Update documentation on zeKernelSuggestMaxCooperativeGroupCount (#412)

* Update documentation on zeKernelSuggestMaxCooperativeGroupCount
* Add set group size to example

Signed-off-by: Aravind Gopalakrishnan <aravind.gopalakrishnan@intel.com>

Author: aravindksg

scripts/core/PROG.rst

@@ -1971,6 +1971,12 @@ cooperative kernel launch may be determined by calling ${x}KernelSuggestMaxCoope
 
 .. parsed-literal::
 
+       // query and set kernel work-group size
+       uint32_t groupSizeX;
+       uint32_t groupSizeY;
+       ${x}KernelSuggestGroupSize(hKernel, imageWidth, imageHeight, 1, &groupSizeX, &groupSizeY, nullptr);
+       ${x}KernelSetGroupSize(hKernel, groupSizeX, groupSizeY, 1);
+
        // query the maximum cooperative kernel launch for the kernel
        uint32_t maxGroupCount;
        ${x}KernelSuggestMaxCooperativeGroupCount(hKernel, &maxGroupCount);

scripts/core/module.yml

@@ -479,6 +479,7 @@ name: SuggestMaxCooperativeGroupCount
 details:
     - "The application may call this function from simultaneous threads."
     - "The implementation of this function should be lock-free."
+    - "Applications are recommended to use $xKernelSuggestGroupSize and $xKernelSetGroupSize first before calling this function and launching cooperative kernels. Otherwise, implementation may return $X_RESULT_ERROR_INVALID_GROUP_SIZE_DIMENSION."
 params:
     - type: "$x_kernel_handle_t"
       name: hKernel