Skip to content

Commit c781052

Browse files
farzonlfarzonl
committed
[DirectX] add support for i64 buffer load/stores
1 parent de33390 commit c781052

File tree

5 files changed

+281
-47
lines changed

5 files changed

+281
-47
lines changed

‎llvm/lib/Target/DirectX/DXILIntrinsicExpansion.cpp

Lines changed: 134 additions & 45 deletions
Original file line numberDiff line numberDiff line change
@@ -25,6 +25,7 @@
2525
#include "llvm/IR/PassManager.h"
2626
#include "llvm/IR/Type.h"
2727
#include "llvm/Pass.h"
28+
#include "llvm/Support/Casting.h"
2829
#include "llvm/Support/ErrorHandling.h"
2930
#include "llvm/Support/MathExtras.h"
3031

@@ -70,15 +71,17 @@ static bool isIntrinsicExpansion(Function &F) {
7071
case Intrinsic::vector_reduce_add:
7172
case Intrinsic::vector_reduce_fadd:
7273
return true;
73-
case Intrinsic::dx_resource_load_typedbuffer:
74-
// We need to handle doubles and vector of doubles.
75-
return F.getReturnType()
76-
->getStructElementType(0)
77-
->getScalarType()
78-
->isDoubleTy();
79-
case Intrinsic::dx_resource_store_typedbuffer:
80-
// We need to handle doubles and vector of doubles.
81-
return F.getFunctionType()->getParamType(2)->getScalarType()->isDoubleTy();
74+
case Intrinsic::dx_resource_load_typedbuffer: {
75+
// We need to handle i64, doubles, and vectors of them.
76+
Type *ScalarTy =
77+
F.getReturnType()->getStructElementType(0)->getScalarType();
78+
return ScalarTy->isDoubleTy() || ScalarTy->isIntegerTy(64);
79+
}
80+
case Intrinsic::dx_resource_store_typedbuffer: {
81+
// We need to handle i64 and doubles and vectors of i64 and doubles.
82+
Type *ScalarTy = F.getFunctionType()->getParamType(2)->getScalarType();
83+
return ScalarTy->isDoubleTy() || ScalarTy->isIntegerTy(64);
84+
}
8285
}
8386
return false;
8487
}
@@ -545,13 +548,15 @@ static bool expandTypedBufferLoadIntrinsic(CallInst *Orig) {
545548
IRBuilder<> Builder(Orig);
546549

547550
Type *BufferTy = Orig->getType()->getStructElementType(0);
548-
assert(BufferTy->getScalarType()->isDoubleTy() &&
549-
"Only expand double or double2");
551+
Type *ScalarTy = BufferTy->getScalarType();
552+
bool IsDouble = ScalarTy->isDoubleTy();
553+
assert(IsDouble || ScalarTy->isIntegerTy(64) &&
554+
"Only expand double or int64 scalars or vectors");
550555

551556
unsigned ExtractNum = 2;
552557
if (auto *VT = dyn_cast<FixedVectorType>(BufferTy)) {
553558
assert(VT->getNumElements() == 2 &&
554-
"TypedBufferLoad double vector has wrong size");
559+
"TypedBufferLoad vector must be size 2");
555560
ExtractNum = 4;
556561
}
557562

@@ -570,22 +575,54 @@ static bool expandTypedBufferLoadIntrinsic(CallInst *Orig) {
570575
ExtractElements.push_back(
571576
Builder.CreateExtractElement(Extract, Builder.getInt32(I)));
572577

573-
// combine into double(s)
578+
// combine into double(s) or int64(s)
574579
Value *Result = PoisonValue::get(BufferTy);
575580
for (unsigned I = 0; I < ExtractNum; I += 2) {
576-
Value *Dbl =
577-
Builder.CreateIntrinsic(Builder.getDoubleTy(), Intrinsic::dx_asdouble,
578-
{ExtractElements[I], ExtractElements[I + 1]});
581+
Value *Combined = nullptr;
582+
if (IsDouble) {
583+
// For doubles, use dx_asdouble intrinsic
584+
Combined =
585+
Builder.CreateIntrinsic(Builder.getDoubleTy(), Intrinsic::dx_asdouble,
586+
{ExtractElements[I], ExtractElements[I + 1]});
587+
} else {
588+
// For int64, manually combine two int32s
589+
// First, zero-extend both values to i64
590+
Value *Lo = Builder.CreateZExt(ExtractElements[I], Builder.getInt64Ty());
591+
Value *Hi =
592+
Builder.CreateZExt(ExtractElements[I + 1], Builder.getInt64Ty());
593+
// Shift the high bits left by 32 bits
594+
Value *ShiftedHi = Builder.CreateShl(Hi, Builder.getInt64(32));
595+
// OR the high and low bits together
596+
Combined = Builder.CreateOr(Lo, ShiftedHi);
597+
}
598+
579599
if (ExtractNum == 4)
580-
Result =
581-
Builder.CreateInsertElement(Result, Dbl, Builder.getInt32(I / 2));
600+
Result = Builder.CreateInsertElement(Result, Combined,
601+
Builder.getInt32(I / 2));
582602
else
583-
Result = Dbl;
603+
Result = Combined;
584604
}
585605

