Reserve slot 0 as padding in all req pools (#24243)

python/sglang/srt/disaggregation/decode.py

@@ -130,13 +130,14 @@ class DecodeReqToTokenPool:
         self.device = device
         self.pre_alloc_size = pre_alloc_size
         with memory_saver_adapter.region(tag=GPU_MEMORY_TYPE_KV_CACHE):
+            # +1 row 0 padding; mirrors ReqToTokenPool / KV pool padding slot 0.
             self.req_to_token = torch.zeros(
-                (size + pre_alloc_size, max_context_len),
+                (size + pre_alloc_size + 1, max_context_len),
                 dtype=torch.int32,
                 device=device,
             )
 
-        self.free_slots = list(range(size + pre_alloc_size))
+        self.free_slots = list(range(1, size + pre_alloc_size + 1))
 
     def write(self, indices, values):
         self.req_to_token[indices] = values
@@ -173,7 +174,7 @@ class DecodeReqToTokenPool:
         req.req_pool_idx = None
 
     def clear(self):
-        self.free_slots = list(range(self.size + self.pre_alloc_size))
+        self.free_slots = list(range(1, self.size + self.pre_alloc_size + 1))
 
 
 class HybridMambaDecodeReqToTokenPool(HybridReqToTokenPool):
@@ -238,7 +239,7 @@ class HybridMambaDecodeReqToTokenPool(HybridReqToTokenPool):
         )
 
     def clear(self):
-        self.free_slots = list(range(self.size + self.pre_alloc_size))
+        self.free_slots = list(range(1, self.size + self.pre_alloc_size + 1))
         self.mamba_pool.clear()
 
 

python/sglang/srt/mem_cache/memory_pool.py

@@ -142,10 +142,14 @@ class ReqToTokenPool:
         self.max_context_len = max_context_len
         self.device = device
         with memory_saver_adapter.region(GPU_MEMORY_TYPE_KV_CACHE):
+            # +1 row for padding slot 0 (mirrors KV pool): cuda-graph padded
+            # batches default req_pool_indices to 0, so routing dummies through
+            # unowned slot 0 keeps req_to_token[0, :] zero and downstream writes
+            # harmless.
             self.req_to_token = torch.zeros(
-                (size, max_context_len), dtype=torch.int32, device=device
+                (size + 1, max_context_len), dtype=torch.int32, device=device
             )
-        self.free_slots = list(range(size))
+        self.free_slots = list(range(1, size + 1))
 
     def write(self, indices, values):
         self.req_to_token[indices] = values
@@ -185,7 +189,7 @@ class ReqToTokenPool:
         req.req_pool_idx = None
 
     def clear(self):
-        self.free_slots = list(range(self.size))
+        self.free_slots = list(range(1, self.size + 1))
 
 
 class MambaPool: