[MLIR][OpenMP] Add conversion support from FIR to LLVM Dialect for OMP DeclareMapper #121005
Conversation
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
TIFitis
requested review from
ergawy,
skatrak,
kiranchandramohan,
agozillon and
kparzysz
llvmbot
added
mlir
flang
|
@llvm/pr-subscribers-flang-openmp @llvm/pr-subscribers-mlir-openmp Author: Akash Banerjee (TIFitis) ChangesAdd conversion support from FIR to LLVM Dialect for OMP DeclareMapper. Full diff: https://github.com/llvm/llvm-project/pull/121005.diff 3 Files Affected:
diff --git a/flang/test/Fir/convert-to-llvm-openmp-and-fir.fir b/flang/test/Fir/convert-to-llvm-openmp-and-fir.fir
index 8e4e1fe824d9f5..82f2aea3ad983c 100644
--- a/flang/test/Fir/convert-to-llvm-openmp-and-fir.fir
+++ b/flang/test/Fir/convert-to-llvm-openmp-and-fir.fir
@@ -936,9 +936,9 @@ func.func @omp_map_info_descriptor_type_conversion(%arg0 : !fir.ref<!fir.box<!fi
%1 = omp.map.info var_ptr(%0 : !fir.llvm_ptr<!fir.ref<i32>>, i32) map_clauses(tofrom) capture(ByRef) -> !fir.llvm_ptr<!fir.ref<i32>> {name = ""}
// CHECK: %[[DESC_MAP:.*]] = omp.map.info var_ptr(%[[ARG_0]] : !llvm.ptr, !llvm.struct<(ptr, i64, i32, i8, i8, i8, i8)>) map_clauses(always, delete) capture(ByRef) members(%[[MEMBER_MAP]] : [0] : !llvm.ptr) -> !llvm.ptr {name = ""}
%2 = omp.map.info var_ptr(%arg0 : !fir.ref<!fir.box<!fir.heap<i32>>>, !fir.box<!fir.heap<i32>>) map_clauses(always, delete) capture(ByRef) members(%1 : [0] : !fir.llvm_ptr<!fir.ref<i32>>) -> !fir.ref<!fir.box<!fir.heap<i32>>> {name = ""}
- // CHECK: omp.target_exit_data map_entries(%[[DESC_MAP]] : !llvm.ptr)
+ // CHECK: omp.target_exit_data map_entries(%[[DESC_MAP]] : !llvm.ptr)
omp.target_exit_data map_entries(%2 : !fir.ref<!fir.box<!fir.heap<i32>>>)
- return
+ return
}
// -----
@@ -956,8 +956,8 @@ func.func @omp_map_info_derived_type_explicit_member_conversion(%arg0 : !fir.ref
%3 = fir.field_index real, !fir.type<_QFderived_type{real:f32,array:!fir.array<10xi32>,int:i32}>
%4 = fir.coordinate_of %arg0, %3 : (!fir.ref<!fir.type<_QFderived_type{real:f32,array:!fir.array<10xi32>,int:i32}>>, !fir.field) -> !fir.ref<f32>
// CHECK: %[[MAP_MEMBER_2:.*]] = omp.map.info var_ptr(%[[GEP_2]] : !llvm.ptr, f32) map_clauses(tofrom) capture(ByRef) -> !llvm.ptr {name = "dtype%real"}
- %5 = omp.map.info var_ptr(%4 : !fir.ref<f32>, f32) map_clauses(tofrom) capture(ByRef) -> !fir.ref<f32> {name = "dtype%real"}
- // CHECK: %[[MAP_PARENT:.*]] = omp.map.info var_ptr(%[[ARG_0]] : !llvm.ptr, !llvm.struct<"_QFderived_type", (f32, array<10 x i32>, i32)>) map_clauses(tofrom) capture(ByRef) members(%[[MAP_MEMBER_1]], %[[MAP_MEMBER_2]] : [2], [0] : !llvm.ptr, !llvm.ptr) -> !llvm.ptr {name = "dtype", partial_map = true}
+ %5 = omp.map.info var_ptr(%4 : !fir.ref<f32>, f32) map_clauses(tofrom) capture(ByRef) -> !fir.ref<f32> {name = "dtype%real"}
