From: Tom Lane <tgl@sss.pgh.pa.us>
Date: Fri, 4 May 2001 18:39:16 +0000 (+0000)
Subject: Seems like we should not hold off cancel/die interrupts while we are
X-Git-Url: http://git.postgresql.org/gitweb/?a=commitdiff_plain;h=dd54fdca2aad836ce9a8f6520c2134b79c0da332;p=users%2Fbernd%2Fpostgres.git

Seems like we should not hold off cancel/die interrupts while we are
running deferred triggers.  They are really part of the regular
transaction, and they could take awhile.
---

diff --git a/src/backend/access/transam/xact.c b/src/backend/access/transam/xact.c
index 92871fa5c3..ae1add8ac0 100644
--- a/src/backend/access/transam/xact.c
+++ b/src/backend/access/transam/xact.c
@@ -1026,9 +1026,6 @@ CommitTransaction(void)
 	if (s->state != TRANS_INPROGRESS)
 		elog(NOTICE, "CommitTransaction and not in in-progress state ");
 
-	/* Prevent cancel/die interrupt while cleaning up */
-	HOLD_INTERRUPTS();
-
 	/*
 	 * Tell the trigger manager that this transaction is about to be
 	 * committed. He'll invoke all trigger deferred until XACT before we
@@ -1036,6 +1033,9 @@ CommitTransaction(void)
 	 */
 	DeferredTriggerEndXact();
 
+	/* Prevent cancel/die interrupt while cleaning up */
+	HOLD_INTERRUPTS();
+
 	/*
 	 * set the current transaction state information appropriately during
 	 * the abort processing