summaryrefslogtreecommitdiff
path: root/lib/CodeGen/CGBuiltin.cpp
diff options
context:
space:
mode:
Diffstat (limited to 'lib/CodeGen/CGBuiltin.cpp')
-rw-r--r--lib/CodeGen/CGBuiltin.cpp212
1 files changed, 21 insertions, 191 deletions
diff --git a/lib/CodeGen/CGBuiltin.cpp b/lib/CodeGen/CGBuiltin.cpp
index 2a6e92e7f3ce9..bc902507c46e5 100644
--- a/lib/CodeGen/CGBuiltin.cpp
+++ b/lib/CodeGen/CGBuiltin.cpp
@@ -1810,12 +1810,12 @@ RValue CodeGenFunction::EmitBuiltinExpr(const FunctionDecl *FD,
case Builtin::BI__atomic_signal_fence:
case Builtin::BI__c11_atomic_thread_fence:
case Builtin::BI__c11_atomic_signal_fence: {
- llvm::SynchronizationScope Scope;
+ llvm::SyncScope::ID SSID;
if (BuiltinID == Builtin::BI__atomic_signal_fence ||
BuiltinID == Builtin::BI__c11_atomic_signal_fence)
- Scope = llvm::SingleThread;
+ SSID = llvm::SyncScope::SingleThread;
else
- Scope = llvm::CrossThread;
+ SSID = llvm::SyncScope::System;
Value *Order = EmitScalarExpr(E->getArg(0));
if (isa<llvm::ConstantInt>(Order)) {
int ord = cast<llvm::ConstantInt>(Order)->getZExtValue();
@@ -1825,17 +1825,16 @@ RValue CodeGenFunction::EmitBuiltinExpr(const FunctionDecl *FD,
break;
case 1: // memory_order_consume
case 2: // memory_order_acquire
- Builder.CreateFence(llvm::AtomicOrdering::Acquire, Scope);
+ Builder.CreateFence(llvm::AtomicOrdering::Acquire, SSID);
break;
case 3: // memory_order_release
- Builder.CreateFence(llvm::AtomicOrdering::Release, Scope);
+ Builder.CreateFence(llvm::AtomicOrdering::Release, SSID);
break;
case 4: // memory_order_acq_rel
- Builder.CreateFence(llvm::AtomicOrdering::AcquireRelease, Scope);
+ Builder.CreateFence(llvm::AtomicOrdering::AcquireRelease, SSID);
break;
case 5: // memory_order_seq_cst
- Builder.CreateFence(llvm::AtomicOrdering::SequentiallyConsistent,
- Scope);
+ Builder.CreateFence(llvm::AtomicOrdering::SequentiallyConsistent, SSID);
break;
}
return RValue::get(nullptr);
@@ -1852,23 +1851,23 @@ RValue CodeGenFunction::EmitBuiltinExpr(const FunctionDecl *FD,
llvm::SwitchInst *SI = Builder.CreateSwitch(Order, ContBB);
Builder.SetInsertPoint(AcquireBB);
- Builder.CreateFence(llvm::AtomicOrdering::Acquire, Scope);
+ Builder.CreateFence(llvm::AtomicOrdering::Acquire, SSID);
Builder.CreateBr(ContBB);
SI->addCase(Builder.getInt32(1), AcquireBB);
SI->addCase(Builder.getInt32(2), AcquireBB);
Builder.SetInsertPoint(ReleaseBB);
- Builder.CreateFence(llvm::AtomicOrdering::Release, Scope);
+ Builder.CreateFence(llvm::AtomicOrdering::Release, SSID);
Builder.CreateBr(ContBB);
SI->addCase(Builder.getInt32(3), ReleaseBB);
Builder.SetInsertPoint(AcqRelBB);
- Builder.CreateFence(llvm::AtomicOrdering::AcquireRelease, Scope);
+ Builder.CreateFence(llvm::AtomicOrdering::AcquireRelease, SSID);