+ // CHECK: %[[MAP_PARENT:.*]] = omp.map.info var_ptr(%[[ARG_0]] : !llvm.ptr, !llvm.struct<"_QFderived_type", (f32, array<10 x i32>, i32)>) map_clauses(tofrom) capture(ByRef) members(%[[MAP_MEMBER_1]], %[[MAP_MEMBER_2]] : [2], [0] : !llvm.ptr, !llvm.ptr) -> !llvm.ptr {name = "dtype", partial_map = true}
%6 = omp.map.info var_ptr(%arg0 : !fir.ref<!fir.type<_QFderived_type{real:f32,array:!fir.array<10xi32>,int:i32}>>, !fir.type<_QFderived_type{real:f32,array:!fir.array<10xi32>,int:i32}>) map_clauses(tofrom) capture(ByRef) members(%2, %5 : [2], [0] : !fir.ref<i32>, !fir.ref<f32>) -> !fir.ref<!fir.type<_QFderived_type{real:f32,array:!fir.array<10xi32>,int:i32}>> {name = "dtype", partial_map = true}
// CHECK: omp.target map_entries(%[[MAP_MEMBER_1]] -> %[[ARG_1:.*]], %[[MAP_MEMBER_2]] -> %[[ARG_2:.*]], %[[MAP_PARENT]] -> %[[ARG_3:.*]] : !llvm.ptr, !llvm.ptr, !llvm.ptr) {
omp.target map_entries(%2 -> %arg1, %5 -> %arg2, %6 -> %arg3 : !fir.ref<i32>, !fir.ref<f32>, !fir.ref<!fir.type<_QFderived_type{real:f32,array:!fir.array<10xi32>,int:i32}>>) {
@@ -1275,3 +1275,22 @@ func.func @map_nested_dtype_alloca_mem2(%arg0 : !fir.ref<!fir.type<_QFRecTy{i:f3
}
return
}
+
+// -----
+
+// CHECK-LABEL: omp.declare_mapper @my_mapper : !llvm.struct<"_QFdeclare_mapperTmy_type", (i32)> {
+omp.declare_mapper @my_mapper : !fir.type<_QFdeclare_mapperTmy_type{data:i32}> {
+// CHECK: ^bb0(%[[VAL_0:.*]]: !llvm.ptr):
+^bb0(%0: !fir.ref<!fir.type<_QFdeclare_mapperTmy_type{data:i32}>>):
+// CHECK: %[[VAL_1:.*]] = llvm.mlir.constant(0 : i32) : i32
+ %1 = fir.field_index data, !fir.type<_QFdeclare_mapperTmy_type{data:i32}>
+// CHECK: %[[VAL_2:.*]] = llvm.getelementptr %[[VAL_0]][0, 0] : (!llvm.ptr) -> !llvm.ptr, !llvm.struct<"_QFdeclare_mapperTmy_type", (i32)>
+ %2 = fir.coordinate_of %0, %1 : (!fir.ref<!fir.type<_QFdeclare_mapperTmy_type{data:i32}>>, !fir.field) -> !fir.ref<i32>
+// CHECK: %[[VAL_3:.*]] = omp.map.info var_ptr(%[[VAL_2]] : !llvm.ptr, i32) map_clauses(tofrom) capture(ByRef) -> !llvm.ptr {name = "var%[[VAL_4:.*]]"}
+ %3 = omp.map.info var_ptr(%2 : !fir.ref<i32>, i32) map_clauses(tofrom) capture(ByRef) -> !fir.ref<i32> {name = "var%data"}
+// CHECK: %[[VAL_5:.*]] = omp.map.info var_ptr(%[[VAL_0]] : !llvm.ptr, !llvm.struct<"_QFdeclare_mapperTmy_type", (i32)>) map_clauses(tofrom) capture(ByRef) members(%[[VAL_3]] : [0] : !llvm.ptr) -> !llvm.ptr {name = "var", partial_map = true}
+ %4 = omp.map.info var_ptr(%0 : !fir.ref<!fir.type<_QFdeclare_mapperTmy_type{data:i32}>>, !fir.type<_QFdeclare_mapperTmy_type{data:i32}>) map_clauses(tofrom) capture(ByRef) members(%3 : [0] : !fir.ref<i32>) -> !fir.ref<!fir.type<_QFdeclare_mapperTmy_type{data:i32}>> {name = "var", partial_map = true}
+// CHECK: omp.declare_mapper_info map_entries(%[[VAL_5]], %[[VAL_3]] : !llvm.ptr, !llvm.ptr)
+ omp.declare_mapper_info map_entries(%4, %3 : !fir.ref<!fir.type<_QFdeclare_mapperTmy_type{data:i32}>>, !fir.ref<i32>)
+// CHECK: }
+}
diff --git a/mlir/lib/Conversion/OpenMPToLLVM/OpenMPToLLVM.cpp b/mlir/lib/Conversion/OpenMPToLLVM/OpenMPToLLVM.cpp
index 58fd3d565fce50..352756db1d4d32 100644
--- a/mlir/lib/Conversion/OpenMPToLLVM/OpenMPToLLVM.cpp
+++ b/mlir/lib/Conversion/OpenMPToLLVM/OpenMPToLLVM.cpp
@@ -186,6 +186,32 @@ struct MapInfoOpConversion : public ConvertOpToLLVMPattern<omp::MapInfoOp> {
}
};
+struct DeclMapperOpConversion
+ : public ConvertOpToLLVMPattern<omp::DeclareMapperOp> {
+ using ConvertOpToLLVMPattern<omp::DeclareMapperOp>::ConvertOpToLLVMPattern;
+ LogicalResult
+ matchAndRewrite(omp::DeclareMapperOp curOp, OpAdaptor adaptor,
+ ConversionPatternRewriter &rewriter) const override {
+ const TypeConverter *converter = ConvertToLLVMPattern::getTypeConverter();
+ SmallVector<NamedAttribute> newAttrs;
+ newAttrs.emplace_back(curOp.getSymNameAttrName(), curOp.getSymNameAttr());
+ newAttrs.emplace_back(
+ curOp.getVarTypeAttrName(),
+ TypeAttr::get(converter->convertType(curOp.getVarType())));
+
+ auto newOp = rewriter.create<omp::DeclareMapperOp>(
+ curOp.getLoc(), TypeRange(), adaptor.getOperands(), newAttrs);
+ rewriter.inlineRegionBefore(curOp.getRegion(), newOp.getRegion(),
+ newOp.getRegion().end());
+ if (failed(rewriter.convertRegionTypes(&newOp.getRegion(),
+ *this->getTypeConverter())))
+ return failure();
+
+ rewriter.eraseOp(curOp);
+ return success();
+ }
+};
+
template <typename OpType>
struct MultiRegionOpConversion : public ConvertOpToLLVMPattern<OpType> {
using ConvertOpToLLVMPattern<OpType>::ConvertOpToLLVMPattern;
@@ -225,20 +251,21 @@ void mlir::configureOpenMPToLLVMConversionLegality(
ConversionTarget &target, const LLVMTypeConverter &typeConverter) {
target.addDynamicallyLegalOp<
omp::AtomicReadOp, omp::AtomicWriteOp, omp::CancellationPointOp,
- omp::CancelOp, omp::CriticalDeclareOp, omp::FlushOp, omp::MapBoundsOp,
- omp::MapInfoOp, omp::OrderedOp, omp::TargetEnterDataOp,
- omp::TargetExitDataOp, omp::TargetUpdateOp, omp::ThreadprivateOp,
- omp::YieldOp>([&](Operation *op) {
+ omp::CancelOp, omp::CriticalDeclareOp, omp::DeclareMapperInfoOp,
+ omp::FlushOp, omp::MapBoundsOp, omp::MapInfoOp, omp::OrderedOp,
+ omp::TargetEnterDataOp, omp::TargetExitDataOp, omp::TargetUpdateOp,
+ omp::ThreadprivateOp, omp::YieldOp>([&](Operation *op) {
return typeConverter.isLegal(op->getOperandTypes()) &&
typeConverter.isLegal(op->getResultTypes());
});
target.addDynamicallyLegalOp<
- omp::AtomicUpdateOp, omp::CriticalOp, omp::DeclareReductionOp,
- omp::DistributeOp, omp::LoopNestOp, omp::LoopOp, omp::MasterOp,
- omp::OrderedRegionOp, omp::ParallelOp, omp::PrivateClauseOp,
- omp::SectionOp, omp::SectionsOp, omp::SimdOp, omp::SingleOp,
- omp::TargetDataOp, omp::TargetOp, omp::TaskgroupOp, omp::TaskloopOp,
- omp::TaskOp, omp::TeamsOp, omp::WsloopOp>([&](Operation *op) {
+ omp::AtomicUpdateOp, omp::CriticalOp, omp::DeclareMapperOp,
+ omp::DeclareReductionOp, omp::DistributeOp, omp::LoopNestOp, omp::LoopOp,
+ omp::MasterOp, omp::OrderedRegionOp, omp::ParallelOp,
+ omp::PrivateClauseOp, omp::SectionOp, omp::SectionsOp, omp::SimdOp,
