database/postgres
1
0
Fork 0
mirror of https://github.com/postgres/postgres.git synced 2025-07-18 17:42:25 +03:00
Code Activity

pgindent run.

Browse Source
This commit is contained in:
Bruce Momjian
2003-08-04 00:43:34 +00:00
parent 63354a0228
commit 089003fb46
554 changed files with 24888 additions and 21245 deletions
Download Patch File Download Diff File Expand all files Collapse all files

4
src/backend/access/common/heaptuple.c
View File

@ -9,7 +9,7 @@
*
*
* IDENTIFICATION
* $Header: /cvsroot/pgsql/src/backend/access/common/heaptuple.c,v 1.84 2003/07/21 20:29:37 tgl Exp $
* $Header: /cvsroot/pgsql/src/backend/access/common/heaptuple.c,v 1.85 2003/08/04 00:43:11 momjian Exp $
*
* NOTES
* The old interface functions have been converted to macros
@ -617,7 +617,7 @@ heap_formtuple(TupleDesc tupleDescriptor,
td->t_natts = numberOfAttributes;
td->t_hoff = hoff;
if (tupleDescriptor->tdhasoid) /* else leave infomask = 0 */
if (tupleDescriptor->tdhasoid) /* else leave infomask = 0 */
td->t_infomask = HEAP_HASOID;
DataFill((char *) td + hoff,

8
src/backend/access/common/indextuple.c
View File

@ -9,7 +9,7 @@
*
*
* IDENTIFICATION
* $Header: /cvsroot/pgsql/src/backend/access/common/indextuple.c,v 1.65 2003/07/21 20:29:37 tgl Exp $
* $Header: /cvsroot/pgsql/src/backend/access/common/indextuple.c,v 1.66 2003/08/04 00:43:11 momjian Exp $
*
*-------------------------------------------------------------------------
*/
@ -162,9 +162,9 @@ index_formtuple(TupleDesc tupleDescriptor,
if ((size & INDEX_SIZE_MASK) != size)
ereport(ERROR,
(errcode(ERRCODE_PROGRAM_LIMIT_EXCEEDED),
errmsg("index tuple requires %lu bytes, maximum size is %lu",
(unsigned long) size,
(unsigned long) INDEX_SIZE_MASK)));
errmsg("index tuple requires %lu bytes, maximum size is %lu",
(unsigned long) size,
(unsigned long) INDEX_SIZE_MASK)));
infomask |= size;

58
src/backend/access/common/printtup.c
View File

@ -9,7 +9,7 @@
* Portions Copyright (c) 1994, Regents of the University of California
*
* IDENTIFICATION
* $Header: /cvsroot/pgsql/src/backend/access/common/printtup.c,v 1.75 2003/07/21 20:29:38 tgl Exp $
* $Header: /cvsroot/pgsql/src/backend/access/common/printtup.c,v 1.76 2003/08/04 00:43:12 momjian Exp $
*
*-------------------------------------------------------------------------
*/
@ -24,13 +24,13 @@
static void printtup_startup(DestReceiver *self, int operation,
TupleDesc typeinfo);
TupleDesc typeinfo);
static void printtup(HeapTuple tuple, TupleDesc typeinfo,
DestReceiver *self);
DestReceiver *self);
static void printtup_20(HeapTuple tuple, TupleDesc typeinfo,
DestReceiver *self);
DestReceiver *self);
static void printtup_internal_20(HeapTuple tuple, TupleDesc typeinfo,
DestReceiver *self);
DestReceiver *self);
static void printtup_shutdown(DestReceiver *self);
static void printtup_destroy(DestReceiver *self);
@ -81,8 +81,8 @@ printtup_create_DR(CommandDest dest, Portal portal)
else
{
/*
* In protocol 2.0 the Bind message does not exist, so there is
* no way for the columns to have different print formats; it's
* In protocol 2.0 the Bind message does not exist, so there is no
* way for the columns to have different print formats; it's
* sufficient to look at the first one.
*/
if (portal->formats && portal->formats[0] != 0)
@ -111,12 +111,13 @@ static void
printtup_startup(DestReceiver *self, int operation, TupleDesc typeinfo)
{
DR_printtup *myState = (DR_printtup *) self;
Portal portal = myState->portal;
Portal portal = myState->portal;
if (PG_PROTOCOL_MAJOR(FrontendProtocol) < 3)
{
/*
* Send portal name to frontend (obsolete cruft, gone in proto 3.0)
* Send portal name to frontend (obsolete cruft, gone in proto
* 3.0)
*
* If portal name not specified, use "blank" portal.
*/
@ -129,8 +130,8 @@ printtup_startup(DestReceiver *self, int operation, TupleDesc typeinfo)
}
/*
* If this is a retrieve, and we are supposed to emit row descriptions,
* then we send back the tuple descriptor of the tuples.
* If this is a retrieve, and we are supposed to emit row
* descriptions, then we send back the tuple descriptor of the tuples.
*/
if (operation == CMD_SELECT && myState->sendDescrip)
{
@ -163,7 +164,7 @@ printtup_startup(DestReceiver *self, int operation, TupleDesc typeinfo)
* or some similar function; it does not contain a full set of fields.
* The targetlist will be NIL when executing a utility function that does
* not have a plan. If the targetlist isn't NIL then it is a Query node's
* targetlist; it is up to us to ignore resjunk columns in it. The formats[]
* targetlist; it is up to us to ignore resjunk columns in it. The formats[]
* array pointer might be NULL (if we are doing Describe on a prepared stmt);
* send zeroes for the format codes in that case.
*/
@ -176,14 +177,14 @@ SendRowDescriptionMessage(TupleDesc typeinfo, List *targetlist, int16 *formats)
int i;
StringInfoData buf;
pq_beginmessage(&buf, 'T'); /* tuple descriptor message type */
pq_sendint(&buf, natts, 2); /* # of attrs in tuples */
pq_beginmessage(&buf, 'T'); /* tuple descriptor message type */
pq_sendint(&buf, natts, 2); /* # of attrs in tuples */
for (i = 0; i < natts; ++i)
{
Oid atttypid = attrs[i]->atttypid;
int32 atttypmod = attrs[i]->atttypmod;
Oid basetype;
Oid atttypid = attrs[i]->atttypid;
int32 atttypmod = attrs[i]->atttypmod;
Oid basetype;
pq_sendstring(&buf, NameStr(attrs[i]->attname));
/* column ID info appears in protocol 3.0 and up */
@ -320,8 +321,8 @@ printtup(HeapTuple tuple, TupleDesc typeinfo, DestReceiver *self)
}
/*
* If we have a toasted datum, forcibly detoast it here to
* avoid memory leakage inside the type's output routine.
* If we have a toasted datum, forcibly detoast it here to avoid
* memory leakage inside the type's output routine.
*/
if (thisState->typisvarlena)
attr = PointerGetDatum(PG_DETOAST_DATUM(origattr));
@ -347,7 +348,7 @@ printtup(HeapTuple tuple, TupleDesc typeinfo, DestReceiver *self)
outputbytes = DatumGetByteaP(FunctionCall2(&thisState->finfo,
attr,
ObjectIdGetDatum(thisState->typelem)));
ObjectIdGetDatum(thisState->typelem)));
/* We assume the result will not have been toasted */
pq_sendint(&buf, VARSIZE(outputbytes) - VARHDRSZ, 4);
pq_sendbytes(&buf, VARDATA(outputbytes),
@ -424,8 +425,8 @@ printtup_20(HeapTuple tuple, TupleDesc typeinfo, DestReceiver *self)
Assert(thisState->format == 0);
/*
* If we have a toasted datum, forcibly detoast it here to
* avoid memory leakage inside the type's output routine.
* If we have a toasted datum, forcibly detoast it here to avoid
* memory leakage inside the type's output routine.
*/
if (thisState->typisvarlena)
attr = PointerGetDatum(PG_DETOAST_DATUM(origattr));
@ -536,9 +537,10 @@ debugtup(HeapTuple tuple, TupleDesc typeinfo, DestReceiver *self)
continue;
getTypeOutputInfo(typeinfo->attrs[i]->atttypid,
&typoutput, &typelem, &typisvarlena);
/*
* If we have a toasted datum, forcibly detoast it here to
* avoid memory leakage inside the type's output routine.
* If we have a toasted datum, forcibly detoast it here to avoid
* memory leakage inside the type's output routine.
*/
if (typisvarlena)
attr = PointerGetDatum(PG_DETOAST_DATUM(origattr));
@ -547,7 +549,7 @@ debugtup(HeapTuple tuple, TupleDesc typeinfo, DestReceiver *self)
value = DatumGetCString(OidFunctionCall3(typoutput,
attr,
ObjectIdGetDatum(typelem),
ObjectIdGetDatum(typelem),
Int32GetDatum(typeinfo->attrs[i]->atttypmod)));
printatt((unsigned) i + 1, typeinfo->attrs[i], value);
@ -627,8 +629,8 @@ printtup_internal_20(HeapTuple tuple, TupleDesc typeinfo, DestReceiver *self)
Assert(thisState->format == 1);
/*
* If we have a toasted datum, forcibly detoast it here to
* avoid memory leakage inside the type's output routine.
* If we have a toasted datum, forcibly detoast it here to avoid
* memory leakage inside the type's output routine.
*/
if (thisState->typisvarlena)
attr = PointerGetDatum(PG_DETOAST_DATUM(origattr));
@ -637,7 +639,7 @@ printtup_internal_20(HeapTuple tuple, TupleDesc typeinfo, DestReceiver *self)
outputbytes = DatumGetByteaP(FunctionCall2(&thisState->finfo,
attr,
ObjectIdGetDatum(thisState->typelem)));
ObjectIdGetDatum(thisState->typelem)));
/* We assume the result will not have been toasted */
pq_sendint(&buf, VARSIZE(outputbytes) - VARHDRSZ, 4);
pq_sendbytes(&buf, VARDATA(outputbytes),

9
src/backend/access/gist/gistscan.c
View File

@ -8,7 +8,7 @@
* Portions Copyright (c) 1994, Regents of the University of California
*
* IDENTIFICATION
* $Header: /cvsroot/pgsql/src/backend/access/gist/gistscan.c,v 1.45 2003/07/28 00:09:14 tgl Exp $
* $Header: /cvsroot/pgsql/src/backend/access/gist/gistscan.c,v 1.46 2003/08/04 00:43:12 momjian Exp $
*
*-------------------------------------------------------------------------
*/
@ -104,11 +104,12 @@ gistrescan(PG_FUNCTION_ARGS)
memmove(s->keyData,
key,
s->numberOfKeys * sizeof(ScanKeyData));
/*
* Play games here with the scan key to use the Consistent
* function for all comparisons: 1) the sk_procedure field
* will now be used to hold the strategy number 2) the
* sk_func field will point to the Consistent function
* function for all comparisons: 1) the sk_procedure field will
* now be used to hold the strategy number 2) the sk_func field
* will point to the Consistent function
*/
for (i = 0; i < s->numberOfKeys; i++)
{

20
src/backend/access/hash/hashfunc.c
View File

@ -8,7 +8,7 @@
*
*
* IDENTIFICATION
* $Header: /cvsroot/pgsql/src/backend/access/hash/hashfunc.c,v 1.36 2003/06/22 22:04:54 tgl Exp $
* $Header: /cvsroot/pgsql/src/backend/access/hash/hashfunc.c,v 1.37 2003/08/04 00:43:12 momjian Exp $
*
* NOTES
* These functions are stored in pg_amproc. For each operator class
@ -60,9 +60,9 @@ hashfloat4(PG_FUNCTION_ARGS)
float4 key = PG_GETARG_FLOAT4(0);
/*
* On IEEE-float machines, minus zero and zero have different bit patterns
* but should compare as equal. We must ensure that they have the same
* hash value, which is most easily done this way:
* On IEEE-float machines, minus zero and zero have different bit
* patterns but should compare as equal. We must ensure that they
* have the same hash value, which is most easily done this way:
*/
if (key == (float4) 0)
PG_RETURN_UINT32(0);
@ -76,9 +76,9 @@ hashfloat8(PG_FUNCTION_ARGS)
float8 key = PG_GETARG_FLOAT8(0);
/*
* On IEEE-float machines, minus zero and zero have different bit patterns
* but should compare as equal. We must ensure that they have the same
* hash value, which is most easily done this way:
* On IEEE-float machines, minus zero and zero have different bit
* patterns but should compare as equal. We must ensure that they
* have the same hash value, which is most easily done this way:
*/
if (key == (float8) 0)
PG_RETURN_UINT32(0);
@ -121,9 +121,9 @@ hashtext(PG_FUNCTION_ARGS)
Datum result;
/*
* Note: this is currently identical in behavior to hashvarlena,
* but it seems likely that we may need to do something different
* in non-C locales. (See also hashbpchar, if so.)
* Note: this is currently identical in behavior to hashvarlena, but
* it seems likely that we may need to do something different in non-C
* locales. (See also hashbpchar, if so.)
*/
result = hash_any((unsigned char *) VARDATA(key),
VARSIZE(key) - VARHDRSZ);

6
src/backend/access/hash/hashovfl.c
View File

@ -8,7 +8,7 @@
*
*
* IDENTIFICATION
* $Header: /cvsroot/pgsql/src/backend/access/hash/hashovfl.c,v 1.35 2003/07/21 20:29:38 tgl Exp $
* $Header: /cvsroot/pgsql/src/backend/access/hash/hashovfl.c,v 1.36 2003/08/04 00:43:12 momjian Exp $
*
* NOTES
* Overflow pages look like ordinary relation pages.
@ -205,8 +205,8 @@ _hash_getovfladdr(Relation rel, Buffer *metabufp)
if (++splitnum >= NCACHED)
ereport(ERROR,
(errcode(ERRCODE_PROGRAM_LIMIT_EXCEEDED),
errmsg("out of overflow pages in hash index \"%s\"",
RelationGetRelationName(rel))));
errmsg("out of overflow pages in hash index \"%s\"",
RelationGetRelationName(rel))));
metap->hashm_ovflpoint = splitnum;
metap->hashm_spares[splitnum] = metap->hashm_spares[splitnum - 1];
metap->hashm_spares[splitnum - 1]--;

16
src/backend/access/heap/heapam.c
View File

@ -8,7 +8,7 @@
*
*
* IDENTIFICATION
* $Header: /cvsroot/pgsql/src/backend/access/heap/heapam.c,v 1.152 2003/07/21 20:29:38 tgl Exp $
* $Header: /cvsroot/pgsql/src/backend/access/heap/heapam.c,v 1.153 2003/08/04 00:43:14 momjian Exp $
*
*
* INTERFACE ROUTINES
@ -1132,6 +1132,7 @@ heap_insert(Relation relation, HeapTuple tup, CommandId cid)
xlhdr.t_natts = tup->t_data->t_natts;
xlhdr.t_infomask = tup->t_data->t_infomask;
xlhdr.t_hoff = tup->t_data->t_hoff;
/*
* note we mark rdata[1] as belonging to buffer; if XLogInsert
* decides to write the whole page to the xlog, we don't need to
@ -1149,9 +1150,9 @@ heap_insert(Relation relation, HeapTuple tup, CommandId cid)
rdata[2].next = NULL;
/*
* If this is the single and first tuple on page, we can reinit the
* page instead of restoring the whole thing. Set flag, and hide
* buffer references from XLogInsert.
* If this is the single and first tuple on page, we can reinit
* the page instead of restoring the whole thing. Set flag, and
* hide buffer references from XLogInsert.
*/
if (ItemPointerGetOffsetNumber(&(tup->t_self)) == FirstOffsetNumber &&
PageGetMaxOffsetNumber(page) == FirstOffsetNumber)
@ -1912,7 +1913,7 @@ log_heap_clean(Relation reln, Buffer buffer, OffsetNumber *unused, int uncnt)
/*
* The unused-offsets array is not actually in the buffer, but pretend
* that it is. When XLogInsert stores the whole buffer, the offsets
* that it is. When XLogInsert stores the whole buffer, the offsets
* array need not be stored too.
*/
rdata[1].buffer = buffer;
@ -1991,9 +1992,10 @@ log_heap_update(Relation reln, Buffer oldbuf, ItemPointerData from,
2 * sizeof(TransactionId));
hsize += 2 * sizeof(TransactionId);
}
/*
* As with insert records, we need not store the rdata[2] segment
* if we decide to store the whole buffer instead.
* As with insert records, we need not store the rdata[2] segment if
* we decide to store the whole buffer instead.
*/
rdata[2].buffer = newbuf;
rdata[2].data = (char *) &xlhdr;

51
src/backend/access/index/indexam.c
View File

@ -8,7 +8,7 @@
*
*
* IDENTIFICATION
* $Header: /cvsroot/pgsql/src/backend/access/index/indexam.c,v 1.67 2003/07/21 20:29:39 tgl Exp $
* $Header: /cvsroot/pgsql/src/backend/access/index/indexam.c,v 1.68 2003/08/04 00:43:15 momjian Exp $
*
* INTERFACE ROUTINES
* index_open - open an index relation by relation OID
@ -300,7 +300,7 @@ index_beginscan(Relation heapRelation,
* index_rescan - (re)start a scan of an index
*
* The caller may specify a new set of scankeys (but the number of keys
* cannot change). To restart the scan without changing keys, pass NULL
* cannot change). To restart the scan without changing keys, pass NULL
* for the key array.
*
* Note that this is also called when first starting an indexscan;
@ -394,8 +394,8 @@ index_restrpos(IndexScanDesc scan)
/*
* We do not reset got_tuple; so if the scan is actually being
* short-circuited by index_getnext, the effective position restoration
* is done by restoring unique_tuple_pos.
* short-circuited by index_getnext, the effective position
* restoration is done by restoring unique_tuple_pos.
*/
scan->unique_tuple_pos = scan->unique_tuple_mark;
@ -427,24 +427,24 @@ index_getnext(IndexScanDesc scan, ScanDirection direction)
}
/*
* If we already got a tuple and it must be unique, there's no need
* to make the index AM look through any additional tuples. (This can
* If we already got a tuple and it must be unique, there's no need to
* make the index AM look through any additional tuples. (This can
* save a useful amount of work in scenarios where there are many dead
* tuples due to heavy update activity.)
*
* To do this we must keep track of the logical scan position
* (before/on/after tuple). Also, we have to be sure to release scan
* resources before returning NULL; if we fail to do so then a multi-index
* scan can easily run the system out of free buffers. We can release
* index-level resources fairly cheaply by calling index_rescan. This
* means there are two persistent states as far as the index AM is
* concerned: on-tuple and rescanned. If we are actually asked to
* re-fetch the single tuple, we have to go through a fresh indexscan
* startup, which penalizes that (infrequent) case.
* resources before returning NULL; if we fail to do so then a
* multi-index scan can easily run the system out of free buffers. We
* can release index-level resources fairly cheaply by calling
* index_rescan. This means there are two persistent states as far as
* the index AM is concerned: on-tuple and rescanned. If we are
* actually asked to re-fetch the single tuple, we have to go through
* a fresh indexscan startup, which penalizes that (infrequent) case.
*/
if (scan->keys_are_unique && scan->got_tuple)
{
int new_tuple_pos = scan->unique_tuple_pos;
int new_tuple_pos = scan->unique_tuple_pos;
if (ScanDirectionIsForward(direction))
{
@ -459,22 +459,23 @@ index_getnext(IndexScanDesc scan, ScanDirection direction)
if (new_tuple_pos == 0)
{
/*
* We are moving onto the unique tuple from having been off it.
* We just fall through and let the index AM do the work. Note
* we should get the right answer regardless of scan direction.
* We are moving onto the unique tuple from having been off
* it. We just fall through and let the index AM do the work.
* Note we should get the right answer regardless of scan
* direction.
*/
scan->unique_tuple_pos = 0; /* need to update position */
scan->unique_tuple_pos = 0; /* need to update position */
}
else
{
/*
* Moving off the tuple; must do amrescan to release index-level
* pins before we return NULL. Since index_rescan will reset
* my state, must save and restore...
* Moving off the tuple; must do amrescan to release
* index-level pins before we return NULL. Since index_rescan
* will reset my state, must save and restore...
*/
int unique_tuple_mark = scan->unique_tuple_mark;
int unique_tuple_mark = scan->unique_tuple_mark;
index_rescan(scan, NULL /* no change to key */);
index_rescan(scan, NULL /* no change to key */ );
scan->keys_are_unique = true;
scan->got_tuple = true;
@ -631,7 +632,7 @@ index_bulk_delete(Relation indexRelation,
*/
IndexBulkDeleteResult *
index_vacuum_cleanup(Relation indexRelation,
IndexVacuumCleanupInfo *info,
IndexVacuumCleanupInfo * info,
IndexBulkDeleteResult *stats)
{
RegProcedure procedure;
@ -649,7 +650,7 @@ index_vacuum_cleanup(Relation indexRelation,
DatumGetPointer(OidFunctionCall3(procedure,
PointerGetDatum(indexRelation),
PointerGetDatum((Pointer) info),
PointerGetDatum((Pointer) stats)));
PointerGetDatum((Pointer) stats)));
return result;
}

55
src/backend/access/nbtree/nbtinsert.c
View File

@ -8,7 +8,7 @@
*
*
* IDENTIFICATION
* $Header: /cvsroot/pgsql/src/backend/access/nbtree/nbtinsert.c,v 1.102 2003/07/28 00:09:14 tgl Exp $
* $Header: /cvsroot/pgsql/src/backend/access/nbtree/nbtinsert.c,v 1.103 2003/08/04 00:43:15 momjian Exp $
*
*-------------------------------------------------------------------------
*/
@ -432,9 +432,9 @@ _bt_insertonpg(Relation rel,
*
* must write-lock that page before releasing write lock on
* current page; else someone else's _bt_check_unique scan
* could fail to see our insertion. write locks on intermediate
* dead pages won't do because we don't know when they will get
* de-linked from the tree.
* could fail to see our insertion. write locks on
* intermediate dead pages won't do because we don't know when
* they will get de-linked from the tree.
*/
Buffer rbuf = InvalidBuffer;
@ -523,9 +523,10 @@ _bt_insertonpg(Relation rel,
/*
* If we are doing this insert because we split a page that was
* the only one on its tree level, but was not the root, it may
* have been the "fast root". We need to ensure that the fast root
* link points at or above the current page. We can safely acquire
* a lock on the metapage here --- see comments for _bt_newroot().
* have been the "fast root". We need to ensure that the fast
* root link points at or above the current page. We can safely
* acquire a lock on the metapage here --- see comments for
* _bt_newroot().
*/
if (split_only_page)
{
@ -1135,7 +1136,7 @@ _bt_checksplitloc(FindSplitData *state, OffsetNumber firstright,
*
* On entry, buf and rbuf are the left and right split pages, which we
* still hold write locks on per the L&Y algorithm. We release the
* write locks once we have write lock on the parent page. (Any sooner,
* write locks once we have write lock on the parent page. (Any sooner,
* and it'd be possible for some other process to try to split or delete
* one of these pages, and get confused because it cannot find the downlink.)
*
@ -1155,19 +1156,19 @@ _bt_insert_parent(Relation rel,
bool is_only)
{
/*
* Here we have to do something Lehman and Yao don't talk about:
* deal with a root split and construction of a new root. If our
* stack is empty then we have just split a node on what had been
* the root level when we descended the tree. If it was still the
* root then we perform a new-root construction. If it *wasn't*
* the root anymore, search to find the next higher level that
* someone constructed meanwhile, and find the right place to insert
* as for the normal case.
* Here we have to do something Lehman and Yao don't talk about: deal
* with a root split and construction of a new root. If our stack is
* empty then we have just split a node on what had been the root
* level when we descended the tree. If it was still the root then we
* perform a new-root construction. If it *wasn't* the root anymore,
* search to find the next higher level that someone constructed
* meanwhile, and find the right place to insert as for the normal
* case.
*
* If we have to search for the parent level, we do so by
* re-descending from the root. This is not super-efficient,
* but it's rare enough not to matter. (This path is also taken
* when called from WAL recovery --- we have no stack in that case.)
* If we have to search for the parent level, we do so by re-descending
* from the root. This is not super-efficient, but it's rare enough
* not to matter. (This path is also taken when called from WAL
* recovery --- we have no stack in that case.)
*/
if (is_root)
{
@ -1222,9 +1223,9 @@ _bt_insert_parent(Relation rel,
/*
* Find the parent buffer and get the parent page.
*
* Oops - if we were moved right then we need to change stack
* item! We want to find parent pointing to where we are,
* right ? - vadim 05/27/97
* Oops - if we were moved right then we need to change stack item!
* We want to find parent pointing to where we are, right ? -
* vadim 05/27/97
*/
ItemPointerSet(&(stack->bts_btitem.bti_itup.t_tid),
bknum, P_HIKEY);
@ -1296,16 +1297,16 @@ _bt_getstackbuf(Relation rel, BTStack stack, int access)
/*
* start = InvalidOffsetNumber means "search the whole page".
* We need this test anyway due to possibility that
* page has a high key now when it didn't before.
* We need this test anyway due to possibility that page has a
* high key now when it didn't before.
*/
if (start < minoff)
start = minoff;
/*
* These loops will check every item on the page --- but in an
* order that's attuned to the probability of where it actually
* is. Scan to the right first, then to the left.
* order that's attuned to the probability of where it
* actually is. Scan to the right first, then to the left.
*/
for (offnum = start;
offnum <= maxoff;

