This is an archive of the discontinued LLVM Phabricator instance.

Differential D73184

[CodeGen] Emit IR for compound assignment with fixed-point operands.
ClosedPublic

Authored by ebevhan on Jan 22 2020, 6:22 AM.

Details

Reviewers
rjmccall
leonardchan
bjope
Commits
rG313461f6d8f9: [CodeGen] Emit IR for compound assignment with fixed-point operands.

Diff Detail

Event Timeline

ebevhan created this revision.Jan 22 2020, 6:22 AM
Herald added a project: Restricted Project. · View Herald TranscriptJan 22 2020, 6:22 AM
Herald added a subscriber: cfe-commits. · View Herald Transcript
Harbormaster completed remote builds in B44581: Diff 239562.Jan 22 2020, 6:24 AM
ebevhan added a reviewer: bjope.Jan 22 2020, 6:31 AM
ebevhan updated this revision to Diff 239795.Jan 23 2020, 12:37 AM

Rebased.

Harbormaster completed remote builds in B44678: Diff 239795.Jan 23 2020, 12:40 AM
leonardchan accepted this revision.Jan 23 2020, 4:06 PM
This revision is now accepted and ready to land.Jan 23 2020, 4:06 PM
rjmccall added a comment.Jan 28 2020, 10:18 AM

Please test the cases where the result is used; this should be handled for you automatically, but still.

ebevhan updated this revision to Diff 255972.Apr 8 2020, 5:08 AM

Rebased.

Harbormaster failed remote builds in B52335: Diff 255972!Apr 8 2020, 5:56 AM
Closed by commit rG313461f6d8f9: [CodeGen] Emit IR for compound assignment with fixed-point operands. (authored by ebevhan). · Explain WhyApr 8 2020, 5:57 AM
This revision was automatically updated to reflect the committed changes.

