrw.32 q0, [r1, #16]
-; CHECK-NEXT:    vmov.f32 s30, s14
-; CHECK-NEXT:    vstrw.32 q1, [r1]
-; CHECK-NEXT:    vmov.f32 s24, s17
-; CHECK-NEXT:    vmov.f32 s27, s22
-; CHECK-NEXT:    vmov.f32 s28, s16
-; CHECK-NEXT:    vmov.f32 s29, s19
-; CHECK-NEXT:    vmov.f32 s31, s21
-; CHECK-NEXT:    vadd.f32 q6, q7, q6
-; CHECK-NEXT:    vmov.f32 s12, s18
-; CHECK-NEXT:    vmov.f32 s14, s20
-; CHECK-NEXT:    vmov.f32 s15, s23
-; CHECK-NEXT:    vadd.f32 q3, q6, q3
-; CHECK-NEXT:    vstrw.32 q3, [r1, #32]
-; CHECK-NEXT:    vpop {d8, d9, d10, d11, d12, d13, d14, d15}
-; CHECK-NEXT:    bx lr
+; CHECK-LV-LABEL: vld3_v16f32:
+; CHECK-LV:       @ %bb.0: @ %entry
+; CHECK-LV-NEXT:    .vsave {d8, d9, d10, d11, d12, d13, d14, d15}
+; CHECK-LV-NEXT:    vpush {d8, d9, d10, d11, d12, d13, d14, d15}
+; CHECK-LV-NEXT:    vldrw.u32 q0, [r0, #64]
+; CHECK-LV-NEXT:    vldrw.u32 q1, [r0, #48]
+; CHECK-LV-NEXT:    vldrw.u32 q4, [r0, #80]
+; CHECK-LV-NEXT:    vldrw.u32 q6, [r0, #176]
+; CHECK-LV-NEXT:    vmov.f32 s10, s2
+; CHECK-LV-NEXT:    vmov.f32 s13, s0
+; CHECK-LV-NEXT:    vmov.f32 s14, s3
+; CHECK-LV-NEXT:    vmov.f32 s8, s4
+; CHECK-LV-NEXT:    vmov.f32 s9, s7
+; CHECK-LV-NEXT:    vmov.f32 s12, s5
+; CHECK-LV-NEXT:    vmov.f32 s15, s18
+; CHECK-LV-NEXT:    vmov.f32 s11, s17
+; CHECK-LV-NEXT:    vadd.f32 q2, q2, q3
+; CHECK-LV-NEXT:    vmov.f32 s0, s6
+; CHECK-LV-NEXT:    vmov.f32 s2, s16
+; CHECK-LV-NEXT:    vldrw.u32 q1, [r0, #16]
+; CHECK-LV-NEXT:    vmov.f32 s3, s19
+; CHECK-LV-NEXT:    vldrw.u32 q3, [r0, #32]
+; CHECK-LV-NEXT:    vadd.f32 q0, q2, q0
+; CHECK-LV-NEXT:    vldrw.u32 q2, [r0]
+; CHECK-LV-NEXT:    vmov.f32 s17, s4
+; CHECK-LV-NEXT:    vmov.f32 s18, s7
+; CHECK-LV-NEXT:    vmov.f32 s22, s6
+; CHECK-LV-NEXT:    vmov.f32 s16, s9
+; CHECK-LV-NEXT:    vmov.f32 s19, s14
+; CHECK-LV-NEXT:    vmov.f32 s20, s8
+; CHECK-LV-NEXT:    vmov.f32 s21, s11
+; CHECK-LV-NEXT:    vmov.f32 s23, s13
+; CHECK-LV-NEXT:    vmov.f32 s4, s10
+; CHECK-LV-NEXT:    vldrw.u32 q2, [r0, #160]
+; CHECK-LV-NEXT:    vmov.f32 s6, s12
+; CHECK-LV-NEXT:    vadd.f32 q4, q5, q4
+; CHECK-LV-NEXT:    vmov.f32 s7, s15
+; CHECK-LV-NEXT:    vldrw.u32 q3, [r0, #144]
+; CHECK-LV-NEXT:    vadd.f32 q1, q4, q1
+; CHECK-LV-NEXT:    vmov.f32 s18, s10
+; CHECK-LV-NEXT:    vmov.f32 s21, s8
+; CHECK-LV-NEXT:    vmov.f32 s22, s11
