[flang][cuda] Lower CUDA shared variable with cuf.shared_memory op #131399
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
llvmbot
added
flang
|
@llvm/pr-subscribers-flang-fir-hlfir Author: Valentin Clement (バレンタイン クレメン) (clementval) ChangesFull diff: https://github.com/llvm/llvm-project/pull/131399.diff 2 Files Affected:
diff --git a/flang/lib/Lower/ConvertVariable.cpp b/flang/lib/Lower/ConvertVariable.cpp
index ab5e6346f8d54..05256fec67241 100644
--- a/flang/lib/Lower/ConvertVariable.cpp
+++ b/flang/lib/Lower/ConvertVariable.cpp
@@ -738,9 +738,11 @@ static mlir::Value createNewLocal(Fortran::lower::AbstractConverter &converter,
auto idxTy = builder.getIndexType();
for (mlir::Value sh : elidedShape)
indices.push_back(builder.createConvert(loc, idxTy, sh));
- mlir::Value alloc = builder.create<cuf::AllocOp>(
- loc, ty, nm, symNm, dataAttr, lenParams, indices);
- return alloc;
+ if (dataAttr.getValue() == cuf::DataAttribute::Shared)
+ return builder.create<cuf::SharedMemoryOp>(loc, ty, nm, symNm, lenParams,
+ indices);
+ return builder.create<cuf::AllocOp>(loc, ty, nm, symNm, dataAttr, lenParams,
+ indices);
}
// Let the builder do all the heavy lifting.
@@ -1032,12 +1034,16 @@ static void instantiateLocal(Fortran::lower::AbstractConverter &converter,
symMap);
if (Fortran::semantics::NeedCUDAAlloc(var.getSymbol())) {
auto *builder = &converter.getFirOpBuilder();
+ cuf::DataAttributeAttr dataAttr =
+ Fortran::lower::translateSymbolCUFDataAttribute(builder->getContext(),
+ var.getSymbol());
mlir::Location loc = converter.getCurrentLocation();
fir::ExtendedValue exv =
converter.getSymbolExtendedValue(var.getSymbol(), &symMap);
auto *sym = &var.getSymbol();
const Fortran::semantics::Scope &owner = sym->owner();
- if (owner.kind() != Fortran::semantics::Scope::Kind::MainProgram) {
+ if (owner.kind() != Fortran::semantics::Scope::Kind::MainProgram &&
+ dataAttr.getValue() != cuf::DataAttribute::Shared) {
converter.getFctCtx().attachCleanup([builder, loc, exv, sym]() {
cuf::DataAttributeAttr dataAttr =
Fortran::lower::translateSymbolCUFDataAttribute(
diff --git a/flang/test/Lower/CUDA/cuda-shared.cuf b/flang/test/Lower/CUDA/cuda-shared.cuf
new file mode 100644
index 0000000000000..0bacc4ec0b71e
--- /dev/null
+++ b/flang/test/Lower/CUDA/cuda-shared.cuf
@@ -0,0 +1,11 @@
+! RUN: bbc -emit-hlfir -fcuda %s -o - | FileCheck %s
+
+attributes(global) subroutine sharedmem()
+ real, shared :: s(32)
+ integer :: t
+ t = threadIdx%x
+ s(t) = t
+end subroutine
+
+! CHECK-LABEL: func.func @_QPsharedmem() attributes {cuf.proc_attr = #cuf.cuda_proc<global>}
+! CHECK: %{{.*}} = cuf.shared_memory !fir.array<32xf32> {bindc_name = "s", uniq_name = "_QFsharedmemEs"} -> !fir.ref<!fir.array<32xf32>>
|
wangzpgi
approved these changes
Mar 14, 2025
clementval
changed the base branch from
users/clementval/cuf_shared_memory_op
to
main
clementval
force-pushed
the
cuf_lower_shared_memory_op
branch
from
2dc9b15 to
abadac9
Compare
clementval
force-pushed
the
cuf_lower_shared_memory_op
branch
from
abadac9 to
74b42bb
Compare
clementval
commented
Mar 16, 2025
|
LLVM Buildbot has detected a new failure on builder Full details are available at: https://lab.llvm.org/buildbot/#/builders/41/builds/5606 Here is the relevant piece of the build log for the reference |
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.
Use
cuf.shared_memoryoperation instead ofcuf.allocfor CUDA shared variable. These variables do not need free operations.