Revision Contents

PathSize
clang/
lib/
CodeGen/
20 lines
Sema/
8 lines
test/
Frontend/
374 lines

Diff 239562

clang/lib/CodeGen/CGExprScalar.cpp

Show First 20 Lines • Show All 3,498 Lines • ▼ Show 20 LinesValue *ScalarExprEmitter::EmitFixedPointBinOp(const BinOpInfo &op) {
// This is either a binary operation where at least one of the operands is // This is either a binary operation where at least one of the operands is
// a fixed-point type, or a unary operation where the operand is a fixed-point // a fixed-point type, or a unary operation where the operand is a fixed-point
// type. The result type of a binary operation is determined by // type. The result type of a binary operation is determined by
// Sema::handleFixedPointConversions(). // Sema::handleFixedPointConversions().
QualType ResultTy = op.Ty; QualType ResultTy = op.Ty;
QualType LHSTy, RHSTy; QualType LHSTy, RHSTy;
if (const auto *BinOp = dyn_cast<BinaryOperator>(op.E)) { if (const auto *BinOp = dyn_cast<BinaryOperator>(op.E)) {
LHSTy = BinOp->getLHS()->getType();
RHSTy = BinOp->getRHS()->getType(); RHSTy = BinOp->getRHS()->getType();
if (const auto *CAO = dyn_cast<CompoundAssignOperator>(BinOp)) {
// For compound assignment, the effective type of the LHS at this point
// is the computation LHS type, not the actual LHS type, and the final
// result type is not the type of the expression but rather the
// computation result type.
LHSTy = CAO->getComputationLHSType();
ResultTy = CAO->getComputationResultType();
} else
LHSTy = BinOp->getLHS()->getType();
} else if (const auto *UnOp = dyn_cast<UnaryOperator>(op.E)) { } else if (const auto *UnOp = dyn_cast<UnaryOperator>(op.E)) {
LHSTy = UnOp->getSubExpr()->getType(); LHSTy = UnOp->getSubExpr()->getType();
RHSTy = UnOp->getSubExpr()->getType(); RHSTy = UnOp->getSubExpr()->getType();
} }
ASTContext &Ctx = CGF.getContext(); ASTContext &Ctx = CGF.getContext();
Value *LHS = op.LHS; Value *LHS = op.LHS;
Value *RHS = op.RHS; Value *RHS = op.RHS;
auto LHSFixedSema = Ctx.getFixedPointSemantics(LHSTy); auto LHSFixedSema = Ctx.getFixedPointSemantics(LHSTy);
auto RHSFixedSema = Ctx.getFixedPointSemantics(RHSTy); auto RHSFixedSema = Ctx.getFixedPointSemantics(RHSTy);
auto ResultFixedSema = Ctx.getFixedPointSemantics(ResultTy); auto ResultFixedSema = Ctx.getFixedPointSemantics(ResultTy);
auto CommonFixedSema = LHSFixedSema.getCommonSemantics(RHSFixedSema); auto CommonFixedSema = LHSFixedSema.getCommonSemantics(RHSFixedSema);
// Convert the operands to the full precision type. // Convert the operands to the full precision type.
Value *FullLHS = EmitFixedPointConversion(LHS, LHSFixedSema, CommonFixedSema, Value *FullLHS = EmitFixedPointConversion(LHS, LHSFixedSema, CommonFixedSema,
op.E->getExprLoc()); op.E->getExprLoc());
Value *FullRHS = EmitFixedPointConversion(RHS, RHSFixedSema, CommonFixedSema, Value *FullRHS = EmitFixedPointConversion(RHS, RHSFixedSema, CommonFixedSema,
op.E->getExprLoc()); op.E->getExprLoc());
// Perform the actual addition. // Perform the actual operation.
Value *Result; Value *Result;
switch (op.Opcode) { switch (op.Opcode) {
case BO_AddAssign:
case BO_Add: { case BO_Add: {
if (ResultFixedSema.isSaturated()) { if (ResultFixedSema.isSaturated()) {
llvm::Intrinsic::ID IID = ResultFixedSema.isSigned() llvm::Intrinsic::ID IID = ResultFixedSema.isSigned()
? llvm::Intrinsic::sadd_sat ? llvm::Intrinsic::sadd_sat
: llvm::Intrinsic::uadd_sat; : llvm::Intrinsic::uadd_sat;
Result = Builder.CreateBinaryIntrinsic(IID, FullLHS, FullRHS); Result = Builder.CreateBinaryIntrinsic(IID, FullLHS, FullRHS);
} else { } else {
Result = Builder.CreateAdd(FullLHS, FullRHS); Result = Builder.CreateAdd(FullLHS, FullRHS);
} }
break; break;
} }
case BO_SubAssign:
case BO_Sub: { case BO_Sub: {
if (ResultFixedSema.isSaturated()) { if (ResultFixedSema.isSaturated()) {
llvm::Intrinsic::ID IID = ResultFixedSema.isSigned() llvm::Intrinsic::ID IID = ResultFixedSema.isSigned()
? llvm::Intrinsic::ssub_sat ? llvm::Intrinsic::ssub_sat
: llvm::Intrinsic::usub_sat; : llvm::Intrinsic::usub_sat;
Result = Builder.CreateBinaryIntrinsic(IID, FullLHS, FullRHS); Result = Builder.CreateBinaryIntrinsic(IID, FullLHS, FullRHS);
} else { } else {
Result = Builder.CreateSub(FullLHS, FullRHS); Result = Builder.CreateSub(FullLHS, FullRHS);
} }
break; break;
} }
case BO_MulAssign:
case BO_Mul: { case BO_Mul: {
llvm::Intrinsic::ID IID; llvm::Intrinsic::ID IID;
if (ResultFixedSema.isSaturated()) if (ResultFixedSema.isSaturated())
IID = ResultFixedSema.isSigned() IID = ResultFixedSema.isSigned()
? llvm::Intrinsic::smul_fix_sat ? llvm::Intrinsic::smul_fix_sat
: llvm::Intrinsic::umul_fix_sat; : llvm::Intrinsic::umul_fix_sat;
else else
IID = ResultFixedSema.isSigned() IID = ResultFixedSema.isSigned()
? llvm::Intrinsic::smul_fix ? llvm::Intrinsic::smul_fix
: llvm::Intrinsic::umul_fix; : llvm::Intrinsic::umul_fix;
Result = Builder.CreateIntrinsic(IID, {FullLHS->getType()}, Result = Builder.CreateIntrinsic(IID, {FullLHS->getType()},
{FullLHS, FullRHS, Builder.getInt32(CommonFixedSema.getScale())}); {FullLHS, FullRHS, Builder.getInt32(CommonFixedSema.getScale())});
break; break;
} }
case BO_DivAssign:
case BO_Div: { case BO_Div: {
llvm::Intrinsic::ID IID; llvm::Intrinsic::ID IID;
if (ResultFixedSema.isSaturated()) if (ResultFixedSema.isSaturated())
IID = ResultFixedSema.isSigned() ? llvm::Intrinsic::sdiv_fix_sat IID = ResultFixedSema.isSigned() ? llvm::Intrinsic::sdiv_fix_sat
: llvm::Intrinsic::udiv_fix_sat; : llvm::Intrinsic::udiv_fix_sat;
else else
IID = ResultFixedSema.isSigned() ? llvm::Intrinsic::sdiv_fix IID = ResultFixedSema.isSigned() ? llvm::Intrinsic::sdiv_fix
: llvm::Intrinsic::udiv_fix; : llvm::Intrinsic::udiv_fix;
Show All 20 Lines case BO_EQ:
return Builder.CreateICmpEQ(FullLHS, FullRHS); return Builder.CreateICmpEQ(FullLHS, FullRHS);
case BO_NE: case BO_NE:
return Builder.CreateICmpNE(FullLHS, FullRHS); return Builder.CreateICmpNE(FullLHS, FullRHS);
case BO_Shl: case BO_Shl:
case BO_Shr: case BO_Shr:
case BO_Cmp: case BO_Cmp:
case BO_LAnd: case BO_LAnd:
case BO_LOr: case BO_LOr:
case BO_MulAssign:
case BO_DivAssign:
case BO_AddAssign:
case BO_SubAssign:
case BO_ShlAssign: case BO_ShlAssign:
case BO_ShrAssign: case BO_ShrAssign:
llvm_unreachable("Found unimplemented fixed point binary operation"); llvm_unreachable("Found unimplemented fixed point binary operation");
case BO_PtrMemD: case BO_PtrMemD:
case BO_PtrMemI: case BO_PtrMemI:
case BO_Rem: case BO_Rem:
case BO_Xor: case BO_Xor:
case BO_And: case BO_And:
▲ Show 20 Lines • Show All 1,295 Lines • Show Last 20 Lines

clang/lib/Sema/SemaExpr.cpp

  • This file is larger than 256 KB, so syntax highlighting is disabled by default.
Show First 20 Lines • Show All 13,088 Lines • ▼ Show 20 Lines if (getLangOpts().CPlusPlus && !RHS.isInvalid()) {
VK = RHS.get()->getValueKind(); VK = RHS.get()->getValueKind();
OK = RHS.get()->getObjectKind(); OK = RHS.get()->getObjectKind();
} }
break; break;
} }
if (ResultTy.isNull() || LHS.isInvalid() || RHS.isInvalid()) if (ResultTy.isNull() || LHS.isInvalid() || RHS.isInvalid())
return ExprError(); return ExprError();
// The LHS is not converted to the result type for fixed-point compound
// assignment as the common type is computed on demand. Reset the CompLHSTy
// to the LHS type we would have gotten after unary conversions.
if (!CompLHSTy.isNull() &&
(LHS.get()->getType()->isFixedPointType() ||
RHS.get()->getType()->isFixedPointType()))
CompLHSTy = UsualUnaryConversions(LHS.get()).get()->getType();
if (ResultTy->isRealFloatingType() && if (ResultTy->isRealFloatingType() &&
(getLangOpts().getFPRoundingMode() != LangOptions::FPR_ToNearest || (getLangOpts().getFPRoundingMode() != LangOptions::FPR_ToNearest ||
getLangOpts().getFPExceptionMode() != LangOptions::FPE_Ignore)) getLangOpts().getFPExceptionMode() != LangOptions::FPE_Ignore))
// Mark the current function as usng floating point constrained intrinsics // Mark the current function as usng floating point constrained intrinsics
if (FunctionDecl *F = dyn_cast<FunctionDecl>(CurContext)) { if (FunctionDecl *F = dyn_cast<FunctionDecl>(CurContext)) {
F->setUsesFPIntrin(true); F->setUsesFPIntrin(true);
} }
▲ Show 20 Lines • Show All 5,010 Lines • Show Last 20 Lines

