open-dsi
/
postgres
mirror of https://github.com/postgres/postgres
1
0
Fork 0
Code Issues Packages Projects Releases Wiki Activity

Reduce PANIC to ERROR in some occasionally-reported btree failure cases.

Browse Source 
This patch changes _bt_split() and _bt_pagedel() to throw a plain ERROR,
rather than PANIC, for several cases that are reported from the field
from time to time:
* right sibling's left-link doesn't match;
* PageAddItem failure during _bt_split();
* parent page's next child isn't right sibling during _bt_pagedel().
In addition the error messages for these cases have been made a bit
more verbose, with additional values included.

The original motivation for PANIC here was to capture core dumps for
subsequent analysis.  But with so many users whose platforms don't capture
core dumps by default, or who are unprepared to analyze them anyway, it's hard
to justify a forced database restart when we can fairly easily detect the
problems before we've reached the critical sections where PANIC would be
necessary.  It is not currently known whether the reports of these messages
indicate well-hidden bugs in Postgres, or are a result of storage-level
malfeasance; the latter possibility suggests that we ought to try to be more
robust even if there is a bug here that's ultimately found.

Backpatch to 8.2.  The code before that is sufficiently different that
it doesn't seem worth the trouble to back-port further.
REL8_4_STABLE
Tom Lane 16 years ago
parent d321ff10c1
commit 27fddbadc4
2 changed files with 135 additions and 52 deletions
Show all changes Ignore whitespace when comparing lines Ignore changes in amount of whitespace Ignore changes in whitespace at EOL
Show Stats Download Patch File Download Diff File
  1. 133
      src/backend/access/nbtree/nbtinsert.c
  2. 54
      src/backend/access/nbtree/nbtpage.c

133
src/backend/access/nbtree/nbtinsert.c
Unescape View File

@ -8,7 +8,7 @@
*
*
* IDENTIFICATION
* $PostgreSQL: pgsql/src/backend/access/nbtree/nbtinsert.c,v 1.170.2.1 2009/10/02 21:14:11 tgl Exp $
* $PostgreSQL: pgsql/src/backend/access/nbtree/nbtinsert.c,v 1.170.2.2 2010/08/29 19:33:29 tgl Exp $
*
*-------------------------------------------------------------------------
*/
@ -72,9 +72,8 @@ static OffsetNumber _bt_findsplitloc(Relation rel, Page page,
static void _bt_checksplitloc(FindSplitData *state,
OffsetNumber firstoldonright, bool newitemonleft,
int dataitemstoleft, Size firstoldonrightsz);
static void _bt_pgaddtup(Relation rel, Page page,
Size itemsize, IndexTuple itup,
OffsetNumber itup_off, const char *where);
static bool _bt_pgaddtup(Page page, Size itemsize, IndexTuple itup,
OffsetNumber itup_off);
static bool _bt_isequal(TupleDesc itupdesc, Page page, OffsetNumber offnum,
int keysz, ScanKey scankey);
static void _bt_vacuum_one_page(Relation rel, Buffer buffer);
@ -651,7 +650,9 @@ _bt_insertonpg(Relation rel,
/* Do the update. No ereport(ERROR) until changes are logged */
START_CRIT_SECTION();
_bt_pgaddtup(rel, page, itemsz, itup, newitemoff, "page");
if (!_bt_pgaddtup(page, itemsz, itup, newitemoff))
elog(PANIC, "failed to add new item to block %u in index \"%s\"",
itup_blkno, RelationGetRelationName(rel));
MarkBufferDirty(buf);
@ -777,6 +778,8 @@ _bt_split(Relation rel, Buffer buf, OffsetNumber firstright,
Page origpage;
Page leftpage,
rightpage;
BlockNumber origpagenumber,
rightpagenumber;
BTPageOpaque ropaque,
lopaque,
oopaque;
@ -792,11 +795,27 @@ _bt_split(Relation rel, Buffer buf, OffsetNumber firstright,
OffsetNumber i;
bool isroot;
/* Acquire a new page to split into */
rbuf = _bt_getbuf(rel, P_NEW, BT_WRITE);
/*
* origpage is the original page to be split. leftpage is a temporary
* buffer that receives the left-sibling data, which will be copied back
* into origpage on success. rightpage is the new page that receives
* the right-sibling data. If we fail before reaching the critical
* section, origpage hasn't been modified and leftpage is only workspace.
* In principle we shouldn't need to worry about rightpage either,
* because it hasn't been linked into the btree page structure; but to
* avoid leaving possibly-confusing junk behind, we are careful to rewrite
* rightpage as zeroes before throwing any error.
*/
origpage = BufferGetPage(buf);
leftpage = PageGetTempPage(origpage);
rightpage = BufferGetPage(rbuf);
origpagenumber = BufferGetBlockNumber(buf);
rightpagenumber = BufferGetBlockNumber(rbuf);
_bt_pageinit(leftpage, BufferGetPageSize(buf));
/* rightpage was already initialized by _bt_getbuf */
@ -821,8 +840,8 @@ _bt_split(Relation rel, Buffer buf, OffsetNumber firstright,
lopaque->btpo_flags &= ~(BTP_ROOT | BTP_SPLIT_END | BTP_HAS_GARBAGE);
ropaque->btpo_flags = lopaque->btpo_flags;
lopaque->btpo_prev = oopaque->btpo_prev;
lopaque->btpo_next = BufferGetBlockNumber(rbuf);
ropaque->btpo_prev = BufferGetBlockNumber(buf);
lopaque->btpo_next = rightpagenumber;
ropaque->btpo_prev = origpagenumber;
ropaque->btpo_next = oopaque->btpo_next;
lopaque->btpo.level = ropaque->btpo.level = oopaque->btpo.level;
/* Since we already have write-lock on both pages, ok to read cycleid */
@ -845,9 +864,12 @@ _bt_split(Relation rel, Buffer buf, OffsetNumber firstright,
item = (IndexTuple) PageGetItem(origpage, itemid);
if (PageAddItem(rightpage, (Item) item, itemsz, rightoff,
false, false) == InvalidOffsetNumber)
elog(PANIC, "failed to add hikey to the right sibling"
{
memset(rightpage, 0, BufferGetPageSize(rbuf));
elog(ERROR, "failed to add hikey to the right sibling"
" while splitting block %u of index \"%s\"",
BufferGetBlockNumber(buf), RelationGetRelationName(rel));
origpagenumber, RelationGetRelationName(rel));
}
rightoff = OffsetNumberNext(rightoff);
}
@ -872,9 +894,12 @@ _bt_split(Relation rel, Buffer buf, OffsetNumber firstright,
}
if (PageAddItem(leftpage, (Item) item, itemsz, leftoff,
false, false) == InvalidOffsetNumber)
elog(PANIC, "failed to add hikey to the left sibling"
{
memset(rightpage, 0, BufferGetPageSize(rbuf));
elog(ERROR, "failed to add hikey to the left sibling"
" while splitting block %u of index \"%s\"",
BufferGetBlockNumber(buf), RelationGetRelationName(rel));
origpagenumber, RelationGetRelationName(rel));
}
leftoff = OffsetNumberNext(leftoff);
/*
@ -896,14 +921,24 @@ _bt_split(Relation rel, Buffer buf, OffsetNumber firstright,
{
if (newitemonleft)
{
_bt_pgaddtup(rel, leftpage, newitemsz, newitem, leftoff,
"left sibling");
if (!_bt_pgaddtup(leftpage, newitemsz, newitem, leftoff))
{
memset(rightpage, 0, BufferGetPageSize(rbuf));
elog(ERROR, "failed to add new item to the left sibling"
" while splitting block %u of index \"%s\"",
origpagenumber, RelationGetRelationName(rel));
}
leftoff = OffsetNumberNext(leftoff);
}
else
{
_bt_pgaddtup(rel, rightpage, newitemsz, newitem, rightoff,
"right sibling");
if (!_bt_pgaddtup(rightpage, newitemsz, newitem, rightoff))
{
memset(rightpage, 0, BufferGetPageSize(rbuf));
elog(ERROR, "failed to add new item to the right sibling"
" while splitting block %u of index \"%s\"",
origpagenumber, RelationGetRelationName(rel));
}
rightoff = OffsetNumberNext(rightoff);
}
}
@ -911,14 +946,24 @@ _bt_split(Relation rel, Buffer buf, OffsetNumber firstright,
/* decide which page to put it on */
if (i < firstright)
{
_bt_pgaddtup(rel, leftpage, itemsz, item, leftoff,
"left sibling");
if (!_bt_pgaddtup(leftpage, itemsz, item, leftoff))
{
memset(rightpage, 0, BufferGetPageSize(rbuf));
elog(ERROR, "failed to add old item to the left sibling"
" while splitting block %u of index \"%s\"",
origpagenumber, RelationGetRelationName(rel));
}
leftoff = OffsetNumberNext(leftoff);
}
else
{
_bt_pgaddtup(rel, rightpage, itemsz, item, rightoff,
"right sibling");
if (!_bt_pgaddtup(rightpage, itemsz, item, rightoff))
{
memset(rightpage, 0, BufferGetPageSize(rbuf));
elog(ERROR, "failed to add old item to the right sibling"
" while splitting block %u of index \"%s\"",
origpagenumber, RelationGetRelationName(rel));
}
rightoff = OffsetNumberNext(rightoff);
}
}
@ -932,8 +977,13 @@ _bt_split(Relation rel, Buffer buf, OffsetNumber firstright,
* not be splitting the page).
*/
Assert(!newitemonleft);
_bt_pgaddtup(rel, rightpage, newitemsz, newitem, rightoff,
"right sibling");
if (!_bt_pgaddtup(rightpage, newitemsz, newitem, rightoff))
{
memset(rightpage, 0, BufferGetPageSize(rbuf));
elog(ERROR, "failed to add new item to the right sibling"
" while splitting block %u of index \"%s\"",
origpagenumber, RelationGetRelationName(rel));
}
rightoff = OffsetNumberNext(rightoff);
}
@ -945,16 +995,19 @@ _bt_split(Relation rel, Buffer buf, OffsetNumber firstright,
* neighbors.
*/
if (!P_RIGHTMOST(ropaque))
if (!P_RIGHTMOST(oopaque))
{
sbuf = _bt_getbuf(rel, ropaque->btpo_next, BT_WRITE);
sbuf = _bt_getbuf(rel, oopaque->btpo_next, BT_WRITE);
spage = BufferGetPage(sbuf);
sopaque = (BTPageOpaque) PageGetSpecialPointer(spage);
if (sopaque->btpo_prev != ropaque->btpo_prev)
elog(PANIC, "right sibling's left-link doesn't match: "
"block %u links to %u instead of expected %u in index \"%s\"",
ropaque->btpo_next, sopaque->btpo_prev, ropaque->btpo_prev,
if (sopaque->btpo_prev != origpagenumber)
{
memset(rightpage, 0, BufferGetPageSize(rbuf));
elog(ERROR, "right sibling's left-link doesn't match: "
"block %u links to %u instead of expected %u in index \"%s\"",
oopaque->btpo_next, sopaque->btpo_prev, origpagenumber,
RelationGetRelationName(rel));
}
/*
* Check to see if we can set the SPLIT_END flag in the right-hand
@ -979,8 +1032,7 @@ _bt_split(Relation rel, Buffer buf, OffsetNumber firstright,
*
* NO EREPORT(ERROR) till right sibling is updated. We can get away with
* not starting the critical section till here because we haven't been
* scribbling on the original page yet, and we don't care about the new
* sibling until it's linked into the btree.
* scribbling on the original page yet; see comments above.
*/
START_CRIT_SECTION();
@ -992,19 +1044,21 @@ _bt_split(Relation rel, Buffer buf, OffsetNumber firstright,
* (in the page management code) that the center of a page always be
* clean, and the most efficient way to guarantee this is just to compact
* the data by reinserting it into a new left page. (XXX the latter
* comment is probably obsolete.)
* comment is probably obsolete; but in any case it's good to not scribble
* on the original page until we enter the critical section.)
*
* We need to do this before writing the WAL record, so that XLogInsert
* can WAL log an image of the page if necessary.
*/
PageRestoreTempPage(leftpage, origpage);
/* leftpage, lopaque must not be used below here */
MarkBufferDirty(buf);
MarkBufferDirty(rbuf);
if (!P_RIGHTMOST(ropaque))
{
sopaque->btpo_prev = BufferGetBlockNumber(rbuf);
sopaque->btpo_prev = rightpagenumber;
MarkBufferDirty(sbuf);
}
@ -1018,8 +1072,8 @@ _bt_split(Relation rel, Buffer buf, OffsetNumber firstright,
XLogRecData *lastrdata;
xlrec.node = rel->rd_node;
xlrec.leftsib = BufferGetBlockNumber(buf);
xlrec.rightsib = BufferGetBlockNumber(rbuf);
xlrec.leftsib = origpagenumber;
xlrec.rightsib = rightpagenumber;
xlrec.rnext = ropaque->btpo_next;
xlrec.level = ropaque->btpo.level;
xlrec.firstright = firstright;
@ -1818,13 +1872,11 @@ _bt_newroot(Relation rel, Buffer lbuf, Buffer rbuf)
* we insert the tuples in order, so that the given itup_off does
* represent the final position of the tuple!
*/
static void
_bt_pgaddtup(Relation rel,
Page page,
static bool
_bt_pgaddtup(Page page,
Size itemsize,
IndexTuple itup,
OffsetNumber itup_off,
const char *where)
OffsetNumber itup_off)
{
BTPageOpaque opaque = (BTPageOpaque) PageGetSpecialPointer(page);
IndexTupleData trunctuple;
@ -1839,8 +1891,9 @@ _bt_pgaddtup(Relation rel,
if (PageAddItem(page, (Item) itup, itemsize, itup_off,
false, false) == InvalidOffsetNumber)
elog(PANIC, "failed to add item to the %s in index \"%s\"",
where, RelationGetRelationName(rel));
return false;
return true;
}
/*

54
src/backend/access/nbtree/nbtpage.c
Unescape View File

@ -9,7 +9,7 @@
*
*
* IDENTIFICATION
* $PostgreSQL: pgsql/src/backend/access/nbtree/nbtpage.c,v 1.113 2009/05/05 19:02:22 tgl Exp $
* $PostgreSQL: pgsql/src/backend/access/nbtree/nbtpage.c,v 1.113.2.1 2010/08/29 19:33:29 tgl Exp $
*
* NOTES
* Postgres btree pages look like ordinary relation pages. The opaque
@ -1033,6 +1033,13 @@ _bt_pagedel(Relation rel, Buffer buf, BTStack stack, bool vacuum_full)
*/
rightsib = opaque->btpo_next;
rbuf = _bt_getbuf(rel, rightsib, BT_WRITE);
page = BufferGetPage(rbuf);
opaque = (BTPageOpaque) PageGetSpecialPointer(page);
if (opaque->btpo_prev != target)
elog(ERROR, "right sibling's left-link doesn't match: "
"block %u links to %u instead of expected %u in index \"%s\"",
rightsib, opaque->btpo_prev, target,
RelationGetRelationName(rel));
/*
* Next find and write-lock the current parent of the target page. This is
@ -1110,6 +1117,38 @@ _bt_pagedel(Relation rel, Buffer buf, BTStack stack, bool vacuum_full)
}
}
/*
* Check that the parent-page index items we're about to delete/overwrite
* contain what we expect. This can fail if the index has become
* corrupt for some reason. We want to throw any error before entering
* the critical section --- otherwise it'd be a PANIC.
*
* The test on the target item is just an Assert because _bt_getstackbuf
* should have guaranteed it has the expected contents. The test on the
* next-child downlink is known to sometimes fail in the field, though.
*/
page = BufferGetPage(pbuf);
opaque = (BTPageOpaque) PageGetSpecialPointer(page);
#ifdef USE_ASSERT_CHECKING
itemid = PageGetItemId(page, poffset);
itup = (IndexTuple) PageGetItem(page, itemid);
Assert(ItemPointerGetBlockNumber(&(itup->t_tid)) == target);
#endif
if (!parent_half_dead)
{
OffsetNumber nextoffset;
nextoffset = OffsetNumberNext(poffset);
itemid = PageGetItemId(page, nextoffset);
itup = (IndexTuple) PageGetItem(page, itemid);
if (ItemPointerGetBlockNumber(&(itup->t_tid)) != rightsib)
elog(ERROR, "right sibling %u of block %u is not next child %u of block %u in index \"%s\"",
rightsib, target, ItemPointerGetBlockNumber(&(itup->t_tid)),
parent, RelationGetRelationName(rel));
}
/*
* Here we begin doing the deletion.
*/
@ -1123,8 +1162,6 @@ _bt_pagedel(Relation rel, Buffer buf, BTStack stack, bool vacuum_full)
* to copy the right sibling's downlink over the target downlink, and then
* delete the following item.
*/
page = BufferGetPage(pbuf);
opaque = (BTPageOpaque) PageGetSpecialPointer(page);
if (parent_half_dead)
{
PageIndexTupleDelete(page, poffset);
@ -1136,23 +1173,16 @@ _bt_pagedel(Relation rel, Buffer buf, BTStack stack, bool vacuum_full)
itemid = PageGetItemId(page, poffset);
itup = (IndexTuple) PageGetItem(page, itemid);
Assert(ItemPointerGetBlockNumber(&(itup->t_tid)) == target);
ItemPointerSet(&(itup->t_tid), rightsib, P_HIKEY);
nextoffset = OffsetNumberNext(poffset);
/* This part is just for double-checking */
itemid = PageGetItemId(page, nextoffset);
itup = (IndexTuple) PageGetItem(page, itemid);
if (ItemPointerGetBlockNumber(&(itup->t_tid)) != rightsib)
elog(PANIC, "right sibling %u of block %u is not next child of %u in index \"%s\"",
rightsib, target, BufferGetBlockNumber(pbuf),
RelationGetRelationName(rel));
PageIndexTupleDelete(page, nextoffset);
}
/*
* Update siblings' side-links. Note the target page's side-links will
* continue to point to the siblings.
* continue to point to the siblings. Asserts here are just rechecking
* things we already verified above.
*/
if (BufferIsValid(lbuf))
{

Write Preview
Loading…
Cancel
Save