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

Make btree index structure adjustments and WAL logging changes needed to 

support btree compaction, as per proposal of a few days ago.  btree index
pages no longer store parent links, instead they have a level indicator
(counting up from zero for leaf pages).  The FixBTree recovery logic is
removed, and replaced by code that detects missing parent-level insertions
during WAL replay.  Also, generate appropriate WAL entries when updating
btree metapage and when building a btree index from scratch.  I believe
btree indexes are now completely WAL-legal for the first time.
initdb forced due to index and WAL changes.
This commit is contained in:
Tom Lane 2003-02-21 00:06:22 +00:00
parent 4df0f1d26f
commit 70508ba7ae
13 changed files with 2179 additions and 1875 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

4
src/backend/access/nbtree/Makefile
View File

@ -4,7 +4,7 @@
# Makefile for access/nbtree
#
# IDENTIFICATION
# $Header: /cvsroot/pgsql/src/backend/access/nbtree/Makefile,v 1.11 2001/07/15 22:48:16 tgl Exp $
# $Header: /cvsroot/pgsql/src/backend/access/nbtree/Makefile,v 1.12 2003/02/21 00:06:21 tgl Exp $
#
#-------------------------------------------------------------------------
@ -13,7 +13,7 @@ top_builddir = ../../../..
include $(top_builddir)/src/Makefile.global
OBJS = nbtcompare.o nbtinsert.o nbtpage.o nbtree.o nbtsearch.o \
nbtstrat.o nbtutils.o nbtsort.o
nbtstrat.o nbtutils.o nbtsort.o nbtxlog.o
all: SUBSYS.o

506
src/backend/access/nbtree/README
View File

@ -1,186 +1,378 @@
$Header: /cvsroot/pgsql/src/backend/access/nbtree/README,v 1.6 2002/10/20 20:47:31 tgl Exp $
$Header: /cvsroot/pgsql/src/backend/access/nbtree/README,v 1.7 2003/02/21 00:06:21 tgl Exp $
This directory contains a correct implementation of Lehman and Yao's
high-concurrency B-tree management algorithm (P. Lehman and S. Yao,
Efficient Locking for Concurrent Operations on B-Trees, ACM Transactions
on Database Systems, Vol 6, No. 4, December 1981, pp 650-670).
on Database Systems, Vol 6, No. 4, December 1981, pp 650-670). We also
use a simplified version of the deletion logic described in Lanin and
Shasha (V. Lanin and D. Shasha, A Symmetric Concurrent B-Tree Algorithm,
Proceedings of 1986 Fall Joint Computer Conference, pp 380-389).
We have made the following changes in order to incorporate their algorithm
The Lehman and Yao algorithm and insertions
-------------------------------------------
We have made the following changes in order to incorporate the L&Y algorithm
into Postgres:
+ The requirement that all btree keys be unique is too onerous,
but the algorithm won't work correctly without it. Fortunately, it is
only necessary that keys be unique on a single tree level, because L&Y
only use the assumption of key uniqueness when re-finding a key in a
parent node (to determine where to insert the key for a split page).
Therefore, we can use the link field to disambiguate multiple
occurrences of the same user key: only one entry in the parent level
will be pointing at the page we had split. (Indeed we need not look at
the real "key" at all, just at the link field.) We can distinguish
items at the leaf level in the same way, by examining their links to
heap tuples; we'd never have two items for the same heap tuple.
The requirement that all btree keys be unique is too onerous,
but the algorithm won't work correctly without it. Fortunately, it is
only necessary that keys be unique on a single tree level, because L&Y
only use the assumption of key uniqueness when re-finding a key in a
parent page (to determine where to insert the key for a split page).
Therefore, we can use the link field to disambiguate multiple
occurrences of the same user key: only one entry in the parent level
will be pointing at the page we had split. (Indeed we need not look at
the real "key" at all, just at the link field.) We can distinguish
items at the leaf level in the same way, by examining their links to
heap tuples; we'd never have two items for the same heap tuple.
+ Lehman and Yao assume that the key range for a subtree S is described
by Ki < v <= Ki+1 where Ki and Ki+1 are the adjacent keys in the parent
node. This does not work for nonunique keys (for example, if we have
enough equal keys to spread across several leaf pages, there *must* be
some equal bounding keys in the first level up). Therefore we assume
Ki <= v <= Ki+1 instead. A search that finds exact equality to a
bounding key in an upper tree level must descend to the left of that
key to ensure it finds any equal keys in the preceding page. An
insertion that sees the high key of its target page is equal to the key
to be inserted has a choice whether or not to move right, since the new
key could go on either page. (Currently, we try to find a page where
there is room for the new key without a split.)
Lehman and Yao assume that the key range for a subtree S is described
by Ki < v <= Ki+1 where Ki and Ki+1 are the adjacent keys in the parent
page. This does not work for nonunique keys (for example, if we have
enough equal keys to spread across several leaf pages, there *must* be
some equal bounding keys in the first level up). Therefore we assume
Ki <= v <= Ki+1 instead. A search that finds exact equality to a
bounding key in an upper tree level must descend to the left of that
key to ensure it finds any equal keys in the preceding page. An
insertion that sees the high key of its target page is equal to the key
to be inserted has a choice whether or not to move right, since the new
key could go on either page. (Currently, we try to find a page where
there is room for the new key without a split.)
+ Lehman and Yao don't require read locks, but assume that in-memory
copies of tree nodes are unshared. Postgres shares in-memory buffers
among backends. As a result, we do page-level read locking on btree
nodes in order to guarantee that no record is modified while we are
examining it. This reduces concurrency but guaranteees correct
behavior. An advantage is that when trading in a read lock for a
write lock, we need not re-read the page after getting the write lock.
Since we're also holding a pin on the shared buffer containing the
page, we know that buffer still contains the page and is up-to-date.
Lehman and Yao don't require read locks, but assume that in-memory
copies of tree pages are unshared. Postgres shares in-memory buffers
among backends. As a result, we do page-level read locking on btree
pages in order to guarantee that no record is modified while we are
examining it. This reduces concurrency but guaranteees correct
behavior. An advantage is that when trading in a read lock for a
write lock, we need not re-read the page after getting the write lock.
Since we're also holding a pin on the shared buffer containing the
page, we know that buffer still contains the page and is up-to-date.
+ We support the notion of an ordered "scan" of an index as well as
insertions, deletions, and simple lookups. A scan in the forward
direction is no problem, we just use the right-sibling pointers that
L&Y require anyway. (Thus, once we have descended the tree to the
correct start point for the scan, the scan looks only at leaf pages
and never at higher tree levels.) To support scans in the backward
direction, we also store a "left sibling" link much like the "right
sibling". (This adds an extra step to the L&Y split algorithm: while
holding the write lock on the page being split, we also lock its former
right sibling to update that page's left-link. This is safe since no
writer of that page can be interested in acquiring a write lock on our
page.) A backwards scan has one additional bit of complexity: after
following the left-link we must account for the possibility that the
left sibling page got split before we could read it. So, we have to
move right until we find a page whose right-link matches the page we
came from.
We support the notion of an ordered "scan" of an index as well as
insertions, deletions, and simple lookups. A scan in the forward
direction is no problem, we just use the right-sibling pointers that
L&Y require anyway. (Thus, once we have descended the tree to the
correct start point for the scan, the scan looks only at leaf pages
and never at higher tree levels.) To support scans in the backward
direction, we also store a "left sibling" link much like the "right
sibling". (This adds an extra step to the L&Y split algorithm: while
holding the write lock on the page being split, we also lock its former
right sibling to update that page's left-link. This is safe since no
writer of that page can be interested in acquiring a write lock on our
page.) A backwards scan has one additional bit of complexity: after
following the left-link we must account for the possibility that the
left sibling page got split before we could read it. So, we have to
move right until we find a page whose right-link matches the page we
came from. (Actually, it's even harder than that; see deletion discussion
below.)
+ Read locks on a page are held for as long as a scan is examining a page.
But nbtree.c arranges to drop the read lock, but not the buffer pin,
on the current page of a scan before control leaves nbtree. When we
come back to resume the scan, we have to re-grab the read lock and
then move right if the current item moved (see _bt_restscan()). Keeping
the pin ensures that the current item cannot move left or be deleted
(see btbulkdelete).
Read locks on a page are held for as long as a scan is examining a page.
But nbtree.c arranges to drop the read lock, but not the buffer pin,
on the current page of a scan before control leaves nbtree. When we
come back to resume the scan, we have to re-grab the read lock and
then move right if the current item moved (see _bt_restscan()). Keeping
the pin ensures that the current item cannot move left or be deleted
(see btbulkdelete).
+ In most cases we release our lock and pin on a page before attempting
to acquire pin and lock on the page we are moving to. In a few places
it is necessary to lock the next page before releasing the current one.
This is safe when moving right or up, but not when moving left or down
(else we'd create the possibility of deadlocks).
In most cases we release our lock and pin on a page before attempting
to acquire pin and lock on the page we are moving to. In a few places
it is necessary to lock the next page before releasing the current one.
This is safe when moving right or up, but not when moving left or down
(else we'd create the possibility of deadlocks).
+ Lehman and Yao fail to discuss what must happen when the root page
becomes full and must be split. Our implementation is to split the
root in the same way that any other page would be split, then construct
a new root page holding pointers to both of the resulting pages (which
now become siblings on level 2 of the tree). The new root page is then
installed by altering the root pointer in the meta-data page (see
below). This works because the root is not treated specially in any
other way --- in particular, searches will move right using its link
pointer if the link is set. Therefore, searches will find the data
that's been moved into the right sibling even if they read the metadata
page before it got updated. This is the same reasoning that makes a
split of a non-root page safe. The locking considerations are similar too.
Lehman and Yao fail to discuss what must happen when the root page
becomes full and must be split. Our implementation is to split the
root in the same way that any other page would be split, then construct
a new root page holding pointers to both of the resulting pages (which
now become siblings on the next level of the tree). The new root page
is then installed by altering the root pointer in the meta-data page (see
below). This works because the root is not treated specially in any
other way --- in particular, searches will move right using its link
pointer if the link is set. Therefore, searches will find the data
that's been moved into the right sibling even if they read the meta-data
page before it got updated. This is the same reasoning that makes a
split of a non-root page safe. The locking considerations are similar too.
+ Lehman and Yao assume fixed-size keys, but we must deal with
variable-size keys. Therefore there is not a fixed maximum number of
keys per page; we just stuff in as many as will fit. When we split a
page, we try to equalize the number of bytes, not items, assigned to
each of the resulting pages. Note we must include the incoming item in
this calculation, otherwise it is possible to find that the incoming
item doesn't fit on the split page where it needs to go!
When an inserter recurses up the tree, splitting internal pages to insert
links to pages inserted on the level below, it is possible that it will
need to access a page above the level that was the root when it began its
descent (or more accurately, the level that was the root when it read the
meta-data page). In this case the stack it made while descending does not
help for finding the correct page. When this happens, we find the correct
place by re-descending the tree until we reach the level one above the
level we need to insert a link to, and then moving right as necessary.
(Typically this will take only two fetches, the meta-data page and the new
root, but in principle there could have been more than one root split
since we saw the root. We can identify the correct tree level by means of
the level numbers stored in each page. The situation is rare enough that
we do not need a more efficient solution.)
In addition, the following things are handy to know:
Lehman and Yao assume fixed-size keys, but we must deal with
variable-size keys. Therefore there is not a fixed maximum number of
keys per page; we just stuff in as many as will fit. When we split a
page, we try to equalize the number of bytes, not items, assigned to
each of the resulting pages. Note we must include the incoming item in
this calculation, otherwise it is possible to find that the incoming
item doesn't fit on the split page where it needs to go!
+ Page zero of every btree is a meta-data page. This page stores
the location of the root page, a pointer to a list of free
pages, and other stuff that's handy to know. (Currently, we
never shrink btree indexes so there are never any free pages.)
The deletion algorithm
----------------------
+ The algorithm assumes we can fit at least three items per page
(a "high key" and two real data items). Therefore it's unsafe
to accept items larger than 1/3rd page size. Larger items would
work sometimes, but could cause failures later on depending on
what else gets put on their page.
Deletions of leaf items are handled by getting a super-exclusive lock on
the target page, so that no other backend has a pin on the page when the
deletion starts. This means no scan is pointing at the page, so no other
backend can lose its place due to the item deletion.
+ This algorithm doesn't guarantee btree consistency after a kernel crash
or hardware failure. To do that, we'd need ordered writes, and UNIX
doesn't support ordered writes (short of fsync'ing every update, which
is too high a price). Rebuilding corrupted indexes during restart
seems more attractive.
The above does not work for deletion of items in internal pages, since
other backends keep no lock nor pin on a page they have descended past.
Instead, when a backend is ascending the tree using its stack, it must
be prepared for the possibility that the item it wants is to the left of
the recorded position (but it can't have moved left out of the recorded
page). Since we hold a lock on the lower page (per L&Y) until we have
re-found the parent item that links to it, we can be assured that the
parent item does still exist and can't have been deleted. Also, because
we are matching downlink page numbers and not data keys, we don't have any
problem with possibly misidentifying the parent item.
+ Deletions are handled by getting a super-exclusive lock on the target
page, so that no other backend has a pin on the page when the deletion
starts. This means no scan is pointing at the page. This is OK for
deleting leaf items, probably not OK for deleting internal nodes;
will need to think harder when it's time to support index compaction.
We consider deleting an entire page from the btree only when it's become
completely empty of items. (Merging partly-full pages would allow better
space reuse, but it seems impractical to move existing data items left or
right to make this happen --- a scan moving in the opposite direction
might miss the items if so. We could do it during VACUUM FULL, though.)
Also, we *never* delete the rightmost page on a tree level (this
restriction simplifies the traversal algorithms, as explained below).
+ "ScanKey" data structures are used in two fundamentally different ways
in this code. Searches for the initial position for a scan, as well as
insertions, use scankeys in which the comparison function is a 3-way
comparator (<0, =0, >0 result). These scankeys are built within the
btree code (eg, by _bt_mkscankey()) and used by _bt_compare(). Once we
are positioned, sequential examination of tuples in a scan is done by
_bt_checkkeys() using scankeys in which the comparison functions return
booleans --- for example, int4lt might be used. These scankeys are the
ones originally passed in from outside the btree code. Same
representation, but different comparison functions!
To delete an empty page, we acquire write lock on its left sibling (if
any), the target page itself, the right sibling (there must be one), and
the parent page, in that order. The parent page must be found using the
same type of search as used to find the parent during an insertion split.
Then we update the side-links in the siblings, mark the target page
deleted, and remove the downlink from the parent, as well as the parent's
upper bounding key for the target (the one separating it from its right
sibling). This causes the target page's key space to effectively belong
to its right sibling. (Neither the left nor right sibling pages need to
change their "high key" if any; so there is no problem with possibly not
having enough space to replace a high key.) The side-links in the target
page are not changed.
Notes about data representation:
(Note: Lanin and Shasha prefer to make the key space move left, but their
argument for doing so hinges on not having left-links, which we have
anyway. So we simplify the algorithm by moving key space right.)
+ The right-sibling link required by L&Y is kept in the page "opaque
data" area, as is the left-sibling link and some flags.
To preserve consistency on the parent level, we cannot merge the key space
of a page into its right sibling unless the right sibling is a child of
the same parent --- otherwise, the parent's key space assignment changes
too, meaning we'd have to make bounding-key updates in its parent, and
perhaps all the way up the tree. Since we can't possibly do that
atomically, we forbid this case. That means that the rightmost child of a
parent node can't be deleted unless it's the only remaining child.
+ We also keep a parent link in the opaque data, but this link is not
very trustworthy because it is not updated when the parent page splits.
Thus, it points to some page on the parent level, but possibly a page
well to the left of the page's actual current parent. In most cases
we do not need this link at all. Normally we return to a parent page
using a stack of entries that are made as we descend the tree, as in L&Y.
There is exactly one case where the stack will not help: concurrent
root splits. If an inserter process needs to split what had been the
root when it started its descent, but finds that that page is no longer
the root (because someone else split it meanwhile), then it uses the
parent link to move up to the next level. This is OK because we do fix
the parent link in a former root page when splitting it. This logic
will work even if the root is split multiple times (even up to creation
of multiple new levels) before an inserter returns to it. The same
could not be said of finding the new root via the metapage, since that
would work only for a single level of added root.
When we delete the last remaining child of a parent page, we mark the
parent page "half-dead" as part of the atomic update that deletes the
child page. This implicitly transfers the parent's key space to its right
sibling (which it must have, since we never delete the overall-rightmost
page of a level). No future insertions into the parent level are allowed
to insert keys into the half-dead page --- they must move right to its
sibling, instead. The parent remains empty and can be deleted in a
separate atomic action. (However, if it's the rightmost child of its own
parent, it might have to stay half-dead for awhile, until it's also the
only child.)
+ The Postgres disk block data format (an array of items) doesn't fit
Lehman and Yao's alternating-keys-and-pointers notion of a disk page,
so we have to play some games.
Note that an empty leaf page is a valid tree state, but an empty interior
page is not legal (an interior page must have children to delegate its
key space to). So an interior page *must* be marked half-dead as soon
as its last child is deleted.
+ On a page that is not rightmost in its tree level, the "high key" is
kept in the page's first item, and real data items start at item 2.
The link portion of the "high key" item goes unused. A page that is
rightmost has no "high key", so data items start with the first item.
Putting the high key at the left, rather than the right, may seem odd,
but it avoids moving the high key as we add data items.
The notion of a half-dead page means that the key space relationship between
the half-dead page's level and its parent's level may be a little out of
whack: key space that appears to belong to the half-dead page's parent on the
parent level may really belong to its right sibling. We can tolerate this,
however, because insertions and deletions on upper tree levels are always
done by reference to child page numbers, not keys. The only cost is that
searches may sometimes descend to the half-dead page and then have to move
right, rather than going directly to the sibling page.
+ On a leaf page, the data items are simply links to (TIDs of) tuples
in the relation being indexed, with the associated key values.
A deleted page cannot be reclaimed immediately, since there may be other
processes waiting to reference it (ie, search processes that just left the
parent, or scans moving right or left from one of the siblings). These
processes must observe that the page is marked dead and recover
accordingly. Searches and forward scans simply follow the right-link
until they find a non-dead page --- this will be where the deleted page's
key-space moved to.
+ On a non-leaf page, the data items are down-links to child pages with
bounding keys. The key in each data item is the *lower* bound for
keys on that child page, so logically the key is to the left of that
downlink. The high key (if present) is the upper bound for the last
downlink. The first data item on each such page has no lower bound
--- or lower bound of minus infinity, if you prefer. The comparison
routines must treat it accordingly. The actual key stored in the
item is irrelevant, and need not be stored at all. This arrangement
corresponds to the fact that an L&Y non-leaf page has one more pointer
than key.
Stepping left in a backward scan is complicated because we must consider
the possibility that the left sibling was just split (meaning we must find
the rightmost page derived from the left sibling), plus the possibility
that the page we were just on has now been deleted and hence isn't in the
sibling chain at all anymore. So the move-left algorithm becomes:
0. Remember the page we are on as the "original page".
1. Follow the original page's left-link (we're done if this is zero).
2. If the current page is live and its right-link matches the "original
page", we are done.
3. Otherwise, move right one or more times looking for a live page whose
right-link matches the "original page". If found, we are done. (In
principle we could scan all the way to the right end of the index, but
in practice it seems better to give up after a small number of tries.
It's unlikely the original page's sibling split more than a few times
while we were in flight to it; if we do not find a matching link in a
few tries, then most likely the original page is deleted.)
4. Return to the "original page". If it is still live, return to step 1
(we guessed wrong about it being deleted, and should restart with its
current left-link). If it is dead, move right until a non-dead page
is found (there must be one, since rightmost pages are never deleted),
mark that as the new "original page", and return to step 1.
This algorithm is correct because the live page found by step 4 will have
the same left keyspace boundary as the page we started from. Therefore,
when we ultimately exit, it must be on a page whose right keyspace
boundary matches the left boundary of where we started --- which is what
we need to be sure we don't miss or re-scan any items.
Notes to operator class implementors:
A deleted page can only be reclaimed once there is no scan or search that
has a reference to it; until then, it must stay in place with its
right-link undisturbed. We implement this by waiting until all
transactions that were running at the time of deletion are dead; which is
overly strong, but is simple to implement within Postgres. When marked
dead, a deleted page is labeled with the next-transaction counter value.
VACUUM can reclaim the page for re-use when this transaction number is
older than the oldest open transaction. (NOTE: VACUUM FULL can reclaim
such pages immediately.)
+ With this implementation, we require each supported datatype to supply
us with a comparison procedure via pg_amproc. This procedure must take
two nonnull values A and B and return an int32 < 0, 0, or > 0 if A < B,
A = B, or A > B, respectively. See nbtcompare.c for examples.
Reclaiming a page doesn't actually change its state on disk --- we simply
record it in the shared-memory free space map, from which it will be
handed out the next time a new page is needed for a page split. The
deleted page's contents will be overwritten by the split operation.
(Note: if we find a deleted page with an extremely old transaction
number, it'd be worthwhile to re-mark it with FrozenTransactionId so that
a later xid wraparound can't cause us to think the page is unreclaimable.
But in more normal situations this would be a waste of a disk write.)
Because we never delete the rightmost page of any level (and in particular
never delete the root), it's impossible for the height of the tree to
decrease. After massive deletions we might have a scenario in which the
tree is "skinny", with several single-page levels below the root.
Operations will still be correct in this case, but we'd waste cycles
descending through the single-page levels. To handle this we use an idea
from Lanin and Shasha: we keep track of the "fast root" level, which is
the lowest single-page level. The meta-data page keeps a pointer to this
level as well as the true root. All ordinary operations initiate their
searches at the fast root not the true root. When we split a page that is
alone on its level or delete the next-to-last page on a level (both cases
are easily detected), we have to make sure that the fast root pointer is
adjusted appropriately. In the split case, we do this work as part of the
atomic update for the insertion into the parent level; in the delete case
as part of the atomic update for the delete (either way, the metapage has
to be the last page locked in the update to avoid deadlock risks). This
avoids race conditions if two such operations are executing concurrently.
VACUUM needs to do a linear scan of an index to search for empty leaf
pages and half-dead parent pages that can be deleted, as well as deleted
pages that can be reclaimed because they are older than all open
transactions.
WAL considerations
------------------
The insertion and deletion algorithms in themselves don't guarantee btree
consistency after a crash. To provide robustness, we depend on WAL
replay. A single WAL entry is effectively an atomic action --- we can
redo it from the log if it fails to complete.
Ordinary item insertions (that don't force a page split) are of course
single WAL entries, since they only affect one page. The same for
leaf-item deletions (if the deletion brings the leaf page to zero items,
it is now a candidate to be deleted, but that is a separate action).
An insertion that causes a page split is logged as a single WAL entry for
the changes occuring on the insertion's level --- including update of the
right sibling's left-link --- followed by a second WAL entry for the
insertion on the parent level (which might itself be a page split, requiring
an additional insertion above that, etc).
For a root split, the followon WAL entry is a "new root" entry rather than
an "insertion" entry, but details are otherwise much the same.
Because insertion involves multiple atomic actions, the WAL replay logic
has to detect the case where a page split isn't followed by a matching
insertion on the parent level, and then do that insertion on its own (and
recursively for any subsequent parent insertion, of course). This is
feasible because the WAL entry for the split contains enough info to know
what must be inserted in the parent level.
When splitting a non-root page that is alone on its level, the required
metapage update (of the "fast root" link) is performed and logged as part
of the insertion into the parent level. When splitting the root page, the
metapage update is handled as part of the "new root" action.
A page deletion is logged as a single WAL entry covering all four
required page updates (target page, left and right siblings, and parent)
as an atomic event. (Any required fast-root link update is also part
of the WAL entry.) If the parent page becomes half-dead but is not
immediately deleted due to a subsequent crash, there is no loss of
consistency, and the empty page will be picked up by the next VACUUM.
Other things that are handy to know
-----------------------------------
Page zero of every btree is a meta-data page. This page stores the
location of the root page --- both the true root and the current effective
root ("fast" root).
The algorithm assumes we can fit at least three items per page
(a "high key" and two real data items). Therefore it's unsafe
to accept items larger than 1/3rd page size. Larger items would
work sometimes, but could cause failures later on depending on
what else gets put on their page.
"ScanKey" data structures are used in two fundamentally different ways
in this code. Searches for the initial position for a scan, as well as
insertions, use scankeys in which the comparison function is a 3-way
comparator (<0, =0, >0 result). These scankeys are built within the
btree code (eg, by _bt_mkscankey()) and used by _bt_compare(). Once we
are positioned, sequential examination of tuples in a scan is done by
_bt_checkkeys() using scankeys in which the comparison functions return
booleans --- for example, int4lt might be used. These scankeys are the
ones originally passed in from outside the btree code. Same
representation, but different comparison functions!
Notes about data representation
-------------------------------
The right-sibling link required by L&Y is kept in the page "opaque
data" area, as is the left-sibling link, the page level, and some flags.
The page level counts upwards from zero at the leaf level, to the tree
depth minus 1 at the root. (Counting up from the leaves ensures that we
don't need to renumber any existing pages when splitting the root.)
The Postgres disk block data format (an array of items) doesn't fit
Lehman and Yao's alternating-keys-and-pointers notion of a disk page,
so we have to play some games.
On a page that is not rightmost in its tree level, the "high key" is
kept in the page's first item, and real data items start at item 2.
The link portion of the "high key" item goes unused. A page that is
rightmost has no "high key", so data items start with the first item.
Putting the high key at the left, rather than the right, may seem odd,
but it avoids moving the high key as we add data items.
On a leaf page, the data items are simply links to (TIDs of) tuples
in the relation being indexed, with the associated key values.
On a non-leaf page, the data items are down-links to child pages with
bounding keys. The key in each data item is the *lower* bound for
keys on that child page, so logically the key is to the left of that
downlink. The high key (if present) is the upper bound for the last
downlink. The first data item on each such page has no lower bound
--- or lower bound of minus infinity, if you prefer. The comparison
routines must treat it accordingly. The actual key stored in the
item is irrelevant, and need not be stored at all. This arrangement
corresponds to the fact that an L&Y non-leaf page has one more pointer
than key.
Notes to operator class implementors
------------------------------------
With this implementation, we require each supported datatype to supply
us with a comparison procedure via pg_amproc. This procedure must take
two nonnull values A and B and return an int32 < 0, 0, or > 0 if A < B,
A = B, or A > B, respectively. See nbtcompare.c for examples.

1199
src/backend/access/nbtree/nbtinsert.c
View File

File diff suppressed because it is too large Load Diff

260
src/backend/access/nbtree/nbtpage.c
View File

@ -9,7 +9,7 @@
*
*
* IDENTIFICATION
* $Header: /cvsroot/pgsql/src/backend/access/nbtree/nbtpage.c,v 1.58 2002/08/06 02:36:33 tgl Exp $
* $Header: /cvsroot/pgsql/src/backend/access/nbtree/nbtpage.c,v 1.59 2003/02/21 00:06:21 tgl Exp $
*
* NOTES
* Postgres btree pages look like ordinary relation pages. The opaque
@ -47,15 +47,16 @@ extern Buffer _bt_fixroot(Relation rel, Buffer oldrootbuf, bool release);
#define USELOCKING (!BuildingBtree && !IsInitProcessingMode())
/*
* _bt_metapinit() -- Initialize the metadata page of a btree.
* _bt_metapinit() -- Initialize the metadata page of a new btree.
*/
void
_bt_metapinit(Relation rel)
{
Buffer buf;
Page pg;
BTMetaPageData metad;
BTMetaPageData *metad;
BTPageOpaque op;
/* can't be sharing this with anyone, now... */
@ -67,18 +68,51 @@ _bt_metapinit(Relation rel)
RelationGetRelationName(rel));
buf = ReadBuffer(rel, P_NEW);
Assert(BufferGetBlockNumber(buf) == BTREE_METAPAGE);
pg = BufferGetPage(buf);
/* NO ELOG(ERROR) from here till newmeta op is logged */
START_CRIT_SECTION();
_bt_pageinit(pg, BufferGetPageSize(buf));
metad.btm_magic = BTREE_MAGIC;
metad.btm_version = BTREE_VERSION;
metad.btm_root = P_NONE;
metad.btm_level = 0;
memcpy((char *) BTPageGetMeta(pg), (char *) &metad, sizeof(metad));
metad = BTPageGetMeta(pg);
metad->btm_magic = BTREE_MAGIC;
metad->btm_version = BTREE_VERSION;
metad->btm_root = P_NONE;
metad->btm_level = 0;
metad->btm_fastroot = P_NONE;
metad->btm_fastlevel = 0;
op = (BTPageOpaque) PageGetSpecialPointer(pg);
op->btpo_flags = BTP_META;
/* XLOG stuff */
if (!rel->rd_istemp)
{
xl_btree_newmeta xlrec;
XLogRecPtr recptr;
XLogRecData rdata[1];
xlrec.node = rel->rd_node;
xlrec.meta.root = metad->btm_root;
xlrec.meta.level = metad->btm_level;
xlrec.meta.fastroot = metad->btm_fastroot;
xlrec.meta.fastlevel = metad->btm_fastlevel;
rdata[0].buffer = InvalidBuffer;
rdata[0].data = (char *) &xlrec;
rdata[0].len = SizeOfBtreeNewmeta;
rdata[0].next = NULL;
recptr = XLogInsert(RM_BTREE_ID, XLOG_BTREE_NEWMETA, rdata);
PageSetLSN(pg, recptr);
PageSetSUI(pg, ThisStartUpID);
}
END_CRIT_SECTION();
WriteBuffer(buf);
/* all done */
@ -102,6 +136,14 @@ _bt_metapinit(Relation rel)
* NOTE that the returned root page will have only a read lock set
* on it even if access = BT_WRITE!
*
* The returned page is not necessarily the true root --- it could be
* a "fast root" (a page that is alone in its level due to deletions).
* Also, if the root page is split while we are "in flight" to it,
* what we will return is the old root, which is now just the leftmost
* page on a probably-not-very-wide level. For most purposes this is
* as good as or better than the true root, so we do not bother to
* insist on finding the true root.
*
* On successful return, the root page is pinned and read-locked.
* The metadata page is not locked or pinned on exit.
*/
@ -162,15 +204,19 @@ _bt_getroot(Relation rel, int access)
rootblkno = BufferGetBlockNumber(rootbuf);
rootpage = BufferGetPage(rootbuf);
_bt_pageinit(rootpage, BufferGetPageSize(rootbuf));
rootopaque = (BTPageOpaque) PageGetSpecialPointer(rootpage);
rootopaque->btpo_prev = rootopaque->btpo_next = P_NONE;
rootopaque->btpo_flags = (BTP_LEAF | BTP_ROOT);
rootopaque->btpo.level = 0;
/* NO ELOG(ERROR) till meta is updated */
START_CRIT_SECTION();
metad->btm_root = rootblkno;
metad->btm_level = 1;
_bt_pageinit(rootpage, BufferGetPageSize(rootbuf));
rootopaque = (BTPageOpaque) PageGetSpecialPointer(rootpage);
rootopaque->btpo_flags |= (BTP_LEAF | BTP_ROOT);
metad->btm_level = 0;
metad->btm_fastroot = rootblkno;
metad->btm_fastlevel = 0;
/* XLOG stuff */
if (!rel->rd_istemp)
@ -180,16 +226,15 @@ _bt_getroot(Relation rel, int access)
XLogRecData rdata;
xlrec.node = rel->rd_node;
xlrec.level = 1;
BlockIdSet(&(xlrec.rootblk), rootblkno);
xlrec.rootblk = rootblkno;
xlrec.level = 0;
rdata.buffer = InvalidBuffer;
rdata.data = (char *) &xlrec;
rdata.len = SizeOfBtreeNewroot;
rdata.next = NULL;
recptr = XLogInsert(RM_BTREE_ID,
XLOG_BTREE_NEWROOT | XLOG_BTREE_LEAF,
&rdata);
recptr = XLogInsert(RM_BTREE_ID, XLOG_BTREE_NEWROOT, &rdata);
PageSetLSN(rootpage, recptr);
PageSetSUI(rootpage, ThisStartUpID);
@ -201,7 +246,11 @@ _bt_getroot(Relation rel, int access)
_bt_wrtnorelbuf(rel, rootbuf);
/* swap write lock for read lock */
/*
* swap root write lock for read lock. There is no danger of
* anyone else accessing the new root page while it's unlocked,
* since no one else knows where it is yet.
*/
LockBuffer(rootbuf, BUFFER_LOCK_UNLOCK);
LockBuffer(rootbuf, BT_READ);
@ -221,86 +270,72 @@ _bt_getroot(Relation rel, int access)
}
else
{
rootblkno = metad->btm_root;
rootblkno = metad->btm_fastroot;
_bt_relbuf(rel, metabuf); /* done with the meta page */
rootbuf = _bt_getbuf(rel, rootblkno, BT_READ);
}
/*
* Race condition: If the root page split between the time we looked
* at the metadata page and got the root buffer, then we got the wrong
* buffer. Release it and try again.
* By here, we have a pin and read lock on the root page, and no
* lock set on the metadata page. Return the root page's buffer.
*/
rootpage = BufferGetPage(rootbuf);
rootopaque = (BTPageOpaque) PageGetSpecialPointer(rootpage);
return rootbuf;
}
if (!P_ISROOT(rootopaque))
/*
* _bt_gettrueroot() -- Get the true root page of the btree.
*
* This is the same as the BT_READ case of _bt_getroot(), except
* we follow the true-root link not the fast-root link.
*
* By the time we acquire lock on the root page, it might have been split and
* not be the true root anymore. This is okay for the present uses of this
* routine; we only really need to be able to move up at least one tree level
* from whatever non-root page we were at. If we ever do need to lock the
* one true root page, we could loop here, re-reading the metapage on each
* failure. (Note that it wouldn't do to hold the lock on the metapage while
* moving to the root --- that'd deadlock against any concurrent root split.)
*/
Buffer
_bt_gettrueroot(Relation rel)
{
Buffer metabuf;
Page metapg;
BTPageOpaque metaopaque;
Buffer rootbuf;
BlockNumber rootblkno;
BTMetaPageData *metad;
metabuf = _bt_getbuf(rel, BTREE_METAPAGE, BT_READ);
metapg = BufferGetPage(metabuf);
metaopaque = (BTPageOpaque) PageGetSpecialPointer(metapg);
metad = BTPageGetMeta(metapg);
if (!(metaopaque->btpo_flags & BTP_META) ||
metad->btm_magic != BTREE_MAGIC)
elog(ERROR, "Index %s is not a btree",
RelationGetRelationName(rel));
if (metad->btm_version != BTREE_VERSION)
elog(ERROR, "Version mismatch on %s: version %d file, version %d code",
RelationGetRelationName(rel),
metad->btm_version, BTREE_VERSION);
/* if no root page initialized yet, fail */
if (metad->btm_root == P_NONE)
{
/*
* It happened, but if root page splitter failed to create new
* root page then we'll go in loop trying to call _bt_getroot
* again and again.
*/
if (FixBTree)
{
Buffer newrootbuf;
check_parent:;
if (BTreeInvalidParent(rootopaque)) /* unupdated! */
{
LockBuffer(rootbuf, BUFFER_LOCK_UNLOCK);
LockBuffer(rootbuf, BT_WRITE);
/* handle concurrent fix of root page */
if (BTreeInvalidParent(rootopaque)) /* unupdated! */
{
elog(WARNING, "bt_getroot[%s]: fixing root page", RelationGetRelationName(rel));
newrootbuf = _bt_fixroot(rel, rootbuf, true);
LockBuffer(newrootbuf, BUFFER_LOCK_UNLOCK);
LockBuffer(newrootbuf, BT_READ);
rootbuf = newrootbuf;
rootpage = BufferGetPage(rootbuf);
rootopaque = (BTPageOpaque) PageGetSpecialPointer(rootpage);
/* New root might be splitted while changing lock */
if (P_ISROOT(rootopaque))
return (rootbuf);
/* rootbuf is read locked */
goto check_parent;
}
else
{
/* someone else already fixed root */
LockBuffer(rootbuf, BUFFER_LOCK_UNLOCK);
LockBuffer(rootbuf, BT_READ);
}
}
/*
* Ok, here we have old root page with btpo_parent pointing to
* upper level - check parent page because of there is good
* chance that parent is root page.
*/
newrootbuf = _bt_getbuf(rel, rootopaque->btpo_parent, BT_READ);
_bt_relbuf(rel, rootbuf);
rootbuf = newrootbuf;
rootpage = BufferGetPage(rootbuf);
rootopaque = (BTPageOpaque) PageGetSpecialPointer(rootpage);
if (P_ISROOT(rootopaque))
return (rootbuf);
/* no luck -:( */
}
/* try again */
_bt_relbuf(rel, rootbuf);
return _bt_getroot(rel, access);
_bt_relbuf(rel, metabuf);
return InvalidBuffer;
}
/*
* By here, we have a correct lock on the root block, its reference
* count is correct, and we have no lock set on the metadata page.
* Return the root block.
*/
rootblkno = metad->btm_root;
_bt_relbuf(rel, metabuf); /* done with the meta page */
rootbuf = _bt_getbuf(rel, rootblkno, BT_READ);
return rootbuf;
}
@ -397,13 +432,14 @@ _bt_wrtnorelbuf(Relation rel, Buffer buf)
/*
* _bt_pageinit() -- Initialize a new page.
*
* On return, the page header is initialized; data space is empty;
* special space is zeroed out.
*/
void
_bt_pageinit(Page page, Size size)
{
PageInit(page, size, sizeof(BTPageOpaqueData));
((BTPageOpaque) PageGetSpecialPointer(page))->btpo_parent =
InvalidBlockNumber;
}
/*
@ -418,9 +454,12 @@ _bt_pageinit(Page page, Size size)
* at least the old root page when you call this, you're making a big
* mistake. On exit, metapage data is correct and we no longer have
* a pin or lock on the metapage.
*
* XXX this is not used for splitting anymore, only in nbtsort.c at the
* completion of btree building.
*/
void
_bt_metaproot(Relation rel, BlockNumber rootbknum, int level)
_bt_metaproot(Relation rel, BlockNumber rootbknum, uint32 level)
{
Buffer metabuf;
Page metap;
@ -431,12 +470,42 @@ _bt_metaproot(Relation rel, BlockNumber rootbknum, int level)
metap = BufferGetPage(metabuf);
metaopaque = (BTPageOpaque) PageGetSpecialPointer(metap);
Assert(metaopaque->btpo_flags & BTP_META);
/* NO ELOG(ERROR) from here till newmeta op is logged */
START_CRIT_SECTION();
metad = BTPageGetMeta(metap);
metad->btm_root = rootbknum;
if (level == 0) /* called from _do_insert */
metad->btm_level += 1;
else
metad->btm_level = level; /* called from btsort */
metad->btm_level = level;
metad->btm_fastroot = rootbknum;
metad->btm_fastlevel = level;
/* XLOG stuff */
if (!rel->rd_istemp)
{
xl_btree_newmeta xlrec;
XLogRecPtr recptr;
XLogRecData rdata[1];
xlrec.node = rel->rd_node;
xlrec.meta.root = metad->btm_root;
xlrec.meta.level = metad->btm_level;
xlrec.meta.fastroot = metad->btm_fastroot;
xlrec.meta.fastlevel = metad->btm_fastlevel;
rdata[0].buffer = InvalidBuffer;
rdata[0].data = (char *) &xlrec;
rdata[0].len = SizeOfBtreeNewmeta;
rdata[0].next = NULL;
recptr = XLogInsert(RM_BTREE_ID, XLOG_BTREE_NEWMETA, rdata);
PageSetLSN(metap, recptr);
PageSetSUI(metap, ThisStartUpID);
}
END_CRIT_SECTION();
_bt_wrtbuf(rel, metabuf);
}
@ -467,6 +536,7 @@ _bt_itemdel(Relation rel, Buffer buf, ItemPointer tid)
xlrec.target.node = rel->rd_node;
xlrec.target.tid = *tid;
rdata[0].buffer = InvalidBuffer;
rdata[0].data = (char *) &xlrec;
rdata[0].len = SizeOfBtreeDelete;

399
src/backend/access/nbtree/nbtree.c
View File

@ -12,21 +12,17 @@
* Portions Copyright (c) 1994, Regents of the University of California
*
* IDENTIFICATION
* $Header: /cvsroot/pgsql/src/backend/access/nbtree/nbtree.c,v 1.94 2002/11/15 01:26:08 momjian Exp $