586606
Value *CheckBit = nullptr;
587607
for (User *U : make_early_inc_range(Orig->users())) {
588-
auto *EVI = cast<ExtractValueInst>(U);
608+
if (auto *Ret = dyn_cast<ReturnInst>(U)) {
609+
// For return instructions, we need to handle the case where the function
610+
// is directly returning the result of the call
611+
Type *RetTy = Ret->getFunction()->getReturnType();
612+
Value *StructRet = PoisonValue::get(RetTy);
613+
StructRet = Builder.CreateInsertValue(StructRet, Result, {0});
614+
Value *CheckBitForRet = Builder.CreateExtractValue(Load, {1});
615+
StructRet = Builder.CreateInsertValue(StructRet, CheckBitForRet, {1});
616+
Ret->setOperand(0, StructRet);
617+
continue;
618+
}
619+
auto *EVI = dyn_cast<ExtractValueInst>(U);
620+
if (!EVI) {
621+
// If it's not a ReturnInst or ExtractValueInst, we don't know how to
622+
// handle it
623+
llvm_unreachable("Unexpected user of typedbufferload");
624+
}
625+
589626
ArrayRef<unsigned> Indices = EVI->getIndices();
590627
assert(Indices.size() == 1);
591628

@@ -609,38 +646,90 @@ static bool expandTypedBufferStoreIntrinsic(CallInst *Orig) {
609646
IRBuilder<> Builder(Orig);
610647

611648
Type *BufferTy = Orig->getFunctionType()->getParamType(2);
612-
assert(BufferTy->getScalarType()->isDoubleTy() &&
613-
"Only expand double or double2");
649+
Type *ScalarTy = BufferTy->getScalarType();
650+
bool IsDouble = ScalarTy->isDoubleTy();
651+
assert((IsDouble || ScalarTy->isIntegerTy(64)) &&
652+
"Only expand double or int64 scalars or vectors");
614653

