Skip to content

[Flang][OpenMP] Make boxed procedure pass aware of OpenMP private ops #118261

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
merged 4 commits into from
Dec 9, 2024
Merged
Show file tree
Hide file tree
Changes from all commits
Commits
File filter

Filter by extension

Filter by extension

Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
17 changes: 15 additions & 2 deletions flang/lib/Optimizer/CodeGen/BoxedProcedure.cpp
Original file line number Diff line number Diff line change
Expand Up @@ -103,6 +103,8 @@ class BoxprocTypeRewriter : public mlir::TypeConverter {
return needsConversion(unwrapRefType(ty));
if (auto t = mlir::dyn_cast<SequenceType>(ty))
return needsConversion(unwrapSequenceType(ty));
if (auto t = mlir::dyn_cast<TypeDescType>(ty))
return needsConversion(t.getOfTy());
return false;
}

Expand Down Expand Up @@ -167,6 +169,9 @@ class BoxprocTypeRewriter : public mlir::TypeConverter {
rec.finalize(ps, cs);
return rec;
});
addConversion([&](TypeDescType ty) {
return TypeDescType::get(convertType(ty.getOfTy()));
});
addArgumentMaterialization(materializeProcedure);
addSourceMaterialization(materializeProcedure);
addTargetMaterialization(materializeProcedure);
Expand Down Expand Up @@ -220,7 +225,6 @@ class BoxedProcedurePass
auto *context = &getContext();
mlir::IRRewriter rewriter(context);
BoxprocTypeRewriter typeConverter(mlir::UnknownLoc::get(context));
mlir::Dialect *firDialect = context->getLoadedDialect("fir");
getModule().walk([&](mlir::Operation *op) {
bool opIsValid = true;
typeConverter.setLocation(op->getLoc());
Expand Down Expand Up @@ -366,13 +370,22 @@ class BoxedProcedurePass
index, toTy, index.getFieldId(), toOnTy, index.getTypeparams());
opIsValid = false;
}
} else if (op->getDialect() == firDialect) {
} else {
rewriter.startOpModification(op);
// Convert the operands if needed
for (auto i : llvm::enumerate(op->getResultTypes()))
if (typeConverter.needsConversion(i.value())) {
auto toTy = typeConverter.convertType(i.value());
op->getResult(i.index()).setType(toTy);
}

// Convert the type attributes if needed
for (const mlir::NamedAttribute &attr : op->getAttrDictionary())
if (auto tyAttr = llvm::dyn_cast<mlir::TypeAttr>(attr.getValue()))
if (typeConverter.needsConversion(tyAttr.getValue())) {
auto toTy = typeConverter.convertType(tyAttr.getValue());
op->setAttr(attr.getName(), mlir::TypeAttr::get(toTy));
}
rewriter.finalizeOpModification(op);
}
// Ensure block arguments are updated if needed.
Expand Down
3 changes: 3 additions & 0 deletions flang/test/Fir/boxproc-2.fir
Original file line number Diff line number Diff line change
Expand Up @@ -13,6 +13,9 @@ func.func private @test3(!fir.boxproc<() -> (!fir.type<a{x:i32, y:f64}>)>) -> no
//CHECK-LABEL: func.func private @test5(((i32) -> f32) -> ())
func.func private @test5(!fir.boxproc<(!fir.boxproc<(i32) -> (f32)>) -> ()>)

//CHECK-LABEL: func.func private @test6(!fir.tdesc<!fir.type<_QMcartesian_2d_objectsTcartesian_2d_objectUnboxProc{base_pde_object:!fir.type<_QMbase_pde_objectsTbase_pde_objectUnboxProc{process_p:(!fir.class<!fir.type<_QMbase_pde_objectsTbase_pde_objectUnboxProc>>) -> !fir.class<!fir.heap<!fir.type<_QMbase_pde_objectsTbase_pde_objectUnboxProc>>>,source_p:(!fir.class<!fir.type<_QMbase_pde_objectsTbase_pde_objectUnboxProc>>, !fir.ref<f32>) -> !fir.class<!fir.heap<!fir.type<_QMbase_pde_objectsTbase_pde_objectUnboxProc>>>}>,c:!fir.box<!fir.heap<!fir.array<?x?xf32>>>,dx:f32,dy:f32}>>) -> !fir.ref<none>
func.func private @test6(!fir.tdesc<!fir.type<_QMcartesian_2d_objectsTcartesian_2d_object{base_pde_object:!fir.type<_QMbase_pde_objectsTbase_pde_object{process_p:!fir.boxproc<(!fir.class<!fir.type<_QMbase_pde_objectsTbase_pde_object>>) -> !fir.class<!fir.heap<!fir.type<_QMbase_pde_objectsTbase_pde_object>>>>,source_p:!fir.boxproc<(!fir.class<!fir.type<_QMbase_pde_objectsTbase_pde_object>>, !fir.ref<f32>) -> !fir.class<!fir.heap<!fir.type<_QMbase_pde_objectsTbase_pde_object>>>>}>,c:!fir.box<!fir.heap<!fir.array<?x?xf32>>>,dx:f32,dy:f32}>>) -> !fir.ref<none>