149
src/backend/access/nbtree/nbtpage.c
View File

@ -9,7 +9,7 @@
*
*
* IDENTIFICATION
* $Header: /cvsroot/pgsql/src/backend/access/nbtree/nbtpage.c,v 1.66 2003/07/21 20:29:39 tgl Exp $
* $Header: /cvsroot/pgsql/src/backend/access/nbtree/nbtpage.c,v 1.67 2003/08/04 00:43:15 momjian Exp $
*
* NOTES
* Postgres btree pages look like ordinary relation pages. The opaque
@ -181,8 +181,8 @@ _bt_getroot(Relation rel, int access)
/*
* Metadata initialized by someone else. In order to
* guarantee no deadlocks, we have to release the metadata
* page and start all over again. (Is that really true?
* But it's hardly worth trying to optimize this case.)
* page and start all over again. (Is that really true? But
* it's hardly worth trying to optimize this case.)
*/
_bt_relbuf(rel, metabuf);
return _bt_getroot(rel, access);
@ -190,8 +190,8 @@ _bt_getroot(Relation rel, int access)
/*
* Get, initialize, write, and leave a lock of the appropriate
* type on the new root page. Since this is the first page in
* the tree, it's a leaf as well as the root.
* type on the new root page. Since this is the first page in the
* tree, it's a leaf as well as the root.
*/
rootbuf = _bt_getbuf(rel, P_NEW, BT_WRITE);
rootblkno = BufferGetBlockNumber(rootbuf);
@ -240,7 +240,7 @@ _bt_getroot(Relation rel, int access)
_bt_wrtnorelbuf(rel, rootbuf);
/*
* swap root write lock for read lock. There is no danger of
* swap root write lock for read lock. There is no danger of
* anyone else accessing the new root page while it's unlocked,
* since no one else knows where it is yet.
*/
@ -284,8 +284,8 @@ _bt_getroot(Relation rel, int access)
}
/*
* By here, we have a pin and read lock on the root page, and no
* lock set on the metadata page. Return the root page's buffer.
* By here, we have a pin and read lock on the root page, and no lock
* set on the metadata page. Return the root page's buffer.
*/
return rootbuf;
}
@ -299,7 +299,7 @@ _bt_getroot(Relation rel, int access)
* By the time we acquire lock on the root page, it might have been split and
* not be the true root anymore. This is okay for the present uses of this
* routine; we only really need to be able to move up at least one tree level
* from whatever non-root page we were at. If we ever do need to lock the
* from whatever non-root page we were at. If we ever do need to lock the
* one true root page, we could loop here, re-reading the metapage on each
* failure. (Note that it wouldn't do to hold the lock on the metapage while
* moving to the root --- that'd deadlock against any concurrent root split.)
@ -406,9 +406,9 @@ _bt_getbuf(Relation rel, BlockNumber blkno, int access)
* First see if the FSM knows of any free pages.
*
* We can't trust the FSM's report unreservedly; we have to check
* that the page is still free. (For example, an already-free page
* could have been re-used between the time the last VACUUM scanned
* it and the time the VACUUM made its FSM updates.)
* that the page is still free. (For example, an already-free
* page could have been re-used between the time the last VACUUM
* scanned it and the time the VACUUM made its FSM updates.)
*/
for (;;)
{
@ -431,10 +431,10 @@ _bt_getbuf(Relation rel, BlockNumber blkno, int access)
/*
* Extend the relation by one page.
*
* We have to use a lock to ensure no one else is extending the rel at
* the same time, else we will both try to initialize the same new
* page. We can skip locking for new or temp relations, however,
* since no one else could be accessing them.
* We have to use a lock to ensure no one else is extending the rel
* at the same time, else we will both try to initialize the same
* new page. We can skip locking for new or temp relations,
* however, since no one else could be accessing them.
*/
needLock = !(rel->rd_isnew || rel->rd_istemp);
@ -444,8 +444,8 @@ _bt_getbuf(Relation rel, BlockNumber blkno, int access)
buf = ReadBuffer(rel, P_NEW);
/*
* Release the file-extension lock; it's now OK for someone else to
* extend the relation some more.
* Release the file-extension lock; it's now OK for someone else
* to extend the relation some more.
*/
if (needLock)
UnlockPage(rel, 0, ExclusiveLock);
@ -484,7 +484,7 @@ _bt_relbuf(Relation rel, Buffer buf)
* and a pin on the buffer.
*
* NOTE: actually, the buffer manager just marks the shared buffer page
* dirty here; the real I/O happens later. This is okay since we are not
* dirty here; the real I/O happens later. This is okay since we are not
* relying on write ordering anyway. The WAL mechanism is responsible for
* guaranteeing correctness after a crash.
*/
@ -534,13 +534,14 @@ _bt_page_recyclable(Page page)
BTPageOpaque opaque;
/*
* It's possible to find an all-zeroes page in an index --- for example,
* a backend might successfully extend the relation one page and then
* crash before it is able to make a WAL entry for adding the page.
* If we find a zeroed page then reclaim it.
* It's possible to find an all-zeroes page in an index --- for
* example, a backend might successfully extend the relation one page
* and then crash before it is able to make a WAL entry for adding the
* page. If we find a zeroed page then reclaim it.
*/
if (PageIsNew(page))
return true;
/*
* Otherwise, recycle if deleted and too old to have any processes
* interested in it.
@ -565,7 +566,7 @@ _bt_page_recyclable(Page page)
* mistake. On exit, metapage data is correct and we no longer have
* a pin or lock on the metapage.
*
* Actually this is not used for splitting on-the-fly anymore. It's only used
* Actually this is not used for splitting on-the-fly anymore. It's only used
* in nbtsort.c at the completion of btree building, where we know we have
* sole access to the index anyway.
*/
@ -623,7 +624,7 @@ _bt_metaproot(Relation rel, BlockNumber rootbknum, uint32 level)
/*
* Delete item(s) from a btree page.
*
* This must only be used for deleting leaf items. Deleting an item on a
* This must only be used for deleting leaf items. Deleting an item on a
* non-leaf page has to be done as part of an atomic action that includes
* deleting the page it points to.
*
@ -646,9 +647,7 @@ _bt_delitems(Relation rel, Buffer buf,
* adjusting item numbers for previous deletions.
*/
for (i = nitems - 1; i >= 0; i--)
{
PageIndexTupleDelete(page, itemnos[i]);
}
/* XLOG stuff */
if (!rel->rd_istemp)
@ -666,8 +665,8 @@ _bt_delitems(Relation rel, Buffer buf,
rdata[0].next = &(rdata[1]);
/*
* The target-offsets array is not in the buffer, but pretend
* that it is. When XLogInsert stores the whole buffer, the offsets
* The target-offsets array is not in the buffer, but pretend that
* it is. When XLogInsert stores the whole buffer, the offsets
* array need not be stored too.
*/
rdata[1].buffer = buf;
@ -701,7 +700,7 @@ _bt_delitems(Relation rel, Buffer buf,
* may currently be trying to follow links leading to the page; they have to
* be allowed to use its right-link to recover. See nbtree/README.
*
* On entry, the target buffer must be pinned and read-locked. This lock and
* On entry, the target buffer must be pinned and read-locked. This lock and
* pin will be dropped before exiting.
*
* Returns the number of pages successfully deleted (zero on failure; could
@ -714,7 +713,7 @@ _bt_delitems(Relation rel, Buffer buf,
int
_bt_pagedel(Relation rel, Buffer buf, bool vacuum_full)
{
BlockNumber target,
BlockNumber target,
leftsib,
rightsib,
parent;
@ -740,17 +739,18 @@ _bt_pagedel(Relation rel, Buffer buf, bool vacuum_full)
BTPageOpaque opaque;
/*
* We can never delete rightmost pages nor root pages. While at it,
* We can never delete rightmost pages nor root pages. While at it,
* check that page is not already deleted and is empty.
*/
page = BufferGetPage(buf);
opaque = (BTPageOpaque) PageGetSpecialPointer(page);
if (P_RIGHTMOST(opaque) || P_ISROOT(opaque) || P_ISDELETED(opaque) ||
P_FIRSTDATAKEY(opaque) <= PageGetMaxOffsetNumber(page))
P_FIRSTDATAKEY(opaque) <= PageGetMaxOffsetNumber(page))
{
_bt_relbuf(rel, buf);
return 0;
}
/*
* Save info about page, including a copy of its high key (it must
* have one, being non-rightmost).
@ -760,12 +760,13 @@ _bt_pagedel(Relation rel, Buffer buf, bool vacuum_full)
leftsib = opaque->btpo_prev;
itemid = PageGetItemId(page, P_HIKEY);
targetkey = CopyBTItem((BTItem) PageGetItem(page, itemid));
/*
* We need to get an approximate pointer to the page's parent page.
* Use the standard search mechanism to search for the page's high key;
* this will give us a link to either the current parent or someplace
* to its left (if there are multiple equal high keys). To avoid
* deadlocks, we'd better drop the target page lock first.
* Use the standard search mechanism to search for the page's high
* key; this will give us a link to either the current parent or
* someplace to its left (if there are multiple equal high keys). To
* avoid deadlocks, we'd better drop the target page lock first.
*/
_bt_relbuf(rel, buf);
/* we need a scan key to do our search, so build one */
@ -775,9 +776,11 @@ _bt_pagedel(Relation rel, Buffer buf, bool vacuum_full)
&lbuf, BT_READ);
/* don't need a pin on that either */
_bt_relbuf(rel, lbuf);
/*
* If we are trying to delete an interior page, _bt_search did more
* than we needed. Locate the stack item pointing to our parent level.
* than we needed. Locate the stack item pointing to our parent
* level.
*/
ilevel = 0;
for (;;)
@ -789,10 +792,12 @@ _bt_pagedel(Relation rel, Buffer buf, bool vacuum_full)
stack = stack->bts_parent;
ilevel++;
}
/*
* We have to lock the pages we need to modify in the standard order:
* moving right, then up. Else we will deadlock against other writers.
*
* moving right, then up. Else we will deadlock against other
* writers.
*
* So, we need to find and write-lock the current left sibling of the
* target page. The sibling that was current a moment ago could have
* split, so we may have to move right. This search could fail if
@ -823,21 +828,24 @@ _bt_pagedel(Relation rel, Buffer buf, bool vacuum_full)
}
else
lbuf = InvalidBuffer;
/*
* Next write-lock the target page itself. It should be okay to take just
* a write lock not a superexclusive lock, since no scans would stop on an
* empty page.
* Next write-lock the target page itself. It should be okay to take
* just a write lock not a superexclusive lock, since no scans would
* stop on an empty page.
*/
buf = _bt_getbuf(rel, target, BT_WRITE);
page = BufferGetPage(buf);
opaque = (BTPageOpaque) PageGetSpecialPointer(page);
/*
* Check page is still empty etc, else abandon deletion. The empty check
* is necessary since someone else might have inserted into it while
* we didn't have it locked; the others are just for paranoia's sake.
* Check page is still empty etc, else abandon deletion. The empty
* check is necessary since someone else might have inserted into it
* while we didn't have it locked; the others are just for paranoia's
* sake.
*/
if (P_RIGHTMOST(opaque) || P_ISROOT(opaque) || P_ISDELETED(opaque) ||
P_FIRSTDATAKEY(opaque) <= PageGetMaxOffsetNumber(page))
P_FIRSTDATAKEY(opaque) <= PageGetMaxOffsetNumber(page))
{
_bt_relbuf(rel, buf);
if (BufferIsValid(lbuf))
@ -846,14 +854,17 @@ _bt_pagedel(Relation rel, Buffer buf, bool vacuum_full)
}
if (opaque->btpo_prev != leftsib)
elog(ERROR, "left link changed unexpectedly");
/*
* And next write-lock the (current) right sibling.
*/
rightsib = opaque->btpo_next;
rbuf = _bt_getbuf(rel, rightsib, BT_WRITE);
/*
* Next find and write-lock the current parent of the target page.
* This is essentially the same as the corresponding step of splitting.
* This is essentially the same as the corresponding step of
* splitting.
*/
ItemPointerSet(&(stack->bts_btitem.bti_itup.t_tid),
target, P_HIKEY);
@ -863,10 +874,11 @@ _bt_pagedel(Relation rel, Buffer buf, bool vacuum_full)
RelationGetRelationName(rel));
parent = stack->bts_blkno;
poffset = stack->bts_offset;
/*
* If the target is the rightmost child of its parent, then we can't
* delete, unless it's also the only child --- in which case the parent
* changes to half-dead status.
* delete, unless it's also the only child --- in which case the
* parent changes to half-dead status.
*/
page = BufferGetPage(pbuf);
opaque = (BTPageOpaque) PageGetSpecialPointer(page);
@ -893,12 +905,13 @@ _bt_pagedel(Relation rel, Buffer buf, bool vacuum_full)
if (OffsetNumberNext(P_FIRSTDATAKEY(opaque)) == maxoff)
parent_one_child = true;
}
/*
* If we are deleting the next-to-last page on the target's level,
* then the rightsib is a candidate to become the new fast root.
* (In theory, it might be possible to push the fast root even further
* down, but the odds of doing so are slim, and the locking considerations
* daunting.)
* then the rightsib is a candidate to become the new fast root. (In
* theory, it might be possible to push the fast root even further
* down, but the odds of doing so are slim, and the locking
* considerations daunting.)
*
* We can safely acquire a lock on the metapage here --- see comments for
* _bt_newroot().
@ -914,12 +927,13 @@ _bt_pagedel(Relation rel, Buffer buf, bool vacuum_full)
metabuf = _bt_getbuf(rel, BTREE_METAPAGE, BT_WRITE);
metapg = BufferGetPage(metabuf);
metad = BTPageGetMeta(metapg);
/*
* The expected case here is btm_fastlevel == targetlevel+1;
* if the fastlevel is <= targetlevel, something is wrong, and we
* choose to overwrite it to fix it.
* if the fastlevel is <= targetlevel, something is wrong, and
* we choose to overwrite it to fix it.
*/
if (metad->btm_fastlevel > targetlevel+1)
if (metad->btm_fastlevel > targetlevel + 1)
{
/* no update wanted */
_bt_relbuf(rel, metabuf);
@ -937,9 +951,9 @@ _bt_pagedel(Relation rel, Buffer buf, bool vacuum_full)
/*
* Update parent. The normal case is a tad tricky because we want to
* delete the target's downlink and the *following* key. Easiest way is
* to copy the right sibling's downlink over the target downlink, and then
* delete the following item.
* delete the target's downlink and the *following* key. Easiest way
* is to copy the right sibling's downlink over the target downlink,
* and then delete the following item.
*/
page = BufferGetPage(pbuf);
opaque = (BTPageOpaque) PageGetSpecialPointer(page);
@ -950,7 +964,7 @@ _bt_pagedel(Relation rel, Buffer buf, bool vacuum_full)
}
else
{
OffsetNumber nextoffset;
OffsetNumber nextoffset;
itemid = PageGetItemId(page, poffset);
btitem = (BTItem) PageGetItem(page, itemid);
@ -968,8 +982,8 @@ _bt_pagedel(Relation rel, Buffer buf, bool vacuum_full)
}
/*
* Update siblings' side-links. Note the target page's side-links will
* continue to point to the siblings.
* Update siblings' side-links. Note the target page's side-links
* will continue to point to the siblings.
*/
if (BufferIsValid(lbuf))
{
@ -1096,10 +1110,11 @@ _bt_pagedel(Relation rel, Buffer buf, bool vacuum_full)
_bt_wrtbuf(rel, lbuf);
/*
* If parent became half dead, recurse to try to delete it. Otherwise,
* if right sibling is empty and is now the last child of the parent,
* recurse to try to delete it. (These cases cannot apply at the same
* time, though the second case might itself recurse to the first.)
* If parent became half dead, recurse to try to delete it.
* Otherwise, if right sibling is empty and is now the last child of
* the parent, recurse to try to delete it. (These cases cannot apply
* at the same time, though the second case might itself recurse to
* the first.)
*/
if (parent_half_dead)
{

90
src/backend/access/nbtree/nbtree.c
View File

@ -12,7 +12,7 @@
* Portions Copyright (c) 1994, Regents of the University of California
*
* IDENTIFICATION
* $Header: /cvsroot/pgsql/src/backend/access/nbtree/nbtree.c,v 1.103 2003/07/21 20:29:39 tgl Exp $
* $Header: /cvsroot/pgsql/src/backend/access/nbtree/nbtree.c,v 1.104 2003/08/04 00:43:15 momjian Exp $
*
*-------------------------------------------------------------------------
*/
@ -580,19 +580,20 @@ btbulkdelete(PG_FUNCTION_ARGS)
/*
* The outer loop iterates over index leaf pages, the inner over items
* on a leaf page. We issue just one _bt_delitems() call per page,
* so as to minimize WAL traffic.
* on a leaf page. We issue just one _bt_delitems() call per page, so
* as to minimize WAL traffic.
*
* Note that we exclusive-lock every leaf page containing data items,
* in sequence left to right. It sounds attractive to only exclusive-lock
* those containing items we need to delete, but unfortunately that
* is not safe: we could then pass a stopped indexscan, which could
* in rare cases lead to deleting the item it needs to find when it
* resumes. (See _bt_restscan --- this could only happen if an indexscan
* stops on a deletable item and then a page split moves that item
* into a page further to its right, which the indexscan will have no
* pin on.) We can skip obtaining exclusive lock on empty pages
* though, since no indexscan could be stopped on those.
* Note that we exclusive-lock every leaf page containing data items, in
* sequence left to right. It sounds attractive to only
* exclusive-lock those containing items we need to delete, but
* unfortunately that is not safe: we could then pass a stopped
* indexscan, which could in rare cases lead to deleting the item it
* needs to find when it resumes. (See _bt_restscan --- this could
* only happen if an indexscan stops on a deletable item and then a
* page split moves that item into a page further to its right, which
* the indexscan will have no pin on.) We can skip obtaining
* exclusive lock on empty pages though, since no indexscan could be
* stopped on those.
*/
buf = _bt_get_endpoint(rel, 0, false);
if (BufferIsValid(buf)) /* check for empty index */
@ -604,7 +605,7 @@ btbulkdelete(PG_FUNCTION_ARGS)
OffsetNumber offnum,
minoff,
maxoff;
BlockNumber nextpage;
BlockNumber nextpage;
CHECK_FOR_INTERRUPTS();
@ -622,12 +623,14 @@ btbulkdelete(PG_FUNCTION_ARGS)
*/
LockBuffer(buf, BUFFER_LOCK_UNLOCK);
LockBufferForCleanup(buf);
/*
* Recompute minoff/maxoff, both of which could have changed
* while we weren't holding the lock.
* Recompute minoff/maxoff, both of which could have
* changed while we weren't holding the lock.
*/
minoff = P_FIRSTDATAKEY(opaque);
maxoff = PageGetMaxOffsetNumber(page);
/*
* Scan over all items to see which ones need deleted
* according to the callback function.
@ -640,7 +643,7 @@ btbulkdelete(PG_FUNCTION_ARGS)
ItemPointer htup;
btitem = (BTItem) PageGetItem(page,
PageGetItemId(page, offnum));
PageGetItemId(page, offnum));
htup = &(btitem->bti_itup.t_tid);
if (callback(htup, callback_state))
{
@ -651,6 +654,7 @@ btbulkdelete(PG_FUNCTION_ARGS)
num_index_tuples += 1;
}
}
/*
* If we need to delete anything, do it and write the buffer;
* else just release the buffer.
@ -662,9 +666,7 @@ btbulkdelete(PG_FUNCTION_ARGS)
_bt_wrtbuf(rel, buf);
}
else
{
_bt_relbuf(rel, buf);
}
/* And advance to next page, if any */
if (nextpage == P_NONE)
break;
@ -712,7 +714,7 @@ btvacuumcleanup(PG_FUNCTION_ARGS)
/* No point in remembering more than MaxFSMPages pages */
maxFreePages = MaxFSMPages;
if ((BlockNumber) maxFreePages > num_pages)
maxFreePages = (int) num_pages + 1; /* +1 to avoid palloc(0) */
maxFreePages = (int) num_pages + 1; /* +1 to avoid palloc(0) */
freePages = (BlockNumber *) palloc(maxFreePages * sizeof(BlockNumber));
nFreePages = 0;
@ -728,10 +730,10 @@ btvacuumcleanup(PG_FUNCTION_ARGS)
* after we start the scan will not be examined; this should be fine,
* since they can't possibly be empty.)
*/
for (blkno = BTREE_METAPAGE+1; blkno < num_pages; blkno++)
for (blkno = BTREE_METAPAGE + 1; blkno < num_pages; blkno++)
{
Buffer buf;
Page page;
Buffer buf;
Page page;
BTPageOpaque opaque;
buf = _bt_getbuf(rel, blkno, BT_READ);
@ -753,7 +755,7 @@ btvacuumcleanup(PG_FUNCTION_ARGS)
P_FIRSTDATAKEY(opaque) > PageGetMaxOffsetNumber(page))
{
/* Empty, try to delete */
int ndel;
int ndel;
/* Run pagedel in a temp context to avoid memory leakage */
MemoryContextReset(mycontext);
@ -768,7 +770,7 @@ btvacuumcleanup(PG_FUNCTION_ARGS)
/*
* During VACUUM FULL it's okay to recycle deleted pages
* immediately, since there can be no other transactions
* scanning the index. Note that we will only recycle the
* scanning the index. Note that we will only recycle the
* current page and not any parent pages that _bt_pagedel
* might have recursed to; this seems reasonable in the name
* of simplicity. (Trying to do otherwise would mean we'd
@ -787,16 +789,16 @@ btvacuumcleanup(PG_FUNCTION_ARGS)
}
/*
* During VACUUM FULL, we truncate off any recyclable pages at the
* end of the index. In a normal vacuum it'd be unsafe to do this
* except by acquiring exclusive lock on the index and then rechecking
* all the pages; doesn't seem worth it.
* During VACUUM FULL, we truncate off any recyclable pages at the end
* of the index. In a normal vacuum it'd be unsafe to do this except
* by acquiring exclusive lock on the index and then rechecking all
* the pages; doesn't seem worth it.
*/
if (info->vacuum_full && nFreePages > 0)
{
BlockNumber new_pages = num_pages;
BlockNumber new_pages = num_pages;
while (nFreePages > 0 && freePages[nFreePages-1] == new_pages-1)
while (nFreePages > 0 && freePages[nFreePages - 1] == new_pages - 1)
{
new_pages--;
pages_deleted--;
@ -810,9 +812,10 @@ btvacuumcleanup(PG_FUNCTION_ARGS)
* Okay to truncate.
*
* First, flush any shared buffers for the blocks we intend to
* delete. FlushRelationBuffers is a bit more than we need for
* this, since it will also write out dirty buffers for blocks we
* aren't deleting, but it's the closest thing in bufmgr's API.
* delete. FlushRelationBuffers is a bit more than we need
* for this, since it will also write out dirty buffers for
* blocks we aren't deleting, but it's the closest thing in
* bufmgr's API.
*/
i = FlushRelationBuffers(rel, new_pages);
if (i < 0)
@ -822,7 +825,8 @@ btvacuumcleanup(PG_FUNCTION_ARGS)
* Do the physical truncation.
*/
new_pages = smgrtruncate(DEFAULT_SMGR, rel, new_pages);
rel->rd_nblocks = new_pages; /* update relcache immediately */
rel->rd_nblocks = new_pages; /* update relcache
* immediately */
rel->rd_targblock = InvalidBlockNumber;
num_pages = new_pages;
}
@ -856,7 +860,7 @@ btvacuumcleanup(PG_FUNCTION_ARGS)
* and so no deletion can have occurred on that page.
*
* On entry, we have a pin but no read lock on the buffer that contained
* the index tuple we stopped the scan on. On exit, we have pin and read
* the index tuple we stopped the scan on. On exit, we have pin and read
* lock on the buffer that now contains that index tuple, and the scandesc's
* current position is updated to point at it.
*/
@ -877,8 +881,8 @@ _bt_restscan(IndexScanDesc scan)
BlockNumber blkno;
/*
* Reacquire read lock on the buffer. (We should still have
* a reference-count pin on it, so need not get that.)
* Reacquire read lock on the buffer. (We should still have a
* reference-count pin on it, so need not get that.)
*/
LockBuffer(buf, BT_READ);
@ -921,11 +925,11 @@ _bt_restscan(IndexScanDesc scan)
/*
* The item we're looking for moved right at least one page, so
* move right. We are careful here to pin and read-lock the next
* non-dead page before releasing the current one. This ensures that
* a concurrent btbulkdelete scan cannot pass our position --- if it
* did, it might be able to reach and delete our target item before
* we can find it again.
* move right. We are careful here to pin and read-lock the next
* non-dead page before releasing the current one. This ensures
* that a concurrent btbulkdelete scan cannot pass our position
* --- if it did, it might be able to reach and delete our target
* item before we can find it again.
*/
if (P_RIGHTMOST(opaque))
elog(ERROR, "failed to re-find previous key in \"%s\"",

91
src/backend/access/nbtree/nbtsearch.c
View File

@ -8,7 +8,7 @@
* Portions Copyright (c) 1994, Regents of the University of California
*
* IDENTIFICATION
* $Header: /cvsroot/pgsql/src/backend/access/nbtree/nbtsearch.c,v 1.77 2003/07/29 22:18:38 tgl Exp $
* $Header: /cvsroot/pgsql/src/backend/access/nbtree/nbtsearch.c,v 1.78 2003/08/04 00:43:15 momjian Exp $
*
*-------------------------------------------------------------------------
*/
@ -64,8 +64,8 @@ _bt_search(Relation rel, int keysz, ScanKey scankey,
/*
* Race -- the page we just grabbed may have split since we read
* its pointer in the parent (or metapage). If it has, we may need
* to move right to its new sibling. Do that.
* its pointer in the parent (or metapage). If it has, we may
* need to move right to its new sibling. Do that.
*/
*bufP = _bt_moveright(rel, *bufP, keysz, scankey, BT_READ);
@ -87,14 +87,14 @@ _bt_search(Relation rel, int keysz, ScanKey scankey,
par_blkno = BufferGetBlockNumber(*bufP);
/*
* We need to save the location of the index entry we chose in
* the parent page on a stack. In case we split the tree, we'll
* use the stack to work back up to the parent page. We also save
* the actual downlink (TID) to uniquely identify the index entry,
* in case it moves right while we're working lower in the
* tree. See the paper by Lehman and Yao for how this is detected
* and handled. (We use the child link to disambiguate duplicate
* keys in the index -- Lehman and Yao disallow duplicate keys.)
* We need to save the location of the index entry we chose in the
* parent page on a stack. In case we split the tree, we'll use
* the stack to work back up to the parent page. We also save the
* actual downlink (TID) to uniquely identify the index entry, in
* case it moves right while we're working lower in the tree. See
* the paper by Lehman and Yao for how this is detected and
* handled. (We use the child link to disambiguate duplicate keys
* in the index -- Lehman and Yao disallow duplicate keys.)
*/
new_stack = (BTStack) palloc(sizeof(BTStackData));
new_stack->bts_blkno = par_blkno;
@ -151,8 +151,8 @@ _bt_moveright(Relation rel,
* might not need to move right; have to scan the page first anyway.)
* It could even have split more than once, so scan as far as needed.
*
* We also have to move right if we followed a link that brought us to
* a dead page.
* We also have to move right if we followed a link that brought us to a
* dead page.
*/
while (!P_RIGHTMOST(opaque) &&
(P_IGNORE(opaque) ||
@ -599,8 +599,8 @@ _bt_first(IndexScanDesc scan, ScanDirection dir)
/*
* At this point we are positioned at the first item >= scan key, or
* possibly at the end of a page on which all the existing items are
* less than the scan key and we know that everything on later
* pages is greater than or equal to scan key.
* less than the scan key and we know that everything on later pages
* is greater than or equal to scan key.
*
* We could step forward in the latter case, but that'd be a waste of
* time if we want to scan backwards. So, it's now time to examine
@ -851,7 +851,8 @@ _bt_step(IndexScanDesc scan, Buffer *bufP, ScanDirection dir)
}
}
}
else /* backwards scan */
else
/* backwards scan */
{
if (offnum > P_FIRSTDATAKEY(opaque))
offnum = OffsetNumberPrev(offnum);
@ -860,9 +861,9 @@ _bt_step(IndexScanDesc scan, Buffer *bufP, ScanDirection dir)
/*
* Walk left to the next page with data. This is much more
* complex than the walk-right case because of the possibility
* that the page to our left splits while we are in flight to it,
* plus the possibility that the page we were on gets deleted
* after we leave it. See nbtree/README for details.
* that the page to our left splits while we are in flight to
* it, plus the possibility that the page we were on gets
* deleted after we leave it. See nbtree/README for details.
*/
for (;;)
{
@ -877,10 +878,11 @@ _bt_step(IndexScanDesc scan, Buffer *bufP, ScanDirection dir)
}
page = BufferGetPage(*bufP);
opaque = (BTPageOpaque) PageGetSpecialPointer(page);
/*
* Okay, we managed to move left to a non-deleted page.
* Done if it's not half-dead and not empty. Else loop back
* and do it all again.
* Done if it's not half-dead and not empty. Else loop
* back and do it all again.
*/
if (!P_IGNORE(opaque))
{
@ -946,17 +948,18 @@ _bt_walk_left(Relation rel, Buffer buf)
buf = _bt_getbuf(rel, blkno, BT_READ);
page = BufferGetPage(buf);
opaque = (BTPageOpaque) PageGetSpecialPointer(page);
/*
* If this isn't the page we want, walk right till we find
* what we want --- but go no more than four hops (an
* arbitrary limit). If we don't find the correct page by then,
* the most likely bet is that the original page got deleted
* and isn't in the sibling chain at all anymore, not that its
* left sibling got split more than four times.
* If this isn't the page we want, walk right till we find what we
* want --- but go no more than four hops (an arbitrary limit).
* If we don't find the correct page by then, the most likely bet
* is that the original page got deleted and isn't in the sibling
* chain at all anymore, not that its left sibling got split more
* than four times.
*
* Note that it is correct to test P_ISDELETED not P_IGNORE
* here, because half-dead pages are still in the sibling
* chain. Caller must reject half-dead pages if wanted.
* Note that it is correct to test P_ISDELETED not P_IGNORE here,
* because half-dead pages are still in the sibling chain. Caller
* must reject half-dead pages if wanted.
*/
tries = 0;
for (;;)
@ -983,8 +986,8 @@ _bt_walk_left(Relation rel, Buffer buf)
if (P_ISDELETED(opaque))
{
/*
* It was deleted. Move right to first nondeleted page
* (there must be one); that is the page that has acquired the
* It was deleted. Move right to first nondeleted page (there
* must be one); that is the page that has acquired the
* deleted one's keyspace, so stepping left from it will take
* us where we want to be.
*/
@ -1001,18 +1004,18 @@ _bt_walk_left(Relation rel, Buffer buf)
if (!P_ISDELETED(opaque))
break;
}
/*
* Now return to top of loop, resetting obknum to
* point to this nondeleted page, and try again.
* Now return to top of loop, resetting obknum to point to
* this nondeleted page, and try again.
*/
}
else
{
/*
* It wasn't deleted; the explanation had better be
* that the page to the left got split or deleted.
* Without this check, we'd go into an infinite loop
* if there's anything wrong.
* It wasn't deleted; the explanation had better be that the
* page to the left got split or deleted. Without this check,
* we'd go into an infinite loop if there's anything wrong.
*/
if (opaque->btpo_prev == lblkno)
elog(ERROR, "could not find left sibling in \"%s\"",
@ -1028,7 +1031,7 @@ _bt_walk_left(Relation rel, Buffer buf)
* _bt_get_endpoint() -- Find the first or last page on a given tree level
*
* If the index is empty, we will return InvalidBuffer; any other failure
* condition causes ereport(). We will not return a dead page.
* condition causes ereport(). We will not return a dead page.
*
* The returned buffer is pinned and read-locked.
*/
@ -1045,8 +1048,8 @@ _bt_get_endpoint(Relation rel, uint32 level, bool rightmost)
/*
* If we are looking for a leaf page, okay to descend from fast root;
* otherwise better descend from true root. (There is no point in being
* smarter about intermediate levels.)
* otherwise better descend from true root. (There is no point in
* being smarter about intermediate levels.)
*/
if (level == 0)
buf = _bt_getroot(rel, BT_READ);
@ -1066,9 +1069,9 @@ _bt_get_endpoint(Relation rel, uint32 level, bool rightmost)
{
/*
* If we landed on a deleted page, step right to find a live page
* (there must be one). Also, if we want the rightmost page,
* step right if needed to get to it (this could happen if the
* page split since we obtained a pointer to it).
* (there must be one). Also, if we want the rightmost page, step
* right if needed to get to it (this could happen if the page
* split since we obtained a pointer to it).
*/
while (P_IGNORE(opaque) ||
(rightmost && !P_RIGHTMOST(opaque)))

6
src/backend/access/nbtree/nbtsort.c
View File

@ -36,7 +36,7 @@
* Portions Copyright (c) 1994, Regents of the University of California
*
* IDENTIFICATION
* $Header: /cvsroot/pgsql/src/backend/access/nbtree/nbtsort.c,v 1.73 2003/07/21 20:29:39 tgl Exp $
* $Header: /cvsroot/pgsql/src/backend/access/nbtree/nbtsort.c,v 1.74 2003/08/04 00:43:15 momjian Exp $
*
*-------------------------------------------------------------------------
*/
@ -93,7 +93,7 @@ typedef struct BTPageState
static void _bt_blnewpage(Relation index, Buffer *buf, Page *page,
uint32 level);
uint32 level);
static BTPageState *_bt_pagestate(Relation index, uint32 level);
static void _bt_slideleft(Relation index, Buffer buf, Page page);
static void _bt_sortaddtup(Page page, Size itemsize,
@ -469,7 +469,7 @@ _bt_buildadd(Relation index, BTPageState *state, BTItem bti)
oopaque->btpo_next = BufferGetBlockNumber(nbuf);
nopaque->btpo_prev = BufferGetBlockNumber(obuf);
nopaque->btpo_next = P_NONE; /* redundant */
nopaque->btpo_next = P_NONE; /* redundant */
}
/*

