[mx] Add missing parameter in mxfp8 kernel

When triton_to_mxfp8_dim1() was called from
_to_mxfp8_dim1_kernel_wrapper(), the scale_calculation_mode
parameter was not passed, resulting in incorrect default
value.

Signed-off-by: Ula Golowicz <urszula.golowicz@intel.com>

Author: ugolowic

torchao/prototype/mx_formats/utils.py

@@ -164,7 +164,9 @@ def _to_mxfp8_dim1_kernel_wrapper(
             ScaleCalculationMode.FLOOR,
             ScaleCalculationMode.RCEIL,
         )
-        a_data, a_scale = triton_to_mxfp8_dim1(a, block_size)
+        a_data, a_scale = triton_to_mxfp8_dim1(
+            a, block_size, scale_calculation_mode.value
+        )
     elif cast_kernel_choice == MXFP8Dim1CastKernelChoice.CUDA:
         assert scale_calculation_mode in (
             ScaleCalculationMode.FLOOR,