opensourcemirror
/
postgresql
mirror of git://git.postgresql.org/git/postgresql.git
1
0
Fork 0
Code Issues Releases Wiki Activity

Fix tuple chain moving bug found by "Hiroshi Inoue" <Inoue@tpf.co.jp>.

This commit is contained in:
Vadim B. Mikheev 1999-05-23 09:10:24 +00:00
parent b14c99d8d6
commit 7e14593d2e
1 changed files with 105 additions and 17 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

122
src/backend/commands/vacuum.c
View File

@ -7,7 +7,7 @@
*
*
* IDENTIFICATION
* $Header: /cvsroot/pgsql/src/backend/commands/vacuum.c,v 1.102 1999/05/10 00:44:59 momjian Exp $
* $Header: /cvsroot/pgsql/src/backend/commands/vacuum.c,v 1.103 1999/05/23 09:10:24 vadim Exp $
*
*-------------------------------------------------------------------------
*/
@ -87,7 +87,7 @@ static void vc_scanheap(VRelStats *vacrelstats, Relation onerel, VPageList vacuu
static void vc_rpfheap(VRelStats *vacrelstats, Relation onerel, VPageList vacuum_pages, VPageList fraged_pages, int nindices, Relation *Irel);
static void vc_vacheap(VRelStats *vacrelstats, Relation onerel, VPageList vpl);
static void vc_vacpage(Page page, VPageDescr vpd);
static void vc_vaconeind(VPageList vpl, Relation indrel, int num_tuples);
static void vc_vaconeind(VPageList vpl, Relation indrel, int num_tuples, int keep_tuples);
static void vc_scanoneind(Relation indrel, int num_tuples);
static void vc_attrstats(Relation onerel, VRelStats *vacrelstats, HeapTuple tuple);
static void vc_bucketcpy(Form_pg_attribute attr, Datum value, Datum *bucket, int16 *bucket_len);
@ -541,7 +541,7 @@ vc_vacone(Oid relid, bool analyze, List *va_cols)
if (vacuum_pages.vpl_num_pages > 0)
{
for (i = 0; i < nindices; i++)
vc_vaconeind(&vacuum_pages, Irel[i], vacrelstats->num_tuples);
vc_vaconeind(&vacuum_pages, Irel[i], vacrelstats->num_tuples, 0);
}
else
/* just scan indices to update statistic */
@ -1042,9 +1042,11 @@ vc_rpfheap(VRelStats *vacrelstats, Relation onerel,
num_fraged_pages,
vacuumed_pages;
int checked_moved,
num_tuples;
num_tuples,
keep_tuples = 0;
bool isempty,
dowrite;
dowrite,
chain_tuple_moved;
struct rusage ru0,
ru1;
@ -1126,6 +1128,7 @@ vc_rpfheap(VRelStats *vacrelstats, Relation onerel,
else
Assert(!isempty);
chain_tuple_moved = false; /* no one chain-tuple was moved off this page, yet */
vpc->vpd_blkno = blkno;
maxoff = PageGetMaxOffsetNumber(page);
for (offnum = FirstOffsetNumber;
@ -1145,11 +1148,39 @@ vc_rpfheap(VRelStats *vacrelstats, Relation onerel,
{
if ((TransactionId)tuple.t_data->t_cmin != myXID)
elog(ERROR, "Invalid XID in t_cmin");
if (tuple.t_data->t_infomask & HEAP_MOVED_OFF)
continue; /* already removed by me */
if (tuple.t_data->t_infomask & HEAP_MOVED_IN)
break;
elog(ERROR, "HEAP_MOVED_OFF/HEAP_MOVED_IN was expected");
elog(ERROR, "HEAP_MOVED_IN was not expected");
/*
* If this (chain) tuple is moved by me already then
* I have to check is it in vpc or not - i.e. is it
* moved while cleaning this page or some previous one.
*/
if (tuple.t_data->t_infomask & HEAP_MOVED_OFF)
{
if (keep_tuples == 0)
continue;
if (chain_tuple_moved) /* some chains was moved while */
{ /* cleaning this page */
Assert(vpc->vpd_offsets_free > 0);
for (i = 0; i < vpc->vpd_offsets_free; i++)
{
if (vpc->vpd_offsets[i] == offnum)
break;
}
if (i >= vpc->vpd_offsets_free) /* not found */
{
vpc->vpd_offsets[vpc->vpd_offsets_free++] = offnum;
keep_tuples--;
}
}
else
{
vpc->vpd_offsets[vpc->vpd_offsets_free++] = offnum;
keep_tuples--;
}
continue;
}
elog(ERROR, "HEAP_MOVED_OFF was expected");
}
/*
@ -1386,9 +1417,15 @@ moving chain: failed to add item with len = %u to page %u",
tuple.t_data->t_infomask |= HEAP_MOVED_OFF;
num_moved++;
/*
* Remember that we moved tuple from the current page
* (corresponding index tuple will be cleaned).
*/
if (Cbuf == buf)
vpc->vpd_offsets[vpc->vpd_offsets_free++] =
ItemPointerGetOffsetNumber(&(tuple.t_self));
else
keep_tuples++;
if (Irel != (Relation *) NULL)
{
@ -1418,6 +1455,7 @@ moving chain: failed to add item with len = %u to page %u",
}
cur_buffer = InvalidBuffer;
pfree(vtmove);
chain_tuple_moved = true;
continue;
}
@ -1532,10 +1570,58 @@ failed to add item with len = %u to page %u (free space %u, nusd %u, noff %u)",
} /* walk along page */
if (offnum < maxoff && keep_tuples > 0)
{
OffsetNumber off;
for (off = OffsetNumberNext(offnum);
off <= maxoff;
off = OffsetNumberNext(off))
{
itemid = PageGetItemId(page, off);
if (!ItemIdIsUsed(itemid))
continue;
tuple.t_data = (HeapTupleHeader) PageGetItem(page, itemid);
if (tuple.t_data->t_infomask & HEAP_XMIN_COMMITTED)
continue;
if ((TransactionId)tuple.t_data->t_cmin != myXID)
elog(ERROR, "Invalid XID in t_cmin (4)");
if (tuple.t_data->t_infomask & HEAP_MOVED_IN)
elog(ERROR, "HEAP_MOVED_IN was not expected (2)");
if (tuple.t_data->t_infomask & HEAP_MOVED_OFF)
{
if (chain_tuple_moved) /* some chains was moved while */
{ /* cleaning this page */
Assert(vpc->vpd_offsets_free > 0);
for (i = 0; i < vpc->vpd_offsets_free; i++)
{
if (vpc->vpd_offsets[i] == off)
break;
}
if (i >= vpc->vpd_offsets_free) /* not found */
{
vpc->vpd_offsets[vpc->vpd_offsets_free++] = off;
Assert(keep_tuples > 0);
keep_tuples--;
}
}
else
{
vpc->vpd_offsets[vpc->vpd_offsets_free++] = off;
Assert(keep_tuples > 0);
keep_tuples--;
}
}
}
}
if (vpc->vpd_offsets_free > 0) /* some tuples were moved */
{
qsort((char *) (vpc->vpd_offsets), vpc->vpd_offsets_free,
if (chain_tuple_moved) /* else - they are ordered */
{
qsort((char *) (vpc->vpd_offsets), vpc->vpd_offsets_free,
sizeof(OffsetNumber), vc_cmp_offno);
}
vc_reappage(&Nvpl, vpc);
WriteBuffer(buf);
}
@ -1559,7 +1645,6 @@ failed to add item with len = %u to page %u (free space %u, nusd %u, noff %u)",
if (num_moved > 0)
{
/*
* We have to commit our tuple' movings before we'll truncate
* relation, but we shouldn't lose our locks. And so - quick hack:
@ -1610,7 +1695,7 @@ failed to add item with len = %u to page %u (free space %u, nusd %u, noff %u)",
else if (tuple.t_data->t_infomask & HEAP_MOVED_OFF)
tuple.t_data->t_infomask |= HEAP_XMIN_INVALID;
else
elog(ERROR, "HEAP_MOVED_OFF/HEAP_MOVED_IN was expected (2)");
elog(ERROR, "HEAP_MOVED_OFF/HEAP_MOVED_IN was expected");
}
}
Assert((*vpp)->vpd_offsets_used == num_tuples);
@ -1647,8 +1732,10 @@ Elapsed %u/%u sec.",
*vpleft = *vpright;
*vpright = vpsave;
}
Assert(keep_tuples >= 0);
for (i = 0; i < nindices; i++)
vc_vaconeind(&Nvpl, Irel[i], vacrelstats->num_tuples);
vc_vaconeind(&Nvpl, Irel[i],
vacrelstats->num_tuples, keep_tuples);
}
/*
@ -1678,7 +1765,7 @@ Elapsed %u/%u sec.",
num_tuples++;
}
else
elog(ERROR, "HEAP_MOVED_OFF was expected");
elog(ERROR, "HEAP_MOVED_OFF was expected (2)");
}
}
@ -1854,7 +1941,7 @@ vc_scanoneind(Relation indrel, int num_tuples)
* pg_class.
*/
static void
vc_vaconeind(VPageList vpl, Relation indrel, int num_tuples)
vc_vaconeind(VPageList vpl, Relation indrel, int num_tuples, int keep_tuples)
{
RetrieveIndexResult res;
IndexScanDesc iscan;
@ -1911,11 +1998,12 @@ vc_vaconeind(VPageList vpl, Relation indrel, int num_tuples)
getrusage(RUSAGE_SELF, &ru1);
elog(MESSAGE_LEVEL, "Index %s: Pages %u; Tuples %u: Deleted %u. Elapsed %u/%u sec.",
indrel->rd_rel->relname.data, num_pages, num_index_tuples, tups_vacuumed,
indrel->rd_rel->relname.data, num_pages,
num_index_tuples - keep_tuples, tups_vacuumed,
ru1.ru_stime.tv_sec - ru0.ru_stime.tv_sec,
ru1.ru_utime.tv_sec - ru0.ru_utime.tv_sec);
if (num_index_tuples != num_tuples)
if (num_index_tuples != num_tuples + keep_tuples)
elog(NOTICE, "Index %s: NUMBER OF INDEX' TUPLES (%u) IS NOT THE SAME AS HEAP' (%u)",
indrel->rd_rel->relname.data, num_index_tuples, num_tuples);