+; CHECK-LV-NEXT:    vmov.f32 s16, s12
+; CHECK-LV-NEXT:    vmov.f32 s17, s15
+; CHECK-LV-NEXT:    vmov.f32 s20, s13
+; CHECK-LV-NEXT:    vmov.f32 s23, s26
+; CHECK-LV-NEXT:    vmov.f32 s19, s25
+; CHECK-LV-NEXT:    vadd.f32 q4, q4, q5
+; CHECK-LV-NEXT:    vmov.f32 s8, s14
+; CHECK-LV-NEXT:    vmov.f32 s10, s24
+; CHECK-LV-NEXT:    vldrw.u32 q3, [r0, #112]
+; CHECK-LV-NEXT:    vmov.f32 s11, s27
+; CHECK-LV-NEXT:    vldrw.u32 q5, [r0, #128]
+; CHECK-LV-NEXT:    vadd.f32 q2, q4, q2
+; CHECK-LV-NEXT:    vldrw.u32 q4, [r0, #96]
+; CHECK-LV-NEXT:    vmov.f32 s25, s12
+; CHECK-LV-NEXT:    vstrw.32 q2, [r1, #48]
+; CHECK-LV-NEXT:    vmov.f32 s26, s15
+; CHECK-LV-NEXT:    vstrw.32 q0, [r1, #16]
+; CHECK-LV-NEXT:    vmov.f32 s30, s14
+; CHECK-LV-NEXT:    vstrw.32 q1, [r1]
+; CHECK-LV-NEXT:    vmov.f32 s24, s17
+; CHECK-LV-NEXT:    vmov.f32 s27, s22
+; CHECK-LV-NEXT:    vmov.f32 s28, s16
+; CHECK-LV-NEXT:    vmov.f32 s29, s19
+; CHECK-LV-NEXT:    vmov.f32 s31, s21
+; CHECK-LV-NEXT:    vadd.f32 q6, q7, q6
+; CHECK-LV-NEXT:    vmov.f32 s12, s18
+; CHECK-LV-NEXT:    vmov.f32 s14, s20
+; CHECK-LV-NEXT:    vmov.f32 s15, s23
+; CHECK-LV-NEXT:    vadd.f32 q3, q6, q3
+; CHECK-LV-NEXT:    vstrw.32 q3, [r1, #32]
+; CHECK-LV-NEXT:    vpop {d8, d9, d10, d11, d12, d13, d14, d15}
+; CHECK-LV-NEXT:    bx lr
+;
+; CHECK-LIS-LABEL: vld3_v16f32:
+; CHECK-LIS:       @ %bb.0: @ %entry
+; CHECK-LIS-NEXT:    .vsave {d8, d9, d10, d11, d12, d13, d14, d15}
+; CHECK-LIS-NEXT:    vpush {d8, d9, d10, d11, d12, d13, d14, d15}
+; CHECK-LIS-NEXT:    vldrw.u32 q0, [r0, #64]
+; CHECK-LIS-NEXT:    vldrw.u32 q1, [r0, #48]
+; CHECK-LIS-NEXT:    vldrw.u32 q4, [r0, #80]
+; CHECK-LIS-NEXT:    vldrw.u32 q6, [r0, #176]
+; CHECK-LIS-NEXT:    vmov.f32 s10, s2
+; CHECK-LIS-NEXT:    vmov.f32 s13, s0
+; CHECK-LIS-NEXT:    vmov.f32 s14, s3
+; CHECK-LIS-NEXT:    vmov.f32 s8, s4
+; CHECK-LIS-NEXT:    vmov.f32 s9, s7
+; CHECK-LIS-NEXT:    vmov.f32 s12, s5
+; CHECK-LIS-NEXT:    vmov.f32 s15, s18
+; CHECK-LIS-NEXT:    vmov.f32 s11, s17
+; CHECK-LIS-NEXT:    vmov.f32 s0, s6
+; CHECK-LIS-NEXT:    vadd.f32 q2, q2, q3
+; CHECK-LIS-NEXT:    vmov.f32 s2, s16
+; CHECK-LIS-NEXT:    vldrw.u32 q1, [r0, #16]
+; CHECK-LIS-NEXT:    vmov.f32 s3, s19
+; CHECK-LIS-NEXT:    vldrw.u32 q4, [r0, #32]
+; CHECK-LIS-NEXT:    vadd.f32 q0, q2, q0
+; CHECK-LIS-NEXT:    vldrw.u32 q2, [r0]
