[fir] Add fir.gentypedesc conversion

[clementval]

Summary

Add conversion pattern for the GenTypeDescOp.
Convert to a global constant with an addressof.

This patch is part of the upstreaming effort from fir-dev branch.

Co-authored-by: Eric Schweitz <[e-mail]>
Co-authored-by: Jean Perier <[e-mail]>

[kiranchandramohan]

flang/lib/Optimizer/CodeGen/CodeGen.cpp

641

Nit: Can you add a bit more documentation on the conversion (global + addressof) and why is it a global (when it was not in FIR).

[schweitz]

flang/lib/Optimizer/CodeGen/CodeGen.cpp

641

This prototype lowering is alluding to one possible implementation.

In FIR, one need not know the exact layout and contents of a "type descriptor". It is an abstraction that can be used to perform type tests, like "isa", perform virtual dispatch (for type bound procedures), etc. (This is like the box abstraction, which allows a box to be lowered to any number of "descriptor" formats, depending on the target runtime, etc.)

When lowering from FIR to LLVM, some sort of (constant) data structure must either be present (and referenced) or be created. Initially there was nothing, so the prototype created a dummy implementation, which is what we have here. The front end has since evolved and is generating its own data structures. (These can be found in the FIR dumps of tests with derived types, for example.) It is not entirely clear how this will be brought together.

This prototype could be replaced with a TODO.

[clementval]

Replace the prototype lowering with not implmented yet.

[kiranchandramohan]

LGTM.

Nit: Are the additional headers required?

[clementval]

In D113766#3134319, @kiranchandramohan wrote:

LGTM.

Nit: Are the additional headers required?

Not anymore. I'm gonna remove them before pushing.

[clementval]

Rebase + remove extra includes

flang/lib/Optimizer/CodeGen/CodeGen.cpp

 //===-- CodeGen.cpp -- bridge to lower to LLVM ----------------------------===//
 //
 // Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions.
 // See https://llvm.org/LICENSE.txt for license information.
 // SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception
 //
 //===----------------------------------------------------------------------===//
 //
 // Coding style: https://mlir.llvm.org/getting_started/DeveloperGuide/
 //
 //===----------------------------------------------------------------------===//
 
 #include "flang/Optimizer/CodeGen/CodeGen.h"
 #include "PassDetail.h"
 #include "flang/ISO_Fortran_binding.h"
+#include "flang/Optimizer/Builder/FIRBuilder.h"
 #include "flang/Optimizer/Dialect/FIRAttr.h"
 #include "flang/Optimizer/Dialect/FIROps.h"
+#include "flang/Optimizer/Support/InternalNames.h"
 #include "mlir/Conversion/ArithmeticToLLVM/ArithmeticToLLVM.h"
 #include "mlir/Conversion/LLVMCommon/Pattern.h"
 #include "mlir/Conversion/StandardToLLVM/ConvertStandardToLLVM.h"
 #include "mlir/IR/BuiltinTypes.h"
 #include "mlir/IR/Matchers.h"
 #include "mlir/Pass/Pass.h"
 #include "llvm/ADT/ArrayRef.h"
 
struct DTEntryOpConversion : public FIROpConversion<fir::DTEntryOp> {
   mlir::LogicalResult
   matchAndRewrite(fir::DTEntryOp dtEnt, OpAdaptor adaptor,
                   mlir::ConversionPatternRewriter &rewriter) const override {
     return rewriter.notifyMatchFailure(
         dtEnt, "fir.dt_entry codegen is not implemented yet");
   }
 };
 
+/// Lower `fir.gentypedesc` to a global constant.
+struct GenTypeDescOpConversion : public FIROpConversion<fir::GenTypeDescOp> {
+  using FIROpConversion::FIROpConversion;
+
+  mlir::LogicalResult
+  matchAndRewrite(fir::GenTypeDescOp gentypedesc, OpAdaptor adaptor,
+                  mlir::ConversionPatternRewriter &rewriter) const override {
+    return rewriter.notifyMatchFailure(
+        gentypedesc, "fir.fir.gentypedesc codegen is not implemented yet");
+  }
+};
+
 /// Lower `fir.has_value` operation to `llvm.return` operation.
 struct HasValueOpConversion : public FIROpConversion<fir::HasValueOp> {
   using FIROpConversion::FIROpConversion;
 
   mlir::LogicalResult
   matchAndRewrite(fir::HasValueOp op, OpAdaptor adaptor,
                   mlir::ConversionPatternRewriter &rewriter) const override {
     rewriter.replaceOpWithNewOp<LLVM::ReturnOp>(op, adaptor.getOperands());
void runOnOperation() override final {
     mlir::OwningRewritePatternList pattern(context);
     pattern.insert<
         AbsentOpConversion, AddcOpConversion, AddrOfOpConversion,
         AllocaOpConversion, BoxAddrOpConversion, BoxDimsOpConversion,
         BoxEleSizeOpConversion, BoxIsAllocOpConversion, BoxIsArrayOpConversion,
         BoxIsPtrOpConversion, BoxRankOpConversion, CallOpConversion,
         ConvertOpConversion, DispatchOpConversion, DispatchTableOpConversion,
         DTEntryOpConversion, DivcOpConversion, ExtractValueOpConversion,
-        HasValueOpConversion, GlobalOpConversion, InsertOnRangeOpConversion,
-        InsertValueOpConversion, IsPresentOpConversion, LoadOpConversion,
-        NegcOpConversion, MulcOpConversion, SelectCaseOpConversion,
-        SelectOpConversion, SelectRankOpConversion, StoreOpConversion,
-        SubcOpConversion, UndefOpConversion, UnreachableOpConversion,
-        ZeroOpConversion>(typeConverter);
+        HasValueOpConversion, GenTypeDescOpConversion, GlobalOpConversion,
+        InsertOnRangeOpConversion, InsertValueOpConversion,
+        IsPresentOpConversion, LoadOpConversion, NegcOpConversion,
+        MulcOpConversion, SelectCaseOpConversion, SelectOpConversion,
+        SelectRankOpConversion, StoreOpConversion, SubcOpConversion,
+        UndefOpConversion, UnreachableOpConversion, ZeroOpConversion>(
+        typeConverter);
     mlir::populateStdToLLVMConversionPatterns(typeConverter, pattern);
     mlir::arith::populateArithmeticToLLVMConversionPatterns(typeConverter,
                                                             pattern);
     mlir::ConversionTarget target{*context};
     target.addLegalDialect<mlir::LLVM::LLVMDialect>();
 
     // required NOPs for applying a full conversion
     target.addLegalOp<mlir::ModuleOp>();

flang/test/Fir/convert-to-llvm-invalid.fir

^bb1:
   %c1_i32 = arith.constant 1 : i32
   br ^bb3
 ^bb2:
   %c2_i32 = arith.constant 2 : i32
   br ^bb3
 ^bb3:
   return
 }
+
+// -----
+
+// Test `fir.gentypedesc` conversion failure. Not implemented yet.
+
+func @gentypedesc() {
+  // expected-error@+1{{failed to legalize operation 'fir.gentypedesc'}}
+  %0 = fir.gentypedesc !fir.type<derived3>
+  return
+}