Fix race between DROP TABLESPACE and checkpointing.

macdice · macdice · commit 26e007939809 · 2022-03-16T17:20:50.000+13:00
Commands like ALTER TABLE SET TABLESPACE may leave files for the next checkpoint to clean up. If such files are not removed by the time DROP TABLESPACE is called, we request a checkpoint so that they are deleted. However, there is presently a window before checkpoint start where new unlink requests won't be scheduled until the following checkpoint. This means that the checkpoint forced by DROP TABLESPACE might not remove the files we expect it to remove, and the following ERROR will be emitted: ERROR: tablespace "mytblspc" is not empty To fix, add a call to AbsorbSyncRequests() just before advancing the unlink cycle counter. This ensures that any unlink requests forwarded prior to checkpoint start (i.e., when ckpt_started is incremented) will be processed by the current checkpoint. Since AbsorbSyncRequests() performs memory allocations, it cannot be called within a critical section, so we also need to move SyncPreCheckpoint() to before CreateCheckPoint()'s critical section. This is an old bug, so back-patch to all supported versions. Author: Nathan Bossart <nathandbossart@gmail.com> Reported-by: Nathan Bossart <nathandbossart@gmail.com> Reviewed-by: Thomas Munro <thomas.munro@gmail.com> Reviewed-by: Andres Freund <andres@anarazel.de> Discussion: https://postgr.es/m/20220215235845.GA2665318%40nathanxps13
diff --git a/src/backend/access/transam/xlog.c b/src/backend/access/transam/xlog.c
@@ -9020,6 +9020,14 @@ CreateCheckPoint(int flags)
 	MemSet(&CheckpointStats, 0, sizeof(CheckpointStats));
 	CheckpointStats.ckpt_start_t = GetCurrentTimestamp();
 
+	/*
+	 * Let smgr prepare for checkpoint; this has to happen outside the
+	 * critical section and before we determine the REDO pointer.  Note that
+	 * smgr must not do anything that'd have to be undone if we decide no
+	 * checkpoint is needed.
+	 */
+	SyncPreCheckpoint();
+
 	/*
 	 * Use a critical section to force system panic if we have trouble.
 	 */
@@ -9034,13 +9042,6 @@ CreateCheckPoint(int flags)
 		LWLockRelease(ControlFileLock);
 	}
 
-	/*
-	 * Let smgr prepare for checkpoint; this has to happen before we determine
-	 * the REDO pointer.  Note that smgr must not do anything that'd have to
-	 * be undone if we decide no checkpoint is needed.
-	 */
-	SyncPreCheckpoint();
-
 	/* Begin filling in the checkpoint WAL record */
 	MemSet(&checkPoint, 0, sizeof(checkPoint));
 	checkPoint.time = (pg_time_t) time(NULL);
diff --git a/src/backend/storage/sync/sync.c b/src/backend/storage/sync/sync.c
@@ -173,14 +173,26 @@ InitSync(void)
  * counter is incremented here.
  *
  * This must be called *before* the checkpoint REDO point is determined.
- * That ensures that we won't delete files too soon.
+ * That ensures that we won't delete files too soon.  Since this calls
+ * AbsorbSyncRequests(), which performs memory allocations, it cannot be
+ * called within a critical section.
  *
  * Note that we can't do anything here that depends on the assumption
  * that the checkpoint will be completed.
  */
 void
 SyncPreCheckpoint(void)
 {
+	/*
+	 * Operations such as DROP TABLESPACE assume that the next checkpoint will
+	 * process all recently forwarded unlink requests, but if they aren't
+	 * absorbed prior to advancing the cycle counter, they won't be processed
+	 * until a future checkpoint.  The following absorb ensures that any
+	 * unlink requests forwarded before the checkpoint began will be processed
+	 * in the current checkpoint.
+	 */
+	AbsorbSyncRequests();
+
 	/*
 	 * Any unlink requests arriving after this point will be assigned the next
 	 * cycle counter, and won't be unlinked until next checkpoint.
