Index: lib/Target/WebAssembly/WebAssemblyExplicitLocals.cpp
===================================================================
--- lib/Target/WebAssembly/WebAssemblyExplicitLocals.cpp
+++ lib/Target/WebAssembly/WebAssemblyExplicitLocals.cpp
@@ -35,7 +35,7 @@
 // for the purpose of testing with lit/llc ONLY.
 // This produces output which is not valid WebAssembly, and is not supported
 // by assemblers/disassemblers and other MC based tools.
-static cl::opt<bool> WasmDisableExplicitLocals(
+cl::opt<bool> WasmDisableExplicitLocals(
     "wasm-disable-explicit-locals", cl::Hidden,
     cl::desc("WebAssembly: output implicit locals in"
              " instruction output for test purposes only."),
Index: lib/Target/WebAssembly/WebAssemblyInstrSIMD.td
===================================================================
--- lib/Target/WebAssembly/WebAssemblyInstrSIMD.td
+++ lib/Target/WebAssembly/WebAssemblyInstrSIMD.td
@@ -271,6 +271,17 @@
 def : Pat<(i32 (vector_extract (v8i16 V128:$vec), (i32 LaneIdx8:$idx))),
           (EXTRACT_LANE_v8i16_u V128:$vec, (i32 LaneIdx8:$idx))>;
 
+// Bitcasts are nops
+// Matching bitcast t1 to t1 causes strange errors, so avoid repeating types
+foreach t1 = [v16i8, v8i16, v4i32, v2i64, v4f32, v2f64] in
+foreach t2 = !foldl(
+  []<ValueType>, [v16i8, v8i16, v4i32, v2i64, v4f32, v2f64],
+  acc, cur, !if(!eq(!cast<string>(t1), !cast<string>(cur)),
+    acc, !listconcat(acc, [cur])
+  )
+) in
+def : Pat<(t1 (bitconvert (t2 V128:$v))), (t1 V128:$v)>;
+
 // arbitrary other BUILD_VECTOR patterns
 def : Pat<(v16i8 (build_vector
             (i32 I32:$x0), (i32 I32:$x1), (i32 I32:$x2), (i32 I32:$x3),
Index: lib/Target/WebAssembly/WebAssemblyRegStackify.cpp
===================================================================
--- lib/Target/WebAssembly/WebAssemblyRegStackify.cpp
+++ lib/Target/WebAssembly/WebAssemblyRegStackify.cpp
@@ -562,6 +562,12 @@
     MachineRegisterInfo &MRI, const WebAssemblyInstrInfo *TII) {
   LLVM_DEBUG(dbgs() << "Move and tee for multi-use:"; Def->dump());
 
+  // If explicit locals are disabled, the tee this function would produce will
+  // never be lowered to tee_local, so do not produce a tee.
+  extern cl::opt<bool> WasmDisableExplicitLocals;
+  if (WasmDisableExplicitLocals)
+    return Def;
+
   // Move Def into place.
   MBB.splice(Insert, &MBB, Def);
   LIS.handleMove(*Def);
Index: lib/Target/WebAssembly/WebAssemblyUtilities.cpp
===================================================================
--- lib/Target/WebAssembly/WebAssemblyUtilities.cpp
+++ lib/Target/WebAssembly/WebAssemblyUtilities.cpp
@@ -79,6 +79,8 @@
   case WebAssembly::TEE_F32_S:
   case WebAssembly::TEE_F64:
   case WebAssembly::TEE_F64_S:
+  case WebAssembly::TEE_V128:
+  case WebAssembly::TEE_V128_S:
     return true;
   default:
     return false;
Index: test/CodeGen/WebAssembly/simd-conversions.ll
===================================================================
--- /dev/null
+++ test/CodeGen/WebAssembly/simd-conversions.ll
@@ -0,0 +1,224 @@
+; RUN: llc < %s -disable-wasm-fallthrough-return-opt -wasm-disable-explicit-locals -wasm-enable-unimplemented-simd -mattr=+simd128,+sign-ext --show-mc-encoding | FileCheck %s --check-prefixes CHECK,SIMD128