* $Header: /cvsroot/pgsql/src/backend/access/nbtree/nbtree.c,v 1.95 2003/02/21 00:06:21 tgl Exp $
*
*-------------------------------------------------------------------------
*/
#include "postgres.h"
#include "access/genam.h"
#include "access/heapam.h"
#include "access/nbtree.h"
#include "catalog/index.h"
#include "executor/executor.h"
#include "miscadmin.h"
#include "storage/sinval.h"
#include "access/xlogutils.h"
/* Working state for btbuild and its callback */
@ -817,396 +813,3 @@ _bt_restscan(IndexScanDesc scan)
ItemPointerSet(current, blkno, offnum);
}
}
static void
_bt_restore_page(Page page, char *from, int len)
{
BTItemData btdata;
Size itemsz;
char *end = from + len;
for (; from < end;)
{
memcpy(&btdata, from, sizeof(BTItemData));
itemsz = IndexTupleDSize(btdata.bti_itup) +
(sizeof(BTItemData) - sizeof(IndexTupleData));
itemsz = MAXALIGN(itemsz);
if (PageAddItem(page, (Item) from, itemsz,
FirstOffsetNumber, LP_USED) == InvalidOffsetNumber)
elog(PANIC, "_bt_restore_page: can't add item to page");
from += itemsz;
}
}
static void
btree_xlog_delete(bool redo, XLogRecPtr lsn, XLogRecord *record)
{
xl_btree_delete *xlrec;
Relation reln;
Buffer buffer;
Page page;
if (!redo || (record->xl_info & XLR_BKP_BLOCK_1))
return;
xlrec = (xl_btree_delete *) XLogRecGetData(record);
reln = XLogOpenRelation(redo, RM_BTREE_ID, xlrec->target.node);
if (!RelationIsValid(reln))
return;
buffer = XLogReadBuffer(false, reln,
ItemPointerGetBlockNumber(&(xlrec->target.tid)));
if (!BufferIsValid(buffer))
elog(PANIC, "btree_delete_redo: block unfound");
page = (Page) BufferGetPage(buffer);
if (PageIsNew((PageHeader) page))
elog(PANIC, "btree_delete_redo: uninitialized page");
if (XLByteLE(lsn, PageGetLSN(page)))
{
UnlockAndReleaseBuffer(buffer);
return;
}
PageIndexTupleDelete(page, ItemPointerGetOffsetNumber(&(xlrec->target.tid)));
PageSetLSN(page, lsn);
PageSetSUI(page, ThisStartUpID);
UnlockAndWriteBuffer(buffer);
return;
}
static void
btree_xlog_insert(bool redo, XLogRecPtr lsn, XLogRecord *record)
{
xl_btree_insert *xlrec;
Relation reln;
Buffer buffer;
Page page;
BTPageOpaque pageop;
if (redo && (record->xl_info & XLR_BKP_BLOCK_1))
return;
xlrec = (xl_btree_insert *) XLogRecGetData(record);
reln = XLogOpenRelation(redo, RM_BTREE_ID, xlrec->target.node);
if (!RelationIsValid(reln))
return;
buffer = XLogReadBuffer(false, reln,
ItemPointerGetBlockNumber(&(xlrec->target.tid)));
if (!BufferIsValid(buffer))
elog(PANIC, "btree_insert_%sdo: block unfound", (redo) ? "re" : "un");
page = (Page) BufferGetPage(buffer);
if (PageIsNew((PageHeader) page))
elog(PANIC, "btree_insert_%sdo: uninitialized page", (redo) ? "re" : "un");
pageop = (BTPageOpaque) PageGetSpecialPointer(page);
if (redo)
{
if (XLByteLE(lsn, PageGetLSN(page)))
{
UnlockAndReleaseBuffer(buffer);
return;
}
if (PageAddItem(page, (Item) ((char *) xlrec + SizeOfBtreeInsert),
record->xl_len - SizeOfBtreeInsert,
ItemPointerGetOffsetNumber(&(xlrec->target.tid)),
LP_USED) == InvalidOffsetNumber)
elog(PANIC, "btree_insert_redo: failed to add item");
PageSetLSN(page, lsn);
PageSetSUI(page, ThisStartUpID);
UnlockAndWriteBuffer(buffer);
}
else
{
if (XLByteLT(PageGetLSN(page), lsn))
elog(PANIC, "btree_insert_undo: bad page LSN");
if (!P_ISLEAF(pageop))
{
UnlockAndReleaseBuffer(buffer);
return;
}
elog(PANIC, "btree_insert_undo: unimplemented");
}
return;
}
static void
btree_xlog_split(bool redo, bool onleft, XLogRecPtr lsn, XLogRecord *record)
{
xl_btree_split *xlrec = (xl_btree_split *) XLogRecGetData(record);
Relation reln;
BlockNumber blkno;
Buffer buffer;
Page page;
BTPageOpaque pageop;
char *op = (redo) ? "redo" : "undo";
bool isleaf = (record->xl_info & XLOG_BTREE_LEAF);
reln = XLogOpenRelation(redo, RM_BTREE_ID, xlrec->target.node);
if (!RelationIsValid(reln))
return;
/* Left (original) sibling */
blkno = (onleft) ? ItemPointerGetBlockNumber(&(xlrec->target.tid)) :
BlockIdGetBlockNumber(&(xlrec->otherblk));
buffer = XLogReadBuffer(false, reln, blkno);
if (!BufferIsValid(buffer))
elog(PANIC, "btree_split_%s: lost left sibling", op);
page = (Page) BufferGetPage(buffer);
if (redo)
_bt_pageinit(page, BufferGetPageSize(buffer));
else if (PageIsNew((PageHeader) page))
elog(PANIC, "btree_split_undo: uninitialized left sibling");
pageop = (BTPageOpaque) PageGetSpecialPointer(page);
if (redo)
{
pageop->btpo_parent = BlockIdGetBlockNumber(&(xlrec->parentblk));
pageop->btpo_prev = BlockIdGetBlockNumber(&(xlrec->leftblk));
if (onleft)
pageop->btpo_next = BlockIdGetBlockNumber(&(xlrec->otherblk));
else
pageop->btpo_next = ItemPointerGetBlockNumber(&(xlrec->target.tid));
pageop->btpo_flags = (isleaf) ? BTP_LEAF : 0;
_bt_restore_page(page, (char *) xlrec + SizeOfBtreeSplit, xlrec->leftlen);
PageSetLSN(page, lsn);
PageSetSUI(page, ThisStartUpID);
UnlockAndWriteBuffer(buffer);
}
else
/* undo */
{
if (XLByteLT(PageGetLSN(page), lsn))
elog(PANIC, "btree_split_undo: bad left sibling LSN");
elog(PANIC, "btree_split_undo: unimplemented");
}
/* Right (new) sibling */
blkno = (onleft) ? BlockIdGetBlockNumber(&(xlrec->otherblk)) :
ItemPointerGetBlockNumber(&(xlrec->target.tid));
buffer = XLogReadBuffer((redo) ? true : false, reln, blkno);
if (!BufferIsValid(buffer))
elog(PANIC, "btree_split_%s: lost right sibling", op);
page = (Page) BufferGetPage(buffer);
if (redo)
_bt_pageinit(page, BufferGetPageSize(buffer));
else if (PageIsNew((PageHeader) page))
elog(PANIC, "btree_split_undo: uninitialized right sibling");
pageop = (BTPageOpaque) PageGetSpecialPointer(page);
if (redo)
{
pageop->btpo_parent = BlockIdGetBlockNumber(&(xlrec->parentblk));
pageop->btpo_prev = (onleft) ?
ItemPointerGetBlockNumber(&(xlrec->target.tid)) :
BlockIdGetBlockNumber(&(xlrec->otherblk));
pageop->btpo_next = BlockIdGetBlockNumber(&(xlrec->rightblk));
pageop->btpo_flags = (isleaf) ? BTP_LEAF : 0;
_bt_restore_page(page,
(char *) xlrec + SizeOfBtreeSplit + xlrec->leftlen,
record->xl_len - SizeOfBtreeSplit - xlrec->leftlen);
PageSetLSN(page, lsn);
PageSetSUI(page, ThisStartUpID);
UnlockAndWriteBuffer(buffer);
}
else
/* undo */
{
if (XLByteLT(PageGetLSN(page), lsn))
elog(PANIC, "btree_split_undo: bad right sibling LSN");
elog(PANIC, "btree_split_undo: unimplemented");
}
if (!redo || (record->xl_info & XLR_BKP_BLOCK_1))
return;
/* Right (next) page */
blkno = BlockIdGetBlockNumber(&(xlrec->rightblk));
if (blkno == P_NONE)
return;
buffer = XLogReadBuffer(false, reln, blkno);
if (!BufferIsValid(buffer))
elog(PANIC, "btree_split_redo: lost next right page");
page = (Page) BufferGetPage(buffer);
if (PageIsNew((PageHeader) page))
elog(PANIC, "btree_split_redo: uninitialized next right page");
if (XLByteLE(lsn, PageGetLSN(page)))
{
UnlockAndReleaseBuffer(buffer);
return;
}
pageop = (BTPageOpaque) PageGetSpecialPointer(page);
pageop->btpo_prev = (onleft) ?
BlockIdGetBlockNumber(&(xlrec->otherblk)) :
ItemPointerGetBlockNumber(&(xlrec->target.tid));
PageSetLSN(page, lsn);
PageSetSUI(page, ThisStartUpID);
UnlockAndWriteBuffer(buffer);
}
static void
btree_xlog_newroot(bool redo, XLogRecPtr lsn, XLogRecord *record)
{
xl_btree_newroot *xlrec = (xl_btree_newroot *) XLogRecGetData(record);
Relation reln;
Buffer buffer;
Page page;
BTPageOpaque pageop;
Buffer metabuf;
Page metapg;
BTMetaPageData md;
if (!redo)
return;
reln = XLogOpenRelation(redo, RM_BTREE_ID, xlrec->node);
if (!RelationIsValid(reln))
return;
buffer = XLogReadBuffer(true, reln, BlockIdGetBlockNumber(&(xlrec->rootblk)));
if (!BufferIsValid(buffer))
elog(PANIC, "btree_newroot_redo: no root page");
metabuf = XLogReadBuffer(false, reln, BTREE_METAPAGE);
if (!BufferIsValid(buffer))
elog(PANIC, "btree_newroot_redo: no metapage");
page = (Page) BufferGetPage(buffer);
_bt_pageinit(page, BufferGetPageSize(buffer));
pageop = (BTPageOpaque) PageGetSpecialPointer(page);
pageop->btpo_flags |= BTP_ROOT;
pageop->btpo_prev = pageop->btpo_next = P_NONE;
pageop->btpo_parent = BTREE_METAPAGE;
if (record->xl_info & XLOG_BTREE_LEAF)
pageop->btpo_flags |= BTP_LEAF;
if (record->xl_len > SizeOfBtreeNewroot)
_bt_restore_page(page,
(char *) xlrec + SizeOfBtreeNewroot,
record->xl_len - SizeOfBtreeNewroot);
PageSetLSN(page, lsn);
PageSetSUI(page, ThisStartUpID);
UnlockAndWriteBuffer(buffer);
metapg = BufferGetPage(metabuf);
_bt_pageinit(metapg, BufferGetPageSize(metabuf));
md.btm_magic = BTREE_MAGIC;
md.btm_version = BTREE_VERSION;
md.btm_root = BlockIdGetBlockNumber(&(xlrec->rootblk));
md.btm_level = xlrec->level;
memcpy((char *) BTPageGetMeta(metapg), (char *) &md, sizeof(md));
pageop = (BTPageOpaque) PageGetSpecialPointer(metapg);
pageop->btpo_flags = BTP_META;
PageSetLSN(metapg, lsn);
PageSetSUI(metapg, ThisStartUpID);
UnlockAndWriteBuffer(metabuf);
}
void
btree_redo(XLogRecPtr lsn, XLogRecord *record)
{
uint8 info = record->xl_info & ~XLR_INFO_MASK;
info &= ~XLOG_BTREE_LEAF;
if (info == XLOG_BTREE_DELETE)
btree_xlog_delete(true, lsn, record);
else if (info == XLOG_BTREE_INSERT)
btree_xlog_insert(true, lsn, record);
else if (info == XLOG_BTREE_SPLIT)
btree_xlog_split(true, false, lsn, record); /* new item on the right */
else if (info == XLOG_BTREE_SPLEFT)
btree_xlog_split(true, true, lsn, record); /* new item on the left */
else if (info == XLOG_BTREE_NEWROOT)
btree_xlog_newroot(true, lsn, record);
else
elog(PANIC, "btree_redo: unknown op code %u", info);
}
void
btree_undo(XLogRecPtr lsn, XLogRecord *record)
{
uint8 info = record->xl_info & ~XLR_INFO_MASK;
info &= ~XLOG_BTREE_LEAF;
if (info == XLOG_BTREE_DELETE)
btree_xlog_delete(false, lsn, record);
else if (info == XLOG_BTREE_INSERT)
btree_xlog_insert(false, lsn, record);
else if (info == XLOG_BTREE_SPLIT)
btree_xlog_split(false, false, lsn, record); /* new item on the right */
else if (info == XLOG_BTREE_SPLEFT)
btree_xlog_split(false, true, lsn, record); /* new item on the left */
else if (info == XLOG_BTREE_NEWROOT)
btree_xlog_newroot(false, lsn, record);
else
elog(PANIC, "btree_undo: unknown op code %u", info);
}
static void
out_target(char *buf, xl_btreetid *target)
{
sprintf(buf + strlen(buf), "node %u/%u; tid %u/%u",
target->node.tblNode, target->node.relNode,
ItemPointerGetBlockNumber(&(target->tid)),
ItemPointerGetOffsetNumber(&(target->tid)));
}
void
btree_desc(char *buf, uint8 xl_info, char *rec)
{
uint8 info = xl_info & ~XLR_INFO_MASK;
info &= ~XLOG_BTREE_LEAF;
if (info == XLOG_BTREE_INSERT)
{
xl_btree_insert *xlrec = (xl_btree_insert *) rec;
strcat(buf, "insert: ");
out_target(buf, &(xlrec->target));
}
else if (info == XLOG_BTREE_DELETE)
{
xl_btree_delete *xlrec = (xl_btree_delete *) rec;
strcat(buf, "delete: ");
out_target(buf, &(xlrec->target));
}
else if (info == XLOG_BTREE_SPLIT || info == XLOG_BTREE_SPLEFT)
{
xl_btree_split *xlrec = (xl_btree_split *) rec;
sprintf(buf + strlen(buf), "split(%s): ",
(info == XLOG_BTREE_SPLIT) ? "right" : "left");
out_target(buf, &(xlrec->target));
sprintf(buf + strlen(buf), "; oth %u; rgh %u",
BlockIdGetBlockNumber(&xlrec->otherblk),
BlockIdGetBlockNumber(&xlrec->rightblk));
}
else if (info == XLOG_BTREE_NEWROOT)
{
xl_btree_newroot *xlrec = (xl_btree_newroot *) rec;
sprintf(buf + strlen(buf), "root: node %u/%u; blk %u",
xlrec->node.tblNode, xlrec->node.relNode,
BlockIdGetBlockNumber(&xlrec->rootblk));
}
else
strcat(buf, "UNKNOWN");
}

133
src/backend/access/nbtree/nbtsearch.c
View File

@ -8,7 +8,7 @@
* Portions Copyright (c) 1994, Regents of the University of California
*
* IDENTIFICATION
* $Header: /cvsroot/pgsql/src/backend/access/nbtree/nbtsearch.c,v 1.72 2002/06/20 20:29:25 momjian Exp $
* $Header: /cvsroot/pgsql/src/backend/access/nbtree/nbtsearch.c,v 1.73 2003/02/21 00:06:21 tgl Exp $
*
*-------------------------------------------------------------------------
*/
@ -895,6 +895,89 @@ _bt_step(IndexScanDesc scan, Buffer *bufP, ScanDirection dir)
return true;
}
/*
* _bt_get_endpoint() -- Find the first or last page on a given tree level
*
* If the index is empty, we will return InvalidBuffer; any other failure
* condition causes elog().
*
* The returned buffer is pinned and read-locked.
*/
Buffer
_bt_get_endpoint(Relation rel, uint32 level, bool rightmost)
{
Buffer buf;
Page page;
BTPageOpaque opaque;
OffsetNumber offnum;
BlockNumber blkno;
BTItem btitem;
IndexTuple itup;
/*
* If we are looking for a leaf page, okay to descend from fast root;
* otherwise better descend from true root. (There is no point in being
* smarter about intermediate levels.)
*/
if (level == 0)
buf = _bt_getroot(rel, BT_READ);
else
buf = _bt_gettrueroot(rel);
if (!BufferIsValid(buf))
{
/* empty index... */
return InvalidBuffer;
}
page = BufferGetPage(buf);
opaque = (BTPageOpaque) PageGetSpecialPointer(page);
for (;;)
{
/*
* If we landed on a deleted page, step right to find a live page
* (there must be one). Also, if we want the rightmost page,
* step right if needed to get to it (this could happen if the
* page split since we obtained a pointer to it).
*/
while (P_ISDELETED(opaque) ||
(rightmost && !P_RIGHTMOST(opaque)))
{
blkno = opaque->btpo_next;
if (blkno == P_NONE)
elog(ERROR, "_bt_get_endpoint: ran off end of btree");
_bt_relbuf(rel, buf);
buf = _bt_getbuf(rel, blkno, BT_READ);
page = BufferGetPage(buf);
opaque = (BTPageOpaque) PageGetSpecialPointer(page);
}
/* Done? */
if (opaque->btpo.level == level)
break;
if (opaque->btpo.level < level)
elog(ERROR, "_bt_get_endpoint: btree level %u not found", level);
/* Step to leftmost or rightmost child page */
if (rightmost)
offnum = PageGetMaxOffsetNumber(page);
else
offnum = P_FIRSTDATAKEY(opaque);
btitem = (BTItem) PageGetItem(page, PageGetItemId(page, offnum));
itup = &(btitem->bti_itup);
blkno = ItemPointerGetBlockNumber(&(itup->t_tid));
_bt_relbuf(rel, buf);
buf = _bt_getbuf(rel, blkno, BT_READ);
page = BufferGetPage(buf);
opaque = (BTPageOpaque) PageGetSpecialPointer(page);
}
return buf;
}
/*
* _bt_endpoint() -- Find the first or last key in the index.
*
@ -910,8 +993,7 @@ _bt_endpoint(IndexScanDesc scan, ScanDirection dir)
Page page;
BTPageOpaque opaque;
ItemPointer current;
OffsetNumber offnum,
maxoff;
OffsetNumber maxoff;
OffsetNumber start;
BlockNumber blkno;
BTItem btitem;
@ -929,7 +1011,7 @@ _bt_endpoint(IndexScanDesc scan, ScanDirection dir)
* simplified version of _bt_search(). We don't maintain a stack
* since we know we won't need it.
*/
buf = _bt_getroot(rel, BT_READ);
buf = _bt_get_endpoint(rel, 0, ScanDirectionIsBackward(dir));
if (!BufferIsValid(buf))
{
@ -942,51 +1024,14 @@ _bt_endpoint(IndexScanDesc scan, ScanDirection dir)
blkno = BufferGetBlockNumber(buf);
page = BufferGetPage(buf);
opaque = (BTPageOpaque) PageGetSpecialPointer(page);
Assert(P_ISLEAF(opaque));
for (;;)
{
if (P_ISLEAF(opaque))
break;
if (ScanDirectionIsForward(dir))
offnum = P_FIRSTDATAKEY(opaque);
else
offnum = PageGetMaxOffsetNumber(page);
btitem = (BTItem) PageGetItem(page, PageGetItemId(page, offnum));
itup = &(btitem->bti_itup);
blkno = ItemPointerGetBlockNumber(&(itup->t_tid));
_bt_relbuf(rel, buf);
buf = _bt_getbuf(rel, blkno, BT_READ);
page = BufferGetPage(buf);
opaque = (BTPageOpaque) PageGetSpecialPointer(page);
/*
* Race condition: If the child page we just stepped onto was just
* split, we need to make sure we're all the way at the right edge
* of the tree. See the paper by Lehman and Yao.
*/
if (ScanDirectionIsBackward(dir) && !P_RIGHTMOST(opaque))
{
do
{
blkno = opaque->btpo_next;
_bt_relbuf(rel, buf);
buf = _bt_getbuf(rel, blkno, BT_READ);
page = BufferGetPage(buf);
opaque = (BTPageOpaque) PageGetSpecialPointer(page);
} while (!P_RIGHTMOST(opaque));
}
}
/* okay, we've got the {left,right}-most page in the tree */
maxoff = PageGetMaxOffsetNumber(page);
if (ScanDirectionIsForward(dir))
{
Assert(P_LEFTMOST(opaque));
/* There could be dead pages to the left, so not this: */
/* Assert(P_LEFTMOST(opaque)); */
start = P_FIRSTDATAKEY(opaque);
}

