static void
PersistentBuild_NonTransactionTruncate(RelFileNode *relFileNode)
{
SMgrRelation smgrRelation;
PersistentFileSysObjName fsObjName;
PersistentFileSysObjName_SetRelationFile(
&fsObjName,
relFileNode,
/* segmentFileNum */ 0,
is_tablespace_shared);
if (Debug_persistent_print)
elog(Persistent_DebugPrintLevel(),
"Non-transaction truncate of '%s'",
PersistentFileSysObjName_ObjectName(&fsObjName));
smgrRelation = smgropen(*relFileNode);
smgrtruncate(
smgrRelation,
0,
/* isTemp */ true,
/* isLocalBuf */ false,
/* persistentTid */ NULL,
/* persistentSerialNum */ 0);
smgrclose(smgrRelation);
}
/*
* FreeSpaceMapTruncateRel - adjust for truncation of a relation.
*
* The caller must hold AccessExclusiveLock on the relation, to ensure that
* other backends receive the smgr invalidation event that this function sends
* before they access the FSM again.
*
* nblocks is the new___ size of the heap.
*/
void
FreeSpaceMapTruncateRel(Relation rel, BlockNumber nblocks)
{
BlockNumber new_nfsmblocks;
FSMAddress first_removed_address;
uint16 first_removed_slot;
Buffer buf;
RelationOpenSmgr(rel);
/*
* If no FSM has been created yet for this relation, there's nothing to
* truncate.
*/
if (!smgrexists(rel->rd_smgr, FSM_FORKNUM))
return;
/* Get the location in the FSM of the first removed heap block */
first_removed_address = fsm_get_location(nblocks, &first_removed_slot);
/*
* Zero out the tail of the last remaining FSM page. If the slot
* representing the first removed heap block is at a page boundary, as the
* first slot on the FSM page that first_removed_address points to, we can
* just truncate that page altogether.
*/
if (first_removed_slot > 0)
{
buf = fsm_readbuf(rel, first_removed_address, false);
if (!BufferIsValid(buf))
return; /* nothing to do; the FSM was already smaller */
LockBuffer(buf, BUFFER_LOCK_EXCLUSIVE);
fsm_truncate_avail(BufferGetPage(buf), first_removed_slot);
MarkBufferDirtyHint(buf, false);
UnlockReleaseBuffer(buf);
new_nfsmblocks = fsm_logical_to_physical(first_removed_address) + 1;
}
else
{
new_nfsmblocks = fsm_logical_to_physical(first_removed_address);
if (smgrnblocks(rel->rd_smgr, FSM_FORKNUM) <= new_nfsmblocks)
return; /* nothing to do; the FSM was already smaller */
}
/* Truncate the unused FSM pages, and send smgr inval message */
smgrtruncate(rel->rd_smgr, FSM_FORKNUM, new_nfsmblocks);
/*
* We might as well update the local smgr_fsm_nblocks setting.
* smgrtruncate sent an smgr cache inval message, which will cause other
* backends to invalidate their copy of smgr_fsm_nblocks, and this one too
* at the next command boundary. But this ensures it isn't outright wrong
* until then.
*/
if (rel->rd_smgr)
rel->rd_smgr->smgr_fsm_nblocks = new_nfsmblocks;
}
void
smgr_redo(XLogRecPtr lsn, XLogRecord *record)
{
uint8 info = record->xl_info & ~XLR_INFO_MASK;
/* Backup blocks are not used in smgr records */
Assert(!(record->xl_info & XLR_BKP_BLOCK_MASK));
if (info == XLOG_SMGR_CREATE)
{
xl_smgr_create *xlrec = (xl_smgr_create *) XLogRecGetData(record);
SMgrRelation reln;
reln = smgropen(xlrec->rnode);
smgrcreate(reln, MAIN_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);
/*
* Forcibly create relation if it doesn't exist (which suggests that
* it was dropped somewhere later in the WAL sequence). As in
* XLogOpenRelation, we prefer to recreate the rel and replay the log
* as best we can until the drop is seen.
*/
smgrcreate(reln, MAIN_FORKNUM, true);
smgrtruncate(reln, MAIN_FORKNUM, xlrec->blkno, false);
/* Also tell xlogutils.c about it */
XLogTruncateRelation(xlrec->rnode, MAIN_FORKNUM, xlrec->blkno);
/* Truncate FSM and VM too */
rel = CreateFakeRelcacheEntry(xlrec->rnode);
if (smgrexists(reln, FSM_FORKNUM))
FreeSpaceMapTruncateRel(rel, xlrec->blkno);
if (smgrexists(reln, VISIBILITYMAP_FORKNUM))
visibilitymap_truncate(rel, xlrec->blkno);
FreeFakeRelcacheEntry(rel);
}
else
elog(PANIC, "smgr_redo: unknown op code %u", info);
}
/*
* visibilitymap_truncate - truncate the visibility map
*
* The caller must hold AccessExclusiveLock on the relation, to ensure that
* other backends receive the smgr invalidation event that this function sends
* before they access the VM again.
*
* nheapblocks is the new size of the heap.
*/
void
visibilitymap_truncate(Relation rel, BlockNumber nheapblocks)
{
BlockNumber newnblocks;
/* last remaining block, byte, and bit */
BlockNumber truncBlock = HEAPBLK_TO_MAPBLOCK(nheapblocks);
uint32 truncByte = HEAPBLK_TO_MAPBYTE(nheapblocks);
uint8 truncBit = HEAPBLK_TO_MAPBIT(nheapblocks);
#ifdef TRACE_VISIBILITYMAP
elog(DEBUG1, "vm_truncate %s %d", RelationGetRelationName(rel), nheapblocks);
#endif
RelationOpenSmgr(rel);
/*
* If no visibility map has been created yet for this relation, there's
* nothing to truncate.
*/
if (!smgrexists(rel->rd_smgr, VISIBILITYMAP_FORKNUM))
return;
/*
* Unless the new size is exactly at a visibility map page boundary, the
* tail bits in the last remaining map page, representing truncated heap
* blocks, need to be cleared. This is not only tidy, but also necessary
* because we don't get a chance to clear the bits if the heap is extended
* again.
*/
if (truncByte != 0 || truncBit != 0)
{
Buffer mapBuffer;
Page page;
char *map;
newnblocks = truncBlock + 1;
mapBuffer = vm_readbuf(rel, truncBlock, false);
if (!BufferIsValid(mapBuffer))
{
/* nothing to do, the file was already smaller */
return;
}
page = BufferGetPage(mapBuffer);
map = PageGetContents(page);
LockBuffer(mapBuffer, BUFFER_LOCK_EXCLUSIVE);
/* Clear out the unwanted bytes. */
MemSet(&map[truncByte + 1], 0, MAPSIZE - (truncByte + 1));
/*
* Mask out the unwanted bits of the last remaining byte.
*
* ((1 << 0) - 1) = 00000000 ((1 << 1) - 1) = 00000001 ... ((1 << 6) -
* 1) = 00111111 ((1 << 7) - 1) = 01111111
*/
map[truncByte] &= (1 << truncBit) - 1;
MarkBufferDirty(mapBuffer);
UnlockReleaseBuffer(mapBuffer);
}
else
newnblocks = truncBlock;
if (smgrnblocks(rel->rd_smgr, VISIBILITYMAP_FORKNUM) <= newnblocks)
{
/* nothing to do, the file was already smaller than requested size */
return;
}
/* Truncate the unused VM pages, and send smgr inval message */
smgrtruncate(rel->rd_smgr, VISIBILITYMAP_FORKNUM, newnblocks);
/*
* We might as well update the local smgr_vm_nblocks setting. smgrtruncate
* sent an smgr cache inval message, which will cause other backends to
* invalidate their copy of smgr_vm_nblocks, and this one too at the next
* command boundary. But this ensures it isn't outright wrong until then.
