Merge pull request #986 from hdelan/discourage-using-sycl-tc-to-generate-kernels

[UR] Discourage using SYCL tc to generate kernels

Author: kbenzie

scripts/core/CUDA.rst

@@ -21,6 +21,14 @@ CUDA device images can be generated by a CUDA-capable compiler toolchain. Most
 CUDA compiler toolchains are capable of generating PTX, SASS and/or bundles of
 PTX and SASS.
 
+When generating device code to be launched using Unified Runtime, it is
+recommended to use a programming model with explicit kernel parameters, such as
+OpenCL or CUDA. This is because kernels generated by a programming model with
+implicit kernel parameters, such as SYCL, cannot guarantee any specific number
+or ordering of kernel parameters. It has been observed that kernel signatures
+for the same SYCL kernel may vary significantly when compiled for different
+architectures.
+
 PTX
 ---
 

scripts/core/HIP.rst

@@ -27,6 +27,14 @@ HIPCC can generate device code for a particular arch using the ``--genco`` flag
 
     $ hipcc --genco hello.cu --amdgpu-target=gfx906 -o hello.hsaco
 
+When generating device code to be launched using Unified Runtime, it is
+recommended to use a programming model with explicit kernel parameters, such as
+OpenCL or HIP. This is because kernels generated by a programming model with
+implicit kernel parameters, such as SYCL, cannot guarantee any specific number
+or ordering of kernel parameters. It has been observed that kernel signatures
+for the same SYCL kernel may vary significantly when compiled for different
+architectures.
+
 UR Programs
 ===========