[Scheduler] Gate SWA eviction on accumulated tokens (#33404)

python/sglang/srt/managers/schedule_batch.py

@@ -3252,14 +3252,23 @@ class ScheduleBatch(ScheduleBatchDisaggregationDecodeMixin):
             )
 
             eviction_interval = max(1, envs.SGLANG_SWA_EVICTION_INTERVAL.get())
-            swa_maintenance_step = (self.forward_iter or 0) % eviction_interval == 0
             self.token_to_kv_pool_allocator.free_group_begin()
             for idx, req in enumerate(self.reqs):
                 if self.forward_mode.is_decode():
                     # We set evict_swa condition here with two reasons:
                     # 1. In overlap scheduler, we cannot evict swa when req.decode_batch_idx == 0 since the prev extend batch is still running.
-                    # 2. Evict swa every eviction_interval iterations to reduce the overhead.
+                    # 2. Evict only once >= eviction_interval tokens have slid
-                    if swa_maintenance_step and req.decode_batch_idx >= 1:
+                    # out of the window, amortizing eviction work while keeping
+                    # each request's overshoot within the interval the pool
+                    # budget reserves. Gating on accumulated tokens (rather
+                    # than an iteration-counter phase) cannot starve because
+                    # seqlen progress is monotonic per KV handle.
+                    if (
+                        req.decode_batch_idx >= 1
+                        and req.kv is not None
+                        and req.seqlen - 1 - sliding_window_size
+                        >= req.kv.swa_evicted_seqlen + eviction_interval
+                    ):
                         self._evict_swa(req, req.seqlen - 1)
 
                     # DSV4-NPU only (no-op elsewhere): the small paged compress-state