+ omp::SingleOp, omp::TargetDataOp, omp::TargetOp, omp::TaskgroupOp,
+ omp::TaskloopOp, omp::TaskOp, omp::TeamsOp,
+ omp::WsloopOp>([&](Operation *op) {
return std::all_of(op->getRegions().begin(), op->getRegions().end(),
[&](Region ®ion) {
return typeConverter.isLegal(®ion);
@@ -257,12 +284,13 @@ void mlir::populateOpenMPToLLVMConversionPatterns(LLVMTypeConverter &converter,
[&](omp::MapBoundsType type) -> Type { return type; });
patterns.add<
- AtomicReadOpConversion, MapInfoOpConversion,
+ AtomicReadOpConversion, DeclMapperOpConversion, MapInfoOpConversion,
MultiRegionOpConversion<omp::DeclareReductionOp>,
MultiRegionOpConversion<omp::PrivateClauseOp>,
RegionLessOpConversion<omp::CancellationPointOp>,
RegionLessOpConversion<omp::CancelOp>,
RegionLessOpConversion<omp::CriticalDeclareOp>,
+ RegionLessOpConversion<omp::DeclareMapperInfoOp>,
RegionLessOpConversion<omp::OrderedOp>,
RegionLessOpConversion<omp::TargetEnterDataOp>,
RegionLessOpConversion<omp::TargetExitDataOp>,
diff --git a/mlir/test/Conversion/OpenMPToLLVM/convert-to-llvmir.mlir b/mlir/test/Conversion/OpenMPToLLVM/convert-to-llvmir.mlir
index 4f37dd16b44dd3..3de0e064a5fd0c 100644
--- a/mlir/test/Conversion/OpenMPToLLVM/convert-to-llvmir.mlir
+++ b/mlir/test/Conversion/OpenMPToLLVM/convert-to-llvmir.mlir
@@ -607,3 +607,16 @@ func.func @omp_taskloop(%arg0: index, %arg1 : memref<i32>) {
}
return
}
+
+// -----
+
+// CHECK-LABEL: omp.declare_mapper @my_mapper : !llvm.struct<"_QFdeclare_mapperTmy_type", (i32)> {
+omp.declare_mapper @my_mapper : !llvm.struct<"_QFdeclare_mapperTmy_type", (i32)> {
+^bb0(%arg0: !llvm.ptr):
+ %0 = llvm.mlir.constant(0 : i32) : i32
+ %1 = llvm.getelementptr %arg0[0, 0] : (!llvm.ptr) -> !llvm.ptr, !llvm.struct<"_QFdeclare_mapperTmy_type", (i32)>
+ %2 = omp.map.info var_ptr(%1 : !llvm.ptr, i32) map_clauses(tofrom) capture(ByRef) -> !llvm.ptr {name = "var%data"}
+ %3 = omp.map.info var_ptr(%arg0 : !llvm.ptr, !llvm.struct<"_QFdeclare_mapperTmy_type", (i32)>) map_clauses(tofrom) capture(ByRef) members(%2 : [0] : !llvm.ptr) -> !llvm.ptr {name = "var", partial_map = true}
+ // CHECK: omp.declare_mapper_info map_entries(%{{.*}}, %{{.*}} : !llvm.ptr, !llvm.ptr)
+ omp.declare_mapper_info map_entries(%3, %2 : !llvm.ptr, !llvm.ptr)
+}
|
ergawy
reviewed
Jan 6, 2025
TIFitis
added a commit
that referenced
this pull request
Jan 28, 2025
…pers This patch adds OpenMPToLLVMIRTranslation support for the OpenMP Declare Mapper directive. Since both MLIR and Clang now support custom mappers, I've made the relative params required instead of optional as well. Depends on #121005
TIFitis
added a commit
that referenced
this pull request
Jan 31, 2025
…pers This patch adds OpenMPToLLVMIRTranslation support for the OpenMP Declare Mapper directive. Since both MLIR and Clang now support custom mappers, I've made the relative params required instead of optional as well. Depends on #121005
TIFitis
added a commit