// CHECK-LABEL: func.func @proc_pointer_component(
// CHECK-SAME: %[[VAL_0:.*]]: (!fir.ref<i32>) -> f32,
// CHECK-SAME: %[[VAL_1:.*]]: !fir.ref<i32>) {
Expand Down
87 changes: 87 additions & 0 deletions flang/test/Fir/boxproc-openmp.fir
Original file line number Diff line number Diff line change
@@ -0,0 +1,87 @@
// RUN: fir-opt --boxed-procedure %s | FileCheck %s
// Test the boxed procedure pass with OpenMP declaration operations.
// Check minimally, only arguments, yields and the private types.

// Test a private declaration with one region (alloc)
//CHECK: omp.private {type = private} @_QFsub1Et1_private_ref_rec__QFsub1Tt : !fir.ref<!fir.type<_QFsub1TtUnboxProc{p1:() -> ()}>> alloc {
omp.private {type = private} @_QFsub1Et1_private_ref_rec__QFsub1Tt : !fir.ref<!fir.type<_QFsub1Tt{p1:!fir.boxproc<() -> ()>}>> alloc {
//CHECK: ^bb0(%{{.*}}: !fir.ref<!fir.type<_QFsub1TtUnboxProc{p1:() -> ()}>>):
^bb0(%arg0: !fir.ref<!fir.type<_QFsub1Tt{p1:!fir.boxproc<() -> ()>}>>):
%c1_i32 = arith.constant 1 : i32
%0 = fir.alloca !fir.type<_QFsub1Tt{p1:!fir.boxproc<() -> ()>}> {bindc_name = "t1", pinned, uniq_name = "_QFsub1Et1"}
%1 = fir.declare %0 {uniq_name = "_QFsub1Et1"} : (!fir.ref<!fir.type<_QFsub1Tt{p1:!fir.boxproc<() -> ()>}>>) -> !fir.ref<!fir.type<_QFsub1Tt{p1:!fir.boxproc<() -> ()>}>>
%2 = fir.embox %1 : (!fir.ref<!fir.type<_QFsub1Tt{p1:!fir.boxproc<() -> ()>}>>) -> !fir.box<!fir.type<_QFsub1Tt{p1:!fir.boxproc<() -> ()>}>>
%3 = fir.address_of(@_QQclXea6256ba131ddd9c2210e68030a0edd3) : !fir.ref<!fir.char<1,49>>
%4 = fir.convert %2 : (!fir.box<!fir.type<_QFsub1Tt{p1:!fir.boxproc<() -> ()>}>>) -> !fir.box<none>
%5 = fir.convert %3 : (!fir.ref<!fir.char<1,49>>) -> !fir.ref<i8>
%6 = fir.call @_FortranAInitialize(%4, %5, %c1_i32) fastmath<contract> : (!fir.box<none>, !fir.ref<i8>, i32) -> none
//CHECK: omp.yield(%{{.*}} : !fir.ref<!fir.type<_QFsub1TtUnboxProc{p1:() -> ()}>>)
omp.yield(%1 : !fir.ref<!fir.type<_QFsub1Tt{p1:!fir.boxproc<() -> ()>}>>)
}
func.func @_QPsub1() {
%0 = fir.alloca !fir.type<_QFsub1Tt{p1:!fir.boxproc<() -> ()>}> {bindc_name = "t1", uniq_name = "_QFsub1Et1"}
%1 = fir.declare %0 {uniq_name = "_QFsub1Et1"} : (!fir.ref<!fir.type<_QFsub1Tt{p1:!fir.boxproc<() -> ()>}>>) -> !fir.ref<!fir.type<_QFsub1Tt{p1:!fir.boxproc<() -> ()>}>>
//CHECK: omp.parallel private(@_QFsub1Et1_private_ref_rec__QFsub1Tt %{{.*}} -> %{{.*}} : !fir.ref<!fir.type<_QFsub1TtUnboxProc{p1:() -> ()}>>) {
omp.parallel private(@_QFsub1Et1_private_ref_rec__QFsub1Tt %1 -> %arg0 : !fir.ref<!fir.type<_QFsub1Tt{p1:!fir.boxproc<() -> ()>}>>) {
%2 = fir.declare %arg0 {uniq_name = "_QFsub1Et1"} : (!fir.ref<!fir.type<_QFsub1Tt{p1:!fir.boxproc<() -> ()>}>>) -> !fir.ref<!fir.type<_QFsub1Tt{p1:!fir.boxproc<() -> ()>}>>
omp.terminator
}
return
}