615654
unsigned ExtractNum = 2;
616655
if (auto *VT = dyn_cast<FixedVectorType>(BufferTy)) {
617656
assert(VT->getNumElements() == 2 &&
618-
"TypedBufferStore double vector has wrong size");
657+
"TypedBufferStore vector must be size 2");
619658
ExtractNum = 4;
620659
}
660+
if (IsDouble) {
661+
Type *SplitElementTy = Builder.getInt32Ty();
662+
if (ExtractNum == 4)
663+
SplitElementTy = VectorType::get(SplitElementTy, 2, false);
664+
665+
// Handle double type(s) - keep original behavior
666+
auto *SplitTy = llvm::StructType::get(SplitElementTy, SplitElementTy);
667+
Value *Split = Builder.CreateIntrinsic(SplitTy, Intrinsic::dx_splitdouble,
668+
{Orig->getOperand(2)});
669+
// create our vector
670+
Value *LowBits = Builder.CreateExtractValue(Split, 0);
671+
Value *HighBits = Builder.CreateExtractValue(Split, 1);
672+
Value *Val;
673+
if (ExtractNum == 2) {
674+
Val = PoisonValue::get(VectorType::get(Builder.getInt32Ty(), 2, false));
675+
Val = Builder.CreateInsertElement(Val, LowBits, Builder.getInt32(0));
676+
Val = Builder.CreateInsertElement(Val, HighBits, Builder.getInt32(1));
677+
} else
678+
Val = Builder.CreateShuffleVector(LowBits, HighBits, {0, 2, 1, 3});
679+
680+
Builder.CreateIntrinsic(Builder.getVoidTy(),
681+
Intrinsic::dx_resource_store_typedbuffer,
682+
{Orig->getOperand(0), Orig->getOperand(1), Val});
683+
} else {
684+
// Handle int64 type(s)
685+
Value *InputVal = Orig->getOperand(2);
686+
Value *Val;
687+
688+
if (ExtractNum == 4) {
689+
// Handle vector of int64
690+
Type *Int32x4Ty = VectorType::get(Builder.getInt32Ty(), 4, false);
691+
Val = PoisonValue::get(Int32x4Ty);
692+
693+
for (unsigned I = 0; I < 2; ++I) {
694+
// Extract each int64 element
695+
Value *Int64Val =
696+
Builder.CreateExtractElement(InputVal, Builder.getInt32(I));
697+
698+
// Get low 32 bits by truncating to i32
699+
Value *LowBits = Builder.CreateTrunc(Int64Val, Builder.getInt32Ty());
700+
701+
// Get high 32 bits by shifting right by 32 and truncating
702+
Value *ShiftedVal = Builder.CreateLShr(Int64Val, Builder.getInt64(32));
703+
Value *HighBits = Builder.CreateTrunc(ShiftedVal, Builder.getInt32Ty());
704+
705+
// Insert into our final vector
706+
Val =
707+
Builder.CreateInsertElement(Val, LowBits, Builder.getInt32(I * 2));
708+
Val = Builder.CreateInsertElement(Val, HighBits,
709+
Builder.getInt32(I * 2 + 1));
710+
}
711+
} else {
712+
// Handle scalar int64
713+
Type *Int32x2Ty = VectorType::get(Builder.getInt32Ty(), 2, false);
714+
Val = PoisonValue::get(Int32x2Ty);
715+
716+
// Get low 32 bits by truncating to i32
717+
Value *LowBits = Builder.CreateTrunc(InputVal, Builder.getInt32Ty());
718+
719+
// Get high 32 bits by shifting right by 32 and truncating
720+
Value *ShiftedVal = Builder.CreateLShr(InputVal, Builder.getInt64(32));
721+
Value *HighBits = Builder.CreateTrunc(ShiftedVal, Builder.getInt32Ty());
722+
723+
// Insert into our final vector
724+
Val = Builder.CreateInsertElement(Val, LowBits, Builder.getInt32(0));
725+
Val = Builder.CreateInsertElement(Val, HighBits, Builder.getInt32(1));
726+
}
727+
728+
Builder.CreateIntrinsic(Builder.getVoidTy(),
729+
Intrinsic::dx_resource_store_typedbuffer,
730+
{Orig->getOperand(0), Orig->getOperand(1), Val});
731+
}
621732

622-
Type *SplitElementTy = Builder.getInt32Ty();
623-
if (ExtractNum == 4)
624-
SplitElementTy = VectorType::get(SplitElementTy, 2, false);
625-
626-
// split our double(s)
627-
auto *SplitTy = llvm::StructType::get(SplitElementTy, SplitElementTy);
628-
Value *Split = Builder.CreateIntrinsic(SplitTy, Intrinsic::dx_splitdouble,
629-
Orig->getOperand(2));
630-
// create our vector
631-
Value *LowBits = Builder.CreateExtractValue(Split, 0);
632-
Value *HighBits = Builder.CreateExtractValue(Split, 1);
633-
Value *Val;
634-
if (ExtractNum == 2) {
635-
Val = PoisonValue::get(VectorType::get(SplitElementTy, 2, false));
636-
Val = Builder.CreateInsertElement(Val, LowBits, Builder.getInt32(0));
637-
Val = Builder.CreateInsertElement(Val, HighBits, Builder.getInt32(1));
638-
} else
639-
Val = Builder.CreateShuffleVector(LowBits, HighBits, {0, 2, 1, 3});
640-
641-
Builder.CreateIntrinsic(Builder.getVoidTy(),
642-
Intrinsic::dx_resource_store_typedbuffer,
643-
{Orig->getOperand(0), Orig->getOperand(1), Val});
644733
Orig->eraseFromParent();
645734
return true;
646735
}

‎llvm/test/CodeGen/DirectX/BufferLoadDouble.ll

Lines changed: 2 additions & 2 deletions
Original file line numberDiff line numberDiff line change
@@ -1,6 +1,6 @@
11
; RUN: opt -S -dxil-intrinsic-expansion %s | FileCheck %s
22

3-
target triple = "dxil-pc-shadermodel6.6-compute"
3+
target triple = "dxil-pc-shadermodel6.2-compute"
44