that referenced
this pull request
Jan 31, 2025
…pers This patch adds OpenMPToLLVMIRTranslation support for the OpenMP Declare Mapper directive. Since both MLIR and Clang now support custom mappers, I've made the relative params required instead of optional as well. Depends on #121005
TIFitis
force-pushed
the
users/akash/mapper_clause_lowering
branch
from
ed3cb3d to
5f892bb
Compare
TIFitis
force-pushed
the
users/akash/mapper_llvm_dialect
branch
from
8a90d20 to
5fe7a97
Compare
TIFitis
added a commit
that referenced
this pull request
Jan 31, 2025
…pers This patch adds OpenMPToLLVMIRTranslation support for the OpenMP Declare Mapper directive. Since both MLIR and Clang now support custom mappers, I've made the relative params required instead of optional as well. Depends on #121005
TIFitis
force-pushed
the
users/akash/mapper_clause_lowering
branch
from
5f892bb to
802c3df
Compare
TIFitis
force-pushed
the
users/akash/mapper_llvm_dialect
branch
from
5fe7a97 to
3302435
Compare
TIFitis
added a commit
that referenced
this pull request
Feb 7, 2025
…pers This patch adds OpenMPToLLVMIRTranslation support for the OpenMP Declare Mapper directive. Since both MLIR and Clang now support custom mappers, I've made the relative params required instead of optional as well. Depends on #121005
TIFitis
force-pushed
the
users/akash/mapper_clause_lowering
branch
from
802c3df to
1b3e235
Compare
TIFitis
force-pushed
the
users/akash/mapper_llvm_dialect
branch
from
3302435 to
8b929a8
Compare
TIFitis
added a commit
that referenced
this pull request
Feb 10, 2025
…pers This patch adds OpenMPToLLVMIRTranslation support for the OpenMP Declare Mapper directive. Since both MLIR and Clang now support custom mappers, I've made the relative params required instead of optional as well. Depends on #121005
TIFitis
force-pushed
the
users/akash/mapper_clause_lowering
branch
from
1b3e235 to
c5b4f50
Compare
TIFitis
force-pushed
the
users/akash/mapper_llvm_dialect
branch
from
8b929a8 to
af3dcbf
Compare
TIFitis
force-pushed
the
users/akash/mapper_clause_lowering
branch
from
c5b4f50 to
e5d0cd9
Compare
TIFitis
force-pushed
the
users/akash/mapper_llvm_dialect
branch
from
af3dcbf to
3ddc972
Compare
TIFitis
added a commit
that referenced
this pull request
Feb 12, 2025
…pers This patch adds OpenMPToLLVMIRTranslation support for the OpenMP Declare Mapper directive. Since both MLIR and Clang now support custom mappers, I've made the relative params required instead of optional as well. Depends on #121005
TIFitis
force-pushed
the
users/akash/mapper_clause_lowering
branch
from
e5d0cd9 to
20a5496
Compare
TIFitis
force-pushed
the
users/akash/mapper_llvm_dialect
branch
from
3ddc972 to
55e38d7
Compare
TIFitis
added a commit
that referenced
this pull request
Feb 18, 2025
…pers This patch adds OpenMPToLLVMIRTranslation support for the OpenMP Declare Mapper directive. Since both MLIR and Clang now support custom mappers, I've made the relative params required instead of optional as well. Depends on #121005
TIFitis
force-pushed
the
users/akash/mapper_clause_lowering
branch
from
20a5496 to
862fda5
Compare
TIFitis
force-pushed
the
users/akash/mapper_llvm_dialect
branch
from
55e38d7 to
ffd2012
Compare
TIFitis
added a commit
that referenced
this pull request
Feb 18, 2025
…pers This patch adds OpenMPToLLVMIRTranslation support for the OpenMP Declare Mapper directive. Since both MLIR and Clang now support custom mappers, I've made the relative params required instead of optional as well. Depends on #121005
wldfngrs
pushed a commit
to wldfngrs/llvm-project
that referenced
this pull request
Feb 19, 2025
…P DeclareMapper (llvm#121005) Add conversion support from FIR to LLVM Dialect for OMP DeclareMapper. Depends on llvm#121001
wldfngrs
pushed a commit
to wldfngrs/llvm-project
that referenced
this pull request
Feb 19, 2025
…pers (llvm#124746) This patch adds OpenMPToLLVMIRTranslation support for the OpenMP Declare Mapper directive. Since both MLIR and Clang now support custom mappers, I've changed the respective function params to no longer be optional as well. Depends on llvm#121005
searlmc1
pushed a commit
to ROCm/llvm-project
that referenced
this pull request
Feb 20, 2025
land and revert: 785a5b4 [MLIR][OpenMP] Add LLVM translation support for OpenMP UserDefinedMappers (llvm#124746) d6ab12c [MLIR][OpenMP] Add conversion support from FIR to LLVM Dialect for OMP DeclareMapper (llvm#121005) 886b2ed [MLIR][OpenMP] Add Lowering support for OpenMP custom mappers in map clause (llvm#121001) ee17955 [MLIR][OpenMP] Add OMP Mapper field to MapInfoOp (llvm#120994)
TIFitis
added a commit
to TIFitis/llvm-project
that referenced
this pull request
Feb 20, 2025
…P DeclareMapper (llvm#121005) Add conversion support from FIR to LLVM Dialect for OMP DeclareMapper. Depends on llvm#121001
TIFitis
added a commit
to TIFitis/llvm-project
that referenced
this pull request
Feb 20, 2025
…pers (llvm#124746) This patch adds OpenMPToLLVMIRTranslation support for the OpenMP Declare Mapper directive. Since both MLIR and Clang now support custom mappers, I've changed the respective function params to no longer be optional as well. Depends on llvm#121005
searlmc1
pushed a commit
to ROCm/llvm-project
that referenced
this pull request
Feb 22, 2025
…P DeclareMapper (llvm#121005) Add conversion support from FIR to LLVM Dialect for OMP DeclareMapper. Depends on llvm#121001
searlmc1
pushed a commit
to ROCm/llvm-project
that referenced
this pull request
Feb 22, 2025
…pers (llvm#124746) This patch adds OpenMPToLLVMIRTranslation support for the OpenMP Declare Mapper directive. Since both MLIR and Clang now support custom mappers, I've changed the respective function params to no longer be optional as well. Depends on llvm#121005
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
Add conversion support from FIR to LLVM Dialect for OMP DeclareMapper.
Depends on #121001