Use <stdexcept> header unconditionally

docs/cccl/development/macro.rst

@@ -360,6 +360,10 @@ CUDA doesn't support exceptions in device code, however, sometimes we need to wr
 
 *Note*: The ``_CCCL_CATCH`` clause must always introduce a named variable, like: ``_CCCL_CATCH(const exception_type& var)``.
 
+.. note::
+
+  ``_CCCL_THROW`` requires to include the ``<stdexcept>`` header, regardless exceptions are enabled or not.
+
 Example:
 
 .. code-block:: c++

libcudacxx/include/cuda/__driver/driver_api.h

@@ -35,6 +35,8 @@
 #    include <dlfcn.h>
 #  endif
 
+#  include <stdexcept>
+
 #  include <cuda.h>
 
 #  include <cuda/std/__cccl/prologue.h>

libcudacxx/include/cuda/__tma/make_tma_descriptor.h

@@ -33,6 +33,8 @@
 #  include <cuda/std/cstdint>
 #  include <cuda/std/span>
 
+#  include <stdexcept>
+
 #  include <driver_types.h>
 
 #  include <dlpack/dlpack.h>