[WebAssembly] Update SIMD binary arithmetic

Add missing SIMD types (v2f64) and binary ops. Also adds
tablegen support for automatically prepending prefix byte to SIMD
opcodes.

Differential Revision: https://reviews.llvm.org/D50292

Patch by Thomas Lively

llvm-svn: 339186
This commit is contained in:
Derek Schuff 2018-08-07 21:24:01 +00:00
parent dca675a0d8
commit 51ed131ed2
15 changed files with 203 additions and 22 deletions

View File

@ -229,7 +229,9 @@ const char *llvm::WebAssembly::TypeToString(MVT Ty) {
case MVT::v16i8:
case MVT::v8i16:
case MVT::v4i32:
case MVT::v2i64:
case MVT::v4f32:
case MVT::v2f64:
return "v128";
case MVT::ExceptRef:
return "except_ref";

View File

@ -50,7 +50,7 @@ MVT WebAssemblyAsmPrinter::getRegType(unsigned RegNo) const {
const TargetRegisterInfo *TRI = Subtarget->getRegisterInfo();
const TargetRegisterClass *TRC = MRI->getRegClass(RegNo);
for (MVT T : {MVT::i32, MVT::i64, MVT::f32, MVT::f64, MVT::v16i8, MVT::v8i16,
MVT::v4i32, MVT::v4f32})
MVT::v4i32, MVT::v2i64, MVT::v4f32, MVT::v2f64})
if (TRI->isTypeLegalForClass(*TRC, T))
return T;
LLVM_DEBUG(errs() << "Unknown type for register number: " << RegNo);
@ -175,7 +175,9 @@ void WebAssemblyAsmPrinter::EmitInstruction(const MachineInstr *MI) {
case WebAssembly::ARGUMENT_v16i8:
case WebAssembly::ARGUMENT_v8i16:
case WebAssembly::ARGUMENT_v4i32:
case WebAssembly::ARGUMENT_v2i64:
case WebAssembly::ARGUMENT_v4f32:
case WebAssembly::ARGUMENT_v2f64:
// These represent values which are live into the function entry, so there's
// no instruction to emit.
break;
@ -186,7 +188,9 @@ void WebAssemblyAsmPrinter::EmitInstruction(const MachineInstr *MI) {
case WebAssembly::FALLTHROUGH_RETURN_v16i8:
case WebAssembly::FALLTHROUGH_RETURN_v8i16:
case WebAssembly::FALLTHROUGH_RETURN_v4i32:
case WebAssembly::FALLTHROUGH_RETURN_v4f32: {
case WebAssembly::FALLTHROUGH_RETURN_v2i64:
case WebAssembly::FALLTHROUGH_RETURN_v4f32:
case WebAssembly::FALLTHROUGH_RETURN_v2f64: {
// These instructions represent the implicit return at the end of a
// function body. The operand is always a pop.
assert(MFI->isVRegStackified(MI->getOperand(0).getReg()));

View File

@ -256,7 +256,9 @@ static void FixEndsAtEndOfFunction(
case MVT::v16i8:
case MVT::v8i16:
case MVT::v4i32:
case MVT::v2i64:
case MVT::v4f32:
case MVT::v2f64:
retType = WebAssembly::ExprType::V128;
break;
case MVT::ExceptRef: retType = WebAssembly::ExprType::ExceptRef; break;

View File

@ -72,7 +72,9 @@ static unsigned GetNonPseudoCallIndirectOpcode(const MachineInstr &MI) {
case PCALL_INDIRECT_v16i8: return CALL_INDIRECT_v16i8;
case PCALL_INDIRECT_v8i16: return CALL_INDIRECT_v8i16;
case PCALL_INDIRECT_v4i32: return CALL_INDIRECT_v4i32;
case PCALL_INDIRECT_v2i64: return CALL_INDIRECT_v2i64;
case PCALL_INDIRECT_v4f32: return CALL_INDIRECT_v4f32;
case PCALL_INDIRECT_v2f64: return CALL_INDIRECT_v2f64;
default: return INSTRUCTION_LIST_END;
}
}
@ -133,4 +135,3 @@ bool WebAssemblyCallIndirectFixup::runOnMachineFunction(MachineFunction &MF) {
return Changed;
}

