[SYCL] Align design document with recent changes (#260)

- Replace "kernel wrapper" with OpenCL kernel
- Remove "OpenCL C++ to SPIR-V built-in function names mapper"

Signed-off-by: Mariya Podchishchaeva <[email protected]>

Author: Fznamznon

sycl/doc/SYCL_compiler_and_runtime_design.md

@@ -33,8 +33,7 @@ header" which provides information like kernel name, parameters order and data
 type for the runtime library.
 - **Middle-end** - transforms the initial LLVM IR* to get consumed by the
 back-end. Today middle-end transformations include just a couple of passes:
-  - OpenCL C++* to SPIR-V* built-in function names mapper
-  - Address space handling pass
+  - Optionally: Address space inference pass
   - TBD: potentially the middle-end optimizer can run any LLVM IR
   transformation with only one limitation: back-end compiler should be able to
   handle transformed LLVM IR.
@@ -109,38 +108,38 @@ invoking kernels (such as `parallel_for`). For example, in the previous code
 snippet above `accessor` `A` is one such captured kernel argument.
 
 To facilitate the mapping of the captures/fields of lambdas/functors to OpenCL
-kernel and overcome OpenCL limitations we added the generation of a "kernel
-wrapper" function inside the compiler. A "kernel wrapper" function contains the
+kernel and overcome OpenCL limitations we added the generation of an OpenCL
+kernel function inside the compiler. An OpenCL kernel function contains the
 body of the SYCL kernel function, receives OpenCL like parameters and
 additionally does some manipulation to initialize captured lambda/functor fields
-with these parameters. In some pseudo code the "kernel wrapper" for the previous
-code snippet above looks like this:
+with these parameters. In some pseudo code the OpenCL kernel function for the
+previous code snippet above looks like this:
 
 ```C++
 
-// Let the lambda expression passed to the parallel_for declare unnamed
-// function object with "Lambda" type.
-
 // SYCL kernel is defined in SYCL headers:
-__attribute__((sycl_kernel)) someSYCLKernel(Lambda lambda) {
-  lambda();
+template <typename KernelName, typename KernelType/*, ...*/>
+__attribute__((sycl_kernel)) void sycl_kernel_function(KernelType KernelFuncObj) {
+  // ...
+  KernelFuncObj();
 }
 
-// Kernel wrapper
-__kernel wrapper(global int* a) {
-  Lambda lambda; // Actually lambda declaration doesn't have a name in AST
-  // Let the lambda have one captured field - accessor A. We need to init it
-  // with global pointer from arguments:
-  lambda.A.__init(a);
-  // Body of SYCL kernel from SYCL headers:
+// Generated OpenCL kernel function
+__kernel KernelName(global int* a) {
+  KernelType KernelFuncObj; // Actually kernel function object declaration
+  // doesn't have a name in AST.
+  // Let the kernel function object have one captured field - accessor A.
+  // We need to init it with global pointer from arguments:
+  KernelFuncObj.A.__init(a);
+  // Body of the SYCL kernel from SYCL headers:
   {
-    lambda();
+    KernelFuncObj();
   }
 }
 
 ```
 
-"Kernel wrapper" is generated by the compiler inside the Sema using AST nodes.
+OpenCL kernel function is generated by the compiler inside the Sema using AST nodes.
 
 ### SYCL support in the driver