Skip to content

Add workgroup size attribute to AMDGPU functions in codegen #4342

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
merged 1 commit into from
Nov 15, 2019
Merged

Add workgroup size attribute to AMDGPU functions in codegen #4342

Changes from all commits
Commits
Show all changes
1 commit
Select commit
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
26 changes: 26 additions & 0 deletions src/codegen/llvm/codegen_amdgpu.cc
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -36,13 +36,39 @@
namespace tvm {
namespace codegen {

namespace {

// calls the device api to get the max threads per block
static inline int DetectROCMmaxThreadsPerBlock() {
TVMContext tvm_ctx;
tvm_ctx.device_type = kDLROCM;
tvm_ctx.device_id = 0;
tvm::runtime::DeviceAPI* api = tvm::runtime::DeviceAPI::Get(tvm_ctx, true);
if (api != nullptr) {
TVMRetValue val;
api->GetAttr(tvm_ctx, tvm::runtime::kExist, &val);
if (val.operator int() == 1) {
tvm::runtime::DeviceAPI::Get(tvm_ctx)->
GetAttr(tvm_ctx, tvm::runtime::kMaxThreadsPerBlock, &val);
return val.operator int();
}
}
LOG(WARNING) << "Cannot get maximum number of threads for AMD codegen";
return 256; // see the discussion at PR #4342 for the choice of default
}

} // namespace

// AMDGPU code generator.
class CodeGenAMDGPU : public CodeGenLLVM {
public:
void AddFunction(const LoweredFunc& f) final {
// add function as void return value
CodeGenLLVM::AddFunctionInternal(f, true);
function_->setCallingConv(llvm::CallingConv::AMDGPU_KERNEL);
std::ostringstream attr;
attr << "1," << DetectROCMmaxThreadsPerBlock();
function_->addFnAttr("amdgpu-flat-work-group-size", attr.str());
}

void VisitStmt_(const Allocate* op) final {
Expand Down