Prevent concurrent SimpleLruTruncate() for any given SLRU.

The SimpleLruTruncate() header comment states the new coding rule.  To
achieve this, add locktype "frozenid" and two LWLocks.  This closes a
rare opportunity for data loss, which manifested as "apparent
wraparound" or "could not access status of transaction" errors.  Data
loss is more likely in pg_multixact, due to released branches' thin
margin between multiStopLimit and multiWrapLimit.  If a user's physical
replication primary logged ":  apparent wraparound" messages, the user
should rebuild standbys of that primary regardless of symptoms.  At less
risk is a cluster having emitted "not accepting commands" errors or
"must be vacuumed" warnings at some point.  One can test a cluster for
this data loss by running VACUUM FREEZE in every database.  Back-patch
to 9.5 (all supported versions).

Discussion: https://postgr.es/m/20190218073103[email protected]

diff --git a/doc/src/sgml/catalogs.sgml b/doc/src/sgml/catalogs.sgml
index f314f93051dd98942d66464b82cc1add466b0c5d..96e55193f19384174b7f3f2733efa76db593b737 100644 (file)
--- a/doc/src/sgml/catalogs.sgml
+++ b/doc/src/sgml/catalogs.sgml
@@ -8886,7 +8886,8 @@ SCRAM-SHA-256$<replaceable>&lt;iteration count&gt;</>:<replaceable>&lt;salt&gt;<
    and general database objects (identified by class OID and object OID,
    in the same way as in <structname>pg_description</structname> or
    <structname>pg_depend</structname>).  Also, the right to extend a
-   relation is represented as a separate lockable object.
+   relation is represented as a separate lockable object, as is the right to
+   update <structname>pg_database</structname>.<structfield>datfrozenxid</structfield>.
    Also, <quote>advisory</> locks can be taken on numbers that have
    user-defined meanings.
   </para>
@@ -8912,6 +8913,7 @@ SCRAM-SHA-256$<replaceable>&lt;iteration count&gt;</>:<replaceable>&lt;salt&gt;<
        Type of the lockable object:
        <literal>relation</>,
        <literal>extend</>,
+       <literal>frozenid</literal>,
        <literal>page</>,
        <literal>tuple</>,
        <literal>transactionid</>,

diff --git a/doc/src/sgml/monitoring.sgml b/doc/src/sgml/monitoring.sgml
index c492114f331f925e54c1d077e0162992f729b494..601616ec2cec6209a65fb899681d9c900281492d 100644 (file)
--- a/doc/src/sgml/monitoring.sgml
+++ b/doc/src/sgml/monitoring.sgml
@@ -845,7 +845,7 @@ postgres   27093  0.0  0.0  30096  2752 ?        Ss   11:34   0:00 postgres: ser
 
       <tbody>
        <row>
-        <entry morerows="62"><literal>LWLock</></entry>
+        <entry morerows="64"><literal>LWLock</></entry>
         <entry><literal>ShmemIndexLock</></entry>
         <entry>Waiting to find or allocate space in shared memory.</entry>
        </row>
@@ -1043,6 +1043,16 @@ postgres   27093  0.0  0.0  30096  2752 ?        Ss   11:34   0:00 postgres: ser
          <entry>Waiting to execute <function>txid_status</function> or update
          the oldest transaction id available to it.</entry>
         </row>
+        <row>
+         <entry><literal>WrapLimitsVacuumLock</literal></entry>
+         <entry>Waiting to update limits on transaction id and multixact
+         consumption.</entry>
+        </row>
+        <row>
+         <entry><literal>NotifyQueueTailLock</literal></entry>
+         <entry>Waiting to update limit on notification message
+         storage.</entry>
+        </row>
         <row>
          <entry><literal>clog</></entry>
          <entry>Waiting for I/O on a clog (transaction status) buffer.</entry>
@@ -1118,7 +1128,7 @@ postgres   27093  0.0  0.0  30096  2752 ?        Ss   11:34   0:00 postgres: ser
          <entry>Waiting for TBM shared iterator lock.</entry>
         </row>
         <row>
-         <entry morerows="9"><literal>Lock</></entry>
+         <entry morerows="10"><literal>Lock</></entry>
          <entry><literal>relation</></entry>
          <entry>Waiting to acquire a lock on a relation.</entry>
         </row>
@@ -1126,6 +1136,12 @@ postgres   27093  0.0  0.0  30096  2752 ?        Ss   11:34   0:00 postgres: ser
          <entry><literal>extend</></entry>
          <entry>Waiting to extend a relation.</entry>
         </row>
+        <row>
+         <entry><literal>frozenid</literal></entry>
+         <entry>Waiting to
+         update <structname>pg_database</structname>.<structfield>datfrozenxid</structfield>
+         and <structname>pg_database</structname>.<structfield>datminmxid</structfield>.</entry>
+        </row>
         <row>
          <entry><literal>page</></entry>
          <entry>Waiting to acquire a lock on page of a relation.</entry>

