static void
gistRedoPageSplitRecord(XLogReaderState *record)
{
XLogRecPtr lsn = record->EndRecPtr;
gistxlogPageSplit *xldata = (gistxlogPageSplit *) XLogRecGetData(record);
Buffer firstbuffer = InvalidBuffer;
Buffer buffer;
Page page;
int i;
bool isrootsplit = false;
/*
* We must hold lock on the first-listed page throughout the action,
* including while updating the left child page (if any). We can unlock
* remaining pages in the list as soon as they've been written, because
* there is no path for concurrent queries to reach those pages without
* first visiting the first-listed page.
*/
/* loop around all pages */
for (i = 0; i < xldata->npage; i++)
{
int flags;
char *data;
Size datalen;
int num;
BlockNumber blkno;
IndexTuple *tuples;
XLogRecGetBlockTag(record, i + 1, NULL, NULL, &blkno);
if (blkno == GIST_ROOT_BLKNO)
{
Assert(i == 0);
isrootsplit = true;
}
buffer = XLogInitBufferForRedo(record, i + 1);
page = (Page) BufferGetPage(buffer);
data = XLogRecGetBlockData(record, i + 1, &datalen);
tuples = decodePageSplitRecord(data, datalen, &num);
/* ok, clear buffer */
if (xldata->origleaf && blkno != GIST_ROOT_BLKNO)
flags = F_LEAF;
else
flags = 0;
GISTInitBuffer(buffer, flags);
/* and fill it */
gistfillbuffer(page, tuples, num, FirstOffsetNumber);
if (blkno == GIST_ROOT_BLKNO)
{
GistPageGetOpaque(page)->rightlink = InvalidBlockNumber;
GistPageSetNSN(page, xldata->orignsn);
GistClearFollowRight(page);
}
else
{
if (i < xldata->npage - 1)
{
BlockNumber nextblkno;
XLogRecGetBlockTag(record, i + 2, NULL, NULL, &nextblkno);
GistPageGetOpaque(page)->rightlink = nextblkno;
}
else
GistPageGetOpaque(page)->rightlink = xldata->origrlink;
GistPageSetNSN(page, xldata->orignsn);
if (i < xldata->npage - 1 && !isrootsplit &&
xldata->markfollowright)
GistMarkFollowRight(page);
else
GistClearFollowRight(page);
}
PageSetLSN(page, lsn);
MarkBufferDirty(buffer);
if (i == 0)
firstbuffer = buffer;
else
UnlockReleaseBuffer(buffer);
}
/* Fix follow-right data on left child page, if any */
if (XLogRecHasBlockRef(record, 0))
gistRedoClearFollowRight(record, 0);
/* Finally, release lock on the first page */
UnlockReleaseBuffer(firstbuffer);
}
/*
* redo any page update (except page split)
*/
static void
gistRedoPageUpdateRecord(XLogReaderState *record)
{
XLogRecPtr lsn = record->EndRecPtr;
gistxlogPageUpdate *xldata = (gistxlogPageUpdate *) XLogRecGetData(record);
Buffer buffer;
Page page;
if (XLogReadBufferForRedo(record, 0, &buffer) == BLK_NEEDS_REDO)
{
char *begin;
char *data;
Size datalen;
int ninserted = 0;
data = begin = XLogRecGetBlockData(record, 0, &datalen);
page = (Page) BufferGetPage(buffer);
if (xldata->ntodelete == 1 && xldata->ntoinsert == 1)
{
/*
* When replacing one tuple with one other tuple, we must use
* PageIndexTupleOverwrite for consistency with gistplacetopage.
*/
OffsetNumber offnum = *((OffsetNumber *) data);
IndexTuple itup;
Size itupsize;
data += sizeof(OffsetNumber);
itup = (IndexTuple) data;
itupsize = IndexTupleSize(itup);
if (!PageIndexTupleOverwrite(page, offnum, (Item) itup, itupsize))
elog(ERROR, "failed to add item to GiST index page, size %d bytes",
(int) itupsize);
data += itupsize;
/* should be nothing left after consuming 1 tuple */
Assert(data - begin == datalen);
/* update insertion count for assert check below */
ninserted++;
}
else if (xldata->ntodelete > 0)
{
/* Otherwise, delete old tuples if any */
OffsetNumber *todelete = (OffsetNumber *) data;
data += sizeof(OffsetNumber) * xldata->ntodelete;
PageIndexMultiDelete(page, todelete, xldata->ntodelete);
if (GistPageIsLeaf(page))
GistMarkTuplesDeleted(page);
}
/* Add new tuples if any */
if (data - begin < datalen)
{
OffsetNumber off = (PageIsEmpty(page)) ? FirstOffsetNumber :
OffsetNumberNext(PageGetMaxOffsetNumber(page));
while (data - begin < datalen)
{
IndexTuple itup = (IndexTuple) data;
Size sz = IndexTupleSize(itup);
OffsetNumber l;
data += sz;
l = PageAddItem(page, (Item) itup, sz, off, false, false);
if (l == InvalidOffsetNumber)
elog(ERROR, "failed to add item to GiST index page, size %d bytes",
(int) sz);
off++;
ninserted++;
}
}
/* Check that XLOG record contained expected number of tuples */
Assert(ninserted == xldata->ntoinsert);
PageSetLSN(page, lsn);
MarkBufferDirty(buffer);
}
/*
* Fix follow-right data on left child page
*
* This must be done while still holding the lock on the target page. Note
* that even if the target page no longer exists, we still attempt to
* replay the change on the child page.
*/
if (XLogRecHasBlockRef(record, 1))
gistRedoClearFollowRight(record, 1);
if (BufferIsValid(buffer))
UnlockReleaseBuffer(buffer);
}
/*
* Write an empty XLOG file, containing only the checkpoint record
* already set up in ControlFile.
*/
static void
WriteEmptyXLOG(void)
{
char *buffer;
XLogPageHeader page;
XLogLongPageHeader longpage;
XLogRecord *record;
pg_crc32 crc;
char path[MAXPGPATH];
int fd;
int nbytes;
FILE *fp = NULL;
char *pch = NULL;
char buf[BUFFER_LEN];
/* Use malloc() to ensure buffer is MAXALIGNED */
buffer = (char *) malloc(XLOG_BLCKSZ);
page = (XLogPageHeader) buffer;
memset(buffer, 0, XLOG_BLCKSZ);
/* Set up the XLOG page header */
page->xlp_magic = XLOG_PAGE_MAGIC;
page->xlp_info = XLP_LONG_HEADER;
page->xlp_tli = ControlFile.checkPointCopy.ThisTimeLineID;
page->xlp_pageaddr.xlogid =
ControlFile.checkPointCopy.redo.xlogid;
page->xlp_pageaddr.xrecoff =
ControlFile.checkPointCopy.redo.xrecoff - SizeOfXLogLongPHD;
longpage = (XLogLongPageHeader) page;
longpage->xlp_sysid = ControlFile.system_identifier;
longpage->xlp_seg_size = XLogSegSize;
longpage->xlp_xlog_blcksz = XLOG_BLCKSZ;
/* Insert the initial checkpoint record */
record = (XLogRecord *) ((char *) page + SizeOfXLogLongPHD);
record->xl_prev.xlogid = 0;
record->xl_prev.xrecoff = 0;
record->xl_xid = InvalidTransactionId;
record->xl_tot_len = SizeOfXLogRecord + sizeof(CheckPoint);
record->xl_len = sizeof(CheckPoint);
record->xl_info = XLOG_CHECKPOINT_SHUTDOWN;
record->xl_rmid = RM_XLOG_ID;
memcpy(XLogRecGetData(record), &ControlFile.checkPointCopy,
sizeof(CheckPoint));
INIT_CRC32C(crc);
COMP_CRC32C(crc, &ControlFile.checkPointCopy, sizeof(CheckPoint));
COMP_CRC32C(crc, (char *) record + sizeof(pg_crc32), SizeOfXLogRecord - sizeof(pg_crc32));
FIN_CRC32C(crc);
record->xl_crc = crc;
/*
* If we make the filespace for transaction files configurable, then
* pg_resetxlog should pick up the XLOG files from the right location.
* Check the flat file for determining the right location of XLOG files
*/
fp = fopen(TXN_FILESPACE_FLATFILE, "r");
if (fp)
{
MemSet(buf, 0, BUFFER_LEN);
if (fgets(buf, BUFFER_LEN, fp))
; /* First line is Filespace OID, skip it */
MemSet(buf, 0, BUFFER_LEN);
if (fgets(buf, BUFFER_LEN, fp))
{
buf[strlen(buf)-1]='\0';
pch = strtok(buf, " "); /* The first part is DBID. Skip it */
pch = strtok(NULL, " ");
sprintf(path,"%s/%s", pch, XLOGDIR);
}
}
else
{
/* No flat file. Use the default pg_system filespace */
sprintf(path, "%s", XLOGDIR);
}
/* Write the first page */
XLogFilePath2(path, ControlFile.checkPointCopy.ThisTimeLineID,
newXlogId, newXlogSeg);
unlink(path);
fd = open(path, O_RDWR | O_CREAT | O_EXCL | PG_BINARY,
S_IRUSR | S_IWUSR);
if (fd < 0)
{
fprintf(stderr, _("%s: could not open file \"%s\": %s\n"),
progname, path, strerror(errno));
exit(1);
}
errno = 0;
if (write(fd, buffer, XLOG_BLCKSZ) != XLOG_BLCKSZ)
{
/* if write didn't set errno, assume problem is no disk space */
if (errno == 0)
errno = ENOSPC;
fprintf(stderr, _("%s: could not write file \"%s\": %s\n"),
progname, path, strerror(errno));
exit(1);
}
/* Fill the rest of the file with zeroes */
memset(buffer, 0, XLOG_BLCKSZ);
for (nbytes = XLOG_BLCKSZ; nbytes < XLogSegSize; nbytes += XLOG_BLCKSZ)
{
errno = 0;
if (write(fd, buffer, XLOG_BLCKSZ) != XLOG_BLCKSZ)
{
if (errno == 0)
errno = ENOSPC;
fprintf(stderr, _("%s: could not write file \"%s\": %s\n"),
progname, path, strerror(errno));
exit(1);
}
}
if (fsync(fd) != 0)
{
fprintf(stderr, _("%s: fsync error: %s\n"), progname, strerror(errno));
exit(1);
}
close(fd);
}
static void
spgRedoMoveLeafs(XLogReaderState *record)
{
XLogRecPtr lsn = record->EndRecPtr;
char *ptr = XLogRecGetData(record);
spgxlogMoveLeafs *xldata = (spgxlogMoveLeafs *) ptr;
SpGistState state;
OffsetNumber *toDelete;
OffsetNumber *toInsert;
int nInsert;
Buffer buffer;
Page page;
XLogRedoAction action;
BlockNumber blknoDst;
XLogRecGetBlockTag(record, 1, NULL, NULL, &blknoDst);
fillFakeState(&state, xldata->stateSrc);
nInsert = xldata->replaceDead ? 1 : xldata->nMoves + 1;
ptr += SizeOfSpgxlogMoveLeafs;
toDelete = (OffsetNumber *) ptr;
ptr += sizeof(OffsetNumber) * xldata->nMoves;
toInsert = (OffsetNumber *) ptr;
ptr += sizeof(OffsetNumber) * nInsert;
/* now ptr points to the list of leaf tuples */
/*
* In normal operation we would have all three pages (source, dest, and
* parent) locked simultaneously; but in WAL replay it should be safe to
* update them one at a time, as long as we do it in the right order.
*/
/* Insert tuples on the dest page (do first, so redirect is valid) */
if (xldata->newPage)
{
buffer = XLogInitBufferForRedo(record, 1);
SpGistInitBuffer(buffer,
SPGIST_LEAF | (xldata->storesNulls ? SPGIST_NULLS : 0));
action = BLK_NEEDS_REDO;
}
else
action = XLogReadBufferForRedo(record, 1, &buffer);
if (action == BLK_NEEDS_REDO)
{
int i;
page = BufferGetPage(buffer);
for (i = 0; i < nInsert; i++)
{
char *leafTuple;
SpGistLeafTupleData leafTupleHdr;
/*
* the tuples are not aligned, so must copy to access the size
* field.
*/
leafTuple = ptr;
memcpy(&leafTupleHdr, leafTuple,
sizeof(SpGistLeafTupleData));
addOrReplaceTuple(page, (Item) leafTuple,
leafTupleHdr.size, toInsert[i]);
ptr += leafTupleHdr.size;
}
PageSetLSN(page, lsn);
MarkBufferDirty(buffer);
}
if (BufferIsValid(buffer))
UnlockReleaseBuffer(buffer);
/* Delete tuples from the source page, inserting a redirection pointer */
if (XLogReadBufferForRedo(record, 0, &buffer) == BLK_NEEDS_REDO)
{
page = BufferGetPage(buffer);
spgPageIndexMultiDelete(&state, page, toDelete, xldata->nMoves,
state.isBuild ? SPGIST_PLACEHOLDER : SPGIST_REDIRECT,
SPGIST_PLACEHOLDER,
blknoDst,
toInsert[nInsert - 1]);
PageSetLSN(page, lsn);
MarkBufferDirty(buffer);
}
if (BufferIsValid(buffer))
UnlockReleaseBuffer(buffer);
/* And update the parent downlink */
if (XLogReadBufferForRedo(record, 2, &buffer) == BLK_NEEDS_REDO)
{
SpGistInnerTuple tuple;
page = BufferGetPage(buffer);
tuple = (SpGistInnerTuple) PageGetItem(page,
PageGetItemId(page, xldata->offnumParent));
spgUpdateNodeLink(tuple, xldata->nodeI,
blknoDst, toInsert[nInsert - 1]);
PageSetLSN(page, lsn);
MarkBufferDirty(buffer);
}
if (BufferIsValid(buffer))
UnlockReleaseBuffer(buffer);
}
static void
spgRedoSplitTuple(XLogReaderState *record)
{
XLogRecPtr lsn = record->EndRecPtr;
char *ptr = XLogRecGetData(record);
spgxlogSplitTuple *xldata = (spgxlogSplitTuple *) ptr;
char *prefixTuple;
SpGistInnerTupleData prefixTupleHdr;
char *postfixTuple;
SpGistInnerTupleData postfixTupleHdr;
Buffer buffer;
Page page;
XLogRedoAction action;
ptr += sizeof(spgxlogSplitTuple);
prefixTuple = ptr;
/* the prefix tuple is unaligned, so make a copy to access its header */
memcpy(&prefixTupleHdr, prefixTuple, sizeof(SpGistInnerTupleData));
ptr += prefixTupleHdr.size;
postfixTuple = ptr;
/* postfix tuple is also unaligned */
memcpy(&postfixTupleHdr, postfixTuple, sizeof(SpGistInnerTupleData));
/*
* In normal operation we would have both pages locked simultaneously; but
* in WAL replay it should be safe to update them one at a time, as long
* as we do it in the right order.
*/
/* insert postfix tuple first to avoid dangling link */
if (!xldata->postfixBlkSame)
{
if (xldata->newPage)
{
buffer = XLogInitBufferForRedo(record, 1);
/* SplitTuple is not used for nulls pages */
SpGistInitBuffer(buffer, 0);
action = BLK_NEEDS_REDO;
}
else
action = XLogReadBufferForRedo(record, 1, &buffer);
if (action == BLK_NEEDS_REDO)
{
page = BufferGetPage(buffer);
addOrReplaceTuple(page, (Item) postfixTuple,
postfixTupleHdr.size, xldata->offnumPostfix);
PageSetLSN(page, lsn);
MarkBufferDirty(buffer);
}
if (BufferIsValid(buffer))
UnlockReleaseBuffer(buffer);
}
/* now handle the original page */
if (XLogReadBufferForRedo(record, 0, &buffer) == BLK_NEEDS_REDO)
{
page = BufferGetPage(buffer);
PageIndexTupleDelete(page, xldata->offnumPrefix);
if (PageAddItem(page, (Item) prefixTuple, prefixTupleHdr.size,
xldata->offnumPrefix, false, false) != xldata->offnumPrefix)
elog(ERROR, "failed to add item of size %u to SPGiST index page",
prefixTupleHdr.size);
if (xldata->postfixBlkSame)
addOrReplaceTuple(page, (Item) postfixTuple,
postfixTupleHdr.size,
xldata->offnumPostfix);
PageSetLSN(page, lsn);
MarkBufferDirty(buffer);
}
if (BufferIsValid(buffer))
UnlockReleaseBuffer(buffer);
}
void
btree_desc(StringInfo buf, XLogReaderState *record)
{
char *rec = XLogRecGetData(record);
uint8 info = XLogRecGetInfo(record) & ~XLR_INFO_MASK;
switch (info)
{
case XLOG_BTREE_INSERT_LEAF:
case XLOG_BTREE_INSERT_UPPER:
case XLOG_BTREE_INSERT_META:
{
xl_btree_insert *xlrec = (xl_btree_insert *) rec;
appendStringInfo(buf, "off %u", xlrec->offnum);
break;
}
case XLOG_BTREE_SPLIT_L:
case XLOG_BTREE_SPLIT_R:
case XLOG_BTREE_SPLIT_L_ROOT:
case XLOG_BTREE_SPLIT_R_ROOT:
{
xl_btree_split *xlrec = (xl_btree_split *) rec;
appendStringInfo(buf, "level %u, firstright %d",
xlrec->level, xlrec->firstright);
break;
}
case XLOG_BTREE_VACUUM:
{
xl_btree_vacuum *xlrec = (xl_btree_vacuum *) rec;
appendStringInfo(buf, "lastBlockVacuumed %u",
xlrec->lastBlockVacuumed);
break;
}
case XLOG_BTREE_DELETE:
{
xl_btree_delete *xlrec = (xl_btree_delete *) rec;
appendStringInfo(buf, "%d items", xlrec->nitems);
break;
}
case XLOG_BTREE_MARK_PAGE_HALFDEAD:
{
xl_btree_mark_page_halfdead *xlrec = (xl_btree_mark_page_halfdead *) rec;
appendStringInfo(buf, "topparent %u; leaf %u; left %u; right %u",
xlrec->topparent, xlrec->leafblk, xlrec->leftblk, xlrec->rightblk);
break;
}
case XLOG_BTREE_UNLINK_PAGE_META:
case XLOG_BTREE_UNLINK_PAGE:
{
xl_btree_unlink_page *xlrec = (xl_btree_unlink_page *) rec;
appendStringInfo(buf, "left %u; right %u; btpo_xact %u; ",
xlrec->leftsib, xlrec->rightsib,
xlrec->btpo_xact);
appendStringInfo(buf, "leafleft %u; leafright %u; topparent %u",
xlrec->leafleftsib, xlrec->leafrightsib,
xlrec->topparent);
break;
}
case XLOG_BTREE_NEWROOT:
{
xl_btree_newroot *xlrec = (xl_btree_newroot *) rec;
appendStringInfo(buf, "lev %u", xlrec->level);
break;
}
case XLOG_BTREE_REUSE_PAGE:
{
xl_btree_reuse_page *xlrec = (xl_btree_reuse_page *) rec;
appendStringInfo(buf, "rel %u/%u/%u; latestRemovedXid %u",
xlrec->node.spcNode, xlrec->node.dbNode,
xlrec->node.relNode, xlrec->latestRemovedXid);
break;
}
}
}
/*
* Decode XLOG_HEAP2_MULTI_INSERT_insert record into multiple tuplebufs.
*
* Currently MULTI_INSERT will always contain the full tuples.
*/
static void
DecodeMultiInsert(LogicalDecodingContext *ctx, XLogRecordBuffer *buf)
{
XLogReaderState *r = buf->record;
xl_heap_multi_insert *xlrec;
int i;
char *data;
char *tupledata;
Size tuplelen;
RelFileNode rnode;
xlrec = (xl_heap_multi_insert *) XLogRecGetData(r);
/* only interested in our database */
XLogRecGetBlockTag(r, 0, &rnode, NULL, NULL);
if (rnode.dbNode != ctx->slot->data.database)
return;
/* output plugin doesn't look for this origin, no need to queue */
if (FilterByOrigin(ctx, XLogRecGetOrigin(r)))
return;
tupledata = XLogRecGetBlockData(r, 0, &tuplelen);
data = tupledata;
for (i = 0; i < xlrec->ntuples; i++)
{
ReorderBufferChange *change;
xl_multi_insert_tuple *xlhdr;
int datalen;
ReorderBufferTupleBuf *tuple;
change = ReorderBufferGetChange(ctx->reorder);
change->action = REORDER_BUFFER_CHANGE_INSERT;
change->origin_id = XLogRecGetOrigin(r);
memcpy(&change->data.tp.relnode, &rnode, sizeof(RelFileNode));
/*
* CONTAINS_NEW_TUPLE will always be set currently as multi_insert
* isn't used for catalogs, but better be future proof.
*
* We decode the tuple in pretty much the same way as DecodeXLogTuple,
* but since the layout is slightly different, we can't use it here.
*/
if (xlrec->flags & XLH_INSERT_CONTAINS_NEW_TUPLE)
{
change->data.tp.newtuple = ReorderBufferGetTupleBuf(ctx->reorder);
tuple = change->data.tp.newtuple;
/* not a disk based tuple */
ItemPointerSetInvalid(&tuple->tuple.t_self);
xlhdr = (xl_multi_insert_tuple *) SHORTALIGN(data);
data = ((char *) xlhdr) + SizeOfMultiInsertTuple;
datalen = xlhdr->datalen;
/*
* We can only figure this out after reassembling the
* transactions.
*/
tuple->tuple.t_tableOid = InvalidOid;
tuple->tuple.t_data = &tuple->t_data.header;
tuple->tuple.t_len = datalen + SizeofHeapTupleHeader;
memset(&tuple->t_data.header, 0, SizeofHeapTupleHeader);
memcpy((char *) &tuple->t_data.header + SizeofHeapTupleHeader,
(char *) data,
datalen);
data += datalen;
tuple->t_data.header.t_infomask = xlhdr->t_infomask;
tuple->t_data.header.t_infomask2 = xlhdr->t_infomask2;
tuple->t_data.header.t_hoff = xlhdr->t_hoff;
}
/*
* Reset toast reassembly state only after the last row in the last
* xl_multi_insert_tuple record emitted by one heap_multi_insert()
* call.
*/
if (xlrec->flags & XLH_INSERT_LAST_IN_MULTI &&
(i + 1) == xlrec->ntuples)
change->data.tp.clear_toast_afterwards = true;
else
change->data.tp.clear_toast_afterwards = false;
ReorderBufferQueueChange(ctx->reorder, XLogRecGetXid(r),
buf->origptr, change);
}
Assert(data == tupledata + tuplelen);
}
void
spg_desc(StringInfo buf, XLogReaderState *record)
{
char *rec = XLogRecGetData(record);
uint8 info = XLogRecGetInfo(record) & ~XLR_INFO_MASK;
switch (info)
{
case XLOG_SPGIST_ADD_LEAF:
{
spgxlogAddLeaf *xlrec = (spgxlogAddLeaf *) rec;
appendStringInfoString(buf, "add leaf to page");
appendStringInfo(buf, "; off %u; headoff %u; parentoff %u",
xlrec->offnumLeaf, xlrec->offnumHeadLeaf,
xlrec->offnumParent);
if (xlrec->newPage)
appendStringInfoString(buf, " (newpage)");
if (xlrec->storesNulls)
appendStringInfoString(buf, " (nulls)");
}
break;
case XLOG_SPGIST_MOVE_LEAFS:
appendStringInfo(buf, "%u leafs",
((spgxlogMoveLeafs *) rec)->nMoves);
break;
case XLOG_SPGIST_ADD_NODE:
appendStringInfo(buf, "off %u",
((spgxlogAddNode *) rec)->offnum);
break;
case XLOG_SPGIST_SPLIT_TUPLE:
appendStringInfo(buf, "prefix off: %u, postfix off: %u (same %d, new %d)",
((spgxlogSplitTuple *) rec)->offnumPrefix,
((spgxlogSplitTuple *) rec)->offnumPostfix,
((spgxlogSplitTuple *) rec)->postfixBlkSame,
((spgxlogSplitTuple *) rec)->newPage
);
break;
case XLOG_SPGIST_PICKSPLIT:
{
spgxlogPickSplit *xlrec = (spgxlogPickSplit *) rec;
appendStringInfo(buf, "ndel %u; nins %u",
xlrec->nDelete, xlrec->nInsert);
if (xlrec->innerIsParent)
appendStringInfoString(buf, " (innerIsParent)");
if (xlrec->isRootSplit)
appendStringInfoString(buf, " (isRootSplit)");
}
break;
case XLOG_SPGIST_VACUUM_LEAF:
/* no further information */
break;
case XLOG_SPGIST_VACUUM_ROOT:
/* no further information */
break;
case XLOG_SPGIST_VACUUM_REDIRECT:
appendStringInfo(buf, "newest XID %u",
((spgxlogVacuumRedirect *) rec)->newestRedirectXid);
break;
}
}
void
heap_desc(StringInfo buf, XLogReaderState *record)
{
char *rec = XLogRecGetData(record);
uint8 info = XLogRecGetInfo(record) & ~XLR_INFO_MASK;
info &= XLOG_HEAP_OPMASK;
if (info == XLOG_HEAP_INSERT)
{
xl_heap_insert *xlrec = (xl_heap_insert *) rec;
appendStringInfo(buf, "off %u", xlrec->offnum);
}
else if (info == XLOG_HEAP_DELETE)
{
xl_heap_delete *xlrec = (xl_heap_delete *) rec;
appendStringInfo(buf, "off %u ", xlrec->offnum);
out_infobits(buf, xlrec->infobits_set);
}
else if (info == XLOG_HEAP_UPDATE)
{
xl_heap_update *xlrec = (xl_heap_update *) rec;
appendStringInfo(buf, "off %u xmax %u ",
xlrec->old_offnum,
xlrec->old_xmax);
out_infobits(buf, xlrec->old_infobits_set);
appendStringInfo(buf, "; new off %u xmax %u",
xlrec->new_offnum,
xlrec->new_xmax);
}
else if (info == XLOG_HEAP_HOT_UPDATE)
{
xl_heap_update *xlrec = (xl_heap_update *) rec;
appendStringInfo(buf, "off %u xmax %u ",
xlrec->old_offnum,
xlrec->old_xmax);
out_infobits(buf, xlrec->old_infobits_set);
appendStringInfo(buf, "; new off %u xmax %u",
xlrec->new_offnum,
xlrec->new_xmax);
}
else if (info == XLOG_HEAP_CONFIRM)
{
xl_heap_confirm *xlrec = (xl_heap_confirm *) rec;
appendStringInfo(buf, "off %u", xlrec->offnum);
}
else if (info == XLOG_HEAP_LOCK)
{
xl_heap_lock *xlrec = (xl_heap_lock *) rec;
appendStringInfo(buf, "off %u: xid %u: flags %u ",
xlrec->offnum, xlrec->locking_xid, xlrec->flags);
out_infobits(buf, xlrec->infobits_set);
}
else if (info == XLOG_HEAP_INPLACE)
{
xl_heap_inplace *xlrec = (xl_heap_inplace *) rec;
appendStringInfo(buf, "off %u", xlrec->offnum);
}
}
/*
* redo any page update (except page split)
*/
static void
gistRedoPageUpdateRecord(XLogRecPtr lsn, XLogRecord *record)
{
char *begin = XLogRecGetData(record);
gistxlogPageUpdate *xldata = (gistxlogPageUpdate *) begin;
Buffer buffer;
Page page;
char *data;
/*
* We need to acquire and hold lock on target page while updating the left
* child page. If we have a full-page image of target page, getting the
* lock is a side-effect of restoring that image. Note that even if the
* target page no longer exists, we'll still attempt to replay the change
* on the child page.
*/
if (record->xl_info & XLR_BKP_BLOCK(0))
buffer = RestoreBackupBlock(lsn, record, 0, false, true);
else
buffer = XLogReadBuffer(xldata->node, xldata->blkno, false);
/* Fix follow-right data on left child page */
if (BlockNumberIsValid(xldata->leftchild))
gistRedoClearFollowRight(lsn, record, 1,
xldata->node, xldata->leftchild);
/* Done if target page no longer exists */
if (!BufferIsValid(buffer))
return;
/* nothing more to do if page was backed up (and no info to do it with) */
if (record->xl_info & XLR_BKP_BLOCK(0))
{
UnlockReleaseBuffer(buffer);
return;
}
page = (Page) BufferGetPage(buffer);
/* nothing more to do if change already applied */
if (lsn <= PageGetLSN(page))
{
UnlockReleaseBuffer(buffer);
return;
}
data = begin + sizeof(gistxlogPageUpdate);
/* Delete old tuples */
if (xldata->ntodelete > 0)
{
int i;
OffsetNumber *todelete = (OffsetNumber *) data;
data += sizeof(OffsetNumber) * xldata->ntodelete;
for (i = 0; i < xldata->ntodelete; i++)
PageIndexTupleDelete(page, todelete[i]);
if (GistPageIsLeaf(page))
GistMarkTuplesDeleted(page);
}
/* add tuples */
if (data - begin < record->xl_len)
{
OffsetNumber off = (PageIsEmpty(page)) ? FirstOffsetNumber :
OffsetNumberNext(PageGetMaxOffsetNumber(page));
while (data - begin < record->xl_len)
{
IndexTuple itup = (IndexTuple) data;
Size sz = IndexTupleSize(itup);
OffsetNumber l;
data += sz;
l = PageAddItem(page, (Item) itup, sz, off, false, false);
if (l == InvalidOffsetNumber)
elog(ERROR, "failed to add item to GiST index page, size %d bytes",
(int) sz);
off++;
}
}
else
{
/*
* special case: leafpage, nothing to insert, nothing to delete, then
* vacuum marks page
*/
if (GistPageIsLeaf(page) && xldata->ntodelete == 0)
GistClearTuplesDeleted(page);
}
if (!GistPageIsLeaf(page) &&
PageGetMaxOffsetNumber(page) == InvalidOffsetNumber &&
xldata->blkno == GIST_ROOT_BLKNO)
{
/*
* all links on non-leaf root page was deleted by vacuum full, so root
* page becomes a leaf
*/
GistPageSetLeaf(page);
}
GistPageGetOpaque(page)->rightlink = InvalidBlockNumber;
PageSetLSN(page, lsn);
PageSetTLI(page, ThisTimeLineID);
MarkBufferDirty(buffer);
UnlockReleaseBuffer(buffer);
}
void
smgr_redo(XLogReaderState *record)
{
XLogRecPtr lsn = record->EndRecPtr;
uint8 info = XLogRecGetInfo(record) & ~XLR_INFO_MASK;
/* Backup blocks are not used in smgr records */
Assert(!XLogRecHasAnyBlockRefs(record));
if (info == XLOG_SMGR_CREATE)
{
xl_smgr_create *xlrec = (xl_smgr_create *) XLogRecGetData(record);
SMgrRelation reln;
reln = smgropen(xlrec->rnode, InvalidBackendId);
smgrcreate(reln, xlrec->forkNum, true);
}
else if (info == XLOG_SMGR_TRUNCATE)
{
xl_smgr_truncate *xlrec = (xl_smgr_truncate *) XLogRecGetData(record);
SMgrRelation reln;
Relation rel;
reln = smgropen(xlrec->rnode, InvalidBackendId);
/*
* Forcibly create relation if it doesn't exist (which suggests that
* it was dropped somewhere later in the WAL sequence). As in
* XLogReadBufferForRedo, we prefer to recreate the rel and replay the
* log as best we can until the drop is seen.
*/
smgrcreate(reln, MAIN_FORKNUM, true);
/*
* Before we perform the truncation, update minimum recovery point to
* cover this WAL record. Once the relation is truncated, there's no
* going back. The buffer manager enforces the WAL-first rule for
* normal updates to relation files, so that the minimum recovery
* point is always updated before the corresponding change in the data
* file is flushed to disk. We have to do the same manually here.
*
* Doing this before the truncation means that if the truncation fails
* for some reason, you cannot start up the system even after restart,
* until you fix the underlying situation so that the truncation will
* succeed. Alternatively, we could update the minimum recovery point
* after truncation, but that would leave a small window where the
* WAL-first rule could be violated.
*/
XLogFlush(lsn);
if ((xlrec->flags & SMGR_TRUNCATE_HEAP) != 0)
{
smgrtruncate(reln, MAIN_FORKNUM, xlrec->blkno);
/* Also tell xlogutils.c about it */
XLogTruncateRelation(xlrec->rnode, MAIN_FORKNUM, xlrec->blkno);
}
/* Truncate FSM and VM too */
rel = CreateFakeRelcacheEntry(xlrec->rnode);
if ((xlrec->flags & SMGR_TRUNCATE_FSM) != 0 &&
smgrexists(reln, FSM_FORKNUM))
FreeSpaceMapTruncateRel(rel, xlrec->blkno);
if ((xlrec->flags & SMGR_TRUNCATE_VM) != 0 &&
smgrexists(reln, VISIBILITYMAP_FORKNUM))
visibilitymap_truncate(rel, xlrec->blkno);
FreeFakeRelcacheEntry(rel);
}
else
elog(PANIC, "smgr_redo: unknown op code %u", info);
}
static void
gistRedoPageSplitRecord(XLogRecPtr lsn, XLogRecord *record)
{
gistxlogPageSplit *xldata = (gistxlogPageSplit *) XLogRecGetData(record);
PageSplitRecord xlrec;
Buffer firstbuffer = InvalidBuffer;
Buffer buffer;
Page page;
int i;
bool isrootsplit = false;
decodePageSplitRecord(&xlrec, record);
/*
* We must hold lock on the first-listed page throughout the action,
* including while updating the left child page (if any). We can unlock
* remaining pages in the list as soon as they've been written, because
* there is no path for concurrent queries to reach those pages without
* first visiting the first-listed page.
*/
/* loop around all pages */
for (i = 0; i < xlrec.data->npage; i++)
{
NewPage *newpage = xlrec.page + i;
int flags;
if (newpage->header->blkno == GIST_ROOT_BLKNO)
{
Assert(i == 0);
isrootsplit = true;
}
buffer = XLogReadBuffer(xlrec.data->node, newpage->header->blkno, true);
Assert(BufferIsValid(buffer));
page = (Page) BufferGetPage(buffer);
/* ok, clear buffer */
if (xlrec.data->origleaf && newpage->header->blkno != GIST_ROOT_BLKNO)
flags = F_LEAF;
else
flags = 0;
GISTInitBuffer(buffer, flags);
/* and fill it */
gistfillbuffer(page, newpage->itup, newpage->header->num, FirstOffsetNumber);
if (newpage->header->blkno == GIST_ROOT_BLKNO)
{
GistPageGetOpaque(page)->rightlink = InvalidBlockNumber;
GistPageSetNSN(page, xldata->orignsn);
GistClearFollowRight(page);
}
else
{
if (i < xlrec.data->npage - 1)
GistPageGetOpaque(page)->rightlink = xlrec.page[i + 1].header->blkno;
else
GistPageGetOpaque(page)->rightlink = xldata->origrlink;
GistPageSetNSN(page, xldata->orignsn);
if (i < xlrec.data->npage - 1 && !isrootsplit &&
xldata->markfollowright)
GistMarkFollowRight(page);
else
GistClearFollowRight(page);
}
PageSetLSN(page, lsn);
PageSetTLI(page, ThisTimeLineID);
MarkBufferDirty(buffer);
if (i == 0)
firstbuffer = buffer;
else
UnlockReleaseBuffer(buffer);
}
/* Fix follow-right data on left child page, if any */
if (BlockNumberIsValid(xldata->leftchild))
gistRedoClearFollowRight(lsn, record, 0,
xldata->node, xldata->leftchild);
/* Finally, release lock on the first page */
UnlockReleaseBuffer(firstbuffer);
}
void
heap_desc(StringInfo buf, XLogRecord *record)
{
char *rec = XLogRecGetData(record);
uint8 info = record->xl_info & ~XLR_INFO_MASK;
info &= XLOG_HEAP_OPMASK;
if (info == XLOG_HEAP_INSERT)
{
xl_heap_insert *xlrec = (xl_heap_insert *) rec;
out_target(buf, &(xlrec->target));
}
else if (info == XLOG_HEAP_DELETE)
{
xl_heap_delete *xlrec = (xl_heap_delete *) rec;
out_target(buf, &(xlrec->target));
appendStringInfoChar(buf, ' ');
out_infobits(buf, xlrec->infobits_set);
}
else if (info == XLOG_HEAP_UPDATE)
{
xl_heap_update *xlrec = (xl_heap_update *) rec;
out_target(buf, &(xlrec->target));
appendStringInfo(buf, " xmax %u ", xlrec->old_xmax);
out_infobits(buf, xlrec->old_infobits_set);
appendStringInfo(buf, "; new tid %u/%u xmax %u",
ItemPointerGetBlockNumber(&(xlrec->newtid)),
ItemPointerGetOffsetNumber(&(xlrec->newtid)),
xlrec->new_xmax);
}
else if (info == XLOG_HEAP_HOT_UPDATE)
{
xl_heap_update *xlrec = (xl_heap_update *) rec;
out_target(buf, &(xlrec->target));
appendStringInfo(buf, " xmax %u ", xlrec->old_xmax);
out_infobits(buf, xlrec->old_infobits_set);
appendStringInfo(buf, "; new tid %u/%u xmax %u",
ItemPointerGetBlockNumber(&(xlrec->newtid)),
ItemPointerGetOffsetNumber(&(xlrec->newtid)),
xlrec->new_xmax);
}
else if (info == XLOG_HEAP_LOCK)
{
xl_heap_lock *xlrec = (xl_heap_lock *) rec;
appendStringInfo(buf, "xid %u: ", xlrec->locking_xid);
out_target(buf, &(xlrec->target));
appendStringInfoChar(buf, ' ');
out_infobits(buf, xlrec->infobits_set);
}
else if (info == XLOG_HEAP_INPLACE)
{
xl_heap_inplace *xlrec = (xl_heap_inplace *) rec;
out_target(buf, &(xlrec->target));
}
}
void
heap2_desc(StringInfo buf, XLogRecord *record)
{
char *rec = XLogRecGetData(record);
uint8 info = record->xl_info & ~XLR_INFO_MASK;
info &= XLOG_HEAP_OPMASK;
if (info == XLOG_HEAP2_CLEAN)
{
xl_heap_clean *xlrec = (xl_heap_clean *) rec;
appendStringInfo(buf, "rel %u/%u/%u; blk %u remxid %u",
xlrec->node.spcNode, xlrec->node.dbNode,
xlrec->node.relNode, xlrec->block,
xlrec->latestRemovedXid);
}
else if (info == XLOG_HEAP2_FREEZE_PAGE)
{
xl_heap_freeze_page *xlrec = (xl_heap_freeze_page *) rec;
appendStringInfo(buf, "rel %u/%u/%u; blk %u; cutoff xid %u ntuples %u",
xlrec->node.spcNode, xlrec->node.dbNode,
xlrec->node.relNode, xlrec->block,
xlrec->cutoff_xid, xlrec->ntuples);
}
else if (info == XLOG_HEAP2_CLEANUP_INFO)
{
xl_heap_cleanup_info *xlrec = (xl_heap_cleanup_info *) rec;
appendStringInfo(buf, "remxid %u", xlrec->latestRemovedXid);
}
else if (info == XLOG_HEAP2_VISIBLE)
{
xl_heap_visible *xlrec = (xl_heap_visible *) rec;
appendStringInfo(buf, "rel %u/%u/%u; blk %u",
xlrec->node.spcNode, xlrec->node.dbNode,
xlrec->node.relNode, xlrec->block);
}
else if (info == XLOG_HEAP2_MULTI_INSERT)
{
xl_heap_multi_insert *xlrec = (xl_heap_multi_insert *) rec;
appendStringInfo(buf, "rel %u/%u/%u; blk %u; %d tuples",
xlrec->node.spcNode, xlrec->node.dbNode, xlrec->node.relNode,
xlrec->blkno, xlrec->ntuples);
}
else if (info == XLOG_HEAP2_LOCK_UPDATED)
{
xl_heap_lock_updated *xlrec = (xl_heap_lock_updated *) rec;
appendStringInfo(buf, "xmax %u msk %04x; ", xlrec->xmax,
xlrec->infobits_set);
out_target(buf, &(xlrec->target));
}
else if (info == XLOG_HEAP2_NEW_CID)
{
xl_heap_new_cid *xlrec = (xl_heap_new_cid *) rec;
out_target(buf, &(xlrec->target));
appendStringInfo(buf, "; cmin: %u, cmax: %u, combo: %u",
xlrec->cmin, xlrec->cmax, xlrec->combocid);
}
}
/*
* redo any page update (except page split)
*/
static void
gistRedoPageUpdateRecord(XLogReaderState *record)
{
XLogRecPtr lsn = record->EndRecPtr;
gistxlogPageUpdate *xldata = (gistxlogPageUpdate *) XLogRecGetData(record);
Buffer buffer;
Page page;
if (XLogReadBufferForRedo(record, 0, &buffer) == BLK_NEEDS_REDO)
{
char *begin;
char *data;
Size datalen;
int ninserted = 0;
data = begin = XLogRecGetBlockData(record, 0, &datalen);
page = (Page) BufferGetPage(buffer);
/* Delete old tuples */
if (xldata->ntodelete > 0)
{
OffsetNumber *todelete = (OffsetNumber *) data;
data += sizeof(OffsetNumber) * xldata->ntodelete;
PageIndexMultiDelete(page, todelete, xldata->ntodelete);
if (GistPageIsLeaf(page))
GistMarkTuplesDeleted(page);
}
/* add tuples */
if (data - begin < datalen)
{
OffsetNumber off = (PageIsEmpty(page)) ? FirstOffsetNumber :
OffsetNumberNext(PageGetMaxOffsetNumber(page));
while (data - begin < datalen)
{
IndexTuple itup = (IndexTuple) data;
Size sz = IndexTupleSize(itup);
OffsetNumber l;
data += sz;
l = PageAddItem(page, (Item) itup, sz, off, false, false);
if (l == InvalidOffsetNumber)
elog(ERROR, "failed to add item to GiST index page, size %d bytes",
(int) sz);
off++;
ninserted++;
}
}
Assert(ninserted == xldata->ntoinsert);
PageSetLSN(page, lsn);
MarkBufferDirty(buffer);
}
/*
* Fix follow-right data on left child page
*
* This must be done while still holding the lock on the target page. Note
* that even if the target page no longer exists, we still attempt to
* replay the change on the child page.