98
src/backend/access/nbtree/nbtsort.c
View File

@ -35,7 +35,7 @@
* Portions Copyright (c) 1994, Regents of the University of California
*
* IDENTIFICATION
* $Header: /cvsroot/pgsql/src/backend/access/nbtree/nbtsort.c,v 1.70 2002/11/15 01:26:08 momjian Exp $
* $Header: /cvsroot/pgsql/src/backend/access/nbtree/nbtsort.c,v 1.71 2003/02/21 00:06:21 tgl Exp $
*
*-------------------------------------------------------------------------
*/
@ -43,6 +43,7 @@
#include "postgres.h"
#include "access/nbtree.h"
#include "miscadmin.h"
#include "utils/tuplesort.h"
@ -76,7 +77,7 @@ typedef struct BTPageState
BTItem btps_minkey; /* copy of minimum key (first item) on
* page */
OffsetNumber btps_lastoff; /* last item offset loaded */
int btps_level; /* tree level (0 = leaf) */
uint32 btps_level; /* tree level (0 = leaf) */
Size btps_full; /* "full" if less than this much free
* space */
struct BTPageState *btps_next; /* link to parent level, if any */
@ -90,8 +91,9 @@ typedef struct BTPageState
0)
static void _bt_blnewpage(Relation index, Buffer *buf, Page *page, int flags);
static BTPageState *_bt_pagestate(Relation index, int flags, int level);
static void _bt_blnewpage(Relation index, Buffer *buf, Page *page,
uint32 level);
static BTPageState *_bt_pagestate(Relation index, uint32 level);
static void _bt_slideleft(Relation index, Buffer buf, Page page);
static void _bt_sortaddtup(Page page, Size itemsize,
BTItem btitem, OffsetNumber itup_off);
@ -179,7 +181,7 @@ _bt_leafbuild(BTSpool *btspool, BTSpool *btspool2)
* allocate a new, clean btree page, not linked to any siblings.
*/
static void
_bt_blnewpage(Relation index, Buffer *buf, Page *page, int flags)
_bt_blnewpage(Relation index, Buffer *buf, Page *page, uint32 level)
{
BTPageOpaque opaque;
@ -192,23 +194,67 @@ _bt_blnewpage(Relation index, Buffer *buf, Page *page, int flags)
/* Initialize BT opaque state */
opaque = (BTPageOpaque) PageGetSpecialPointer(*page);
opaque->btpo_prev = opaque->btpo_next = P_NONE;
opaque->btpo_flags = flags;
opaque->btpo.level = level;
opaque->btpo_flags = (level > 0) ? 0 : BTP_LEAF;
/* Make the P_HIKEY line pointer appear allocated */
((PageHeader) *page)->pd_lower += sizeof(ItemIdData);
}
/*
* emit a completed btree page, and release the lock and pin on it.
* This is essentially _bt_wrtbuf except we also emit a WAL record.
*/
static void
_bt_blwritepage(Relation index, Buffer buf)
{
Page pg = BufferGetPage(buf);
/* NO ELOG(ERROR) from here till newpage op is logged */
START_CRIT_SECTION();
/* XLOG stuff */
if (!index->rd_istemp)
{
xl_btree_newpage xlrec;
XLogRecPtr recptr;
XLogRecData rdata[2];
xlrec.node = index->rd_node;
xlrec.blkno = BufferGetBlockNumber(buf);
rdata[0].buffer = InvalidBuffer;
rdata[0].data = (char *) &xlrec;
rdata[0].len = SizeOfBtreeNewpage;
rdata[0].next = &(rdata[1]);
rdata[1].buffer = buf;
rdata[1].data = (char *) pg;
rdata[1].len = BLCKSZ;
rdata[1].next = NULL;
recptr = XLogInsert(RM_BTREE_ID, XLOG_BTREE_NEWPAGE, rdata);
PageSetLSN(pg, recptr);
PageSetSUI(pg, ThisStartUpID);
}
END_CRIT_SECTION();
_bt_wrtbuf(index, buf);
}
/*
* allocate and initialize a new BTPageState. the returned structure
* is suitable for immediate use by _bt_buildadd.
*/
static BTPageState *
_bt_pagestate(Relation index, int flags, int level)
_bt_pagestate(Relation index, uint32 level)
{
BTPageState *state = (BTPageState *) palloc0(sizeof(BTPageState));
/* create initial page */
_bt_blnewpage(index, &(state->btps_buf), &(state->btps_page), flags);
_bt_blnewpage(index, &(state->btps_buf), &(state->btps_page), level);
state->btps_minkey = (BTItem) NULL;
/* initialize lastoff so first item goes into P_FIRSTKEY */
@ -365,9 +411,8 @@ _bt_buildadd(Relation index, BTPageState *state, BTItem bti)
ItemId hii;
BTItem obti;
/* Create new page */
_bt_blnewpage(index, &nbuf, &npage,
(state->btps_level > 0) ? 0 : BTP_LEAF);
/* Create new page on same level */
_bt_blnewpage(index, &nbuf, &npage, state->btps_level);
/*
* We copy the last item on the page into the new page, and then
@ -396,10 +441,8 @@ _bt_buildadd(Relation index, BTPageState *state, BTItem bti)
* btree level.
*/
if (state->btps_next == (BTPageState *) NULL)
{
state->btps_next =
_bt_pagestate(index, 0, state->btps_level + 1);
}
state->btps_next = _bt_pagestate(index, state->btps_level + 1);
Assert(state->btps_minkey != NULL);
ItemPointerSet(&(state->btps_minkey->bti_itup.t_tid),
BufferGetBlockNumber(obuf), P_HIKEY);
@ -414,16 +457,7 @@ _bt_buildadd(Relation index, BTPageState *state, BTItem bti)
state->btps_minkey = _bt_formitem(&(obti->bti_itup));
/*
* Set the sibling links for both pages, and parent links too.
*
* It's not necessary to set the parent link at all, because it's
* only used for handling concurrent root splits, but we may as
* well do it as a debugging aid. Note we set new page's link as
* well as old's, because if the new page turns out to be the last
* of the level, _bt_uppershutdown won't change it. The links may
* be out of date by the time the build finishes, but that's OK;
* they need only point to a left-sibling of the true parent. See
* the README file for more info.
* Set the sibling links for both pages.
*/
{
BTPageOpaque oopaque = (BTPageOpaque) PageGetSpecialPointer(opage);
@ -431,9 +465,7 @@ _bt_buildadd(Relation index, BTPageState *state, BTItem bti)
oopaque->btpo_next = BufferGetBlockNumber(nbuf);
nopaque->btpo_prev = BufferGetBlockNumber(obuf);
nopaque->btpo_next = P_NONE;
oopaque->btpo_parent = nopaque->btpo_parent =
BufferGetBlockNumber(state->btps_next->btps_buf);
nopaque->btpo_next = P_NONE; /* redundant */
}
/*
@ -441,7 +473,7 @@ _bt_buildadd(Relation index, BTPageState *state, BTItem bti)
* can give up our lock (if we had one; most likely BuildingBtree
* is set, so we aren't locking).
*/
_bt_wrtbuf(index, obuf);
_bt_blwritepage(index, obuf);
/*
* Reset last_off to point to new page
@ -519,7 +551,7 @@ _bt_uppershutdown(Relation index, BTPageState *state)
* slid back one slot. Then we can dump out the page.
*/
_bt_slideleft(index, s->btps_buf, s->btps_page);
_bt_wrtbuf(index, s->btps_buf);
_bt_blwritepage(index, s->btps_buf);
}
}
@ -603,7 +635,7 @@ _bt_load(Relation index, BTSpool *btspool, BTSpool *btspool2)
/* When we see first tuple, create first index page */
if (state == NULL)
state = _bt_pagestate(index, BTP_LEAF, 0);
state = _bt_pagestate(index, 0);
if (load1)
{
@ -623,13 +655,13 @@ _bt_load(Relation index, BTSpool *btspool, BTSpool *btspool2)
_bt_freeskey(indexScanKey);
}
else
/* merge is unnecessary */
{
/* merge is unnecessary */
while (bti = (BTItem) tuplesort_getindextuple(btspool->sortstate, true, &should_free), bti != (BTItem) NULL)
{
/* When we see first tuple, create first index page */
if (state == NULL)
state = _bt_pagestate(index, BTP_LEAF, 0);
state = _bt_pagestate(index, 0);
_bt_buildadd(index, state, bti);
if (should_free)

780
src/backend/access/nbtree/nbtxlog.c Normal file
View File

@ -0,0 +1,780 @@
/*-------------------------------------------------------------------------
*
* nbtxlog.c
* WAL replay logic for btrees.
*
*
* Portions Copyright (c) 1996-2002, PostgreSQL Global Development Group
* Portions Copyright (c) 1994, Regents of the University of California
*
* IDENTIFICATION
* $Header: /cvsroot/pgsql/src/backend/access/nbtree/nbtxlog.c,v 1.1 2003/02/21 00:06:21 tgl Exp $
*
*-------------------------------------------------------------------------
*/
#include "postgres.h"
#include "access/nbtree.h"
#include "access/xlogutils.h"
/*
* We must keep track of expected insertions due to page splits, and apply
* them manually if they are not seen in the WAL log during replay. This
* makes it safe for page insertion to be a multiple-WAL-action process.
*
* The data structure is a simple linked list --- this should be good enough,
* since we don't expect a page split to remain incomplete for long.
*/
typedef struct bt_incomplete_split
{
RelFileNode node; /* the index */
BlockNumber leftblk; /* left half of split */
BlockNumber rightblk; /* right half of split */
bool is_root; /* we split the root */
} bt_incomplete_split;
static List *incomplete_splits;
static void
log_incomplete_split(RelFileNode node, BlockNumber leftblk,
BlockNumber rightblk, bool is_root)
{
bt_incomplete_split *split = palloc(sizeof(bt_incomplete_split));
split->node = node;
split->leftblk = leftblk;
split->rightblk = rightblk;
split->is_root = is_root;
incomplete_splits = lappend(incomplete_splits, split);
}
static void
forget_matching_split(Relation reln, RelFileNode node,
BlockNumber insertblk, OffsetNumber offnum,
bool is_root)
{
Buffer buffer;
Page page;
BTItem btitem;
BlockNumber rightblk;
List *l;
/* Get downlink TID from page */
buffer = XLogReadBuffer(false, reln, insertblk);
if (!BufferIsValid(buffer))
elog(PANIC, "forget_matching_split: block unfound");
page = (Page) BufferGetPage(buffer);
btitem = (BTItem) PageGetItem(page, PageGetItemId(page, offnum));
rightblk = ItemPointerGetBlockNumber(&(btitem->bti_itup.t_tid));
Assert(ItemPointerGetOffsetNumber(&(btitem->bti_itup.t_tid)) == P_HIKEY);
UnlockAndReleaseBuffer(buffer);
foreach(l, incomplete_splits)
{
bt_incomplete_split *split = (bt_incomplete_split *) lfirst(l);
if (RelFileNodeEquals(node, split->node) &&
rightblk == split->rightblk)
{
if (is_root != split->is_root)
elog(LOG, "forget_matching_split: fishy is_root data");
incomplete_splits = lremove(split, incomplete_splits);
break; /* need not look further */
}
}
}
static void
_bt_restore_page(Page page, char *from, int len)
{
BTItemData btdata;
Size itemsz;
char *end = from + len;
for (; from < end;)
{
memcpy(&btdata, from, sizeof(BTItemData));
itemsz = IndexTupleDSize(btdata.bti_itup) +
(sizeof(BTItemData) - sizeof(IndexTupleData));
itemsz = MAXALIGN(itemsz);
if (PageAddItem(page, (Item) from, itemsz,
FirstOffsetNumber, LP_USED) == InvalidOffsetNumber)
elog(PANIC, "_bt_restore_page: can't add item to page");
from += itemsz;
}
}
static void
_bt_restore_meta(Relation reln, XLogRecPtr lsn,
BlockNumber root, uint32 level,
BlockNumber fastroot, uint32 fastlevel)
{
Buffer metabuf;
Page metapg;
BTMetaPageData *md;
BTPageOpaque pageop;
metabuf = XLogReadBuffer(true, reln, BTREE_METAPAGE);
if (!BufferIsValid(metabuf))
elog(PANIC, "_bt_restore_meta: no metapage");
metapg = BufferGetPage(metabuf);
_bt_pageinit(metapg, BufferGetPageSize(metabuf));
md = BTPageGetMeta(metapg);
md->btm_magic = BTREE_MAGIC;
md->btm_version = BTREE_VERSION;
md->btm_root = root;
md->btm_level = level;
md->btm_fastroot = fastroot;
md->btm_fastlevel = fastlevel;
pageop = (BTPageOpaque) PageGetSpecialPointer(metapg);
pageop->btpo_flags = BTP_META;
PageSetLSN(metapg, lsn);
PageSetSUI(metapg, ThisStartUpID);
UnlockAndWriteBuffer(metabuf);
}
static void
btree_xlog_insert(bool redo, bool isleaf, bool ismeta,
XLogRecPtr lsn, XLogRecord *record)
{
xl_btree_insert *xlrec = (xl_btree_insert *) XLogRecGetData(record);
Relation reln;
Buffer buffer;
Page page;
BTPageOpaque pageop;
char *datapos;
int datalen;
xl_btree_metadata md;
datapos = (char *) xlrec + SizeOfBtreeInsert;
datalen = record->xl_len - SizeOfBtreeInsert;
if (ismeta)
{
memcpy(&md, datapos, sizeof(xl_btree_metadata));
datapos += sizeof(xl_btree_metadata);
datalen -= sizeof(xl_btree_metadata);
}
if (redo && (record->xl_info & XLR_BKP_BLOCK_1) && !ismeta &&
incomplete_splits == NIL)
return; /* nothing to do */
reln = XLogOpenRelation(redo, RM_BTREE_ID, xlrec->target.node);
if (!RelationIsValid(reln))
return;
if (!redo || !(record->xl_info & XLR_BKP_BLOCK_1))
{
buffer = XLogReadBuffer(false, reln,
ItemPointerGetBlockNumber(&(xlrec->target.tid)));
if (!BufferIsValid(buffer))
elog(PANIC, "btree_insert_%sdo: block unfound", (redo) ? "re" : "un");
page = (Page) BufferGetPage(buffer);
if (PageIsNew((PageHeader) page))
elog(PANIC, "btree_insert_%sdo: uninitialized page", (redo) ? "re" : "un");
pageop = (BTPageOpaque) PageGetSpecialPointer(page);
if (redo)
{
if (XLByteLE(lsn, PageGetLSN(page)))
{
UnlockAndReleaseBuffer(buffer);
}
else
{
if (PageAddItem(page, (Item) datapos, datalen,
ItemPointerGetOffsetNumber(&(xlrec->target.tid)),
LP_USED) == InvalidOffsetNumber)
elog(PANIC, "btree_insert_redo: failed to add item");
PageSetLSN(page, lsn);
PageSetSUI(page, ThisStartUpID);
UnlockAndWriteBuffer(buffer);
}
}
else
{
if (XLByteLT(PageGetLSN(page), lsn))
elog(PANIC, "btree_insert_undo: bad page LSN");
if (!P_ISLEAF(pageop))
{
UnlockAndReleaseBuffer(buffer);
}
else
{
elog(PANIC, "btree_insert_undo: unimplemented");
}
}
}
if (redo) /* metapage changes not undoable */
{
if (ismeta)
_bt_restore_meta(reln, lsn,
md.root, md.level,
md.fastroot, md.fastlevel);
}
/* Forget any split this insertion completes */
if (redo && !isleaf && incomplete_splits != NIL)
{
forget_matching_split(reln, xlrec->target.node,
ItemPointerGetBlockNumber(&(xlrec->target.tid)),
ItemPointerGetOffsetNumber(&(xlrec->target.tid)),
false);
}
}
static void
btree_xlog_split(bool redo, bool onleft, bool isroot,
XLogRecPtr lsn, XLogRecord *record)
{
xl_btree_split *xlrec = (xl_btree_split *) XLogRecGetData(record);
Relation reln;
BlockNumber targetblk;
BlockNumber leftsib;
BlockNumber rightsib;
Buffer buffer;
Page page;
BTPageOpaque pageop;
char *op = (redo) ? "redo" : "undo";
reln = XLogOpenRelation(redo, RM_BTREE_ID, xlrec->target.node);
if (!RelationIsValid(reln))
return;
targetblk = ItemPointerGetBlockNumber(&(xlrec->target.tid));
leftsib = (onleft) ? targetblk : xlrec->otherblk;
rightsib = (onleft) ? xlrec->otherblk : targetblk;
/* Left (original) sibling */
buffer = XLogReadBuffer(false, reln, leftsib);
if (!BufferIsValid(buffer))
elog(PANIC, "btree_split_%s: lost left sibling", op);
page = (Page) BufferGetPage(buffer);
if (redo)
_bt_pageinit(page, BufferGetPageSize(buffer));
else if (PageIsNew((PageHeader) page))
elog(PANIC, "btree_split_undo: uninitialized left sibling");
pageop = (BTPageOpaque) PageGetSpecialPointer(page);
if (redo)
{
pageop->btpo_prev = xlrec->leftblk;
pageop->btpo_next = rightsib;
pageop->btpo.level = xlrec->level;
pageop->btpo_flags = (xlrec->level == 0) ? BTP_LEAF : 0;
_bt_restore_page(page,
(char *) xlrec + SizeOfBtreeSplit,
xlrec->leftlen);
PageSetLSN(page, lsn);
PageSetSUI(page, ThisStartUpID);
UnlockAndWriteBuffer(buffer);
}
else
{
/* undo */
if (XLByteLT(PageGetLSN(page), lsn))
elog(PANIC, "btree_split_undo: bad left sibling LSN");
elog(PANIC, "btree_split_undo: unimplemented");
}
/* Right (new) sibling */
buffer = XLogReadBuffer((redo) ? true : false, reln, rightsib);
if (!BufferIsValid(buffer))
elog(PANIC, "btree_split_%s: lost right sibling", op);
page = (Page) BufferGetPage(buffer);
if (redo)
_bt_pageinit(page, BufferGetPageSize(buffer));
else if (PageIsNew((PageHeader) page))
elog(PANIC, "btree_split_undo: uninitialized right sibling");
pageop = (BTPageOpaque) PageGetSpecialPointer(page);
if (redo)
{
pageop->btpo_prev = leftsib;
pageop->btpo_next = xlrec->rightblk;
pageop->btpo.level = xlrec->level;
pageop->btpo_flags = (xlrec->level == 0) ? BTP_LEAF : 0;
_bt_restore_page(page,
(char *) xlrec + SizeOfBtreeSplit + xlrec->leftlen,
record->xl_len - SizeOfBtreeSplit - xlrec->leftlen);
PageSetLSN(page, lsn);
PageSetSUI(page, ThisStartUpID);
UnlockAndWriteBuffer(buffer);
}
else
{
/* undo */
if (XLByteLT(PageGetLSN(page), lsn))
elog(PANIC, "btree_split_undo: bad right sibling LSN");
elog(PANIC, "btree_split_undo: unimplemented");
}
/* Fix left-link of right (next) page */
if (redo && !(record->xl_info & XLR_BKP_BLOCK_1))
{
if (xlrec->rightblk != P_NONE)
{
buffer = XLogReadBuffer(false, reln, xlrec->rightblk);
if (!BufferIsValid(buffer))
elog(PANIC, "btree_split_redo: lost next right page");
page = (Page) BufferGetPage(buffer);
if (PageIsNew((PageHeader) page))
elog(PANIC, "btree_split_redo: uninitialized next right page");
if (XLByteLE(lsn, PageGetLSN(page)))
{
UnlockAndReleaseBuffer(buffer);
}
else
{
pageop = (BTPageOpaque) PageGetSpecialPointer(page);
pageop->btpo_prev = rightsib;
PageSetLSN(page, lsn);
PageSetSUI(page, ThisStartUpID);
UnlockAndWriteBuffer(buffer);
}
}
}
/* Forget any split this insertion completes */
if (redo && xlrec->level > 0 && incomplete_splits != NIL)
{
forget_matching_split(reln, xlrec->target.node,
ItemPointerGetBlockNumber(&(xlrec->target.tid)),
ItemPointerGetOffsetNumber(&(xlrec->target.tid)),
false);
}
/* The job ain't done till the parent link is inserted... */
log_incomplete_split(xlrec->target.node,
leftsib, rightsib, isroot);
}
static void
btree_xlog_delete(bool redo, XLogRecPtr lsn, XLogRecord *record)
{
xl_btree_delete *xlrec;
Relation reln;
Buffer buffer;
Page page;
if (!redo || (record->xl_info & XLR_BKP_BLOCK_1))
return;
xlrec = (xl_btree_delete *) XLogRecGetData(record);
reln = XLogOpenRelation(redo, RM_BTREE_ID, xlrec->target.node);
if (!RelationIsValid(reln))
return;
buffer = XLogReadBuffer(false, reln,
ItemPointerGetBlockNumber(&(xlrec->target.tid)));
if (!BufferIsValid(buffer))
elog(PANIC, "btree_delete_redo: block unfound");
page = (Page) BufferGetPage(buffer);
if (PageIsNew((PageHeader) page))
elog(PANIC, "btree_delete_redo: uninitialized page");
if (XLByteLE(lsn, PageGetLSN(page)))
{
UnlockAndReleaseBuffer(buffer);
return;
}
PageIndexTupleDelete(page, ItemPointerGetOffsetNumber(&(xlrec->target.tid)));
PageSetLSN(page, lsn);
PageSetSUI(page, ThisStartUpID);
UnlockAndWriteBuffer(buffer);
}
static void
btree_xlog_newroot(bool redo, XLogRecPtr lsn, XLogRecord *record)
{
xl_btree_newroot *xlrec = (xl_btree_newroot *) XLogRecGetData(record);
Relation reln;
Buffer buffer;
Page page;
BTPageOpaque pageop;
if (!redo)
return; /* not undoable */
reln = XLogOpenRelation(redo, RM_BTREE_ID, xlrec->node);
if (!RelationIsValid(reln))
return;
buffer = XLogReadBuffer(true, reln, xlrec->rootblk);
if (!BufferIsValid(buffer))
elog(PANIC, "btree_newroot_redo: no root page");
page = (Page) BufferGetPage(buffer);
_bt_pageinit(page, BufferGetPageSize(buffer));
pageop = (BTPageOpaque) PageGetSpecialPointer(page);
pageop->btpo_flags = BTP_ROOT;
pageop->btpo_prev = pageop->btpo_next = P_NONE;
pageop->btpo.level = xlrec->level;
if (xlrec->level == 0)
pageop->btpo_flags |= BTP_LEAF;
if (record->xl_len > SizeOfBtreeNewroot)
_bt_restore_page(page,
(char *) xlrec + SizeOfBtreeNewroot,
record->xl_len - SizeOfBtreeNewroot);
PageSetLSN(page, lsn);
PageSetSUI(page, ThisStartUpID);
UnlockAndWriteBuffer(buffer);
_bt_restore_meta(reln, lsn,
xlrec->rootblk, xlrec->level,
xlrec->rootblk, xlrec->level);
/* Check to see if this satisfies any incomplete insertions */
if (record->xl_len > SizeOfBtreeNewroot &&
incomplete_splits != NIL)
{
forget_matching_split(reln, xlrec->node,
xlrec->rootblk,
P_FIRSTKEY,
true);
}
}
static void
btree_xlog_newmeta(bool redo, XLogRecPtr lsn, XLogRecord *record)
{
xl_btree_newmeta *xlrec = (xl_btree_newmeta *) XLogRecGetData(record);
Relation reln;
if (!redo)
return; /* not undoable */
reln = XLogOpenRelation(redo, RM_BTREE_ID, xlrec->node);
if (!RelationIsValid(reln))
return;
_bt_restore_meta(reln, lsn,
xlrec->meta.root, xlrec->meta.level,
xlrec->meta.fastroot, xlrec->meta.fastlevel);
}
static void
btree_xlog_newpage(bool redo, XLogRecPtr lsn, XLogRecord *record)
{
xl_btree_newpage *xlrec = (xl_btree_newpage *) XLogRecGetData(record);
Relation reln;
Buffer buffer;
Page page;
if (!redo || (record->xl_info & XLR_BKP_BLOCK_1))
return;
reln = XLogOpenRelation(redo, RM_BTREE_ID, xlrec->node);
if (!RelationIsValid(reln))
return;
buffer = XLogReadBuffer(true, reln, xlrec->blkno);
if (!BufferIsValid(buffer))
elog(PANIC, "btree_newpage_redo: block unfound");
page = (Page) BufferGetPage(buffer);
Assert(record->xl_len == SizeOfBtreeNewpage + BLCKSZ);
memcpy(page, (char *) xlrec + SizeOfBtreeNewpage, BLCKSZ);
PageSetLSN(page, lsn);
PageSetSUI(page, ThisStartUpID);
UnlockAndWriteBuffer(buffer);
}
void
btree_redo(XLogRecPtr lsn, XLogRecord *record)
{
uint8 info = record->xl_info & ~XLR_INFO_MASK;
switch (info)
{
case XLOG_BTREE_INSERT_LEAF:
btree_xlog_insert(true, true, false, lsn, record);
break;
case XLOG_BTREE_INSERT_UPPER:
btree_xlog_insert(true, false, false, lsn, record);
break;
case XLOG_BTREE_INSERT_META:
btree_xlog_insert(true, false, true, lsn, record);
break;
case XLOG_BTREE_SPLIT_L:
btree_xlog_split(true, true, false, lsn, record);
break;
case XLOG_BTREE_SPLIT_R:
btree_xlog_split(true, false, false, lsn, record);
break;
case XLOG_BTREE_SPLIT_L_ROOT:
btree_xlog_split(true, true, true, lsn, record);
break;
case XLOG_BTREE_SPLIT_R_ROOT:
btree_xlog_split(true, false, true, lsn, record);
break;
case XLOG_BTREE_DELETE:
btree_xlog_delete(true, lsn, record);
break;
case XLOG_BTREE_DELETE_PAGE:
case XLOG_BTREE_DELETE_PAGE_META:
// ???
break;
case XLOG_BTREE_NEWROOT:
btree_xlog_newroot(true, lsn, record);
break;
case XLOG_BTREE_NEWMETA:
btree_xlog_newmeta(true, lsn, record);
break;
case XLOG_BTREE_NEWPAGE:
btree_xlog_newpage(true, lsn, record);
break;
default:
elog(PANIC, "btree_redo: unknown op code %u", info);
}
}
void
btree_undo(XLogRecPtr lsn, XLogRecord *record)
{
uint8 info = record->xl_info & ~XLR_INFO_MASK;
switch (info)
{
case XLOG_BTREE_INSERT_LEAF:
btree_xlog_insert(false, true, false, lsn, record);
break;
case XLOG_BTREE_INSERT_UPPER:
btree_xlog_insert(false, false, false, lsn, record);
break;
case XLOG_BTREE_INSERT_META:
btree_xlog_insert(false, false, true, lsn, record);
break;
case XLOG_BTREE_SPLIT_L:
btree_xlog_split(false, true, false, lsn, record);
break;
case XLOG_BTREE_SPLIT_R:
btree_xlog_split(false, false, false, lsn, record);
break;
case XLOG_BTREE_SPLIT_L_ROOT:
btree_xlog_split(false, true, true, lsn, record);
break;
case XLOG_BTREE_SPLIT_R_ROOT:
btree_xlog_split(false, false, true, lsn, record);
break;
case XLOG_BTREE_DELETE:
btree_xlog_delete(false, lsn, record);
break;
case XLOG_BTREE_DELETE_PAGE:
case XLOG_BTREE_DELETE_PAGE_META:
// ???
break;
case XLOG_BTREE_NEWROOT:
btree_xlog_newroot(false, lsn, record);
break;
case XLOG_BTREE_NEWMETA:
btree_xlog_newmeta(false, lsn, record);
break;
case XLOG_BTREE_NEWPAGE:
btree_xlog_newpage(false, lsn, record);
break;
default:
elog(PANIC, "btree_undo: unknown op code %u", info);
}
}
static void
out_target(char *buf, xl_btreetid *target)
{
sprintf(buf + strlen(buf), "node %u/%u; tid %u/%u",
target->node.tblNode, target->node.relNode,
ItemPointerGetBlockNumber(&(target->tid)),
ItemPointerGetOffsetNumber(&(target->tid)));
}
void
btree_desc(char *buf, uint8 xl_info, char *rec)
{
uint8 info = xl_info & ~XLR_INFO_MASK;
switch (info)
{
case XLOG_BTREE_INSERT_LEAF:
{
xl_btree_insert *xlrec = (xl_btree_insert *) rec;
strcat(buf, "insert: ");
out_target(buf, &(xlrec->target));
break;
}
case XLOG_BTREE_INSERT_UPPER:
{
xl_btree_insert *xlrec = (xl_btree_insert *) rec;
strcat(buf, "insert_upper: ");
out_target(buf, &(xlrec->target));
break;
}
case XLOG_BTREE_INSERT_META:
{
xl_btree_insert *xlrec = (xl_btree_insert *) rec;
strcat(buf, "insert_meta: ");
out_target(buf, &(xlrec->target));
break;
}
case XLOG_BTREE_SPLIT_L:
{
xl_btree_split *xlrec = (xl_btree_split *) rec;
strcat(buf, "split_l: ");
out_target(buf, &(xlrec->target));
sprintf(buf + strlen(buf), "; oth %u; rgh %u",
xlrec->otherblk, xlrec->rightblk);
break;
}
case XLOG_BTREE_SPLIT_R:
{
xl_btree_split *xlrec = (xl_btree_split *) rec;
strcat(buf, "split_r: ");
out_target(buf, &(xlrec->target));
sprintf(buf + strlen(buf), "; oth %u; rgh %u",
xlrec->otherblk, xlrec->rightblk);
break;
}
case XLOG_BTREE_SPLIT_L_ROOT:
{
xl_btree_split *xlrec = (xl_btree_split *) rec;
strcat(buf, "split_l_root: ");
out_target(buf, &(xlrec->target));
sprintf(buf + strlen(buf), "; oth %u; rgh %u",
xlrec->otherblk, xlrec->rightblk);
break;
}
case XLOG_BTREE_SPLIT_R_ROOT:
{
xl_btree_split *xlrec = (xl_btree_split *) rec;
strcat(buf, "split_r_root: ");
out_target(buf, &(xlrec->target));
sprintf(buf + strlen(buf), "; oth %u; rgh %u",
xlrec->otherblk, xlrec->rightblk);
break;
}
case XLOG_BTREE_DELETE:
{
xl_btree_delete *xlrec = (xl_btree_delete *) rec;
strcat(buf, "delete: ");
out_target(buf, &(xlrec->target));
break;
}
case XLOG_BTREE_DELETE_PAGE:
case XLOG_BTREE_DELETE_PAGE_META:
{
xl_btree_delete_page *xlrec = (xl_btree_delete_page *) rec;
strcat(buf, "delete_page: ");
out_target(buf, &(xlrec->target));
sprintf(buf + strlen(buf), "; dead %u; left %u; right %u",
xlrec->deadblk, xlrec->leftblk, xlrec->rightblk);
break;
}
case XLOG_BTREE_NEWROOT:
{
xl_btree_newroot *xlrec = (xl_btree_newroot *) rec;
sprintf(buf + strlen(buf), "newroot: node %u/%u; root %u lev %u",
xlrec->node.tblNode, xlrec->node.relNode,
xlrec->rootblk, xlrec->level);
break;
}
case XLOG_BTREE_NEWMETA:
{
xl_btree_newmeta *xlrec = (xl_btree_newmeta *) rec;
sprintf(buf + strlen(buf), "newmeta: node %u/%u; root %u lev %u fast %u lev %u",
xlrec->node.tblNode, xlrec->node.relNode,
xlrec->meta.root, xlrec->meta.level,
xlrec->meta.fastroot, xlrec->meta.fastlevel);
break;
}
case XLOG_BTREE_NEWPAGE:
{
xl_btree_newpage *xlrec = (xl_btree_newpage *) rec;
sprintf(buf + strlen(buf), "newpage: node %u/%u; page %u",
xlrec->node.tblNode, xlrec->node.relNode,
xlrec->blkno);
break;
}
default:
strcat(buf, "UNKNOWN");
break;
}
}
void
btree_xlog_startup(void)
{
incomplete_splits = NIL;
}
void
btree_xlog_cleanup(void)
{
List *l;
foreach(l, incomplete_splits)
{
bt_incomplete_split *split = (bt_incomplete_split *) lfirst(l);
Relation reln;
Buffer lbuf,
rbuf;
Page lpage,
rpage;
BTPageOpaque lpageop,
rpageop;
bool is_only;
reln = XLogOpenRelation(true, RM_BTREE_ID, split->node);
if (!RelationIsValid(reln))
continue;
lbuf = XLogReadBuffer(false, reln, split->leftblk);
if (!BufferIsValid(lbuf))
elog(PANIC, "btree_xlog_cleanup: left block unfound");
lpage = (Page) BufferGetPage(lbuf);
lpageop = (BTPageOpaque) PageGetSpecialPointer(lpage);
rbuf = XLogReadBuffer(false, reln, split->rightblk);
if (!BufferIsValid(rbuf))
elog(PANIC, "btree_xlog_cleanup: right block unfound");
rpage = (Page) BufferGetPage(rbuf);
rpageop = (BTPageOpaque) PageGetSpecialPointer(rpage);
/* if the two pages are all of their level, it's a only-page split */
is_only = P_LEFTMOST(lpageop) && P_RIGHTMOST(rpageop);
_bt_insert_parent(reln, lbuf, rbuf, (BTStack) NULL,
split->is_root, is_only);
}
incomplete_splits = NIL;
}

37
src/backend/access/transam/rmgr.c
View File

@ -3,7 +3,7 @@
*
* Resource managers definition
*
* $Header: /cvsroot/pgsql/src/backend/access/transam/rmgr.c,v 1.9 2001/08/25 18:52:41 tgl Exp $
* $Header: /cvsroot/pgsql/src/backend/access/transam/rmgr.c,v 1.10 2003/02/21 00:06:22 tgl Exp $
*/
#include "postgres.h"
@ -19,21 +19,22 @@
#include "commands/sequence.h"
RmgrData RmgrTable[] = {
{"XLOG", xlog_redo, xlog_undo, xlog_desc},
{"Transaction", xact_redo, xact_undo, xact_desc},
{"Storage", smgr_redo, smgr_undo, smgr_desc},
{"CLOG", clog_redo, clog_undo, clog_desc},
{"Reserved 4", NULL, NULL, NULL},
{"Reserved 5", NULL, NULL, NULL},
{"Reserved 6", NULL, NULL, NULL},
{"Reserved 7", NULL, NULL, NULL},
{"Reserved 8", NULL, NULL, NULL},
{"Reserved 9", NULL, NULL, NULL},
{"Heap", heap_redo, heap_undo, heap_desc},
{"Btree", btree_redo, btree_undo, btree_desc},
{"Hash", hash_redo, hash_undo, hash_desc},
{"Rtree", rtree_redo, rtree_undo, rtree_desc},
{"Gist", gist_redo, gist_undo, gist_desc},
{"Sequence", seq_redo, seq_undo, seq_desc}
RmgrData RmgrTable[RM_MAX_ID+1] = {
{"XLOG", xlog_redo, xlog_undo, xlog_desc, NULL, NULL},
{"Transaction", xact_redo, xact_undo, xact_desc, NULL, NULL},
{"Storage", smgr_redo, smgr_undo, smgr_desc, NULL, NULL},
{"CLOG", clog_redo, clog_undo, clog_desc, NULL, NULL},
{"Reserved 4", NULL, NULL, NULL, NULL, NULL},
{"Reserved 5", NULL, NULL, NULL, NULL, NULL},
{"Reserved 6", NULL, NULL, NULL, NULL, NULL},
{"Reserved 7", NULL, NULL, NULL, NULL, NULL},
{"Reserved 8", NULL, NULL, NULL, NULL, NULL},
{"Reserved 9", NULL, NULL, NULL, NULL, NULL},
{"Heap", heap_redo, heap_undo, heap_desc, NULL, NULL},
{"Btree", btree_redo, btree_undo, btree_desc,
btree_xlog_startup, btree_xlog_cleanup},
{"Hash", hash_redo, hash_undo, hash_desc, NULL, NULL},
{"Rtree", rtree_redo, rtree_undo, rtree_desc, NULL, NULL},
{"Gist", gist_redo, gist_undo, gist_desc, NULL, NULL},
{"Sequence", seq_redo, seq_undo, seq_desc, NULL, NULL}
};

58
src/backend/access/transam/xlog.c
View File

@ -7,7 +7,7 @@
* Portions Copyright (c) 1996-2002, PostgreSQL Global Development Group
* Portions Copyright (c) 1994, Regents of the University of California
*
* $Header: /cvsroot/pgsql/src/backend/access/transam/xlog.c,v 1.111 2003/01/25 03:06:04 tgl Exp $
* $Header: /cvsroot/pgsql/src/backend/access/transam/xlog.c,v 1.112 2003/02/21 00:06:22 tgl Exp $
*
*-------------------------------------------------------------------------
*/
@ -1203,16 +1203,6 @@ XLogFlush(XLogRecPtr record)
XLogRecPtr WriteRqstPtr;
XLogwrtRqst WriteRqst;
if (XLOG_DEBUG)
{
elog(LOG, "XLogFlush%s%s: request %X/%X; write %X/%X; flush %X/%X",
(IsBootstrapProcessingMode()) ? "(bootstrap)" : "",
(InRedo) ? "(redo)" : "",
record.xlogid, record.xrecoff,
LogwrtResult.Write.xlogid, LogwrtResult.Write.xrecoff,
LogwrtResult.Flush.xlogid, LogwrtResult.Flush.xrecoff);
}
/* Disabled during REDO */
if (InRedo)
return;
@ -1221,6 +1211,15 @@ XLogFlush(XLogRecPtr record)
if (XLByteLE(record, LogwrtResult.Flush))
return;
if (XLOG_DEBUG)
{
elog(LOG, "XLogFlush%s: request %X/%X; write %X/%X; flush %X/%X",
(IsBootstrapProcessingMode()) ? "(bootstrap)" : "",
record.xlogid, record.xrecoff,
LogwrtResult.Write.xlogid, LogwrtResult.Write.xrecoff,
LogwrtResult.Flush.xlogid, LogwrtResult.Flush.xrecoff);
}
START_CRIT_SECTION();
/*
@ -2515,6 +2514,12 @@ StartupXLOG(void)
elog(LOG, "database system was interrupted at %s",
str_time(ControlFile->time));
/* This is just to allow attaching to startup process with a debugger */
#ifdef XLOG_REPLAY_DELAY
if (XLOG_DEBUG && ControlFile->state != DB_SHUTDOWNED)
sleep(60);
#endif
/*
* Get the last valid checkpoint record. If the latest one according
* to pg_control is broken, try the next-to-last one.
@ -2578,14 +2583,23 @@ StartupXLOG(void)
/* REDO */
if (InRecovery)
{
int rmid;
elog(LOG, "database system was not properly shut down; "
"automatic recovery in progress");
ControlFile->state = DB_IN_RECOVERY;
ControlFile->time = time(NULL);
UpdateControlFile();
/* Start up the recovery environment */
XLogInitRelationCache();
for (rmid = 0; rmid <= RM_MAX_ID; rmid++)
{
if (RmgrTable[rmid].rm_startup != NULL)
RmgrTable[rmid].rm_startup();
}
/* Is REDO required ? */
if (XLByteLT(checkPoint.redo, RecPtr))
record = ReadRecord(&(checkPoint.redo), PANIC, buffer);
@ -2737,7 +2751,25 @@ StartupXLOG(void)
if (InRecovery)
{
int rmid;
/*
* Allow resource managers to do any required cleanup.
*/
for (rmid = 0; rmid <= RM_MAX_ID; rmid++)
{
if (RmgrTable[rmid].rm_cleanup != NULL)
RmgrTable[rmid].rm_cleanup();
}
/* suppress in-transaction check in CreateCheckPoint */
MyLastRecPtr.xrecoff = 0;
MyXactMadeXLogEntry = false;
MyXactMadeTempRelUpdate = false;
/*
* Perform a new checkpoint to update our recovery activity to disk.
*
* In case we had to use the secondary checkpoint, make sure that
* it will still be shown as the secondary checkpoint after this
* CreateCheckPoint operation; we don't want the broken primary
@ -2745,6 +2777,10 @@ StartupXLOG(void)
*/
ControlFile->checkPoint = checkPointLoc;
CreateCheckPoint(true, true);
/*
* Close down recovery environment
*/
XLogCloseRelationCache();
}

570
src/include/access/nbtree.h
View File

@ -7,7 +7,7 @@
* Portions Copyright (c) 1996-2002, PostgreSQL Global Development Group
* Portions Copyright (c) 1994, Regents of the University of California
*
* $Id: nbtree.h,v 1.63 2002/07/02 05:48:44 momjian Exp $
* $Id: nbtree.h,v 1.64 2003/02/21 00:06:22 tgl Exp $
*
*-------------------------------------------------------------------------
*/
@ -22,46 +22,55 @@
/*
* BTPageOpaqueData -- At the end of every page, we store a pointer
* to both siblings in the tree. This is used to do forward/backward
* index scans. See Lehman and Yao's paper for more
* info. In addition, we need to know what type of page this is
* (leaf or internal), and whether the page is available for reuse.
* index scans. The next-page link is also critical for recovery when
* a search has navigated to the wrong page due to concurrent page splits
* or deletions; see src/backend/access/nbtree/README for more info.
*
* We also store a back-link to the parent page, but this cannot be trusted
* very far since it does not get updated when the parent is split.
* See backend/access/nbtree/README for details.
* In addition, we store the page's btree level (counting upwards from
* zero at a leaf page) as well as some flag bits indicating the page type
* and status. If the page is deleted, we replace the level with the
* next-transaction-ID value indicating when it is safe to reclaim the page.
*
* NOTE: the BTP_LEAF flag bit is redundant since level==0 could be tested
* instead.
*/
typedef struct BTPageOpaqueData
{
BlockNumber btpo_prev; /* used for backward index scans */
BlockNumber btpo_next; /* used for forward index scans */
BlockNumber btpo_parent; /* pointer to parent, but not updated on
* parent split */
uint16 btpo_flags; /* LEAF?, ROOT?, FREE?, META?, REORDER? */
BlockNumber btpo_prev; /* left sibling, or P_NONE if leftmost */
BlockNumber btpo_next; /* right sibling, or P_NONE if rightmost */
union
{
uint32 level; /* tree level --- zero for leaf pages */
TransactionId xact; /* next transaction ID, if deleted */
} btpo;
uint16 btpo_flags; /* flag bits, see below */
} BTPageOpaqueData;
typedef BTPageOpaqueData *BTPageOpaque;
/* Bits defined in btpo_flags */
#define BTP_LEAF (1 << 0) /* leaf page, if not internal page */
#define BTP_LEAF (1 << 0) /* leaf page, i.e. not internal page */
#define BTP_ROOT (1 << 1) /* root page (has no parent) */
#define BTP_FREE (1 << 2) /* page not in use */
#define BTP_DELETED (1 << 2) /* page has been deleted from tree */
#define BTP_META (1 << 3) /* meta-page */
#define BTP_REORDER (1 << 4) /* items need reordering */
/*
* The Meta page is always the first page in the btree index.
* Its primary purpose is to point to the location of the btree root page.
* We also point to the "fast" root, which is the current effective root;
* see README for discussion.
*/
typedef struct BTMetaPageData
{
uint32 btm_magic;
uint32 btm_version;
BlockNumber btm_root;
int32 btm_level;
uint32 btm_magic; /* should contain BTREE_MAGIC */
uint32 btm_version; /* should contain BTREE_VERSION */
BlockNumber btm_root; /* current root location */
uint32 btm_level; /* tree level of the root page */
BlockNumber btm_fastroot; /* current "fast" root location */
uint32 btm_fastlevel; /* tree level of the "fast" root page */
} BTMetaPageData;
#define BTPageGetMeta(p) \
@ -69,12 +78,7 @@ typedef struct BTMetaPageData
#define BTREE_METAPAGE 0 /* first page is meta */
#define BTREE_MAGIC 0x053162 /* magic number of btree pages */
#define BTreeInvalidParent(opaque) \
(opaque->btpo_parent == InvalidBlockNumber || \
opaque->btpo_parent == BTREE_METAPAGE)
#define BTREE_VERSION 1
#define BTREE_VERSION 2 /* current version number */
/*
* We actually need to be able to fit three items on every page,
@ -84,6 +88,295 @@ typedef struct BTMetaPageData
((PageGetPageSize(page) - \
sizeof(PageHeaderData) - \
MAXALIGN(sizeof(BTPageOpaqueData))) / 3 - sizeof(ItemIdData))
/*
* BTItems are what we store in the btree. Each item is an index tuple,
* including key and pointer values. (In some cases either the key or the
* pointer may go unused, see backend/access/nbtree/README for details.)
*
* Old comments:
* In addition, we must guarantee that all tuples in the index are unique,
* in order to satisfy some assumptions in Lehman and Yao. The way that we
* do this is by generating a new OID for every insertion that we do in the
* tree. This adds eight bytes to the size of btree index tuples. Note
* that we do not use the OID as part of a composite key; the OID only
* serves as a unique identifier for a given index tuple (logical position
* within a page).
*
* New comments:
* actually, we must guarantee that all tuples in A LEVEL
* are unique, not in ALL INDEX. So, we can use bti_itup->t_tid
* as unique identifier for a given index tuple (logical position
* within a level). - vadim 04/09/97
*/
typedef struct BTItemData
{
IndexTupleData bti_itup;
} BTItemData;
typedef BTItemData *BTItem;
/*
* For XLOG: size without alignment. Sizeof works as long as
* IndexTupleData has exactly 8 bytes.
*/
#define SizeOfBTItem sizeof(BTItemData)
/* Test whether items are the "same" per the above notes */
#define BTItemSame(i1, i2) ( (i1)->bti_itup.t_tid.ip_blkid.bi_hi == \
(i2)->bti_itup.t_tid.ip_blkid.bi_hi && \
(i1)->bti_itup.t_tid.ip_blkid.bi_lo == \
(i2)->bti_itup.t_tid.ip_blkid.bi_lo && \
(i1)->bti_itup.t_tid.ip_posid == \
(i2)->bti_itup.t_tid.ip_posid )
/*
* In general, the btree code tries to localize its knowledge about
* page layout to a couple of routines. However, we need a special
* value to indicate "no page number" in those places where we expect
* page numbers. We can use zero for this because we never need to
* make a pointer to the metadata page.
*/
#define P_NONE 0
/*
* Macros to test whether a page is leftmost or rightmost on its tree level,
* as well as other state info kept in the opaque data.
*/
#define P_LEFTMOST(opaque) ((opaque)->btpo_prev == P_NONE)
#define P_RIGHTMOST(opaque) ((opaque)->btpo_next == P_NONE)
#define P_ISLEAF(opaque) ((opaque)->btpo_flags & BTP_LEAF)
#define P_ISROOT(opaque) ((opaque)->btpo_flags & BTP_ROOT)
#define P_ISDELETED(opaque) ((opaque)->btpo_flags & BTP_DELETED)
/*
* Lehman and Yao's algorithm requires a ``high key'' on every non-rightmost
* page. The high key is not a data key, but gives info about what range of
* keys is supposed to be on this page. The high key on a page is required
* to be greater than or equal to any data key that appears on the page.
* If we find ourselves trying to insert a key > high key, we know we need
* to move right (this should only happen if the page was split since we
* examined the parent page).
*
* Our insertion algorithm guarantees that we can use the initial least key
* on our right sibling as the high key. Once a page is created, its high
* key changes only if the page is split.
*
* On a non-rightmost page, the high key lives in item 1 and data items
* start in item 2. Rightmost pages have no high key, so we store data
* items beginning in item 1.
*/
#define P_HIKEY ((OffsetNumber) 1)
#define P_FIRSTKEY ((OffsetNumber) 2)
#define P_FIRSTDATAKEY(opaque) (P_RIGHTMOST(opaque) ? P_HIKEY : P_FIRSTKEY)
/*
* XLOG records for btree operations
*
* XLOG allows to store some information in high 4 bits of log
* record xl_info field
*/
#define XLOG_BTREE_INSERT_LEAF 0x00 /* add btitem without split */
#define XLOG_BTREE_INSERT_UPPER 0x10 /* same, on a non-leaf page */
#define XLOG_BTREE_INSERT_META 0x20 /* same, plus update metapage */
#define XLOG_BTREE_SPLIT_L 0x30 /* add btitem with split */
#define XLOG_BTREE_SPLIT_R 0x40 /* as above, new item on right */
#define XLOG_BTREE_SPLIT_L_ROOT 0x50 /* add btitem with split of root */
#define XLOG_BTREE_SPLIT_R_ROOT 0x60 /* as above, new item on right */
#define XLOG_BTREE_DELETE 0x70 /* delete leaf btitem */
#define XLOG_BTREE_DELETE_PAGE 0x80 /* delete an entire page */
#define XLOG_BTREE_DELETE_PAGE_META 0x90 /* same, plus update metapage */
#define XLOG_BTREE_NEWROOT 0xA0 /* new root page */
#define XLOG_BTREE_NEWMETA 0xB0 /* update metadata page */
#define XLOG_BTREE_NEWPAGE 0xC0 /* new index page during build */
/*
* All that we need to find changed index tuple
*/
typedef struct xl_btreetid
{
RelFileNode node;
ItemPointerData tid; /* changed tuple id */
} xl_btreetid;
/*
* All that we need to regenerate the meta-data page
*/
typedef struct xl_btree_metadata
{
BlockNumber root;
uint32 level;
BlockNumber fastroot;
uint32 fastlevel;
} xl_btree_metadata;
/*
* This is what we need to know about simple (without split) insert.
*
* This data record is used for INSERT_LEAF, INSERT_UPPER, INSERT_META.
* Note that INSERT_META implies it's not a leaf page.
*/
typedef struct xl_btree_insert
{
xl_btreetid target; /* inserted tuple id */
/* xl_btree_metadata FOLLOWS IF XLOG_BTREE_INSERT_META */
/* BTITEM FOLLOWS AT END OF STRUCT */
} xl_btree_insert;
#define SizeOfBtreeInsert (offsetof(xl_btreetid, tid) + SizeOfIptrData)
/*
* On insert with split we save items of both left and right siblings
* and restore content of both pages from log record. This way takes less
* xlog space than the normal approach, because if we did it standardly,
* XLogInsert would almost always think the right page is new and store its
* whole page image.
*
* Note: the four XLOG_BTREE_SPLIT xl_info codes all use this data record.
* The _L and _R variants indicate whether the inserted btitem went into the
* left or right split page (and thus, whether otherblk is the right or left
* page of the split pair). The _ROOT variants indicate that we are splitting
* the root page, and thus that a newroot record rather than an insert or
* split record should follow. Note that a split record never carries a
* metapage update --- we'll do that in the parent-level update.
*/
typedef struct xl_btree_split
{
xl_btreetid target; /* inserted tuple id */
BlockNumber otherblk; /* second block participated in split: */
/* first one is stored in target' tid */
BlockNumber leftblk; /* prev/left block */
BlockNumber rightblk; /* next/right block */
uint32 level; /* tree level of page being split */
uint16 leftlen; /* len of left page items below */
/* LEFT AND RIGHT PAGES TUPLES FOLLOW AT THE END */
} xl_btree_split;
#define SizeOfBtreeSplit (offsetof(xl_btree_split, leftlen) + sizeof(uint16))
/*
* This is what we need to know about delete of an individual leaf btitem
*/
typedef struct xl_btree_delete
{
xl_btreetid target; /* deleted tuple id */
} xl_btree_delete;
#define SizeOfBtreeDelete (offsetof(xl_btreetid, tid) + SizeOfIptrData)
/*
* This is what we need to know about deletion of a btree page. The target
* identifies the tuple removed from the parent page (note that we remove
* this tuple's downlink and the *following* tuple's key). Note we do not
* store any content for the deleted page --- it is just rewritten as empty
* during recovery.
*/
typedef struct xl_btree_delete_page
{
xl_btreetid target; /* deleted tuple id in parent page */
BlockNumber deadblk; /* child block being deleted */
BlockNumber leftblk; /* child block's left sibling, if any */
BlockNumber rightblk; /* child block's right sibling */
/* xl_btree_metadata FOLLOWS IF XLOG_BTREE_DELETE_PAGE_META */
} xl_btree_delete_page;
#define SizeOfBtreeDeletePage (offsetof(xl_btree_delete_page, rightblk) + sizeof(BlockNumber))
/*
* New root log record. There are zero btitems if this is to establish an
* empty root, or two if it is the result of splitting an old root.
*
* Note that although this implies rewriting the metadata page, we don't need
* an xl_btree_metadata record --- the rootblk and level are sufficient.
*/
typedef struct xl_btree_newroot
{
RelFileNode node;
BlockNumber rootblk; /* location of new root */
uint32 level; /* its tree level */
/* 0 or 2 BTITEMS FOLLOW AT END OF STRUCT */
} xl_btree_newroot;
#define SizeOfBtreeNewroot (offsetof(xl_btree_newroot, level) + sizeof(uint32))
/*
* New metapage log record. This is not issued during routine operations;
* it's only used when initializing an empty index and at completion of
* index build.
*/
typedef struct xl_btree_newmeta
{
RelFileNode node;
xl_btree_metadata meta;
} xl_btree_newmeta;
#define SizeOfBtreeNewmeta (sizeof(xl_btree_newmeta))
/*
* New index page log record. This is only used while building a new index.
*/
typedef struct xl_btree_newpage
{
RelFileNode node;
BlockNumber blkno; /* location of new page */
/* entire page contents follow at end of record */
} xl_btree_newpage;
#define SizeOfBtreeNewpage (offsetof(xl_btree_newpage, blkno) + sizeof(BlockNumber))
/*
* Operator strategy numbers -- ordering of these is <, <=, =, >=, >
*/
#define BTLessStrategyNumber 1
#define BTLessEqualStrategyNumber 2
#define BTEqualStrategyNumber 3
#define BTGreaterEqualStrategyNumber 4
#define BTGreaterStrategyNumber 5
#define BTMaxStrategyNumber 5
/*
* When a new operator class is declared, we require that the user
* supply us with an amproc procedure for determining whether, for
* two keys a and b, a < b, a = b, or a > b. This routine must
* return < 0, 0, > 0, respectively, in these three cases. Since we
* only have one such proc in amproc, it's number 1.
*/
#define BTORDER_PROC 1
/*
* We need to be able to tell the difference between read and write
* requests for pages, in order to do locking correctly.
*/
#define BT_READ BUFFER_LOCK_SHARE
#define BT_WRITE BUFFER_LOCK_EXCLUSIVE
/*
* BTStackData -- As we descend a tree, we push the (location, downlink)
* pairs from internal pages onto a private stack. If we split a
* leaf, we use this stack to walk back up the tree and insert data
* into parent pages (and possibly to split them, too). Lehman and
* Yao's update algorithm guarantees that under no circumstances can
* our private stack give us an irredeemably bad picture up the tree.
* Again, see the paper for details.
*/
typedef struct BTStackData
{
BlockNumber bts_blkno;
OffsetNumber bts_offset;
BTItemData bts_btitem;
struct BTStackData *bts_parent;
} BTStackData;
typedef BTStackData *BTStack;
/*
* BTScanOpaqueData is used to remember which buffers we're currently
* examining in the scan. We keep these buffers pinned (but not locked,
@ -116,212 +409,6 @@ typedef struct BTScanOpaqueData
typedef BTScanOpaqueData *BTScanOpaque;
/*
* BTItems are what we store in the btree. Each item is an index tuple,
* including key and pointer values. (In some cases either the key or the
* pointer may go unused, see backend/access/nbtree/README for details.)
*
* Old comments:
* In addition, we must guarantee that all tuples in the index are unique,
* in order to satisfy some assumptions in Lehman and Yao. The way that we
* do this is by generating a new OID for every insertion that we do in the
* tree. This adds eight bytes to the size of btree index tuples. Note
* that we do not use the OID as part of a composite key; the OID only
* serves as a unique identifier for a given index tuple (logical position
* within a page).
*
* New comments:
* actually, we must guarantee that all tuples in A LEVEL
* are unique, not in ALL INDEX. So, we can use bti_itup->t_tid
* as unique identifier for a given index tuple (logical position
* within a level). - vadim 04/09/97
*/
typedef struct BTItemData
{
IndexTupleData bti_itup;
} BTItemData;
typedef BTItemData *BTItem;
/*
* For XLOG: size without alignement. Sizeof works as long as
* IndexTupleData has exactly 8 bytes.
*/
#define SizeOfBTItem sizeof(BTItemData)
/* Test whether items are the "same" per the above notes */
#define BTItemSame(i1, i2) ( (i1)->bti_itup.t_tid.ip_blkid.bi_hi == \
(i2)->bti_itup.t_tid.ip_blkid.bi_hi && \
(i1)->bti_itup.t_tid.ip_blkid.bi_lo == \
(i2)->bti_itup.t_tid.ip_blkid.bi_lo && \
(i1)->bti_itup.t_tid.ip_posid == \
(i2)->bti_itup.t_tid.ip_posid )
/*
* BTStackData -- As we descend a tree, we push the (key, pointer)
* pairs from internal nodes onto a private stack. If we split a
* leaf, we use this stack to walk back up the tree and insert data
* into parent nodes (and possibly to split them, too). Lehman and
* Yao's update algorithm guarantees that under no circumstances can
* our private stack give us an irredeemably bad picture up the tree.
* Again, see the paper for details.
*/
typedef struct BTStackData
{
BlockNumber bts_blkno;
OffsetNumber bts_offset;
BTItemData bts_btitem;
struct BTStackData *bts_parent;
} BTStackData;
typedef BTStackData *BTStack;
/*
* We need to be able to tell the difference between read and write
* requests for pages, in order to do locking correctly.
*/
#define BT_READ BUFFER_LOCK_SHARE
#define BT_WRITE BUFFER_LOCK_EXCLUSIVE
/*
* In general, the btree code tries to localize its knowledge about
* page layout to a couple of routines. However, we need a special
* value to indicate "no page number" in those places where we expect
* page numbers. We can use zero for this because we never need to
* make a pointer to the metadata page.
*/
#define P_NONE 0
/*
* Macros to test whether a page is leftmost or rightmost on its tree level,
* as well as other state info kept in the opaque data.
*/
#define P_LEFTMOST(opaque) ((opaque)->btpo_prev == P_NONE)
#define P_RIGHTMOST(opaque) ((opaque)->btpo_next == P_NONE)
#define P_ISLEAF(opaque) ((opaque)->btpo_flags & BTP_LEAF)
#define P_ISROOT(opaque) ((opaque)->btpo_flags & BTP_ROOT)
/*
* Lehman and Yao's algorithm requires a ``high key'' on every non-rightmost
* page. The high key is not a data key, but gives info about what range of
* keys is supposed to be on this page. The high key on a page is required
* to be greater than or equal to any data key that appears on the page.
* If we find ourselves trying to insert a key > high key, we know we need
* to move right (this should only happen if the page was split since we
* examined the parent page).
*
* Our insertion algorithm guarantees that we can use the initial least key
* on our right sibling as the high key. Once a page is created, its high
* key changes only if the page is split.
*
* On a non-rightmost page, the high key lives in item 1 and data items
* start in item 2. Rightmost pages have no high key, so we store data
* items beginning in item 1.
*/
#define P_HIKEY ((OffsetNumber) 1)
#define P_FIRSTKEY ((OffsetNumber) 2)
#define P_FIRSTDATAKEY(opaque) (P_RIGHTMOST(opaque) ? P_HIKEY : P_FIRSTKEY)
/*
* XLOG allows to store some information in high 4 bits of log
* record xl_info field
*/
#define XLOG_BTREE_DELETE 0x00 /* delete btitem */
#define XLOG_BTREE_INSERT 0x10 /* add btitem without split */
#define XLOG_BTREE_SPLIT 0x20 /* add btitem with split */
#define XLOG_BTREE_SPLEFT 0x30 /* as above + flag that new btitem */
/* goes to the left sibling */
#define XLOG_BTREE_NEWROOT 0x40 /* new root page */
#define XLOG_BTREE_LEAF 0x80 /* leaf/internal page was changed */
/*
* All what we need to find changed index tuple
*/
typedef struct xl_btreetid
{
RelFileNode node;
ItemPointerData tid; /* changed tuple id */
} xl_btreetid;
/*
* This is what we need to know about delete
*/
typedef struct xl_btree_delete
{
xl_btreetid target; /* deleted tuple id */
} xl_btree_delete;
#define SizeOfBtreeDelete (offsetof(xl_btreetid, tid) + SizeOfIptrData)
/*
* This is what we need to know about pure (without split) insert
*/
typedef struct xl_btree_insert
{
xl_btreetid target; /* inserted tuple id */
/* BTITEM FOLLOWS AT END OF STRUCT */
} xl_btree_insert;
#define SizeOfBtreeInsert (offsetof(xl_btreetid, tid) + SizeOfIptrData)
/*
* On insert with split we save items of both left and right siblings
* and restore content of both pages from log record
*/
typedef struct xl_btree_split
{
xl_btreetid target; /* inserted tuple id */
BlockIdData otherblk; /* second block participated in split: */
/* first one is stored in target' tid */
BlockIdData parentblk; /* parent block */
BlockIdData leftblk; /* prev left block */
BlockIdData rightblk; /* next right block */
uint16 leftlen; /* len of left page items below */
/* LEFT AND RIGHT PAGES ITEMS FOLLOW AT THE END */
} xl_btree_split;
#define SizeOfBtreeSplit (offsetof(xl_btree_split, leftlen) + sizeof(uint16))
/*
* New root log record.
*/
typedef struct xl_btree_newroot
{
RelFileNode node;
int32 level;
BlockIdData rootblk;
/* 0 or 2 BTITEMS FOLLOW AT END OF STRUCT */
} xl_btree_newroot;
#define SizeOfBtreeNewroot (offsetof(xl_btree_newroot, rootblk) + sizeof(BlockIdData))
/*
* Operator strategy numbers -- ordering of these is <, <=, =, >=, >
*/
#define BTLessStrategyNumber 1
#define BTLessEqualStrategyNumber 2
#define BTEqualStrategyNumber 3
#define BTGreaterEqualStrategyNumber 4
#define BTGreaterStrategyNumber 5
#define BTMaxStrategyNumber 5
/*
* When a new operator class is declared, we require that the user
* supply us with an amproc procedure for determining whether, for
* two keys a and b, a < b, a = b, or a > b. This routine must
* return < 0, 0, > 0, respectively, in these three cases. Since we
* only have one such proc in amproc, it's number 1.
*/
#define BTORDER_PROC 1
/*
* prototypes for functions in nbtree.c (external entry points for btree)
*/
@ -340,27 +427,26 @@ extern Datum btmarkpos(PG_FUNCTION_ARGS);
extern Datum btrestrpos(PG_FUNCTION_ARGS);
extern Datum btbulkdelete(PG_FUNCTION_ARGS);
extern void btree_redo(XLogRecPtr lsn, XLogRecord *record);
extern void btree_undo(XLogRecPtr lsn, XLogRecord *record);
extern void btree_desc(char *buf, uint8 xl_info, char *rec);
/*
* prototypes for functions in nbtinsert.c
*/
extern InsertIndexResult _bt_doinsert(Relation rel, BTItem btitem,
bool index_is_unique, Relation heapRel);
extern void _bt_insert_parent(Relation rel, Buffer buf, Buffer rbuf,
BTStack stack, bool is_root, bool is_only);
/*
* prototypes for functions in nbtpage.c
*/
extern void _bt_metapinit(Relation rel);
extern Buffer _bt_getroot(Relation rel, int access);
extern Buffer _bt_gettrueroot(Relation rel);
extern Buffer _bt_getbuf(Relation rel, BlockNumber blkno, int access);
extern void _bt_relbuf(Relation rel, Buffer buf);
extern void _bt_wrtbuf(Relation rel, Buffer buf);
extern void _bt_wrtnorelbuf(Relation rel, Buffer buf);
extern void _bt_pageinit(Page page, Size size);
extern void _bt_metaproot(Relation rel, BlockNumber rootbknum, int level);
extern void _bt_metaproot(Relation rel, BlockNumber rootbknum, uint32 level);
extern void _bt_itemdel(Relation rel, Buffer buf, ItemPointer tid);
/*
@ -377,6 +463,7 @@ extern int32 _bt_compare(Relation rel, int keysz, ScanKey scankey,
extern bool _bt_next(IndexScanDesc scan, ScanDirection dir);
extern bool _bt_first(IndexScanDesc scan, ScanDirection dir);
extern bool _bt_step(IndexScanDesc scan, Buffer *bufP, ScanDirection dir);
extern Buffer _bt_get_endpoint(Relation rel, uint32 level, bool rightmost);
/*
* prototypes for functions in nbtstrat.c
@ -407,4 +494,13 @@ extern void _bt_spooldestroy(BTSpool *btspool);
extern void _bt_spool(BTItem btitem, BTSpool *btspool);
extern void _bt_leafbuild(BTSpool *btspool, BTSpool *spool2);
/*
* prototypes for functions in nbtxlog.c
*/
extern void btree_redo(XLogRecPtr lsn, XLogRecord *record);
extern void btree_undo(XLogRecPtr lsn, XLogRecord *record);
extern void btree_desc(char *buf, uint8 xl_info, char *rec);
extern void btree_xlog_startup(void);
extern void btree_xlog_cleanup(void);
#endif /* NBTREE_H */

6
src/include/access/xlog.h
View File

@ -6,7 +6,7 @@
* Portions Copyright (c) 1996-2002, PostgreSQL Global Development Group
* Portions Copyright (c) 1994, Regents of the University of California
*
* $Id: xlog.h,v 1.40 2002/11/15 02:44:57 momjian Exp $
* $Id: xlog.h,v 1.41 2003/02/21 00:06:22 tgl Exp $
*/
#ifndef XLOG_H
#define XLOG_H
@ -145,10 +145,12 @@ typedef XLogPageHeaderData *XLogPageHeader;
*/
typedef struct RmgrData
{
char *rm_name;
const char *rm_name;
void (*rm_redo) (XLogRecPtr lsn, XLogRecord *rptr);
void (*rm_undo) (XLogRecPtr lsn, XLogRecord *rptr);
void (*rm_desc) (char *buf, uint8 xl_info, char *rec);
void (*rm_startup) (void);
void (*rm_cleanup) (void);
} RmgrData;
extern RmgrData RmgrTable[];

4
src/include/catalog/catversion.h
View File

@ -37,7 +37,7 @@
* Portions Copyright (c) 1996-2002, PostgreSQL Global Development Group
* Portions Copyright (c) 1994, Regents of the University of California
*
* $Id: catversion.h,v 1.177 2003/02/16 02:30:39 tgl Exp $
* $Id: catversion.h,v 1.178 2003/02/21 00:06:22 tgl Exp $
*
*-------------------------------------------------------------------------
*/
@ -53,6 +53,6 @@
*/
/* yyyymmddN */
#define CATALOG_VERSION_NO 200302151
#define CATALOG_VERSION_NO 200302171
#endif