+; RUN: llc < %s -disable-wasm-fallthrough-return-opt -wasm-disable-explicit-locals -mattr=+simd128,+sign-ext --show-mc-encoding | FileCheck %s --check-prefixes CHECK,SIMD128-VM
+; RUN: llc < %s -disable-wasm-fallthrough-return-opt -wasm-disable-explicit-locals -mattr=-simd128,+sign-ext --show-mc-encoding | FileCheck %s --check-prefixes CHECK,NO-SIMD128
+
+; Test that bitcasts between vector types are lowered to zero instructions
+
+target datalayout = "e-m:e-p:32:32-i64:64-n32:64-S128"
+target triple = "wasm32-unknown-unknown"
+
+; CHECK-LABEL: v16i8_to_v16i8:
+define <16 x i8> @v16i8_to_v16i8(<16 x i8> %v) {
+  %res = bitcast <16 x i8> %v to <16 x i8>
+  ret <16 x i8> %res
+}
+
+; CHECK-LABEL: v16i8_to_v8i16:
+define <8 x i16> @v16i8_to_v8i16(<16 x i8> %v) {
+  %res = bitcast <16 x i8> %v to <8 x i16>
+  ret <8 x i16> %res
+}
+
+; CHECK-LABEL: v16i8_to_v4i32:
+define <4 x i32> @v16i8_to_v4i32(<16 x i8> %v) {
+  %res = bitcast <16 x i8> %v to <4 x i32>
+  ret <4 x i32> %res
+}
+
+; CHECK-LABEL: v16i8_to_v2i64:
+define <2 x i64> @v16i8_to_v2i64(<16 x i8> %v) {
+  %res = bitcast <16 x i8> %v to <2 x i64>
+  ret <2 x i64> %res
+}
+
+; CHECK-LABEL: v16i8_to_v4f32:
+define <4 x float> @v16i8_to_v4f32(<16 x i8> %v) {
+  %res = bitcast <16 x i8> %v to <4 x float>
+  ret <4 x float> %res
+}
+
+; CHECK-LABEL: v16i8_to_v2f64:
+define <2 x double> @v16i8_to_v2f64(<16 x i8> %v) {
+  %res = bitcast <16 x i8> %v to <2 x double>
+  ret <2 x double> %res
+}
+
+; CHECK-LABEL: v8i16_to_v16i8:
+define <16 x i8> @v8i16_to_v16i8(<8 x i16> %v) {
+  %res = bitcast <8 x i16> %v to <16 x i8>
+  ret <16 x i8> %res
+}
+
+; CHECK-LABEL: v8i16_to_v8i16:
+define <8 x i16> @v8i16_to_v8i16(<8 x i16> %v) {
+  %res = bitcast <8 x i16> %v to <8 x i16>
+  ret <8 x i16> %res
+}
+
+; CHECK-LABEL: v8i16_to_v4i32:
+define <4 x i32> @v8i16_to_v4i32(<8 x i16> %v) {
+  %res = bitcast <8 x i16> %v to <4 x i32>
+  ret <4 x i32> %res
+}
+
+; CHECK-LABEL: v8i16_to_v2i64:
+define <2 x i64> @v8i16_to_v2i64(<8 x i16> %v) {
+  %res = bitcast <8 x i16> %v to <2 x i64>
+  ret <2 x i64> %res
+}
+
+; CHECK-LABEL: v8i16_to_v4f32:
+define <4 x float> @v8i16_to_v4f32(<8 x i16> %v) {
+  %res = bitcast <8 x i16> %v to <4 x float>
+  ret <4 x float> %res
+}
+
+; CHECK-LABEL: v8i16_to_v2f64:
+define <2 x double> @v8i16_to_v2f64(<8 x i16> %v) {
+  %res = bitcast <8 x i16> %v to <2 x double>
+  ret <2 x double> %res
+}
+
+; CHECK-LABEL: v4i32_to_v16i8:
+define <16 x i8> @v4i32_to_v16i8(<4 x i32> %v) {
+  %res = bitcast <4 x i32> %v to <16 x i8>
+  ret <16 x i8> %res
+}
+
+; CHECK-LABEL: v4i32_to_v8i16:
+define <8 x i16> @v4i32_to_v8i16(<4 x i32> %v) {
+  %res = bitcast <4 x i32> %v to <8 x i16>
+  ret <8 x i16> %res
+}
+
+; CHECK-LABEL: v4i32_to_v4i32:
+define <4 x i32> @v4i32_to_v4i32(<4 x i32> %v) {
+  %res = bitcast <4 x i32> %v to <4 x i32>
+  ret <4 x i32> %res
+}
+
+; CHECK-LABEL: v4i32_to_v2i64:
+define <2 x i64> @v4i32_to_v2i64(<4 x i32> %v) {
+  %res = bitcast <4 x i32> %v to <2 x i64>
+  ret <2 x i64> %res
+}
+
+; CHECK-LABEL: v4i32_to_v4f32:
+define <4 x float> @v4i32_to_v4f32(<4 x i32> %v) {
+  %res = bitcast <4 x i32> %v to <4 x float>
+  ret <4 x float> %res
+}
+
+; CHECK-LABEL: v4i32_to_v2f64:
+define <2 x double> @v4i32_to_v2f64(<4 x i32> %v) {
+  %res = bitcast <4 x i32> %v to <2 x double>
+  ret <2 x double> %res
+}
+
+; CHECK-LABEL: v2i64_to_v16i8:
+define <16 x i8> @v2i64_to_v16i8(<2 x i64> %v) {
+  %res = bitcast <2 x i64> %v to <16 x i8>
+  ret <16 x i8> %res
+}
+
+; CHECK-LABEL: v2i64_to_v8i16:
+define <8 x i16> @v2i64_to_v8i16(<2 x i64> %v) {
+  %res = bitcast <2 x i64> %v to <8 x i16>
+  ret <8 x i16> %res
+}
+
+; CHECK-LABEL: v2i64_to_v4i32:
+define <4 x i32> @v2i64_to_v4i32(<2 x i64> %v) {
+  %res = bitcast <2 x i64> %v to <4 x i32>
+  ret <4 x i32> %res
+}
+
+; CHECK-LABEL: v2i64_to_v2i64:
+define <2 x i64> @v2i64_to_v2i64(<2 x i64> %v) {
+  %res = bitcast <2 x i64> %v to <2 x i64>
+  ret <2 x i64> %res
+}
+
+; CHECK-LABEL: v2i64_to_v4f32:
+define <4 x float> @v2i64_to_v4f32(<2 x i64> %v) {
+  %res = bitcast <2 x i64> %v to <4 x float>
+  ret <4 x float> %res
+}
+
+; CHECK-LABEL: v2i64_to_v2f64:
+define <2 x double> @v2i64_to_v2f64(<2 x i64> %v) {
+  %res = bitcast <2 x i64> %v to <2 x double>
+  ret <2 x double> %res
+}
+
+; CHECK-LABEL: v4f32_to_v16i8:
+define <16 x i8> @v4f32_to_v16i8(<4 x float> %v) {
+  %res = bitcast <4 x float> %v to <16 x i8>
+  ret <16 x i8> %res
+}
+
+; CHECK-LABEL: v4f32_to_v8i16:
+define <8 x i16> @v4f32_to_v8i16(<4 x float> %v) {
+  %res = bitcast <4 x float> %v to <8 x i16>
+  ret <8 x i16> %res
+}
+
+; CHECK-LABEL: v4f32_to_v4i32:
+define <4 x i32> @v4f32_to_v4i32(<4 x float> %v) {
+  %res = bitcast <4 x float> %v to <4 x i32>
+  ret <4 x i32> %res
+}
+
+; CHECK-LABEL: v4f32_to_v2i64:
+define <2 x i64> @v4f32_to_v2i64(<4 x float> %v) {
+  %res = bitcast <4 x float> %v to <2 x i64>
+  ret <2 x i64> %res
+}
+
+; CHECK-LABEL: v4f32_to_v4f32:
+define <4 x float> @v4f32_to_v4f32(<4 x float> %v) {
+  %res = bitcast <4 x float> %v to <4 x float>
+  ret <4 x float> %res
+}
+
+; CHECK-LABEL: v4f32_to_v2f64:
+define <2 x double> @v4f32_to_v2f64(<4 x float> %v) {
+  %res = bitcast <4 x float> %v to <2 x double>
+  ret <2 x double> %res
+}
+
+; CHECK-LABEL: v2f64_to_v16i8:
+define <16 x i8> @v2f64_to_v16i8(<2 x double> %v) {
+  %res = bitcast <2 x double> %v to <16 x i8>
+  ret <16 x i8> %res
+}
+
+; CHECK-LABEL: v2f64_to_v8i16:
+define <8 x i16> @v2f64_to_v8i16(<2 x double> %v) {
+  %res = bitcast <2 x double> %v to <8 x i16>
+  ret <8 x i16> %res
+}
+
+; CHECK-LABEL: v2f64_to_v4i32:
+define <4 x i32> @v2f64_to_v4i32(<2 x double> %v) {
+  %res = bitcast <2 x double> %v to <4 x i32>
+  ret <4 x i32> %res
+}
+
+; CHECK-LABEL: v2f64_to_v2i64:
+define <2 x i64> @v2f64_to_v2i64(<2 x double> %v) {
+  %res = bitcast <2 x double> %v to <2 x i64>
+  ret <2 x i64> %res
+}
+
+; CHECK-LABEL: v2f64_to_v4f32:
+define <4 x float> @v2f64_to_v4f32(<2 x double> %v) {
+  %res = bitcast <2 x double> %v to <4 x float>
+  ret <4 x float> %res
+}
+
+; CHECK-LABEL: v2f64_to_v2f64:
+define <2 x double> @v2f64_to_v2f64(<2 x double> %v) {
+  %res = bitcast <2 x double> %v to <2 x double>
+  ret <2 x double> %res
+}