/*
* Register new buffer for generic xlog record.
*
* Returns pointer to the page's image in the GenericXLogState, which
* is what the caller should modify.
*
* If the buffer is already registered, just return its existing entry.
* (It's not very clear what to do with the flags in such a case, but
* for now we stay with the original flags.)
*/
Page
GenericXLogRegisterBuffer(GenericXLogState *state, Buffer buffer, int flags)
{
int block_id;
/* Search array for existing entry or first unused slot */
for (block_id = 0; block_id < MAX_GENERIC_XLOG_PAGES; block_id++)
{
PageData *page = &state->pages[block_id];
if (BufferIsInvalid(page->buffer))
{
/* Empty slot, so use it (there cannot be a match later) */
page->buffer = buffer;
page->flags = flags;
memcpy(page->image, BufferGetPage(buffer), BLCKSZ);
return (Page) page->image;
}
else if (page->buffer == buffer)
{
/*
* Buffer is already registered. Just return the image, which is
* already prepared.
*/
return (Page) page->image;
}
}
elog(ERROR, "maximum number %d of generic xlog buffers is exceeded",
MAX_GENERIC_XLOG_PAGES);
/* keep compiler quiet */
return NULL;
}
/*
* Release resources associated with a BrinBuildState.
*/
static void
terminate_brin_buildstate(BrinBuildState *state)
{
/*
* Release the last index buffer used. We might as well ensure that
* whatever free space remains in that page is available in FSM, too.
*/
if (!BufferIsInvalid(state->bs_currentInsertBuf))
{
Page page;
Size freespace;
BlockNumber blk;
page = BufferGetPage(state->bs_currentInsertBuf);
freespace = PageGetFreeSpace(page);
blk = BufferGetBlockNumber(state->bs_currentInsertBuf);
ReleaseBuffer(state->bs_currentInsertBuf);
RecordPageWithFreeSpace(state->bs_irel, blk, freespace);
FreeSpaceMapVacuumRange(state->bs_irel, blk, blk + 1);
}
brin_free_desc(state->bs_bdesc);
pfree(state->bs_dtuple);
pfree(state);
}
/*
* Release resources associated with a BrinBuildState.
*/
static void
terminate_brin_buildstate(BrinBuildState *state)
{
/* release the last index buffer used */
if (!BufferIsInvalid(state->bs_currentInsertBuf))
{
Page page;
page = BufferGetPage(state->bs_currentInsertBuf);
RecordPageWithFreeSpace(state->bs_irel,
BufferGetBlockNumber(state->bs_currentInsertBuf),
PageGetFreeSpace(page));
ReleaseBuffer(state->bs_currentInsertBuf);
}
brin_free_desc(state->bs_bdesc);
pfree(state->bs_dtuple);
pfree(state);
}
/*
* Bulk deletion of all index entries pointing to a set of heap tuples.
* The set of target tuples is specified via a callback routine that tells
* whether any given heap tuple (identified by ItemPointer) is being deleted.
*
* This function also deletes the tuples that are moved by split to other
* bucket.
*
* Result: a palloc'd struct containing statistical info for VACUUM displays.
*/
IndexBulkDeleteResult *
hashbulkdelete(IndexVacuumInfo *info, IndexBulkDeleteResult *stats,
IndexBulkDeleteCallback callback, void *callback_state)
{
Relation rel = info->index;
double tuples_removed;
double num_index_tuples;
double orig_ntuples;
Bucket orig_maxbucket;
Bucket cur_maxbucket;
Bucket cur_bucket;
Buffer metabuf = InvalidBuffer;
HashMetaPage metap;
HashMetaPage cachedmetap;
tuples_removed = 0;
num_index_tuples = 0;
/*
* We need a copy of the metapage so that we can use its hashm_spares[]
* values to compute bucket page addresses, but a cached copy should be
* good enough. (If not, we'll detect that further down and refresh the
* cache as necessary.)