*/
if (XLogRecHasBlockRef(record, 1))
gistRedoClearFollowRight(record, 1);
if (BufferIsValid(buffer))
UnlockReleaseBuffer(buffer);
}
void
heap2_desc(StringInfo buf, XLogReaderState *record)
{
char *rec = XLogRecGetData(record);
uint8 info = XLogRecGetInfo(record) & ~XLR_INFO_MASK;
info &= XLOG_HEAP_OPMASK;
if (info == XLOG_HEAP2_CLEAN)
{
xl_heap_clean *xlrec = (xl_heap_clean *) rec;
appendStringInfo(buf, "remxid %u", xlrec->latestRemovedXid);
}
else if (info == XLOG_HEAP2_FREEZE_PAGE)
{
xl_heap_freeze_page *xlrec = (xl_heap_freeze_page *) rec;
appendStringInfo(buf, "cutoff xid %u ntuples %u",
xlrec->cutoff_xid, xlrec->ntuples);
}
else if (info == XLOG_HEAP2_CLEANUP_INFO)
{
xl_heap_cleanup_info *xlrec = (xl_heap_cleanup_info *) rec;
appendStringInfo(buf, "remxid %u", xlrec->latestRemovedXid);
}
else if (info == XLOG_HEAP2_VISIBLE)
{
xl_heap_visible *xlrec = (xl_heap_visible *) rec;
appendStringInfo(buf, "cutoff xid %u flags %d",
xlrec->cutoff_xid, xlrec->flags);
}
else if (info == XLOG_HEAP2_MULTI_INSERT)
{
xl_heap_multi_insert *xlrec = (xl_heap_multi_insert *) rec;
appendStringInfo(buf, "%d tuples", xlrec->ntuples);
}
else if (info == XLOG_HEAP2_LOCK_UPDATED)
{
xl_heap_lock_updated *xlrec = (xl_heap_lock_updated *) rec;
appendStringInfo(buf, "off %u: xmax %u: flags %u ",
xlrec->offnum, xlrec->xmax, xlrec->flags);
out_infobits(buf, xlrec->infobits_set);
}
else if (info == XLOG_HEAP2_NEW_CID)
{
xl_heap_new_cid *xlrec = (xl_heap_new_cid *) rec;
appendStringInfo(buf, "rel %u/%u/%u; tid %u/%u",
xlrec->target_node.spcNode,
xlrec->target_node.dbNode,
xlrec->target_node.relNode,
ItemPointerGetBlockNumber(&(xlrec->target_tid)),
ItemPointerGetOffsetNumber(&(xlrec->target_tid)));
appendStringInfo(buf, "; cmin: %u, cmax: %u, combo: %u",
xlrec->cmin, xlrec->cmax, xlrec->combocid);
}
}
/*
* TABLESPACE resource manager's routines
*/
void
tblspc_redo(XLogRecPtr beginLoc, XLogRecPtr lsn, XLogRecord *record)
{
uint8 info = record->xl_info & ~XLR_INFO_MASK;
if (info == XLOG_TBLSPC_CREATE)
{
xl_tblspc_create_rec *xlrec = (xl_tblspc_create_rec *) XLogRecGetData(record);
char *location = xlrec->ts_path;
char *linkloc;
char *sublocation;
struct stat st;
/*
* MPP-2333: Try to create the target directory if it does not exist.
*/
if (stat(location, &st) < 0)
{
if (mkdir(location, 0700) != 0)
{
ereport(ERROR,
(errcode_for_file_access(),
errmsg("could not create location directory \"%s\": %m",
location)));
}
}
/*
* Attempt to coerce target directory to safe permissions. If this
* fails, it doesn't exist or has the wrong owner.
*/
if (chmod(location, 0700) != 0)
ereport(ERROR,
(errcode_for_file_access(),
errmsg("could not set permissions on directory \"%s\": %m",
location)));
/* Create segment subdirectory. */
sublocation = palloc(MAXPGPATH);
sprintf(sublocation,"%s/seg%d",location,Gp_segment);
if (mkdir(sublocation, 0700) != 0)
ereport(ERROR,
(errcode_for_file_access(),
errmsg("could not create subdirectory \"%s\": %m",
sublocation)));
/* Create or re-create the PG_VERSION file in the target directory */
set_short_version(sublocation, NULL, false);
/* Create the symlink if not already present */
linkloc = (char *) palloc(10 + 10 + 1);
sprintf(linkloc, "pg_tblspc/%u", xlrec->ts_id);
if (symlink(sublocation, linkloc) < 0)
{
if (errno != EEXIST)
ereport(ERROR,
(errcode_for_file_access(),
errmsg("could not create symbolic link \"%s\": %m",
linkloc)));
}
pfree(sublocation);
pfree(linkloc);
}
else if (info == XLOG_TBLSPC_DROP)
{
xl_tblspc_drop_rec *xlrec = (xl_tblspc_drop_rec *) XLogRecGetData(record);
if (!remove_tablespace_directories(xlrec->ts_id, true, xlrec->ts_path))
ereport(ERROR,
(errcode(ERRCODE_OBJECT_NOT_IN_PREREQUISITE_STATE),
errmsg("tablespace %u is not empty",
xlrec->ts_id)));
}
else
elog(PANIC, "tblspc_redo: unknown op code %u", info);
}
/*
* CommitTS resource manager's routines
*/
void
commit_ts_redo(XLogReaderState *record)
{
uint8 info = XLogRecGetInfo(record) & ~XLR_INFO_MASK;
/* Backup blocks are not used in commit_ts records */
Assert(!XLogRecHasAnyBlockRefs(record));
if (info == COMMIT_TS_ZEROPAGE)
{
int pageno;
int slotno;
memcpy(&pageno, XLogRecGetData(record), sizeof(int));
LWLockAcquire(CommitTsControlLock, LW_EXCLUSIVE);
slotno = ZeroCommitTsPage(pageno, false);
SimpleLruWritePage(CommitTsCtl, slotno);
Assert(!CommitTsCtl->shared->page_dirty[slotno]);
LWLockRelease(CommitTsControlLock);
}
else if (info == COMMIT_TS_TRUNCATE)
{
int pageno;
memcpy(&pageno, XLogRecGetData(record), sizeof(int));
/*
* During XLOG replay, latest_page_number isn't set up yet; insert a
* suitable value to bypass the sanity test in SimpleLruTruncate.
*/
CommitTsCtl->shared->latest_page_number = pageno;
SimpleLruTruncate(CommitTsCtl, pageno);
}
else if (info == COMMIT_TS_SETTS)
{
xl_commit_ts_set *setts = (xl_commit_ts_set *) XLogRecGetData(record);
int nsubxids;
TransactionId *subxids;
nsubxids = ((XLogRecGetDataLen(record) - SizeOfCommitTsSet) /
sizeof(TransactionId));
if (nsubxids > 0)
{
subxids = palloc(sizeof(TransactionId) * nsubxids);
memcpy(subxids,
XLogRecGetData(record) + SizeOfCommitTsSet,
sizeof(TransactionId) * nsubxids);
}
else
subxids = NULL;
TransactionTreeSetCommitTsData(setts->mainxid, nsubxids, subxids,
setts->timestamp, setts->nodeid, true);
if (subxids)
pfree(subxids);
}
else
elog(PANIC, "commit_ts_redo: unknown op code %u", info);
}
/*
* Handle rmgr XACT_ID records for DecodeRecordIntoReorderBuffer().
*/
static void
DecodeXactOp(LogicalDecodingContext *ctx, XLogRecordBuffer *buf)
{
SnapBuild *builder = ctx->snapshot_builder;
ReorderBuffer *reorder = ctx->reorder;
XLogReaderState *r = buf->record;
uint8 info = XLogRecGetInfo(r) & XLOG_XACT_OPMASK;
/* no point in doing anything yet, data could not be decoded anyway */
if (SnapBuildCurrentState(builder) < SNAPBUILD_FULL_SNAPSHOT)
return;
switch (info)
{
case XLOG_XACT_COMMIT:
case XLOG_XACT_COMMIT_PREPARED:
{
xl_xact_commit *xlrec;
xl_xact_parsed_commit parsed;
TransactionId xid;
xlrec = (xl_xact_commit *) XLogRecGetData(r);
ParseCommitRecord(XLogRecGetInfo(buf->record), xlrec, &parsed);
if (!TransactionIdIsValid(parsed.twophase_xid))
xid = XLogRecGetXid(r);
else
xid = parsed.twophase_xid;
DecodeCommit(ctx, buf, &parsed, xid);
break;
}
case XLOG_XACT_ABORT:
case XLOG_XACT_ABORT_PREPARED:
{
xl_xact_abort *xlrec;
xl_xact_parsed_abort parsed;
TransactionId xid;
xlrec = (xl_xact_abort *) XLogRecGetData(r);
ParseAbortRecord(XLogRecGetInfo(buf->record), xlrec, &parsed);
if (!TransactionIdIsValid(parsed.twophase_xid))
xid = XLogRecGetXid(r);
else
xid = parsed.twophase_xid;
DecodeAbort(ctx, buf, &parsed, xid);
break;
}
case XLOG_XACT_ASSIGNMENT:
{
xl_xact_assignment *xlrec;
int i;
TransactionId *sub_xid;
xlrec = (xl_xact_assignment *) XLogRecGetData(r);
sub_xid = &xlrec->xsub[0];
for (i = 0; i < xlrec->nsubxacts; i++)
{
ReorderBufferAssignChild(reorder, xlrec->xtop,
*(sub_xid++), buf->origptr);
}
break;
}
case XLOG_XACT_PREPARE:
/*
* Currently decoding ignores PREPARE TRANSACTION and will just
* decode the transaction when the COMMIT PREPARED is sent or
* throw away the transaction's contents when a ROLLBACK PREPARED
* is received. In the future we could add code to expose prepared
* transactions in the changestream allowing for a kind of
* distributed 2PC.
*/
break;
default:
elog(ERROR, "unexpected RM_XACT_ID record type: %u", info);
}
}
/*
* Write an empty XLOG file, containing only the checkpoint record
* already set up in ControlFile.
*/
static void
WriteEmptyXLOG(void)
{
char *buffer;
XLogPageHeader page;
XLogLongPageHeader longpage;
XLogRecord *record;
pg_crc32 crc;
char path[MAXPGPATH];
int fd;
int nbytes;
/* Use malloc() to ensure buffer is MAXALIGNED */
buffer = (char *) malloc(XLOG_BLCKSZ);
page = (XLogPageHeader) buffer;
memset(buffer, 0, XLOG_BLCKSZ);
/* Set up the XLOG page header */
page->xlp_magic = XLOG_PAGE_MAGIC;
page->xlp_info = XLP_LONG_HEADER;
page->xlp_tli = ControlFile.checkPointCopy.ThisTimeLineID;
page->xlp_pageaddr = ControlFile.checkPointCopy.redo - SizeOfXLogLongPHD;
longpage = (XLogLongPageHeader) page;
longpage->xlp_sysid = ControlFile.system_identifier;
longpage->xlp_seg_size = XLogSegSize;
longpage->xlp_xlog_blcksz = XLOG_BLCKSZ;
/* Insert the initial checkpoint record */
record = (XLogRecord *) ((char *) page + SizeOfXLogLongPHD);
record->xl_prev = 0;
record->xl_xid = InvalidTransactionId;
record->xl_tot_len = SizeOfXLogRecord + sizeof(CheckPoint);
record->xl_len = sizeof(CheckPoint);
record->xl_info = XLOG_CHECKPOINT_SHUTDOWN;
record->xl_rmid = RM_XLOG_ID;
memcpy(XLogRecGetData(record), &ControlFile.checkPointCopy,
sizeof(CheckPoint));
INIT_CRC32(crc);
COMP_CRC32(crc, &ControlFile.checkPointCopy, sizeof(CheckPoint));
COMP_CRC32(crc, (char *) record, offsetof(XLogRecord, xl_crc));
FIN_CRC32(crc);
record->xl_crc = crc;
/* Write the first page */
XLogFilePath(path, ControlFile.checkPointCopy.ThisTimeLineID, newXlogSegNo);
unlink(path);
fd = open(path, O_RDWR | O_CREAT | O_EXCL | PG_BINARY,
S_IRUSR | S_IWUSR);
if (fd < 0)
{
fprintf(stderr, _("%s: could not open file \"%s\": %s\n"),
progname, path, strerror(errno));
exit(1);
}
errno = 0;
if (write(fd, buffer, XLOG_BLCKSZ) != XLOG_BLCKSZ)
{
/* if write didn't set errno, assume problem is no disk space */
if (errno == 0)
errno = ENOSPC;
fprintf(stderr, _("%s: could not write file \"%s\": %s\n"),
progname, path, strerror(errno));
exit(1);
}
/* Fill the rest of the file with zeroes */
memset(buffer, 0, XLOG_BLCKSZ);
for (nbytes = XLOG_BLCKSZ; nbytes < XLogSegSize; nbytes += XLOG_BLCKSZ)
{
errno = 0;
if (write(fd, buffer, XLOG_BLCKSZ) != XLOG_BLCKSZ)
{
if (errno == 0)
errno = ENOSPC;
fprintf(stderr, _("%s: could not write file \"%s\": %s\n"),
progname, path, strerror(errno));
exit(1);
}
}
if (fsync(fd) != 0)
{
fprintf(stderr, _("%s: fsync error: %s\n"), progname, strerror(errno));
exit(1);
}
close(fd);
}
static void
spgRedoAddLeaf(XLogReaderState *record)
{
XLogRecPtr lsn = record->EndRecPtr;
char *ptr = XLogRecGetData(record);
spgxlogAddLeaf *xldata = (spgxlogAddLeaf *) ptr;
char *leafTuple;
SpGistLeafTupleData leafTupleHdr;
Buffer buffer;
Page page;
XLogRedoAction action;
ptr += sizeof(spgxlogAddLeaf);
leafTuple = ptr;
/* the leaf tuple is unaligned, so make a copy to access its header */
memcpy(&leafTupleHdr, leafTuple, sizeof(SpGistLeafTupleData));
/*
* In normal operation we would have both current and parent pages locked
* simultaneously; but in WAL replay it should be safe to update the leaf
* page before updating the parent.
*/
if (xldata->newPage)
{
buffer = XLogInitBufferForRedo(record, 0);
SpGistInitBuffer(buffer,
SPGIST_LEAF | (xldata->storesNulls ? SPGIST_NULLS : 0));
action = BLK_NEEDS_REDO;
}
else
action = XLogReadBufferForRedo(record, 0, &buffer);
if (action == BLK_NEEDS_REDO)
{
page = BufferGetPage(buffer);
/* insert new tuple */
if (xldata->offnumLeaf != xldata->offnumHeadLeaf)
{
/* normal cases, tuple was added by SpGistPageAddNewItem */
addOrReplaceTuple(page, (Item) leafTuple, leafTupleHdr.size,
xldata->offnumLeaf);
/* update head tuple's chain link if needed */
if (xldata->offnumHeadLeaf != InvalidOffsetNumber)
{
SpGistLeafTuple head;
head = (SpGistLeafTuple) PageGetItem(page,
PageGetItemId(page, xldata->offnumHeadLeaf));
Assert(head->nextOffset == leafTupleHdr.nextOffset);
head->nextOffset = xldata->offnumLeaf;
}
}
else
{
/* replacing a DEAD tuple */
PageIndexTupleDelete(page, xldata->offnumLeaf);
if (PageAddItem(page,
(Item) leafTuple, leafTupleHdr.size,
xldata->offnumLeaf, false, false) != xldata->offnumLeaf)
elog(ERROR, "failed to add item of size %u to SPGiST index page",
leafTupleHdr.size);
}
PageSetLSN(page, lsn);
MarkBufferDirty(buffer);
}
if (BufferIsValid(buffer))
UnlockReleaseBuffer(buffer);
/* update parent downlink if necessary */
if (xldata->offnumParent != InvalidOffsetNumber)
{
if (XLogReadBufferForRedo(record, 1, &buffer) == BLK_NEEDS_REDO)
{
SpGistInnerTuple tuple;
BlockNumber blknoLeaf;
XLogRecGetBlockTag(record, 0, NULL, NULL, &blknoLeaf);
page = BufferGetPage(buffer);
tuple = (SpGistInnerTuple) PageGetItem(page,
PageGetItemId(page, xldata->offnumParent));
spgUpdateNodeLink(tuple, xldata->nodeI,
blknoLeaf, xldata->offnumLeaf);
PageSetLSN(page, lsn);
MarkBufferDirty(buffer);
}
if (BufferIsValid(buffer))
UnlockReleaseBuffer(buffer);
}
}
/*
* CRC-check an XLOG record. We do not believe the contents of an XLOG
* record (other than to the minimal extent of computing the amount of
* data to read in) until we've checked the CRCs.
*
* We assume all of the record has been read into memory at *record.
