[Bugfix] Correctly handle the output shape in multimodal attention (#5443)

### What this PR does / why we need it?
Fix https://github.com/vllm-project/vllm-ascend/issues/5297, for
`AscendMMEncoderAttention` forward, we should keep the output shape
consistence with the input

- vLLM version: release/v0.13.0
- vLLM main:
81786c8774

---------

Signed-off-by: wangli <wangli858794774@gmail.com>

tests/e2e/conftest.py

@@ -781,6 +781,11 @@ PROMPT_CONFIGS = {
             "fps": 1,
         },
     },
+    "hunyuan-vl": {
+        "model": "Tencent-Hunyuan/HunyuanOCR",
+        "prompt_fn": hunyuan_prompt,
+        "mm_processor_kwargs": {},
+    },
 }