// Test a private declaration with all regions (alloc, copy, dealloc)
//CHECK: omp.private {type = firstprivate} @_QFsub2Et1_firstprivate_ref_box_heap_rec__QFsub2Tt :
//CHECK-SAME: !fir.ref<!fir.box<!fir.heap<!fir.type<_QFsub2TtUnboxProc{p1:() -> ()}>>>> alloc {
omp.private {type = firstprivate} @_QFsub2Et1_firstprivate_ref_box_heap_rec__QFsub2Tt : !fir.ref<!fir.box<!fir.heap<!fir.type<_QFsub2Tt{p1:!fir.boxproc<() -> ()>}>>>> alloc {
//CHECK: ^bb0(%{{.*}}: !fir.ref<!fir.box<!fir.heap<!fir.type<_QFsub2TtUnboxProc{p1:() -> ()}>>>>):
^bb0(%arg0: !fir.ref<!fir.box<!fir.heap<!fir.type<_QFsub2Tt{p1:!fir.boxproc<() -> ()>}>>>>):
%0 = fir.alloca !fir.box<!fir.heap<!fir.type<_QFsub2Tt{p1:!fir.boxproc<() -> ()>}>>> {bindc_name = "t1", pinned, uniq_name = "_QFsub2Et1"}
%1 = fir.declare %0 {fortran_attrs = #fir.var_attrs<allocatable>, uniq_name = "_QFsub2Et1"} : (!fir.ref<!fir.box<!fir.heap<!fir.type<_QFsub2Tt{p1:!fir.boxproc<() -> ()>}>>>>) -> !fir.ref<!fir.box<!fir.heap<!fir.type<_QFsub2Tt{p1:!fir.boxproc<() -> ()>}>>>>
//CHECK: omp.yield(%{{.*}} : !fir.ref<!fir.box<!fir.heap<!fir.type<_QFsub2TtUnboxProc{p1:() -> ()}>>>>)
omp.yield(%1 : !fir.ref<!fir.box<!fir.heap<!fir.type<_QFsub2Tt{p1:!fir.boxproc<() -> ()>}>>>>)
} copy {
//CHECK: ^bb0(%{{.*}}: !fir.ref<!fir.box<!fir.heap<!fir.type<_QFsub2TtUnboxProc{p1:() -> ()}>>>>,
//CHECK-SAME: %{{.*}}: !fir.ref<!fir.box<!fir.heap<!fir.type<_QFsub2TtUnboxProc{p1:() -> ()}>>>>):
^bb0(%arg0: !fir.ref<!fir.box<!fir.heap<!fir.type<_QFsub2Tt{p1:!fir.boxproc<() -> ()>}>>>>, %arg1: !fir.ref<!fir.box<!fir.heap<!fir.type<_QFsub2Tt{p1:!fir.boxproc<() -> ()>}>>>>):
%c5_i32 = arith.constant 5 : i32
%0 = fir.load %arg0 : !fir.ref<!fir.box<!fir.heap<!fir.type<_QFsub2Tt{p1:!fir.boxproc<() -> ()>}>>>>
%1 = fir.box_addr %0 : (!fir.box<!fir.heap<!fir.type<_QFsub2Tt{p1:!fir.boxproc<() -> ()>}>>>) -> !fir.heap<!fir.type<_QFsub2Tt{p1:!fir.boxproc<() -> ()>}>>
%2 = fir.embox %1 : (!fir.heap<!fir.type<_QFsub2Tt{p1:!fir.boxproc<() -> ()>}>>) -> !fir.box<!fir.type<_QFsub2Tt{p1:!fir.boxproc<() -> ()>}>>
%3 = fir.address_of(@_QQclXea) : !fir.ref<!fir.char<1,8>>
%4 = fir.convert %arg1 : (!fir.ref<!fir.box<!fir.heap<!fir.type<_QFsub2Tt{p1:!fir.boxproc<() -> ()>}>>>>) -> !fir.ref<!fir.box<none>>
%5 = fir.convert %2 : (!fir.box<!fir.type<_QFsub2Tt{p1:!fir.boxproc<() -> ()>}>>) -> !fir.box<none>
%6 = fir.convert %3 : (!fir.ref<!fir.char<1,8>>) -> !fir.ref<i8>
%7 = fir.call @_FortranAAssign(%4, %5, %6, %c5_i32) : (!fir.ref<!fir.box<none>>, !fir.box<none>, !fir.ref<i8>, i32) -> none
