25 const std::string& kernel_name,
33 std::vector<std::string>& arguments_device
56 std::string
warp_shuffle_xor(
const std::string& value,
const std::string& lane_mask)
const override;
63 std::unique_ptr<codegen::LanguageExtension>
Definition structured_sdfg.h:30
Originally started out as a handler for additional source files, like hardware-accelerator kernels th...
Definition code_snippet_factory.h:69
Definition language_extension.h:39
Definition gpu_offload_base_dispatcher.h:19
bool is_device_pointer_storage(const types::StorageType &storage) const override
std::string kernel_header_file_extension() const override
File extension for the kernel's include header ("cu.h"/"rocm.h").
std::string kernel_file_extension() const override
File extension for the kernel translation unit ("cu"/"rocm.cpp").
void emit_target_global_includes(codegen::CodeSnippetFactory &library_snippet_factory) override
static constexpr const char * KERNEL_SNIPPET_FILE_EXT
Definition rocm_offload_dispatcher_strategy.h:50
void emit_target_header_declarations(codegen::PrettyPrinter &kernel_header_stream) override
Target-specific includes/declarations emitted into the kernel include header (device fp16/bf16 intrin...
void dispatch_kernel_launch_error_check(codegen::PrettyPrinter &stream, const codegen::LanguageExtension &language_extension)
std::unique_ptr< codegen::LanguageExtension > create_kernel_language_extension(StructuredSDFG &sdfg, const StructuredLoop &loop) override
bool is_matching_kernel_language_extension(const StructuredSDFG &sdfg, const StructuredLoop &loop, codegen::LanguageExtension &language_extension) override
static constexpr const char * KERNEL_SNIPPET_HEADER_EXT
Definition rocm_offload_dispatcher_strategy.h:51
codegen::TargetType get_instrumentation_kernel_target_type() const override
ROCMOffloadDispatcherStrategy()
void dispatch_kernel_call(codegen::PrettyPrinter &main_stream, const std::string &kernel_name, codegen::LanguageExtension &host_language_ext, symbolic::Expression &num_blocks_x, symbolic::Expression &num_blocks_y, symbolic::Expression &num_blocks_z, symbolic::Expression &block_size_x, symbolic::Expression &block_size_y, symbolic::Expression &block_size_z, std::vector< std::string > &arguments_device) override
int get_warp_size() const override
std::string warp_shuffle_xor(const std::string &value, const std::string &lane_mask) const override
Cross-lane XOR butterfly shuffle of value by lane_mask (CUDA/HIP __shfl_xor_sync).
Base class for structured loop constructs.
Definition structured_loop.h:244
Represents the storage location and management characteristics of data.
Definition type.h:64
SymEngine::RCP< const SymEngine::Basic > Expression
Reference-counted pointer to a symbolic expression (SymEngine Basic).
Definition symbolic.h:83
StringEnum TargetType
Definition instrumentation_info.h:32
Definition rocm_library_node_rewriter_pass.h:9
Symbolic system for integer-based symbol analysis.