224
src/backend/access/nbtree/nbtxlog.c
View File

@ -8,7 +8,7 @@
* Portions Copyright (c) 1994, Regents of the University of California
*
* IDENTIFICATION
* $Header: /cvsroot/pgsql/src/backend/access/nbtree/nbtxlog.c,v 1.3 2003/02/23 22:43:08 tgl Exp $
* $Header: /cvsroot/pgsql/src/backend/access/nbtree/nbtxlog.c,v 1.4 2003/08/04 00:43:15 momjian Exp $
*
*-------------------------------------------------------------------------
*/
@ -29,10 +29,10 @@
typedef struct bt_incomplete_split
{
RelFileNode node; /* the index */
BlockNumber leftblk; /* left half of split */
BlockNumber rightblk; /* right half of split */
BlockNumber leftblk; /* left half of split */
BlockNumber rightblk; /* right half of split */
bool is_root; /* we split the root */
} bt_incomplete_split;
} bt_incomplete_split;
static List *incomplete_splits;
@ -107,7 +107,7 @@ _bt_restore_page(Page page, char *from, int len)
}
static void
_bt_restore_meta(Relation reln, XLogRecPtr lsn,
_bt_restore_meta(Relation reln, XLogRecPtr lsn,
BlockNumber root, uint32 level,
BlockNumber fastroot, uint32 fastlevel)
{
@ -172,7 +172,7 @@ btree_xlog_insert(bool redo, bool isleaf, bool ismeta,
if (!redo || !(record->xl_info & XLR_BKP_BLOCK_1))
{
buffer = XLogReadBuffer(false, reln,
ItemPointerGetBlockNumber(&(xlrec->target.tid)));
ItemPointerGetBlockNumber(&(xlrec->target.tid)));
if (!BufferIsValid(buffer))
elog(PANIC, "btree_insert_%sdo: block unfound", (redo) ? "re" : "un");
page = (Page) BufferGetPage(buffer);
@ -183,13 +183,11 @@ btree_xlog_insert(bool redo, bool isleaf, bool ismeta,
if (redo)
{
if (XLByteLE(lsn, PageGetLSN(page)))
{
UnlockAndReleaseBuffer(buffer);
}
else
{
if (PageAddItem(page, (Item) datapos, datalen,
ItemPointerGetOffsetNumber(&(xlrec->target.tid)),
ItemPointerGetOffsetNumber(&(xlrec->target.tid)),
LP_USED) == InvalidOffsetNumber)
elog(PANIC, "btree_insert_redo: failed to add item");
@ -204,13 +202,9 @@ btree_xlog_insert(bool redo, bool isleaf, bool ismeta,
elog(PANIC, "btree_insert_undo: bad page LSN");
if (!P_ISLEAF(pageop))
{
UnlockAndReleaseBuffer(buffer);
}
else
{
elog(PANIC, "btree_insert_undo: unimplemented");
}
}
}
@ -226,8 +220,8 @@ btree_xlog_insert(bool redo, bool isleaf, bool ismeta,
if (redo && !isleaf && incomplete_splits != NIL)
{
forget_matching_split(reln, xlrec->target.node,
ItemPointerGetBlockNumber(&(xlrec->target.tid)),
ItemPointerGetOffsetNumber(&(xlrec->target.tid)),
ItemPointerGetBlockNumber(&(xlrec->target.tid)),
ItemPointerGetOffsetNumber(&(xlrec->target.tid)),
false);
}
}
@ -238,9 +232,9 @@ btree_xlog_split(bool redo, bool onleft, bool isroot,
{
xl_btree_split *xlrec = (xl_btree_split *) XLogRecGetData(record);
Relation reln;
BlockNumber targetblk;
BlockNumber leftsib;
BlockNumber rightsib;
BlockNumber targetblk;
BlockNumber leftsib;
BlockNumber rightsib;
Buffer buffer;
Page page;
BTPageOpaque pageop;
@ -338,9 +332,7 @@ btree_xlog_split(bool redo, bool onleft, bool isroot,
elog(PANIC, "btree_split_redo: uninitialized next right page");
if (XLByteLE(lsn, PageGetLSN(page)))
{
UnlockAndReleaseBuffer(buffer);
}
else
{
pageop = (BTPageOpaque) PageGetSpecialPointer(page);
@ -357,8 +349,8 @@ btree_xlog_split(bool redo, bool onleft, bool isroot,
if (redo && xlrec->level > 0 && incomplete_splits != NIL)
{
forget_matching_split(reln, xlrec->target.node,
ItemPointerGetBlockNumber(&(xlrec->target.tid)),
ItemPointerGetOffsetNumber(&(xlrec->target.tid)),
ItemPointerGetBlockNumber(&(xlrec->target.tid)),
ItemPointerGetOffsetNumber(&(xlrec->target.tid)),
false);
}
@ -422,10 +414,10 @@ btree_xlog_delete_page(bool redo, bool ismeta,
{
xl_btree_delete_page *xlrec = (xl_btree_delete_page *) XLogRecGetData(record);
Relation reln;
BlockNumber parent;
BlockNumber target;
BlockNumber leftsib;
BlockNumber rightsib;
BlockNumber parent;
BlockNumber target;
BlockNumber leftsib;
BlockNumber rightsib;
Buffer buffer;
Page page;
BTPageOpaque pageop;
@ -451,9 +443,7 @@ btree_xlog_delete_page(bool redo, bool ismeta,
if (PageIsNew((PageHeader) page))
elog(PANIC, "btree_delete_page_redo: uninitialized parent page");
if (XLByteLE(lsn, PageGetLSN(page)))
{
UnlockAndReleaseBuffer(buffer);
}
else
{
OffsetNumber poffset;
@ -469,7 +459,7 @@ btree_xlog_delete_page(bool redo, bool ismeta,
{
ItemId itemid;
BTItem btitem;
OffsetNumber nextoffset;
OffsetNumber nextoffset;
itemid = PageGetItemId(page, poffset);
btitem = (BTItem) PageGetItem(page, itemid);
@ -494,9 +484,7 @@ btree_xlog_delete_page(bool redo, bool ismeta,
if (PageIsNew((PageHeader) page))
elog(PANIC, "btree_delete_page_redo: uninitialized right sibling");
if (XLByteLE(lsn, PageGetLSN(page)))
{
UnlockAndReleaseBuffer(buffer);
}
else
{
pageop = (BTPageOpaque) PageGetSpecialPointer(page);
@ -520,9 +508,7 @@ btree_xlog_delete_page(bool redo, bool ismeta,
if (PageIsNew((PageHeader) page))
elog(PANIC, "btree_delete_page_redo: uninitialized left sibling");
if (XLByteLE(lsn, PageGetLSN(page)))
{
UnlockAndReleaseBuffer(buffer);
}
else
{
pageop = (BTPageOpaque) PageGetSpecialPointer(page);
@ -799,116 +785,116 @@ btree_desc(char *buf, uint8 xl_info, char *rec)
switch (info)
{
case XLOG_BTREE_INSERT_LEAF:
{
xl_btree_insert *xlrec = (xl_btree_insert *) rec;
{
xl_btree_insert *xlrec = (xl_btree_insert *) rec;
strcat(buf, "insert: ");
out_target(buf, &(xlrec->target));
break;
}
strcat(buf, "insert: ");
out_target(buf, &(xlrec->target));
break;
}
case XLOG_BTREE_INSERT_UPPER:
{
xl_btree_insert *xlrec = (xl_btree_insert *) rec;
{
xl_btree_insert *xlrec = (xl_btree_insert *) rec;
strcat(buf, "insert_upper: ");
out_target(buf, &(xlrec->target));
break;
}
strcat(buf, "insert_upper: ");
out_target(buf, &(xlrec->target));
break;
}
case XLOG_BTREE_INSERT_META:
{
xl_btree_insert *xlrec = (xl_btree_insert *) rec;
{
xl_btree_insert *xlrec = (xl_btree_insert *) rec;
strcat(buf, "insert_meta: ");
out_target(buf, &(xlrec->target));
break;
}
strcat(buf, "insert_meta: ");
out_target(buf, &(xlrec->target));
break;
}
case XLOG_BTREE_SPLIT_L:
{
xl_btree_split *xlrec = (xl_btree_split *) rec;
{
xl_btree_split *xlrec = (xl_btree_split *) rec;
strcat(buf, "split_l: ");
out_target(buf, &(xlrec->target));
sprintf(buf + strlen(buf), "; oth %u; rgh %u",
xlrec->otherblk, xlrec->rightblk);
break;
}
strcat(buf, "split_l: ");
out_target(buf, &(xlrec->target));
sprintf(buf + strlen(buf), "; oth %u; rgh %u",
xlrec->otherblk, xlrec->rightblk);
break;
}
case XLOG_BTREE_SPLIT_R:
{
xl_btree_split *xlrec = (xl_btree_split *) rec;
{
xl_btree_split *xlrec = (xl_btree_split *) rec;
strcat(buf, "split_r: ");
out_target(buf, &(xlrec->target));
sprintf(buf + strlen(buf), "; oth %u; rgh %u",
xlrec->otherblk, xlrec->rightblk);
break;
}
strcat(buf, "split_r: ");
out_target(buf, &(xlrec->target));
sprintf(buf + strlen(buf), "; oth %u; rgh %u",
xlrec->otherblk, xlrec->rightblk);
break;
}
case XLOG_BTREE_SPLIT_L_ROOT:
{
xl_btree_split *xlrec = (xl_btree_split *) rec;
{
xl_btree_split *xlrec = (xl_btree_split *) rec;
strcat(buf, "split_l_root: ");
out_target(buf, &(xlrec->target));
sprintf(buf + strlen(buf), "; oth %u; rgh %u",
xlrec->otherblk, xlrec->rightblk);
break;
}
strcat(buf, "split_l_root: ");
out_target(buf, &(xlrec->target));
sprintf(buf + strlen(buf), "; oth %u; rgh %u",
xlrec->otherblk, xlrec->rightblk);
break;
}
case XLOG_BTREE_SPLIT_R_ROOT:
{
xl_btree_split *xlrec = (xl_btree_split *) rec;
{
xl_btree_split *xlrec = (xl_btree_split *) rec;
strcat(buf, "split_r_root: ");
out_target(buf, &(xlrec->target));
sprintf(buf + strlen(buf), "; oth %u; rgh %u",
xlrec->otherblk, xlrec->rightblk);
break;
}
strcat(buf, "split_r_root: ");
out_target(buf, &(xlrec->target));
sprintf(buf + strlen(buf), "; oth %u; rgh %u",
xlrec->otherblk, xlrec->rightblk);
break;
}
case XLOG_BTREE_DELETE:
{
xl_btree_delete *xlrec = (xl_btree_delete *) rec;
{
xl_btree_delete *xlrec = (xl_btree_delete *) rec;
sprintf(buf + strlen(buf), "delete: node %u/%u; blk %u",
xlrec->node.tblNode, xlrec->node.relNode, xlrec->block);
break;
}
sprintf(buf + strlen(buf), "delete: node %u/%u; blk %u",
xlrec->node.tblNode, xlrec->node.relNode, xlrec->block);
break;
}
case XLOG_BTREE_DELETE_PAGE:
case XLOG_BTREE_DELETE_PAGE_META:
{
xl_btree_delete_page *xlrec = (xl_btree_delete_page *) rec;
{
xl_btree_delete_page *xlrec = (xl_btree_delete_page *) rec;
strcat(buf, "delete_page: ");
out_target(buf, &(xlrec->target));
sprintf(buf + strlen(buf), "; dead %u; left %u; right %u",
xlrec->deadblk, xlrec->leftblk, xlrec->rightblk);
break;
}
strcat(buf, "delete_page: ");
out_target(buf, &(xlrec->target));
sprintf(buf + strlen(buf), "; dead %u; left %u; right %u",
xlrec->deadblk, xlrec->leftblk, xlrec->rightblk);
break;
}
case XLOG_BTREE_NEWROOT:
{
xl_btree_newroot *xlrec = (xl_btree_newroot *) rec;
{
xl_btree_newroot *xlrec = (xl_btree_newroot *) rec;
sprintf(buf + strlen(buf), "newroot: node %u/%u; root %u lev %u",
xlrec->node.tblNode, xlrec->node.relNode,
xlrec->rootblk, xlrec->level);
break;
}
sprintf(buf + strlen(buf), "newroot: node %u/%u; root %u lev %u",
xlrec->node.tblNode, xlrec->node.relNode,
xlrec->rootblk, xlrec->level);
break;
}
case XLOG_BTREE_NEWMETA:
{
xl_btree_newmeta *xlrec = (xl_btree_newmeta *) rec;
{
xl_btree_newmeta *xlrec = (xl_btree_newmeta *) rec;
sprintf(buf + strlen(buf), "newmeta: node %u/%u; root %u lev %u fast %u lev %u",
xlrec->node.tblNode, xlrec->node.relNode,
xlrec->meta.root, xlrec->meta.level,
xlrec->meta.fastroot, xlrec->meta.fastlevel);
break;
}
sprintf(buf + strlen(buf), "newmeta: node %u/%u; root %u lev %u fast %u lev %u",
xlrec->node.tblNode, xlrec->node.relNode,
xlrec->meta.root, xlrec->meta.level,
xlrec->meta.fastroot, xlrec->meta.fastlevel);
break;
}
case XLOG_BTREE_NEWPAGE:
{
xl_btree_newpage *xlrec = (xl_btree_newpage *) rec;
{
xl_btree_newpage *xlrec = (xl_btree_newpage *) rec;
sprintf(buf + strlen(buf), "newpage: node %u/%u; page %u",
xlrec->node.tblNode, xlrec->node.relNode,
xlrec->blkno);
break;
}
sprintf(buf + strlen(buf), "newpage: node %u/%u; page %u",
xlrec->node.tblNode, xlrec->node.relNode,
xlrec->blkno);
break;
}
default:
strcat(buf, "UNKNOWN");
break;

10
src/backend/access/rtree/rtscan.c
View File

@ -8,7 +8,7 @@
*
*
* IDENTIFICATION
* $Header: /cvsroot/pgsql/src/backend/access/rtree/Attic/rtscan.c,v 1.45 2003/07/28 00:09:14 tgl Exp $
* $Header: /cvsroot/pgsql/src/backend/access/rtree/Attic/rtscan.c,v 1.46 2003/08/04 00:43:15 momjian Exp $
*
*-------------------------------------------------------------------------
*/
@ -109,10 +109,10 @@ rtrescan(PG_FUNCTION_ARGS)
s->numberOfKeys * sizeof(ScanKeyData));
/*
* Scans on internal pages use different operators than they
* do on leaf pages. For example, if the user wants all boxes
* that exactly match (x1,y1,x2,y2), then on internal pages we
* need to find all boxes that contain (x1,y1,x2,y2).
* Scans on internal pages use different operators than they do on
* leaf pages. For example, if the user wants all boxes that
* exactly match (x1,y1,x2,y2), then on internal pages we need to
* find all boxes that contain (x1,y1,x2,y2).
*/
for (i = 0; i < s->numberOfKeys; i++)
{

4
src/backend/access/transam/clog.c
View File

@ -13,7 +13,7 @@
* Portions Copyright (c) 1996-2002, PostgreSQL Global Development Group
* Portions Copyright (c) 1994, Regents of the University of California
*
* $Header: /cvsroot/pgsql/src/backend/access/transam/clog.c,v 1.16 2003/06/11 22:37:45 momjian Exp $
* $Header: /cvsroot/pgsql/src/backend/access/transam/clog.c,v 1.17 2003/08/04 00:43:15 momjian Exp $
*
*-------------------------------------------------------------------------
*/
@ -73,7 +73,7 @@
static SlruCtlData ClogCtlData;
static SlruCtl ClogCtl = &ClogCtlData;
static int ZeroCLOGPage(int pageno, bool writeXlog);
static bool CLOGPagePrecedes(int page1, int page2);

6
src/backend/access/transam/rmgr.c
View File

@ -3,7 +3,7 @@
*
* Resource managers definition
*
* $Header: /cvsroot/pgsql/src/backend/access/transam/rmgr.c,v 1.10 2003/02/21 00:06:22 tgl Exp $
* $Header: /cvsroot/pgsql/src/backend/access/transam/rmgr.c,v 1.11 2003/08/04 00:43:15 momjian Exp $
*/
#include "postgres.h"
@ -19,7 +19,7 @@
#include "commands/sequence.h"
RmgrData RmgrTable[RM_MAX_ID+1] = {
RmgrData RmgrTable[RM_MAX_ID + 1] = {
{"XLOG", xlog_redo, xlog_undo, xlog_desc, NULL, NULL},
{"Transaction", xact_redo, xact_undo, xact_desc, NULL, NULL},
{"Storage", smgr_redo, smgr_undo, smgr_desc, NULL, NULL},
@ -32,7 +32,7 @@ RmgrData RmgrTable[RM_MAX_ID+1] = {
{"Reserved 9", NULL, NULL, NULL, NULL, NULL},
{"Heap", heap_redo, heap_undo, heap_desc, NULL, NULL},
{"Btree", btree_redo, btree_undo, btree_desc,
btree_xlog_startup, btree_xlog_cleanup},
btree_xlog_startup, btree_xlog_cleanup},
{"Hash", hash_redo, hash_undo, hash_desc, NULL, NULL},
{"Rtree", rtree_redo, rtree_undo, rtree_desc, NULL, NULL},
{"Gist", gist_redo, gist_undo, gist_desc, NULL, NULL},

94
src/backend/access/transam/slru.c
View File

@ -6,7 +6,7 @@
* Portions Copyright (c) 1996-2003, PostgreSQL Global Development Group
* Portions Copyright (c) 1994, Regents of the University of California
*
* $Header: /cvsroot/pgsql/src/backend/access/transam/slru.c,v 1.3 2003/07/28 00:09:14 tgl Exp $
* $Header: /cvsroot/pgsql/src/backend/access/transam/slru.c,v 1.4 2003/08/04 00:43:15 momjian Exp $
*
*-------------------------------------------------------------------------
*/
@ -93,7 +93,7 @@ typedef enum
SLRU_PAGE_CLEAN, /* page is valid and not dirty */
SLRU_PAGE_DIRTY, /* page is valid but needs write */
SLRU_PAGE_WRITE_IN_PROGRESS /* page is being written out */
} SlruPageStatus;
} SlruPageStatus;
/*
* Shared-memory state
@ -117,7 +117,7 @@ typedef struct SlruSharedData
* swapping out the latest page.
*/
int latest_page_number;
} SlruSharedData;
} SlruSharedData;
typedef SlruSharedData *SlruShared;
@ -145,7 +145,7 @@ typedef enum
SLRU_SEEK_FAILED,
SLRU_READ_FAILED,
SLRU_WRITE_FAILED
} SlruErrorCause;
} SlruErrorCause;
static SlruErrorCause slru_errcause;
static int slru_errno;
@ -166,9 +166,9 @@ SimpleLruShmemSize(void)
{
return MAXALIGN(sizeof(SlruSharedData)) + BLCKSZ * NUM_CLOG_BUFFERS
#ifdef EXEC_BACKEND
+ MAXALIGN(sizeof(SlruLockData))
+ MAXALIGN(sizeof(SlruLockData))
#endif
;
;
}
void
@ -183,12 +183,14 @@ SimpleLruInit(SlruCtl ctl, const char *name, const char *subdir)
shared = (SlruShared) ptr;
#ifdef EXEC_BACKEND
/*
* Locks are in shared memory
*/
locks = (SlruLock)(ptr + MAXALIGN(sizeof(SlruSharedData)) +
BLCKSZ * NUM_CLOG_BUFFERS);
locks = (SlruLock) (ptr + MAXALIGN(sizeof(SlruSharedData)) +
BLCKSZ * NUM_CLOG_BUFFERS);
#else
/*
* Locks are in private memory
*/
@ -199,7 +201,7 @@ SimpleLruInit(SlruCtl ctl, const char *name, const char *subdir)
if (!IsUnderPostmaster)
/* Initialize locks and shared memory area */
/* Initialize locks and shared memory area */
{
char *bufptr;
int slotno;
@ -210,8 +212,8 @@ SimpleLruInit(SlruCtl ctl, const char *name, const char *subdir)
memset(shared, 0, sizeof(SlruSharedData));
bufptr = (char *)shared + MAXALIGN(sizeof(SlruSharedData));
bufptr = (char *) shared + MAXALIGN(sizeof(SlruSharedData));
for (slotno = 0; slotno < NUM_CLOG_BUFFERS; slotno++)
{
locks->BufferLocks[slotno] = LWLockAssign();
@ -247,7 +249,7 @@ int
SimpleLruZeroPage(SlruCtl ctl, int pageno)
{
int slotno;
SlruShared shared = (SlruShared) ctl->shared;
SlruShared shared = (SlruShared) ctl->shared;
/* Find a suitable buffer slot for the page */
slotno = SlruSelectLRUPage(ctl, pageno);
@ -285,7 +287,7 @@ SimpleLruZeroPage(SlruCtl ctl, int pageno)
char *
SimpleLruReadPage(SlruCtl ctl, int pageno, TransactionId xid, bool forwrite)
{
SlruShared shared = (SlruShared) ctl->shared;
SlruShared shared = (SlruShared) ctl->shared;
/* Outer loop handles restart if we lose the buffer to someone else */
for (;;)
@ -383,7 +385,7 @@ SimpleLruWritePage(SlruCtl ctl, int slotno)
{
int pageno;
bool ok;
SlruShared shared = (SlruShared) ctl->shared;
SlruShared shared = (SlruShared) ctl->shared;
/* Do nothing if page does not need writing */
if (shared->page_status[slotno] != SLRU_PAGE_DIRTY &&
@ -539,13 +541,13 @@ SlruPhysicalWritePage(SlruCtl ctl, int pageno, int slotno)
* possible for this to need to happen when writing a page that's not
* first in its segment; we assume the OS can cope with that. (Note:
* it might seem that it'd be okay to create files only when
* SimpleLruZeroPage is called for the first page of a segment. However,
* if after a crash and restart the REDO logic elects to replay the
* log from a checkpoint before the latest one, then it's possible
* that we will get commands to set transaction status of transactions
* that have already been truncated from the commit log. Easiest way
* to deal with that is to accept references to nonexistent files here
* and in SlruPhysicalReadPage.)
* SimpleLruZeroPage is called for the first page of a segment.
* However, if after a crash and restart the REDO logic elects to
* replay the log from a checkpoint before the latest one, then it's
* possible that we will get commands to set transaction status of
* transactions that have already been truncated from the commit log.
* Easiest way to deal with that is to accept references to
* nonexistent files here and in SlruPhysicalReadPage.)
*/
fd = BasicOpenFile(path, O_RDWR | PG_BINARY, S_IRUSR | S_IWUSR);
if (fd < 0)
@ -608,37 +610,37 @@ SlruReportIOError(SlruCtl ctl, int pageno, TransactionId xid)
case SLRU_OPEN_FAILED:
ereport(ERROR,
(errcode_for_file_access(),
errmsg("could not access status of transaction %u", xid),
errmsg("could not access status of transaction %u", xid),
errdetail("open of file \"%s\" failed: %m",
path)));
break;
case SLRU_CREATE_FAILED:
ereport(ERROR,
(errcode_for_file_access(),
errmsg("could not access status of transaction %u", xid),
errmsg("could not access status of transaction %u", xid),
errdetail("creation of file \"%s\" failed: %m",
path)));
break;
case SLRU_SEEK_FAILED:
ereport(ERROR,
(errcode_for_file_access(),
errmsg("could not access status of transaction %u", xid),
errdetail("lseek of file \"%s\", offset %u failed: %m",
path, offset)));
errmsg("could not access status of transaction %u", xid),
errdetail("lseek of file \"%s\", offset %u failed: %m",
path, offset)));
break;
case SLRU_READ_FAILED:
ereport(ERROR,
(errcode_for_file_access(),
errmsg("could not access status of transaction %u", xid),
errdetail("read of file \"%s\", offset %u failed: %m",
path, offset)));
errmsg("could not access status of transaction %u", xid),
errdetail("read of file \"%s\", offset %u failed: %m",
path, offset)));
break;
case SLRU_WRITE_FAILED:
ereport(ERROR,
(errcode_for_file_access(),
errmsg("could not access status of transaction %u", xid),
errdetail("write of file \"%s\", offset %u failed: %m",
path, offset)));
errmsg("could not access status of transaction %u", xid),
errdetail("write of file \"%s\", offset %u failed: %m",
path, offset)));
break;
default:
/* can't get here, we trust */
@ -665,6 +667,7 @@ static int
SlruSelectLRUPage(SlruCtl ctl, int pageno)
{
SlruShared shared = (SlruShared) ctl->shared;
/* Outer loop handles restart after I/O */
for (;;)
{
@ -689,7 +692,7 @@ SlruSelectLRUPage(SlruCtl ctl, int pageno)
if (shared->page_status[slotno] == SLRU_PAGE_EMPTY)
return slotno;
if (shared->page_lru_count[slotno] > bestcount &&
shared->page_number[slotno] != shared->latest_page_number)
shared->page_number[slotno] != shared->latest_page_number)
{
bestslot = slotno;
bestcount = shared->page_lru_count[slotno];
@ -705,12 +708,12 @@ SlruSelectLRUPage(SlruCtl ctl, int pageno)
/*
* We need to do I/O. Normal case is that we have to write it
* out, but it's possible in the worst case to have selected a
* read-busy page. In that case we use SimpleLruReadPage to wait for
* the read to complete.
* read-busy page. In that case we use SimpleLruReadPage to wait
* for the read to complete.
*/
if (shared->page_status[bestslot] == SLRU_PAGE_READ_IN_PROGRESS)
(void) SimpleLruReadPage(ctl, shared->page_number[bestslot],
InvalidTransactionId, false);
InvalidTransactionId, false);
else
SimpleLruWritePage(ctl, bestslot);
@ -747,10 +750,11 @@ SimpleLruFlush(SlruCtl ctl, bool checkpoint)
for (slotno = 0; slotno < NUM_CLOG_BUFFERS; slotno++)
{
SimpleLruWritePage(ctl, slotno);
/*
* When called during a checkpoint,
* we cannot assert that the slot is clean now, since another
* process might have re-dirtied it already. That's okay.
* When called during a checkpoint, we cannot assert that the slot
* is clean now, since another process might have re-dirtied it
* already. That's okay.
*/
Assert(checkpoint ||
shared->page_status[slotno] == SLRU_PAGE_EMPTY ||
@ -792,10 +796,10 @@ SimpleLruTruncate(SlruCtl ctl, int cutoffPage)
CreateCheckPoint(false, true);
/*
* Scan shared memory and remove any pages preceding the cutoff
* page, to ensure we won't rewrite them later. (Any dirty pages
* should have been flushed already during the checkpoint, we're just
* being extra careful here.)
* Scan shared memory and remove any pages preceding the cutoff page,
* to ensure we won't rewrite them later. (Any dirty pages should
* have been flushed already during the checkpoint, we're just being
* extra careful here.)
*/
LWLockAcquire(ctl->locks->ControlLock, LW_EXCLUSIVE);
@ -870,7 +874,7 @@ SlruScanDirectory(SlruCtl ctl, int cutoffPage, bool doDeletions)
if (cldir == NULL)
ereport(ERROR,
(errcode_for_file_access(),
errmsg("could not open directory \"%s\": %m", ctl->Dir)));
errmsg("could not open directory \"%s\": %m", ctl->Dir)));
errno = 0;
while ((clde = readdir(cldir)) != NULL)
@ -898,7 +902,7 @@ SlruScanDirectory(SlruCtl ctl, int cutoffPage, bool doDeletions)
if (errno)
ereport(ERROR,
(errcode_for_file_access(),
errmsg("could not read directory \"%s\": %m", ctl->Dir)));
errmsg("could not read directory \"%s\": %m", ctl->Dir)));
closedir(cldir);
return found;

