[Mlir-commits] [mlir] [mlir][emitc] Add 'emitc.switch' op to the dialect (PR #102331)
llvmlistbot at llvm.org
llvmlistbot at llvm.org
Wed Aug 7 09:49:12 PDT 2024
llvmbot wrote:
<!--LLVM PR SUMMARY COMMENT-->
@llvm/pr-subscribers-mlir-emitc
Author: Andrey Timonin (EtoAndruwa)
<details>
<summary>Changes</summary>
This PR is continuation of the [previous one](https://github.com/llvm/llvm-project/pull/101478). As a result, the `emitc::SwitchOp` op was developed inspired by `scf::IndexSwitchOp`.
Main points of PR:
- Added the `emitc::SwitchOp` op to the EmitC dialect + CppEmitter
- Corresponding tests were added
- Conversion from the SCF dialect to the EmitC dialect will be supported after the approvement this/adjusted version of the `emitc::SwitchOp` op
---
Patch is 23.63 KiB, truncated to 20.00 KiB below, full version: https://github.com/llvm/llvm-project/pull/102331.diff
6 Files Affected:
- (modified) mlir/include/mlir/Dialect/EmitC/IR/EmitC.h (+3)
- (modified) mlir/include/mlir/Dialect/EmitC/IR/EmitC.td (+88-1)
- (modified) mlir/lib/Dialect/EmitC/IR/EmitC.cpp (+208)
- (modified) mlir/lib/Target/Cpp/TranslateToCpp.cpp (+46-3)
- (modified) mlir/test/Target/Cpp/invalid.mlir (+129)
- (added) mlir/test/Target/Cpp/switch.mlir (+129)
``````````diff
diff --git a/mlir/include/mlir/Dialect/EmitC/IR/EmitC.h b/mlir/include/mlir/Dialect/EmitC/IR/EmitC.h
index 87a4078f280f65..212271ea813080 100644
--- a/mlir/include/mlir/Dialect/EmitC/IR/EmitC.h
+++ b/mlir/include/mlir/Dialect/EmitC/IR/EmitC.h
@@ -47,6 +47,9 @@ bool isSupportedFloatType(mlir::Type type);
/// Determines whether \p type is a emitc.size_t/ssize_t type.
bool isPointerWideType(mlir::Type type);
+/// Determines whether \p type is a valid integer type for SwitchOp.
+bool isIntegerSwitchOperandType(Type type);
+
} // namespace emitc
} // namespace mlir
diff --git a/mlir/include/mlir/Dialect/EmitC/IR/EmitC.td b/mlir/include/mlir/Dialect/EmitC/IR/EmitC.td
index 16fa58ce47013d..4d0cedb91a3915 100644
--- a/mlir/include/mlir/Dialect/EmitC/IR/EmitC.td
+++ b/mlir/include/mlir/Dialect/EmitC/IR/EmitC.td
@@ -54,6 +54,8 @@ def CExpression : NativeOpTrait<"emitc::CExpression">;
def IntegerIndexOrOpaqueType : Type<CPred<"emitc::isIntegerIndexOrOpaqueType($_self)">,
"integer, index or opaque type supported by EmitC">;
def FloatIntegerIndexOrOpaqueType : AnyTypeOf<[EmitCFloatType, IntegerIndexOrOpaqueType]>;
+def IntegerSwitchOperandType : Type<CPred<"emitc::isIntegerSwitchOperandType($_self)">,
+"integer type for switch operation">;
def EmitC_AddOp : EmitC_BinaryOp<"add", [CExpression]> {
let summary = "Addition operation";
@@ -1188,7 +1190,7 @@ def EmitC_AssignOp : EmitC_Op<"assign", []> {
}
def EmitC_YieldOp : EmitC_Op<"yield",
- [Pure, Terminator, ParentOneOf<["ExpressionOp", "IfOp", "ForOp"]>]> {
+ [Pure, Terminator, ParentOneOf<["ExpressionOp", "IfOp", "ForOp", "SwitchOp"]>]> {
let summary = "Block termination operation";
let description = [{
The `emitc.yield` terminates its parent EmitC op's region, optionally yielding
@@ -1302,5 +1304,90 @@ def EmitC_SubscriptOp : EmitC_Op<"subscript", []> {
let assemblyFormat = "$value `[` $indices `]` attr-dict `:` functional-type(operands, results)";
}
+def EmitC_SwitchOp : EmitC_Op<"switch", [RecursiveMemoryEffects,
+ SingleBlockImplicitTerminator<"emitc::YieldOp">,
+ DeclareOpInterfaceMethods<RegionBranchOpInterface,
+ ["getRegionInvocationBounds",
+ "getEntrySuccessorRegions"]>]> {
+ let summary = "Switch operation";
+ let description = [{
+ The `emitc.switch` is a control-flow operation that branches to one of
+ the given regions based on the values of the argument and the cases. The
+ argument is always of type integer (singed or unsigned), excluding i8 and i1.
