bader updated this revision to Diff 333853. bader added a comment. Applied code review comments.
Repository: rG LLVM Github Monorepo CHANGES SINCE LAST ACTION https://reviews.llvm.org/D99488/new/ https://reviews.llvm.org/D99488 Files: clang/docs/SYCLSupport.md
Index: clang/docs/SYCLSupport.md =================================================================== --- clang/docs/SYCLSupport.md +++ clang/docs/SYCLSupport.md @@ -813,6 +813,111 @@ The SPIR-V specific functions are implemented in for the SYCL host device here: `sycl/source/spirv_ops.cpp`. +### Address spaces handling + +SYCL specification uses C++ classes to represent pointers to disjoint memory +regions on an accelerator to enable compilation with standard C++ toolchain and +SYCL compiler toolchain. Section 3.8.2 of SYCL 2020 specification defines +[memory model](https://www.khronos.org/registry/SYCL/specs/sycl-2020/html/sycl-2020.html#_sycl_device_memory_model), +section 4.7.7 - [address space classes](https://www.khronos.org/registry/SYCL/specs/sycl-2020/html/sycl-2020.html#_address_space_classes) +and section 5.9 covers [address space deduction](https://www.khronos.org/registry/SYCL/specs/sycl-2020/html/sycl-2020.html#_address_space_deduction). + +The main address space semantic difference of SYCL mode from OpenCL is that +SYCL doesn't perform address space qualifier inference detailed in +[OpenCL C v3.0 s6.7.8](https://www.khronos.org/registry/OpenCL/specs/3.0-unified/html/OpenCL_C.html#addr-spaces-inference). + +Similar to other single-source C++-based GPU programming modes like +OpenMP/CUDA/HIP, SYCL uses clang's "default" address space for types with no +address space attributes. This design has two important features: keeps the type system consistent with C++ on one hand and enable tools for emitting device code aligned with SPIR memory model (and other GPU targets). + +So inside a function, this variable declaration: + +```C++ +int var; +``` + +SYCL device compiler turns into + +```C++ +VarDecl var 'int' +``` + +OpenCL compiler turn into + +```C++ +VarDecl var '__private int' +``` + +Changing variable type has massive and destructive effect in C++. For instance +this does not compile in C++ for OpenCL mode: + +```C++ +template<typename T1, typename T2> +struct is_same { + static constexpr int value = 0; +}; + +template<typename T> +struct is_same<T, T> { + static constexpr int value = 1; +}; + +void foo(int p) { + static_assert(is_same<decltype(p), int>::value, "int is not an int?"); // Fails: p is '__private int' != 'int' + static_assert(is_same<decltype(&p), int*>::value, "int* is not an int*?"); // Fails: p is '__private int*' != '__generic int*' +} +``` + +`multi_ptr` class implementation example: + +``` C++ +// check that SYCL mode is ON and we can use non-standard decorations +#if defined(__SYCL_DEVICE_ONLY__) +// GPU/accelerator implementation +template <typename T, address_space AS> class multi_ptr { + // DecoratedType applies corresponding address space attribute to the type T + // DecoratedType<T, global_space>::type == "__attribute__((opencl_global)) T" + // See sycl/include/CL/sycl/access/access.hpp for more details + using pointer_t = typename DecoratedType<T, AS>::type *; + + pointer_t m_Pointer; + public: + pointer_t get() { return m_Pointer; } + T& operator* () { return *reinterpret_cast<T*>(m_Pointer); } +} +#else +// CPU/host implementation +template <typename T, address_space AS> class multi_ptr { + T *m_Pointer; // regular undecorated pointer + public: + T *get() { return m_Pointer; } + T& operator* () { return *m_Pointer; } +} +#endif +``` + +Depending on the compiler mode `multi_ptr` will either decorate internal data +with address space attribute or not. + +To utilize existing clang's functionality, we re-use following OpenCL address +space attributes for decoration pointers implementation: + +| Address space attribute | SYCL address_space enumeration | +|-------------------------|--------------------------------| +| `__attribute__((opencl_global))` | global_space, constant_space | +| `__attribute__((opencl_local))` | local_space | +| `__attribute__((opencl_private))` | private_space | + + TODO: add support for `__attribute__((opencl_global_host))` and + `__attribute__((opencl_global_device))`. + +Default address space represents "Generic-memory", which is a virtual address +space which overlaps the global, local and private address spaces. SYCL mode +enables conversion to/from default address space from/to address space +attributed type. + +SPIR target allocates SYCL namespace scope variables in global address space. + ### Compiler/Runtime interface ## SYCL Runtime architecture
_______________________________________________ cfe-commits mailing list cfe-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
