[MLIR][OpenMP] Introduce overlapped record type map support

This PR introduces a new additional type of map lowering for record types that Clang currently supports, in which a user can map a top-level record type and then individual members with different mapping, effectively creating a sort of "overlapping" mapping that we attempt to cut around.

This is currently most predominantly used in Fortran, when mapping descriptors and there data, we map the descriptor and its data with separate map modifiers and "cut around" the pointer data, so that wedo not overwrite it unless the runtime deems it a neccesary action based on its reference counting mechanism. However, it is a mechanism that will come in handy/trigger when a user explitily maps a record type (derived type or structure) and then explicitly maps a member with a different map type.

These additions were predominantly in the OpenMPToLLVMIRTranslation.cpp file and phase, however, one Flang test that checks end-to-end IR compilation (as far as we care for now at least) was altered.

2/3 required PRs to enable declare target to mapping, should look at PR 3/3 to check for full green passes (this one will fail a number due to some dependencies).

Co-authored-by: Raghu Maddhipatla [email protected]

Author: agozillon

flang/include/flang/Optimizer/Builder/FIRBuilder.h

@@ -969,7 +969,6 @@ void genLifetimeEnd(mlir::OpBuilder &builder, mlir::Location loc,
 mlir::Value getDescriptorWithNewBaseAddress(fir::FirOpBuilder &builder,
                                             mlir::Location loc, mlir::Value box,
                                             mlir::Value newAddr);
-
 } // namespace fir::factory
 
 #endif // FORTRAN_OPTIMIZER_BUILDER_FIRBUILDER_H

llvm/lib/Frontend/OpenMP/OMPIRBuilder.cpp

@@ -7585,8 +7585,12 @@ static Expected<Function *> createOutlinedFunction(
     // preceding mapped arguments that refer to the same global that may be
     // seperate segments. To prevent this, we defer global processing until all
     // other processing has been performed.
+<<<<<<< HEAD
     if (llvm::isa<llvm::GlobalValue, llvm::GlobalObject, llvm::GlobalVariable>(
             removeASCastIfPresent(Input))) {
+=======
+    if (llvm::isa<llvm::GlobalValue, llvm::GlobalObject, llvm::GlobalVariable>(removeASCastIfPresent(Input))) {
+>>>>>>> 1d1536f712a6 ([Flang][OpenMP] Additional global address space modifications for device)
       DeferredReplacement.push_back(std::make_pair(Input, InputCopy));
       continue;
     }

mlir/lib/Target/LLVMIR/Dialect/OpenMP/OpenMPToLLVMIRTranslation.cpp

@@ -3992,6 +3992,9 @@ static void collectMapDataFromMapOperands(
   }
 }
 
+// remember to remove the movebefore that might have been added in one of the
+// PRs and is now deprecated...
+
 static int getMapDataMemberIdx(MapInfoData &mapData, omp::MapInfoOp memberOp) {
   auto *res = llvm::find(mapData.MapClause, memberOp);
   assert(res != mapData.MapClause.end() &&
@@ -4049,7 +4052,7 @@ static void sortMapIndices(llvm::SmallVectorImpl<size_t> &indices,
 
 static omp::MapInfoOp getFirstOrLastMappedMemberPtr(omp::MapInfoOp mapInfo,
                                                     bool first) {
-  ArrayAttr indexAttr = mapInfo.getMembersIndexAttr();
+  mlir::ArrayAttr indexAttr = mapInfo.getMembersIndexAttr();
   // Only 1 member has been mapped, we can return it.
   if (indexAttr.size() == 1)
     return cast<omp::MapInfoOp>(mapInfo.getMembers()[0].getDefiningOp());