+ If the type is not specified, then i32 will be used by default.
+
+ The operation always has a "default" region and any number of case regions
+ denoted by integer constants. Control-flow transfers to the case region
+ whose constant value equals the value of the argument. If the argument does
+ not equal any of the case values, control-flow transfer to the "default"
+ region.
+
+ The operation does not return any value. Moreover, case regions and
+ default region must be terminated using the `emitc.yield` operation.
+
+ Example:
+
+ ```mlir
+ // Unspecified type for the cases (default i32 set).
+ %0 = "emitc.variable"(){value = 42 : i32} : () -> i32
+ emitc.switch %0
+ case 2: {
+ %1 = emitc.call_opaque "func_b" () : () -> i32
+ emitc.yield
+ }
+ case 5: {
+ %2 = emitc.call_opaque "func_a" () : () -> i32
+ emitc.yield
+ }
+ default : {
+ %3 = "emitc.variable"(){value = 42.0 : f32} : () -> f32
+ %4 = "emitc.variable"(){value = 42.0 : f32} : () -> f32
+
+ emitc.call_opaque "func2" (%3) : (f32) -> ()
+ emitc.call_opaque "func3" (%3, %4) { args = [1 : index, 0 : index] } : (f32, f32) -> ()
+ emitc.yield
+ }
+ ...
+ // Specified type for the cases (i16).
+ %0 = "emitc.variable"(){value = 42 : i16} : () -> i16
+ emitc.switch %0 : i16
+ case 2: {
+ %1 = emitc.call_opaque "func_b" () : () -> i32
+ emitc.yield
+ }
+ case 5: {
+ %2 = emitc.call_opaque "func_a" () : () -> i32
+ emitc.yield
+ }
+ default : {
+ %3 = "emitc.variable"(){value = 42.0 : f32} : () -> f32
+ %4 = "emitc.variable"(){value = 42.0 : f32} : () -> f32
+
+ emitc.call_opaque "func2" (%3) : (f32) -> ()
+ emitc.call_opaque "func3" (%3, %4) { args = [1 : index, 0 : index] } : (f32, f32) -> ()
+ emitc.yield
+ }
+ ```
+ }];
+
+ let arguments = (ins IntegerSwitchOperandType:$arg, DenseI64ArrayAttr:$cases);
+ let results = (outs);
+ let regions = (region SizedRegion<1>:$defaultRegion,
+ VariadicRegion<SizedRegion<1>>:$caseRegions);
+
+ let extraClassDeclaration = [{
+ /// Get the number of cases.
+ unsigned getNumCases();
+
+ /// Get the default region body.
+ Block &getDefaultBlock();
+
+ /// Get the body of a case region.
+ Block &getCaseBlock(unsigned idx);
+ }];
+
+ let hasCustomAssemblyFormat = 1;
+ let hasVerifier = 1;
+}
#endif // MLIR_DIALECT_EMITC_IR_EMITC
diff --git a/mlir/lib/Dialect/EmitC/IR/EmitC.cpp b/mlir/lib/Dialect/EmitC/IR/EmitC.cpp
index 70e3e728e01950..216911b2870ebc 100644
--- a/mlir/lib/Dialect/EmitC/IR/EmitC.cpp
+++ b/mlir/lib/Dialect/EmitC/IR/EmitC.cpp
@@ -131,6 +131,12 @@ bool mlir::emitc::isPointerWideType(Type type) {
type);
}
+bool mlir::emitc::isIntegerSwitchOperandType(Type type) {
+ auto intType = llvm::dyn_cast<IntegerType>(type);
+ return isSupportedIntegerType(type) && intType.getWidth() != 1 &&
+ intType.getWidth() != 8;
+}
+
/// Check that the type of the initial value is compatible with the operations
/// result type.
static LogicalResult verifyInitializationAttribute(Operation *op,
@@ -1096,6 +1102,208 @@ GetGlobalOp::verifySymbolUses(SymbolTableCollection &symbolTable) {
return success();
}
+//===----------------------------------------------------------------------===//
+// SwitchOp
+//===----------------------------------------------------------------------===//
+
+/// Parse the case regions and values.