*/
cachedmetap = _hash_getcachedmetap(rel, &metabuf, false);
Assert(cachedmetap != NULL);
orig_maxbucket = cachedmetap->hashm_maxbucket;
orig_ntuples = cachedmetap->hashm_ntuples;
/* Scan the buckets that we know exist */
cur_bucket = 0;
cur_maxbucket = orig_maxbucket;
loop_top:
while (cur_bucket <= cur_maxbucket)
{
BlockNumber bucket_blkno;
BlockNumber blkno;
Buffer bucket_buf;
Buffer buf;
HashPageOpaque bucket_opaque;
Page page;
bool split_cleanup = false;
/* Get address of bucket's start page */
bucket_blkno = BUCKET_TO_BLKNO(cachedmetap, cur_bucket);
blkno = bucket_blkno;
/*
* We need to acquire a cleanup lock on the primary bucket page to out
* wait concurrent scans before deleting the dead tuples.
*/
buf = ReadBufferExtended(rel, MAIN_FORKNUM, blkno, RBM_NORMAL, info->strategy);
LockBufferForCleanup(buf);
_hash_checkpage(rel, buf, LH_BUCKET_PAGE);
page = BufferGetPage(buf);
bucket_opaque = (HashPageOpaque) PageGetSpecialPointer(page);
/*
* If the bucket contains tuples that are moved by split, then we need
* to delete such tuples. We can't delete such tuples if the split
* operation on bucket is not finished as those are needed by scans.
*/
if (!H_BUCKET_BEING_SPLIT(bucket_opaque) &&
H_NEEDS_SPLIT_CLEANUP(bucket_opaque))
{
split_cleanup = true;
/*
* This bucket might have been split since we last held a lock on
* the metapage. If so, hashm_maxbucket, hashm_highmask and
* hashm_lowmask might be old enough to cause us to fail to remove
* tuples left behind by the most recent split. To prevent that,
* now that the primary page of the target bucket has been locked
* (and thus can't be further split), check whether we need to
* update our cached metapage data.
*/
Assert(bucket_opaque->hasho_prevblkno != InvalidBlockNumber);
if (bucket_opaque->hasho_prevblkno > cachedmetap->hashm_maxbucket)
{
cachedmetap = _hash_getcachedmetap(rel, &metabuf, true);
Assert(cachedmetap != NULL);
}
}
bucket_buf = buf;
hashbucketcleanup(rel, cur_bucket, bucket_buf, blkno, info->strategy,
cachedmetap->hashm_maxbucket,
cachedmetap->hashm_highmask,
cachedmetap->hashm_lowmask, &tuples_removed,
&num_index_tuples, split_cleanup,
callback, callback_state);
_hash_dropbuf(rel, bucket_buf);
/* Advance to next bucket */
cur_bucket++;
}
if (BufferIsInvalid(metabuf))
metabuf = _hash_getbuf(rel, HASH_METAPAGE, HASH_NOLOCK, LH_META_PAGE);
/* Write-lock metapage and check for split since we started */
LockBuffer(metabuf, BUFFER_LOCK_EXCLUSIVE);
metap = HashPageGetMeta(BufferGetPage(metabuf));
if (cur_maxbucket != metap->hashm_maxbucket)
{
/* There's been a split, so process the additional bucket(s) */
LockBuffer(metabuf, BUFFER_LOCK_UNLOCK);
cachedmetap = _hash_getcachedmetap(rel, &metabuf, true);
Assert(cachedmetap != NULL);
cur_maxbucket = cachedmetap->hashm_maxbucket;
goto loop_top;
}
/* Okay, we're really done. Update tuple count in metapage. */
START_CRIT_SECTION();
if (orig_maxbucket == metap->hashm_maxbucket &&
orig_ntuples == metap->hashm_ntuples)
{
/*
* No one has split or inserted anything since start of scan, so
* believe our count as gospel.
*/
metap->hashm_ntuples = num_index_tuples;
}
else
{
/*
* Otherwise, our count is untrustworthy since we may have
* double-scanned tuples in split buckets. Proceed by dead-reckoning.
* (Note: we still return estimated_count = false, because using this
* count is better than not updating reltuples at all.)
*/
if (metap->hashm_ntuples > tuples_removed)
metap->hashm_ntuples -= tuples_removed;
else
metap->hashm_ntuples = 0;
num_index_tuples = metap->hashm_ntuples;
}
MarkBufferDirty(metabuf);
/* XLOG stuff */
if (RelationNeedsWAL(rel))
{
xl_hash_update_meta_page xlrec;
XLogRecPtr recptr;
xlrec.ntuples = metap->hashm_ntuples;
XLogBeginInsert();
XLogRegisterData((char *) &xlrec, SizeOfHashUpdateMetaPage);
XLogRegisterBuffer(0, metabuf, REGBUF_STANDARD);
recptr = XLogInsert(RM_HASH_ID, XLOG_HASH_UPDATE_META_PAGE);
PageSetLSN(BufferGetPage(metabuf), recptr);
}
END_CRIT_SECTION();
_hash_relbuf(rel, metabuf);
/* return statistics */
if (stats == NULL)
stats = (IndexBulkDeleteResult *) palloc0(sizeof(IndexBulkDeleteResult));
stats->estimated_count = false;
stats->num_index_tuples = num_index_tuples;
stats->tuples_removed += tuples_removed;
/* hashvacuumcleanup will fill in num_pages */
return stats;
}
/*
* Apply changes represented by GenericXLogState to the actual buffers,
* and emit a generic xlog record.
*/
XLogRecPtr
GenericXLogFinish(GenericXLogState *state)
{
XLogRecPtr lsn;
int i;
if (state->isLogged)
{
/* Logged relation: make xlog record in critical section. */
XLogBeginInsert();
START_CRIT_SECTION();
for (i = 0; i < MAX_GENERIC_XLOG_PAGES; i++)
{
PageData *pageData = &state->pages[i];
Page page;
PageHeader pageHeader;
if (BufferIsInvalid(pageData->buffer))
continue;
page = BufferGetPage(pageData->buffer);
pageHeader = (PageHeader) pageData->image;
if (pageData->flags & GENERIC_XLOG_FULL_IMAGE)
{
/*
* A full-page image does not require us to supply any xlog
* data. Just apply the image, being careful to zero the
* "hole" between pd_lower and pd_upper in order to avoid
* divergence between actual page state and what replay would
* produce.
*/
memcpy(page, pageData->image, pageHeader->pd_lower);
memset(page + pageHeader->pd_lower, 0,
pageHeader->pd_upper - pageHeader->pd_lower);
memcpy(page + pageHeader->pd_upper,
pageData->image + pageHeader->pd_upper,
BLCKSZ - pageHeader->pd_upper);
XLogRegisterBuffer(i, pageData->buffer,
REGBUF_FORCE_IMAGE | REGBUF_STANDARD);
}
else
{
/*
* In normal mode, calculate delta and write it as xlog data
* associated with this page.
*/
computeDelta(pageData, page, (Page) pageData->image);
/* Apply the image, with zeroed "hole" as above */
memcpy(page, pageData->image, pageHeader->pd_lower);
memset(page + pageHeader->pd_lower, 0,
pageHeader->pd_upper - pageHeader->pd_lower);
memcpy(page + pageHeader->pd_upper,
pageData->image + pageHeader->pd_upper,
BLCKSZ - pageHeader->pd_upper);
XLogRegisterBuffer(i, pageData->buffer, REGBUF_STANDARD);
XLogRegisterBufData(i, pageData->delta, pageData->deltaLen);
}
}
/* Insert xlog record */
lsn = XLogInsert(RM_GENERIC_ID, 0);
/* Set LSN and mark buffers dirty */
for (i = 0; i < MAX_GENERIC_XLOG_PAGES; i++)
{
PageData *pageData = &state->pages[i];
if (BufferIsInvalid(pageData->buffer))
continue;
PageSetLSN(BufferGetPage(pageData->buffer), lsn);
MarkBufferDirty(pageData->buffer);
}
END_CRIT_SECTION();
}
else
{
/* Unlogged relation: skip xlog-related stuff */
START_CRIT_SECTION();
for (i = 0; i < MAX_GENERIC_XLOG_PAGES; i++)
{
PageData *pageData = &state->pages[i];
if (BufferIsInvalid(pageData->buffer))
continue;
memcpy(BufferGetPage(pageData->buffer),
pageData->image,
BLCKSZ);
/* We don't worry about zeroing the "hole" in this case */
MarkBufferDirty(pageData->buffer);
}
END_CRIT_SECTION();
/* We don't have a LSN to return, in this case */
lsn = InvalidXLogRecPtr;
}
for (i = 0; i < MAX_GENERIC_XLOG_PAGES; i++)
pfree(state->pages[i].image);
pfree(state);
return lsn;
}