58
src/backend/access/transam/xact.c
View File

@ -8,7 +8,7 @@
*
*
* IDENTIFICATION
* $Header: /cvsroot/pgsql/src/backend/access/transam/xact.c,v 1.149 2003/07/21 20:29:39 tgl Exp $
* $Header: /cvsroot/pgsql/src/backend/access/transam/xact.c,v 1.150 2003/08/04 00:43:15 momjian Exp $
*
* NOTES
* Transaction aborts can now occur two ways:
@ -92,7 +92,7 @@
* AbortTransactionBlock
*
* These are invoked only in response to a user "BEGIN WORK", "COMMIT",
* or "ROLLBACK" command. The tricky part about these functions
* or "ROLLBACK" command. The tricky part about these functions
* is that they are called within the postgres main loop, in between
* the StartTransactionCommand() and CommitTransactionCommand().
*
@ -197,8 +197,8 @@ static TransactionStateData CurrentTransactionStateData = {
0, /* scan command id */
0x0, /* start time */
TRANS_DEFAULT, /* transaction state */
TBLOCK_DEFAULT /* transaction block state from
the client perspective */
TBLOCK_DEFAULT /* transaction block state from the client
* perspective */
};
TransactionState CurrentTransactionState = &CurrentTransactionStateData;
@ -359,7 +359,7 @@ GetCurrentTransactionStartTimeUsec(int *msec)
* TransactionIdIsCurrentTransactionId
*
* During bootstrap, we cheat and say "it's not my transaction ID" even though
* it is. Along with transam.c's cheat to say that the bootstrap XID is
* it is. Along with transam.c's cheat to say that the bootstrap XID is
* already committed, this causes the tqual.c routines to see previously
* inserted tuples as committed, which is what we need during bootstrap.
*/
@ -561,13 +561,13 @@ RecordTransactionCommit(void)
/*
* We must mark the transaction committed in clog if its XID
* appears either in permanent rels or in local temporary rels.
* We test this by seeing if we made transaction-controlled
* entries *OR* local-rel tuple updates. Note that if we made
* only the latter, we have not emitted an XLOG record for our
* commit, and so in the event of a crash the clog update might be
* lost. This is okay because no one else will ever care whether
* we committed.
* appears either in permanent rels or in local temporary rels. We
* test this by seeing if we made transaction-controlled entries
* *OR* local-rel tuple updates. Note that if we made only the
* latter, we have not emitted an XLOG record for our commit, and
* so in the event of a crash the clog update might be lost. This
* is okay because no one else will ever care whether we
* committed.
*/
if (MyLastRecPtr.xrecoff != 0 || MyXactMadeTempRelUpdate)
TransactionIdCommit(xid);
@ -755,9 +755,9 @@ AtAbort_Memory(void)
{
/*
* Make sure we are in a valid context (not a child of
* TopTransactionContext...). Note that it is possible for this
* code to be called when we aren't in a transaction at all; go
* directly to TopMemoryContext in that case.
* TopTransactionContext...). Note that it is possible for this code
* to be called when we aren't in a transaction at all; go directly to
* TopMemoryContext in that case.
*/
if (TopTransactionContext != NULL)
{
@ -891,8 +891,8 @@ CommitTransaction(void)
DeferredTriggerEndXact();
/*
* Similarly, let ON COMMIT management do its thing before we start
* to commit.
* Similarly, let ON COMMIT management do its thing before we start to
* commit.
*/
PreCommit_on_commit_actions();
@ -953,10 +953,10 @@ CommitTransaction(void)
* noncritical resource releasing.
*
* The ordering of operations is not entirely random. The idea is:
* release resources visible to other backends (eg, files, buffer pins);
* then release locks; then release backend-local resources. We want
* to release locks at the point where any backend waiting for us will
* see our transaction as being fully cleaned up.
* release resources visible to other backends (eg, files, buffer
* pins); then release locks; then release backend-local resources.
* We want to release locks at the point where any backend waiting for
* us will see our transaction as being fully cleaned up.
*/
smgrDoPendingDeletes(true);
@ -1064,7 +1064,7 @@ AbortTransaction(void)
}
/*
* Post-abort cleanup. See notes in CommitTransaction() concerning
* Post-abort cleanup. See notes in CommitTransaction() concerning
* ordering.
*/
@ -1194,8 +1194,8 @@ StartTransactionCommand(void)
}
/*
* We must switch to TopTransactionContext before returning. This
* is already done if we called StartTransaction, otherwise not.
* We must switch to TopTransactionContext before returning. This is
* already done if we called StartTransaction, otherwise not.
*/
Assert(TopTransactionContext != NULL);
MemoryContextSwitchTo(TopTransactionContext);
@ -1370,9 +1370,10 @@ PreventTransactionChain(void *stmtNode, const char *stmtType)
if (IsTransactionBlock())
ereport(ERROR,
(errcode(ERRCODE_ACTIVE_SQL_TRANSACTION),
/* translator: %s represents an SQL statement name */
/* translator: %s represents an SQL statement name */
errmsg("%s cannot run inside a transaction block",
stmtType)));
/*
* Are we inside a function call? If the statement's parameter block
* was allocated in QueryContext, assume it is an interactive command.
@ -1381,8 +1382,8 @@ PreventTransactionChain(void *stmtNode, const char *stmtType)
if (!MemoryContextContains(QueryContext, stmtNode))
ereport(ERROR,
(errcode(ERRCODE_ACTIVE_SQL_TRANSACTION),
/* translator: %s represents an SQL statement name */
errmsg("%s cannot be executed from a function", stmtType)));
/* translator: %s represents an SQL statement name */
errmsg("%s cannot be executed from a function", stmtType)));
/* If we got past IsTransactionBlock test, should be in default state */
if (CurrentTransactionState->blockState != TBLOCK_DEFAULT)
elog(ERROR, "cannot prevent transaction chain");
@ -1414,6 +1415,7 @@ RequireTransactionChain(void *stmtNode, const char *stmtType)
*/
if (IsTransactionBlock())
return;
/*
* Are we inside a function call? If the statement's parameter block
* was allocated in QueryContext, assume it is an interactive command.
@ -1423,7 +1425,7 @@ RequireTransactionChain(void *stmtNode, const char *stmtType)
return;
ereport(ERROR,
(errcode(ERRCODE_NO_ACTIVE_SQL_TRANSACTION),
/* translator: %s represents an SQL statement name */
/* translator: %s represents an SQL statement name */
errmsg("%s may only be used in BEGIN/END transaction blocks",
stmtType)));
}