*/
static bool
RecordIsValid(XLogRecord *record, XLogRecPtr recptr)
{
pg_crc32 crc;
int i;
uint32 len = record->xl_len;
BkpBlock bkpb;
char *blk;
/* First the rmgr data */
INIT_CRC32C(crc);
COMP_CRC32C(crc, XLogRecGetData(record), len);
/* Add in the backup blocks, if any */
blk = (char *) XLogRecGetData(record) + len;
for (i = 0; i < XLR_MAX_BKP_BLOCKS; i++)
{
uint32 blen;
if (!(record->xl_info & XLR_SET_BKP_BLOCK(i)))
continue;
memcpy(&bkpb, blk, sizeof(BkpBlock));
if (bkpb.hole_offset + bkpb.hole_length > BLCKSZ)
{
printf("incorrect hole size in record at %X/%X\n",
recptr.xlogid, recptr.xrecoff);
return false;
}
blen = sizeof(BkpBlock) + BLCKSZ - bkpb.hole_length;
COMP_CRC32C(crc, blk, blen);
blk += blen;
}
/* skip total xl_tot_len check if physical log has been removed. */
#if PG_VERSION_NUM < 80300 || PG_VERSION_NUM >= 90200
if (record->xl_info & XLR_BKP_BLOCK_MASK)
#else
if (!(record->xl_info & XLR_BKP_REMOVABLE) ||
record->xl_info & XLR_BKP_BLOCK_MASK)
#endif
{
/* Check that xl_tot_len agrees with our calculation */
if (blk != (char *) record + record->xl_tot_len)
{
printf("incorrect total length in record at %X/%X\n",
recptr.xlogid, recptr.xrecoff);
return false;
}
}
/* Finally include the record header */
COMP_CRC32C(crc, (char *) record + sizeof(pg_crc32),
SizeOfXLogRecord - sizeof(pg_crc32));
FIN_CRC32C(crc);
if (!EQ_CRC32C(record->xl_crc, crc))
{
printf("incorrect resource manager data checksum in record at %X/%X\n",
recptr.xlogid, recptr.xrecoff);
return false;
}
return true;
}
static void
spgRedoAddNode(XLogReaderState *record)
{
XLogRecPtr lsn = record->EndRecPtr;
char *ptr = XLogRecGetData(record);
spgxlogAddNode *xldata = (spgxlogAddNode *) ptr;
char *innerTuple;
SpGistInnerTupleData innerTupleHdr;
SpGistState state;
Buffer buffer;
Page page;
XLogRedoAction action;
ptr += sizeof(spgxlogAddNode);
innerTuple = ptr;
/* the tuple is unaligned, so make a copy to access its header */
memcpy(&innerTupleHdr, innerTuple, sizeof(SpGistInnerTupleData));
fillFakeState(&state, xldata->stateSrc);
if (!XLogRecHasBlockRef(record, 1))
{
/* update in place */
Assert(xldata->parentBlk == -1);
if (XLogReadBufferForRedo(record, 0, &buffer) == BLK_NEEDS_REDO)
{
page = BufferGetPage(buffer);
PageIndexTupleDelete(page, xldata->offnum);
if (PageAddItem(page, (Item) innerTuple, innerTupleHdr.size,
xldata->offnum,
false, false) != xldata->offnum)
elog(ERROR, "failed to add item of size %u to SPGiST index page",
innerTupleHdr.size);
PageSetLSN(page, lsn);
MarkBufferDirty(buffer);
}
if (BufferIsValid(buffer))
UnlockReleaseBuffer(buffer);
}
else
{
BlockNumber blkno;
BlockNumber blknoNew;
XLogRecGetBlockTag(record, 0, NULL, NULL, &blkno);
XLogRecGetBlockTag(record, 1, NULL, NULL, &blknoNew);
/*
* In normal operation we would have all three pages (source, dest,
* and parent) locked simultaneously; but in WAL replay it should be
* safe to update them one at a time, as long as we do it in the right
* order. We must insert the new tuple before replacing the old tuple
* with the redirect tuple.
*/
/* Install new tuple first so redirect is valid */
if (xldata->newPage)
{
/* AddNode is not used for nulls pages */
buffer = XLogInitBufferForRedo(record, 1);
SpGistInitBuffer(buffer, 0);
action = BLK_NEEDS_REDO;
}
else
action = XLogReadBufferForRedo(record, 1, &buffer);
if (action == BLK_NEEDS_REDO)
{
page = BufferGetPage(buffer);
addOrReplaceTuple(page, (Item) innerTuple,
innerTupleHdr.size, xldata->offnumNew);
/*
* If parent is in this same page, update it now.
*/
if (xldata->parentBlk == 1)
{
SpGistInnerTuple parentTuple;
parentTuple = (SpGistInnerTuple) PageGetItem(page,
PageGetItemId(page, xldata->offnumParent));
spgUpdateNodeLink(parentTuple, xldata->nodeI,
blknoNew, xldata->offnumNew);
}
PageSetLSN(page, lsn);
MarkBufferDirty(buffer);
}
if (BufferIsValid(buffer))
UnlockReleaseBuffer(buffer);
/* Delete old tuple, replacing it with redirect or placeholder tuple */
if (XLogReadBufferForRedo(record, 0, &buffer) == BLK_NEEDS_REDO)
{
SpGistDeadTuple dt;
page = BufferGetPage(buffer);
if (state.isBuild)
dt = spgFormDeadTuple(&state, SPGIST_PLACEHOLDER,
InvalidBlockNumber,
InvalidOffsetNumber);
else
dt = spgFormDeadTuple(&state, SPGIST_REDIRECT,
blknoNew,
xldata->offnumNew);
PageIndexTupleDelete(page, xldata->offnum);
if (PageAddItem(page, (Item) dt, dt->size,
xldata->offnum,
false, false) != xldata->offnum)
elog(ERROR, "failed to add item of size %u to SPGiST index page",
dt->size);
if (state.isBuild)
SpGistPageGetOpaque(page)->nPlaceholder++;
else
SpGistPageGetOpaque(page)->nRedirection++;
/*
* If parent is in this same page, update it now.
*/
if (xldata->parentBlk == 0)
{
SpGistInnerTuple parentTuple;
parentTuple = (SpGistInnerTuple) PageGetItem(page,
PageGetItemId(page, xldata->offnumParent));
spgUpdateNodeLink(parentTuple, xldata->nodeI,
blknoNew, xldata->offnumNew);
}
PageSetLSN(page, lsn);
MarkBufferDirty(buffer);
}
if (BufferIsValid(buffer))
UnlockReleaseBuffer(buffer);
/*
* Update parent downlink (if we didn't do it as part of the source or
* destination page update already).
*/
if (xldata->parentBlk == 2)
{
if (XLogReadBufferForRedo(record, 2, &buffer) == BLK_NEEDS_REDO)
{
SpGistInnerTuple parentTuple;
page = BufferGetPage(buffer);
parentTuple = (SpGistInnerTuple) PageGetItem(page,
PageGetItemId(page, xldata->offnumParent));
spgUpdateNodeLink(parentTuple, xldata->nodeI,
blknoNew, xldata->offnumNew);
PageSetLSN(page, lsn);
MarkBufferDirty(buffer);
}
if (BufferIsValid(buffer))
UnlockReleaseBuffer(buffer);
}
}
}
static void
ginRedoUpdateMetapage(XLogReaderState *record)
{
XLogRecPtr lsn = record->EndRecPtr;
ginxlogUpdateMeta *data = (ginxlogUpdateMeta *) XLogRecGetData(record);
Buffer metabuffer;
Page metapage;
Buffer buffer;
/*
* Restore the metapage. This is essentially the same as a full-page
* image, so restore the metapage unconditionally without looking at the
* LSN, to avoid torn page hazards.
*/
metabuffer = XLogInitBufferForRedo(record, 0);
Assert(BufferGetBlockNumber(metabuffer) == GIN_METAPAGE_BLKNO);
metapage = BufferGetPage(metabuffer);
GinInitPage(metapage, GIN_META, BufferGetPageSize(metabuffer));
memcpy(GinPageGetMeta(metapage), &data->metadata, sizeof(GinMetaPageData));
PageSetLSN(metapage, lsn);
MarkBufferDirty(metabuffer);
if (data->ntuples > 0)
{
/*
* insert into tail page
*/
if (XLogReadBufferForRedo(record, 1, &buffer) == BLK_NEEDS_REDO)
{
Page page = BufferGetPage(buffer);
OffsetNumber off;
int i;
Size tupsize;
char *payload;
IndexTuple tuples;
Size totaltupsize;
payload = XLogRecGetBlockData(record, 1, &totaltupsize);
tuples = (IndexTuple) payload;
if (PageIsEmpty(page))
off = FirstOffsetNumber;
else
off = OffsetNumberNext(PageGetMaxOffsetNumber(page));
for (i = 0; i < data->ntuples; i++)
{
tupsize = IndexTupleSize(tuples);
if (PageAddItem(page, (Item) tuples, tupsize, off,
false, false) == InvalidOffsetNumber)
elog(ERROR, "failed to add item to index page");
tuples = (IndexTuple) (((char *) tuples) + tupsize);
off++;
}
Assert(payload + totaltupsize == (char *) tuples);
/*
* Increase counter of heap tuples
*/
GinPageGetOpaque(page)->maxoff++;
PageSetLSN(page, lsn);
MarkBufferDirty(buffer);
}
if (BufferIsValid(buffer))
UnlockReleaseBuffer(buffer);
}
else if (data->prevTail != InvalidBlockNumber)
{
/*
* New tail
*/
if (XLogReadBufferForRedo(record, 1, &buffer) == BLK_NEEDS_REDO)
{
Page page = BufferGetPage(buffer);
GinPageGetOpaque(page)->rightlink = data->newRightlink;
PageSetLSN(page, lsn);
MarkBufferDirty(buffer);
}
if (BufferIsValid(buffer))
UnlockReleaseBuffer(buffer);
}
UnlockReleaseBuffer(metabuffer);
}
static void
spgRedoPickSplit(XLogReaderState *record)
{
XLogRecPtr lsn = record->EndRecPtr;
char *ptr = XLogRecGetData(record);
spgxlogPickSplit *xldata = (spgxlogPickSplit *) ptr;
char *innerTuple;
SpGistInnerTupleData innerTupleHdr;
SpGistState state;
OffsetNumber *toDelete;
OffsetNumber *toInsert;
uint8 *leafPageSelect;
Buffer srcBuffer;
Buffer destBuffer;
Buffer innerBuffer;
Page srcPage;
Page destPage;
Page page;
int i;
BlockNumber blknoInner;
XLogRedoAction action;
XLogRecGetBlockTag(record, 2, NULL, NULL, &blknoInner);
fillFakeState(&state, xldata->stateSrc);
ptr += SizeOfSpgxlogPickSplit;
toDelete = (OffsetNumber *) ptr;
ptr += sizeof(OffsetNumber) * xldata->nDelete;
toInsert = (OffsetNumber *) ptr;
ptr += sizeof(OffsetNumber) * xldata->nInsert;
leafPageSelect = (uint8 *) ptr;
ptr += sizeof(uint8) * xldata->nInsert;
innerTuple = ptr;
/* the inner tuple is unaligned, so make a copy to access its header */
memcpy(&innerTupleHdr, innerTuple, sizeof(SpGistInnerTupleData));
ptr += innerTupleHdr.size;
/* now ptr points to the list of leaf tuples */
if (xldata->isRootSplit)
{
/* when splitting root, we touch it only in the guise of new inner */
srcBuffer = InvalidBuffer;
srcPage = NULL;
}
else if (xldata->initSrc)
{
/* just re-init the source page */
srcBuffer = XLogInitBufferForRedo(record, 0);
srcPage = (Page) BufferGetPage(srcBuffer);
SpGistInitBuffer(srcBuffer,
SPGIST_LEAF | (xldata->storesNulls ? SPGIST_NULLS : 0));
/* don't update LSN etc till we're done with it */
}
else
{
/*
* Delete the specified tuples from source page. (In case we're in
* Hot Standby, we need to hold lock on the page till we're done
* inserting leaf tuples and the new inner tuple, else the added
* redirect tuple will be a dangling link.)