//CHECK: omp.yield(%{{.*}} : !fir.ref<!fir.box<!fir.heap<!fir.type<_QFsub2TtUnboxProc{p1:() -> ()}>>>>)
omp.yield(%arg1 : !fir.ref<!fir.box<!fir.heap<!fir.type<_QFsub2Tt{p1:!fir.boxproc<() -> ()>}>>>>)
} dealloc {
//CHECK: ^bb0(%{{.*}}: !fir.ref<!fir.box<!fir.heap<!fir.type<_QFsub2TtUnboxProc{p1:() -> ()}>>>>):
^bb0(%arg0: !fir.ref<!fir.box<!fir.heap<!fir.type<_QFsub2Tt{p1:!fir.boxproc<() -> ()>}>>>>):
%c5_i32 = arith.constant 5 : i32
%false = arith.constant false
%0 = fir.absent !fir.box<none>
%1 = fir.address_of(@_QQclXea) : !fir.ref<!fir.char<1,8>>
%2 = fir.convert %arg0 : (!fir.ref<!fir.box<!fir.heap<!fir.type<_QFsub2Tt{p1:!fir.boxproc<() -> ()>}>>>>) -> !fir.ref<!fir.box<none>>
%3 = fir.convert %1 : (!fir.ref<!fir.char<1,8>>) -> !fir.ref<i8>
%4 = fir.call @_FortranAAllocatableDeallocate(%2, %false, %0, %3, %c5_i32) fastmath<contract> : (!fir.ref<!fir.box<none>>, i1, !fir.box<none>, !fir.ref<i8>, i32) -> i32
omp.yield
}
func.func @_QPsub2() {
%0 = fir.alloca !fir.box<!fir.heap<!fir.type<_QFsub2Tt{p1:!fir.boxproc<() -> ()>}>>> {bindc_name = "t1", uniq_name = "_QFsub2Et1"}
%1 = fir.declare %0 {fortran_attrs = #fir.var_attrs<allocatable>, uniq_name = "_QFsub2Et1"} : (!fir.ref<!fir.box<!fir.heap<!fir.type<_QFsub2Tt{p1:!fir.boxproc<() -> ()>}>>>>) -> !fir.ref<!fir.box<!fir.heap<!fir.type<_QFsub2Tt{p1:!fir.boxproc<() -> ()>}>>>>
//CHECK: omp.parallel private(@_QFsub2Et1_firstprivate_ref_box_heap_rec__QFsub2Tt %{{.*}} -> %{{.*}} :
//CHECK-SAME: !fir.ref<!fir.box<!fir.heap<!fir.type<_QFsub2TtUnboxProc{p1:() -> ()}>>>>) {
omp.parallel private(@_QFsub2Et1_firstprivate_ref_box_heap_rec__QFsub2Tt %1 -> %arg0 : !fir.ref<!fir.box<!fir.heap<!fir.type<_QFsub2Tt{p1:!fir.boxproc<() -> ()>}>>>>) {
%2 = fir.declare %arg0 {fortran_attrs = #fir.var_attrs<allocatable>, uniq_name = "_QFsub2Et1"} : (!fir.ref<!fir.box<!fir.heap<!fir.type<_QFsub2Tt{p1:!fir.boxproc<() -> ()>}>>>>) -> !fir.ref<!fir.box<!fir.heap<!fir.type<_QFsub2Tt{p1:!fir.boxproc<() -> ()>}>>>>
omp.terminator
}
return
}
func.func private @_FortranAInitialize(!fir.box<none>, !fir.ref<i8>, i32) -> none attributes {fir.runtime}
fir.global linkonce @_QQclXea constant : !fir.char<1,8> {
%0 = fir.string_lit "pp.f90\00"(8) : !fir.char<1,8>
fir.has_value %0 : !fir.char<1,8>
}
func.func private @_FortranAAllocatableDeallocate(!fir.ref<!fir.box<none>>, i1, !fir.box<none>, !fir.ref<i8>, i32) -> i32 attributes {fir.runtime}
func.func private @_FortranAAssign(!fir.ref<!fir.box<none>>, !fir.box<none>, !fir.ref<i8>, i32) -> none attributes {fir.runtime}
Loading