+static ParseResult
+parseSwitchCases(OpAsmParser &parser, DenseI64ArrayAttr &cases,
+ SmallVectorImpl<std::unique_ptr<Region>> &caseRegions) {
+ SmallVector<int64_t> caseValues;
+ while (succeeded(parser.parseOptionalKeyword("case"))) {
+ int64_t value;
+ Region ®ion = *caseRegions.emplace_back(std::make_unique<Region>());
+
+ if (parser.parseInteger(value) || parser.parseColon() ||
+ parser.parseRegion(region, /*arguments=*/{}))
+ return failure();
+ caseValues.push_back(value);
+ }
+ cases = parser.getBuilder().getDenseI64ArrayAttr(caseValues);
+ return success();
+}
+
+/// Print the case regions and values.
+static void printSwitchCases(OpAsmPrinter &parser, Operation *op,
+ DenseI64ArrayAttr cases, RegionRange caseRegions) {
+ for (auto [value, region] : llvm::zip(cases.asArrayRef(), caseRegions)) {
+ parser.printNewline();
+ parser << "case " << value << ": ";
+ parser.printRegion(*region, /*printEntryBlockArgs=*/false);
+ }
+ return;
+}
+
+ParseResult SwitchOp::parse(OpAsmParser &parser, OperationState &result) {
+ OpAsmParser::UnresolvedOperand arg;
+ DenseI64ArrayAttr casesAttr;
+ SmallVector<std::unique_ptr<Region>, 2> caseRegionsRegions;
+ std::unique_ptr<Region> defaultRegionRegion = std::make_unique<Region>();
+
+ if (parser.parseOperand(arg))
+ return failure();
+
+ Type argType = parser.getBuilder().getI32Type();
+ // Parse optional type, else assume i32.
+ if (!parser.parseOptionalColon() && parser.parseType(argType))
+ return failure();
+
+ {
+ auto loc = parser.getCurrentLocation();
+ if (parser.parseOptionalAttrDict(result.attributes))
+ return failure();
+
+ if (failed(verifyInherentAttrs(result.name, result.attributes, [&]() {
+ return parser.emitError(loc)
+ << "'" << result.name.getStringRef() << "' op ";
+ })))
+ return failure();
+ }
+ {
+ auto odsResult = parseSwitchCases(parser, casesAttr, caseRegionsRegions);
+
+ if (odsResult)
+ return failure();
+
+ result.getOrAddProperties<SwitchOp::Properties>().cases = casesAttr;
+ }
+ if (parser.parseKeyword("default") || parser.parseColon())
+ return failure();
+
+ if (parser.parseRegion(*defaultRegionRegion))
+ return failure();
+
+ result.addRegion(std::move(defaultRegionRegion));
+ result.addRegions(caseRegionsRegions);
+
+ if (parser.resolveOperand(arg, argType, result.operands))
+ return failure();
+
+ return success();
+}
+
+void SwitchOp::print(OpAsmPrinter &parser) {
+ parser << ' ';
+ parser << getArg();
+ SmallVector<StringRef, 2> elidedAttrs;
+ elidedAttrs.push_back("cases");
+ parser.printOptionalAttrDict((*this)->getAttrs(), elidedAttrs);
+ parser << ' ';
+ printSwitchCases(parser, *this, getCasesAttr(), getCaseRegions());
+ parser.printNewline();
+ parser << "default";
+ parser << ' ';
+ parser.printRegion(getDefaultRegion(), /*printEntryBlockArgs=*/true,
+ /*printBlockTerminators=*/true);
+
+ return;
+}
+
+static LogicalResult verifyRegion(emitc::SwitchOp op, Region ®ion,
+ const Twine &name) {
+ auto yield = dyn_cast<emitc::YieldOp>(region.front().back());
+ if (!yield)
+ return op.emitOpError("expected region to end with emitc.yield, but got ")
+ << region.front().back().getName();
+
+ if (yield.getNumOperands() != 0) {
+ return (op.emitOpError("expected each region to return ")
+ << "0 values, but " << name << " returns "
+ << yield.getNumOperands())
+ .attachNote(yield.getLoc())
+ << "see yield operation here";
+ }
+ return success();
+}
+
+LogicalResult emitc::SwitchOp::verify() {
+ if (!isIntegerSwitchOperandType(getArg().getType()))
+ return emitOpError("unsupported type ") << getArg().getType();
+
+ if (getCases().size() != getCaseRegions().size()) {
+ return emitOpError("has ")
+ << getCaseRegions().size() << " case regions but "
+ << getCases().size() << " case values";
+ }
+
+ DenseSet<int64_t> valueSet;
+ for (int64_t value : getCases())
+ if (!valueSet.insert(value).second)
+ return emitOpError("has duplicate case value: ") << value;
+
+ if (failed(verifyRegion(*this, getDefaultRegion(), "default region")))
+ return failure();
+
+ for (auto [idx, caseRegion] : llvm::enumerate(getCaseRegions()))
+ if (failed(verifyRegion(*this, caseRegion, "case region #" + Twine(idx))))
+ return failure();
+
+ return success();
+}
+
+unsigned emitc::SwitchOp::getNumCases() { return getCases().size(); }
+
+Block &emitc::SwitchOp::getDefaultBlock() { return getDefaultRegion().front(); }
+
+Block &emitc::SwitchOp::getCaseBlock(unsigned idx) {
+ assert(idx < getNumCases() && "case index out-of-bounds");
+ return getCaseRegions()[idx].front();
+}
+
+void SwitchOp::getSuccessorRegions(
+ RegionBranchPoint point, SmallVectorImpl<RegionSuccessor> &successors) {
+ llvm::copy(getRegions(), std::back_inserter(successors));
+ return;
+}
+
+void SwitchOp::getEntrySuccessorRegions(
+ ArrayRef<Attribute> operands,
+ SmallVectorImpl<RegionSuccessor> &successors) {
+ FoldAdaptor adaptor(operands, *this);
+
+ // If a constant was not provided, all regions are possible successors.
+ auto arg = dyn_cast_or_null<IntegerAttr>(adaptor.getArg());
+ if (!arg) {
+ llvm::copy(getRegions(), std::back_inserter(successors));
+ return;
+ }
+
+ // Otherwise, try to find a case with a matching value. If not, the
+ // default region is the only successor.
+ for (auto [caseValue, caseRegion] : llvm::zip(getCases(), getCaseRegions())) {
+ if (caseValue == arg.getInt()) {
+ successors.emplace_back(&caseRegion);
+ return;
+ }
+ }
+ successors.emplace_back(&getDefaultRegion());
+ return;
+}
+
+void SwitchOp::getRegionInvocationBounds(
+ ArrayRef<Attribute> operands, SmallVectorImpl<InvocationBounds> &bounds) {
+ auto operandValue = llvm::dyn_cast_or_null<IntegerAttr>(operands.front());
+ if (!operandValue) {
+ // All regions are invoked at most once.
+ bounds.append(getNumRegions(), InvocationBounds(/*lb=*/0, /*ub=*/1));
+ return;
+ }
+
+ unsigned liveIndex = getNumRegions() - 1;
+ const auto *iteratorToInt = llvm::find(getCases(), operandValue.getInt());
+
+ liveIndex = iteratorToInt != getCases().end()
+ ? std::distance(getCases().begin(), iteratorToInt)
+ : liveIndex;
+
+ for (unsigned regIndex = 0, regNum = getNumRegions(); regIndex < regNum;
+ ++regIndex)
+ bounds.emplace_back(/*lb=*/0, /*ub=*/regIndex == liveIndex);
+
+ return;
+}
+
//===----------------------------------------------------------------------===//
// TableGen'd op method definitions
//===----------------------------------------------------------------------===//
diff --git a/mlir/lib/Target/Cpp/TranslateToCpp.cpp b/mlir/lib/Target/Cpp/TranslateToCpp.cpp
index 1dadb9dd691e70..f315e91015ef5d 100644
--- a/mlir/lib/Target/Cpp/TranslateToCpp.cpp
+++ b/mlir/lib/Target/Cpp/TranslateToCpp.cpp
@@ -449,6 +449,49 @@ static LogicalResult printOperation(CppEmitter &emitter, emitc::SubOp subOp) {