+; CHECK-LIS-NEXT:    vmov.f32 s13, s4
+; CHECK-LIS-NEXT:    vmov.f32 s14, s7
+; CHECK-LIS-NEXT:    vmov.f32 s22, s6
+; CHECK-LIS-NEXT:    vmov.f32 s12, s9
+; CHECK-LIS-NEXT:    vmov.f32 s15, s18
+; CHECK-LIS-NEXT:    vmov.f32 s20, s8
+; CHECK-LIS-NEXT:    vmov.f32 s21, s11
+; CHECK-LIS-NEXT:    vmov.f32 s23, s17
+; CHECK-LIS-NEXT:    vadd.f32 q3, q5, q3
+; CHECK-LIS-NEXT:    vmov.f32 s4, s10
+; CHECK-LIS-NEXT:    vmov.f32 s7, s19
+; CHECK-LIS-NEXT:    vldrw.u32 q2, [r0, #160]
+; CHECK-LIS-NEXT:    vmov.f32 s6, s16
+; CHECK-LIS-NEXT:    vadd.f32 q1, q3, q1
+; CHECK-LIS-NEXT:    vldrw.u32 q3, [r0, #144]
+; CHECK-LIS-NEXT:    vmov.f32 s18, s10
+; CHECK-LIS-NEXT:    vmov.f32 s21, s8
+; CHECK-LIS-NEXT:    vmov.f32 s22, s11
+; CHECK-LIS-NEXT:    vmov.f32 s16, s12
+; CHECK-LIS-NEXT:    vmov.f32 s17, s15
+; CHECK-LIS-NEXT:    vmov.f32 s20, s13
+; CHECK-LIS-NEXT:    vmov.f32 s23, s26
+; CHECK-LIS-NEXT:    vmov.f32 s19, s25
+; CHECK-LIS-NEXT:    vmov.f32 s8, s14
+; CHECK-LIS-NEXT:    vadd.f32 q4, q4, q5
+; CHECK-LIS-NEXT:    vmov.f32 s10, s24
+; CHECK-LIS-NEXT:    vldrw.u32 q3, [r0, #112]
+; CHECK-LIS-NEXT:    vmov.f32 s11, s27
+; CHECK-LIS-NEXT:    vldrw.u32 q6, [r0, #128]
+; CHECK-LIS-NEXT:    vadd.f32 q2, q4, q2
+; CHECK-LIS-NEXT:    vldrw.u32 q4, [r0, #96]
+; CHECK-LIS-NEXT:    vmov.f32 s21, s12
+; CHECK-LIS-NEXT:    vstrw.32 q2, [r1, #48]
+; CHECK-LIS-NEXT:    vmov.f32 s22, s15
+; CHECK-LIS-NEXT:    vstrw.32 q0, [r1, #16]
+; CHECK-LIS-NEXT:    vmov.f32 s30, s14
+; CHECK-LIS-NEXT:    vstrw.32 q1, [r1]
+; CHECK-LIS-NEXT:    vmov.f32 s20, s17
+; CHECK-LIS-NEXT:    vmov.f32 s23, s26
+; CHECK-LIS-NEXT:    vmov.f32 s28, s16
+; CHECK-LIS-NEXT:    vmov.f32 s29, s19
+; CHECK-LIS-NEXT:    vmov.f32 s31, s25
+; CHECK-LIS-NEXT:    vadd.f32 q5, q7, q5
+; CHECK-LIS-NEXT:    vmov.f32 s12, s18
+; CHECK-LIS-NEXT:    vmov.f32 s14, s24
+; CHECK-LIS-NEXT:    vmov.f32 s15, s27
+; CHECK-LIS-NEXT:    vadd.f32 q3, q5, q3
+; CHECK-LIS-NEXT:    vstrw.32 q3, [r1, #32]
+; CHECK-LIS-NEXT:    vpop {d8, d9, d10, d11, d12, d13, d14, d15}
+; CHECK-LIS-NEXT:    bx lr
 entry:
   %l1 = load <48 x float>, ptr %src, align 4
   %s1 = shufflevector <48 x float> %l1, <48 x float> undef, <16 x i32> <i32 0, i32 3, i32 6, i32 9, i32 12, i32 15, i32 18, i32 21, i32 24, i32 27, i32 30, i32 33, i32 36, i32 39, i32 42, i32 45>