View File

@ -134,7 +134,9 @@ private:
case MVT::v16i8:
case MVT::v8i16:
case MVT::v4i32:
case MVT::v2i64:
case MVT::v4f32:
case MVT::v2f64:
if (Subtarget->hasSIMD128())
return VT;
break;
@ -678,10 +680,18 @@ bool WebAssemblyFastISel::fastLowerArguments() {
Opc = WebAssembly::ARGUMENT_v4i32;
RC = &WebAssembly::V128RegClass;
break;
case MVT::v2i64:
Opc = WebAssembly::ARGUMENT_v2i64;
RC = &WebAssembly::V128RegClass;
break;
case MVT::v4f32:
Opc = WebAssembly::ARGUMENT_v4f32;
RC = &WebAssembly::V128RegClass;
break;
case MVT::v2f64:
Opc = WebAssembly::ARGUMENT_v2f64;
RC = &WebAssembly::V128RegClass;
break;
case MVT::ExceptRef:
Opc = WebAssembly::ARGUMENT_EXCEPT_REF;
RC = &WebAssembly::EXCEPT_REFRegClass;
@ -782,11 +792,21 @@ bool WebAssemblyFastISel::selectCall(const Instruction *I) {
IsDirect ? WebAssembly::CALL_v4i32 : WebAssembly::PCALL_INDIRECT_v4i32;
ResultReg = createResultReg(&WebAssembly::V128RegClass);
break;
case MVT::v2i64:
Opc =
IsDirect ? WebAssembly::CALL_v2i64 : WebAssembly::PCALL_INDIRECT_v2i64;
ResultReg = createResultReg(&WebAssembly::V128RegClass);
break;
case MVT::v4f32:
Opc =
IsDirect ? WebAssembly::CALL_v4f32 : WebAssembly::PCALL_INDIRECT_v4f32;
ResultReg = createResultReg(&WebAssembly::V128RegClass);
break;
case MVT::v2f64:
Opc =
IsDirect ? WebAssembly::CALL_v2f64 : WebAssembly::PCALL_INDIRECT_v2f64;
ResultReg = createResultReg(&WebAssembly::V128RegClass);
break;
case MVT::ExceptRef:
Opc = IsDirect ? WebAssembly::CALL_EXCEPT_REF
: WebAssembly::PCALL_INDIRECT_EXCEPT_REF;
@ -1297,9 +1317,15 @@ bool WebAssemblyFastISel::selectRet(const Instruction *I) {
case MVT::v4i32:
Opc = WebAssembly::RETURN_v4i32;
break;
case MVT::v2i64:
Opc = WebAssembly::RETURN_v2i64;
break;
case MVT::v4f32:
Opc = WebAssembly::RETURN_v4f32;
break;
case MVT::v2f64:
Opc = WebAssembly::RETURN_v2f64;
break;
case MVT::ExceptRef:
Opc = WebAssembly::RETURN_EXCEPT_REF;
break;

View File

@ -59,7 +59,9 @@ WebAssemblyTargetLowering::WebAssemblyTargetLowering(
addRegisterClass(MVT::v16i8, &WebAssembly::V128RegClass);
addRegisterClass(MVT::v8i16, &WebAssembly::V128RegClass);
addRegisterClass(MVT::v4i32, &WebAssembly::V128RegClass);
addRegisterClass(MVT::v2i64, &WebAssembly::V128RegClass);
addRegisterClass(MVT::v4f32, &WebAssembly::V128RegClass);
addRegisterClass(MVT::v2f64, &WebAssembly::V128RegClass);
}
// Compute derived properties from the register classes.
computeRegisterProperties(Subtarget->getRegisterInfo());

View File

@ -91,7 +91,9 @@ let Uses = [SP32, SP64], isCall = 1 in {
defm "" : SIMD_CALL<v16i8, "i8x16.">;
defm "" : SIMD_CALL<v8i16, "i16x8.">;
defm "" : SIMD_CALL<v4i32, "i32x4.">;
defm "" : SIMD_CALL<v2i64, "i64x2.">;
defm "" : SIMD_CALL<v4f32, "f32x4.">;
defm "" : SIMD_CALL<v2f64, "f64x2.">;
defm CALL_VOID : I<(outs), (ins function32_op:$callee, variable_ops),
(outs), (ins function32_op:$callee),
@ -132,8 +134,12 @@ def : Pat<(v8i16 (WebAssemblycall1 (WebAssemblywrapper tglobaladdr:$callee))),
(CALL_v8i16 tglobaladdr:$callee)>, Requires<[HasSIMD128]>;
def : Pat<(v4i32 (WebAssemblycall1 (WebAssemblywrapper tglobaladdr:$callee))),
(CALL_v4i32 tglobaladdr:$callee)>, Requires<[HasSIMD128]>;
def : Pat<(v2i64 (WebAssemblycall1 (WebAssemblywrapper tglobaladdr:$callee))),
(CALL_v2i64 tglobaladdr:$callee)>, Requires<[HasSIMD128]>;
def : Pat<(v4f32 (WebAssemblycall1 (WebAssemblywrapper tglobaladdr:$callee))),
(CALL_v4f32 tglobaladdr:$callee)>, Requires<[HasSIMD128]>;
def : Pat<(v2f64 (WebAssemblycall1 (WebAssemblywrapper tglobaladdr:$callee))),
(CALL_v2f64 tglobaladdr:$callee)>, Requires<[HasSIMD128]>;
def : Pat<(ExceptRef
(WebAssemblycall1 (WebAssemblywrapper tglobaladdr:$callee))),
(CALL_EXCEPT_REF tglobaladdr:$callee)>;
@ -155,8 +161,12 @@ def : Pat<(v8i16 (WebAssemblycall1 (WebAssemblywrapper texternalsym:$callee))),
(CALL_v8i16 texternalsym:$callee)>, Requires<[HasSIMD128]>;
def : Pat<(v4i32 (WebAssemblycall1 (WebAssemblywrapper texternalsym:$callee))),
(CALL_v4i32 texternalsym:$callee)>, Requires<[HasSIMD128]>;
def : Pat<(v2i64 (WebAssemblycall1 (WebAssemblywrapper texternalsym:$callee))),
(CALL_v2i64 texternalsym:$callee)>, Requires<[HasSIMD128]>;
def : Pat<(v4f32 (WebAssemblycall1 (WebAssemblywrapper texternalsym:$callee))),
(CALL_v4f32 texternalsym:$callee)>, Requires<[HasSIMD128]>;
def : Pat<(v2f64 (WebAssemblycall1 (WebAssemblywrapper texternalsym:$callee))),
(CALL_v2f64 texternalsym:$callee)>, Requires<[HasSIMD128]>;
def : Pat<(ExceptRef
(WebAssemblycall1 (WebAssemblywrapper texternalsym:$callee))),
(CALL_EXCEPT_REF texternalsym:$callee)>;

View File

@ -124,7 +124,9 @@ let isReturn = 1 in {
defm "": SIMD_RETURN<v16i8>;
defm "": SIMD_RETURN<v8i16>;
defm "": SIMD_RETURN<v4i32>;
defm "": SIMD_RETURN<v2i64>;
defm "": SIMD_RETURN<v4f32>;
defm "": SIMD_RETURN<v2f64>;
defm RETURN_VOID : NRI<(outs), (ins), [(WebAssemblyreturn)], "return", 0x0f>;