return printBinaryOperation(emitter, operation, "-");
}
+static LogicalResult emitAllExceptLast(CppEmitter &emitter,
+ raw_indented_ostream &os,
+ Region ®ion) {
+ for (Region::OpIterator iteratorOp = region.op_begin(), end = region.op_end();
+ std::next(iteratorOp) != end; ++iteratorOp) {
+ if (failed(emitter.emitOperation(*iteratorOp, /*trailingSemicolon=*/true)))
+ return failure();
+ }
+ os << "break;\n";
+ return success();
+}
+
+static LogicalResult printOperation(CppEmitter &emitter,
+ emitc::SwitchOp switchOp) {
+ raw_indented_ostream &os = emitter.ostream();
+ auto iteratorCaseValues = switchOp.getCases().begin();
+ size_t caseIndex = 0;
+
+ os << "\nswitch(" << emitter.getOrCreateName(switchOp.getArg()) << ") {";
+
+ while (caseIndex != switchOp.getNumCases()) {
+ os << "\ncase "
+ << "(" << *(iteratorCaseValues++) << ")"
+ << ": {\n";
+ os.indent();
+
+ if (failed(emitAllExceptLast(
+ emitter, os, *(switchOp.getCaseBlock(caseIndex++)).getParent())))
+ return failure();
+
+ os.unindent() << "}";
+ }
+
+ os << "\ndefault: {\n";
+ os.indent();
+
+ if (failed(emitAllExceptLast(emitter, os, switchOp.getDefaultRegion())))
+ return failure();
+
+ os.unindent() << "}";
+ return success();
+}
+
static LogicalResult printOperation(CppEmitter &emitter, emitc::CmpOp cmpOp) {
Operation *operation = cmpOp.getOperation();
@@ -998,7 +1041,7 @@ static LogicalResult printFunctionBody(CppEmitter &emitter,
// trailing semicolon is handled within the printOperation function.
bool trailingSemicolon =
!isa<cf::CondBranchOp, emitc::DeclareFuncOp, emitc::ForOp,
- emitc::IfOp, emitc::VerbatimOp>(op);
+ emitc::IfOp, emitc::SwitchOp, emitc::VerbatimOp>(op);
if (failed(emitter.emitOperation(
op, /*trailingSemicolon=*/trailingSemicolon)))
@@ -1509,8 +1552,8 @@ LogicalResult CppEmitter::emitOperation(Operation &op, bool trailingSemicolon) {
emitc::GlobalOp, emitc::IfOp, emitc::IncludeOp,
emitc::LogicalAndOp, emitc::LogicalNotOp, emitc::LogicalOrOp,
emitc::MulOp, emitc::RemOp, emitc::ReturnOp, emitc::SubOp,
- emitc::UnaryMinusOp, emitc::UnaryPlusOp, emitc::VariableOp,
- emitc::VerbatimOp>(
+ emitc::SwitchOp, emitc::UnaryMinusOp, emitc::UnaryPlusOp,
+ emitc::VariableOp, emitc::VerbatimOp>(
[&](auto op) { return printOperation(*this, op); })
// Func ops.
.Case<func::CallOp, func::FuncOp, func::ReturnOp>(
diff --git a/mlir/test/Target/Cpp/invalid.mlir b/mlir/test/Target/Cpp/invalid.mlir
index 513371a09cde1d..1b40ac9a5f3c9b 100644
--- a/mlir/test/Target/Cpp/invalid.mlir
+++ b/mlir/test/Target/Cpp/invalid.mlir
@@ -80,8 +80,137 @@ func.func @array_as_result(%arg: !emitc.array<4xi8>) -> (!emitc.array<4xi8>) {
}
// -----
+
func.func @ptr_to_array() {
// expected-error at +1 {{cannot emit pointer to array type '!emitc.ptr<!emitc.array<9xi16>>'}}
%v = "emitc.variable"(){value = #emitc.opaque<"NULL">} : () -> !emitc.ptr<!emitc.array<9xi16>>
return
}
+
+// -----
+
+func.func @emitc_switch() {
+ %0 = "emitc.variable"(){value = 1 : ui16} : () -> ui16
+
+ // expected-error at +1 {{'emitc.switch' op expected region to end with emitc.yield, but got emitc.call_opaque}}
+ emitc.switch %0 : ui16
+ case 2: {
+ %1 = emitc.call_opaque "func_b" () : () -> i32
+ }
+ case 5: {
+ %2 = emitc.call_opaque "func_a" () : () -> i32
+ emitc.yield
+ }
+ default : {
+ %3 = "emitc.variable"(){value = 42.0 : f32} : () -> f32
+ %4 = "emitc.variable"(){value = 42.0 : f32} : () -> f32
