[Fix] Vacuous marker writes in the cache tests, and an undebited Mamba admission slot

python/sglang/srt/managers/schedule_policy.py

@@ -1222,7 +1222,10 @@ def add_one_req(
         total_tokens = cand_extend_input_len + max_new + self.page_size
         # Shared Mamba pool: fold the new mamba state's shared-gap cost into
         # `total_tokens` so both `rem_total_tokens` gates reflect the joint budget.
-        total_tokens += self._mamba_gap_budget_for_req(req)
+        # Read before `init_load_back` binds `req.mamba_pool_idx` — after that
+        # this returns 0, so the debit sites below reuse the value.
+        mamba_gap_reserve = self._mamba_gap_budget_for_req(req)
+        total_tokens += mamba_gap_reserve
 
         # adjusting the input_tokens based on host_hit_length and page_size
         real_input_tokens = cand_extend_input_len - req.host_hit_length
@@ -1369,7 +1372,7 @@ def add_one_req(
                         CLIP_MAX_NEW_TOKENS,
                     ),
                     req.retracted_stain,
-                    mamba_gap_reserve=self._mamba_gap_budget_for_req(req),
+                    mamba_gap_reserve=mamba_gap_reserve,
                 )
                 self._account_prefill_cache_admission(req, prefix_len)
             else:
@@ -1416,7 +1419,7 @@ def add_one_req(
                     trunc_len,
                     0,
                     req.retracted_stain,
-                    mamba_gap_reserve=self._mamba_gap_budget_for_req(req),
+                    mamba_gap_reserve=mamba_gap_reserve,
                 )
                 self._account_prefill_cache_admission(req, prefix_len)
 

test/registered/unit/mem_cache/test_unified_radix_cache_unittest.py

@@ -4982,10 +4982,9 @@ def _get_full_kv_pool(self, allocator):
     def _fill_full_kv(self, allocator, indices, marker):
         kv_pool = self._get_full_kv_pool(allocator)
         layer_id = kv_pool.start_layer
-        k_buf = kv_pool.get_key_buffer(layer_id)
-        v_buf = kv_pool.get_value_buffer(layer_id)
-        k_buf[indices].fill_(marker)
-        v_buf[indices].fill_(marker + 1)
+        # `buf[indices]` is an advanced-index copy — assign, never `fill_()`.
+        kv_pool.get_key_buffer(layer_id)[indices] = marker
+        kv_pool.get_value_buffer(layer_id)[indices] = marker + 1
 
     def _snapshot_full_kv(self, allocator, indices):
         kv_pool = self._get_full_kv_pool(allocator)
@@ -5000,9 +4999,9 @@ def _fill_mamba_state(self, req_to_token_pool, indices, marker):
             return
         mamba_indices = indices.reshape(-1)
         mamba_cache = req_to_token_pool.mamba_pool.mamba_cache
-        mamba_cache.temporal[:, mamba_indices].fill_(marker)
+        mamba_cache.temporal[:, mamba_indices] = marker
         for offset, conv_buf in enumerate(mamba_cache.conv, start=1):
-            conv_buf[:, mamba_indices].fill_(marker + offset)
+            conv_buf[:, mamba_indices] = marker + offset
 
     def _snapshot_mamba_state(self, req_to_token_pool, indices):
         mamba_indices = indices.reshape(-1)