rdelaage
/
postgresql
mirror of https://git.postgresql.org/git/postgresql.git
1
0
Fork 0
Code Issues Packages Projects Releases Wiki Activity

Fix race condition in GIN posting tree page deletion. 

If a page is deleted, and reused for something else, just as a search is
following a rightlink to it from its left sibling, the search would continue
scanning whatever the new contents of the page are. That could lead to
incorrect query results, or even something more curious if the page is
reused for a different kind of a page.

To fix, modify the search algorithm to lock the next page before releasing
the previous one, and refrain from deleting pages from the leftmost branch
of the tree.

Add a new Concurrency section to the README, explaining why this works.
There is a lot more one could say about concurrency in GIN, but that's for
another patch.

Backpatch to all supported versions.
This commit is contained in:
Heikki Linnakangas 2013-11-08 22:21:42 +02:00
parent 636b868f17
commit ac4ab97ec0
5 changed files with 121 additions and 59 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

50
src/backend/access/gin/README
View File

@ -210,6 +210,56 @@ fit on one pending-list page must have those pages to itself, even if this
results in wasting much of the space on the preceding page and the last
page for the tuple.)
Concurrency
-----------
The entry tree and each posting tree is a B-tree, with right-links connecting
sibling pages at the same level. This is the same structure that is used in
the regular B-tree indexam (invented by Lehman & Yao), but we don't support
scanning a GIN trees backwards, so we don't need left-links.
To avoid deadlocks, B-tree pages must always be locked in the same order:
left to right, and bottom to top. When searching, the tree is traversed from
top to bottom, so the lock on the parent page must be released before
descending to the next level. Concurrent page splits move the keyspace to
right, so after following a downlink, the page actually containing the key
we're looking for might be somewhere to the right of the page we landed on.
In that case, we follow the right-links until we find the page we're looking
for.
To delete a page, the page's left sibling, the target page, and its parent,
are locked in that order, and the page is marked as deleted. However, a
concurrent search might already have read a pointer to the page, and might be
just about to follow it. A page can be reached via the right-link of its left
sibling, or via its downlink in the parent.
To prevent a backend from reaching a deleted page via a right-link, when
following a right-link the lock on the previous page is not released until
the lock on next page has been acquired.
The downlink is more tricky. A search descending the tree must release the
lock on the parent page before locking the child, or it could deadlock with
a concurrent split of the child page; a page split locks the parent, while
already holding a lock on the child page. However, posting trees are only
fully searched from left to right, starting from the leftmost leaf. (The
tree-structure is only needed by insertions, to quickly find the correct
insert location). So as long as we don't delete the leftmost page on each
level, a search can never follow a downlink to page that's about to be
deleted.
The previous paragraph's reasoning only applies to searches, and only to
posting trees. To protect from inserters following a downlink to a deleted
page, vacuum simply locks out all concurrent insertions to the posting tree,
by holding a super-exclusive lock on the posting tree root. Inserters hold a
pin on the root page, but searches do not, so while new searches cannot begin
while root page is locked, any already-in-progress scans can continue
concurrently with vacuum. In the entry tree, we never delete pages.
(This is quite different from the mechanism the btree indexam uses to make
page-deletions safe; it stamps the deleted pages with an XID and keeps the
deleted pages around with the right-link intact until all concurrent scans
have finished.)
Limitations
-----------

53
src/backend/access/gin/ginbtree.c
View File

@ -112,10 +112,8 @@ ginFindLeafPage(GinBtree btree, GinBtreeStack *stack)
/* rightmost page */
break;
stack->buffer = ginStepRight(stack->buffer, btree->index, access);
stack->blkno = rightlink;
LockBuffer(stack->buffer, GIN_UNLOCK);
stack->buffer = ReleaseAndReadBuffer(stack->buffer, btree->index, stack->blkno);
LockBuffer(stack->buffer, access);
page = BufferGetPage(stack->buffer);
}
@ -148,6 +146,41 @@ ginFindLeafPage(GinBtree btree, GinBtreeStack *stack)
}
}
/*
* Step right from current page.
*
* The next page is locked first, before releasing the current page. This is
* crucial to protect from concurrent page deletion (see comment in
* ginDeletePage).
*/
Buffer
ginStepRight(Buffer buffer, Relation index, int lockmode)
{
Buffer nextbuffer;
Page page = BufferGetPage(buffer);
bool isLeaf = GinPageIsLeaf(page);
bool isData = GinPageIsData(page);
BlockNumber blkno = GinPageGetOpaque(page)->rightlink;
nextbuffer = ReadBuffer(index, blkno);
LockBuffer(nextbuffer, lockmode);
UnlockReleaseBuffer(buffer);
/* Sanity check that the page we stepped to is of similar kind. */
page = BufferGetPage(nextbuffer);
if (isLeaf != GinPageIsLeaf(page) || isData != GinPageIsData(page))
elog(ERROR, "right sibling of GIN page is of different type");
/*
* Given the proper lock sequence above, we should never land on a
* deleted page.
*/
if (GinPageIsDeleted(page))
elog(ERROR, "right sibling of GIN page was deleted");
return nextbuffer;
}
void
freeGinBtreeStack(GinBtreeStack *stack)
{
@ -235,12 +268,12 @@ ginFindParents(GinBtree btree, GinBtreeStack *stack,
while ((offset = btree->findChildPtr(btree, page, stack->blkno, InvalidOffsetNumber)) == InvalidOffsetNumber)
{
blkno = GinPageGetOpaque(page)->rightlink;
LockBuffer(buffer, GIN_UNLOCK);
ReleaseBuffer(buffer);
if (blkno == InvalidBlockNumber)
{
UnlockReleaseBuffer(buffer);
break;
buffer = ReadBuffer(btree->index, blkno);
LockBuffer(buffer, GIN_EXCLUSIVE);
}
buffer = ginStepRight(buffer, btree->index, GIN_EXCLUSIVE);
page = BufferGetPage(buffer);
}
@ -444,23 +477,21 @@ ginInsertValue(GinBtree btree, GinBtreeStack *stack, GinStatsData *buildStats)
{
BlockNumber rightlink = GinPageGetOpaque(page)->rightlink;
LockBuffer(parent->buffer, GIN_UNLOCK);
if (rightlink == InvalidBlockNumber)
{
/*
* rightmost page, but we don't find parent, we should use
* plain search...
*/
LockBuffer(parent->buffer, GIN_UNLOCK);
ginFindParents(btree, stack, rootBlkno);
parent = stack->parent;
Assert(parent != NULL);
break;
}
parent->buffer = ginStepRight(parent->buffer, btree->index, GIN_EXCLUSIVE);
parent->blkno = rightlink;
parent->buffer = ReleaseAndReadBuffer(parent->buffer, btree->index, parent->blkno);
LockBuffer(parent->buffer, GIN_EXCLUSIVE);
page = BufferGetPage(parent->buffer);
}

31
src/backend/access/gin/ginget.c
View File

@ -105,16 +105,11 @@ moveRightIfItNeeded(GinBtreeData *btree, GinBtreeStack *stack)
/*
* We scanned the whole page, so we should take right page
*/
stack->blkno = GinPageGetOpaque(page)->rightlink;
if (GinPageRightMost(page))
return false; /* no more pages */
LockBuffer(stack->buffer, GIN_UNLOCK);
stack->buffer = ReleaseAndReadBuffer(stack->buffer,
btree->index,
stack->blkno);
LockBuffer(stack->buffer, GIN_SHARE);
stack->buffer = ginStepRight(stack->buffer, btree->index, GIN_SHARE);
stack->blkno = BufferGetBlockNumber(stack->buffer);
stack->off = FirstOffsetNumber;
}
@ -132,7 +127,6 @@ scanPostingTree(Relation index, GinScanEntry scanEntry,
GinPostingTreeScan *gdi;
Buffer buffer;
Page page;
BlockNumber blkno;
/* Descend to the leftmost leaf page */
gdi = ginPrepareScanPostingTree(index, rootPostingTree, TRUE);
@ -162,10 +156,7 @@ scanPostingTree(Relation index, GinScanEntry scanEntry,
if (GinPageRightMost(page))
break; /* no more pages */
blkno = GinPageGetOpaque(page)->rightlink;
LockBuffer(buffer, GIN_UNLOCK);
buffer = ReleaseAndReadBuffer(buffer, index, blkno);
LockBuffer(buffer, GIN_SHARE);
buffer = ginStepRight(buffer, index, GIN_SHARE);
}
UnlockReleaseBuffer(buffer);
@ -543,7 +534,6 @@ static void
entryGetNextItem(GinState *ginstate, GinScanEntry entry)
{
Page page;
BlockNumber blkno;
for (;;)
{
@ -561,23 +551,18 @@ entryGetNextItem(GinState *ginstate, GinScanEntry entry)
* It's needed to go by right link. During that we should refind
* first ItemPointer greater that stored
*/
blkno = GinPageGetOpaque(page)->rightlink;
LockBuffer(entry->buffer, GIN_UNLOCK);
if (blkno == InvalidBlockNumber)
if (GinPageRightMost(page))
{
ReleaseBuffer(entry->buffer);
UnlockReleaseBuffer(entry->buffer);
ItemPointerSetInvalid(&entry->curItem);
entry->buffer = InvalidBuffer;
entry->isFinished = TRUE;
return;
}
entry->buffer = ReleaseAndReadBuffer(entry->buffer,
ginstate->index,
blkno);
LockBuffer(entry->buffer, GIN_SHARE);
entry->buffer = ginStepRight(entry->buffer,
ginstate->index,
GIN_SHARE);
page = BufferGetPage(entry->buffer);
entry->offset = InvalidOffsetNumber;

45
src/backend/access/gin/ginvacuum.c
View File

@ -240,6 +240,9 @@ ginVacuumPostingTreeLeaves(GinVacuumState *gvs, BlockNumber blkno, bool isRoot,
return hasVoidPage;
}
/*
* Delete a posting tree page.
*/
static void
ginDeletePage(GinVacuumState *gvs, BlockNumber deleteBlkno, BlockNumber leftBlkno,
BlockNumber parentBlkno, OffsetNumber myoff, bool isParentRoot)
@ -249,39 +252,35 @@ ginDeletePage(GinVacuumState *gvs, BlockNumber deleteBlkno, BlockNumber leftBlkn
Buffer pBuffer;
Page page,
parentPage;
BlockNumber rightlink;
/*
* Lock the pages in the same order as an insertion would, to avoid
* deadlocks: left, then right, then parent.
*/
lBuffer = ReadBufferExtended(gvs->index, MAIN_FORKNUM, leftBlkno,
RBM_NORMAL, gvs->strategy);
dBuffer = ReadBufferExtended(gvs->index, MAIN_FORKNUM, deleteBlkno,
RBM_NORMAL, gvs->strategy);
if (leftBlkno != InvalidBlockNumber)
lBuffer = ReadBufferExtended(gvs->index, MAIN_FORKNUM, leftBlkno,
RBM_NORMAL, gvs->strategy);
else
lBuffer = InvalidBuffer;
pBuffer = ReadBufferExtended(gvs->index, MAIN_FORKNUM, parentBlkno,
RBM_NORMAL, gvs->strategy);
LockBuffer(lBuffer, GIN_EXCLUSIVE);
LockBuffer(dBuffer, GIN_EXCLUSIVE);
if (!isParentRoot) /* parent is already locked by
* LockBufferForCleanup() */
LockBuffer(pBuffer, GIN_EXCLUSIVE);
if (leftBlkno != InvalidBlockNumber)
LockBuffer(lBuffer, GIN_EXCLUSIVE);
START_CRIT_SECTION();
if (leftBlkno != InvalidBlockNumber)
{
BlockNumber rightlink;
/* Unlink the page by changing left sibling's rightlink */
page = BufferGetPage(dBuffer);
rightlink = GinPageGetOpaque(page)->rightlink;
page = BufferGetPage(dBuffer);
rightlink = GinPageGetOpaque(page)->rightlink;
page = BufferGetPage(lBuffer);
GinPageGetOpaque(page)->rightlink = rightlink;
}
page = BufferGetPage(lBuffer);
GinPageGetOpaque(page)->rightlink = rightlink;
/* Delete downlink from parent */
parentPage = BufferGetPage(pBuffer);
#ifdef USE_ASSERT_CHECKING
do
@ -363,10 +362,7 @@ ginDeletePage(GinVacuumState *gvs, BlockNumber deleteBlkno, BlockNumber leftBlkn
if (!isParentRoot)
LockBuffer(pBuffer, GIN_UNLOCK);
ReleaseBuffer(pBuffer);
if (leftBlkno != InvalidBlockNumber)
UnlockReleaseBuffer(lBuffer);
UnlockReleaseBuffer(lBuffer);
UnlockReleaseBuffer(dBuffer);
END_CRIT_SECTION();
@ -435,10 +431,9 @@ ginScanToDelete(GinVacuumState *gvs, BlockNumber blkno, bool isRoot,
if (GinPageGetOpaque(page)->maxoff < FirstOffsetNumber)
{
/* the page is empty */
if (!(me->leftBlkno == InvalidBlockNumber && GinPageRightMost(page)))
/* we never delete the left- or rightmost branch */
if (me->leftBlkno != InvalidBlockNumber && !GinPageRightMost(page))
{
/* we never delete right most branch */
Assert(!isRoot);
ginDeletePage(gvs, blkno, me->leftBlkno, me->parent->blkno, myoff, me->parent->isRoot);
meDelete = TRUE;

1
src/include/access/gin_private.h
View File

@ -516,6 +516,7 @@ typedef struct GinBtreeData
extern GinBtreeStack *ginPrepareFindLeafPage(GinBtree btree, BlockNumber blkno);
extern GinBtreeStack *ginFindLeafPage(GinBtree btree, GinBtreeStack *stack);
extern Buffer ginStepRight(Buffer buffer, Relation index, int lockmode);
extern void freeGinBtreeStack(GinBtreeStack *stack);
extern void ginInsertValue(GinBtree btree, GinBtreeStack *stack,
GinStatsData *buildStats);