+
+ emitc.call_opaque "func2" (%3) : (f32) -> ()
+ emitc.call_opaque "func3" (%3, %4) { args = [1 : index, 0 : index] } : (f32, f32) -> ()
+ emitc.yield
+ }
+ return
+}
+
+// -----
+
+func.func @emitc_switch() {
+ %0 = "emitc.variable"(){value = 1 : ui16} : () -> ui16
+
+ // expected-error at +1 {{'emitc.switch' op expected region to end with emitc.yield, but got emitc.call_opaque}}
+ emitc.switch %0 : ui16
+ case 2: {
+ %1 = emitc.call_opaque "func_b" () : () -> i32
+ emitc.yield
+ }
+ case 5: {
+ %2 = emitc.call_opaque "func_a" () : () -> i32
+ emitc.yield
+ }
+ default : {
+ %3 = "emitc.variable"(){value = 42.0 : f32} : () -> f32
+ %4 = "emitc.variable"(){value = 42.0 : f32} : () -> f32
+
+ emitc.call_opaque "func2" (%3) : (f32) -> ()
+ emitc.call_opaque "func3" (%3, %4) { args = [1 : index, 0 : index] } : (f32, f32) -> ()
+ }
+ return
+}
+
+// -----
+
+func.func @emitc_switch() {
+ %0 = "emitc.variable"(){value = 1 : i8} : () -> i8
+
+ // expected-error at +1 {{'emitc.switch' op operand #0 must be integer type for switch operation, but got 'i8'}}
+ emitc.switch %0 : i8
+ case 2: {
+ %1 = emitc.call_opaque "func_b" () : () -> i32
+ emitc.yield
+ }
+ case 5: {
+ %2 = emitc.call_opaque "func_a" () : () -> i32
+ emitc.yield
+ }
+ default : {
+ %3 = "emitc.variable"(){value = 42.0 : f32} : () -> f32
+ %4 = "emitc.variable"(){value = 42.0 : f32} : () -> f32
+
+ emitc.call_opaque "func2" (%3) : (f32) -> ()
+ emitc.call_opaque "func3" (%3, %4) { args = [1 : index, 0 : index] } : (f32, f32) -> ()
+ emitc.yield
+ }
+ return
+}
+
+// -----
+
+func.func @emitc_switch() {
+ %0 = "emitc.variable"(){value = 1 : i1} : () -> i1
+
+ // expected-error at +1 {{'emitc.switch' op operand #0 must be integer type for switch operation, but got 'i1'}}
+ emitc.switch %0 : i1
+ case 2: {
+ %1 = emitc.call_opaque "func_b" () : () -> i32
+ emitc.yield
+ }
+ case 5: {
+ %2 = emitc.call_opaque "func_a" () : () -> i32
+ emitc.yield
+ }
+ default : {
+ %3 = "emitc.variable"(){value = 42.0 : f32} : () -> f32
+ %4 = "emitc.variable"(){value = 42.0 : f32} : () -> f32
+
+ emitc.call_opaque "func2" (%3) : (f32) -> ()
+ emitc.call_opaque "func3" (%3, %4) { args = [1 : index, 0 : index] } : (f32, f32) -> ()
+ emitc.yield
+ }
+ return
+}
+
+// -----
+
+func.func @emitc_switch() {
+ %0 = "emitc.variable"(){value = 1 : i16} : () -> i16
+
+ emitc.switch %0 : i16
+ case 2: {
+ %1 = emitc.call_opaque "func_b" () : () -> i32
+ emitc.yield
+ }
+ // expected-error at +1 {{custom op 'emitc.switch' expected integer value}}
+ case : {
+ %2 = emitc.call_opaque "func_a" () : () -> i32
+ emitc.yield
+ }
+ default : {
+ %3 = "emitc.variable"(){value = 42.0 : f32} : () -> f32
+ %4 = "emitc.variable"(){value = 42.0 : f32} : () -> f32
+
+ emitc.call_opaque "func2" (%3) : (f32) -> ()
+ emitc.call_opaque "func3" (%3, %4) { args = [1 : index, 0 : index] } : (f32, f32) -> ()
+ emitc.yield
+ }
+ return
+}
diff --git a/mlir/test/Target/Cpp/switch.mlir b/mlir/test/Target/Cpp/switch.mlir
new file mode 100644
index 00000000000000..251c13bc188f8d
--- /dev/null
+++ b/mlir/test/Target/Cpp/switch.mlir
@@ -0,0 +1,129 @@
+// RUN: mlir-translate -mlir-to-cpp %s | FileCheck %s
+
+// CHECK-LABEL: int32_t v1 = 1...
[truncated]
``````````
</details>
https://github.com/llvm/llvm-project/pull/102331
More information about the Mlir-commits
mailing list