Skip to content

[flang][cuda] Set address space for constant variables #163430

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.

Sign up for GitHub

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
clementval merged 4 commits into from
Oct 14, 2025
Merged

[flang][cuda] Set address space for constant variables #163430

Show file tree
Hide file tree
Changes from 1 commit
Commits
Show all changes
4 commits
Select commit Hold shift + click to select a range
2fc7512
[flang][cuda] Set address space for constant variables
clementval Oct 14, 2025
91e74bd
Fix address cast in gpu mod
clementval Oct 14, 2025
a8b0bee
Add assertion
clementval Oct 14, 2025
781c293
Remove data layout
clementval Oct 14, 2025
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
3 changes: 2 additions & 1 deletion flang/lib/Optimizer/CodeGen/CodeGen.cpp
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -3231,7 +3231,8 @@ struct GlobalOpConversion : public fir::FIROpConversion<fir::GlobalOp> {

if (global.getDataAttr() &&
*global.getDataAttr() == cuf::DataAttribute::Constant)
TODO(global.getLoc(), "CUDA Fortran CONSTANT variable code generation");
g.setAddrSpace(
static_cast<unsigned>(mlir::NVVM::NVVMMemorySpace::Constant));

rewriter.eraseOp(global);
return mlir::success();
Expand Down
19 changes: 19 additions & 0 deletions flang/test/Fir/CUDA/cuda-code-gen.mlir
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -284,3 +284,22 @@ module attributes {gpu.container_module, dlti.dl_spec = #dlti.dl_spec<#dlti.dl_e
// CHECK-LABEL: llvm.func @_QQxxx()
// CHECK: llvm.alloca %{{.*}} x !llvm.struct<(ptr, i64, i32, i8, i8, i8, i8, array<2 x array<3 x i64>>)> {alignment = 8 : i64} : (i32) -> !llvm.ptr
// CHECK-NOT: llvm.call @_FortranACUFAllocDescriptor

// -----

fir.global @_QMkernelsEinitial_val {data_attr = #cuf.cuda<constant>} : i32 {
%0 = fir.zero_bits i32
fir.has_value %0 : i32
}
func.func @_QMkernelsPassign(%arg0: !fir.ref<i32> {cuf.data_attr = #cuf.cuda<device>, fir.bindc_name = "a"}) attributes {cuf.proc_attr = #cuf.cuda_proc<global>} {
%1 = fir.address_of(@_QMkernelsEinitial_val) : !fir.ref<i32>
%14 = fir.load %1 : !fir.ref<i32>
fir.store %14 to %arg0 : !fir.ref<i32>
return
}

// CHECK: llvm.mlir.global external @_QMkernelsEinitial_val() {addr_space = 4 : i32} : i32
// CHECK-LABEL: llvm.func @_QMkernelsPassign
// CHECK: %[[ADDROF:.*]] = llvm.mlir.addressof @_QMkernelsEinitial_val : !llvm.ptr<4>
// CHECK: %{{.*}} = llvm.addrspacecast %[[ADDROF]] : !llvm.ptr<4> to !llvm.ptr