clang/test/Frontend/fixed_point_compound.c

  • This file was added.
// RUN: %clang_cc1 -ffixed-point -S -emit-llvm %s -o - | FileCheck %s --check-prefixes=CHECK,SIGNED
// RUN: %clang_cc1 -ffixed-point -fpadding-on-unsigned-fixed-point -S -emit-llvm %s -o - | FileCheck %s --check-prefixes=CHECK,UNSIGNED
short _Fract shf;
_Accum a;
unsigned _Fract uf;
unsigned long _Accum ula;
_Sat short _Fract sshf;
_Sat _Accum sa;
_Sat unsigned _Fract suf;
_Sat unsigned long _Accum sula;
int i;
unsigned int u;
signed char c;
// CHECK-LABEL: @Addition(
void Addition() {
// CHECK: [[TMP0:%.*]] = load i32, i32* @a, align 4
// CHECK-NEXT: [[TMP1:%.*]] = load i8, i8* @shf, align 1
// CHECK-NEXT: [[RESIZE:%.*]] = sext i8 [[TMP1]] to i32
// CHECK-NEXT: [[UPSCALE:%.*]] = shl i32 [[RESIZE]], 8
// CHECK-NEXT: [[TMP2:%.*]] = add i32 [[UPSCALE]], [[TMP0]]
// CHECK-NEXT: [[DOWNSCALE:%.*]] = ashr i32 [[TMP2]], 8
// CHECK-NEXT: [[RESIZE1:%.*]] = trunc i32 [[DOWNSCALE]] to i8
// CHECK-NEXT: store i8 [[RESIZE1]], i8* @shf, align 1
shf += a;
// CHECK: [[TMP3:%.*]] = load i16, i16* @uf, align 2
// CHECK-NEXT: [[TMP4:%.*]] = load i32, i32* @a, align 4
// SIGNED-NEXT: [[RESIZE2:%.*]] = sext i32 [[TMP4]] to i33
// SIGNED-NEXT: [[UPSCALE3:%.*]] = shl i33 [[RESIZE2]], 1
// SIGNED-NEXT: [[RESIZE4:%.*]] = zext i16 [[TMP3]] to i33
// SIGNED-NEXT: [[TMP5:%.*]] = add i33 [[UPSCALE3]], [[RESIZE4]]
// SIGNED-NEXT: [[DOWNSCALE5:%.*]] = ashr i33 [[TMP5]], 1
// SIGNED-NEXT: [[RESIZE6:%.*]] = trunc i33 [[DOWNSCALE5]] to i32
// SIGNED-NEXT: store i32 [[RESIZE6]], i32* @a, align 4
// UNSIGNED-NEXT: [[RESIZE2:%.*]] = zext i16 [[TMP3]] to i32
// UNSIGNED-NEXT: [[TMP5:%.*]] = add i32 [[TMP4]], [[RESIZE2]]
// UNSIGNED-NEXT: store i32 [[TMP5]], i32* @a, align 4
a += uf;
// CHECK: [[TMP6:%.*]] = load i64, i64* @ula, align 8
// CHECK-NEXT: [[TMP7:%.*]] = load i16, i16* @uf, align 2
// CHECK-NEXT: [[RESIZE7:%.*]] = zext i16 [[TMP7]] to i64
// CHECK-NEXT: [[UPSCALE8:%.*]] = shl i64 [[RESIZE7]], 16
// CHECK-NEXT: [[TMP8:%.*]] = add i64 [[UPSCALE8]], [[TMP6]]
// CHECK-NEXT: [[DOWNSCALE9:%.*]] = lshr i64 [[TMP8]], 16
// CHECK-NEXT: [[RESIZE10:%.*]] = trunc i64 [[DOWNSCALE9]] to i16
// CHECK-NEXT: store i16 [[RESIZE10]], i16* @uf, align 2
uf += ula;
// CHECK: [[TMP9:%.*]] = load i8, i8* @shf, align 1
// CHECK-NEXT: [[TMP10:%.*]] = load i64, i64* @ula, align 8
// SIGNED-NEXT: [[RESIZE11:%.*]] = zext i64 [[TMP10]] to i65
// SIGNED-NEXT: [[RESIZE12:%.*]] = sext i8 [[TMP9]] to i65
// SIGNED-NEXT: [[UPSCALE13:%.*]] = shl i65 [[RESIZE12]], 25
// SIGNED-NEXT: [[TMP11:%.*]] = add i65 [[RESIZE11]], [[UPSCALE13]]
// SIGNED-NEXT: [[DOWNSCALE14:%.*]] = ashr i65 [[TMP11]], 1
// SIGNED-NEXT: [[RESIZE15:%.*]] = trunc i65 [[DOWNSCALE14]] to i64
// SIGNED-NEXT: [[UPSCALE16:%.*]] = shl i64 [[RESIZE15]], 1
// SIGNED-NEXT: store i64 [[UPSCALE16]], i64* @ula, align 8
// UNSIGNED-NEXT: [[RESIZE7:%.*]] = sext i8 [[TMP9]] to i64
// UNSIGNED-NEXT: [[UPSCALE8:%.*]] = shl i64 [[RESIZE7]], 24
// UNSIGNED-NEXT: [[TMP11:%.*]] = add i64 [[TMP10]], [[UPSCALE8]]
// UNSIGNED-NEXT: store i64 [[TMP11]], i64* @ula, align 8
ula += shf;
// CHECK: [[TMP12:%.*]] = load i8, i8* @shf, align 1
// CHECK-NEXT: [[TMP13:%.*]] = load i16, i16* @uf, align 2
// SIGNED-NEXT: [[RESIZE17:%.*]] = zext i16 [[TMP13]] to i17
