[Bugfix][CPU] Fix CPU backend ROPE dispatch for VL models (#30829)

Signed-off-by: jiang1.li <jiang1.li@intel.com>
Signed-off-by: Li, Jiang <bigpyj64@gmail.com>
Co-authored-by: gemini-code-assist[bot] <176961590+gemini-code-assist[bot]@users.noreply.github.com>
(cherry picked from commit 0cd53536)

vllm/model_executor/layers/rotary_embedding/common.py

@@ -264,6 +264,15 @@ class ApplyRotaryEmb(CustomOp):
 
         return output
 
+    def forward_cpu(
+        self,
+        x: torch.Tensor,
+        cos: torch.Tensor,
+        sin: torch.Tensor,
+    ) -> torch.Tensor:
+        # TODO (bigPYJ1151): need to enable fused CPU ROPE here
+        return self.forward_native(x, cos, sin)
+
     def extra_repr(self) -> str:
         s = f"is_neox_style={self.is_neox_style}"
         s += f"enable_fp32_compute={self.enable_fp32_compute}"