*/
if (rel->rd_smgr)
rel->rd_smgr->smgr_vm_nblocks = newnblocks;
}
static int
FileRepPrimary_ResyncWrite(FileRepResyncHashEntry_s *entry)
{
int status = STATUS_OK;
Page page;
Buffer buf;
BlockNumber numBlocks;
BlockNumber blkno;
SMgrRelation smgr_relation;
char relidstr[OIDCHARS + 1 + OIDCHARS + 1 + OIDCHARS + 1];
XLogRecPtr loc;
int count = 0;
int thresholdCount = 0;
bool mirrorDataLossOccurred = FALSE;
switch (entry->relStorageMgr)
{
case PersistentFileSysRelStorageMgr_BufferPool:
switch (entry->mirrorDataSynchronizationState)
{
case MirroredRelDataSynchronizationState_BufferPoolScanIncremental:
case MirroredRelDataSynchronizationState_FullCopy:
smgr_relation = smgropen(entry->relFileNode);
numBlocks = smgrnblocks(smgr_relation);
snprintf(relidstr, sizeof(relidstr), "%u/%u/%u",
smgr_relation->smgr_rnode.spcNode,
smgr_relation->smgr_rnode.dbNode,
smgr_relation->smgr_rnode.relNode);
if (Debug_filerep_print)
elog(LOG, "resync buffer pool relation '%s' number of blocks '%d' ",
relidstr, numBlocks);
thresholdCount = Min(numBlocks, 1024);
/*
* required in order to report how many blocks were synchronized
* if gp_persistent_relation_node does not return that information
*/
if (entry->mirrorBufpoolResyncChangedPageCount == 0)
{
entry->mirrorBufpoolResyncChangedPageCount = numBlocks - entry->mirrorBufpoolResyncCkptBlockNum;
}
for (blkno = entry->mirrorBufpoolResyncCkptBlockNum; blkno < numBlocks; blkno++)
{
XLogRecPtr endResyncLSN = (isFullResync() ?
FileRepResync_GetEndFullResyncLSN() :
FileRepResync_GetEndIncrResyncLSN());
#ifdef FAULT_INJECTOR
FaultInjector_InjectFaultIfSet(
FileRepResyncWorkerRead,
DDLNotSpecified,
"", //databaseName
""); // tableName
#endif
FileRepResync_SetReadBufferRequest();
buf = ReadBuffer_Resync(smgr_relation, blkno, relidstr);
FileRepResync_ResetReadBufferRequest();
LockBuffer(buf, BUFFER_LOCK_EXCLUSIVE);
page = BufferGetPage(buf);
loc = PageGetLSN(page);
if (Debug_filerep_print)
{
elog(LOG,
"full resync buffer pool identifier '%s' num blocks '%d' blkno '%d' lsn begin change tracking '%s(%u/%u)' "
"lsn page '%s(%u/%u)' lsn end change tracking '%s(%u/%u)' ",
relidstr,
numBlocks,
blkno,
XLogLocationToString(&entry->mirrorBufpoolResyncCkptLoc),
entry->mirrorBufpoolResyncCkptLoc.xlogid,
entry->mirrorBufpoolResyncCkptLoc.xrecoff,
XLogLocationToString(&loc),
loc.xlogid,
loc.xrecoff,
XLogLocationToString(&endResyncLSN),
endResyncLSN.xlogid,
endResyncLSN.xrecoff);
}
else
{
char tmpBuf[FILEREP_MAX_LOG_DESCRIPTION_LEN];
snprintf(tmpBuf, sizeof(tmpBuf),
"full resync buffer pool identifier '%s' num blocks '%d' blkno '%d' lsn begin change tracking '%s(%u/%u)' ",
relidstr,
numBlocks,
blkno,
XLogLocationToString(&entry->mirrorBufpoolResyncCkptLoc),
entry->mirrorBufpoolResyncCkptLoc.xlogid,
entry->mirrorBufpoolResyncCkptLoc.xrecoff);
FileRep_InsertConfigLogEntry(tmpBuf);
snprintf(tmpBuf, sizeof(tmpBuf),
"full resync buffer pool identifier '%s' lsn page '%s(%u/%u)' lsn end change tracking '%s(%u/%u)' ",
relidstr,
XLogLocationToString(&loc),
loc.xlogid,
loc.xrecoff,
XLogLocationToString(&endResyncLSN),
endResyncLSN.xlogid,
endResyncLSN.xrecoff);
FileRep_InsertConfigLogEntry(tmpBuf);
}
if (XLByteLE(PageGetLSN(page), endResyncLSN) &&
XLByteLE(entry->mirrorBufpoolResyncCkptLoc, PageGetLSN(page)))
{
smgrwrite(smgr_relation,
blkno,
(char *)BufferGetBlock(buf),
FALSE);
}
#ifdef FAULT_INJECTOR
FaultInjector_InjectFaultIfSet(
FileRepResyncWorker,
DDLNotSpecified,
"", // databaseName
""); // tableName
#endif
UnlockReleaseBuffer(buf);
if (count > thresholdCount)
{
count = 0;
FileRepSubProcess_ProcessSignals();
if (! (FileRepSubProcess_GetState() == FileRepStateReady &&
dataState == DataStateInResync))
{
mirrorDataLossOccurred = TRUE;
break;
}
}
else
count++;
}
if (mirrorDataLossOccurred)
break;
if (entry->mirrorDataSynchronizationState != MirroredRelDataSynchronizationState_FullCopy)
{
LockRelationForResyncExtension(&smgr_relation->smgr_rnode, ExclusiveLock);
numBlocks = smgrnblocks(smgr_relation);
smgrtruncate(smgr_relation,
numBlocks,
TRUE /* isTemp, TRUE means to not record in XLOG */,
FALSE /* isLocalBuf */,
&entry->persistentTid,
entry->persistentSerialNum);
UnlockRelationForResyncExtension(&smgr_relation->smgr_rnode, ExclusiveLock);
}
smgrimmedsync(smgr_relation);
smgrclose(smgr_relation);
smgr_relation = NULL;
break;
case MirroredRelDataSynchronizationState_None:
case MirroredRelDataSynchronizationState_DataSynchronized:
break;
default:
ereport(LOG,
(errmsg("could not resynchronize relation '%u/%u/%u' "
"mirror synchronization state:'%s(%d)' ",
entry->relFileNode.relNode,
entry->relFileNode.spcNode,
entry->relFileNode.dbNode,
MirroredRelDataSynchronizationState_Name(entry->mirrorDataSynchronizationState),
entry->mirrorDataSynchronizationState)));
break;
}
break;
case PersistentFileSysRelStorageMgr_AppendOnly:
{
MirroredAppendOnlyOpen mirroredOpen;
int primaryError;
bool mirrorDataLossOccurred;
char *buffer = NULL;
int64 endOffset = entry->mirrorAppendOnlyNewEof;
int64 startOffset = entry->mirrorAppendOnlyLossEof;
int32 bufferLen = 0;
int retval = 0;
switch (entry->mirrorDataSynchronizationState)
{
case MirroredRelDataSynchronizationState_AppendOnlyCatchup:
case MirroredRelDataSynchronizationState_FullCopy:
/*
* required in order to report how many blocks were synchronized
* if gp_persistent_relation_node does not return that information
*/
if (entry->mirrorBufpoolResyncChangedPageCount == 0)
{
entry->mirrorBufpoolResyncChangedPageCount = (endOffset - startOffset) / BLCKSZ;
}
/*
* The MirroredAppendOnly_OpenResynchonize routine knows we are a resynch worker and
* will open BOTH, but write only the MIRROR!!!
*/
MirroredAppendOnly_OpenResynchonize(
&mirroredOpen,
&entry->relFileNode,
entry->segmentFileNum,
startOffset,
&primaryError,
&mirrorDataLossOccurred);
if (primaryError != 0)
{
ereport(ERROR,
(errcode_for_file_access(),
errmsg("could not open file %u/%u/%u.%u : %s",
entry->relFileNode.dbNode,
entry->relFileNode.spcNode,
entry->relFileNode.relNode,
entry->segmentFileNum,
strerror(primaryError))));
break;
}
if (mirrorDataLossOccurred)
break;
/* AO and CO Data Store writes 64k size by default */
bufferLen = (Size) Min(2*BLCKSZ, endOffset - startOffset);
buffer = (char*) palloc(bufferLen);
if (buffer == NULL)
ereport(ERROR,
(errcode(ERRCODE_OUT_OF_MEMORY),
(errmsg("not enough memory for resynchronization"))));
MemSet(buffer, 0, bufferLen);
while (startOffset < endOffset)
{
retval = MirroredAppendOnly_Read(
&mirroredOpen,
buffer,
bufferLen);
if (retval != bufferLen)
{
ereport(ERROR,
(errcode_for_file_access(),
errmsg("could not read from position:" INT64_FORMAT " in file %u/%u/%u.%u : %m",
startOffset,
entry->relFileNode.dbNode,
entry->relFileNode.spcNode,
entry->relFileNode.relNode,
entry->segmentFileNum)));
break;
}
MirroredAppendOnly_Append(
&mirroredOpen,
buffer,
bufferLen,
&primaryError,
&mirrorDataLossOccurred);
if (mirrorDataLossOccurred)
break;
Assert(primaryError == 0); // No primary writes as resync worker.
startOffset += bufferLen;
/* AO and CO Data Store writes 64k size by default */
bufferLen = (Size) Min(2*BLCKSZ, endOffset - startOffset);
}
if (buffer)
{
pfree(buffer);
buffer = NULL;
}
if (mirrorDataLossOccurred)
break;
/* Flush written data on Mirror */
MirroredAppendOnly_Flush(
&mirroredOpen,
&primaryError,
&mirrorDataLossOccurred);
if (mirrorDataLossOccurred)
break;
Assert(primaryError == 0); // Not flushed on primary as resync worker.
