/*
* Adds array of item pointers to tuple's posting list or
* creates posting tree and tuple pointed to tree in a case
* of not enough space. Max size of tuple is defined in
* GinFormTuple().
*/
static IndexTuple
addItemPointersToTuple(Relation index, GinState *ginstate,
GinBtreeStack *stack, IndexTuple old,
ItemPointerData *items, uint32 nitem,
GinStatsData *buildStats)
{
Datum key = gin_index_getattr(ginstate, old);
OffsetNumber attnum = gintuple_get_attrnum(ginstate, old);
IndexTuple res = GinFormTuple(index, ginstate, attnum, key,
NULL, nitem + GinGetNPosting(old),
false);
if (res)
{
/* good, small enough */
uint32 newnitem;
newnitem = ginMergeItemPointers(GinGetPosting(res),
GinGetPosting(old),
GinGetNPosting(old),
items, nitem);
/* merge might have eliminated some duplicate items */
GinShortenTuple(res, newnitem);
}
else
{
BlockNumber postingRoot;
GinPostingTreeScan *gdi;
/* posting list becomes big, so we need to make posting's tree */
res = GinFormTuple(index, ginstate, attnum, key, NULL, 0, true);
postingRoot = createPostingTree(index, GinGetPosting(old), GinGetNPosting(old));
GinSetPostingTree(res, postingRoot);
gdi = ginPrepareScanPostingTree(index, postingRoot, FALSE);
gdi->btree.isBuild = (buildStats != NULL);
ginInsertItemPointer(gdi, items, nitem, buildStats);
pfree(gdi);
/* During index build, count the newly-added data page */
if (buildStats)
buildStats->nDataPages++;
}
return res;
}
/*
* Adds array of item pointers to tuple's posting list, or
* creates posting tree and tuple pointing to tree in case
* of not enough space. Max size of tuple is defined in
* GinFormTuple(). Returns a new, modified index tuple.
* items[] must be in sorted order with no duplicates.
*/
static IndexTuple
addItemPointersToLeafTuple(GinState *ginstate,
IndexTuple old,
ItemPointerData *items, uint32 nitem,
GinStatsData *buildStats, Buffer buffer)
{
OffsetNumber attnum;
Datum key;
GinNullCategory category;
IndexTuple res;
ItemPointerData *newItems,
*oldItems;
int oldNPosting,
newNPosting;
GinPostingList *compressedList;
Assert(!GinIsPostingTree(old));
attnum = gintuple_get_attrnum(ginstate, old);
key = gintuple_get_key(ginstate, old, &category);
/* merge the old and new posting lists */
oldItems = ginReadTuple(ginstate, attnum, old, &oldNPosting);
newItems = ginMergeItemPointers(items, nitem,
oldItems, oldNPosting,
&newNPosting);
/* Compress the posting list, and try to a build tuple with room for it */
res = NULL;
compressedList = ginCompressPostingList(newItems, newNPosting, GinMaxItemSize,
NULL);
pfree(newItems);
if (compressedList)
{
res = GinFormTuple(ginstate, attnum, key, category,
(char *) compressedList,
SizeOfGinPostingList(compressedList),
newNPosting,
false);
pfree(compressedList);
}
if (!res)
{
/* posting list would be too big, convert to posting tree */
BlockNumber postingRoot;
/*
* Initialize posting tree with the old tuple's posting list. It's
* surely small enough to fit on one posting-tree page, and should
* already be in order with no duplicates.
*/
postingRoot = createPostingTree(ginstate->index,
oldItems,
oldNPosting,
buildStats,
buffer);
/* Now insert the TIDs-to-be-added into the posting tree */
ginInsertItemPointers(ginstate->index, postingRoot,
items, nitem,
buildStats);
/* And build a new posting-tree-only result tuple */
res = GinFormTuple(ginstate, attnum, key, category, NULL, 0, 0, true);
GinSetPostingTree(res, postingRoot);
}
pfree(oldItems);
return res;
}
/*
* Adds array of item pointers to tuple's posting list, or
* creates posting tree and tuple pointing to tree in case
* of not enough space. Max size of tuple is defined in
* GinFormTuple(). Returns a new, modified index tuple.
* items[] must be in sorted order with no duplicates.
*/
static IndexTuple
addItemPointersToLeafTuple(GinState *ginstate,
IndexTuple old,
ItemPointerData *items, uint32 nitem,
GinStatsData *buildStats)
{
OffsetNumber attnum;
Datum key;
GinNullCategory category;
IndexTuple res;
Assert(!GinIsPostingTree(old));
attnum = gintuple_get_attrnum(ginstate, old);
key = gintuple_get_key(ginstate, old, &category);
/* try to build tuple with room for all the items */
res = GinFormTuple(ginstate, attnum, key, category,
NULL, nitem + GinGetNPosting(old),
false);
if (res)
{
/* good, small enough */
uint32 newnitem;
/* fill in the posting list with union of old and new TIDs */
newnitem = ginMergeItemPointers(GinGetPosting(res),
GinGetPosting(old),
GinGetNPosting(old),
items, nitem);
/* merge might have eliminated some duplicate items */
GinShortenTuple(res, newnitem);
}
else
{
/* posting list would be too big, convert to posting tree */
BlockNumber postingRoot;
/*
* Initialize posting tree with the old tuple's posting list. It's
* surely small enough to fit on one posting-tree page, and should
* already be in order with no duplicates.
*/
postingRoot = createPostingTree(ginstate->index,
GinGetPosting(old),
GinGetNPosting(old),
buildStats);
/* Now insert the TIDs-to-be-added into the posting tree */
ginInsertItemPointers(ginstate->index, postingRoot,
items, nitem,
buildStats);
/* And build a new posting-tree-only result tuple */
res = GinFormTuple(ginstate, attnum, key, category, NULL, 0, true);
GinSetPostingTree(res, postingRoot);
}
return res;
}
static void
ginRedoRecompress(Page page, ginxlogRecompressDataLeaf *data)
{
int actionno;
int segno;
GinPostingList *oldseg;
Pointer segmentend;
char *walbuf;
int totalsize;
/*
* If the page is in pre-9.4 format, convert to new___ format first.
*/
if (!GinPageIsCompressed(page))
{
ItemPointer uncompressed = (ItemPointer) GinDataPageGetData(page);
int nuncompressed = GinPageGetOpaque(page)->maxoff;
int npacked;
GinPostingList *plist;
plist = ginCompressPostingList(uncompressed, nuncompressed,
BLCKSZ, &npacked);
Assert(npacked == nuncompressed);
totalsize = SizeOfGinPostingList(plist);
memcpy(GinDataLeafPageGetPostingList(page), plist, totalsize);
GinDataPageSetDataSize(page, totalsize);
GinPageSetCompressed(page);
GinPageGetOpaque(page)->maxoff = InvalidOffsetNumber;
}
oldseg = GinDataLeafPageGetPostingList(page);
segmentend = (Pointer) oldseg + GinDataLeafPageGetPostingListSize(page);
segno = 0;
walbuf = ((char *) data) + sizeof(ginxlogRecompressDataLeaf);
for (actionno = 0; actionno < data->nactions; actionno++)
{
uint8 a_segno = *((uint8 *) (walbuf++));
uint8 a_action = *((uint8 *) (walbuf++));
GinPostingList *newseg = NULL;
int newsegsize = 0;
ItemPointerData *items = NULL;
uint16 nitems = 0;
ItemPointerData *olditems;
int nolditems;
ItemPointerData *newitems;
int nnewitems;
int segsize;
Pointer segptr;
int szleft;
/* Extract all the information we need from the WAL record */
if (a_action == GIN_SEGMENT_INSERT ||
a_action == GIN_SEGMENT_REPLACE)
{
newseg = (GinPostingList *) walbuf;
newsegsize = SizeOfGinPostingList(newseg);
walbuf += SHORTALIGN(newsegsize);
}
if (a_action == GIN_SEGMENT_ADDITEMS)
{
memcpy(&nitems, walbuf, sizeof(uint16));
walbuf += sizeof(uint16);
items = (ItemPointerData *) walbuf;
walbuf += nitems * sizeof(ItemPointerData);
}
/* Skip to the segment that this action concerns */
Assert(segno <= a_segno);
while (segno < a_segno)
{
oldseg = GinNextPostingListSegment(oldseg);
segno++;
}
/*
* ADDITEMS action is handled like REPLACE, but the new___ segment to
* replace the old one is reconstructed using the old segment from
* disk and the new___ items from the WAL record.
*/
if (a_action == GIN_SEGMENT_ADDITEMS)
{
int npacked;
olditems = ginPostingListDecode(oldseg, &nolditems);
newitems = ginMergeItemPointers(items, nitems,
olditems, nolditems,
&nnewitems);
Assert(nnewitems == nolditems + nitems);
newseg = ginCompressPostingList(newitems, nnewitems,
BLCKSZ, &npacked);
Assert(npacked == nnewitems);
newsegsize = SizeOfGinPostingList(newseg);
a_action = GIN_SEGMENT_REPLACE;
}
segptr = (Pointer) oldseg;
if (segptr != segmentend)
segsize = SizeOfGinPostingList(oldseg);
else
{
/*
* Positioned after the last existing segment. Only INSERTs
* expected here.
*/
Assert(a_action == GIN_SEGMENT_INSERT);
segsize = 0;
}
szleft = segmentend - segptr;
switch (a_action)
{
case GIN_SEGMENT_DELETE:
memmove(segptr, segptr + segsize, szleft - segsize);
segmentend -= segsize;
segno++;
break;
case GIN_SEGMENT_INSERT:
/* make room for the new___ segment */
memmove(segptr + newsegsize, segptr, szleft);
/* copy the new___ segment in place */
memcpy(segptr, newseg, newsegsize);
segmentend += newsegsize;
segptr += newsegsize;
break;
case GIN_SEGMENT_REPLACE:
/* shift the segments that follow */
memmove(segptr + newsegsize,
segptr + segsize,
szleft - segsize);
/* copy the replacement segment in place */
memcpy(segptr, newseg, newsegsize);
segmentend -= segsize;
segmentend += newsegsize;
segptr += newsegsize;
segno++;
break;
default:
elog(ERROR, "unexpected GIN leaf action: %u", a_action);
}
oldseg = (GinPostingList *) segptr;
}
totalsize = segmentend - (Pointer) GinDataLeafPageGetPostingList(page);
GinDataPageSetDataSize(page, totalsize);
}
/*
* Redo recompression of posting list. Doing all the changes in-place is not
* always possible, because it might require more space than we've on the page.
* Instead, once modification is required we copy unprocessed tail of the page
* into separately allocated chunk of memory for further reading original
* versions of segments. Thanks to that we don't bother about moving page data
* in-place.
*/
static void
ginRedoRecompress(Page page, ginxlogRecompressDataLeaf *data)
{
int actionno;
int segno;
GinPostingList *oldseg;
Pointer segmentend;
char *walbuf;
int totalsize;
Pointer tailCopy = NULL;
Pointer writePtr;
Pointer segptr;
/*
* If the page is in pre-9.4 format, convert to new format first.
*/
if (!GinPageIsCompressed(page))
{
ItemPointer uncompressed = (ItemPointer) GinDataPageGetData(page);
int nuncompressed = GinPageGetOpaque(page)->maxoff;
int npacked;
/*
* Empty leaf pages are deleted as part of vacuum, but leftmost and
* rightmost pages are never deleted. So, pg_upgrade'd from pre-9.4
* instances might contain empty leaf pages, and we need to handle
* them correctly.
*/
if (nuncompressed > 0)
{
GinPostingList *plist;
plist = ginCompressPostingList(uncompressed, nuncompressed,
BLCKSZ, &npacked);
totalsize = SizeOfGinPostingList(plist);
Assert(npacked == nuncompressed);
memcpy(GinDataLeafPageGetPostingList(page), plist, totalsize);
}
else
{
totalsize = 0;
}
GinDataPageSetDataSize(page, totalsize);
GinPageSetCompressed(page);
GinPageGetOpaque(page)->maxoff = InvalidOffsetNumber;
}
oldseg = GinDataLeafPageGetPostingList(page);
writePtr = (Pointer) oldseg;
segmentend = (Pointer) oldseg + GinDataLeafPageGetPostingListSize(page);
segno = 0;
walbuf = ((char *) data) + sizeof(ginxlogRecompressDataLeaf);
for (actionno = 0; actionno < data->nactions; actionno++)
{
uint8 a_segno = *((uint8 *) (walbuf++));
uint8 a_action = *((uint8 *) (walbuf++));
GinPostingList *newseg = NULL;
int newsegsize = 0;
ItemPointerData *items = NULL;
uint16 nitems = 0;
ItemPointerData *olditems;
int nolditems;
ItemPointerData *newitems;
int nnewitems;
int segsize;
/* Extract all the information we need from the WAL record */
if (a_action == GIN_SEGMENT_INSERT ||
a_action == GIN_SEGMENT_REPLACE)
{
newseg = (GinPostingList *) walbuf;
newsegsize = SizeOfGinPostingList(newseg);
walbuf += SHORTALIGN(newsegsize);
}
if (a_action == GIN_SEGMENT_ADDITEMS)
{
memcpy(&nitems, walbuf, sizeof(uint16));
walbuf += sizeof(uint16);
items = (ItemPointerData *) walbuf;
walbuf += nitems * sizeof(ItemPointerData);
}
/* Skip to the segment that this action concerns */
Assert(segno <= a_segno);
while (segno < a_segno)
{
/*
* Once modification is started and page tail is copied, we've
* to copy unmodified segments.
*/
segsize = SizeOfGinPostingList(oldseg);
if (tailCopy)
{
Assert(writePtr + segsize < PageGetSpecialPointer(page));
memcpy(writePtr, (Pointer) oldseg, segsize);
}
writePtr += segsize;
oldseg = GinNextPostingListSegment(oldseg);
segno++;
}
/*
* ADDITEMS action is handled like REPLACE, but the new segment to
* replace the old one is reconstructed using the old segment from
* disk and the new items from the WAL record.
*/
if (a_action == GIN_SEGMENT_ADDITEMS)
{
int npacked;
olditems = ginPostingListDecode(oldseg, &nolditems);
newitems = ginMergeItemPointers(items, nitems,
olditems, nolditems,
&nnewitems);
Assert(nnewitems == nolditems + nitems);
newseg = ginCompressPostingList(newitems, nnewitems,
BLCKSZ, &npacked);
Assert(npacked == nnewitems);
newsegsize = SizeOfGinPostingList(newseg);
a_action = GIN_SEGMENT_REPLACE;
}
segptr = (Pointer) oldseg;
if (segptr != segmentend)
segsize = SizeOfGinPostingList(oldseg);
else
{
/*
* Positioned after the last existing segment. Only INSERTs
* expected here.
*/
Assert(a_action == GIN_SEGMENT_INSERT);
segsize = 0;
}
/*
* We're about to start modification of the page. So, copy tail of the
* page if it's not done already.
*/
if (!tailCopy && segptr != segmentend)
{
int tailSize = segmentend - segptr;
tailCopy = (Pointer) palloc(tailSize);
memcpy(tailCopy, segptr, tailSize);
segptr = tailCopy;
oldseg = (GinPostingList *) segptr;
segmentend = segptr + tailSize;
}
switch (a_action)
{
case GIN_SEGMENT_DELETE:
segptr += segsize;
segno++;
break;
case GIN_SEGMENT_INSERT:
/* copy the new segment in place */
Assert(writePtr + newsegsize <= PageGetSpecialPointer(page));
memcpy(writePtr, newseg, newsegsize);
writePtr += newsegsize;
break;
case GIN_SEGMENT_REPLACE:
/* copy the new version of segment in place */
Assert(writePtr + newsegsize <= PageGetSpecialPointer(page));
memcpy(writePtr, newseg, newsegsize);
writePtr += newsegsize;
segptr += segsize;
segno++;
break;
default:
elog(ERROR, "unexpected GIN leaf action: %u", a_action);
}
oldseg = (GinPostingList *) segptr;
}
/* Copy the rest of unmodified segments if any. */
segptr = (Pointer) oldseg;
if (segptr != segmentend && tailCopy)
{
int restSize = segmentend - segptr;
Assert(writePtr + restSize <= PageGetSpecialPointer(page));
memcpy(writePtr, segptr, restSize);
writePtr += restSize;
}
totalsize = writePtr - (Pointer) GinDataLeafPageGetPostingList(page);
GinDataPageSetDataSize(page, totalsize);
}