55
define void @loadf64() {
66
; check the handle from binding is unchanged
@@ -88,4 +88,4 @@ define void @loadf64WithCheckBit() {
8888
; CHECK-NOT: extractvalue { double, i1 }
8989
%cb = extractvalue {double, i1} %load0, 1
9090
ret void
91-
}
91+
}

‎llvm/test/CodeGen/DirectX/BufferLoadInt64.ll

Lines changed: 56 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -0,0 +1,56 @@
1+
; NOTE: Assertions have been autogenerated by utils/update_test_checks.py UTC_ARGS: --version 5
2+
; RUN: opt -S -dxil-intrinsic-expansion %s | FileCheck %s
3+
4+
target triple = "dxil-pc-shadermodel6.2-compute"
5+
6+
define { i64, i1 } @loadi64() {
7+
; CHECK-LABEL: define { i64, i1 } @loadi64() {
8+
; CHECK-NEXT: [[BUFFER:%.*]] = tail call target("dx.TypedBuffer", i64, 1, 0, 0) @llvm.dx.resource.handlefrombinding.tdx.TypedBuffer_i64_1_0_0t(i32 0, i32 0, i32 1, i32 0, i1 false, ptr null)
9+
; CHECK-NEXT: [[TMP1:%.*]] = call { <2 x i32>, i1 } @llvm.dx.resource.load.typedbuffer.v2i32.tdx.TypedBuffer_i64_1_0_0t(target("dx.TypedBuffer", i64, 1, 0, 0) [[BUFFER]], i32 0)
10+
; CHECK-NEXT: [[TMP2:%.*]] = extractvalue { <2 x i32>, i1 } [[TMP1]], 0
11+
; CHECK-NEXT: [[TMP3:%.*]] = extractelement <2 x i32> [[TMP2]], i32 0
12+
; CHECK-NEXT: [[TMP4:%.*]] = extractelement <2 x i32> [[TMP2]], i32 1
13+
; CHECK-NEXT: [[TMP5:%.*]] = zext i32 [[TMP3]] to i64
14+
; CHECK-NEXT: [[TMP6:%.*]] = zext i32 [[TMP4]] to i64
15+
; CHECK-NEXT: [[TMP7:%.*]] = shl i64 [[TMP6]], 32
16+
; CHECK-NEXT: [[TMP8:%.*]] = or i64 [[TMP5]], [[TMP7]]
17+
; CHECK-NEXT: [[TMP9:%.*]] = insertvalue { i64, i1 } poison, i64 [[TMP8]], 0
18+
; CHECK-NEXT: [[TMP10:%.*]] = extractvalue { <2 x i32>, i1 } [[TMP1]], 1
19+
; CHECK-NEXT: [[TMP11:%.*]] = insertvalue { i64, i1 } [[TMP9]], i1 [[TMP10]], 1
20+
; CHECK-NEXT: ret { i64, i1 } [[TMP11]]
21+
;
22+
%buffer = tail call target("dx.TypedBuffer", i64, 1, 0, 0) @llvm.dx.resource.handlefrombinding.tdx.TypedBuffer_i64_1_0_0t(i32 0, i32 0, i32 1, i32 0, i1 false, ptr null)
23+
%result = call { i64, i1 } @llvm.dx.resource.load.typedbuffer.tdx.TypedBuffer_i64_1_0_0t(
24+
target("dx.TypedBuffer", i64, 1, 0, 0) %buffer, i32 0)
25+
ret { i64, i1 } %result
26+
}
27+
28+
define { <2 x i64>, i1 } @loadv2i64() {
