[Bugfix] Fix async copy bug under single expert scenario (#3005)

Add missing barrier when no implicit synchonize by `repeat_interleave`
is available. Otherwise, the `non_blocking=True` copy of `output_splits`
and `input_splits` from NPU may failed to complete before later
`async_all_to_all` uses them.

### What this PR does / why we need it?

### Does this PR introduce _any_ user-facing change?

### How was this patch tested?

- vLLM version: v0.10.2
- vLLM main:
vllm-project/vllm@ef7eefe

Signed-off-by: sdmyzlp <lrwei2@petalmail.com>

Author: sdmyzlp

vllm_ascend/ops/moe/token_dispatcher.py

@@ -639,6 +639,10 @@ def _preprocess(self, topk_ids: torch.Tensor) -> torch.Tensor:
             self.global_input_tokens_local_experts_indices = torch.repeat_interleave(
                 self.expert_ids_per_ep_rank,
                 self.num_global_tokens_per_local_expert.ravel())
+        else:
+            # TODO: This full synchronization can be a performance bottleneck.
+            # A more granular sync (e.g., blocking D2H copies) should be investigated.
+            torch.npu.synchronize()
 
         return num_tokens_per_local_expert