Builder.CreateBr(ContBB);
SI->addCase(Builder.getInt32(4), AcqRelBB);
Builder.SetInsertPoint(SeqCstBB);
- Builder.CreateFence(llvm::AtomicOrdering::SequentiallyConsistent, Scope);
+ Builder.CreateFence(llvm::AtomicOrdering::SequentiallyConsistent, SSID);
Builder.CreateBr(ContBB);
SI->addCase(Builder.getInt32(5), SeqCstBB);
@@ -2956,9 +2955,8 @@ static llvm::VectorType *GetNeonType(CodeGenFunction *CGF,
return llvm::VectorType::get(CGF->Int8Ty, V1Ty ? 1 : (8 << IsQuad));
case NeonTypeFlags::Int16:
case NeonTypeFlags::Poly16:
- return llvm::VectorType::get(CGF->Int16Ty, V1Ty ? 1 : (4 << IsQuad));
case NeonTypeFlags::Float16:
- return llvm::VectorType::get(CGF->HalfTy, V1Ty ? 1 : (4 << IsQuad));
+ return llvm::VectorType::get(CGF->Int16Ty, V1Ty ? 1 : (4 << IsQuad));
case NeonTypeFlags::Int32:
return llvm::VectorType::get(CGF->Int32Ty, V1Ty ? 1 : (2 << IsQuad));
case NeonTypeFlags::Int64:
@@ -2981,8 +2979,6 @@ static llvm::VectorType *GetFloatNeonType(CodeGenFunction *CGF,
NeonTypeFlags IntTypeFlags) {
int IsQuad = IntTypeFlags.isQuad();
switch (IntTypeFlags.getEltType()) {
- case NeonTypeFlags::Int16:
- return llvm::VectorType::get(CGF->HalfTy, (4 << IsQuad));
case NeonTypeFlags::Int32:
return llvm::VectorType::get(CGF->FloatTy, (2 << IsQuad));
case NeonTypeFlags::Int64:
@@ -3130,80 +3126,55 @@ static const NeonIntrinsicInfo ARMSIMDIntrinsicMap [] = {
NEONMAP1(vcvt_f16_f32, arm_neon_vcvtfp2hf, 0),
NEONMAP1(vcvt_f32_f16, arm_neon_vcvthf2fp, 0),
NEONMAP0(vcvt_f32_v),
- NEONMAP2(vcvt_n_f16_v, arm_neon_vcvtfxu2fp, arm_neon_vcvtfxs2fp, 0),
NEONMAP2(vcvt_n_f32_v, arm_neon_vcvtfxu2fp, arm_neon_vcvtfxs2fp, 0),
- NEONMAP1(vcvt_n_s16_v, arm_neon_vcvtfp2fxs, 0),
NEONMAP1(vcvt_n_s32_v, arm_neon_vcvtfp2fxs, 0),
NEONMAP1(vcvt_n_s64_v, arm_neon_vcvtfp2fxs, 0),
- NEONMAP1(vcvt_n_u16_v, arm_neon_vcvtfp2fxu, 0),
NEONMAP1(vcvt_n_u32_v, arm_neon_vcvtfp2fxu, 0),
NEONMAP1(vcvt_n_u64_v, arm_neon_vcvtfp2fxu, 0),
- NEONMAP0(vcvt_s16_v),
NEONMAP0(vcvt_s32_v),
NEONMAP0(vcvt_s64_v),
- NEONMAP0(vcvt_u16_v),
NEONMAP0(vcvt_u32_v),
NEONMAP0(vcvt_u64_v),
- NEONMAP1(vcvta_s16_v, arm_neon_vcvtas, 0),
NEONMAP1(vcvta_s32_v, arm_neon_vcvtas, 0),
NEONMAP1(vcvta_s64_v, arm_neon_vcvtas, 0),
NEONMAP1(vcvta_u32_v, arm_neon_vcvtau, 0),
NEONMAP1(vcvta_u64_v, arm_neon_vcvtau, 0),
- NEONMAP1(vcvtaq_s16_v, arm_neon_vcvtas, 0),
NEONMAP1(vcvtaq_s32_v, arm_neon_vcvtas, 0),
NEONMAP1(vcvtaq_s64_v, arm_neon_vcvtas, 0),
- NEONMAP1(vcvtaq_u16_v, arm_neon_vcvtau, 0),
NEONMAP1(vcvtaq_u32_v, arm_neon_vcvtau, 0),
NEONMAP1(vcvtaq_u64_v, arm_neon_vcvtau, 0),
- NEONMAP1(vcvtm_s16_v, arm_neon_vcvtms, 0),
NEONMAP1(vcvtm_s32_v, arm_neon_vcvtms, 0),
NEONMAP1(vcvtm_s64_v, arm_neon_vcvtms, 0),
- NEONMAP1(vcvtm_u16_v, arm_neon_vcvtmu, 0),
