fix deepseek prefix cache

Signed-off-by: wangxiyuan <wangxiyuan1007@gmail.com>

Author: wangxiyuan

vllm_ascend/platform.py

@@ -247,8 +247,9 @@ def check_and_update_config(cls, vllm_config: VllmConfig) -> None:
             if cache_config.block_size is None:
                 cache_config.block_size = 128
             else:
-                cache_config.block_size = cdiv(cache_config.block_size,
-                                               64) * 64
+                if not vllm_config.model_config.is_deepseek_mla:
+                    cache_config.block_size = cdiv(cache_config.block_size,
+                                                   64) * 64
 
             if cache_config.enable_prefix_caching and cache_config.block_size != 128:
                 logger.warning(