251
src/backend/access/transam/xlog.c
View File

@ -7,7 +7,7 @@
* Portions Copyright (c) 1996-2002, PostgreSQL Global Development Group
* Portions Copyright (c) 1994, Regents of the University of California
*
* $Header: /cvsroot/pgsql/src/backend/access/transam/xlog.c,v 1.120 2003/07/28 00:09:14 tgl Exp $
* $Header: /cvsroot/pgsql/src/backend/access/transam/xlog.c,v 1.121 2003/08/04 00:43:15 momjian Exp $
*
*-------------------------------------------------------------------------
*/
@ -1046,8 +1046,8 @@ XLogWrite(XLogwrtRqst WriteRqst)
if (close(openLogFile) != 0)
ereport(PANIC,
(errcode_for_file_access(),
errmsg("close of log file %u, segment %u failed: %m",
openLogId, openLogSeg)));
errmsg("close of log file %u, segment %u failed: %m",
openLogId, openLogSeg)));
openLogFile = -1;
}
XLByteToPrevSeg(LogwrtResult.Write, openLogId, openLogSeg);
@ -1162,8 +1162,8 @@ XLogWrite(XLogwrtRqst WriteRqst)
if (close(openLogFile) != 0)
ereport(PANIC,
(errcode_for_file_access(),
errmsg("close of log file %u, segment %u failed: %m",
openLogId, openLogSeg)));
errmsg("close of log file %u, segment %u failed: %m",
openLogId, openLogSeg)));
openLogFile = -1;
}
if (openLogFile < 0)
@ -1266,7 +1266,7 @@ XLogFlush(XLogRecPtr record)
XLogCtlInsert *Insert = &XLogCtl->Insert;
uint32 freespace = INSERT_FREESPACE(Insert);
if (freespace < SizeOfXLogRecord) /* buffer is full */
if (freespace < SizeOfXLogRecord) /* buffer is full */
WriteRqstPtr = XLogCtl->xlblocks[Insert->curridx];
else
{
@ -1449,8 +1449,8 @@ XLogFileInit(uint32 log, uint32 seg,
if (fd < 0)
ereport(PANIC,
(errcode_for_file_access(),
errmsg("open of \"%s\" (log file %u, segment %u) failed: %m",
path, log, seg)));
errmsg("open of \"%s\" (log file %u, segment %u) failed: %m",
path, log, seg)));
return (fd);
}
@ -1563,14 +1563,14 @@ XLogFileOpen(uint32 log, uint32 seg, bool econt)
{
ereport(LOG,
(errcode_for_file_access(),
errmsg("open of \"%s\" (log file %u, segment %u) failed: %m",
path, log, seg)));
errmsg("open of \"%s\" (log file %u, segment %u) failed: %m",
path, log, seg)));
return (fd);
}
ereport(PANIC,
(errcode_for_file_access(),
errmsg("open of \"%s\" (log file %u, segment %u) failed: %m",
path, log, seg)));
errmsg("open of \"%s\" (log file %u, segment %u) failed: %m",
path, log, seg)));
}
return (fd);
@ -1621,8 +1621,8 @@ MoveOfflineLogs(uint32 log, uint32 seg, XLogRecPtr endptr)
if (xldir == NULL)
ereport(PANIC,
(errcode_for_file_access(),
errmsg("could not open transaction log directory \"%s\": %m",
XLogDir)));
errmsg("could not open transaction log directory \"%s\": %m",
XLogDir)));
sprintf(lastoff, "%08X%08X", log, seg);
@ -1654,15 +1654,15 @@ MoveOfflineLogs(uint32 log, uint32 seg, XLogRecPtr endptr)
true))
{
ereport(LOG,
(errmsg("recycled transaction log file \"%s\"",
xlde->d_name)));
(errmsg("recycled transaction log file \"%s\"",
xlde->d_name)));
}
else
{
/* No need for any more future segments... */
ereport(LOG,
(errmsg("removing transaction log file \"%s\"",
xlde->d_name)));
(errmsg("removing transaction log file \"%s\"",
xlde->d_name)));
unlink(path);
}
}
@ -1672,8 +1672,8 @@ MoveOfflineLogs(uint32 log, uint32 seg, XLogRecPtr endptr)
if (errno)
ereport(PANIC,
(errcode_for_file_access(),
errmsg("could not read transaction log directory \"%s\": %m",
XLogDir)));
errmsg("could not read transaction log directory \"%s\": %m",
XLogDir)));
closedir(xldir);
}
@ -1746,8 +1746,8 @@ RecordIsValid(XLogRecord *record, XLogRecPtr recptr, int emode)
if (!EQ_CRC64(record->xl_crc, crc))
{
ereport(emode,
(errmsg("bad resource manager data checksum in record at %X/%X",
recptr.xlogid, recptr.xrecoff)));
(errmsg("bad resource manager data checksum in record at %X/%X",
recptr.xlogid, recptr.xrecoff)));
return (false);
}
@ -1769,8 +1769,8 @@ RecordIsValid(XLogRecord *record, XLogRecPtr recptr, int emode)
if (!EQ_CRC64(cbuf, crc))
{
ereport(emode,
(errmsg("bad checksum of backup block %d in record at %X/%X",
i + 1, recptr.xlogid, recptr.xrecoff)));
(errmsg("bad checksum of backup block %d in record at %X/%X",
i + 1, recptr.xlogid, recptr.xrecoff)));
return (false);
}
blk += sizeof(BkpBlock) + BLCKSZ;
@ -1931,7 +1931,7 @@ got_record:;
{
ereport(emode,
(errmsg("invalid resource manager id %u at %X/%X",
record->xl_rmid, RecPtr->xlogid, RecPtr->xrecoff)));
record->xl_rmid, RecPtr->xlogid, RecPtr->xrecoff)));
goto next_record_is_invalid;
}
nextRecord = NULL;
@ -2063,7 +2063,7 @@ ValidXLOGHeader(XLogPageHeader hdr, int emode, bool checkSUI)
{
ereport(emode,
(errmsg("unexpected pageaddr %X/%X in log file %u, segment %u, offset %u",
hdr->xlp_pageaddr.xlogid, hdr->xlp_pageaddr.xrecoff,
hdr->xlp_pageaddr.xlogid, hdr->xlp_pageaddr.xrecoff,
readId, readSeg, readOff)));
return false;
}
@ -2084,7 +2084,7 @@ ValidXLOGHeader(XLogPageHeader hdr, int emode, bool checkSUI)
hdr->xlp_sui > lastReadSUI + 512)
{
ereport(emode,
/* translator: SUI = startup id */
/* translator: SUI = startup id */
(errmsg("out-of-sequence SUI %u (after %u) in log file %u, segment %u, offset %u",
hdr->xlp_sui, lastReadSUI,
readId, readSeg, readOff)));
@ -2235,8 +2235,8 @@ ReadControlFile(void)
ereport(FATAL,
(errmsg("database files are incompatible with server"),
errdetail("The database cluster was initialized with PG_CONTROL_VERSION %d,"
" but the server was compiled with PG_CONTROL_VERSION %d.",
ControlFile->pg_control_version, PG_CONTROL_VERSION),
" but the server was compiled with PG_CONTROL_VERSION %d.",
ControlFile->pg_control_version, PG_CONTROL_VERSION),
errhint("It looks like you need to initdb.")));
/* Now check the CRC. */
INIT_CRC64(crc);
@ -2265,75 +2265,75 @@ ReadControlFile(void)
ereport(FATAL,
(errmsg("database files are incompatible with server"),
errdetail("The database cluster was initialized with CATALOG_VERSION_NO %d,"
" but the server was compiled with CATALOG_VERSION_NO %d.",
ControlFile->catalog_version_no, CATALOG_VERSION_NO),
" but the server was compiled with CATALOG_VERSION_NO %d.",
ControlFile->catalog_version_no, CATALOG_VERSION_NO),
errhint("It looks like you need to initdb.")));
if (ControlFile->blcksz != BLCKSZ)
ereport(FATAL,
(errmsg("database files are incompatible with server"),
errdetail("The database cluster was initialized with BLCKSZ %d,"
" but the server was compiled with BLCKSZ %d.",
ControlFile->blcksz, BLCKSZ),
errhint("It looks like you need to recompile or initdb.")));
errdetail("The database cluster was initialized with BLCKSZ %d,"
" but the server was compiled with BLCKSZ %d.",
ControlFile->blcksz, BLCKSZ),
errhint("It looks like you need to recompile or initdb.")));
if (ControlFile->relseg_size != RELSEG_SIZE)
ereport(FATAL,
(errmsg("database files are incompatible with server"),
errdetail("The database cluster was initialized with RELSEG_SIZE %d,"
" but the server was compiled with RELSEG_SIZE %d.",
" but the server was compiled with RELSEG_SIZE %d.",
ControlFile->relseg_size, RELSEG_SIZE),
errhint("It looks like you need to recompile or initdb.")));
errhint("It looks like you need to recompile or initdb.")));
if (ControlFile->nameDataLen != NAMEDATALEN)
ereport(FATAL,
(errmsg("database files are incompatible with server"),
errdetail("The database cluster was initialized with NAMEDATALEN %d,"
" but the server was compiled with NAMEDATALEN %d.",
" but the server was compiled with NAMEDATALEN %d.",
ControlFile->nameDataLen, NAMEDATALEN),
errhint("It looks like you need to recompile or initdb.")));
errhint("It looks like you need to recompile or initdb.")));
if (ControlFile->funcMaxArgs != FUNC_MAX_ARGS)
ereport(FATAL,
(errmsg("database files are incompatible with server"),
errdetail("The database cluster was initialized with FUNC_MAX_ARGS %d,"
" but the server was compiled with FUNC_MAX_ARGS %d.",
" but the server was compiled with FUNC_MAX_ARGS %d.",
ControlFile->funcMaxArgs, FUNC_MAX_ARGS),
errhint("It looks like you need to recompile or initdb.")));
errhint("It looks like you need to recompile or initdb.")));
#ifdef HAVE_INT64_TIMESTAMP
if (ControlFile->enableIntTimes != TRUE)
ereport(FATAL,
(errmsg("database files are incompatible with server"),
errdetail("The database cluster was initialized without HAVE_INT64_TIMESTAMP"
" but the server was compiled with HAVE_INT64_TIMESTAMP."),
errhint("It looks like you need to recompile or initdb.")));
" but the server was compiled with HAVE_INT64_TIMESTAMP."),
errhint("It looks like you need to recompile or initdb.")));
#else
if (ControlFile->enableIntTimes != FALSE)
ereport(FATAL,
(errmsg("database files are incompatible with server"),
errdetail("The database cluster was initialized with HAVE_INT64_TIMESTAMP"
" but the server was compiled without HAVE_INT64_TIMESTAMP."),
errhint("It looks like you need to recompile or initdb.")));
" but the server was compiled without HAVE_INT64_TIMESTAMP."),
errhint("It looks like you need to recompile or initdb.")));
#endif
if (ControlFile->localeBuflen != LOCALE_NAME_BUFLEN)
ereport(FATAL,
(errmsg("database files are incompatible with server"),
errdetail("The database cluster was initialized with LOCALE_NAME_BUFLEN %d,"
" but the server was compiled with LOCALE_NAME_BUFLEN %d.",
" but the server was compiled with LOCALE_NAME_BUFLEN %d.",
ControlFile->localeBuflen, LOCALE_NAME_BUFLEN),
errhint("It looks like you need to recompile or initdb.")));
errhint("It looks like you need to recompile or initdb.")));
if (setlocale(LC_COLLATE, ControlFile->lc_collate) == NULL)
ereport(FATAL,
(errmsg("database files are incompatible with operating system"),
errdetail("The database cluster was initialized with LC_COLLATE \"%s\","
" which is not recognized by setlocale().",
ControlFile->lc_collate),
errhint("It looks like you need to initdb or install locale support.")));
(errmsg("database files are incompatible with operating system"),
errdetail("The database cluster was initialized with LC_COLLATE \"%s\","
" which is not recognized by setlocale().",
ControlFile->lc_collate),
errhint("It looks like you need to initdb or install locale support.")));
if (setlocale(LC_CTYPE, ControlFile->lc_ctype) == NULL)
ereport(FATAL,
(errmsg("database files are incompatible with operating system"),
errdetail("The database cluster was initialized with LC_CTYPE \"%s\","
" which is not recognized by setlocale().",
ControlFile->lc_ctype),
errhint("It looks like you need to initdb or install locale support.")));
(errmsg("database files are incompatible with operating system"),
errdetail("The database cluster was initialized with LC_CTYPE \"%s\","
" which is not recognized by setlocale().",
ControlFile->lc_ctype),
errhint("It looks like you need to initdb or install locale support.")));
/* Make the fixed locale settings visible as GUC variables, too */
SetConfigOption("lc_collate", ControlFile->lc_collate,
@ -2602,10 +2602,10 @@ StartupXLOG(void)
str_time(ControlFile->time))));
else if (ControlFile->state == DB_IN_RECOVERY)
ereport(LOG,
(errmsg("database system was interrupted while in recovery at %s",
str_time(ControlFile->time)),
errhint("This probably means that some data is corrupted and"
" you will have to use the last backup for recovery.")));
(errmsg("database system was interrupted while in recovery at %s",
str_time(ControlFile->time)),
errhint("This probably means that some data is corrupted and"
" you will have to use the last backup for recovery.")));
else if (ControlFile->state == DB_IN_PRODUCTION)
ereport(LOG,
(errmsg("database system was interrupted at %s",
@ -2637,12 +2637,12 @@ StartupXLOG(void)
checkPointLoc = ControlFile->prevCheckPoint;
ereport(LOG,
(errmsg("using previous checkpoint record at %X/%X",
checkPointLoc.xlogid, checkPointLoc.xrecoff)));
checkPointLoc.xlogid, checkPointLoc.xrecoff)));
InRecovery = true; /* force recovery even if SHUTDOWNED */
}
else
ereport(PANIC,
(errmsg("could not locate a valid checkpoint record")));
(errmsg("could not locate a valid checkpoint record")));
}
LastRec = RecPtr = checkPointLoc;
memcpy(&checkPoint, XLogRecGetData(record), sizeof(CheckPoint));
@ -2665,11 +2665,12 @@ StartupXLOG(void)
ShmemVariableCache->oidCount = 0;
/*
* If it was a shutdown checkpoint, then any following WAL entries were
* created under the next StartUpID; if it was a regular checkpoint then
* any following WAL entries were created under the same StartUpID.
* We must replay WAL entries using the same StartUpID they were created
* under, so temporarily adopt that SUI (see also xlog_redo()).
* If it was a shutdown checkpoint, then any following WAL entries
* were created under the next StartUpID; if it was a regular
* checkpoint then any following WAL entries were created under the
* same StartUpID. We must replay WAL entries using the same StartUpID
* they were created under, so temporarily adopt that SUI (see also
* xlog_redo()).
*/
if (wasShutdown)
ThisStartUpID = checkPoint.ThisStartUpID + 1;
@ -2690,7 +2691,7 @@ StartupXLOG(void)
{
if (wasShutdown)
ereport(PANIC,
(errmsg("invalid redo/undo record in shutdown checkpoint")));
(errmsg("invalid redo/undo record in shutdown checkpoint")));
InRecovery = true;
}
else if (ControlFile->state != DB_SHUTDOWNED)
@ -2699,7 +2700,7 @@ StartupXLOG(void)
/* REDO */
if (InRecovery)
{
int rmid;
int rmid;
ereport(LOG,
(errmsg("database system was not properly shut down; "
@ -2791,8 +2792,8 @@ StartupXLOG(void)
/*
* Tricky point here: readBuf contains the *last* block that the
* LastRec record spans, not the one it starts in. The last block
* is indeed the one we want to use.
* LastRec record spans, not the one it starts in. The last block is
* indeed the one we want to use.
*/
Assert(readOff == (XLogCtl->xlblocks[0].xrecoff - BLCKSZ) % XLogSegSize);
memcpy((char *) Insert->currpage, readBuf, BLCKSZ);
@ -2818,11 +2819,12 @@ StartupXLOG(void)
else
{
/*
* Whenever Write.LogwrtResult points to exactly the end of a page,
* Write.curridx must point to the *next* page (see XLogWrite()).
* Whenever Write.LogwrtResult points to exactly the end of a
* page, Write.curridx must point to the *next* page (see
* XLogWrite()).
*
* Note: it might seem we should do AdvanceXLInsertBuffer() here,
* but we can't since we haven't yet determined the correct StartUpID
* Note: it might seem we should do AdvanceXLInsertBuffer() here, but
* we can't since we haven't yet determined the correct StartUpID
* to put into the new page's header. The first actual attempt to
* insert a log record will advance the insert state.
*/
@ -2859,7 +2861,7 @@ StartupXLOG(void)
if (InRecovery)
{
int rmid;
int rmid;
/*
* Allow resource managers to do any required cleanup.
@ -2885,14 +2887,15 @@ StartupXLOG(void)
ThisStartUpID = ControlFile->checkPointCopy.ThisStartUpID;
/*
* Perform a new checkpoint to update our recovery activity to disk.
* Perform a new checkpoint to update our recovery activity to
* disk.
*
* Note that we write a shutdown checkpoint. This is correct since
* the records following it will use SUI one more than what is shown
* in the checkpoint's ThisStartUpID.
* the records following it will use SUI one more than what is
* shown in the checkpoint's ThisStartUpID.
*
* In case we had to use the secondary checkpoint, make sure that
* it will still be shown as the secondary checkpoint after this
* In case we had to use the secondary checkpoint, make sure that it
* will still be shown as the secondary checkpoint after this
* CreateCheckPoint operation; we don't want the broken primary
* checkpoint to become prevCheckPoint...
*/
@ -2907,10 +2910,10 @@ StartupXLOG(void)
else
{
/*
* If we are not doing recovery, then we saw a checkpoint with nothing
* after it, and we can safely use StartUpID equal to one more than
* the checkpoint's SUI. But just for paranoia's sake, check against
* pg_control too.
* If we are not doing recovery, then we saw a checkpoint with
* nothing after it, and we can safely use StartUpID equal to one
* more than the checkpoint's SUI. But just for paranoia's sake,
* check against pg_control too.
*/
ThisStartUpID = checkPoint.ThisStartUpID;
if (ThisStartUpID < ControlFile->checkPointCopy.ThisStartUpID)
@ -2923,7 +2926,8 @@ StartupXLOG(void)
PreallocXlogFiles(EndOfLog);
/*
* Advance StartUpID to one more than the highest value used previously.
* Advance StartUpID to one more than the highest value used
* previously.
*/
ThisStartUpID++;
XLogCtl->ThisStartUpID = ThisStartUpID;
@ -2973,9 +2977,9 @@ ReadCheckpointRecord(XLogRecPtr RecPtr,
if (!XRecOffIsValid(RecPtr.xrecoff))
{
ereport(LOG,
/* translator: %s is "primary" or "secondary" */
/* translator: %s is "primary" or "secondary" */
(errmsg("invalid %s checkpoint link in control file",
(whichChkpt == 1) ? gettext("primary") : gettext("secondary"))));
(whichChkpt == 1) ? gettext("primary") : gettext("secondary"))));
return NULL;
}
@ -2984,34 +2988,34 @@ ReadCheckpointRecord(XLogRecPtr RecPtr,
if (record == NULL)
{
ereport(LOG,
/* translator: %s is "primary" or "secondary" */
/* translator: %s is "primary" or "secondary" */
(errmsg("invalid %s checkpoint record",
(whichChkpt == 1) ? gettext("primary") : gettext("secondary"))));
(whichChkpt == 1) ? gettext("primary") : gettext("secondary"))));
return NULL;
}
if (record->xl_rmid != RM_XLOG_ID)
{
ereport(LOG,
/* translator: %s is "primary" or "secondary" */
(errmsg("invalid resource manager id in %s checkpoint record",
(whichChkpt == 1) ? gettext("primary") : gettext("secondary"))));
/* translator: %s is "primary" or "secondary" */
(errmsg("invalid resource manager id in %s checkpoint record",
(whichChkpt == 1) ? gettext("primary") : gettext("secondary"))));
return NULL;
}
if (record->xl_info != XLOG_CHECKPOINT_SHUTDOWN &&
record->xl_info != XLOG_CHECKPOINT_ONLINE)
{
ereport(LOG,
/* translator: %s is "primary" or "secondary" */
/* translator: %s is "primary" or "secondary" */
(errmsg("invalid xl_info in %s checkpoint record",
(whichChkpt == 1) ? gettext("primary") : gettext("secondary"))));
(whichChkpt == 1) ? gettext("primary") : gettext("secondary"))));
return NULL;
}
if (record->xl_len != sizeof(CheckPoint))
{
ereport(LOG,
/* translator: %s is "primary" or "secondary" */
/* translator: %s is "primary" or "secondary" */
(errmsg("invalid length of %s checkpoint record",
(whichChkpt == 1) ? gettext("primary") : gettext("secondary"))));
(whichChkpt == 1) ? gettext("primary") : gettext("secondary"))));
return NULL;
}
return record;
@ -3112,10 +3116,11 @@ CreateCheckPoint(bool shutdown, bool force)
if (MyXactMadeXLogEntry)
ereport(ERROR,
(errcode(ERRCODE_ACTIVE_SQL_TRANSACTION),
errmsg("checkpoint cannot be made inside transaction block")));
errmsg("checkpoint cannot be made inside transaction block")));
/*
* Acquire CheckpointLock to ensure only one checkpoint happens at a time.
* Acquire CheckpointLock to ensure only one checkpoint happens at a
* time.
*
* The CheckpointLock can be held for quite a while, which is not good
* because we won't respond to a cancel/die request while waiting for
@ -3149,14 +3154,15 @@ CreateCheckPoint(bool shutdown, bool force)
LWLockAcquire(WALInsertLock, LW_EXCLUSIVE);
/*
* If this isn't a shutdown or forced checkpoint, and we have not inserted
* any XLOG records since the start of the last checkpoint, skip the
* checkpoint. The idea here is to avoid inserting duplicate checkpoints
* when the system is idle. That wastes log space, and more importantly it
* exposes us to possible loss of both current and previous checkpoint
* records if the machine crashes just as we're writing the update.
* (Perhaps it'd make even more sense to checkpoint only when the previous
* checkpoint record is in a different xlog page?)
* If this isn't a shutdown or forced checkpoint, and we have not
* inserted any XLOG records since the start of the last checkpoint,
* skip the checkpoint. The idea here is to avoid inserting duplicate
* checkpoints when the system is idle. That wastes log space, and
* more importantly it exposes us to possible loss of both current and
* previous checkpoint records if the machine crashes just as we're
* writing the update. (Perhaps it'd make even more sense to
* checkpoint only when the previous checkpoint record is in a
* different xlog page?)
*
* We have to make two tests to determine that nothing has happened since
* the start of the last checkpoint: current insertion point must
@ -3204,12 +3210,13 @@ CreateCheckPoint(bool shutdown, bool force)
* Here we update the shared RedoRecPtr for future XLogInsert calls;
* this must be done while holding the insert lock AND the info_lck.
*
* Note: if we fail to complete the checkpoint, RedoRecPtr will be
* left pointing past where it really needs to point. This is okay;
* the only consequence is that XLogInsert might back up whole buffers
* that it didn't really need to. We can't postpone advancing RedoRecPtr
* because XLogInserts that happen while we are dumping buffers must
* assume that their buffer changes are not included in the checkpoint.
* Note: if we fail to complete the checkpoint, RedoRecPtr will be left
* pointing past where it really needs to point. This is okay; the
* only consequence is that XLogInsert might back up whole buffers
* that it didn't really need to. We can't postpone advancing
* RedoRecPtr because XLogInserts that happen while we are dumping
* buffers must assume that their buffer changes are not included in
* the checkpoint.
*/
{
/* use volatile pointer to prevent code rearrangement */
@ -3538,15 +3545,15 @@ assign_xlog_sync_method(const char *method, bool doit, bool interactive)
if (pg_fsync(openLogFile) != 0)
ereport(PANIC,
(errcode_for_file_access(),
errmsg("fsync of log file %u, segment %u failed: %m",
openLogId, openLogSeg)));
errmsg("fsync of log file %u, segment %u failed: %m",
openLogId, openLogSeg)));
if (open_sync_bit != new_sync_bit)
{
if (close(openLogFile) != 0)
ereport(PANIC,
(errcode_for_file_access(),
errmsg("close of log file %u, segment %u failed: %m",
openLogId, openLogSeg)));
errmsg("close of log file %u, segment %u failed: %m",
openLogId, openLogSeg)));
openLogFile = -1;
}
}
@ -3570,16 +3577,16 @@ issue_xlog_fsync(void)
if (pg_fsync(openLogFile) != 0)
ereport(PANIC,
(errcode_for_file_access(),
errmsg("fsync of log file %u, segment %u failed: %m",
openLogId, openLogSeg)));
errmsg("fsync of log file %u, segment %u failed: %m",
openLogId, openLogSeg)));
break;
#ifdef HAVE_FDATASYNC
case SYNC_METHOD_FDATASYNC:
if (pg_fdatasync(openLogFile) != 0)
ereport(PANIC,
(errcode_for_file_access(),
errmsg("fdatasync of log file %u, segment %u failed: %m",
openLogId, openLogSeg)));
errmsg("fdatasync of log file %u, segment %u failed: %m",
openLogId, openLogSeg)));
break;
#endif
case SYNC_METHOD_OPEN: