/*
* Read item pointers from leaf entry tuple.
*
* Returns a palloc'd array of ItemPointers. The number of items is returned
* in *nitems.
*/
ItemPointer
ginReadTuple(GinState *ginstate, OffsetNumber attnum, IndexTuple itup,
int *nitems)
{
Pointer ptr = GinGetPosting(itup);
int nipd = GinGetNPosting(itup);
ItemPointer ipd;
int ndecoded;
if (GinItupIsCompressed(itup))
{
if (nipd > 0)
{
ipd = ginPostingListDecode((GinPostingList *) ptr, &ndecoded);
if (nipd != ndecoded)
elog(ERROR, "number of items mismatch in GIN entry tuple, %d in tuple header, %d decoded",
nipd, ndecoded);
}
else
{
ipd = palloc(0);
}
}
else
{
ipd = (ItemPointer) palloc(sizeof(ItemPointerData) * nipd);
memcpy(ipd, ptr, sizeof(ItemPointerData) * nipd);
}
*nitems = nipd;
return ipd;
}
/*
* returns modified page or NULL if page isn't modified.
* Function works with original page until first change is occurred,
* then page is copied into temporary one.
*/
static Page
ginVacuumEntryPage(GinVacuumState *gvs, Buffer buffer, BlockNumber *roots, uint32 *nroot)
{
Page origpage = BufferGetPage(buffer),
tmppage;
OffsetNumber i,
maxoff = PageGetMaxOffsetNumber(origpage);
tmppage = origpage;
*nroot = 0;
for (i = FirstOffsetNumber; i <= maxoff; i++)
{
IndexTuple itup = (IndexTuple) PageGetItem(tmppage, PageGetItemId(tmppage, i));
if (GinIsPostingTree(itup))
{
/*
* store posting tree's roots for further processing, we can't
* vacuum it just now due to risk of deadlocks with scans/inserts
*/
roots[*nroot] = GinGetDownlink(itup);
(*nroot)++;
}
else if (GinGetNPosting(itup) > 0)
{
int nitems;
ItemPointer items_orig;
bool free_items_orig;
ItemPointer items;
/* Get list of item pointers from the tuple. */
if (GinItupIsCompressed(itup))
{
items_orig = ginPostingListDecode((GinPostingList *) GinGetPosting(itup), &nitems);
free_items_orig = true;
}
else
{
items_orig = (ItemPointer) GinGetPosting(itup);
nitems = GinGetNPosting(itup);
free_items_orig = false;
}
/* Remove any items from the list that need to be vacuumed. */
items = ginVacuumItemPointers(gvs, items_orig, nitems, &nitems);
if (free_items_orig)
pfree(items_orig);
/* If any item pointers were removed, recreate the tuple. */
if (items)
{
OffsetNumber attnum;
Datum key;
GinNullCategory category;
GinPostingList *plist;
int plistsize;
if (nitems > 0)
{
plist = ginCompressPostingList(items, nitems, GinMaxItemSize, NULL);
plistsize = SizeOfGinPostingList(plist);
}
else
{
plist = NULL;
plistsize = 0;
}
/*
* if we already created a temporary page, make changes in
* place
*/
if (tmppage == origpage)
{
/*
* On first difference, create a temporary copy of the
* page and copy the tuple's posting list to it.
*/
tmppage = PageGetTempPageCopy(origpage);
/* set itup pointer to new page */
itup = (IndexTuple) PageGetItem(tmppage, PageGetItemId(tmppage, i));
}
attnum = gintuple_get_attrnum(&gvs->ginstate, itup);
key = gintuple_get_key(&gvs->ginstate, itup, &category);
itup = GinFormTuple(&gvs->ginstate, attnum, key, category,
(char *) plist, plistsize,
nitems, true);
if (plist)
pfree(plist);
PageIndexTupleDelete(tmppage, i);
if (PageAddItem(tmppage, (Item) itup, IndexTupleSize(itup), i, false, false) != i)
elog(ERROR, "failed to add item to index page in \"%s\"",
RelationGetRelationName(gvs->index));
pfree(itup);
pfree(items);
}
}
}
return (tmppage == origpage) ? NULL : tmppage;
}
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);
}
/*
* Encode a posting list.
*
* The encoded list is returned in a palloc'd struct, which will be at most
* 'maxsize' bytes in size. The number items in the returned segment is
* returned in *nwritten. If it's not equal to nipd, not all the items fit
* in 'maxsize', and only the first *nwritten were encoded.
*
* The allocated size of the returned struct is short-aligned, and the padding
* byte at the end, if any, is zero.
*/
GinPostingList *
ginCompressPostingList(const ItemPointer ipd, int nipd, int maxsize,
int *nwritten)
{
uint64 prev;
int totalpacked = 0;
int maxbytes;
GinPostingList *result;
unsigned char *ptr;
unsigned char *endptr;
maxsize = SHORTALIGN_DOWN(maxsize);
result = palloc(maxsize);
maxbytes = maxsize - offsetof(GinPostingList, bytes);
Assert(maxbytes > 0);
/* Store the first special item */
result->first = ipd[0];
prev = itemptr_to_uint64(&result->first);
ptr = result->bytes;
endptr = result->bytes + maxbytes;
for (totalpacked = 1; totalpacked < nipd; totalpacked++)
{
uint64 val = itemptr_to_uint64(&ipd[totalpacked]);
uint64 delta = val - prev;
Assert(val > prev);
if (endptr - ptr >= 6)
encode_varbyte(delta, &ptr);
else
{
/*
* There are less than 6 bytes left. Have to check if the next
* item fits in that space before writing it out.