*/
srcPage = NULL;
if (XLogReadBufferForRedo(record, 0, &srcBuffer) == BLK_NEEDS_REDO)
{
srcPage = BufferGetPage(srcBuffer);
/*
* We have it a bit easier here than in doPickSplit(), because we
* know the inner tuple's location already, so we can inject the
* correct redirection tuple now.
*/
if (!state.isBuild)
spgPageIndexMultiDelete(&state, srcPage,
toDelete, xldata->nDelete,
SPGIST_REDIRECT,
SPGIST_PLACEHOLDER,
blknoInner,
xldata->offnumInner);
else
spgPageIndexMultiDelete(&state, srcPage,
toDelete, xldata->nDelete,
SPGIST_PLACEHOLDER,
SPGIST_PLACEHOLDER,
InvalidBlockNumber,
InvalidOffsetNumber);
/* don't update LSN etc till we're done with it */
}
}
/* try to access dest page if any */
if (!XLogRecHasBlockRef(record, 1))
{
destBuffer = InvalidBuffer;
destPage = NULL;
}
else if (xldata->initDest)
{
/* just re-init the dest page */
destBuffer = XLogInitBufferForRedo(record, 1);
destPage = (Page) BufferGetPage(destBuffer);
SpGistInitBuffer(destBuffer,
SPGIST_LEAF | (xldata->storesNulls ? SPGIST_NULLS : 0));
/* don't update LSN etc till we're done with it */
}
else
{
/*
* We could probably release the page lock immediately in the
* full-page-image case, but for safety let's hold it till later.
*/
if (XLogReadBufferForRedo(record, 1, &destBuffer) == BLK_NEEDS_REDO)
destPage = (Page) BufferGetPage(destBuffer);
else
destPage = NULL; /* don't do any page updates */
}
/* restore leaf tuples to src and/or dest page */
for (i = 0; i < xldata->nInsert; i++)
{
char *leafTuple;
SpGistLeafTupleData leafTupleHdr;
/* the tuples are not aligned, so must copy to access the size field. */
leafTuple = ptr;
memcpy(&leafTupleHdr, leafTuple, sizeof(SpGistLeafTupleData));
ptr += leafTupleHdr.size;
page = leafPageSelect[i] ? destPage : srcPage;
if (page == NULL)
continue; /* no need to touch this page */
addOrReplaceTuple(page, (Item) leafTuple, leafTupleHdr.size,
toInsert[i]);
}
/* Now update src and dest page LSNs if needed */
if (srcPage != NULL)
{
PageSetLSN(srcPage, lsn);
MarkBufferDirty(srcBuffer);
}
if (destPage != NULL)
{
PageSetLSN(destPage, lsn);
MarkBufferDirty(destBuffer);
}
/* restore new inner tuple */
if (xldata->initInner)
{
innerBuffer = XLogInitBufferForRedo(record, 2);
SpGistInitBuffer(innerBuffer, (xldata->storesNulls ? SPGIST_NULLS : 0));
action = BLK_NEEDS_REDO;
}
else
action = XLogReadBufferForRedo(record, 2, &innerBuffer);
if (action == BLK_NEEDS_REDO)
{
page = BufferGetPage(innerBuffer);
addOrReplaceTuple(page, (Item) innerTuple, innerTupleHdr.size,
xldata->offnumInner);
/* if inner is also parent, update link while we're here */
if (xldata->innerIsParent)
{
SpGistInnerTuple parent;
parent = (SpGistInnerTuple) PageGetItem(page,
PageGetItemId(page, xldata->offnumParent));
spgUpdateNodeLink(parent, xldata->nodeI,
blknoInner, xldata->offnumInner);
}
PageSetLSN(page, lsn);
MarkBufferDirty(innerBuffer);
}
if (BufferIsValid(innerBuffer))
UnlockReleaseBuffer(innerBuffer);
/*
* Now we can release the leaf-page locks. It's okay to do this before
* updating the parent downlink.
*/
if (BufferIsValid(srcBuffer))
UnlockReleaseBuffer(srcBuffer);
if (BufferIsValid(destBuffer))
UnlockReleaseBuffer(destBuffer);
/* update parent downlink, unless we did it above */
if (XLogRecHasBlockRef(record, 3))
{
Buffer parentBuffer;
if (XLogReadBufferForRedo(record, 3, &parentBuffer) == BLK_NEEDS_REDO)
{
SpGistInnerTuple parent;
page = BufferGetPage(parentBuffer);
parent = (SpGistInnerTuple) PageGetItem(page,
PageGetItemId(page, xldata->offnumParent));
spgUpdateNodeLink(parent, xldata->nodeI,
blknoInner, xldata->offnumInner);
PageSetLSN(page, lsn);
MarkBufferDirty(parentBuffer);
}
if (BufferIsValid(parentBuffer))
UnlockReleaseBuffer(parentBuffer);
}
else
Assert(xldata->innerIsParent || xldata->isRootSplit);
}
/*
* Get the latestRemovedXid from the heap pages pointed at by the index
* tuples being deleted. See also btree_xlog_delete_get_latestRemovedXid,
* on which this function is based.
*/
static TransactionId
gistRedoDeleteRecordGetLatestRemovedXid(XLogReaderState *record)
{
gistxlogDelete *xlrec = (gistxlogDelete *) XLogRecGetData(record);
OffsetNumber *todelete;
Buffer ibuffer,
hbuffer;
Page ipage,
hpage;
RelFileNode rnode;
BlockNumber blkno;
ItemId iitemid,
hitemid;
IndexTuple itup;
HeapTupleHeader htuphdr;
BlockNumber hblkno;
OffsetNumber hoffnum;
TransactionId latestRemovedXid = InvalidTransactionId;
int i;
/*
* If there's nothing running on the standby we don't need to derive a
* full latestRemovedXid value, so use a fast path out of here. This
* returns InvalidTransactionId, and so will conflict with all HS
* transactions; but since we just worked out that that's zero people,
* it's OK.
*
* XXX There is a race condition here, which is that a new backend might
* start just after we look. If so, it cannot need to conflict, but this
* coding will result in throwing a conflict anyway.
*/
if (CountDBBackends(InvalidOid) == 0)
return latestRemovedXid;
/*
* In what follows, we have to examine the previous state of the index
* page, as well as the heap page(s) it points to. This is only valid if
* WAL replay has reached a consistent database state; which means that
* the preceding check is not just an optimization, but is *necessary*. We
* won't have let in any user sessions before we reach consistency.
*/
if (!reachedConsistency)
elog(PANIC, "gistRedoDeleteRecordGetLatestRemovedXid: cannot operate with inconsistent data");
/*
* Get index page. If the DB is consistent, this should not fail, nor
* should any of the heap page fetches below. If one does, we return
* InvalidTransactionId to cancel all HS transactions. That's probably
* overkill, but it's safe, and certainly better than panicking here.
*/
XLogRecGetBlockTag(record, 0, &rnode, NULL, &blkno);
ibuffer = XLogReadBufferExtended(rnode, MAIN_FORKNUM, blkno, RBM_NORMAL);
if (!BufferIsValid(ibuffer))
return InvalidTransactionId;
LockBuffer(ibuffer, BUFFER_LOCK_EXCLUSIVE);
ipage = (Page) BufferGetPage(ibuffer);
/*
* Loop through the deleted index items to obtain the TransactionId from
* the heap items they point to.
*/
todelete = (OffsetNumber *) ((char *) xlrec + SizeOfGistxlogDelete);
for (i = 0; i < xlrec->ntodelete; i++)
{
/*
* Identify the index tuple about to be deleted
*/
iitemid = PageGetItemId(ipage, todelete[i]);
itup = (IndexTuple) PageGetItem(ipage, iitemid);
/*
* Locate the heap page that the index tuple points at
*/
hblkno = ItemPointerGetBlockNumber(&(itup->t_tid));
hbuffer = XLogReadBufferExtended(xlrec->hnode, MAIN_FORKNUM, hblkno, RBM_NORMAL);
if (!BufferIsValid(hbuffer))
{
UnlockReleaseBuffer(ibuffer);
return InvalidTransactionId;
}
LockBuffer(hbuffer, BUFFER_LOCK_SHARE);
hpage = (Page) BufferGetPage(hbuffer);
/*
* Look up the heap tuple header that the index tuple points at by
* using the heap node supplied with the xlrec. We can't use
* heap_fetch, since it uses ReadBuffer rather than XLogReadBuffer.
* Note that we are not looking at tuple data here, just headers.
*/
hoffnum = ItemPointerGetOffsetNumber(&(itup->t_tid));
hitemid = PageGetItemId(hpage, hoffnum);
/*
* Follow any redirections until we find something useful.
*/
while (ItemIdIsRedirected(hitemid))
{
hoffnum = ItemIdGetRedirect(hitemid);
hitemid = PageGetItemId(hpage, hoffnum);
CHECK_FOR_INTERRUPTS();
}
/*
* If the heap item has storage, then read the header and use that to
* set latestRemovedXid.
*
* Some LP_DEAD items may not be accessible, so we ignore them.
*/
if (ItemIdHasStorage(hitemid))
{
htuphdr = (HeapTupleHeader) PageGetItem(hpage, hitemid);
HeapTupleHeaderAdvanceLatestRemovedXid(htuphdr, &latestRemovedXid);
}
else if (ItemIdIsDead(hitemid))
{
/*
* Conjecture: if hitemid is dead then it had xids before the xids
* marked on LP_NORMAL items. So we just ignore this item and move
* onto the next, for the purposes of calculating
* latestRemovedxids.
*/
}
else
Assert(!ItemIdIsUsed(hitemid));
UnlockReleaseBuffer(hbuffer);
}
UnlockReleaseBuffer(ibuffer);
/*
* If all heap tuples were LP_DEAD then we will be returning
* InvalidTransactionId here, which avoids conflicts. This matches
* existing logic which assumes that LP_DEAD tuples must already be older
* than the latestRemovedXid on the cleanup record that set them as
* LP_DEAD, hence must already have generated a conflict.
*/
return latestRemovedXid;
}