29+
; CHECK-LABEL: define { <2 x i64>, i1 } @loadv2i64() {
30+
; CHECK-NEXT: [[BUFFER:%.*]] = tail call target("dx.TypedBuffer", <2 x i64>, 1, 0, 0) @llvm.dx.resource.handlefrombinding.tdx.TypedBuffer_v2i64_1_0_0t(i32 0, i32 0, i32 1, i32 0, i1 false, ptr null)
31+
; CHECK-NEXT: [[TMP1:%.*]] = call { <4 x i32>, i1 } @llvm.dx.resource.load.typedbuffer.v4i32.tdx.TypedBuffer_v2i64_1_0_0t(target("dx.TypedBuffer", <2 x i64>, 1, 0, 0) [[BUFFER]], i32 0)
32+
; CHECK-NEXT: [[TMP2:%.*]] = extractvalue { <4 x i32>, i1 } [[TMP1]], 0
33+
; CHECK-NEXT: [[TMP3:%.*]] = extractelement <4 x i32> [[TMP2]], i32 0
34+
; CHECK-NEXT: [[TMP4:%.*]] = extractelement <4 x i32> [[TMP2]], i32 1
35+
; CHECK-NEXT: [[TMP5:%.*]] = extractelement <4 x i32> [[TMP2]], i32 2
36+
; CHECK-NEXT: [[TMP6:%.*]] = extractelement <4 x i32> [[TMP2]], i32 3
37+
; CHECK-NEXT: [[TMP7:%.*]] = zext i32 [[TMP3]] to i64
38+
; CHECK-NEXT: [[TMP8:%.*]] = zext i32 [[TMP4]] to i64
39+
; CHECK-NEXT: [[TMP9:%.*]] = shl i64 [[TMP8]], 32
40+
; CHECK-NEXT: [[TMP10:%.*]] = or i64 [[TMP7]], [[TMP9]]
41+
; CHECK-NEXT: [[TMP11:%.*]] = insertelement <2 x i64> poison, i64 [[TMP10]], i32 0
42+
; CHECK-NEXT: [[TMP12:%.*]] = zext i32 [[TMP5]] to i64
43+
; CHECK-NEXT: [[TMP13:%.*]] = zext i32 [[TMP6]] to i64
44+
; CHECK-NEXT: [[TMP14:%.*]] = shl i64 [[TMP13]], 32
45+
; CHECK-NEXT: [[TMP15:%.*]] = or i64 [[TMP12]], [[TMP14]]
46+
; CHECK-NEXT: [[TMP16:%.*]] = insertelement <2 x i64> [[TMP11]], i64 [[TMP15]], i32 1
47+
; CHECK-NEXT: [[TMP17:%.*]] = insertvalue { <2 x i64>, i1 } poison, <2 x i64> [[TMP16]], 0
48+
; CHECK-NEXT: [[TMP18:%.*]] = extractvalue { <4 x i32>, i1 } [[TMP1]], 1
49+
; CHECK-NEXT: [[TMP19:%.*]] = insertvalue { <2 x i64>, i1 } [[TMP17]], i1 [[TMP18]], 1
50+
; CHECK-NEXT: ret { <2 x i64>, i1 } [[TMP19]]
51+
;
52+
%buffer = tail call target("dx.TypedBuffer", <2 x i64>, 1, 0, 0) @llvm.dx.resource.handlefrombinding.tdx.TypedBuffer_v2i64_1_0_0t(i32 0, i32 0, i32 1, i32 0, i1 false, ptr null)
53+
%result = call { <2 x i64>, i1 } @llvm.dx.resource.load.typedbuffer.tdx.TypedBuffer_v2i64_1_0_0t(
54+
target("dx.TypedBuffer", <2 x i64>, 1, 0, 0) %buffer, i32 0)
55+
ret { <2 x i64>, i1 } %result
56+
}