NEONMAP1(vcvtm_u32_v, arm_neon_vcvtmu, 0),
NEONMAP1(vcvtm_u64_v, arm_neon_vcvtmu, 0),
- NEONMAP1(vcvtmq_s16_v, arm_neon_vcvtms, 0),
NEONMAP1(vcvtmq_s32_v, arm_neon_vcvtms, 0),
NEONMAP1(vcvtmq_s64_v, arm_neon_vcvtms, 0),
- NEONMAP1(vcvtmq_u16_v, arm_neon_vcvtmu, 0),
NEONMAP1(vcvtmq_u32_v, arm_neon_vcvtmu, 0),
NEONMAP1(vcvtmq_u64_v, arm_neon_vcvtmu, 0),
- NEONMAP1(vcvtn_s16_v, arm_neon_vcvtns, 0),
NEONMAP1(vcvtn_s32_v, arm_neon_vcvtns, 0),
NEONMAP1(vcvtn_s64_v, arm_neon_vcvtns, 0),
- NEONMAP1(vcvtn_u16_v, arm_neon_vcvtnu, 0),
NEONMAP1(vcvtn_u32_v, arm_neon_vcvtnu, 0),
NEONMAP1(vcvtn_u64_v, arm_neon_vcvtnu, 0),
- NEONMAP1(vcvtnq_s16_v, arm_neon_vcvtns, 0),
NEONMAP1(vcvtnq_s32_v, arm_neon_vcvtns, 0),
NEONMAP1(vcvtnq_s64_v, arm_neon_vcvtns, 0),
- NEONMAP1(vcvtnq_u16_v, arm_neon_vcvtnu, 0),
NEONMAP1(vcvtnq_u32_v, arm_neon_vcvtnu, 0),
NEONMAP1(vcvtnq_u64_v, arm_neon_vcvtnu, 0),
- NEONMAP1(vcvtp_s16_v, arm_neon_vcvtps, 0),
NEONMAP1(vcvtp_s32_v, arm_neon_vcvtps, 0),
NEONMAP1(vcvtp_s64_v, arm_neon_vcvtps, 0),
- NEONMAP1(vcvtp_u16_v, arm_neon_vcvtpu, 0),
NEONMAP1(vcvtp_u32_v, arm_neon_vcvtpu, 0),
NEONMAP1(vcvtp_u64_v, arm_neon_vcvtpu, 0),
- NEONMAP1(vcvtpq_s16_v, arm_neon_vcvtps, 0),
NEONMAP1(vcvtpq_s32_v, arm_neon_vcvtps, 0),
NEONMAP1(vcvtpq_s64_v, arm_neon_vcvtps, 0),
- NEONMAP1(vcvtpq_u16_v, arm_neon_vcvtpu, 0),
NEONMAP1(vcvtpq_u32_v, arm_neon_vcvtpu, 0),
NEONMAP1(vcvtpq_u64_v, arm_neon_vcvtpu, 0),
NEONMAP0(vcvtq_f32_v),
- NEONMAP2(vcvtq_n_f16_v, arm_neon_vcvtfxu2fp, arm_neon_vcvtfxs2fp, 0),
NEONMAP2(vcvtq_n_f32_v, arm_neon_vcvtfxu2fp, arm_neon_vcvtfxs2fp, 0),
- NEONMAP1(vcvtq_n_s16_v, arm_neon_vcvtfp2fxs, 0),
NEONMAP1(vcvtq_n_s32_v, arm_neon_vcvtfp2fxs, 0),
NEONMAP1(vcvtq_n_s64_v, arm_neon_vcvtfp2fxs, 0),
- NEONMAP1(vcvtq_n_u16_v, arm_neon_vcvtfp2fxu, 0),
NEONMAP1(vcvtq_n_u32_v, arm_neon_vcvtfp2fxu, 0),
NEONMAP1(vcvtq_n_u64_v, arm_neon_vcvtfp2fxu, 0),
- NEONMAP0(vcvtq_s16_v),
NEONMAP0(vcvtq_s32_v),
NEONMAP0(vcvtq_s64_v),
- NEONMAP0(vcvtq_u16_v),
NEONMAP0(vcvtq_u32_v),
NEONMAP0(vcvtq_u64_v),
NEONMAP0(vext_v),
@@ -3366,27 +3337,19 @@ static const NeonIntrinsicInfo AArch64SIMDIntrinsicMap[] = {
NEONMAP1(vcnt_v, ctpop, Add1ArgType),
NEONMAP1(vcntq_v, ctpop, Add1ArgType),
NEONMAP1(vcvt_f16_f32, aarch64_neon_vcvtfp2hf, 0),
- NEONMAP0(vcvt_f16_v),
NEONMAP1(vcvt_f32_f16, aarch64_neon_vcvthf2fp, 0),
NEONMAP0(vcvt_f32_v),
- NEONMAP2(vcvt_n_f16_v, aarch64_neon_vcvtfxu2fp, aarch64_neon_vcvtfxs2fp, 0),
NEONMAP2(vcvt_n_f32_v, aarch64_neon_vcvtfxu2fp, aarch64_neon_vcvtfxs2fp, 0),
NEONMAP2(vcvt_n_f64_v, aarch64_neon_vcvtfxu2fp, aarch64_neon_vcvtfxs2fp, 0),
- NEONMAP1(vcvt_n_s16_v, aarch64_neon_vcvtfp2fxs, 0),
NEONMAP1(vcvt_n_s32_v, aarch64_neon_vcvtfp2fxs, 0),
NEONMAP1(vcvt_n_s64_v, aarch64_neon_vcvtfp2fxs, 0),
- NEONMAP1(vcvt_n_u16_v, aarch64_neon_vcvtfp2fxu, 0),
NEONMAP1(vcvt_n_u32_v, aarch64_neon_vcvtfp2fxu, 0),
NEONMAP1(vcvt_n_u64_v, aarch64_neon_vcvtfp2fxu, 0),
- NEONMAP0(vcvtq_f16_v),
NEONMAP0(vcvtq_f32_v),
- NEONMAP2(vcvtq_n_f16_v, aarch64_neon_vcvtfxu2fp, aarch64_neon_vcvtfxs2fp, 0),
NEONMAP2(vcvtq_n_f32_v, aarch64_neon_vcvtfxu2fp, aarch64_neon_vcvtfxs2fp, 0),
NEONMAP2(vcvtq_n_f64_v, aarch64_neon_vcvtfxu2fp, aarch64_neon_vcvtfxs2fp, 0),
- NEONMAP1(vcvtq_n_s16_v, aarch64_neon_vcvtfp2fxs, 0),
NEONMAP1(vcvtq_n_s32_v, aarch64_neon_vcvtfp2fxs, 0),
NEONMAP1(vcvtq_n_s64_v, aarch64_neon_vcvtfp2fxs, 0),
- NEONMAP1(vcvtq_n_u16_v, aarch64_neon_vcvtfp2fxu, 0),
NEONMAP1(vcvtq_n_u32_v, aarch64_neon_vcvtfp2fxu, 0),
NEONMAP1(vcvtq_n_u64_v, aarch64_neon_vcvtfp2fxu, 0),
NEONMAP1(vcvtx_f32_v, aarch64_neon_fcvtxn, AddRetType | Add1ArgType),
@@ -3855,20 +3818,9 @@ Value *CodeGenFunction::EmitCommonNeonBuiltinExpr(
case NEON::BI__builtin_neon_vcageq_v:
case NEON::BI__builtin_neon_vcagt_v:
case NEON::BI__builtin_neon_vcagtq_v: {
- llvm::Type *Ty;
- switch (VTy->getScalarSizeInBits()) {
- default: llvm_unreachable("unexpected type");
- case 32:
- Ty = FloatTy;
- break;
- case 64:
- Ty = DoubleTy;
- break;
- case 16:
- Ty = HalfTy;
- break;
- }
- llvm::Type *VecFlt = llvm::VectorType::get(Ty, VTy->getNumElements());
+ llvm::Type *VecFlt = llvm::VectorType::get(
+ VTy->getScalarSizeInBits() == 32 ? FloatTy : DoubleTy,
+ VTy->getNumElements());
llvm::Type *Tys[] = { VTy, VecFlt };
Function *F = CGM.getIntrinsic(LLVMIntrinsic, Tys);
return EmitNeonCall(F, Ops, NameHint);
@@ -3885,16 +3837,8 @@ Value *CodeGenFunction::EmitCommonNeonBuiltinExpr(
Ty = GetNeonType(this, NeonTypeFlags(NeonTypeFlags::Float32, false, Quad));
return Usgn ? Builder.CreateUIToFP(Ops[0], Ty, "vcvt")
: Builder.CreateSIToFP(Ops[0], Ty, "vcvt");
- case NEON::BI__builtin_neon_vcvt_f16_v:
- case NEON::BI__builtin_neon_vcvtq_f16_v:
- Ops[0] = Builder.CreateBitCast(Ops[0], Ty);
- Ty = GetNeonType(this, NeonTypeFlags(NeonTypeFlags::Float16, false, Quad));
- return Usgn ? Builder.CreateUIToFP(Ops[0], Ty, "vcvt")
- : Builder.CreateSIToFP(Ops[0], Ty, "vcvt");
- case NEON::BI__builtin_neon_vcvt_n_f16_v:
case NEON::BI__builtin_neon_vcvt_n_f32_v:
case NEON::BI__builtin_neon_vcvt_n_f64_v:
- case NEON::BI__builtin_neon_vcvtq_n_f16_v:
case NEON::BI__builtin_neon_vcvtq_n_f32_v:
case NEON::BI__builtin_neon_vcvtq_n_f64_v: {
llvm::Type *Tys[2] = { GetFloatNeonType(this, Type), Ty };
@@ -3902,15 +3846,11 @@ Value *CodeGenFunction::EmitCommonNeonBuiltinExpr(
Function *F = CGM.getIntrinsic(Int, Tys);
return EmitNeonCall(F, Ops, "vcvt_n");
}
- case NEON::BI__builtin_neon_vcvt_n_s16_v:
case NEON::BI__builtin_neon_vcvt_n_s32_v:
- case NEON::BI__builtin_neon_vcvt_n_u16_v:
case NEON::BI__builtin_neon_vcvt_n_u32_v:
case NEON::BI__builtin_neon_vcvt_n_s64_v:
case NEON::BI__builtin_neon_vcvt_n_u64_v:
- case NEON::BI__builtin_neon_vcvtq_n_s16_v:
case NEON::BI__builtin_neon_vcvtq_n_s32_v:
- case NEON::BI__builtin_neon_vcvtq_n_u16_v:
case NEON::BI__builtin_neon_vcvtq_n_u32_v:
case NEON::BI__builtin_neon_vcvtq_n_s64_v:
case NEON::BI__builtin_neon_vcvtq_n_u64_v: {
@@ -3922,63 +3862,44 @@ Value *CodeGenFunction::EmitCommonNeonBuiltinExpr(
case NEON::BI__builtin_neon_vcvt_u32_v:
case NEON::BI__builtin_neon_vcvt_s64_v:
case NEON::BI__builtin_neon_vcvt_u64_v:
- case NEON::BI__builtin_neon_vcvt_s16_v:
- case NEON::BI__builtin_neon_vcvt_u16_v:
case NEON::BI__builtin_neon_vcvtq_s32_v:
case NEON::BI__builtin_neon_vcvtq_u32_v:
case NEON::BI__builtin_neon_vcvtq_s64_v:
- case NEON::BI__builtin_neon_vcvtq_u64_v:
- case NEON::BI__builtin_neon_vcvtq_s16_v:
- case NEON::BI__builtin_neon_vcvtq_u16_v: {
+ case NEON::BI__builtin_neon_vcvtq_u64_v: {
Ops[0] = Builder.CreateBitCast(Ops[0], GetFloatNeonType(this, Type));
return Usgn ? Builder.CreateFPToUI(Ops[0], Ty, "vcvt")
: Builder.CreateFPToSI(Ops[0], Ty, "vcvt");
}
- case NEON::BI__builtin_neon_vcvta_s16_v:
case NEON::BI__builtin_neon_vcvta_s32_v:
case NEON::BI__builtin_neon_vcvta_s64_v:
case NEON::BI__builtin_neon_vcvta_u32_v:
case NEON::BI__builtin_neon_vcvta_u64_v:
- case NEON::BI__builtin_neon_vcvtaq_s16_v:
case NEON::BI__builtin_neon_vcvtaq_s32_v:
case NEON::BI__builtin_neon_vcvtaq_s64_v:
- case NEON::BI__builtin_neon_vcvtaq_u16_v:
case NEON::BI__builtin_neon_vcvtaq_u32_v:
case NEON::BI__builtin_neon_vcvtaq_u64_v:
- case NEON::BI__builtin_neon_vcvtn_s16_v:
case NEON::BI__builtin_neon_vcvtn_s32_v:
case NEON::BI__builtin_neon_vcvtn_s64_v:
- case NEON::BI__builtin_neon_vcvtn_u16_v:
case NEON::BI__builtin_neon_vcvtn_u32_v:
case NEON::BI__builtin_neon_vcvtn_u64_v:
- case NEON::BI__builtin_neon_vcvtnq_s16_v:
case NEON::BI__builtin_neon_vcvtnq_s32_v:
case NEON::BI__builtin_neon_vcvtnq_s64_v:
- case NEON::BI__builtin_neon_vcvtnq_u16_v:
case NEON::BI__builtin_neon_vcvtnq_u32_v:
case NEON::BI__builtin_neon_vcvtnq_u64_v:
- case NEON::BI__builtin_neon_vcvtp_s16_v:
case NEON::BI__builtin_neon_vcvtp_s32_v:
case NEON::BI__builtin_neon_vcvtp_s64_v:
- case NEON::BI__builtin_neon_vcvtp_u16_v:
case NEON::BI__builtin_neon_vcvtp_u32_v:
case NEON::BI__builtin_neon_vcvtp_u64_v:
- case NEON::BI__builtin_neon_vcvtpq_s16_v:
case NEON::BI__builtin_neon_vcvtpq_s32_v:
case NEON::BI__builtin_neon_vcvtpq_s64_v:
- case NEON::BI__builtin_neon_vcvtpq_u16_v:
case NEON::BI__builtin_neon_vcvtpq_u32_v:
case NEON::BI__builtin_neon_vcvtpq_u64_v:
- case NEON::BI__builtin_neon_vcvtm_s16_v:
case NEON::BI__builtin_neon_vcvtm_s32_v:
case NEON::BI__builtin_neon_vcvtm_s64_v:
- case NEON::BI__builtin_neon_vcvtm_u16_v:
case NEON::BI__builtin_neon_vcvtm_u32_v:
case NEON::BI__builtin_neon_vcvtm_u64_v:
- case NEON::BI__builtin_neon_vcvtmq_s16_v:
case NEON::BI__builtin_neon_vcvtmq_s32_v:
case NEON::BI__builtin_neon_vcvtmq_s64_v:
- case NEON::BI__builtin_neon_vcvtmq_u16_v:
case NEON::BI__builtin_neon_vcvtmq_u32_v:
case NEON::BI__builtin_neon_vcvtmq_u64_v: {
llvm::Type *Tys[2] = { Ty, GetFloatNeonType(this, Type) };
@@ -6188,9 +6109,7 @@ Value *CodeGenFunction::EmitAArch64BuiltinExpr(unsigned BuiltinID,
Ops[2] = EmitNeonSplat(Ops[2], cast<ConstantInt>(Ops[3]));
return Builder.CreateCall(F, {Ops[2], Ops[1], Ops[0]});
}
- case NEON::BI__builtin_neon_vfmah_lane_f16:
case NEON::BI__builtin_neon_vfmas_lane_f32:
- case NEON::BI__builtin_neon_vfmah_laneq_f16:
case NEON::BI__builtin_neon_vfmas_laneq_f32:
case NEON::BI__builtin_neon_vfmad_lane_f64:
case NEON::BI__builtin_neon_vfmad_laneq_f64: {
@@ -6365,25 +6284,18 @@ Value *CodeGenFunction::EmitAArch64BuiltinExpr(unsigned BuiltinID,
case NEON::BI__builtin_neon_vcvt_u32_v:
case NEON::BI__builtin_neon_vcvt_s64_v:
case NEON::BI__builtin_neon_vcvt_u64_v:
- case NEON::BI__builtin_neon_vcvt_s16_v:
- case NEON::BI__builtin_neon_vcvt_u16_v:
case NEON::BI__builtin_neon_vcvtq_s32_v:
case NEON::BI__builtin_neon_vcvtq_u32_v:
case NEON::BI__builtin_neon_vcvtq_s64_v:
- case NEON::BI__builtin_neon_vcvtq_u64_v:
- case NEON::BI__builtin_neon_vcvtq_s16_v:
- case NEON::BI__builtin_neon_vcvtq_u16_v: {
+ case NEON::BI__builtin_neon_vcvtq_u64_v: {
Ops[0] = Builder.CreateBitCast(Ops[0], GetFloatNeonType(this, Type));
if (usgn)
return Builder.CreateFPToUI(Ops[0], Ty);
return Builder.CreateFPToSI(Ops[0], Ty);
}
- case NEON::BI__builtin_neon_vcvta_s16_v:
case NEON::BI__builtin_neon_vcvta_s32_v:
- case NEON::BI__builtin_neon_vcvtaq_s16_v:
case NEON::BI__builtin_neon_vcvtaq_s32_v:
case NEON::BI__builtin_neon_vcvta_u32_v:
- case NEON::BI__builtin_neon_vcvtaq_u16_v:
case NEON::BI__builtin_neon_vcvtaq_u32_v:
case NEON::BI__builtin_neon_vcvta_s64_v:
case NEON::BI__builtin_neon_vcvtaq_s64_v:
@@ -6393,13 +6305,9 @@ Value *CodeGenFunction::EmitAArch64BuiltinExpr(unsigned BuiltinID,
llvm::Type *Tys[2] = { Ty, GetFloatNeonType(this, Type) };
return EmitNeonCall(CGM.getIntrinsic(Int, Tys), Ops, "vcvta");
}
- case NEON::BI__builtin_neon_vcvtm_s16_v:
case NEON::BI__builtin_neon_vcvtm_s32_v:
- case NEON::BI__builtin_neon_vcvtmq_s16_v:
case NEON::BI__builtin_neon_vcvtmq_s32_v:
- case NEON::BI__builtin_neon_vcvtm_u16_v:
case NEON::BI__builtin_neon_vcvtm_u32_v:
- case NEON::BI__builtin_neon_vcvtmq_u16_v:
case NEON::BI__builtin_neon_vcvtmq_u32_v:
case NEON::BI__builtin_neon_vcvtm_s64_v:
case NEON::BI__builtin_neon_vcvtmq_s64_v:
@@ -6409,13 +6317,9 @@ Value *CodeGenFunction::EmitAArch64BuiltinExpr(unsigned BuiltinID,
llvm::Type *Tys[2] = { Ty, GetFloatNeonType(this, Type) };
return EmitNeonCall(CGM.getIntrinsic(Int, Tys), Ops, "vcvtm");
}
- case NEON::BI__builtin_neon_vcvtn_s16_v:
case NEON::BI__builtin_neon_vcvtn_s32_v:
- case NEON::BI__builtin_neon_vcvtnq_s16_v:
case NEON::BI__builtin_neon_vcvtnq_s32_v:
- case NEON::BI__builtin_neon_vcvtn_u16_v:
case NEON::BI__builtin_neon_vcvtn_u32_v:
- case NEON::BI__builtin_neon_vcvtnq_u16_v:
case NEON::BI__builtin_neon_vcvtnq_u32_v:
case NEON::BI__builtin_neon_vcvtn_s64_v:
case NEON::BI__builtin_neon_vcvtnq_s64_v:
@@ -6425,13 +6329,9 @@ Value *CodeGenFunction::EmitAArch64BuiltinExpr(unsigned BuiltinID,
llvm::Type *Tys[2] = { Ty, GetFloatNeonType(this, Type) };
return EmitNeonCall(CGM.getIntrinsic(Int, Tys), Ops, "vcvtn");
}
- case NEON::BI__builtin_neon_vcvtp_s16_v:
case NEON::BI__builtin_neon_vcvtp_s32_v:
- case NEON::BI__builtin_neon_vcvtpq_s16_v:
case NEON::BI__builtin_neon_vcvtpq_s32_v:
- case NEON::BI__builtin_neon_vcvtp_u16_v:
case NEON::BI__builtin_neon_vcvtp_u32_v:
- case NEON::BI__builtin_neon_vcvtpq_u16_v:
case NEON::BI__builtin_neon_vcvtpq_u32_v:
case NEON::BI__builtin_neon_vcvtp_s64_v:
case NEON::BI__builtin_neon_vcvtpq_s64_v:
@@ -6604,24 +6504,6 @@ Value *CodeGenFunction::EmitAArch64BuiltinExpr(unsigned BuiltinID,
Ops[0] = EmitNeonCall(CGM.getIntrinsic(Int, Tys), Ops, "vmaxv");
return Builder.CreateTrunc(Ops[0], Int16Ty);
}
- case NEON::BI__builtin_neon_vmaxv_f16: {
- Int = Intrinsic::aarch64_neon_fmaxv;
- Ty = HalfTy;
- VTy = llvm::VectorType::get(HalfTy, 4);
- llvm::Type *Tys[2] = { Ty, VTy };
- Ops.push_back(EmitScalarExpr(E->getArg(0)));
- Ops[0] = EmitNeonCall(CGM.getIntrinsic(Int, Tys), Ops, "vmaxv");
- return Builder.CreateTrunc(Ops[0], HalfTy);
- }
- case NEON::BI__builtin_neon_vmaxvq_f16: {
- Int = Intrinsic::aarch64_neon_fmaxv;
- Ty = HalfTy;
- VTy = llvm::VectorType::get(HalfTy, 8);
- llvm::Type *Tys[2] = { Ty, VTy };
- Ops.push_back(EmitScalarExpr(E->getArg(0)));
- Ops[0] = EmitNeonCall(CGM.getIntrinsic(Int, Tys), Ops, "vmaxv");
- return Builder.CreateTrunc(Ops[0], HalfTy);
- }
case NEON::BI__builtin_neon_vminv_u8: {
Int = Intrinsic::aarch64_neon_uminv;
Ty = Int32Ty;
@@ -6694,60 +6576,6 @@ Value *CodeGenFunction::EmitAArch64BuiltinExpr(unsigned BuiltinID,
Ops[0] = EmitNeonCall(CGM.getIntrinsic(Int, Tys), Ops, "vminv");
return Builder.CreateTrunc(Ops[0], Int16Ty);
}
- case NEON::BI__builtin_neon_vminv_f16: {
- Int = Intrinsic::aarch64_neon_fminv;
- Ty = HalfTy;
- VTy = llvm::VectorType::get(HalfTy, 4);
- llvm::Type *Tys[2] = { Ty, VTy };
- Ops.push_back(EmitScalarExpr(E->getArg(0)));
- Ops[0] = EmitNeonCall(CGM.getIntrinsic(Int, Tys), Ops, "vminv");
- return Builder.CreateTrunc(Ops[0], HalfTy);
- }
- case NEON::BI__builtin_neon_vminvq_f16: {
- Int = Intrinsic::aarch64_neon_fminv;
- Ty = HalfTy;
- VTy = llvm::VectorType::get(HalfTy, 8);
- llvm::Type *Tys[2] = { Ty, VTy };
- Ops.push_back(EmitScalarExpr(E->getArg(0)));
- Ops[0] = EmitNeonCall(CGM.getIntrinsic(Int, Tys), Ops, "vminv");
- return Builder.CreateTrunc(Ops[0], HalfTy);
- }
- case NEON::BI__builtin_neon_vmaxnmv_f16: {
- Int = Intrinsic::aarch64_neon_fmaxnmv;
- Ty = HalfTy;
- VTy = llvm::VectorType::get(HalfTy, 4);
- llvm::Type *Tys[2] = { Ty, VTy };
- Ops.push_back(EmitScalarExpr(E->getArg(0)));
- Ops[0] = EmitNeonCall(CGM.getIntrinsic(Int, Tys), Ops, "vmaxnmv");
- return Builder.CreateTrunc(Ops[0], HalfTy);
- }
- case NEON::BI__builtin_neon_vmaxnmvq_f16: {
- Int = Intrinsic::aarch64_neon_fmaxnmv;
- Ty = HalfTy;
- VTy = llvm::VectorType::get(HalfTy, 8);
- llvm::Type *Tys[2] = { Ty, VTy };
- Ops.push_back(EmitScalarExpr(E->getArg(0)));
- Ops[0] = EmitNeonCall(CGM.getIntrinsic(Int, Tys), Ops, "vmaxnmv");
- return Builder.CreateTrunc(Ops[0], HalfTy);
- }
- case NEON::BI__builtin_neon_vminnmv_f16: {
- Int = Intrinsic::aarch64_neon_fminnmv;
- Ty = HalfTy;
- VTy = llvm::VectorType::get(HalfTy, 4);
- llvm::Type *Tys[2] = { Ty, VTy };
- Ops.push_back(EmitScalarExpr(E->getArg(0)));
- Ops[0] = EmitNeonCall(CGM.getIntrinsic(Int, Tys), Ops, "vminnmv");
- return Builder.CreateTrunc(Ops[0], HalfTy);
- }
- case NEON::BI__builtin_neon_vminnmvq_f16: {
- Int = Intrinsic::aarch64_neon_fminnmv;
- Ty = HalfTy;
- VTy = llvm::VectorType::get(HalfTy, 8);
- llvm::Type *Tys[2] = { Ty, VTy };
- Ops.push_back(EmitScalarExpr(E->getArg(0)));
- Ops[0] = EmitNeonCall(CGM.getIntrinsic(Int, Tys), Ops, "vminnmv");
- return Builder.CreateTrunc(Ops[0], HalfTy);
- }
case NEON::BI__builtin_neon_vmul_n_f64: {
Ops[0] = Builder.CreateBitCast(Ops[0], DoubleTy);
Value *RHS = Builder.CreateBitCast(EmitScalarExpr(E->getArg(1)), DoubleTy);
@@ -7506,6 +7334,8 @@ Value *CodeGenFunction::EmitX86BuiltinExpr(unsigned BuiltinID,
AVX512PF,
AVX512VBMI,
AVX512IFMA,
+ AVX5124VNNIW, // TODO implement this fully
+ AVX5124FMAPS, // TODO implement this fully
AVX512VPOPCNTDQ,
MAX
};
@@ -8208,13 +8038,13 @@ Value *CodeGenFunction::EmitX86BuiltinExpr(unsigned BuiltinID,
case X86::BI__faststorefence: {
return Builder.CreateFence(llvm::AtomicOrdering::SequentiallyConsistent,
- llvm::CrossThread);
+ llvm::SyncScope::System);
}
case X86::BI_ReadWriteBarrier:
case X86::BI_ReadBarrier:
case X86::BI_WriteBarrier: {
return Builder.CreateFence(llvm::AtomicOrdering::SequentiallyConsistent,
- llvm::SingleThread);
+ llvm::SyncScope::SingleThread);
}
case X86::BI_BitScanForward:
case X86::BI_BitScanForward64:
generated by cgit v1.2.3 (git 2.25.1) at 2025-06-29 01:58:42 +0000