[Bugfix] fix --scheduling-policy=priority & n>1 crashes engine (#29764)

Signed-off-by: chaunceyjiang <chaunceyjiang@gmail.com>
Signed-off-by: Nick Hill <nhill@redhat.com>
Co-authored-by: Nick Hill <nhill@redhat.com>

vllm/v1/core/sched/request_queue.py

@@ -137,31 +137,30 @@ class PriorityRequestQueue(RequestQueue):
     """
     A priority queue that supports heap operations.
 
-    Requests with a smaller value of `priority` are processed first.
+    Respects the ordering defined in the Request class, where
+    requests with a smaller value of `priority` are processed first.
     If multiple requests have the same priority, the one with the earlier
     `arrival_time` is processed first.
     """
 
     def __init__(self) -> None:
-        self._heap: list[tuple[int, float, Request]] = []
+        self._heap: list[Request] = []
 
     def add_request(self, request: Request) -> None:
         """Add a request to the queue according to priority policy."""
-        heapq.heappush(self._heap, (request.priority, request.arrival_time, request))
+        heapq.heappush(self._heap, request)
 
     def pop_request(self) -> Request:
         """Pop a request from the queue according to priority policy."""
         if not self._heap:
             raise IndexError("pop from empty heap")
-        _, _, request = heapq.heappop(self._heap)
-        return request
+        return heapq.heappop(self._heap)
 
     def peek_request(self) -> Request:
         """Peek at the next request in the queue without removing it."""
         if not self._heap:
             raise IndexError("peek from empty heap")
-        _, _, request = self._heap[0]
-        return request
+        return self._heap[0]
 
     def prepend_request(self, request: Request) -> None:
         """Add a request to the queue according to priority policy.
@@ -180,15 +179,13 @@ class PriorityRequestQueue(RequestQueue):
 
     def remove_request(self, request: Request) -> None:
         """Remove a specific request from the queue."""
-        self._heap = [(p, t, r) for p, t, r in self._heap if r != request]
+        self._heap.remove(request)
         heapq.heapify(self._heap)
 
     def remove_requests(self, requests: Iterable[Request]) -> None:
         """Remove multiple specific requests from the queue."""
-        requests_to_remove = set(requests)
-        self._heap = [
-            (p, t, r) for p, t, r in self._heap if r not in requests_to_remove
-        ]
+        requests_to_remove = requests if isinstance(requests, set) else set(requests)
+        self._heap = [r for r in self._heap if r not in requests_to_remove]
         heapq.heapify(self._heap)
 
     def __bool__(self) -> bool:
@@ -203,8 +200,7 @@ class PriorityRequestQueue(RequestQueue):
         """Iterate over the queue according to priority policy."""
         heap_copy = self._heap[:]
         while heap_copy:
-            _, _, request = heapq.heappop(heap_copy)
-            yield request
+            yield heapq.heappop(heap_copy)
 
     def __reversed__(self) -> Iterator[Request]:
         """Iterate over the queue in reverse priority order."""