// SIGNED-NEXT: [[RESIZE18:%.*]] = sext i8 [[TMP12]] to i17
// SIGNED-NEXT: [[UPSCALE19:%.*]] = shl i17 [[RESIZE18]], 9
// SIGNED-NEXT: [[TMP14:%.*]] = add i17 [[RESIZE17]], [[UPSCALE19]]
// SIGNED-NEXT: [[DOWNSCALE20:%.*]] = ashr i17 [[TMP14]], 1
// SIGNED-NEXT: [[RESIZE21:%.*]] = trunc i17 [[DOWNSCALE20]] to i16
// SIGNED-NEXT: [[UPSCALE22:%.*]] = shl i16 [[RESIZE21]], 1
// SIGNED-NEXT: store i16 [[UPSCALE22]], i16* @uf, align 2
// UNSIGNED-NEXT: [[RESIZE9:%.*]] = sext i8 [[TMP12]] to i16
// UNSIGNED-NEXT: [[UPSCALE10:%.*]] = shl i16 [[RESIZE9]], 8
// UNSIGNED-NEXT: [[TMP14:%.*]] = add i16 [[TMP13]], [[UPSCALE10]]
// UNSIGNED-NEXT: store i16 [[TMP14]], i16* @uf, align 2
uf += shf;
// CHECK: [[TMP15:%.*]] = load i8, i8* @shf, align 1
// CHECK-NEXT: [[TMP16:%.*]] = load i32, i32* @a, align 4
// CHECK-NEXT: [[RESIZE23:%.*]] = sext i8 [[TMP15]] to i32
// CHECK-NEXT: [[UPSCALE24:%.*]] = shl i32 [[RESIZE23]], 8
// CHECK-NEXT: [[TMP17:%.*]] = add i32 [[TMP16]], [[UPSCALE24]]
// CHECK-NEXT: store i32 [[TMP17]], i32* @a, align 4
a += shf;
// CHECK: [[TMP18:%.*]] = load i32, i32* @i, align 4
// CHECK-NEXT: [[TMP19:%.*]] = load i32, i32* @a, align 4
// CHECK-NEXT: [[RESIZE25:%.*]] = sext i32 [[TMP19]] to i47
// CHECK-NEXT: [[RESIZE26:%.*]] = sext i32 [[TMP18]] to i47
// CHECK-NEXT: [[UPSCALE27:%.*]] = shl i47 [[RESIZE26]], 15
// CHECK-NEXT: [[TMP20:%.*]] = add i47 [[RESIZE25]], [[UPSCALE27]]
// CHECK-NEXT: [[RESIZE28:%.*]] = trunc i47 [[TMP20]] to i32
// CHECK-NEXT: store i32 [[RESIZE28]], i32* @a, align 4
a += i;
// CHECK: [[TMP21:%.*]] = load i32, i32* @u, align 4
// CHECK-NEXT: [[TMP22:%.*]] = load i32, i32* @a, align 4
// CHECK-NEXT: [[RESIZE29:%.*]] = sext i32 [[TMP22]] to i48
// CHECK-NEXT: [[RESIZE30:%.*]] = zext i32 [[TMP21]] to i48
// CHECK-NEXT: [[UPSCALE31:%.*]] = shl i48 [[RESIZE30]], 15
// CHECK-NEXT: [[TMP23:%.*]] = add i48 [[RESIZE29]], [[UPSCALE31]]
// CHECK-NEXT: [[RESIZE32:%.*]] = trunc i48 [[TMP23]] to i32
// CHECK-NEXT: store i32 [[RESIZE32]], i32* @a, align 4
a += u;
// CHECK: [[TMP24:%.*]] = load i32, i32* @i, align 4
// CHECK-NEXT: [[TMP25:%.*]] = load i64, i64* @ula, align 8
// SIGNED-NEXT: [[RESIZE33:%.*]] = zext i64 [[TMP25]] to i65
// SIGNED-NEXT: [[RESIZE34:%.*]] = sext i32 [[TMP24]] to i65
// SIGNED-NEXT: [[UPSCALE35:%.*]] = shl i65 [[RESIZE34]], 32
// SIGNED-NEXT: [[TMP26:%.*]] = add i65 [[RESIZE33]], [[UPSCALE35]]
// SIGNED-NEXT: [[RESIZE36:%.*]] = trunc i65 [[TMP26]] to i64
// SIGNED-NEXT: store i64 [[RESIZE36]], i64* @ula, align 8
// UNSIGNED-NEXT: [[RESIZE21:%.*]] = sext i32 [[TMP24]] to i64
// UNSIGNED-NEXT: [[UPSCALE22:%.*]] = shl i64 [[RESIZE21]], 31
// UNSIGNED-NEXT: [[TMP26:%.*]] = add i64 [[TMP25]], [[UPSCALE22]]
// UNSIGNED-NEXT: store i64 [[TMP26]], i64* @ula, align 8
ula += i;
// CHECK: [[TMP27:%.*]] = load i64, i64* @ula, align 8
// CHECK-NEXT: [[TMP28:%.*]] = load i32, i32* @i, align 4
// SIGNED-NEXT: [[RESIZE37:%.*]] = sext i32 [[TMP28]] to i65
// SIGNED-NEXT: [[UPSCALE38:%.*]] = shl i65 [[RESIZE37]], 32
// SIGNED-NEXT: [[RESIZE39:%.*]] = zext i64 [[TMP27]] to i65
// SIGNED-NEXT: [[TMP29:%.*]] = add i65 [[UPSCALE38]], [[RESIZE39]]
// SIGNED-NEXT: [[RESIZE40:%.*]] = trunc i65 [[TMP29]] to i64
// SIGNED-NEXT: [[DOWNSCALE41:%.*]] = lshr i64 [[RESIZE40]], 32
