[NFC][docs] Add AMDGPU documentation for LIBOMPTARGET_STACK_SIZE

Add documentation w.r.t. changes by #72606, which allows to set the dynamic callstack size.
llvm · Nov 28, 2023 · 19fa276 · 19fa276
1 parent a8ac930
commit 19fa276
Show file tree

Hide file tree

Showing 2 changed files with 5 additions and 3 deletions.
diff --git a/llvm/docs/AMDGPUUsage.rst b/llvm/docs/AMDGPUUsage.rst
@@ -820,6 +820,8 @@ supported for the ``amdgcn`` target.
   that reach other lanes or by explicitly constructing the scratch buffer descriptor,
   triggers undefined behavior when it modifies the scratch values of other lanes.
   The compiler may assume that such modifications do not occur.
+  When using code object V5 ``LIBOMPTARGET_STACK_SIZE`` may be used to provide the
+  private segment size in bytes, for cases where a dynamic stack is used.
 
 **Constant 32-bit**
   *TODO*

diff --git a/openmp/docs/design/Runtimes.rst b/openmp/docs/design/Runtimes.rst
@@ -1016,9 +1016,9 @@ default. The solution is to add an explicit map clause in the target region.
 LIBOMPTARGET_STACK_SIZE
 """""""""""""""""""""""
 
-This environment variable sets the stack size in bytes for the CUDA plugin. This
-can be used to increase or decrease the standard amount of memory reserved for
-each thread's stack.
+This environment variable sets the stack size in bytes for the AMDGPU and CUDA
+plugins. This can be used to increase or decrease the standard amount of memory
+reserved for each thread's stack.
 
 LIBOMPTARGET_HEAP_SIZE
 """""""""""""""""""""""