[FEAT] Enhance CLI argument parsing for prefix caching: add error handling for invalid hash algorithms and introduce xxhash support

Signed-off-by: LuminolT <[email protected]>

Author: LuminolT

tests/v1/engine/test_engine_args.py

@@ -9,6 +9,7 @@
 from vllm.engine.arg_utils import EngineArgs
 from vllm.usage.usage_lib import UsageContext
 from vllm.utils.argparse_utils import FlexibleArgumentParser
+from vllm.utils.hashing import _xxhash
 
 
 def test_prefix_caching_from_cli():
@@ -42,6 +43,16 @@ def test_prefix_caching_from_cli():
     vllm_config = EngineArgs.from_cli_args(args=args).create_engine_config()
     assert vllm_config.cache_config.prefix_caching_hash_algo == "sha256"
 
+    # an invalid hash algorithm raises an error
+    parser.exit_on_error = False
+    with pytest.raises(ArgumentError):
+        args = parser.parse_args(["--prefix-caching-hash-algo", "invalid"])
+
+
+@pytest.mark.skipif(_xxhash is None, reason="xxhash not installed")
+def test_prefix_caching_xxhash_from_cli():
+    parser = EngineArgs.add_cli_args(FlexibleArgumentParser())
+
     # set hash algorithm to xxhash (pickle)
     args = parser.parse_args(["--prefix-caching-hash-algo", "xxhash"])
     vllm_config = EngineArgs.from_cli_args(args=args).create_engine_config()
@@ -52,11 +63,6 @@ def test_prefix_caching_from_cli():
     vllm_config = EngineArgs.from_cli_args(args=args).create_engine_config()
     assert vllm_config.cache_config.prefix_caching_hash_algo == "xxhash_cbor"
 
-    # an invalid hash algorithm raises an error
-    parser.exit_on_error = False
-    with pytest.raises(ArgumentError):
-        args = parser.parse_args(["--prefix-caching-hash-algo", "invalid"])
-
 
 def test_defaults_with_usage_context():
     engine_args = EngineArgs(model="facebook/opt-125m")