[Core][v1] Unify allocating slots in prefill and decode in KV cache manager (#12608)

As mentioned in RFC https://github.com/vllm-project/vllm/issues/12254,
this PR achieves the task: combine allocate_slots and append_slots.

There should be no functionality change, except that in decode, also
raise exception when num_tokens is zero (like prefill), and change the
unit test case accordingly.

@comaniac @rickyyx @WoosukKwon @youkaichao @heheda12345 @simon-mo

---------

Signed-off-by: Shawn Du <shawnd200@outlook.com>

tests/v1/core/test_prefix_caching.py

@@ -164,7 +164,7 @@ def test_decode():
     req0.num_computed_tokens = 55
     for _ in range(4):
         req0.append_output_token_ids(8)
-    new_blocks = manager.append_slots(req0, 4)
+    new_blocks = manager.allocate_slots(req0, 4)
     assert new_blocks is not None and len(new_blocks) == 0
     assert manager.req_to_blocks[req0.request_id][-2].block_hash is None
 
@@ -175,7 +175,7 @@ def test_decode():
     # the preallocated block.
     for _ in range(5 + 10):
         req0.append_output_token_ids(7)
-    new_blocks = manager.append_slots(req0, 15)
+    new_blocks = manager.allocate_slots(req0, 15)
     assert new_blocks is not None and len(new_blocks) == 0
     assert manager.req_to_blocks[req0.request_id][-2].block_hash is not None
 
@@ -185,7 +185,7 @@ def test_decode():
     # the preallocated block.
     for _ in range(6 + 11):
         req0.append_output_token_ids(12)
-    new_blocks = manager.append_slots(req0, 17)
+    new_blocks = manager.allocate_slots(req0, 17)
     # Plus one preallocated block.
     assert new_blocks is not None and len(new_blocks) == 2
 
@@ -395,12 +395,14 @@ def test_preallocate_blocks(num_preallocate_tokens: int, block_size: int):
     req.num_computed_tokens = block_size
     assert len(blocks) == 1 + num_preallocated_blocks
 
-    # Assume all computed.
+    # Assume all computed, only when num_preallocate_tokens > 0, we need to
-    manager.append_slots(req, block_size * (len(blocks) - 1))
+    # consume the previously preallocated blocks.
+    if num_preallocated_blocks > 0:
+        manager.allocate_slots(req, block_size * (len(blocks) - 1))
         req.num_computed_tokens = block_size * len(blocks)
 
     # Append 1 block.
-    blocks = manager.append_slots(req, block_size)
+    blocks = manager.allocate_slots(req, block_size)
     assert len(blocks) == 1 + num_preallocated_blocks
 
 
@@ -503,7 +505,7 @@ def test_mm_prefix_caching():
     # Append slots without allocating a new block.
     for _ in range(5):
         req0.append_output_token_ids(8)
-    new_blocks = manager.append_slots(req0, 5)
+    new_blocks = manager.allocate_slots(req0, 5)
     assert new_blocks is not None and len(new_blocks) == 0
 
     # The just completed block should have hashes with extra keys.
@@ -603,7 +605,7 @@ def test_reset_prefix_cache():
     unique_token_ids = [3] * 7
     all_token_ids = full_block_token_ids + unique_token_ids
     req0 = make_request("0", all_token_ids)
-    blocks = manager.allocate_slots(req0, 55, [])
+    blocks = manager.allocate_slots(req0, 55)
     assert [b.block_id for b in blocks] == [0, 1, 2, 3]
 
     unique_token_ids = [4] * 7
@@ -639,7 +641,7 @@ def test_uncache_blocks():
     )
 
     req0 = make_request("0", list(range(30)))
-    blocks = manager.allocate_slots(req0, 30, [])
+    blocks = manager.allocate_slots(req0, 30)
     assert [b.block_id for b in blocks] == [0, 1]
     assert len(manager.cached_block_hash_to_block) == 1
 
@@ -648,7 +650,7 @@ def test_uncache_blocks():
     # Simulate speculative tokens.
     for _ in range(5):
         req0.append_output_token_ids(8)
-    manager.append_slots(req0, 5)
+    manager.allocate_slots(req0, 5)
     assert len(manager.cached_block_hash_to_block) == 2
 
     # After sampling, assuming only 1 token is accepted.

vllm/v1/core/kv_cache_manager.py

@@ -1,5 +1,5 @@
 from collections import defaultdict
-from typing import Dict, Iterable, List, Optional, Tuple
+from typing import DefaultDict, Dict, Iterable, List, Optional, Tuple
 
 from vllm.logger import init_logger
 from vllm.utils import cdiv
@@ -67,7 +67,8 @@ class KVCacheManager:
         # Mapping from request ID to blocks to track the blocks allocated
         # for each request, so that we can free the blocks when the request
         # is finished.
-        self.req_to_blocks: Dict[str, List[KVCacheBlock]] = {}
+        self.req_to_blocks: DefaultDict[str,
+                                        List[KVCacheBlock]] = defaultdict(list)
 
     @property
     def usage(self) -> float:
@@ -115,33 +116,75 @@ class KVCacheManager:
         num_computed_tokens = len(computed_blocks) * self.block_size
         return computed_blocks, num_computed_tokens
 
-    def append_slots(
+    def allocate_slots(
         self,
         request: Request,
         num_tokens: int,
+        new_computed_blocks: Optional[List[KVCacheBlock]] = None
     ) -> Optional[List[KVCacheBlock]]:
-        """Append slots to the block table of the request.
+        """Add slots for a request with new tokens to append.
-        We first append slots to already allocated blocks. If the allocated
-        blocks are not enough, we allocate new blocks.
 
         Args:
-            request: The request to append slots.
+            request: The request to allocate slots.
-            num_tokens: The number of tokens to append.
+            num_tokens: The number of tokens to allocate. Note that this does
+                not include the tokens that have already been computed.
+            new_computed_blocks: A list of new computed blocks just hitting the
+                prefix caching.
+
+        Blocks layout:
+        -----------------------------------------------------------------------
+        | < computed > | < new computed > |    < new >    | < pre-allocated > |
+        -----------------------------------------------------------------------
+        |                  < required >                   |
+        --------------------------------------------------
+        |                    < full >                  |
+        ------------------------------------------------
+                                          | <new full> |
+                                          --------------
+        The following *_blocks are illustrated in this layout.
 
         Returns:
-            A list of new blocks if new blocks are allocated, or None
+            A list of new allocated blocks.
-            if new blocks are required but cannot be allocated.
         """
-        num_required_blocks = cdiv(request.num_computed_tokens + num_tokens,
+        if num_tokens == 0:
+            raise ValueError("num_tokens must be greater than 0")
+
+        new_computed_blocks = new_computed_blocks or []
+
+        # The number of computed tokens is the number of computed tokens plus
+        # the new prefix caching hits
+        num_computed_tokens = (request.num_computed_tokens +
+                               len(new_computed_blocks) * self.block_size)
+        num_required_blocks = cdiv(num_computed_tokens + num_tokens,
                                    self.block_size)
         req_blocks = self.req_to_blocks[request.request_id]
+        num_new_blocks = (num_required_blocks - len(req_blocks) -
+                          len(new_computed_blocks))
 
-        num_new_blocks = num_required_blocks - len(req_blocks)
+        # If a computed block of a request is an eviction candidate (in the
-        if num_new_blocks > self.free_block_queue.num_free_blocks:
+        # free queue and ref_cnt == 0), it cannot be counted as a free block
-            # Need to allocate new blocks due to insufficient pre-allocated
+        # when allocating this request.
-            # slots, but we cannot allocate new blocks due to the limit.
+        num_evictable_computed_blocks = sum(1 for blk in new_computed_blocks
+                                            if blk.ref_cnt == 0)
+        if (num_new_blocks > self.free_block_queue.num_free_blocks -
+                num_evictable_computed_blocks):
+            # Cannot allocate new blocks
             return None
 
+        # Touch the computed blocks to make sure they won't be evicted.
+        if self.enable_caching:
+            self._touch(new_computed_blocks)
+        else:
+            assert not new_computed_blocks, (
+                "Computed blocks should be empty when "
+                "prefix caching is disabled")
+
+        # Append the new computed blocks to the request blocks until now to
+        # avoid the case where the new blocks cannot be allocated.
+        req_blocks.extend(new_computed_blocks)
+
+        # Start to handle new blocks
+
         if num_new_blocks <= 0:
             # No new block is needed.
             new_blocks = []
@@ -160,112 +203,29 @@ class KVCacheManager:
             )
             assert num_new_blocks > 0
 
+            # Concatenate the computed block IDs and the new block IDs.
             new_blocks = self._get_new_blocks(num_new_blocks)
             req_blocks.extend(new_blocks)
 
         if not self.enable_caching:
             return new_blocks
 
-        num_computed_full_blocks = (request.num_computed_tokens //
-                                    self.block_size)
-
         # NOTE(rickyx): We are assuming the `num_tokens` are actual
         # tokens rather than lookahead slots (e.g. for speculative decoding).
         # TODO(rickyx): When supporting speculative decoding, we will need to
         # differentiate between them so that we can know how many blocks are
         # full after appending the actual tokens.
-        num_full_blocks_after_append = (request.num_computed_tokens +
+        num_full_blocks = (num_computed_tokens + num_tokens) // self.block_size
-                                        num_tokens) // self.block_size
+        num_computed_full_blocks = num_computed_tokens // self.block_size
-        assert num_full_blocks_after_append <= len(req_blocks)
+        new_full_blocks = req_blocks[num_computed_full_blocks:num_full_blocks]
-
-        new_full_blocks = req_blocks[
-            num_computed_full_blocks:num_full_blocks_after_append]
         if new_full_blocks:
             self._cache_full_blocks(
                 request=request,
                 blk_start_idx=num_computed_full_blocks,
-                full_blocks=new_full_blocks,
-                prev_block=req_blocks[num_computed_full_blocks - 1]
-                if num_computed_full_blocks >= 1 else None,
-            )
-
-        return new_blocks
-
-    def allocate_slots(
-        self,
-        request: Request,
-        num_tokens: int,
-        computed_blocks: List[KVCacheBlock],
-    ) -> Optional[List[KVCacheBlock]]:
-        """Allocate slots for a new request.
-
-        Args:
-            request: The request to allocate slots.
-            num_tokens: The number of tokens to allocate. Note that this does
-                not include the tokens that have already been computed.
-            computed_blocks: A list of computed blocks.
-
-        Returns:
-            A list of new allocated blocks.
-        """
-        if num_tokens == 0:
-            raise ValueError(
-                f"num_tokens must be greater than 0, got {num_tokens}")
-
-        # If a computed block of a request is an eviction candidate (in the
-        # free queue and ref_cnt == 0), it cannot be counted as a free block
-        # when allocating this request.
-        num_evictable_computed_blocks = sum(1 for blk in computed_blocks
-                                            if blk.ref_cnt == 0)
-
-        num_required_blocks = cdiv(num_tokens, self.block_size)
-        if (num_required_blocks > self.free_block_queue.num_free_blocks -
-                num_evictable_computed_blocks):
-            # Cannot allocate new blocks.
-            return None
-
-        # Touch the computed blocks to make sure they won't be evicted.
-        if self.enable_caching:
-            self._touch(computed_blocks)
-        else:
-            assert not computed_blocks, (
-                "Computed blocks should be empty when "
-                "prefix caching is disabled")
-
-        # Determine the number of new blocks to allocate considering
-        # preallocated blocks.
-        num_new_blocks = min(
-            num_required_blocks + self.num_preallocate_blocks,
-            self.free_block_queue.num_free_blocks,
-            # Should not exceed the maximum number of blocks per request.
-            # This is especially because the block table has the shape
-            # [..., max_num_blocks_per_req].
-            # TODO(woosuk): Check and reject requests if
-            # num_prompt_tokens + max_tokens > max_model_len.
-            self.max_num_blocks_per_req - len(computed_blocks),
-        )
-        assert num_new_blocks > 0
-
-        # Concatenate the computed block IDs and the new block IDs.
-        new_blocks = self._get_new_blocks(num_new_blocks)
-        self.req_to_blocks[request.request_id] = computed_blocks + new_blocks
-
-        if not self.enable_caching:
-            return new_blocks
-
-        num_computed_tokens = len(computed_blocks) * self.block_size
-        num_full_blocks = (num_computed_tokens + num_tokens) // self.block_size
-
-        new_full_blocks = self.req_to_blocks[
-            request.request_id][len(computed_blocks):num_full_blocks]
-        if new_full_blocks:
-            self._cache_full_blocks(
-                request=request,
-                blk_start_idx=len(computed_blocks),
                 # The new full blocks are the full blocks that are not computed.
                 full_blocks=new_full_blocks,
-                prev_block=computed_blocks[-1] if computed_blocks else None,
+                prev_block=(req_blocks[num_computed_full_blocks - 1]
-            )
+                            if num_computed_full_blocks > 0 else None))
 
         return new_blocks
 

vllm/v1/core/scheduler.py

@@ -138,7 +138,7 @@ class Scheduler:
             assert num_new_tokens > 0
 
             while True:
-                new_blocks = self.kv_cache_manager.append_slots(
+                new_blocks = self.kv_cache_manager.allocate_slots(
                     request, num_new_tokens)
                 if new_blocks is None:
                     # The request cannot be scheduled.