// SIGNED-NEXT: [[RESIZE42:%.*]] = trunc i64 [[DOWNSCALE41]] to i32
// SIGNED-NEXT: store i32 [[RESIZE42]], i32* @i, align 4
// UNSIGNED-NEXT: [[RESIZE23:%.*]] = sext i32 [[TMP28]] to i64
// UNSIGNED-NEXT: [[UPSCALE24:%.*]] = shl i64 [[RESIZE23]], 31
// UNSIGNED-NEXT: [[TMP29:%.*]] = add i64 [[UPSCALE24]], [[TMP27]]
// UNSIGNED-NEXT: [[DOWNSCALE25:%.*]] = lshr i64 [[TMP29]], 31
// UNSIGNED-NEXT: [[RESIZE26:%.*]] = trunc i64 [[DOWNSCALE25]] to i32
// UNSIGNED-NEXT: store i32 [[RESIZE26]], i32* @i, align 4
i += ula;
// CHECK: [[TMP30:%.*]] = load i32, i32* @a, align 4
// CHECK-NEXT: [[TMP31:%.*]] = load i8, i8* @c, align 1
// CHECK-NEXT: [[CONV:%.*]] = sext i8 [[TMP31]] to i32
// CHECK-NEXT: [[RESIZE43:%.*]] = sext i32 [[CONV]] to i47
// CHECK-NEXT: [[UPSCALE44:%.*]] = shl i47 [[RESIZE43]], 15
// CHECK-NEXT: [[RESIZE45:%.*]] = sext i32 [[TMP30]] to i47
// CHECK-NEXT: [[TMP32:%.*]] = add i47 [[UPSCALE44]], [[RESIZE45]]
// CHECK-NEXT: [[RESIZE46:%.*]] = trunc i47 [[TMP32]] to i32
// CHECK-NEXT: [[TMP33:%.*]] = icmp slt i32 [[RESIZE46]], 0
// CHECK-NEXT: [[TMP34:%.*]] = add i32 [[RESIZE46]], 32767
// CHECK-NEXT: [[TMP35:%.*]] = select i1 [[TMP33]], i32 [[TMP34]], i32 [[RESIZE46]]
// CHECK-NEXT: [[DOWNSCALE47:%.*]] = ashr i32 [[TMP35]], 15
// CHECK-NEXT: [[RESIZE48:%.*]] = trunc i32 [[DOWNSCALE47]] to i8
// CHECK-NEXT: store i8 [[RESIZE48]], i8* @c, align 1
c += a;
// CHECK: [[TMP36:%.*]] = load i32, i32* @i, align 4
// CHECK-NEXT: [[TMP37:%.*]] = load i32, i32* @sa, align 4
// CHECK-NEXT: [[RESIZE47:%.*]] = sext i32 [[TMP37]] to i47
// CHECK-NEXT: [[RESIZE48:%.*]] = sext i32 [[TMP36]] to i47
// CHECK-NEXT: [[UPSCALE49:%.*]] = shl i47 [[RESIZE48]], 15
// CHECK-NEXT: [[TMP38:%.*]] = call i47 @llvm.sadd.sat.i47(i47 [[RESIZE47]], i47 [[UPSCALE49]])
// CHECK-NEXT: [[TMP39:%.*]] = icmp sgt i47 [[TMP38]], 2147483647
// CHECK-NEXT: [[SATMAX:%.*]] = select i1 [[TMP39]], i47 2147483647, i47 [[TMP38]]
// CHECK-NEXT: [[TMP40:%.*]] = icmp slt i47 [[SATMAX]], -2147483648
// CHECK-NEXT: [[SATMIN:%.*]] = select i1 [[TMP40]], i47 -2147483648, i47 [[SATMAX]]
// CHECK-NEXT: [[RESIZE50:%.*]] = trunc i47 [[SATMIN]] to i32
// CHECK-NEXT: store i32 [[RESIZE50]], i32* @sa, align 4
sa += i;
// CHECK: [[TMP41:%.*]] = load i32, i32* @sa, align 4
// CHECK-NEXT: [[TMP42:%.*]] = load i8, i8* @c, align 1
// CHECK-NEXT: [[CONV53:%.*]] = sext i8 [[TMP42]] to i32
// CHECK-NEXT: [[RESIZE54:%.*]] = sext i32 [[CONV53]] to i47
// CHECK-NEXT: [[UPSCALE55:%.*]] = shl i47 [[RESIZE54]], 15
// CHECK-NEXT: [[RESIZE56:%.*]] = sext i32 [[TMP41]] to i47
// CHECK-NEXT: [[TMP43:%.*]] = call i47 @llvm.sadd.sat.i47(i47 [[UPSCALE55]], i47 [[RESIZE56]])
// CHECK-NEXT: [[TMP44:%.*]] = icmp sgt i47 [[TMP43]], 2147483647
// CHECK-NEXT: [[SATMAX57:%.*]] = select i1 [[TMP44]], i47 2147483647, i47 [[TMP43]]
// CHECK-NEXT: [[TMP45:%.*]] = icmp slt i47 [[SATMAX57]], -2147483648
// CHECK-NEXT: [[SATMIN58:%.*]] = select i1 [[TMP45]], i47 -2147483648, i47 [[SATMAX57]]
// CHECK-NEXT: [[RESIZE59:%.*]] = trunc i47 [[SATMIN58]] to i32
// CHECK-NEXT: [[TMP46:%.*]] = icmp slt i32 [[RESIZE59]], 0
// CHECK-NEXT: [[TMP47:%.*]] = add i32 [[RESIZE59]], 32767
// CHECK-NEXT: [[TMP48:%.*]] = select i1 [[TMP46]], i32 [[TMP47]], i32 [[RESIZE59]]
// CHECK-NEXT: [[DOWNSCALE60:%.*]] = ashr i32 [[TMP48]], 15
// CHECK-NEXT: [[RESIZE61:%.*]] = trunc i32 [[DOWNSCALE60]] to i8
// CHECK-NEXT: store i8 [[RESIZE61]], i8* @c, align 1
c += sa;
// CHECK: [[TMP47:%.*]] = load i32, i32* @u, align 4
// CHECK-NEXT: [[TMP48:%.*]] = load i64, i64* @sula, align 8
// SIGNED-NEXT: [[RESIZE55:%.*]] = zext i32 [[TMP47]] to i64
// SIGNED-NEXT: [[UPSCALE56:%.*]] = shl i64 [[RESIZE55]], 32
// SIGNED-NEXT: [[TMP49:%.*]] = call i64 @llvm.uadd.sat.i64(i64 [[TMP48]], i64 [[UPSCALE56]])
// SIGNED-NEXT: store i64 [[TMP49]], i64* @sula, align 8
// UNSIGNED-NEXT: [[RESIZE39:%.*]] = trunc i64 [[TMP48]] to i63
// UNSIGNED-NEXT: [[RESIZE40:%.*]] = zext i32 [[TMP47]] to i63
// UNSIGNED-NEXT: [[UPSCALE41:%.*]] = shl i63 [[RESIZE40]], 31
// UNSIGNED-NEXT: [[TMP49:%.*]] = call i63 @llvm.uadd.sat.i63(i63 [[RESIZE39]], i63 [[UPSCALE41]])
// UNSIGNED-NEXT: [[RESIZE42:%.*]] = zext i63 [[TMP49]] to i64
// UNSIGNED-NEXT: store i64 [[RESIZE42]], i64* @sula, align 8
sula += u;
// CHECK: [[TMP50:%.*]] = load i16, i16* @suf, align 2
// CHECK-NEXT: [[TMP51:%.*]] = load i8, i8* @sshf, align 1
// SIGNED-NEXT: [[RESIZE57:%.*]] = sext i8 [[TMP51]] to i17
// SIGNED-NEXT: [[UPSCALE58:%.*]] = shl i17 [[RESIZE57]], 9
// SIGNED-NEXT: [[RESIZE59:%.*]] = zext i16 [[TMP50]] to i17
// SIGNED-NEXT: [[TMP52:%.*]] = call i17 @llvm.sadd.sat.i17(i17 [[UPSCALE58]], i17 [[RESIZE59]])
// SIGNED-NEXT: [[DOWNSCALE60:%.*]] = ashr i17 [[TMP52]], 1
// SIGNED-NEXT: [[RESIZE61:%.*]] = trunc i17 [[DOWNSCALE60]] to i16
// SIGNED-NEXT: [[DOWNSCALE62:%.*]] = ashr i16 [[RESIZE61]], 8
// SIGNED-NEXT: [[RESIZE63:%.*]] = trunc i16 [[DOWNSCALE62]] to i8
// SIGNED-NEXT: store i8 [[RESIZE63]], i8* @sshf, align 1
// UNSIGNED-NEXT: [[RESIZE43:%.*]] = sext i8 [[TMP51]] to i16
// UNSIGNED-NEXT: [[UPSCALE44:%.*]] = shl i16 [[RESIZE43]], 8
// UNSIGNED-NEXT: [[TMP52:%.*]] = call i16 @llvm.sadd.sat.i16(i16 [[UPSCALE44]], i16 [[TMP50]])
// UNSIGNED-NEXT: [[DOWNSCALE45:%.*]] = ashr i16 [[TMP52]], 8
// UNSIGNED-NEXT: [[RESIZE46:%.*]] = trunc i16 [[DOWNSCALE45]] to i8
// UNSIGNED-NEXT: store i8 [[RESIZE46]], i8* @sshf, align 1
sshf += suf;
}
// Subtraction, multiplication and division should work about the same, so
// just make sure we can do them.
// CHECK-LABEL: @Subtraction(
void Subtraction() {
// CHECK: [[TMP0:%.*]] = load i16, i16* @uf, align 2
// CHECK-NEXT: [[TMP1:%.*]] = load i32, i32* @a, align 4
// SIGNED-NEXT: [[RESIZE:%.*]] = sext i32 [[TMP1]] to i33
// SIGNED-NEXT: [[UPSCALE:%.*]] = shl i33 [[RESIZE]], 1
// SIGNED-NEXT: [[RESIZE1:%.*]] = zext i16 [[TMP0]] to i33
// SIGNED-NEXT: [[TMP2:%.*]] = sub i33 [[UPSCALE]], [[RESIZE1]]
// SIGNED-NEXT: [[DOWNSCALE:%.*]] = ashr i33 [[TMP2]], 1
// SIGNED-NEXT: [[RESIZE2:%.*]] = trunc i33 [[DOWNSCALE]] to i32
// SIGNED-NEXT: store i32 [[RESIZE2]], i32* @a, align 4
// UNSIGNED-NEXT: [[RESIZE:%.*]] = zext i16 [[TMP0]] to i32
// UNSIGNED-NEXT: [[TMP2:%.*]] = sub i32 [[TMP1]], [[RESIZE]]
// UNSIGNED-NEXT: store i32 [[TMP2]], i32* @a, align 4
a -= uf;
// CHECK: [[TMP3:%.*]] = load i32, i32* @i, align 4
// CHECK-NEXT: [[TMP4:%.*]] = load i32, i32* @a, align 4
// CHECK-NEXT: [[RESIZE3:%.*]] = sext i32 [[TMP4]] to i47
// CHECK-NEXT: [[RESIZE4:%.*]] = sext i32 [[TMP3]] to i47
// CHECK-NEXT: [[UPSCALE5:%.*]] = shl i47 [[RESIZE4]], 15
// CHECK-NEXT: [[TMP5:%.*]] = sub i47 [[RESIZE3]], [[UPSCALE5]]
// CHECK-NEXT: [[RESIZE6:%.*]] = trunc i47 [[TMP5]] to i32
// CHECK-NEXT: store i32 [[RESIZE6]], i32* @a, align 4
a -= i;
// CHECK: [[TMP6:%.*]] = load i32, i32* @sa, align 4
// CHECK-NEXT: [[TMP7:%.*]] = load i8, i8* @c, align 1
// CHECK-NEXT: [[CONV:%.*]] = sext i8 [[TMP7]] to i32
// CHECK-NEXT: [[RESIZE7:%.*]] = sext i32 [[CONV]] to i47
// CHECK-NEXT: [[UPSCALE8:%.*]] = shl i47 [[RESIZE7]], 15
// CHECK-NEXT: [[RESIZE9:%.*]] = sext i32 [[TMP6]] to i47
// CHECK-NEXT: [[TMP8:%.*]] = call i47 @llvm.ssub.sat.i47(i47 [[UPSCALE8]], i47 [[RESIZE9]])
// CHECK-NEXT: [[TMP9:%.*]] = icmp sgt i47 [[TMP8]], 2147483647
// CHECK-NEXT: [[SATMAX:%.*]] = select i1 [[TMP9]], i47 2147483647, i47 [[TMP8]]
// CHECK-NEXT: [[TMP10:%.*]] = icmp slt i47 [[SATMAX]], -2147483648
// CHECK-NEXT: [[SATMIN:%.*]] = select i1 [[TMP10]], i47 -2147483648, i47 [[SATMAX]]
// CHECK-NEXT: [[RESIZE10:%.*]] = trunc i47 [[SATMIN]] to i32
// CHECK-NEXT: [[TMP11:%.*]] = icmp slt i32 [[RESIZE10]], 0
// CHECK-NEXT: [[TMP12:%.*]] = add i32 [[RESIZE10]], 32767
// CHECK-NEXT: [[TMP13:%.*]] = select i1 [[TMP11]], i32 [[TMP12]], i32 [[RESIZE10]]
// CHECK-NEXT: [[DOWNSCALE11:%.*]] = ashr i32 [[TMP13]], 15
// CHECK-NEXT: [[RESIZE12:%.*]] = trunc i32 [[DOWNSCALE11]] to i8
// CHECK-NEXT: store i8 [[RESIZE12]], i8* @c, align 1
c -= sa;
}
// CHECK-LABEL: @Multiplication(
void Multiplication() {
// CHECK: [[TMP0:%.*]] = load i16, i16* @uf, align 2
// CHECK-NEXT: [[TMP1:%.*]] = load i32, i32* @a, align 4
// SIGNED-NEXT: [[RESIZE:%.*]] = sext i32 [[TMP1]] to i33
// SIGNED-NEXT: [[UPSCALE:%.*]] = shl i33 [[RESIZE]], 1
// SIGNED-NEXT: [[RESIZE1:%.*]] = zext i16 [[TMP0]] to i33
// SIGNED-NEXT: [[TMP2:%.*]] = call i33 @llvm.smul.fix.i33(i33 [[UPSCALE]], i33 [[RESIZE1]], i32 16)
// SIGNED-NEXT: [[DOWNSCALE:%.*]] = ashr i33 [[TMP2]], 1
// SIGNED-NEXT: [[RESIZE2:%.*]] = trunc i33 [[DOWNSCALE]] to i32
// SIGNED-NEXT: store i32 [[RESIZE2]], i32* @a, align 4
// UNSIGNED-NEXT: [[RESIZE:%.*]] = zext i16 [[TMP0]] to i32
// UNSIGNED-NEXT: [[TMP2:%.*]] = call i32 @llvm.smul.fix.i32(i32 [[TMP1]], i32 [[RESIZE]], i32 15)
// UNSIGNED-NEXT: store i32 [[TMP2]], i32* @a, align 4
a *= uf;
// CHECK: [[TMP3:%.*]] = load i32, i32* @i, align 4
// CHECK-NEXT: [[TMP4:%.*]] = load i32, i32* @a, align 4
// CHECK-NEXT: [[RESIZE3:%.*]] = sext i32 [[TMP4]] to i47
// CHECK-NEXT: [[RESIZE4:%.*]] = sext i32 [[TMP3]] to i47
// CHECK-NEXT: [[UPSCALE5:%.*]] = shl i47 [[RESIZE4]], 15
// CHECK-NEXT: [[TMP5:%.*]] = call i47 @llvm.smul.fix.i47(i47 [[RESIZE3]], i47 [[UPSCALE5]], i32 15)
// CHECK-NEXT: [[RESIZE6:%.*]] = trunc i47 [[TMP5]] to i32
// CHECK-NEXT: store i32 [[RESIZE6]], i32* @a, align 4
a *= i;
// CHECK: [[TMP6:%.*]] = load i32, i32* @sa, align 4
// CHECK-NEXT: [[TMP7:%.*]] = load i8, i8* @c, align 1
// CHECK-NEXT: [[CONV:%.*]] = sext i8 [[TMP7]] to i32
// CHECK-NEXT: [[RESIZE7:%.*]] = sext i32 [[CONV]] to i47
// CHECK-NEXT: [[UPSCALE8:%.*]] = shl i47 [[RESIZE7]], 15
// CHECK-NEXT: [[RESIZE9:%.*]] = sext i32 [[TMP6]] to i47
// CHECK-NEXT: [[TMP8:%.*]] = call i47 @llvm.smul.fix.sat.i47(i47 [[UPSCALE8]], i47 [[RESIZE9]], i32 15)
// CHECK-NEXT: [[TMP9:%.*]] = icmp sgt i47 [[TMP8]], 2147483647
// CHECK-NEXT: [[SATMAX:%.*]] = select i1 [[TMP9]], i47 2147483647, i47 [[TMP8]]
// CHECK-NEXT: [[TMP10:%.*]] = icmp slt i47 [[SATMAX]], -2147483648
// CHECK-NEXT: [[SATMIN:%.*]] = select i1 [[TMP10]], i47 -2147483648, i47 [[SATMAX]]
// CHECK-NEXT: [[RESIZE10:%.*]] = trunc i47 [[SATMIN]] to i32
// CHECK-NEXT: [[TMP11:%.*]] = icmp slt i32 [[RESIZE10]], 0
// CHECK-NEXT: [[TMP12:%.*]] = add i32 [[RESIZE10]], 32767
// CHECK-NEXT: [[TMP13:%.*]] = select i1 [[TMP11]], i32 [[TMP12]], i32 [[RESIZE10]]
// CHECK-NEXT: [[DOWNSCALE11:%.*]] = ashr i32 [[TMP13]], 15
// CHECK-NEXT: [[RESIZE12:%.*]] = trunc i32 [[DOWNSCALE11]] to i8
// CHECK-NEXT: store i8 [[RESIZE12]], i8* @c, align 1
c *= sa;
}
// CHECK-LABEL: @Division(
void Division() {
// CHECK: [[TMP0:%.*]] = load i16, i16* @uf, align 2
// CHECK-NEXT: [[TMP1:%.*]] = load i32, i32* @a, align 4
// SIGNED-NEXT: [[RESIZE:%.*]] = sext i32 [[TMP1]] to i33
// SIGNED-NEXT: [[UPSCALE:%.*]] = shl i33 [[RESIZE]], 1
// SIGNED-NEXT: [[RESIZE1:%.*]] = zext i16 [[TMP0]] to i33
// SIGNED-NEXT: [[TMP2:%.*]] = call i33 @llvm.sdiv.fix.i33(i33 [[UPSCALE]], i33 [[RESIZE1]], i32 16)
// SIGNED-NEXT: [[DOWNSCALE:%.*]] = ashr i33 [[TMP2]], 1
// SIGNED-NEXT: [[RESIZE2:%.*]] = trunc i33 [[DOWNSCALE]] to i32
// SIGNED-NEXT: store i32 [[RESIZE2]], i32* @a, align 4
// UNSIGNED-NEXT: [[RESIZE:%.*]] = zext i16 [[TMP0]] to i32
// UNSIGNED-NEXT: [[TMP2:%.*]] = call i32 @llvm.sdiv.fix.i32(i32 [[TMP1]], i32 [[RESIZE]], i32 15)
// UNSIGNED-NEXT: store i32 [[TMP2]], i32* @a, align 4
a /= uf;
// CHECK: [[TMP3:%.*]] = load i32, i32* @i, align 4
// CHECK-NEXT: [[TMP4:%.*]] = load i32, i32* @a, align 4
// CHECK-NEXT: [[RESIZE3:%.*]] = sext i32 [[TMP4]] to i47
// CHECK-NEXT: [[RESIZE4:%.*]] = sext i32 [[TMP3]] to i47
// CHECK-NEXT: [[UPSCALE5:%.*]] = shl i47 [[RESIZE4]], 15
// CHECK-NEXT: [[TMP5:%.*]] = call i47 @llvm.sdiv.fix.i47(i47 [[RESIZE3]], i47 [[UPSCALE5]], i32 15)
// CHECK-NEXT: [[RESIZE6:%.*]] = trunc i47 [[TMP5]] to i32
// CHECK-NEXT: store i32 [[RESIZE6]], i32* @a, align 4
a /= i;
// CHECK: [[TMP6:%.*]] = load i32, i32* @sa, align 4
// CHECK-NEXT: [[TMP7:%.*]] = load i8, i8* @c, align 1
// CHECK-NEXT: [[CONV:%.*]] = sext i8 [[TMP7]] to i32
// CHECK-NEXT: [[RESIZE7:%.*]] = sext i32 [[CONV]] to i47
// CHECK-NEXT: [[UPSCALE8:%.*]] = shl i47 [[RESIZE7]], 15
// CHECK-NEXT: [[RESIZE9:%.*]] = sext i32 [[TMP6]] to i47
// CHECK-NEXT: [[TMP8:%.*]] = call i47 @llvm.sdiv.fix.sat.i47(i47 [[UPSCALE8]], i47 [[RESIZE9]], i32 15)
// CHECK-NEXT: [[TMP9:%.*]] = icmp sgt i47 [[TMP8]], 2147483647
// CHECK-NEXT: [[SATMAX:%.*]] = select i1 [[TMP9]], i47 2147483647, i47 [[TMP8]]
// CHECK-NEXT: [[TMP10:%.*]] = icmp slt i47 [[SATMAX]], -2147483648
// CHECK-NEXT: [[SATMIN:%.*]] = select i1 [[TMP10]], i47 -2147483648, i47 [[SATMAX]]
// CHECK-NEXT: [[RESIZE10:%.*]] = trunc i47 [[SATMIN]] to i32
// CHECK-NEXT: [[TMP11:%.*]] = icmp slt i32 [[RESIZE10]], 0
// CHECK-NEXT: [[TMP12:%.*]] = add i32 [[RESIZE10]], 32767
// CHECK-NEXT: [[TMP13:%.*]] = select i1 [[TMP11]], i32 [[TMP12]], i32 [[RESIZE10]]
// CHECK-NEXT: [[DOWNSCALE11:%.*]] = ashr i32 [[TMP13]], 15
// CHECK-NEXT: [[RESIZE12:%.*]] = trunc i32 [[DOWNSCALE11]] to i8
// CHECK-NEXT: store i8 [[RESIZE12]], i8* @c, align 1
c /= sa;
}