diff --git a/src/backend/access/transam/slru.c b/src/backend/access/transam/slru.c
index 56ca77c407446d539accd8ad32ed110ec1da730d..c7e3a76b35aeef51f06004da57ca08232ef47f34 100644 (file)
--- a/src/backend/access/transam/slru.c
+++ b/src/backend/access/transam/slru.c
@@ -1164,6 +1164,14 @@ SimpleLruFlush(SlruCtl ctl, bool allow_redirtied)
 
 /*
  * Remove all segments before the one holding the passed page number
+ *
+ * All SLRUs prevent concurrent calls to this function, either with an LWLock
+ * or by calling it only as part of a checkpoint.  Mutual exclusion must begin
+ * before computing cutoffPage.  Mutual exclusion must end after any limit
+ * update that would permit other backends to write fresh data into the
+ * segment immediately preceding the one containing cutoffPage.  Otherwise,
+ * when the SLRU is quite full, SimpleLruTruncate() might delete that segment
+ * after it has accrued freshly-written data.
  */
 void
 SimpleLruTruncate(SlruCtl ctl, int cutoffPage)

diff --git a/src/backend/access/transam/subtrans.c b/src/backend/access/transam/subtrans.c
index f6406611301f8f7bdb320665947ee45a264f3c86..cfcab811a823226b565824520bec461d3df40bf6 100644 (file)
--- a/src/backend/access/transam/subtrans.c
+++ b/src/backend/access/transam/subtrans.c
@@ -347,8 +347,8 @@ ExtendSUBTRANS(TransactionId newestXact)
 /*
  * Remove all SUBTRANS segments before the one holding the passed transaction ID
  *
- * This is normally called during checkpoint, with oldestXact being the
- * oldest TransactionXmin of any running transaction.
+ * oldestXact is the oldest TransactionXmin of any running transaction.  This
+ * is called only during checkpoint.
  */
 void
 TruncateSUBTRANS(TransactionId oldestXact)

diff --git a/src/backend/commands/async.c b/src/backend/commands/async.c
index 9b5273a6de826e36d175dcb5152a05ca85fbeac1..3d1f60c632a673d15e7f8817d0e984759fab954d 100644 (file)
--- a/src/backend/commands/async.c
+++ b/src/backend/commands/async.c
@@ -224,19 +224,22 @@ typedef struct QueueBackendStatus
 /*
  * Shared memory state for LISTEN/NOTIFY (excluding its SLRU stuff)
  *
- * The AsyncQueueControl structure is protected by the AsyncQueueLock.
+ * The AsyncQueueControl structure is protected by the AsyncQueueLock and
+ * NotifyQueueTailLock.
  *
- * When holding the lock in SHARED mode, backends may only inspect their own
- * entries as well as the head and tail pointers. Consequently we can allow a
- * backend to update its own record while holding only SHARED lock (since no
- * other backend will inspect it).
+ * When holding AsyncQueueLock in SHARED mode, backends may only inspect their
+ * own entries as well as the head and tail pointers. Consequently we can
+ * allow a backend to update its own record while holding only SHARED lock
+ * (since no other backend will inspect it).
  *
- * When holding the lock in EXCLUSIVE mode, backends can inspect the entries
- * of other backends and also change the head and tail pointers.
+ * When holding AsyncQueueLock in EXCLUSIVE mode, backends can inspect the
+ * entries of other backends and also change the head pointer. When holding
+ * both AsyncQueueLock and NotifyQueueTailLock in EXCLUSIVE mode, backends can
+ * change the tail pointer.
  *
  * AsyncCtlLock is used as the control lock for the pg_notify SLRU buffers.
- * In order to avoid deadlocks, whenever we need both locks, we always first
- * get AsyncQueueLock and then AsyncCtlLock.
+ * In order to avoid deadlocks, whenever we need multiple locks, we first get
+ * NotifyQueueTailLock, then AsyncQueueLock, and lastly AsyncCtlLock.
  *
  * Each backend uses the backend[] array entry with index equal to its
  * BackendId (which can range from 1 to MaxBackends).  We rely on this to make
@@ -2013,6 +2016,10 @@ asyncQueueAdvanceTail(void)
    int         newtailpage;
    int         boundary;
 
+   /* Restrict task to one backend per cluster; see SimpleLruTruncate(). */
+   LWLockAcquire(NotifyQueueTailLock, LW_EXCLUSIVE);
+
+   /* Compute the new tail. */
    LWLockAcquire(AsyncQueueLock, LW_EXCLUSIVE);
    min = QUEUE_HEAD;
    for (i = 1; i <= MaxBackends; i++)
@@ -2021,7 +2028,6 @@ asyncQueueAdvanceTail(void)
            min = QUEUE_POS_MIN(min, QUEUE_BACKEND_POS(i));
    }
    oldtailpage = QUEUE_POS_PAGE(QUEUE_TAIL);
-   QUEUE_TAIL = min;
    LWLockRelease(AsyncQueueLock);
 
    /*
@@ -2041,6 +2047,17 @@ asyncQueueAdvanceTail(void)
         */
        SimpleLruTruncate(AsyncCtl, newtailpage);
    }
+
+   /*
+    * Advertise the new tail.  This changes asyncQueueIsFull()'s verdict for
+    * the segment immediately prior to the new tail, allowing fresh data into
+    * that segment.
+    */
+   LWLockAcquire(AsyncQueueLock, LW_EXCLUSIVE);
+   QUEUE_TAIL = min;
+   LWLockRelease(AsyncQueueLock);
+
+   LWLockRelease(NotifyQueueTailLock);
 }
 
 /*

diff --git a/src/backend/commands/vacuum.c b/src/backend/commands/vacuum.c
index 83e807e32aaa89600f1359cffd86b3a14a41c3c5..fa5a3a34bf43c42b4acbc017da61cf727cb61dc6 100644 (file)
--- a/src/backend/commands/vacuum.c
+++ b/src/backend/commands/vacuum.c
@@ -933,6 +933,14 @@ vac_update_datfrozenxid(void)
    bool        bogus = false;
    bool        dirty = false;
 
+   /*
+    * Restrict this task to one backend per database.  This avoids race
+    * conditions that would move datfrozenxid or datminmxid backward.  It
+    * avoids calling vac_truncate_clog() with a datfrozenxid preceding a
+    * datfrozenxid passed to an earlier vac_truncate_clog() call.
+    */
+   LockDatabaseFrozenIds(ExclusiveLock);
+
    /*
     * Initialize the "min" calculation with GetOldestXmin, which is a
     * reasonable approximation to the minimum relfrozenxid for not-yet-
@@ -1097,6 +1105,9 @@ vac_truncate_clog(TransactionId frozenXID,
    bool        bogus = false;
    bool        frozenAlreadyWrapped = false;
 
