fix_test_auto_prefix_cache_support

Signed-off-by: Huamin Li <3ericli@gmail.com>

Author: hl475

tests/models/language/pooling/test_auto_prefix_cache_support.py

@@ -26,7 +26,8 @@ def test_classify_models(
     with vllm_runner(model,
                      max_model_len=512,
                      dtype=dtype,
-                     enable_prefix_caching=True) as vllm_model:
+                     enable_prefix_caching=True,
+                     enable_chunked_prefill=True) as vllm_model:
         cache_config = vllm_model.llm.llm_engine.cache_config
         assert cache_config.enable_prefix_caching
         vllm_outputs = vllm_model.classify(example_prompts)
@@ -63,6 +64,7 @@ def test_embed_models(
             runner="pooling",
             max_model_len=None,
             enable_prefix_caching=True,
+            enable_chunked_prefill=True,
     ) as vllm_model:
         cache_config = vllm_model.llm.llm_engine.cache_config
         assert cache_config.enable_prefix_caching