*/
unsigned char buf[6];
unsigned char *p = buf;
encode_varbyte(delta, &p);
if (p - buf > (endptr - ptr))
break; /* output is full */
memcpy(ptr, buf, p - buf);
ptr += (p - buf);
}
prev = val;
}
result->nbytes = ptr - result->bytes;
/*
* If we wrote an odd number of bytes, zero out the padding byte at the
* end.
*/
if (result->nbytes != SHORTALIGN(result->nbytes))
result->bytes[result->nbytes] = 0;
if (nwritten)
*nwritten = totalpacked;
Assert(SizeOfGinPostingList(result) <= maxsize);
/*
* Check that the encoded segment decodes back to the original items.
*/
#if defined (CHECK_ENCODING_ROUNDTRIP)
{
int ndecoded;
ItemPointer tmp = ginPostingListDecode(result, &ndecoded);
int i;
Assert(ndecoded == totalpacked);
for (i = 0; i < ndecoded; i++)
Assert(memcmp(&tmp[i], &ipd[i], sizeof(ItemPointerData)) == 0);
pfree(tmp);
}
#endif
return result;
}
Datum
gin_leafpage_items(PG_FUNCTION_ARGS)
{
bytea *raw_page = PG_GETARG_BYTEA_P(0);
int raw_page_size;
FuncCallContext *fctx;
gin_leafpage_items_state *inter_call_data;
if (!superuser())
ereport(ERROR,
(errcode(ERRCODE_INSUFFICIENT_PRIVILEGE),
(errmsg("must be superuser to use raw page functions"))));
raw_page_size = VARSIZE(raw_page) - VARHDRSZ;
if (SRF_IS_FIRSTCALL())
{
TupleDesc tupdesc;
MemoryContext mctx;
Page page;
GinPageOpaque opaq;
if (raw_page_size < BLCKSZ)
ereport(ERROR,
(errcode(ERRCODE_INVALID_PARAMETER_VALUE),
errmsg("input page too small (%d bytes)", raw_page_size)));
page = VARDATA(raw_page);
if (PageGetSpecialSize(page) != MAXALIGN(sizeof(GinPageOpaqueData)))
ereport(ERROR,
(errcode(ERRCODE_INVALID_PARAMETER_VALUE),
errmsg("input page is not a valid GIN data leaf page"),
errdetail("Special size %d, expected %d",
(int) PageGetSpecialSize(page),
(int) MAXALIGN(sizeof(GinPageOpaqueData)))));
opaq = (GinPageOpaque) PageGetSpecialPointer(page);
if (opaq->flags != (GIN_DATA | GIN_LEAF | GIN_COMPRESSED))
ereport(ERROR,
(errcode(ERRCODE_INVALID_PARAMETER_VALUE),
errmsg("input page is not a compressed GIN data leaf page"),
errdetail("Flags %04X, expected %04X",
opaq->flags,
(GIN_DATA | GIN_LEAF | GIN_COMPRESSED))));
fctx = SRF_FIRSTCALL_INIT();
mctx = MemoryContextSwitchTo(fctx->multi_call_memory_ctx);
inter_call_data = palloc(sizeof(gin_leafpage_items_state));
/* Build a tuple descriptor for our result type */
if (get_call_result_type(fcinfo, NULL, &tupdesc) != TYPEFUNC_COMPOSITE)
elog(ERROR, "return type must be a row type");
inter_call_data->tupd = tupdesc;
inter_call_data->seg = GinDataLeafPageGetPostingList(page);
inter_call_data->lastseg = (GinPostingList *)
(((char *) inter_call_data->seg) +
GinDataLeafPageGetPostingListSize(page));
fctx->user_fctx = inter_call_data;
MemoryContextSwitchTo(mctx);
}
fctx = SRF_PERCALL_SETUP();
inter_call_data = fctx->user_fctx;
if (inter_call_data->seg != inter_call_data->lastseg)
{
GinPostingList *cur = inter_call_data->seg;
HeapTuple resultTuple;
Datum result;
Datum values[3];
bool nulls[3];
int ndecoded,
i;
ItemPointer tids;
Datum *tids_datum;
memset(nulls, 0, sizeof(nulls));
values[0] = ItemPointerGetDatum(&cur->first);
values[1] = UInt16GetDatum(cur->nbytes);
/* build an array of decoded item pointers */
tids = ginPostingListDecode(cur, &ndecoded);
tids_datum = (Datum *) palloc(ndecoded * sizeof(Datum));
for (i = 0; i < ndecoded; i++)
tids_datum[i] = ItemPointerGetDatum(&tids[i]);
values[2] = PointerGetDatum(construct_array(tids_datum,
ndecoded,
TIDOID,
sizeof(ItemPointerData),
false, 's'));
pfree(tids_datum);
pfree(tids);
/* Build and return the result tuple. */
resultTuple = heap_form_tuple(inter_call_data->tupd, values, nulls);
result = HeapTupleGetDatum(resultTuple);
inter_call_data->seg = GinNextPostingListSegment(cur);
SRF_RETURN_NEXT(fctx, result);
}
else
SRF_RETURN_DONE(fctx);
}