+   /* Restrict task to one backend per cluster; see SimpleLruTruncate(). */
+   LWLockAcquire(WrapLimitsVacuumLock, LW_EXCLUSIVE);
+
    /* init oldest datoids to sync with my frozenXID/minMulti values */
    oldestxid_datoid = MyDatabaseId;
    minmulti_datoid = MyDatabaseId;
@@ -1206,6 +1217,8 @@ vac_truncate_clog(TransactionId frozenXID,
     */
    SetTransactionIdLimit(frozenXID, oldestxid_datoid);
    SetMultiXactIdLimit(minMulti, minmulti_datoid, false);
+
+   LWLockRelease(WrapLimitsVacuumLock);
 }
 
 

diff --git a/src/backend/storage/lmgr/lmgr.c b/src/backend/storage/lmgr/lmgr.c
index 10e2d028fa8a537bc1bf73b4e103e4042daeca01..9510442628c8644ab86cb2e31b9bfefc0ebcefe2 100644 (file)
--- a/src/backend/storage/lmgr/lmgr.c
+++ b/src/backend/storage/lmgr/lmgr.c
@@ -412,6 +412,21 @@ UnlockRelationForExtension(Relation relation, LOCKMODE lockmode)
    LockRelease(&tag, lockmode, false);
 }
 
+/*
+ *     LockDatabaseFrozenIds
+ *
+ * This allows one backend per database to execute vac_update_datfrozenxid().
+ */
+void
+LockDatabaseFrozenIds(LOCKMODE lockmode)
+{
+   LOCKTAG     tag;
+
+   SET_LOCKTAG_DATABASE_FROZEN_IDS(tag, MyDatabaseId);
+
+   (void) LockAcquire(&tag, lockmode, false, false);
+}
+
 /*
  *     LockPage
  *
@@ -1015,6 +1030,11 @@ DescribeLockTag(StringInfo buf, const LOCKTAG *tag)
                             tag->locktag_field2,
                             tag->locktag_field1);
            break;
+       case LOCKTAG_DATABASE_FROZEN_IDS:
+           appendStringInfo(buf,
+                            _("pg_database.datfrozenxid of database %u"),
+                            tag->locktag_field1);
+           break;
        case LOCKTAG_PAGE:
            appendStringInfo(buf,
                             _("page %u of relation %u of database %u"),

diff --git a/src/backend/storage/lmgr/lwlock.c b/src/backend/storage/lmgr/lwlock.c
index f40af6a1d47ca857566e54757931d23ba96c433f..1965e093b04e55a596e4f57f86e431efeae7daae 100644 (file)
--- a/src/backend/storage/lmgr/lwlock.c
+++ b/src/backend/storage/lmgr/lwlock.c
@@ -495,7 +495,7 @@ RegisterLWLockTranches(void)
 
    if (LWLockTrancheArray == NULL)
    {
-       LWLockTranchesAllocated = 64;
+       LWLockTranchesAllocated = 128;
        LWLockTrancheArray = (char **)
            MemoryContextAllocZero(TopMemoryContext,
                                   LWLockTranchesAllocated * sizeof(char *));

diff --git a/src/backend/storage/lmgr/lwlocknames.txt b/src/backend/storage/lmgr/lwlocknames.txt
index e6025ecedb3ba34e7579a4115510ed294d4f8f6d..04a1786d372f1366715e9754ba6a6b68307a7985 100644 (file)
--- a/src/backend/storage/lmgr/lwlocknames.txt
+++ b/src/backend/storage/lmgr/lwlocknames.txt
@@ -50,3 +50,5 @@ OldSnapshotTimeMapLock                42
 BackendRandomLock                  43
 LogicalRepWorkerLock               44
 CLogTruncationLock                 45
+WrapLimitsVacuumLock               46
+NotifyQueueTailLock                    47