‎llvm/test/CodeGen/DirectX/BufferStoreDouble.ll

Lines changed: 43 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -45,3 +45,46 @@ define void @storev2f64(<2 x double> %0) {
4545
<2 x double> %0)
4646
ret void
4747
}
48+
49+
define { double, i1 } @loadAndReturnf64() {
50+
; CHECK-LABEL: define { double, i1 } @loadAndReturnf64() {
51+
; CHECK-NEXT: [[BUFFER:%.*]] = tail call target("dx.TypedBuffer", double, 1, 0, 0) @llvm.dx.resource.handlefrombinding.tdx.TypedBuffer_f64_1_0_0t(i32 0, i32 0, i32 1, i32 0, i1 false, ptr null)
52+
; CHECK-NEXT: [[TMP1:%.*]] = call { <2 x i32>, i1 } @llvm.dx.resource.load.typedbuffer.v2i32.tdx.TypedBuffer_f64_1_0_0t(target("dx.TypedBuffer", double, 1, 0, 0) [[BUFFER]], i32 0)
53+
; CHECK-NEXT: [[TMP2:%.*]] = extractvalue { <2 x i32>, i1 } [[TMP1]], 0
54+
; CHECK-NEXT: [[TMP3:%.*]] = extractelement <2 x i32> [[TMP2]], i32 0
55+
; CHECK-NEXT: [[TMP4:%.*]] = extractelement <2 x i32> [[TMP2]], i32 1
56+
; CHECK-NEXT: [[TMP5:%.*]] = call double @llvm.dx.asdouble.i32(i32 [[TMP3]], i32 [[TMP4]])
57+
; CHECK-NEXT: [[TMP6:%.*]] = insertvalue { double, i1 } poison, double [[TMP5]], 0
58+
; CHECK-NEXT: [[TMP7:%.*]] = extractvalue { <2 x i32>, i1 } [[TMP1]], 1
59+
; CHECK-NEXT: [[TMP8:%.*]] = insertvalue { double, i1 } [[TMP6]], i1 [[TMP7]], 1
60+
; CHECK-NEXT: ret { double, i1 } [[TMP8]]
61+
;
62+
%buffer = tail call target("dx.TypedBuffer", double, 1, 0, 0) @llvm.dx.resource.handlefrombinding.tdx.TypedBuffer_f64_1_0_0t(i32 0, i32 0, i32 1, i32 0, i1 false, ptr null)
63+
%result = call { double, i1 } @llvm.dx.resource.load.typedbuffer.tdx.TypedBuffer_f64_1_0_0t(
64+
target("dx.TypedBuffer", double, 1, 0, 0) %buffer, i32 0)
65+
ret { double, i1 } %result
66+
}
67+
68+
define { <2 x double>, i1 } @loadAndReturnv2f64() {
69+
; CHECK-LABEL: define { <2 x double>, i1 } @loadAndReturnv2f64() {
70+
; CHECK-NEXT: [[BUFFER:%.*]] = tail call target("dx.TypedBuffer", <2 x double>, 1, 0, 0) @llvm.dx.resource.handlefrombinding.tdx.TypedBuffer_v2f64_1_0_0t(i32 0, i32 0, i32 1, i32 0, i1 false, ptr null)
71+
; CHECK-NEXT: [[TMP1:%.*]] = call { <4 x i32>, i1 } @llvm.dx.resource.load.typedbuffer.v4i32.tdx.TypedBuffer_v2f64_1_0_0t(target("dx.TypedBuffer", <2 x double>, 1, 0, 0) [[BUFFER]], i32 0)
72+
; CHECK-NEXT: [[TMP2:%.*]] = extractvalue { <4 x i32>, i1 } [[TMP1]], 0
73+
; CHECK-NEXT: [[TMP3:%.*]] = extractelement <4 x i32> [[TMP2]], i32 0
74+
; CHECK-NEXT: [[TMP4:%.*]] = extractelement <4 x i32> [[TMP2]], i32 1
75+
; CHECK-NEXT: [[TMP5:%.*]] = extractelement <4 x i32> [[TMP2]], i32 2
76+
; CHECK-NEXT: [[TMP6:%.*]] = extractelement <4 x i32> [[TMP2]], i32 3
77+
; CHECK-NEXT: [[TMP7:%.*]] = call double @llvm.dx.asdouble.i32(i32 [[TMP3]], i32 [[TMP4]])
78+
; CHECK-NEXT: [[TMP8:%.*]] = insertelement <2 x double> poison, double [[TMP7]], i32 0
79+
; CHECK-NEXT: [[TMP9:%.*]] = call double @llvm.dx.asdouble.i32(i32 [[TMP5]], i32 [[TMP6]])
80+
; CHECK-NEXT: [[TMP10:%.*]] = insertelement <2 x double> [[TMP8]], double [[TMP9]], i32 1
81+
; CHECK-NEXT: [[TMP11:%.*]] = insertvalue { <2 x double>, i1 } poison, <2 x double> [[TMP10]], 0
82+
; CHECK-NEXT: [[TMP12:%.*]] = extractvalue { <4 x i32>, i1 } [[TMP1]], 1
83+
; CHECK-NEXT: [[TMP13:%.*]] = insertvalue { <2 x double>, i1 } [[TMP11]], i1 [[TMP12]], 1
84+
; CHECK-NEXT: ret { <2 x double>, i1 } [[TMP13]]
85+
;
86+
%buffer = tail call target("dx.TypedBuffer", <2 x double>, 1, 0, 0) @llvm.dx.resource.handlefrombinding.tdx.TypedBuffer_v2f64_1_0_0t(i32 0, i32 0, i32 1, i32 0, i1 false, ptr null)
87+
%result = call { <2 x double>, i1 } @llvm.dx.resource.load.typedbuffer.tdx.TypedBuffer_v2f64_1_0_0t(
88+
target("dx.TypedBuffer", <2 x double>, 1, 0, 0) %buffer, i32 0)
89+
ret { <2 x double>, i1 } %result
90+
}

0 commit comments

Comments
 (0)