View File

@ -60,9 +60,9 @@ multiclass NRI<dag oops, dag iops, list<dag> pattern, string asmstr = "",
multiclass SIMD_I<dag oops_r, dag iops_r, dag oops_s, dag iops_s,
list<dag> pattern_r, string asmstr_r = "",
string asmstr_s = "", bits<32> inst = -1> {
string asmstr_s = "", bits<32> simdop = -1> {
defm "" : I<oops_r, iops_r, oops_s, iops_s, pattern_r, asmstr_r, asmstr_s,
inst>,
!or(0xfd00, !and(0xff, simdop))>,
Requires<[HasSIMD128]>;
}
@ -119,31 +119,45 @@ multiclass BinaryFP<SDNode node, string name, bits<32> f32Inst,
!strconcat("f64.", !strconcat(name, "\t$dst, $lhs, $rhs")),
!strconcat("f64.", name), f64Inst>;
}
multiclass SIMDBinary<SDNode node, SDNode fnode, string name> {
multiclass SIMDBinaryInt<SDNode node, string name, bits<32> baseInst> {
defm _I8x16 : SIMD_I<(outs V128:$dst), (ins V128:$lhs, V128:$rhs),
(outs), (ins),
[(set (v16i8 V128:$dst), (node V128:$lhs, V128:$rhs))],
!strconcat("i8x16.",
!strconcat(name, "\t$dst, $lhs, $rhs")),
!strconcat("i8x16.", name)>;
!strconcat("i8x16.", name), baseInst>;
defm _I16x8 : SIMD_I<(outs V128:$dst), (ins V128:$lhs, V128:$rhs),
(outs), (ins),
[(set (v8i16 V128:$dst), (node V128:$lhs, V128:$rhs))],
!strconcat("i16x8.",
!strconcat(name, "\t$dst, $lhs, $rhs")),
!strconcat("i16x8.", name)>;
!strconcat("i16x8.", name), !add(baseInst, 1)>;
defm _I32x4 : SIMD_I<(outs V128:$dst), (ins V128:$lhs, V128:$rhs),
(outs), (ins),
[(set (v4i32 V128:$dst), (node V128:$lhs, V128:$rhs))],
!strconcat("i32x4.",
!strconcat(name, "\t$dst, $lhs, $rhs")),
!strconcat("i32x4.", name)>;
!strconcat("i32x4.", name), !add(baseInst, 2)>;
defm _I64x2 : SIMD_I<(outs V128:$dst), (ins V128:$lhs, V128:$rhs),
(outs), (ins),
[(set (v2i64 V128:$dst), (node V128:$lhs, V128:$rhs))],
!strconcat("i64x2.",
!strconcat(name, "\t$dst, $lhs, $rhs")),
!strconcat("i64x2.", name), !add(baseInst, 3)>;
}
multiclass SIMDBinaryFP<SDNode node, string name, bits<32> baseInst> {
defm _F32x4 : SIMD_I<(outs V128:$dst), (ins V128:$lhs, V128:$rhs),
(outs), (ins),
[(set (v4f32 V128:$dst), (fnode V128:$lhs, V128:$rhs))],
[(set (v4f32 V128:$dst), (node V128:$lhs, V128:$rhs))],
!strconcat("f32x4.",
!strconcat(name, "\t$dst, $lhs, $rhs")),
!strconcat("f32x4.", name)>;
!strconcat("f32x4.", name), baseInst>;
defm _F64x2 : SIMD_I<(outs V128:$dst), (ins V128:$lhs, V128:$rhs),
(outs), (ins),
[(set (v2f64 V128:$dst), (node V128:$lhs, V128:$rhs))],
!strconcat("f64x2.",
!strconcat(name, "\t$dst, $lhs, $rhs")),
!strconcat("f64x2.", name), !add(baseInst, 1)>;
}
multiclass ComparisonInt<CondCode cond, string name, bits<32> i32Inst, bits<32> i64Inst> {
defm _I32 : I<(outs I32:$dst), (ins I32:$lhs, I32:$rhs), (outs), (ins),

View File

@ -172,7 +172,9 @@ defm "": ARGUMENT<EXCEPT_REF>;
defm "": SIMD_ARGUMENT<v16i8>;
defm "": SIMD_ARGUMENT<v8i16>;
defm "": SIMD_ARGUMENT<v4i32>;
defm "": SIMD_ARGUMENT<v2i64>;
defm "": SIMD_ARGUMENT<v4f32>;
defm "": SIMD_ARGUMENT<v2f64>;
let Defs = [ARGUMENTS] in {

View File

@ -12,8 +12,19 @@
///
//===----------------------------------------------------------------------===//
let isCommutable = 1 in {
defm ADD : SIMDBinary<add, fadd, "add ">;
defm MUL: SIMDBinary<mul, fmul, "mul ">;
} // isCommutable = 1
defm SUB: SIMDBinary<sub, fsub, "sub ">;
let Defs = [ARGUMENTS] in {
let isCommutable = 1 in
defm ADD : SIMDBinaryInt<add, "add ", 24>;
defm SUB : SIMDBinaryInt<sub, "sub ", 28>;
let isCommutable = 1 in
defm MUL : SIMDBinaryInt<mul, "mul ", 32>;
let isCommutable = 1 in
defm ADD : SIMDBinaryFP<fadd, "add ", 122>;
defm SUB : SIMDBinaryFP<fsub, "sub ", 124>;
defm DIV : SIMDBinaryFP<fdiv, "div ", 126>;
let isCommutable = 1 in
defm MUL : SIMDBinaryFP<fmul, "mul ", 128>;
} // Defs = [ARGUMENTS]

View File

@ -192,11 +192,21 @@ bool WebAssemblyPeephole::runOnMachineFunction(MachineFunction &MF) {
MI, MBB, MF, MFI, MRI, TII, WebAssembly::FALLTHROUGH_RETURN_v4i32,
WebAssembly::COPY_V128);
break;
case WebAssembly::RETURN_v2i64:
Changed |= MaybeRewriteToFallthrough(
MI, MBB, MF, MFI, MRI, TII, WebAssembly::FALLTHROUGH_RETURN_v2i64,
WebAssembly::COPY_V128);
break;
case WebAssembly::RETURN_v4f32:
Changed |= MaybeRewriteToFallthrough(
MI, MBB, MF, MFI, MRI, TII, WebAssembly::FALLTHROUGH_RETURN_v4f32,
WebAssembly::COPY_V128);
break;
case WebAssembly::RETURN_v2f64:
Changed |= MaybeRewriteToFallthrough(
MI, MBB, MF, MFI, MRI, TII, WebAssembly::FALLTHROUGH_RETURN_v2f64,
WebAssembly::COPY_V128);
break;
case WebAssembly::RETURN_VOID:
Changed |= MaybeRewriteToFallthrough(
MI, MBB, MF, MFI, MRI, TII, WebAssembly::FALLTHROUGH_RETURN_VOID,

View File

@ -63,6 +63,6 @@ def I32 : WebAssemblyRegClass<[i32], 32, (add FP32, SP32, I32_0)>;
def I64 : WebAssemblyRegClass<[i64], 64, (add FP64, SP64, I64_0)>;
def F32 : WebAssemblyRegClass<[f32], 32, (add F32_0)>;
def F64 : WebAssemblyRegClass<[f64], 64, (add F64_0)>;
def V128 : WebAssemblyRegClass<[v4f32, v4i32, v16i8, v8i16], 128, (add V128_0)>;
def V128 : WebAssemblyRegClass<[v4f32, v2f64, v2i64, v4i32, v16i8, v8i16], 128,
(add V128_0)>;
def EXCEPT_REF : WebAssemblyRegClass<[ExceptRef], 0, (add EXCEPT_REF_0)>;

View File

@ -34,7 +34,9 @@ bool WebAssembly::isArgument(const MachineInstr &MI) {
case WebAssembly::ARGUMENT_v16i8:
case WebAssembly::ARGUMENT_v8i16:
case WebAssembly::ARGUMENT_v4i32:
case WebAssembly::ARGUMENT_v2i64:
case WebAssembly::ARGUMENT_v4f32:
case WebAssembly::ARGUMENT_v2f64:
return true;
default:
return false;
@ -88,7 +90,9 @@ bool WebAssembly::isCallDirect(const MachineInstr &MI) {
case WebAssembly::CALL_v16i8:
case WebAssembly::CALL_v8i16:
case WebAssembly::CALL_v4i32:
case WebAssembly::CALL_v2i64:
case WebAssembly::CALL_v4f32:
case WebAssembly::CALL_v2f64:
case WebAssembly::CALL_EXCEPT_REF:
return true;
default:
@ -106,7 +110,9 @@ bool WebAssembly::isCallIndirect(const MachineInstr &MI) {
case WebAssembly::CALL_INDIRECT_v16i8:
case WebAssembly::CALL_INDIRECT_v8i16:
case WebAssembly::CALL_INDIRECT_v4i32:
case WebAssembly::CALL_INDIRECT_v2i64:
case WebAssembly::CALL_INDIRECT_v4f32:
case WebAssembly::CALL_INDIRECT_v2f64:
case WebAssembly::CALL_INDIRECT_EXCEPT_REF:
return true;
default:

View File

@ -8,10 +8,6 @@
target datalayout = "e-m:e-p:32:32-i64:64-n32:64-S128"
target triple = "wasm32-unknown-unknown"
declare i32 @llvm.ctlz.i32(i32, i1)
declare i32 @llvm.cttz.i32(i32, i1)
declare i32 @llvm.ctpop.i32(i32)
; ==============================================================================
; 16 x i8
; ==============================================================================
@ -120,6 +116,42 @@ define <4 x i32> @mul_v4i32(<4 x i32> %x, <4 x i32> %y) {
ret <4 x i32> %a
}
; ==============================================================================
; 2 x i64
; ==============================================================================
; CHECK-LABEL: add_v2i64
; NO-SIMD128-NOT: i64x2
; SIMD128: .param v128, v128{{$}}
; SIMD128: .result v128{{$}}
; SIMD128: i64x2.add $push0=, $0, $1{{$}}
; SIMD128: return $pop0{{$}}
define <2 x i64> @add_v2i64(<2 x i64> %x, <2 x i64> %y) {
%a = add <2 x i64> %x, %y
ret <2 x i64> %a
}
; CHECK-LABEL: sub_v2i64
; NO-SIMD128-NOT: i64x2
; SIMD128: .param v128, v128{{$}}
; SIMD128: .result v128{{$}}
; SIMD128: i64x2.sub $push0=, $0, $1{{$}}
; SIMD128: return $pop0{{$}}
define <2 x i64> @sub_v2i64(<2 x i64> %x, <2 x i64> %y) {
%a = sub <2 x i64> %x, %y
ret <2 x i64> %a
}
; CHECK-LABEL: mul_v2i64
; NO-SIMD128-NOT: i64x2
; SIMD128: .param v128, v128{{$}}
; SIMD128: .result v128{{$}}
; SIMD128: i64x2.mul $push0=, $0, $1{{$}}
; SIMD128: return $pop0{{$}}
define <2 x i64> @mul_v2i64(<2 x i64> %x, <2 x i64> %y) {
%a = mul <2 x i64> %x, %y
ret <2 x i64> %a
}
; ==============================================================================
; 4 x float
; ==============================================================================
@ -145,6 +177,17 @@ define <4 x float> @sub_v4f32(<4 x float> %x, <4 x float> %y) {
ret <4 x float> %a
}
; CHECK-LABEL: div_v4f32
; NO-SIMD128-NOT: f32x4
; SIMD128: .param v128, v128{{$}}
; SIMD128: .result v128{{$}}
; SIMD128: f32x4.div $push0=, $0, $1{{$}}
; SIMD128: return $pop0{{$}}
define <4 x float> @div_v4f32(<4 x float> %x, <4 x float> %y) {
%a = fdiv <4 x float> %x, %y
ret <4 x float> %a
}
; CHECK-LABEL: mul_v4f32
; NO-SIMD128-NOT: f32x4
; SIMD128: .param v128, v128{{$}}
@ -156,3 +199,49 @@ define <4 x float> @mul_v4f32(<4 x float> %x, <4 x float> %y) {
ret <4 x float> %a
}
; ==============================================================================
; 2 x double
; ==============================================================================
; CHECK-LABEL: add_v2f64
; NO-SIMD128-NOT: f64x2
; SIMD128: .param v128, v128{{$}}
; SIMD128: .result v128{{$}}
; SIMD128: f64x2.add $push0=, $0, $1{{$}}
; SIMD128: return $pop0{{$}}
define <2 x double> @add_v2f64(<2 x double> %x, <2 x double> %y) {
%a = fadd <2 x double> %x, %y
ret <2 x double> %a
}
; CHECK-LABEL: sub_v2f64
; NO-SIMD128-NOT: f64x2
; SIMD128: .param v128, v128{{$}}
; SIMD128: .result v128{{$}}
; SIMD128: f64x2.sub $push0=, $0, $1{{$}}
; SIMD128: return $pop0{{$}}
define <2 x double> @sub_v2f64(<2 x double> %x, <2 x double> %y) {
%a = fsub <2 x double> %x, %y
ret <2 x double> %a
}
; CHECK-LABEL: div_v2f64
; NO-SIMD128-NOT: f64x2
; SIMD128: .param v128, v128{{$}}
; SIMD128: .result v128{{$}}
; SIMD128: f64x2.div $push0=, $0, $1{{$}}
; SIMD128: return $pop0{{$}}
define <2 x double> @div_v2f64(<2 x double> %x, <2 x double> %y) {
%a = fdiv <2 x double> %x, %y
ret <2 x double> %a
}
; CHECK-LABEL: mul_v2f64
; NO-SIMD128-NOT: f64x2
; SIMD128: .param v128, v128{{$}}
; SIMD128: .result v128{{$}}
; SIMD128: f64x2.mul $push0=, $0, $1{{$}}
; SIMD128: return $pop0{{$}}
define <2 x double> @mul_v2f64(<2 x double> %x, <2 x double> %y) {
%a = fmul <2 x double> %x, %y
ret <2 x double> %a
}