diff --git a/src/backend/utils/adt/lockfuncs.c b/src/backend/utils/adt/lockfuncs.c
index 9e0a8ab79d40e2148df8f6c9ead1e5bed7a8b314..4ed5bf15f9e2bfc8b23541c7f05252e94f1309f1 100644 (file)
--- a/src/backend/utils/adt/lockfuncs.c
+++ b/src/backend/utils/adt/lockfuncs.c
@@ -26,6 +26,7 @@
 const char *const LockTagTypeNames[] = {
    "relation",
    "extend",
+   "frozenid",
    "page",
    "tuple",
    "transactionid",
@@ -245,6 +246,17 @@ pg_lock_status(PG_FUNCTION_ARGS)
                nulls[8] = true;
                nulls[9] = true;
                break;
+           case LOCKTAG_DATABASE_FROZEN_IDS:
+               values[1] = ObjectIdGetDatum(instance->locktag.locktag_field1);
+               nulls[2] = true;
+               nulls[3] = true;
+               nulls[4] = true;
+               nulls[5] = true;
+               nulls[6] = true;
+               nulls[7] = true;
+               nulls[8] = true;
+               nulls[9] = true;
+               break;
            case LOCKTAG_PAGE:
                values[1] = ObjectIdGetDatum(instance->locktag.locktag_field1);
                values[2] = ObjectIdGetDatum(instance->locktag.locktag_field2);

diff --git a/src/include/storage/lmgr.h b/src/include/storage/lmgr.h
index 0b923227a223028fb8143b685edaa67cf686faa5..a37107903a6f4d141dd6a945245b04ea28a2a92b 100644 (file)
--- a/src/include/storage/lmgr.h
+++ b/src/include/storage/lmgr.h
@@ -57,6 +57,9 @@ extern bool ConditionalLockRelationForExtension(Relation relation,
                                    LOCKMODE lockmode);
 extern int RelationExtensionLockWaiterCount(Relation relation);
 
+/* Lock to recompute pg_database.datfrozenxid in the current database */
+extern void LockDatabaseFrozenIds(LOCKMODE lockmode);
+
 /* Lock a page (currently only used within indexes) */
 extern void LockPage(Relation relation, BlockNumber blkno, LOCKMODE lockmode);
 extern bool ConditionalLockPage(Relation relation, BlockNumber blkno, LOCKMODE lockmode);

diff --git a/src/include/storage/lock.h b/src/include/storage/lock.h
index 1cd3685fae84f0785121a91aa74c069a125d46c9..cf1bed6ea4c15cb079b5014958f426b26d1ed7f8 100644 (file)
--- a/src/include/storage/lock.h
+++ b/src/include/storage/lock.h
@@ -141,6 +141,8 @@ typedef enum LockTagType
    /* ID info for a relation is DB OID + REL OID; DB OID = 0 if shared */
    LOCKTAG_RELATION_EXTEND,    /* the right to extend a relation */
    /* same ID info as RELATION */
+   LOCKTAG_DATABASE_FROZEN_IDS,    /* pg_database.datfrozenxid */
+   /* ID info for frozen IDs is DB OID */
    LOCKTAG_PAGE,               /* one page of a relation */
    /* ID info for a page is RELATION info + BlockNumber */
    LOCKTAG_TUPLE,              /* one physical tuple */
@@ -206,6 +208,14 @@ typedef struct LOCKTAG
     (locktag).locktag_type = LOCKTAG_RELATION_EXTEND, \
     (locktag).locktag_lockmethodid = DEFAULT_LOCKMETHOD)
 
+#define SET_LOCKTAG_DATABASE_FROZEN_IDS(locktag,dboid) \
+   ((locktag).locktag_field1 = (dboid), \
+    (locktag).locktag_field2 = 0, \
+    (locktag).locktag_field3 = 0, \
+    (locktag).locktag_field4 = 0, \
+    (locktag).locktag_type = LOCKTAG_DATABASE_FROZEN_IDS, \
+    (locktag).locktag_lockmethodid = DEFAULT_LOCKMETHOD)
+
 #define SET_LOCKTAG_PAGE(locktag,dboid,reloid,blocknum) \
    ((locktag).locktag_field1 = (dboid), \
     (locktag).locktag_field2 = (reloid), \