/* Close Primary and Mirror */
MirroredAppendOnly_Close(
&mirroredOpen,
&mirrorDataLossOccurred);
break;
case MirroredRelDataSynchronizationState_None:
case MirroredRelDataSynchronizationState_DataSynchronized:
break;
default:
ereport(LOG,
(errmsg("could not resynchronize relation '%u/%u/%u' "
"mirror synchronization state:'%s(%d)' ",
entry->relFileNode.relNode,
entry->relFileNode.spcNode,
entry->relFileNode.dbNode,
MirroredRelDataSynchronizationState_Name(entry->mirrorDataSynchronizationState),
entry->mirrorDataSynchronizationState)));
break;
}
break;
} //case
default:
Assert(0);
break;
} //switch
if (mirrorDataLossOccurred)
status = STATUS_ERROR;
return status;
}
/*
* RelationTruncate
* Physically truncate a relation to the specified number of blocks.
*
* This includes getting rid of any buffers for the blocks that are to be
* dropped.
*/
void
RelationTruncate(Relation rel, BlockNumber nblocks)
{
bool fsm;
bool vm;
/* Open it at the smgr level if not already done */
RelationOpenSmgr(rel);
/*
* Make sure smgr_targblock etc aren't pointing somewhere past new end
*/
rel->rd_smgr->smgr_targblock = InvalidBlockNumber;
rel->rd_smgr->smgr_fsm_nblocks = InvalidBlockNumber;
rel->rd_smgr->smgr_vm_nblocks = InvalidBlockNumber;
/* Truncate the FSM first if it exists */
fsm = smgrexists(rel->rd_smgr, FSM_FORKNUM);
if (fsm)
FreeSpaceMapTruncateRel(rel, nblocks);
/* Truncate the visibility map too if it exists. */
vm = smgrexists(rel->rd_smgr, VISIBILITYMAP_FORKNUM);
if (vm)
visibilitymap_truncate(rel, nblocks);
/*
* We WAL-log the truncation before actually truncating, which means
* trouble if the truncation fails. If we then crash, the WAL replay
* likely isn't going to succeed in the truncation either, and cause a
* PANIC. It's tempting to put a critical section here, but that cure
* would be worse than the disease. It would turn a usually harmless
* failure to truncate, that might spell trouble at WAL replay, into a
* certain PANIC.
*/
if (!rel->rd_istemp)
{
/*
* Make an XLOG entry reporting the file truncation.
*/
XLogRecPtr lsn;
XLogRecData rdata;
xl_smgr_truncate xlrec;
xlrec.blkno = nblocks;
xlrec.rnode = rel->rd_node;
rdata.data = (char *) &xlrec;
rdata.len = sizeof(xlrec);
rdata.buffer = InvalidBuffer;
rdata.next = NULL;
lsn = XLogInsert(RM_SMGR_ID, XLOG_SMGR_TRUNCATE, &rdata);
/*
* Flush, because otherwise the truncation of the main relation might
* hit the disk before the WAL record, and the truncation of the FSM
* or visibility map. If we crashed during that window, we'd be left
* with a truncated heap, but the FSM or visibility map would still
* contain entries for the non-existent heap pages.
*/
if (fsm || vm)
XLogFlush(lsn);
}
/* Do the real work */
smgrtruncate(rel->rd_smgr, MAIN_FORKNUM, nblocks, rel->rd_istemp);
}
/*
* FreeSpaceMapTruncateRel - adjust for truncation of a relation.
*
* The caller must hold AccessExclusiveLock on the relation, to ensure that
* other backends receive the smgr invalidation event that this function sends
* before they access the FSM again.
*
* nblocks is the new size of the heap.
*/
void
FreeSpaceMapTruncateRel(Relation rel, BlockNumber nblocks)
{
BlockNumber new_nfsmblocks;
FSMAddress first_removed_address;
uint16 first_removed_slot;
Buffer buf;
RelationOpenSmgr(rel);
/*
* If no FSM has been created yet for this relation, there's nothing to
* truncate.
*/
if (!smgrexists(rel->rd_smgr, FSM_FORKNUM))
return;
/* Get the location in the FSM of the first removed heap block */
first_removed_address = fsm_get_location(nblocks, &first_removed_slot);
/*
* Zero out the tail of the last remaining FSM page. If the slot
* representing the first removed heap block is at a page boundary, as the
* first slot on the FSM page that first_removed_address points to, we can
* just truncate that page altogether.
*/
if (first_removed_slot > 0)
{
buf = fsm_readbuf(rel, first_removed_address, false);
if (!BufferIsValid(buf))
return; /* nothing to do; the FSM was already smaller */
LockBuffer(buf, BUFFER_LOCK_EXCLUSIVE);
/* NO EREPORT(ERROR) from here till changes are logged */
START_CRIT_SECTION();
fsm_truncate_avail(BufferGetPage(buf), first_removed_slot);
/*
* Truncation of a relation is WAL-logged at a higher-level, and we
* will be called at WAL replay. But if checksums are enabled, we need
* to still write a WAL record to protect against a torn page, if the
* page is flushed to disk before the truncation WAL record. We cannot
* use MarkBufferDirtyHint here, because that will not dirty the page
* during recovery.
*/
MarkBufferDirty(buf);
if (!InRecovery && RelationNeedsWAL(rel) && XLogHintBitIsNeeded())
log_newpage_buffer(buf, false);
END_CRIT_SECTION();
UnlockReleaseBuffer(buf);
new_nfsmblocks = fsm_logical_to_physical(first_removed_address) + 1;
}
else
{
new_nfsmblocks = fsm_logical_to_physical(first_removed_address);
if (smgrnblocks(rel->rd_smgr, FSM_FORKNUM) <= new_nfsmblocks)
return; /* nothing to do; the FSM was already smaller */
}
/* Truncate the unused FSM pages, and send smgr inval message */
smgrtruncate(rel->rd_smgr, FSM_FORKNUM, new_nfsmblocks);
/*
* We might as well update the local smgr_fsm_nblocks setting.
* smgrtruncate sent an smgr cache inval message, which will cause other
* backends to invalidate their copy of smgr_fsm_nblocks, and this one too
* at the next command boundary. But this ensures it isn't outright wrong
* until then.
*/
if (rel->rd_smgr)
rel->rd_smgr->smgr_fsm_nblocks = new_nfsmblocks;
/*
* Update upper-level FSM pages to account for the truncation. This is
* important because the just-truncated pages were likely marked as
* all-free, and would be preferentially selected.
*/
FreeSpaceMapVacuumRange(rel, nblocks, InvalidBlockNumber);
}
void
smgr_redo(XLogRecPtr lsn, XLogRecord *record)
{
uint8 info = record->xl_info & ~XLR_INFO_MASK;
/* Backup blocks are not used in smgr records */
Assert(!(record->xl_info & XLR_BKP_BLOCK_MASK));
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
* XLogReadBuffer, 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);
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 (smgrexists(reln, FSM_FORKNUM))
FreeSpaceMapTruncateRel(rel, xlrec->blkno);
if (smgrexists(reln, VISIBILITYMAP_FORKNUM))
visibilitymap_truncate(rel, xlrec->blkno);
FreeFakeRelcacheEntry(rel);
}
else
elog(PANIC, "smgr_redo: unknown op code %u", info);
}
/*
* visibilitymap_test - truncate the visibility map
*/
void
visibilitymap_truncate(Relation rel, BlockNumber nheapblocks)
{
BlockNumber newnblocks;
/* last remaining block, byte, and bit */
BlockNumber truncBlock = HEAPBLK_TO_MAPBLOCK(nheapblocks);
uint32 truncByte = HEAPBLK_TO_MAPBYTE(nheapblocks);
uint8 truncBit = HEAPBLK_TO_MAPBIT(nheapblocks);
#ifdef TRACE_VISIBILITYMAP
elog(DEBUG1, "vm_truncate %s %d", RelationGetRelationName(rel), nheapblocks);
#endif
/*
* If no visibility map has been created yet for this relation, there's
* nothing to truncate.
*/
if (!smgrexists(rel->rd_smgr, VISIBILITYMAP_FORKNUM))
return;
/*
* Unless the new size is exactly at a visibility map page boundary, the
* tail bits in the last remaining map page, representing truncated heap
* blocks, need to be cleared. This is not only tidy, but also necessary
* because we don't get a chance to clear the bits if the heap is extended
* again.
*/
if (truncByte != 0 || truncBit != 0)
{
Buffer mapBuffer;
Page page;
char *map;
newnblocks = truncBlock + 1;
mapBuffer = vm_readbuf(rel, truncBlock, false);
if (!BufferIsValid(mapBuffer))
{
/* nothing to do, the file was already smaller */
return;
}
page = BufferGetPage(mapBuffer);
map = PageGetContents(page);
LockBuffer(mapBuffer, BUFFER_LOCK_EXCLUSIVE);
/* Clear out the unwanted bytes. */
MemSet(&map[truncByte + 1], 0, MAPSIZE - (truncByte + 1));
/*
* Mask out the unwanted bits of the last remaining byte.
*
* ((1 << 0) - 1) = 00000000 ((1 << 1) - 1) = 00000001 ... ((1 << 6) -
* 1) = 00111111 ((1 << 7) - 1) = 01111111
*/
map[truncByte] &= (1 << truncBit) - 1;
MarkBufferDirty(mapBuffer);
UnlockReleaseBuffer(mapBuffer);
}
else
newnblocks = truncBlock;
if (smgrnblocks(rel->rd_smgr, VISIBILITYMAP_FORKNUM) < newnblocks)
{
/* nothing to do, the file was already smaller than requested size */
return;
}
smgrtruncate(rel->rd_smgr, VISIBILITYMAP_FORKNUM, newnblocks,
rel->rd_istemp);
/*
* Need to invalidate the relcache entry, because rd_vm_nblocks seen by
* other backends is no longer valid.
*/
if (!InRecovery)
CacheInvalidateRelcache(rel);
rel->rd_vm_nblocks = newnblocks;
}
/*
* reindex_index - This routine is used to recreate a single index
*/
void
reindex_index(Oid indexId)
{
Relation iRel,
heapRelation;
IndexInfo *indexInfo;
Oid heapId;
bool inplace;
/*
* Open our index relation and get an exclusive lock on it.
*
* Note: for REINDEX INDEX, doing this before opening the parent heap
* relation means there's a possibility for deadlock failure against
* another xact that is doing normal accesses to the heap and index.
* However, it's not real clear why you'd be wanting to do REINDEX INDEX
* on a table that's in active use, so I'd rather have the protection of
* making sure the index is locked down. In the REINDEX TABLE and
* REINDEX DATABASE cases, there is no problem because caller already
* holds exclusive lock on the parent table.
*/
iRel = index_open(indexId);
LockRelation(iRel, AccessExclusiveLock);
/* Get OID of index's parent table */
heapId = iRel->rd_index->indrelid;
/* Open and lock the parent heap relation */
heapRelation = heap_open(heapId, AccessExclusiveLock);
SetReindexProcessing(heapId, indexId);
/*
* If it's a shared index, we must do inplace processing (because we
* have no way to update relfilenode in other databases). Otherwise
* we can do it the normal transaction-safe way.
*
* Since inplace processing isn't crash-safe, we only allow it in a
* standalone backend. (In the REINDEX TABLE and REINDEX DATABASE cases,
* the caller should have detected this.)
*/
inplace = iRel->rd_rel->relisshared;
if (inplace && IsUnderPostmaster)
ereport(ERROR,
(errcode(ERRCODE_INSUFFICIENT_PRIVILEGE),
errmsg("shared index \"%s\" can only be reindexed in stand-alone mode",
RelationGetRelationName(iRel))));
/* Fetch info needed for index_build */
indexInfo = BuildIndexInfo(iRel);
if (inplace)
{
/*
* Release any buffers associated with this index. If they're
* dirty, they're just dropped without bothering to flush to disk.
*/
DropRelationBuffers(iRel);
/* Now truncate the actual data and set blocks to zero */
smgrtruncate(DEFAULT_SMGR, iRel, 0);
iRel->rd_nblocks = 0;
iRel->rd_targblock = InvalidBlockNumber;
}
else
{
/*
* We'll build a new physical relation for the index.
*/
setNewRelfilenode(iRel);
}
/* Initialize the index and rebuild */
index_build(heapRelation, iRel, indexInfo);
/*
* index_build will close both the heap and index relations (but not
* give up the locks we hold on them). So we're done.
*/
SetReindexProcessing(InvalidOid, InvalidOid);
}