[Core][Perf] Only invoke save_new_computed_blocks when computed blocks are not empty (#27799)

Signed-off-by: Jialin Ouyang <Jialin.Ouyang@gmail.com>

vllm/v1/core/kv_cache_manager.py

@@ -306,6 +306,7 @@ class KVCacheManager:
                 "Computed blocks should be empty when prefix caching is disabled"
             )
 
+        if new_computed_block_list is not self.empty_kv_cache_blocks.blocks:
             # Append the new computed blocks to the request blocks until now to
             # avoid the case where the new blocks cannot be allocated.
             self.coordinator.save_new_computed_blocks(

vllm/v1/core/single_type_kv_cache_manager.py

@@ -151,7 +151,7 @@ class SingleTypeKVCacheManager(ABC):
             num_tokens: The total number of tokens that need to be cached
                 (including tokens that are already cached).
         """
-        num_cached_blocks = self.num_cached_block[request.request_id]
+        num_cached_blocks = self.num_cached_block.get(request.request_id, 0)
         num_full_blocks = num_tokens // self.block_size
 
         if num_cached_blocks >= num_full_blocks: