This moves sample scan support to below tableam. It's not optional as
there is, in contrast to e.g. bitmap heap scans, no alternative way to
perform tablesample queries. If an AM can't deal with the block based
API, it will have to throw an ERROR.
The tableam callbacks for this are block based, but given the current
TsmRoutine interface, that seems to be required.
The new interface doesn't require TsmRoutines to perform visibility
checks anymore - that requires the TsmRoutine to know details about
the AM, which we want to avoid. To continue to allow taking the
returned number of tuples account SampleScanState now has a donetuples
field (which previously e.g. existed in SystemRowsSamplerData), which
is only incremented after the visibility check succeeds.
Author: Andres Freund
Discussion: https://postgr.es/m/20180703070645.wchpu5muyto5n647@alap3.anarazel.de
Commit 29b64d1d mishandled skipping over truncated high key attributes
during row comparisons. The row comparison key matching loop would loop
forever when a truncated attribute was encountered for a row compare
subkey. Fix by following the example of other code in the loop: advance
the current subkey, or break out of the loop when the last subkey is
reached.
Add test coverage for the relevant _bt_check_rowcompare() code path.
The new test case is somewhat tied to nbtree implementation details,
which isn't ideal, but seems unavoidable.
We can set this up once and for all in subquery_planner's initial survey
of the flattened rangetable, rather than incrementally adjusting it in
build_simple_rel. The previous approach made it rather hard to reason
about exactly when the value would be available, and we were definitely
using it in some places before the final value was computed.
Noted while fooling around with Amit Langote's patch to delay creation
of inheritance child rels. That didn't break this code, but it made it
even more fragile, IMO.
An anti-wraparound vacuum has to be by definition aggressive as it needs
to work on all the pages of a relation. However it can happen that due
to some concurrent activity an anti-wraparound vacuum is marked as
non-aggressive, which makes it redundant with a previous run, and
it is actually useless as an anti-wraparound vacuum should process all
the pages of a relation. This commit makes such vacuums to be skipped.
An anti-wraparound vacuum not aggressive can be found easily by mixing
low values of autovacuum_freeze_max_age (to control anti-wraparound) and
autovacuum_freeze_table_age (to control the aggressiveness).
28a8fa9 has added some extra logging printing all the possible
combinations of anti-wraparound and aggressive vacuums, which now gets
simplified as an anti-wraparound vacuum also non-aggressive gets
skipped.
Per discussion mainly between Andrew Dunstan, Robert Haas, Álvaro
Herrera, Kyotaro Horiguchi, Masahiko Sawada, and myself.
Author: Kyotaro Horiguchi, Michael Paquier
Reviewed-by: Andrew Dunstan, Álvaro Herrera
Discussion: https://postgr.es/m/20180914153554.562muwr3uwujno75@alvherre.pgsql
This just moves the table/matview[/toast] determination of relation
size to a callback, and uses a copy of the existing logic to implement
that callback for heap.
It probably would make sense to also move the index specific logic
into a callback, so the metapage handling (and probably more) can be
index specific. But that's a separate task.
Author: Andres Freund
Discussion: https://postgr.es/m/20180703070645.wchpu5muyto5n647@alap3.anarazel.de
This is a relatively straightforward move of the current
implementation to sit below tableam. As the current analyze sampling
implementation is pretty inherently block based, the tableam analyze
interface is as well. It might make sense to generalize that at some
point, but that seems like a larger project that shouldn't be
undertaken at the same time as the introduction of tableam.
Author: Andres Freund
Discussion: https://postgr.es/m/20180703070645.wchpu5muyto5n647@alap3.anarazel.de
Previously, the planner created RangeTblEntry and RelOptInfo structs
for every partition of a partitioned table, even though many of them
might later be deemed uninteresting thanks to partition pruning logic.
This incurred significant overhead when there are many partitions.
Arrange to postpone creation of these data structures until after
we've processed the query enough to identify restriction quals for
the partitioned table, and then apply partition pruning before not
after creation of each partition's data structures. In this way
we need not open the partition relations at all for partitions that
the planner has no real interest in.
For queries that can be proven at plan time to access only a small
number of partitions, this patch improves the practical maximum
number of partitions from under 100 to perhaps a few thousand.
Amit Langote, reviewed at various times by Dilip Kumar, Jesper Pedersen,
Yoshikazu Imai, and David Rowley
Discussion: https://postgr.es/m/9d7c5112-cb99-6a47-d3be-cf1ee6862a1d@lab.ntt.co.jp
Commit d85e0f366a tried to fix memory alignment issues in serialization
and deserialization of pg_mcv_list values, but it was a few bricks shy.
The arrays of uint16 indexes in serialized items was not aligned, and
the both the values and isnull flags were using the same pointer.
Per investigation by Tom Lane on gaur.
While trying to plan a partitionwise join, we may be faced with cases
where one or both input partitions for a particular segment of the join
have been pruned away. In HEAD and v11, this is problematic because
earlier processing didn't bother to make a pruned RelOptInfo fully
valid. With an upcoming patch to make partition pruning more efficient,
this'll be even more problematic because said RelOptInfo won't exist at
all.
The existing code attempts to deal with this by retroactively making the
RelOptInfo fully valid, but that causes crashes under GEQO because join
planning is done in a short-lived memory context. In v11 we could
probably have fixed this by switching to the planner's main context
while fixing up the RelOptInfo, but that idea doesn't scale well to the
upcoming patch. It would be better not to mess with the base-relation
data structures during join planning, anyway --- that's just a recipe
for order-of-operations bugs.
In many cases, though, we don't actually need the child RelOptInfo,
because if the input is certainly empty then the join segment's result
is certainly empty, so we can skip making a join plan altogether. (The
existing code ultimately arrives at the same conclusion, but only after
doing a lot more work.) This approach works except when the pruned-away
partition is on the nullable side of a LEFT, ANTI, or FULL join, and the
other side isn't pruned. But in those cases the existing code leaves a
lot to be desired anyway --- the correct output is just the result of
the unpruned side of the join, but we were emitting a useless outer join
against a dummy Result. Pending somebody writing code to handle that
more nicely, let's just abandon the partitionwise-join optimization in
such cases.
When the modified code skips making a join plan, it doesn't make a
join RelOptInfo either; this requires some upper-level code to
cope with nulls in part_rels[] arrays. We would have had to have
that anyway after the upcoming patch.
Back-patch to v11 since the crash is demonstrable there.
Discussion: https://postgr.es/m/8305.1553884377@sss.pgh.pa.us
This is an SQL-standard feature that allows creating columns that are
computed from expressions rather than assigned, similar to a view or
materialized view but on a column basis.
This implements one kind of generated column: stored (computed on
write). Another kind, virtual (computed on read), is planned for the
future, and some room is left for it.
Reviewed-by: Michael Paquier <michael@paquier.xyz>
Reviewed-by: Pavel Stehule <pavel.stehule@gmail.com>
Discussion: https://www.postgresql.org/message-id/flat/b151f851-4019-bdb1-699e-ebab07d2f40a@2ndquadrant.com
Blind attempt at fixing ia64, hppa an sparc builds.
The serialized representation of MCV lists did not enforce proper memory
alignment for internal fields, resulting in deserialization issues on
platforms that are more sensitive to this (ia64, sparc and hppa).
This forces a catalog version bump, because the layout of serialized
pg_mcv_list changes.
Broken since 7300a699.
Especially, warn about the hazards of mishandling the backup_label
file. Adjust a couple of server messages to be more clear about
the hazards associated with removing backup_label files, too.
David Steele and Robert Haas, reviewed by Laurenz Albe, Martín
Marqués, Peter Eisentraut, and Magnus Hagander.
Discussion: http://postgr.es/m/7d85c387-000e-16f0-e00b-50bf83c22127@pgmasters.net
The previous code was adding pointers to transient variables to a
list, but by the time the list was read, the variable might be gone,
depending on the compiler. Fix it by making copies in the proper
memory context.
This adds the CONCURRENTLY option to the REINDEX command. A REINDEX
CONCURRENTLY on a specific index creates a new index (like CREATE
INDEX CONCURRENTLY), then renames the old index away and the new index
in place and adjusts the dependencies, and then drops the old
index (like DROP INDEX CONCURRENTLY). The REINDEX command also has
the capability to run its other variants (TABLE, DATABASE) with the
CONCURRENTLY option (but not SYSTEM).
The reindexdb command gets the --concurrently option.
Author: Michael Paquier, Andreas Karlsson, Peter Eisentraut
Reviewed-by: Andres Freund, Fujii Masao, Jim Nasby, Sergei Kornilov
Discussion: https://www.postgresql.org/message-id/flat/60052986-956b-4478-45ed-8bd119e9b9cf%402ndquadrant.com#74948a1044c56c5e817a5050f554ddee
This moves the responsibility for:
- creating the storage necessary for a relation, including creating a
new relfilenode for a relation with existing storage
- non-transactional truncation of a relation
- VACUUM FULL / CLUSTER's rewrite of a table
below tableam.
This is fairly straight forward, with a bit of complexity smattered in
to move the computation of xid / multixid horizons below the AM, as
they don't make sense for every table AM.
Author: Andres Freund
Discussion: https://postgr.es/m/20180703070645.wchpu5muyto5n647@alap3.anarazel.de
There were multiple issues in deserialization of pg_mcv_list values.
Firstly, the data is loaded from syscache, but the deserialization was
performed after ReleaseSysCache(), at which point the data might have
already disappeared. Fixed by moving the calls in statext_mcv_load,
and using the same NULL-handling code as existing stats.
Secondly, the deserialized representation used pointers into the
serialized representation. But that is also unsafe, because the data
may disappear at any time. Fixed by reworking and simplifying the
deserialization code to always copy all the data.
And thirdly, when deserializing values for types passed by value, the
code simply did memcpy(d,s,typlen) which however does not work on
bigendian machines. Fixed by using fetch_att/store_att_byval.
Provide GetTopFullTransactionId() and GetCurrentFullTransactionId().
The intended users of these interfaces are access methods that use
xids for visibility checks but don't want to have to go back and
"freeze" existing references some time later before the 32 bit xid
counter wraps around.
Use a new struct to serialize the transaction state for parallel
query, because FullTransactionId doesn't fit into the previous
serialization scheme very well.
Author: Thomas Munro
Reviewed-by: Heikki Linnakangas
Discussion: https://postgr.es/m/CAA4eK1%2BMv%2Bmb0HFfWM9Srtc6MVe160WFurXV68iAFMcagRZ0dQ%40mail.gmail.com
Instead of inferring epoch progress from xids and checkpoints,
introduce a 64 bit FullTransactionId type and use it to track xid
generation. This fixes an unlikely bug where the epoch is reported
incorrectly if the range of active xids wraps around more than once
between checkpoints.
The only user-visible effect of this commit is to correct the epoch
used by txid_current() and txid_status(), also visible with
pg_controldata, in those rare circumstances. It also creates some
basic infrastructure so that later patches can use 64 bit
transaction IDs in more places.
The new type is a struct that we pass by value, as a form of strong
typedef. This prevents the sort of accidental confusion between
TransactionId and FullTransactionId that would be possible if we
were to use a plain old uint64.
Author: Thomas Munro
Reported-by: Amit Kapila
Reviewed-by: Andres Freund, Tom Lane, Heikki Linnakangas
Discussion: https://postgr.es/m/CAA4eK1%2BMv%2Bmb0HFfWM9Srtc6MVe160WFurXV68iAFMcagRZ0dQ%40mail.gmail.com
To support building indexes over tables of different AMs, the scans to
do so need to be routed through the table AM. While moving a fair
amount of code, nearly all the changes are just moving code to below a
callback.
Currently the range based interface wouldn't make much sense for non
block based table AMs. But that seems aceptable for now.
Author: Andres Freund
Discussion: https://postgr.es/m/20180703070645.wchpu5muyto5n647@alap3.anarazel.de
The MCV build should always call get_mincount_for_mcv_list(), as the
there is no other logic to decide whether the MCV list represents all
the data. So just remove the (ngroups > nitems) condition.
Also, when building MCV lists, the number of items was limited by the
statistics target (i.e. up to 10000). But when deserializing the MCV
list, a different value (8192) was used to check the input, causing
an error. Simply ensure that the same value is used in both places.
This should have been included in 7300a69950, but I forgot to include it
in that commit.
Introduce a third extended statistic type, supported by the CREATE
STATISTICS command - MCV lists, a generalization of the statistic
already built and used for individual columns.
Compared to the already supported types (n-distinct coefficients and
functional dependencies), MCV lists are more complex, include column
values and allow estimation of much wider range of common clauses
(equality and inequality conditions, IS NULL, IS NOT NULL etc.).
Similarly to the other types, a new pseudo-type (pg_mcv_list) is used.
Author: Tomas Vondra
Reviewed-by: Dean Rasheed, David Rowley, Mark Dilger, Alvaro Herrera
Discussion: https://postgr.es/m/dfdac334-9cf2-2597-fb27-f0fb3753f435@2ndquadrant.com
In the dim past, the planner kept the fully-processed version of the query
targetlist (the result of preprocess_targetlist) in grouping_planner's
local variable "tlist", and only grudgingly passed it to individual other
routines as needed. Later we discovered a need to still have it available
after grouping_planner finishes, and invented the root->processed_tlist
field for that purpose, but it wasn't used internally to grouping_planner;
the tlist was still being passed around separately in the same places as
before.
Now comes a proposed patch to allow appendrel expansion to add entries
to the processed tlist, well after preprocess_targetlist has finished
its work. To avoid having to pass around the tlist explicitly, it's
proposed to allow appendrel expansion to modify root->processed_tlist.
That makes aliasing the tlist with assorted parameters and local
variables really scary. It would accidentally work as long as the
tlist is initially nonempty, because then the List header won't move
around, but it's not exactly hard to think of ways for that to break.
Aliased values are poor programming practice anyway.
Hence, get rid of local variables and parameters that can be identified
with root->processed_tlist, in favor of just using that field directly.
And adjust comments to match. (Some of the new comments speak as though
it's already possible for appendrel expansion to modify the tlist; that's
not true yet, but will happen in a later patch.)
Discussion: https://postgr.es/m/9d7c5112-cb99-6a47-d3be-cf1ee6862a1d@lab.ntt.co.jp
Column references are not allowed in default expressions and partition
bound expressions, and are restricted as such once the transformation of
their expressions is done. However, trying to use more complex column
references can lead to confusing error messages. For example, trying to
use a two-field column reference name for default expressions and
partition bounds leads to "missing FROM-clause entry for table", which
makes no sense in their respective context.
In order to make the errors generated more useful, this commit adds more
verbose messages when transforming column references depending on the
context. This has a little consequence though: for example an
expression using an aggregate with a column reference as argument would
cause an error to be generated for the column reference, while the
aggregate was the problem reported before this commit because column
references get transformed first.
The confusion exists for default expressions for a long time, and the
problem is new as of v12 for partition bounds. Still per the lack of
complaints on the matter no backpatch is done.
The patch has been written by Amit Langote and me, and Tom Lane has
provided the improvement of the documentation for default expressions on
the CREATE TABLE page.
Author: Amit Langote, Michael Paquier
Reviewed-by: Tom Lane
Discussion: https://postgr.es/m/20190326020853.GM2558@paquier.xyz
The transaction ID returned by GetNextXidAndEpoch() is in the future,
so we can't attempt to access its status or we might try to read a
CLOG page that doesn't exist. The > vs >= confusion probably stemmed
from the choice of a variable name containing the word "last" instead
of "next", so fix that too.
Back-patch to 10 where the function arrived.
Author: Thomas Munro
Discussion: https://postgr.es/m/CA%2BhUKG%2Buua_BV5cyfsioKVN2d61Lukg28ECsWTXKvh%3DBtN2DPA%40mail.gmail.com
Relations dropped in a single transaction are tracked in a list of
unowned relations. With large number of dropped relations this resulted
in poor performance at the end of a transaction, when the relations are
removed from the singly linked list one by one.
Commit b4166911 attempted to address this issue (particularly when it
happens during recovery) by removing the relations in a reverse order,
resulting in O(1) lookups in the list of unowned relations. This did
not work reliably, though, and it was possible to trigger the O(N^2)
behavior in various ways.
Instead of trying to remove the relations in a specific order with
respect to the linked list, which seems rather fragile, switch to a
regular doubly linked. That allows us to remove relations cheaply no
matter where in the list they are.
As b4166911 was a bugfix, backpatched to all supported versions, do the
same thing here.
Reviewed-by: Alvaro Herrera
Discussion: https://www.postgresql.org/message-id/flat/80c27103-99e4-1d0c-642c-d9f3b94aaa0a%402ndquadrant.com
Backpatch-through: 9.4
Previously the xid horizon was only computed during WAL replay. That
had two major problems:
1) It relied on knowing what the table pointed to looks like. That was
easy enough before the introducing of tableam (we knew it had to be
heap, although some trickery around logging the heap relfilenodes
was required). But to properly handle table AMs we need
per-database catalog access to look up the AM handler, which
recovery doesn't allow.
2) Not knowing the xid horizon also makes it hard to support logical
decoding on standbys. When on a catalog table, we need to be able
to conflict with slots that have an xid horizon that's too old. But
computing the horizon by visiting the heap only works once
consistency is reached, but we always need to be able to detect
conflicts.
There's also a secondary problem, in that the current method performs
redundant work on every standby. But that's counterbalanced by
potentially computing the value when not necessary (either because
there's no standby, or because there's no connected backends).
Solve 1) and 2) by moving computation of the xid horizon to the
primary and by involving tableam in the computation of the horizon.
To address the potentially increased overhead, increase the efficiency
of the xid horizon computation for heap by sorting the tids, and
eliminating redundant buffer accesses. When prefetching is available,
additionally perform prefetching of buffers. As this is more of a
maintenance task, rather than something routinely done in every read
only query, we add an arbitrary 10 to the effective concurrency -
thereby using IO concurrency, when not globally enabled. That's
possibly not the perfect formula, but seems good enough for now.
Bumps WAL format, as latestRemovedXid is now part of the records, and
the heap's relfilenode isn't anymore.
Author: Andres Freund, Amit Khandekar, Robert Haas
Reviewed-By: Robert Haas
Discussion:
https://postgr.es/m/20181212204154.nsxf3gzqv3gesl32@alap3.anarazel.dehttps://postgr.es/m/20181214014235.dal5ogljs3bmlq44@alap3.anarazel.dehttps://postgr.es/m/20180703070645.wchpu5muyto5n647@alap3.anarazel.de
ALTER INDEX .. ATTACH PARTITION fails if the partitioned table where the
index is defined contains more dropped columns than its partition, with
this message:
ERROR: incorrect attribute map
The cause was that one caller of CompareIndexInfo was passing the number
of attributes of the partition rather than the parent, which confused
the length check. Repair.
This can cause pg_upgrade to fail when used on such a database. Leave
some more objects around after regression tests, so that the case is
detected by pg_upgrade test suite.
Remove some spurious empty lines noticed while looking for other cases
of the same problem.
Discussion: https://postgr.es/m/20190326213924.GA2322@alvherre.pgsql
Up to now, otherrel RelOptInfos were built at the same time as baserel
RelOptInfos, thanks to recursion in build_simple_rel(). However,
nothing in query_planner's preprocessing cares at all about otherrels,
only baserels, so we don't really need to build them until just before
we enter make_one_rel. This has two benefits:
* create_lateral_join_info did a lot of extra work to propagate
lateral-reference information from parents to the correct children.
But if we delay creation of the children till after that, it's
trivial (and much harder to break, too).
* Since we have all the restriction quals correctly assigned to
parent appendrels by this point, it'll be possible to do plan-time
pruning and never make child RelOptInfos at all for partitions that
can be pruned away. That's not done here, but will be later on.
Amit Langote, reviewed at various times by Dilip Kumar, Jesper Pedersen,
Yoshikazu Imai, and David Rowley
Discussion: https://postgr.es/m/9d7c5112-cb99-6a47-d3be-cf1ee6862a1d@lab.ntt.co.jp
Commit caf626b2c missed that the relevant reloptions entry needs
to be moved from the intRelOpts[] array to realRelOpts[].
Somewhat surprisingly, it seems to work anyway, perhaps because
the desired default and limit values are all integers. We ought
to have either a simpler data structure or better cross-checking
here, but that's for another patch.
Nikolay Shaplov
Discussion: https://postgr.es/m/4861742.12LTaSB3sv@x200m
We've been creating duplicate RTEs for partitioned tables just
because we do so for regular inheritance parent tables. But unlike
regular-inheritance parents which are themselves regular tables
and thus need to be scanned, partitioned tables don't need the
extra RTE.
This makes the conditions for building a child RTE the same as those
for building an AppendRelInfo, allowing minor simplification in
expand_single_inheritance_child. Since the planner's actual processing
is driven off the AppendRelInfo list, nothing much changes beyond that,
we just have one fewer useless RTE entry.
Amit Langote, reviewed and hacked a bit by me
Discussion: https://postgr.es/m/9d7c5112-cb99-6a47-d3be-cf1ee6862a1d@lab.ntt.co.jp
These instances were apparently trying to carry the const qualifier
from the arguments through the complex casts, but for that the const
qualifier was misplaced.
After 71bdc99d0d, "tableam: Add helper for indexes to check if a
corresponding table tuples exist." there's no in-core user left. As
there's unlikely to be an external user, and such an external user
could easily be adjusted to use table_index_fetch_tuple_check(),
remove heap_hot_search().
Per complaint from Peter Geoghegan
Author: Andres Freund
Discussion: https://postgr.es/m/CAH2-Wzn0Oq4ftJrTqRAsWy2WGjv0QrJcwoZ+yqWsF_Z5vjUBFw@mail.gmail.com
Since 7c079d7, partition bounds are able to use generalized expression
syntax when processed, treating "minvalue" and "maxvalue" as specific
cases as they get passed down for transformation as a column references.
The checks for infinite bounds in range expressions have been lax
though, causing crashes when trying to use column reference names with
more than one field. Here is an example causing a crash:
CREATE TABLE list_parted (a int) PARTITION BY LIST (a);
CREATE TABLE part_list_crash PARTITION OF list_parted
FOR VALUES IN (somename.somename);
Note that the creation of the second relation should fail as partition
bounds cannot have column references in their expressions, so when
finding an expression which does not match the expected infinite bounds,
then this commit lets the generic transformation machinery check after
it. The error message generated in this case references as well a
missing RTE, which is confusing. This problem will be treated
separately as it impacts as well default expressions for some time, and
for now only the cases where a crash can happen are fixed.
While on it, extend the set of regression tests in place for list
partition bounds and add an extra set for range partition bounds.
Reported-by: Alexander Lakhin
Author: Michael Paquier
Reviewed-by: Amit Langote
Discussion: https://postgr.es/m/15668-0377b1981aa1a393@postgresql.org
This primarily is to allow WHERE CURRENT OF to continue to work as it
currently does. It's not clear to me that these semantics make sense
for every AM, but it works for the in-core heap, and the out of core
zheap. We can refine it further at a later point if necessary.
Author: Andres Freund
Discussion: https://postgr.es/m/20180703070645.wchpu5muyto5n647@alap3.anarazel.de
We were getting just DEFAULT_INEQ_SEL for comparisons such as
"ctid >= constant", but it's possible to do a lot better if we don't
mind some assumptions about the table's tuple density being reasonably
uniform. There are already assumptions much like that elsewhere in
the planner, so that hardly seems like much of an objection.
Extracted from a patch set that also proposes to introduce a special
executor node type for such queries. Not sure if that's going to make
it into v12, but improving the selectivity estimate is useful
independently of that.
Edmund Horner, reviewed by David Rowley
Discussion: https://postgr.es/m/CAMyN-kB-nFTkF=VA_JPwFNo08S0d-Yk0F741S2B7LDmYAi8eyA@mail.gmail.com
If there's only one child relation, the Append or MergeAppend isn't
doing anything useful, and can be elided. It does have a purpose
during planning though, which is to serve as a buffer between parent
and child Var numbering. Therefore we keep it all the way through
to setrefs.c, and get rid of it only after fixing references in the
plan level(s) above it. This works largely the same as setrefs.c's
ancient hack to get rid of no-op SubqueryScan nodes, and can even
share some code with that.
Note the change to make setrefs.c use apply_tlist_labeling rather than
ad-hoc code. This has the effect of propagating the child's resjunk
and ressortgroupref labels, which formerly weren't propagated when
removing a SubqueryScan. Doing that is demonstrably necessary for
the [Merge]Append cases, and seems harmless for SubqueryScan, if only
because trivial_subqueryscan is afraid to collapse cases where the
resjunk marking differs. (I suspect that restriction could now be
removed, though it's unclear that it'd make any new matches possible,
since the outer query can't have references to a child resjunk column.)
David Rowley, reviewed by Alvaro Herrera and Tomas Vondra
Discussion: https://postgr.es/m/CAKJS1f_7u8ATyJ1JGTMHFoKDvZdeF-iEBhs+sM_SXowOr9cArg@mail.gmail.com
Add additional heuristics to the algorithm for locating an optimal split
location. New logic identifies localized monotonically increasing
values in indexes with multiple columns. When this insertion pattern is
detected, page splits split just after the new item that provoked a page
split (or apply leaf fillfactor in the style of a rightmost page split).
This optimization is a variation of the long established leaf fillfactor
optimization used during rightmost page splits.
50/50 page splits are only appropriate with a pattern of truly random
insertions, where the average space utilization ends up at 65% - 70%.
Without this patch, affected cases have leaf pages that are no more than
about 50% full on average. Future insertions can never make use of the
free space left behind. With this patch, affected cases have leaf pages
that are about 90% full on average (assuming a fillfactor of 90).
Localized monotonically increasing insertion patterns are presumed to be
fairly common in real-world applications. There is a fair amount of
anecdotal evidence for this. Both pg_depend system catalog indexes
(pg_depend_depender_index and pg_depend_reference_index) are at least
20% smaller after the regression tests are run when the optimization is
available. Furthermore, many of the indexes created by a fair use
implementation of TPC-C for Postgres are consistently about 40% smaller
when the optimization is available.
Note that even pg_upgrade'd v3 indexes make use of this optimization.
Author: Peter Geoghegan
Reviewed-By: Heikki Linnakangas
Discussion: https://postgr.es/m/CAH2-WzkpKeZJrXvR_p7VSY1b-s85E3gHyTbZQzR0BkJ5LrWF_A@mail.gmail.com
This uses the same progress reporting infrastructure added in commit
c16dc1aca5 and extends it to these
additional cases. We lack the ability to track the internal progress
of sorts and index builds so the information reported is
coarse-grained for some parts of the operation, but it still seems
like a significant improvement over having nothing at all.
Tatsuro Yamada, reviewed by Thomas Munro, Masahiko Sawada, Michael
Paquier, Jeff Janes, Alvaro Herrera, Rafia Sabih, and by me. A fair
amount of polishing also by me.
Discussion: http://postgr.es/m/59A77072.3090401@lab.ntt.co.jp
Non-backtracking flex parsers work faster than backtracking ones. So, this
commit gets rid of backtracking in jsonpath_scan.l. That required explicit
handling of some cases as well as manual backtracking for some cases. More
regression tests for numerics are added.
Discussion: https://mail.google.com/mail/u/0?ik=a20b091faa&view=om&permmsgid=msg-f%3A1628425344167939063
Author: John Naylor, Nikita Gluknov, Alexander Korotkov
This commit include formatting improvements, renamings and comments. Also,
it makes jsonpath_scan.l be more uniform with other our lexers. Firstly,
states names are renamed to more short alternatives. Secondly, <INITIAL>
prefix removed from the rules. Corresponding rules are moved to the tail, so
they would anyway work only in initial state.
Author: Alexander Korotkov
Reviewed-by: John Naylor
Coverity complained that simple8b_encode() might read beyond the end of
the 'diffs' array, in the loop to encode the integers. That was a false
positive, because we never get into the loop in modes 0 or 1, and the
array is large enough for all the other modes. But I admit it's very
subtle, so it's not surprising that Coverity didn't see it, and it's not
very obvious to humans either. Refactor it, so that the second loop
re-computes the differences, instead of carrying them over from the first
loop in the 'diffs' array. This way, the 'diffs' array is not needed
anymore. It makes no measurable difference in performance, and seems more
straightforward this way.
Also, improve the comments in simple8b_encode(): fix the comment about its
return value that was flat-out wrong, and explain the condition when it
returns EMPTY_CODEWORD better.
In the passing, move the 'selector' from the codeword's low bits to the
high bits. It doesn't matter much, but looking at the original paper, and
googling around for other Simple-8b implementations, that's how it's
usually done.
Per Coverity, and Tom Lane's report off-list.
This is essentially the tableam version of heapam_fetch(),
i.e. fetching a tuple identified by a tid, performing visibility
checks.
Note that this different from table_index_fetch_tuple(), which is for
index lookups. It therefore has to handle a tid pointing to an earlier
version of a tuple if the AM uses an optimization like heap's HOT. Add
comments to that end.
This commit removes the stats_relation argument from heap_fetch, as
it's been unused for a long time.
Author: Andres Freund
Reviewed-By: Haribabu Kommi
Discussion: https://postgr.es/m/20180703070645.wchpu5muyto5n647@alap3.anarazel.de
Previously those directly performed a heap_insert(). Use
table_insert() instead. The input slot of those routines is not of
the target relation - we could fix that by copying if necessary, but
that'd not be beneficial for performance. As those codepaths don't
access any AM specific tuple fields (say xmin/xmax), there's no need
to use an AM specific slot.
Author: Andres Freund
Reviewed-By: Haribabu Kommi
Discussion: https://postgr.es/m/20180703070645.wchpu5muyto5n647@alap3.anarazel.de
Commit 8aa9dd74b didn't quite finish the job in this area after all,
because DROP ROLE has a code path distinct from DROP OWNED BY, and
it was still reporting dependent objects in whatever order the index
scan returned them in.
Buildfarm experience shows that index ordering of equal-keyed objects is
significantly less stable than before in the wake of using heap TIDs as
tie-breakers. So if we try to hide the unstable ordering by suppressing
DETAIL reports, we're just going to end up having to do that for every
DROP that reports multiple objects. That's not great from a coverage
or problem-detection standpoint, and it's something we'll inevitably
forget in future patches, leading to more iterations of fixing-an-
unstable-result. So let's just bite the bullet and sort here too.
Discussion: https://postgr.es/m/E1h6eep-0001Mw-Vd@gemulon.postgresql.org
It doesn't make sense to consider the possibility that there will only
be one candidate split point when choosing among split points to find
the split with the lowest penalty. This is a vestige of an earlier
version of the patch that became commit fab25024.
Issue spotted while rereviewing coverage of the nbtree patch series
using gcov.
Since its introduction in 19dc233c, current_logfiles has been assigned
the same permissions as a log file, which can be enforced with
log_file_mode. This setup can lead to incompatibility problems with
group access permissions as current_logfiles is not located in the log
directory, but at the root of the data folder. Hence, if group
permissions are used but log_file_mode is more restrictive, a backup
with a user in the group having read access could fail even if the log
directory is located outside of the data folder.
Per discussion with the folks mentioned below, we have concluded that
current_logfiles should not be treated as a log file as it only stores
metadata related to log files, and that it should use the same
permissions as all other files in the data directory. This solution has
the merit to be simple and fixes all the interaction problems between
group access and log_file_mode.
Author: Haribabu Kommi
Reviewed-by: Stephen Frost, Robert Haas, Tom Lane, Michael Paquier
Discussion: https://postgr.es/m/CAJrrPGcEotF1P7AWoeQyD3Pqr-0xkQg_Herv98DjbaMj+naozw@mail.gmail.com
Backpatch-through: 11, where group access has been added.
Add command variants COMMIT AND CHAIN and ROLLBACK AND CHAIN, which
start new transactions with the same transaction characteristics as the
just finished one, per SQL standard.
Support for transaction chaining in PL/pgSQL is also added. This
functionality is especially useful when running COMMIT in a loop in
PL/pgSQL.
Reviewed-by: Fabien COELHO <coelho@cri.ensmp.fr>
Discussion: https://www.postgresql.org/message-id/flat/28536681-324b-10dc-ade8-ab46f7645a5a@2ndquadrant.com
This adds new, required, table AM callbacks for insert/delete/update
and lock_tuple. To be able to reasonably use those, the EvalPlanQual
mechanism had to be adapted, moving more logic into the AM.
Previously both delete/update/lock call-sites and the EPQ mechanism had
to have awareness of the specific tuple format to be able to fetch the
latest version of a tuple. Obviously that needs to be abstracted
away. To do so, move the logic that find the latest row version into
the AM. lock_tuple has a new flag argument,
TUPLE_LOCK_FLAG_FIND_LAST_VERSION, that forces it to lock the last
version, rather than the current one. It'd have been possible to do
so via a separate callback as well, but finding the last version
usually also necessitates locking the newest version, making it
sensible to combine the two. This replaces the previous use of
EvalPlanQualFetch(). Additionally HeapTupleUpdated, which previously
signaled either a concurrent update or delete, is now split into two,
to avoid callers needing AM specific knowledge to differentiate.
The move of finding the latest row version into tuple_lock means that
encountering a row concurrently moved into another partition will now
raise an error about "tuple to be locked" rather than "tuple to be
updated/deleted" - which is accurate, as that always happens when
locking rows. While possible slightly less helpful for users, it seems
like an acceptable trade-off.
As part of this commit HTSU_Result has been renamed to TM_Result, and
its members been expanded to differentiated between updating and
deleting. HeapUpdateFailureData has been renamed to TM_FailureData.
The interface to speculative insertion is changed so nodeModifyTable.c
does not have to set the speculative token itself anymore. Instead
there's a version of tuple_insert, tuple_insert_speculative, that
performs the speculative insertion (without requiring a flag to signal
that fact), and the speculative insertion is either made permanent
with table_complete_speculative(succeeded = true) or aborted with
succeeded = false).
Note that multi_insert is not yet routed through tableam, nor is
COPY. Changing multi_insert requires changes to copy.c that are large
enough to better be done separately.
Similarly, although simpler, CREATE TABLE AS and CREATE MATERIALIZED
VIEW are also only going to be adjusted in a later commit.
Author: Andres Freund and Haribabu Kommi
Discussion:
https://postgr.es/m/20180703070645.wchpu5muyto5n647@alap3.anarazel.dehttps://postgr.es/m/20190313003903.nwvrxi7rw3ywhdel@alap3.anarazel.dehttps://postgr.es/m/20160812231527.GA690404@alvherre.pgsql
Previously we were using the SQL:2003 definition, which doesn't allow
this, but that creates a serious dump/restore gotcha: there is no
setting of xmloption that will allow all valid XML data. Hence,
switch to the 2006 definition.
Since libxml doesn't accept <!DOCTYPE> directives in the mode we
use for CONTENT parsing, the implementation is to detect <!DOCTYPE>
in the input and switch to DOCUMENT parsing mode. This should not
cost much, because <!DOCTYPE> should be close to the front of the
input if it's there at all. It's possible that this causes the
error messages for malformed input to be slightly different than
they were before, if said input includes <!DOCTYPE>; but that does
not seem like a big problem.
In passing, buy back a few cycles in parsing of large XML documents
by not doing strlen() of the whole input in parse_xml_decl().
Back-patch because dump/restore failures are not nice. This change
shouldn't break any cases that worked before, so it seems safe to
back-patch.
Chapman Flack (revised a bit by me)
Discussion: https://postgr.es/m/CAN-V+g-6JqUQEQZ55Q3toXEN6d5Ez5uvzL4VR+8KtvJKj31taw@mail.gmail.com
Teach nbtree forward index scans to check the high key before moving to
the right sibling page in the hope of finding that it isn't actually
necessary to do so. The new check may indicate that the scan definitely
cannot find matching tuples to the right, ending the scan immediately.
We already opportunistically force a similar "continuescan orientated"
key check of the final non-pivot tuple when it's clear that it cannot be
returned to the scan due to being dead-to-all. The new high key check
is complementary.
The new approach for forward scans is more effective than checking the
final non-pivot tuple, especially with composite indexes and non-unique
indexes. The improvements to the logic for picking a split point added
by commit fab25024 make it likely that relatively dissimilar high keys
will appear on a page. A distinguishing key value that can only appear
on non-pivot tuples on the right sibling page will often be present in
leaf page high keys.
Since forcing the final item to be key checked no longer makes any
difference in the case of forward scans, the existing extra key check is
now only used for backwards scans. Backward scans continue to
opportunistically check the final non-pivot tuple, which is actually the
first non-pivot tuple on the page (not the last).
Note that even pg_upgrade'd v3 indexes make use of this optimization.
Author: Peter Geoghegan, Heikki Linnakangas
Reviewed-By: Heikki Linnakangas
Discussion: https://postgr.es/m/CAH2-WzkOmUduME31QnuTFpimejuQoiZ-HOf0pOWeFZNhTMctvA@mail.gmail.com
Postpone most of the effort of constructing PartitionedRelPruneInfos
until after we have found out whether run-time pruning is needed at all.
This costs very little duplicated effort (basically just an extra
find_base_rel() call per partition) and saves quite a bit when we
can't do run-time pruning.
Also, merge the first loop (for building relid_subpart_map) into
the second loop, since we don't need the map to be valid during
that loop.
Amit Langote
Discussion: https://postgr.es/m/9d7c5112-cb99-6a47-d3be-cf1ee6862a1d@lab.ntt.co.jp
I (tgl) remain dubious that it's a good idea for PartitionDirectory
to hold a pin on a relcache entry throughout planning, rather than
copying the data or using some kind of refcount scheme. However, it's
certainly the responsibility of the PartitionDirectory code to ensure
that what it's handing back is a stable data structure, not that of
its caller. So this is a pretty clear oversight in commit 898e5e329,
and one that can cost a lot of performance when there are many
partitions.
Amit Langote (extracted from a much larger patch set)
Discussion: https://postgr.es/m/CA+TgmoY3bRmGB6-DUnoVy5fJoreiBJ43rwMrQRCdPXuKt4Ykaw@mail.gmail.com
Discussion: https://postgr.es/m/9d7c5112-cb99-6a47-d3be-cf1ee6862a1d@lab.ntt.co.jp
There were more large constants that needed UINT64CONST. And one variable
was declared as "int", when it needed to be uint64. These bugs were only
visible on 32-bit systems; clearly I should've tested on one, given that
this code does a lot of work with 64-bit integers.
Also, in the test "huge distances" test, the code created some values with
random distances between them, but the test logic didn't take into account
the possibility that the random distance was exactly 1. That never actually
happens with the seed we're using, but let's be tidy.
To do this, we scan GiST two times. In the first pass we make note of
empty leaf pages and internal pages. At second pass we scan through
internal pages, looking for downlinks to the empty pages.
Deleting internal pages is still not supported, like in nbtree, the last
child of an internal page is never deleted. That means that if you have a
workload where new keys are always inserted to different area than where
old keys are removed, the index will still grow without bound. But the rate
of growth will be an order of magnitude slower than before.
Author: Andrey Borodin
Discussion: https://www.postgresql.org/message-id/B1E4DF12-6CD3-4706-BDBD-BF3283328F60@yandex-team.ru
The set is implemented as a B-tree, with a compact representation at leaf
items, using Simple-8b algorithm, so that clusters of nearby values use
less memory.
The IntegerSet isn't used for anything yet, aside from the test code, but
we have two patches in the works that would benefit from this: A patch to
allow GiST vacuum to delete empty pages, and a patch to reduce heap
VACUUM's memory usage, by storing the list of dead TIDs more efficiently
and lifting the 1 GB limit on its size.
This includes a unit test module, in src/test/modules/test_integerset.
It can be used to verify correctness, as a regression test, but if you run
it manully, it can also print memory usage and execution time of some of
the tests.
Author: Heikki Linnakangas, Andrey Borodin
Reviewed-by: Julien Rouhaud
Discussion: https://www.postgresql.org/message-id/b5e82599-1966-5783-733c-1a947ddb729f@iki.fi
This adds a flag "deterministic" to collations. If that is false,
such a collation disables various optimizations that assume that
strings are equal only if they are byte-wise equal. That then allows
use cases such as case-insensitive or accent-insensitive comparisons
or handling of strings with different Unicode normal forms.
This functionality is only supported with the ICU provider. At least
glibc doesn't appear to have any locales that work in a
nondeterministic way, so it's not worth supporting this for the libc
provider.
The term "deterministic comparison" in this context is from Unicode
Technical Standard #10
(https://unicode.org/reports/tr10/#Deterministic_Comparison).
This patch makes changes in three areas:
- CREATE COLLATION DDL changes and system catalog changes to support
this new flag.
- Many executor nodes and auxiliary code are extended to track
collations. Previously, this code would just throw away collation
information, because the eventually-called user-defined functions
didn't use it since they only cared about equality, which didn't
need collation information.
- String data type functions that do equality comparisons and hashing
are changed to take the (non-)deterministic flag into account. For
comparison, this just means skipping various shortcuts and tie
breakers that use byte-wise comparison. For hashing, we first need
to convert the input string to a canonical "sort key" using the ICU
analogue of strxfrm().
Reviewed-by: Daniel Verite <daniel@manitou-mail.org>
Reviewed-by: Peter Geoghegan <pg@bowt.ie>
Discussion: https://www.postgresql.org/message-id/flat/1ccc668f-4cbc-0bef-af67-450b47cdfee7@2ndquadrant.com
Trying to call the function with the top-most parent of a partition tree
was leading to a crash. In this case the correct result is to return
the top-most parent itself.
Reported-by: Álvaro Herrera
Author: Michael Paquier
Reviewed-by: Amit Langote
Discussion: https://postgr.es/m/20190322032612.GA323@alvherre.pgsql
When DefineIndex recurses to create constraints on partitions, it needs
to use the value returned by index_constraint_create to set up partition
dependencies. However, in the course of fixing the DEPENDENCY_INTERNAL_AUTO
mess, commit 1d92a0c9f7 introduced some code to that function that
clobbered the return value, causing the recorded OID to be of the wrong
object. Close examination of pg_depend after creating the tables leads
to indescribable objects :-( My sin (in commit bdc3d7fa23, while
preparing for DDL deparsing in event triggers) was to use a variable
name for the return value that's typically used for throwaway objects in
dependency-setting calls ("referenced"). Fix by changing the variable
names to match extended practice (the return value is "myself" rather
than "referenced".)
The pg_upgrade test notices the problem (in an indirect way: the pg_dump
outputs are in different order), but only if you create the objects in a
specific way that wasn't being used in the existing tests. Add a stanza
to leave some objects around that shows the bug.
Catversion bump because preexisting databases might have bogus pg_depend
entries.
Discussion: https://postgr.es/m/20190318204235.GA30360@alvherre.pgsql
These commands allow the argument type list to be omitted if there is
just one object that matches by name. However, if that syntax was
used with DROP IF EXISTS and there was more than one match, you got
a "function ... does not exist, skipping" notice message rather than a
truthful complaint about the ambiguity. This was basically due to
poor factorization and a rats-nest of logic, so refactor the relevant
lookup code to make it cleaner.
Note that this amounts to narrowing the scope of which sorts of
error conditions IF EXISTS will bypass. Per discussion, we only
intend it to skip no-such-object cases, not multiple-possible-matches
cases.
Per bug #15572 from Ash Marath. Although this definitely seems like
a bug, it's not clear that people would thank us for changing the
behavior in minor releases, so no back-patch.
David Rowley, reviewed by Julien Rouhaud and Pavel Stehule
Discussion: https://postgr.es/m/15572-ed1b9ed09503de8a@postgresql.org
LDAP servers can be advertised on a network with RFC 2782 DNS SRV
records. The OpenLDAP command-line tools automatically try to find
servers that way, if no server name is provided by the user. Teach
PostgreSQL to do the same using OpenLDAP's support functions, when
building with OpenLDAP.
For now, we assume that HAVE_LDAP_INITIALIZE (an OpenLDAP extension
available since OpenLDAP 2.0 and also present in Apple LDAP) implies
that you also have ldap_domain2hostlist() (which arrived in the same
OpenLDAP version and is also present in Apple LDAP).
Author: Thomas Munro
Reviewed-by: Daniel Gustafsson
Discussion: https://postgr.es/m/CAEepm=2hAnSfhdsd6vXsM6VZVN0br-FbAZ-O+Swk18S5HkCP=A@mail.gmail.com
This finishes a task we left undone in commit f1ad067fc, by extending
the delete-in-descending-OID-order rule to deletions triggered by
DROP OWNED BY. We've coped with machine-dependent deletion orders
one time too many, and the new issues caused by Peter G's recent
nbtree hacking seem like the last straw.
Discussion: https://postgr.es/m/E1h6eep-0001Mw-Vd@gemulon.postgresql.org
Teach nbtree to give some consideration to how "distinguishing"
candidate leaf page split points are. This should not noticeably affect
the balance of free space within each half of the split, while still
making suffix truncation truncate away significantly more attributes on
average.
The logic for choosing a leaf split point now uses a fallback mode in
the case where the page is full of duplicates and it isn't possible to
find even a minimally distinguishing split point. When the page is full
of duplicates, the split should pack the left half very tightly, while
leaving the right half mostly empty. Our assumption is that logical
duplicates will almost always be inserted in ascending heap TID order
with v4 indexes. This strategy leaves most of the free space on the
half of the split that will likely be where future logical duplicates of
the same value need to be placed.
The number of cycles added is not very noticeable. This is important
because deciding on a split point takes place while at least one
exclusive buffer lock is held. We avoid using authoritative insertion
scankey comparisons to save cycles, unlike suffix truncation proper. We
use a faster binary comparison instead.
Note that even pg_upgrade'd v3 indexes make use of these optimizations.
Benchmarking has shown that even v3 indexes benefit, despite the fact
that suffix truncation will only truncate non-key attributes in INCLUDE
indexes. Grouping relatively similar tuples together is beneficial in
and of itself, since it reduces the number of leaf pages that must be
accessed by subsequent index scans.
Author: Peter Geoghegan
Reviewed-By: Heikki Linnakangas
Discussion: https://postgr.es/m/CAH2-WzmmoLNQOj9mAD78iQHfWLJDszHEDrAzGTUMG3mVh5xWPw@mail.gmail.com
Make nbtree treat all index tuples as having a heap TID attribute.
Index searches can distinguish duplicates by heap TID, since heap TID is
always guaranteed to be unique. This general approach has numerous
benefits for performance, and is prerequisite to teaching VACUUM to
perform "retail index tuple deletion".
Naively adding a new attribute to every pivot tuple has unacceptable
overhead (it bloats internal pages), so suffix truncation of pivot
tuples is added. This will usually truncate away the "extra" heap TID
attribute from pivot tuples during a leaf page split, and may also
truncate away additional user attributes. This can increase fan-out,
especially in a multi-column index. Truncation can only occur at the
attribute granularity, which isn't particularly effective, but works
well enough for now. A future patch may add support for truncating
"within" text attributes by generating truncated key values using new
opclass infrastructure.
Only new indexes (BTREE_VERSION 4 indexes) will have insertions that
treat heap TID as a tiebreaker attribute, or will have pivot tuples
undergo suffix truncation during a leaf page split (on-disk
compatibility with versions 2 and 3 is preserved). Upgrades to version
4 cannot be performed on-the-fly, unlike upgrades from version 2 to
version 3. contrib/amcheck continues to work with version 2 and 3
indexes, while also enforcing stricter invariants when verifying version
4 indexes. These stricter invariants are the same invariants described
by "3.1.12 Sequencing" from the Lehman and Yao paper.
A later patch will enhance the logic used by nbtree to pick a split
point. This patch is likely to negatively impact performance without
smarter choices around the precise point to split leaf pages at. Making
these two mostly-distinct sets of enhancements into distinct commits
seems like it might clarify their design, even though neither commit is
particularly useful on its own.
The maximum allowed size of new tuples is reduced by an amount equal to
the space required to store an extra MAXALIGN()'d TID in a new high key
during leaf page splits. The user-facing definition of the "1/3 of a
page" restriction is already imprecise, and so does not need to be
revised. However, there should be a compatibility note in the v12
release notes.
Author: Peter Geoghegan
Reviewed-By: Heikki Linnakangas, Alexander Korotkov
Discussion: https://postgr.es/m/CAH2-WzkVb0Kom=R+88fDFb=JSxZMFvbHVC6Mn9LJ2n=X=kS-Uw@mail.gmail.com
Use dedicated struct to represent nbtree insertion scan keys. Having a
dedicated struct makes the difference between search type scankeys and
insertion scankeys a lot clearer, and simplifies the signature of
several related functions. This is based on a suggestion by Andrey
Lepikhov.
Streamline how unique index insertions cache binary search progress.
Cache the state of in-progress binary searches within _bt_check_unique()
for later instead of having callers avoid repeating the binary search in
an ad-hoc manner. This makes it easy to add a new optimization:
_bt_check_unique() now falls out of its loop immediately in the common
case where it's already clear that there couldn't possibly be a
duplicate.
The new _bt_check_unique() scheme makes it a lot easier to manage cached
binary search effort afterwards, from within _bt_findinsertloc(). This
is needed for the upcoming patch to make nbtree tuples unique by
treating heap TID as a final tiebreaker column. Unique key binary
searches need to restore lower and upper bounds. They cannot simply
continue to use the >= lower bound as the offset to insert at, because
the heap TID tiebreaker column must be used in comparisons for the
restored binary search (unlike the original _bt_check_unique() binary
search, where scankey's heap TID column must be omitted).
Author: Peter Geoghegan, Heikki Linnakangas
Reviewed-By: Heikki Linnakangas, Andrey Lepikhov
Discussion: https://postgr.es/m/CAH2-WzmE6AhUdk9NdWBf4K3HjWXZBX3+umC7mH7+WDrKcRtsOw@mail.gmail.com
Jsonpath grammar and scanner are both quite small. It doesn't worth complexity
to compile them separately. This commit makes grammar and scanner be compiled
at once. Therefore, jsonpath_gram.h and jsonpath_gram.h are no longer needed.
This commit also does some reorganization of code in jsonpath_gram.y.
Discussion: https://postgr.es/m/d47b2023-3ecb-5f04-d253-d557547cf74f%402ndQuadrant.com
Commit 6f6a6d8b1 introduced a delay of up to 2 seconds if we're trying
to request a checkpoint but the checkpointer hasn't started yet (or,
much less likely, our kill() call fails). However buildfarm experience
shows that that's not quite enough for slow or heavily-loaded machines.
There's no good reason to assume that the checkpointer won't start
eventually, so we may as well make the timeout much longer, say 60 sec.
However, if the caller didn't say CHECKPOINT_WAIT, it seems like a bad
idea to be waiting at all, much less for as long as 60 sec. We can
remove the need for that, and make this whole thing more robust, by
adjusting the code so that the existence of a pending checkpoint
request is clear from the contents of shared memory, and making sure
that the checkpointer process will notice it at startup even if it did
not get a signal. In this way there's no need for a non-CHECKPOINT_WAIT
call to wait at all; if it can't send the signal, it can nonetheless
assume that the checkpointer will eventually service the request.
A potential downside of this change is that "kill -INT" on the checkpointer
process is no longer enough to trigger a checkpoint, should anyone be
relying on something so hacky. But there's no obvious reason to do it
like that rather than issuing a plain old CHECKPOINT command, so we'll
assume that nobody is. There doesn't seem to be a way to preserve this
undocumented quasi-feature without introducing race conditions.
Since a principal reason for messing with this is to prevent intermittent
buildfarm failures, back-patch to all supported branches.
Discussion: https://postgr.es/m/27830.1552752475@sss.pgh.pa.us
Typedef name should be both unique and non-intersect with variable names
across all the sources. That makes both pg_indent and debuggers happy.
Discussion: https://postgr.es/m/23865.1552936099%40sss.pgh.pa.us
Running ALTER TABLE on any table will check if a TOAST table needs to be
added. On shared tables, this would previously fail, thus effectively
disabling ALTER TABLE for those tables. On (non-shared) system
catalogs, on the other hand, it would add a TOAST table, even though we
don't really want TOAST tables on some system catalogs. In some cases,
it would also fail with an error "AccessExclusiveLock required to add
toast table.", depending on what locks the ALTER TABLE actions had
already taken.
So instead, just ignore attempts to add TOAST tables to such tables,
outside of bootstrap mode, pretending they don't need one.
This allows running ALTER TABLE on such tables without messing up the
TOAST situation. Legitimate uses for ALTER TABLE on system catalogs
include setting reloptions (say, fillfactor or autovacuum settings).
(All this still requires allow_system_table_mods, which is independent
of this.)
Discussion: https://www.postgresql.org/message-id/flat/e49f825b-fb25-0bc8-8afc-d5ad895c7975@2ndquadrant.com
Unrecognized attribute names are supposed to be ignored. But the code
would error out on an unrecognized attribute value even if it did not
recognize the attribute name. So unrecognized attributes wouldn't
really be ignored unless the value happened to be one that matched a
recognized value. This would break some important cases where the
attribute would be processed by ucol_open() directly. Fix that and
add a test case.
The restructured code should also avoid compiler warnings about
initializing a UColAttribute value to -1, because the type might be an
unsigned enum. (reported by Andres Freund)
Aggregates have acquired a dozen or so optional attributes in recent
years for things like parallel query and moving-aggregate mode; the
lack of an OR REPLACE option to add or change these for an existing
agg makes extension upgrades gratuitously hard. Rectify.
Commit f2dec34e1 changed things so that printtup's output stringinfo
buffer was allocated outside the per-row temporary context, not inside
it. This creates a need to free that buffer explicitly when the temp
context is freed, but that was overlooked. In most cases, this is all
happening inside a portal or executor context that will go away shortly
anyhow, but that's not always true. Notably, the stringinfo ends up
getting leaked when JDBC uses row-at-a-time fetches. For a query
that returns wide rows, that adds up after awhile.
Per bug #15700 from Matthias Otterbach. Back-patch to v11 where the
faulty code was added.
Discussion: https://postgr.es/m/15700-8c408321a87d56bb@postgresql.org
Like commit f41551f61f, this aims
to make it easier to add non-Boolean options to VACUUM (or, in
this case, to ANALYZE). Instead of building up a bitmap of
options directly in the parser, build up a list of DefElem
objects and let ExecVacuum() sort it out; right now, we make
no use of the fact that a DefElem can carry an associated value,
but it will be easy to make that change in the future.
Masahiko Sawada
Discussion: http://postgr.es/m/CAD21AoATE4sn0jFFH3NcfUZXkU2BMbjBWB_kDj-XWYA-LXDcQA@mail.gmail.com
Many places need both, so this allows a few functions to take one
fewer parameter. More importantly, as soon as we add a VACUUM
option that takes a non-Boolean parameter, we need to replace
'int options' with a struct, and it seems better to think
of adding more fields to VacuumParams rather than passing around
both VacuumParams and a separate struct as well.
Patch by me, reviewed by Masahiko Sawada
Discussion: http://postgr.es/m/CA+Tgmob6g6-s50fyv8E8he7APfwCYYJ4z0wbZC2yZeSz=26CYQ@mail.gmail.com
In RI_FKey_pk_upd_check_required(), we check among other things
whether the old and new key are equal, so that we don't need to run
cascade actions when nothing has actually changed. This was using the
equality operator. But the effect of this is that if a value in the
primary key is changed to one that "looks" different but compares as
equal, the update is not propagated. (Examples are float -0 and 0 and
case-insensitive text.) This appears to violate the SQL standard, and
it also behaves inconsistently if in a multicolumn key another key is
also updated that would cause the row to compare as not equal.
To fix, if we are looking at the PK table in ri_KeysEqual(), then do a
bytewise comparison similar to record_image_eq() instead of using the
equality operators. This only makes a difference for ON UPDATE
CASCADE, but for consistency we treat all changes to the PK the same. For
the FK table, we continue to use the equality operators.
Discussion: https://www.postgresql.org/message-id/flat/3326fc2e-bc02-d4c5-e3e5-e54da466e89a@2ndquadrant.com
ce6afc6 has begun the refactoring work by plugging pg_rewind into a
central routine to update the control file, and left around two extra
copies, with one in xlog.c for the backend and one in pg_resetwal.c. By
adding an extra option to the central routine in controldata_utils.c to
control if a flush of the control file needs to be done, it is proving
to be straight-forward to make xlog.c and pg_resetwal.c use the central
code path at the condition of moving the wait event tracking there.
Hence, this allows to have only one central code path to update the
control file, shaving the code from the duplicates.
This refactoring actually fixes a problem in pg_resetwal. Previously,
the control file was first removed before being recreated. So if a
crash happened between the moment the file was removed and the moment
the file was created, then it would have been possible to not have a
control file anymore in the database folder.
Author: Fabien Coelho
Reviewed-by: Michael Paquier
Discussion: https://postgr.es/m/alpine.DEB.2.21.1903170935210.2506@lancre
Starting in ICU 54, collation customization attributes can be
specified in the locale string, for example
"@colStrength=primary;colCaseLevel=yes". Add support for this for
older ICU versions as well, by adding some minimal parsing of the
attributes in the locale string and calling ucol_setAttribute() on
them. This is essentially what never ICU versions do internally in
ucol_open(). This was we can offer this functionality in a consistent
way in all ICU versions supported by PostgreSQL.
Also add some tests for ICU collation customization.
Reported-by: Daniel Verite <daniel@manitou-mail.org>
Discussion: https://www.postgresql.org/message-id/0270ebd4-f67c-8774-1a5a-91adfb9bb41f@2ndquadrant.com
Add support of numeric error suppression to jsonpath as it's required by
standard. This commit doesn't use PG_TRY()/PG_CATCH() in order to implement
that. Instead, it provides internal versions of numeric functions used, which
support error suppression.
Discussion: https://postgr.es/m/fcc6fc6a-b497-f39a-923d-aa34d0c588e8%402ndQuadrant.com
Author: Alexander Korotkov, Nikita Glukhov
Reviewed-by: Tomas Vondra
SQL 2016 standards among other things contains set of SQL/JSON features for
JSON processing inside of relational database. The core of SQL/JSON is JSON
path language, allowing access parts of JSON documents and make computations
over them. This commit implements partial support JSON path language as
separate datatype called "jsonpath". The implementation is partial because
it's lacking datetime support and suppression of numeric errors. Missing
features will be added later by separate commits.
Support of SQL/JSON features requires implementation of separate nodes, and it
will be considered in subsequent patches. This commit includes following
set of plain functions, allowing to execute jsonpath over jsonb values:
* jsonb_path_exists(jsonb, jsonpath[, jsonb, bool]),
* jsonb_path_match(jsonb, jsonpath[, jsonb, bool]),
* jsonb_path_query(jsonb, jsonpath[, jsonb, bool]),
* jsonb_path_query_array(jsonb, jsonpath[, jsonb, bool]).
* jsonb_path_query_first(jsonb, jsonpath[, jsonb, bool]).
This commit also implements "jsonb @? jsonpath" and "jsonb @@ jsonpath", which
are wrappers over jsonpath_exists(jsonb, jsonpath) and jsonpath_predicate(jsonb,
jsonpath) correspondingly. These operators will have an index support
(implemented in subsequent patches).
Catversion bumped, to add new functions and operators.
Code was written by Nikita Glukhov and Teodor Sigaev, revised by me.
Documentation was written by Oleg Bartunov and Liudmila Mantrova. The work
was inspired by Oleg Bartunov.
Discussion: https://postgr.es/m/fcc6fc6a-b497-f39a-923d-aa34d0c588e8%402ndQuadrant.com
Author: Nikita Glukhov, Teodor Sigaev, Alexander Korotkov, Oleg Bartunov, Liudmila Mantrova
Reviewed-by: Tomas Vondra, Andrew Dunstan, Pavel Stehule, Alexander Korotkov
Some buildfarm members using CLOBBER_CACHE_ALWAYS have been having OOM
problems of late. Commit 2455ab488 addressed this problem by recovering
space transiently used within RelationBuildPartitionDesc, but it turns
out that leaves quite a lot on the table, because other subroutines of
RelationBuildDesc also leak memory like mad. Let's move the temp-context
management into RelationBuildDesc so that leakage from the other
subroutines is also recovered.
I examined this issue by arranging for postgres.c to dump the size of
MessageContext just before resetting it in each command cycle, and
then running the update.sql regression test (which is one of the two
that are seeing buildfarm OOMs) with and without CLOBBER_CACHE_ALWAYS.
Before 2455ab488, the peak space usage with CCA was as much as 250MB.
That patch got it down to ~80MB, but with this patch it's about 0.5MB,
and indeed the space usage now seems nearly indistinguishable from a
non-CCA build.
RelationBuildDesc's traditional behavior of not worrying about leaking
transient data is of many years' standing, so I'm pretty hesitant to
change that without more evidence that it'd be useful in a normal build.
(So far as I can see, non-CCA memory consumption is about the same with
or without this change, whuch if anything suggests that it isn't useful.)
Hence, configure the patch so that we recover space only when
CLOBBER_CACHE_ALWAYS or CLOBBER_CACHE_RECURSIVELY is defined. However,
that choice can be overridden at compile time, in case somebody would
like to do some performance testing and try to develop evidence for
changing that decision.
It's possible that we ought to back-patch this change, but in the
absence of back-branch OOM problems in the buildfarm, I'm not in
a hurry to do that.
Discussion: https://postgr.es/m/CA+TgmoY3bRmGB6-DUnoVy5fJoreiBJ43rwMrQRCdPXuKt4Ykaw@mail.gmail.com
Add a separate walreceiver API function walrcv_server_version() to get
the version of the remote server, instead of doing it as part of
walrcv_identify_system(). This allows the server version to be
available even for uses that don't call IDENTIFY_SYSTEM, and it seems
cleaner anyway.
This is for an upcoming patch, not currently used.
Reviewed-by: Michael Paquier <michael@paquier.xyz>
Discussion: https://www.postgresql.org/message-id/20190115071359.GF1433@paquier.xyz
Previously, the SERIALIZABLE isolation level prevented parallel query
from being used. Allow the two features to be used together by
sharing the leader's SERIALIZABLEXACT with parallel workers.
An extra per-SERIALIZABLEXACT LWLock is introduced to make it safe to
share, and new logic is introduced to coordinate the early release
of the SERIALIZABLEXACT required for the SXACT_FLAG_RO_SAFE
optimization, as follows:
The first backend to observe the SXACT_FLAG_RO_SAFE flag (set by
some other transaction) will 'partially release' the SERIALIZABLEXACT,
meaning that the conflicts and locks it holds are released, but the
SERIALIZABLEXACT itself will remain active because other backends
might still have a pointer to it.
Whenever any backend notices the SXACT_FLAG_RO_SAFE flag, it clears
its own MySerializableXact variable and frees local resources so that
it can skip SSI checks for the rest of the transaction. In the
special case of the leader process, it transfers the SERIALIZABLEXACT
to a new variable SavedSerializableXact, so that it can be completely
released at the end of the transaction after all workers have exited.
Remove the serializable_okay flag added to CreateParallelContext() by
commit 9da0cc35, because it's now redundant.
Author: Thomas Munro
Reviewed-by: Haribabu Kommi, Robert Haas, Masahiko Sawada, Kevin Grittner
Discussion: https://postgr.es/m/CAEepm=0gXGYhtrVDWOTHS8SQQy_=S9xo+8oCxGLWZAOoeJ=yzQ@mail.gmail.com
The idea was to generate all the junk in a destroyable subcontext rather
than leaking it in the caller's context, but partition_bounds_create was
still being called in the caller's context, allowing plenty of scope for
leakage. Also, get_rel_relkind() was still being called in the rel's
rd_pdcxt, creating a risk of session-lifespan memory wastage.
Simplify the logic a bit while at it. Also, reduce rd_pdcxt to
ALLOCSET_SMALL_SIZES, since it seems likely to not usually be big.
Probably something like this needs to be back-patched into v11,
but for now let's get some buildfarm testing on this.
Discussion: https://postgr.es/m/15943.1552601288@sss.pgh.pa.us
The assertions added by commits 34ea1ab7f et al found another problem:
set_dummy_rel_pathlist and mark_dummy_rel were failing to label
the dummy paths they create with the correct outer_relids, in case
the relation is necessarily parameterized due to having lateral
references in its tlist. It's likely that this has no user-visible
consequences in production builds, at the moment; but still an assertion
failure is a bad thing, so back-patch the fix.
Per bug #15694 from Roman Zharkov (via Alexander Lakhin)
and an independent report by Tushar Ahuja.
Discussion: https://postgr.es/m/15694-74f2ca97e7044f7f@postgresql.org
Discussion: https://postgr.es/m/7d72ab20-c725-3ce2-f99d-4e64dd8a0de6@enterprisedb.com
In normal builds, this isn't very important, because the leaks go
into fairly short-lived contexts, but under CLOBBER_CACHE_ALWAYS,
this can result in leaking hundreds of megabytes into MessageContext,
which probably explains recent failures on hyrax.
This may or may not be the best long-term strategy for dealing
with this leak, but we can change it later if we come up with
something better. For now, do this to make the buildfarm green
again (hopefully). Commit 898e5e3290
seems to have exacerbated this problem for reasons that are not
quite clear, but I don't believe it's actually the cause.
Discussion: http://postgr.es/m/CA+TgmoY3bRmGB6-DUnoVy5fJoreiBJ43rwMrQRCdPXuKt4Ykaw@mail.gmail.com
Commit a6417078c missed updating some comments in transam.h about
reservation of high OIDs for development purposes. Also tamp down
an over-optimistic comment there about how easy it'd be to change
FirstNormalObjectId.
Earlier, commit 09568ec3d failed to update bki.sgml for the split
between genbki.pl-assigned OIDs and those assigned during initdb.
Also fix genbki.pl so that it will complain if it overruns
that split. It's possible that doing so would have no very bad
consequences, but that's no excuse for not detecting it.
If existing CHECK or NOT NULL constraints preclude the presence
of nulls, we need not look to see whether any are present.
Sergei Kornilov, reviewed by Stephen Frost, Ildar Musin, David Rowley,
and by me.
Discussion: http://postgr.es/m/81911511895540@web58j.yandex.ru
The current tool name is too restrictive and focuses only on verifying
checksums. As more options to control checksums for an offline cluster
are planned to be added, switch to a more generic name. Documentation
as well as all past references to the tool are updated.
Author: Michael Paquier
Reviewed-by: Michael Banck, Fabien Coelho, Seigei Kornilov
Discussion: https://postgr.es/m/20181221201616.GD4974@nighthawk.caipicrew.dd-dns.de
Commit 40dae7ec53, which made the nbtree page split algorithm more
robust, made _bt_insert_parent() only unlock the right child of the
parent page before inserting a new downlink into the parent. Update a
comment from the Berkeley days claiming that both left and right child
pages are unlocked before the new downlink actually gets inserted.
The claim that it is okay to release both locks early based on Lehman
and Yao's say-so never made much sense. Lehman and Yao must sometimes
"couple" buffer locks across a pair of internal pages when relocating a
downlink, unlike the corresponding code within _bt_getstack().
The SQL:2016 standard adds support for the hyperbolic functions
sinh(), cosh(), and tanh(). POSIX has long required libm to
provide those functions as well as their inverses asinh(),
acosh(), atanh(). Hence, let's just expose the libm functions
to the SQL level. As with the trig functions, we only implement
versions for float8, not numeric.
For the moment, we'll assume that all platforms actually do have
these functions; if experience teaches otherwise, some autoconf
effort may be needed.
SQL:2016 also adds support for base-10 logarithm, but with the
function name log10(), whereas the name we've long used is log().
Add aliases named log10() for the float8 and numeric versions.
Lætitia Avrot
Discussion: https://postgr.es/m/CAB_COdguG22LO=rnxDQ2DW1uzv8aQoUzyDQNJjrR4k00XSgm5w@mail.gmail.com
In the v11-era commits that taught genbki.pl to resolve symbolic
OID references in the initial catalog data, we didn't bother to
make every last reference symbolic; some of the catalogs have so
few initial rows that it didn't seem worthwhile.
However, the new project policy that OIDs assigned by new patches
should be automatically renumberable changes this calculus.
A patch that wants to add a row in one of these catalogs would have
a problem when the OID it assigns gets renumbered. Hence, do the
mop-up work needed to make all OID references in initial data be
symbolic, and establish an associated project policy that we'll
never again write a hard-wired OID reference there.
No catversion bump since the contents of postgres.bki aren't
actually changed by this commit.
Discussion: https://postgr.es/m/CAH2-WzmMTGMcPuph4OvsO7Ykut0AOCF_i-=eaochT0dd2BN9CQ@mail.gmail.com
This commit adds a Perl script renumber_oids.pl, which can reassign a
range of manually-assigned OIDs to someplace else by modifying OID
fields of the catalog *.dat files and OID-assigning macros in the
catalog *.h files.
Up to now, we've encouraged new patches that need manually-assigned
OIDs to use OIDs just above the range of existing OIDs. Predictably,
this leads to patches stepping on each others' toes, as whichever
one gets committed first creates an OID conflict that other patch
author(s) have to resolve manually. With the availability of
renumber_oids.pl, we can eliminate a lot of this hassle.
The new project policy, therefore, is:
* Encourage new patches to use high OIDs (the documentation suggests
choosing a block of OIDs at random in 8000..9999).
* After feature freeze in each development cycle, run renumber_oids.pl
to move all such OIDs down to lower numbers, thus freeing the high OID
range for the next development cycle.
This plan should greatly reduce the risk of OID collisions between
concurrently-developed patches. Also, if such a collision happens
anyway, we have the option to resolve it without much effort by doing
an off-schedule OID renumbering to get the first-committed patch out
of the way. Or a patch author could use renumber_oids.pl to change
their patch's assignments without much pain.
This approach does put a premium on not hard-wiring any OID values
in places where renumber_oids.pl and genbki.pl can't fix them.
Project practice in that respect seems to be pretty good already,
but a follow-on patch will sand down some rough edges.
John Naylor and Tom Lane, per an idea of Peter Geoghegan's
Discussion: https://postgr.es/m/CAH2-WzmMTGMcPuph4OvsO7Ykut0AOCF_i-=eaochT0dd2BN9CQ@mail.gmail.com
In commit 960df2a971 ("Correctly assess parallel-safety of tlists when
SRFs are used."), the testing of scan/join target was done incorrectly,
which caused a plan-quality problem. Backpatch through to v11 where
the aforementioned commit went in, since this is a regression from v10.
Author: Etsuro Fujita
Reviewed-by: Robert Haas and Tom Lane
Discussion: https://postgr.es/m/5C75303E.8020303@lab.ntt.co.jp
Historically guc.c has just refused examples like set work_mem = '30.1GB',
but it seems more useful for it to take that and round off the value to
some reasonable approximation of what the user said. Just rounding to
the parameter's native unit would work, but it would lead to rather
silly-looking settings, such as 31562138kB for this example. Instead
let's round to the nearest multiple of the next smaller unit (if any),
producing 30822MB.
Also, do the units conversion math in floating point and round to integer
(if needed) only at the end. This produces saner results for inputs that
aren't exact multiples of the parameter's native unit, and removes another
difference in the behavior for integer vs. float parameters.
In passing, document the ability to use hex or octal input where it
ought to be documented.
Discussion: https://postgr.es/m/1798.1552165479@sss.pgh.pa.us
Further buildfarm testing shows that on the machines that are failing
ac75959cd's test case, what we're actually getting from strtod("-infinity")
is a syntax error (endptr == value) not ERANGE at all. This test case
is not worth carrying two sets of expected output for, so just remove it,
and revert commit b212245f9's misguided attempt to work around the platform
dependency.
Discussion: https://postgr.es/m/E1h33xk-0001Og-Gs@gemulon.postgresql.org
Previously ParallelTableScanDescData was just a member in BTShared,
but after c2fe139c2 that doesn't guarantee sufficient alignment as
specific AMs might (are likely to) need atomic variables in the
struct.
One might think that MAXALIGNing would be sufficient, but as a
comment in shm_toc_allocate() explains, that's not enough. For now,
copy the hack described there.
For parallel sequential scans no such change is needed, as its
allocations go through shm_toc_allocate().
An alternative approach would have been to allocate the parallel scan
descriptor in a separate TOC entry, but there seems little benefit in
doing so.
Per buildfarm member dromedary.
Author: Andres Freund
Discussion: https://postgr.es/m/20190311203126.ty5gbfz42gjbm6i6@alap3.anarazel.de
Too allow table accesses to be not directly dependent on heap, several
new abstractions are needed. Specifically:
1) Heap scans need to be generalized into table scans. Do this by
introducing TableScanDesc, which will be the "base class" for
individual AMs. This contains the AM independent fields from
HeapScanDesc.
The previous heap_{beginscan,rescan,endscan} et al. have been
replaced with a table_ version.
There's no direct replacement for heap_getnext(), as that returned
a HeapTuple, which is undesirable for a other AMs. Instead there's
table_scan_getnextslot(). But note that heap_getnext() lives on,
it's still used widely to access catalog tables.
This is achieved by new scan_begin, scan_end, scan_rescan,
scan_getnextslot callbacks.
2) The portion of parallel scans that's shared between backends need
to be able to do so without the user doing per-AM work. To achieve
that new parallelscan_{estimate, initialize, reinitialize}
callbacks are introduced, which operate on a new
ParallelTableScanDesc, which again can be subclassed by AMs.
As it is likely that several AMs are going to be block oriented,
block oriented callbacks that can be shared between such AMs are
provided and used by heap. table_block_parallelscan_{estimate,
intiialize, reinitialize} as callbacks, and
table_block_parallelscan_{nextpage, init} for use in AMs. These
operate on a ParallelBlockTableScanDesc.
3) Index scans need to be able to access tables to return a tuple, and
there needs to be state across individual accesses to the heap to
store state like buffers. That's now handled by introducing a
sort-of-scan IndexFetchTable, which again is intended to be
subclassed by individual AMs (for heap IndexFetchHeap).
The relevant callbacks for an AM are index_fetch_{end, begin,
reset} to create the necessary state, and index_fetch_tuple to
retrieve an indexed tuple. Note that index_fetch_tuple
implementations need to be smarter than just blindly fetching the
tuples for AMs that have optimizations similar to heap's HOT - the
currently alive tuple in the update chain needs to be fetched if
appropriate.
Similar to table_scan_getnextslot(), it's undesirable to continue
to return HeapTuples. Thus index_fetch_heap (might want to rename
that later) now accepts a slot as an argument. Core code doesn't
have a lot of call sites performing index scans without going
through the systable_* API (in contrast to loads of heap_getnext
calls and working directly with HeapTuples).
Index scans now store the result of a search in
IndexScanDesc->xs_heaptid, rather than xs_ctup->t_self. As the
target is not generally a HeapTuple anymore that seems cleaner.
To be able to sensible adapt code to use the above, two further
callbacks have been introduced:
a) slot_callbacks returns a TupleTableSlotOps* suitable for creating
slots capable of holding a tuple of the AMs
type. table_slot_callbacks() and table_slot_create() are based
upon that, but have additional logic to deal with views, foreign
tables, etc.
While this change could have been done separately, nearly all the
call sites that needed to be adapted for the rest of this commit
also would have been needed to be adapted for
table_slot_callbacks(), making separation not worthwhile.
b) tuple_satisfies_snapshot checks whether the tuple in a slot is
currently visible according to a snapshot. That's required as a few
places now don't have a buffer + HeapTuple around, but a
slot (which in heap's case internally has that information).
Additionally a few infrastructure changes were needed:
I) SysScanDesc, as used by systable_{beginscan, getnext} et al. now
internally uses a slot to keep track of tuples. While
systable_getnext() still returns HeapTuples, and will so for the
foreseeable future, the index API (see 1) above) now only deals with
slots.
The remainder, and largest part, of this commit is then adjusting all
scans in postgres to use the new APIs.
Author: Andres Freund, Haribabu Kommi, Alvaro Herrera
Discussion:
https://postgr.es/m/20180703070645.wchpu5muyto5n647@alap3.anarazel.dehttps://postgr.es/m/20160812231527.GA690404@alvherre.pgsql
... as well as its implementation from backend/access/hash/hashfunc.c to
backend/utils/hash/hashfn.c.
access/hash is the place for the hash index AM, not really appropriate
for generic facilities, which is what hash_any is; having things the old
way meant that anything using hash_any had to include the AM's include
file, pointlessly polluting its namespace with unrelated, unnecessary
cruft.
Also move the HTEqual strategy number to access/stratnum.h from
access/hash.h.
To avoid breaking third-party extension code, add an #include
"utils/hashutils.h" to access/hash.h. (An easily removed line by
committers who enjoy their asbestos suits to protect them from angry
extension authors.)
Discussion: https://postgr.es/m/201901251935.ser5e4h6djt2@alvherre.pgsql
Instead, just proceed with the infinity or zero result that it should
return for overflow/underflow. This avoids a platform dependency,
in that various versions of strtod are inconsistent about whether they
signal ERANGE for a value that's specified as infinity.
It's possible this won't be enough to remove the buildfarm failures
we're seeing from ac75959cd, in which case I'll take out the infinity
test case that commit added. But first let's see if we can fix it.
Discussion: https://postgr.es/m/E1h33xk-0001Og-Gs@gemulon.postgresql.org
93473c6 has removed openLogOff, changing on the way the error message
which is used to report partial writes to WAL segments. The
newly-introduced error message used the offset up to which the write has
happened, keeping always the same total length to write. This changes
the error message so as the number of bytes left to write are reported.
Reported-by: Michael Paquier
Author: Robert Haas
Discussion: https://postgr.es/m/20190306235251.GA17293@paquier.xyz
This reverts commit bd09503e63.
Per discussion, it seems like what we should do instead is to
reduce the default value of autovacuum_vacuum_cost_delay by the
same factor. That's functionally equivalent as long as the
platform can accurately service the smaller delay request, which
should be true on anything released in the last 10 years or more.
And smaller, more-closely-spaced delays are better in terms of
providing a steady I/O load.
Discussion: https://postgr.es/m/28720.1552101086@sss.pgh.pa.us
This change makes it possible to specify sub-millisecond delays,
which work well on most modern platforms, though that was not true
when the cost-delay feature was designed.
To support this without breaking existing configuration entries,
improve guc.c to allow floating-point GUCs to have units. Also,
allow "us" (microseconds) as an input/output unit for time-unit GUCs.
(It's not allowed as a base unit, at least not yet.)
Likewise change the autovacuum_vacuum_cost_delay reloption to be
floating-point; this forces a catversion bump because the layout of
StdRdOptions changes.
This patch doesn't in itself change the default values or allowed
ranges for these parameters, and it should not affect the behavior
for any already-allowed setting for them.
Discussion: https://postgr.es/m/1798.1552165479@sss.pgh.pa.us
This should reduce confusion in cases where we've applied a units
conversion, so that the number being reported (and the quoted range
limits) are in some other units than what the user gave in the
setting we're rejecting.
Some of the changes here assume that float GUCs can have units,
which isn't true just yet, but will be shortly.
Discussion: https://postgr.es/m/3811.1552169665@sss.pgh.pa.us
None of the code that uses GUC values is really prepared for them to
hold NaN, but parse_real() didn't have any defense against accepting
such a value. Treat it the same as a syntax error.
I haven't attempted to analyze the exact consequences of setting any
of the float GUCs to NaN, but since they're quite unlikely to be good,
this seems like a back-patchable bug fix.
Note: we don't need an explicit test for +-Infinity because those will
be rejected by existing range checks. I added a regression test for
that in HEAD, but not older branches because the spelling of the value
in the error message will be platform-dependent in branches where we
don't always use port/snprintf.c.
Discussion: https://postgr.es/m/1798.1552165479@sss.pgh.pa.us
Similarly to B-tree, GiST index access method gets support of INCLUDE
attributes. These attributes aren't used for tree navigation and aren't
present in non-leaf pages. But they are present in leaf pages and can be
fetched during index-only scan.
The point of having INCLUDE attributes in GiST indexes is slightly different
from the point of having them in B-tree. The main point of INCLUDE attributes
in B-tree is to define UNIQUE constraint over part of attributes enabled for
index-only scan. In GiST the main point of INCLUDE attributes is to use
index-only scan for attributes, whose data types don't have GiST opclasses.
Discussion: https://postgr.es/m/73A1A452-AD5F-40D4-BD61-978622FF75C1%40yandex-team.ru
Author: Andrey Borodin, with small changes by me
Reviewed-by: Andreas Karlsson
This allows a login to require both that the cn of the certificate
matches (like authentication type cert) *and* that another
authentication method (such as password or kerberos) succeeds as well.
The old value of clientcert=1 maps to the new clientcert=verify-ca,
clientcert=0 maps to the new clientcert=no-verify, and the new option
erify-full will add the validation of the CN.
Author: Julian Markwort, Marius Timmer
Reviewed by: Magnus Hagander, Thomas Munro
This adds a column that counts how many checksum failures have occurred
on files belonging to a specific database. Both checksum failures
during normal backend processing and those created when a base backup
detects a checksum failure are counted.
Author: Magnus Hagander
Reviewed by: Julien Rouhaud
When the timezone is UTC, timestamptz and timestamp are binary coercible
in both directions. See b8a18ad485 and
c22ecc6562 for the previous attempt in
this problem space. Skip the table rewrite; for now, continue to
needlessly rewrite any index on an affected column.
Reviewed by Simon Riggs and Tom Lane.
Discussion: https://postgr.es/m/20190226061450.GA1665944@rfd.leadboat.com
This fixes two sets of issues related to the use of transient files in
the backend:
1) OpenTransientFile() has been used in some code paths with read-write
flags while read-only is sufficient, so switch those calls to be
read-only where necessary. These have been reported by Joe Conway.
2) When opening transient files, it is up to the caller to close the
file descriptors opened. In error code paths, CloseTransientFile() gets
called to clean up things before issuing an error. However in normal
exit paths, a lot of callers of CloseTransientFile() never actually
reported errors, which could leave a file descriptor open without
knowing about it. This is an issue I complained about a couple of
times, but never had the courage to write and submit a patch, so here we
go.
Note that one frontend code path is impacted by this commit so as an
error is issued when fetching control file data, making backend and
frontend to be treated consistently.
Reported-by: Joe Conway, Michael Paquier
Author: Michael Paquier
Reviewed-by: Álvaro Herrera, Georgios Kokolatos, Joe Conway
Discussion: https://postgr.es/m/20190301023338.GD1348@paquier.xyz
Discussion: https://postgr.es/m/c49b69ec-e2f7-ff33-4f17-0eaa4f2cef27@joeconway.com
Certain libxml2 versions (such as the 2.7.6 commonly seen in older
distributions, but apparently only on x86_64) contain a bug that causes
xmlCopyNode, when called on a XML_DOCUMENT_NODE, to return a node that
xmlFreeNode crashes on. Arrange to call xmlFreeDoc instead of
xmlFreeNode for those nodes.
Per buildfarm members lapwing and grison.
Author: Pavel Stehule, light editing by Álvaro.
Discussion: https://postgr.es/m/20190308024436.GA2374@alvherre.pgsql
Correctly process nodes of more types than previously. In some cases,
nodes were being ignored (nothing was output); in other cases, trying to
return them resulted in errors about unrecognized nodes. In yet other
cases, necessary escaping (of XML special characters) was not being
done. Fix all those (as far as the authors could find) and add
regression tests cases verifying the new behavior.
I (Álvaro) was of two minds about backpatching these changes. They do
seem bugfixes that would benefit most users of the affected functions;
but on the other hand it would change established behavior in minor
releases, so it seems prudent not to.
Authors: Pavel Stehule, Markus Winand, Chapman Flack
Discussion:
https://postgr.es/m/CAFj8pRA6J25CtAZ2TuRvxK3gat7-bBUYh0rfE2yM7Hj9GD14Dg@mail.gmail.comhttps://postgr.es/m/8BDB0627-2105-4564-AA76-7849F028B96E@winand.at
The elephant in the room as pointed out by Chapman Flack, not fixed in
this commit, is that we still have XMLTABLE operating on XPath 1.0
instead of the standard-mandated XQuery (or even its subset XPath 2.0).
Fixing that is a major undertaking, however.
When we introduced separate ProjectSetPath nodes for application of
set-returning functions in v10, we inadvertently broke some cases where
we're supposed to recognize that the result of a subquery is known to be
empty (contain zero rows). That's because IS_DUMMY_REL was just looking
for a childless AppendPath without allowing for a ProjectSetPath being
possibly stuck on top. In itself, this didn't do anything much worse
than produce slightly worse plans for some corner cases.
Then in v11, commit 11cf92f6e rearranged things to allow the scan/join
targetlist to be applied directly to partial paths before they get
gathered. But it inserted a short-circuit path for dummy relations
that was a little too short: it failed to insert a ProjectSetPath node
at all for a targetlist containing set-returning functions, resulting in
bogus "set-valued function called in context that cannot accept a set"
errors, as reported in bug #15669 from Madelaine Thibaut.
The best way to fix this mess seems to be to reimplement IS_DUMMY_REL
so that it drills down through any ProjectSetPath nodes that might be
there (and it seems like we'd better allow for ProjectionPath as well).
While we're at it, make it look at rel->pathlist not cheapest_total_path,
so that it gives the right answer independently of whether set_cheapest
has been done lately. That dependency looks pretty shaky in the context
of code like apply_scanjoin_target_to_paths, and even if it's not broken
today it'd certainly bite us at some point. (Nastily, unsafe use of the
old coding would almost always work; the hazard comes down to possibly
looking through a dangling pointer, and only once in a blue moon would
you find something there that resulted in the wrong answer.)
It now looks like it was a mistake for IS_DUMMY_REL to be a macro: if
there are any extensions using it, they'll continue to use the old
inadequate logic until they're recompiled, after which they'll fail
to load into server versions predating this fix. Hopefully there are
few such extensions.
Having fixed IS_DUMMY_REL, the special path for dummy rels in
apply_scanjoin_target_to_paths is unnecessary as well as being wrong,
so we can just drop it.
Also change a few places that were testing for partitioned-ness of a
planner relation but not using IS_PARTITIONED_REL for the purpose; that
seems unsafe as well as inconsistent, plus it required an ugly hack in
apply_scanjoin_target_to_paths.
In passing, save a few cycles in apply_scanjoin_target_to_paths by
skipping processing of pre-existing paths for partitioned rels,
and do some cosmetic cleanup and comment adjustment in that function.
I renamed IS_DUMMY_PATH to IS_DUMMY_APPEND with the intention of breaking
any code that might be using it, since in almost every case that would
be wrong; IS_DUMMY_REL is what to be using instead.
In HEAD, also make set_dummy_rel_pathlist static (since it's no longer
used from outside allpaths.c), and delete is_dummy_plan, since it's no
longer used anywhere.
Back-patch as appropriate into v11 and v10.
Tom Lane and Julien Rouhaud
Discussion: https://postgr.es/m/15669-02fb3296cca26203@postgresql.org
We still require AccessExclusiveLock on the partition itself, because
otherwise an insert that violates the newly-imposed partition
constraint could be in progress at the same time that we're changing
that constraint; only the lock level on the parent relation is
weakened.
To make this safe, we have to cope with (at least) three separate
problems. First, relevant DDL might commit while we're in the process
of building a PartitionDesc. If so, find_inheritance_children() might
see a new partition while the RELOID system cache still has the old
partition bound cached, and even before invalidation messages have
been queued. To fix that, if we see that the pg_class tuple seems to
be missing or to have a null relpartbound, refetch the value directly
from the table. We can't get the wrong value, because DETACH PARTITION
still requires AccessExclusiveLock throughout; if we ever want to
change that, this will need more thought. In testing, I found it quite
difficult to hit even the null-relpartbound case; the race condition
is extremely tight, but the theoretical risk is there.
Second, successive calls to RelationGetPartitionDesc might not return
the same answer. The query planner will get confused if lookup up the
PartitionDesc for a particular relation does not return a consistent
answer for the entire duration of query planning. Likewise, query
execution will get confused if the same relation seems to have a
different PartitionDesc at different times. Invent a new
PartitionDirectory concept and use it to ensure consistency. This
ensures that a single invocation of either the planner or the executor
sees the same view of the PartitionDesc from beginning to end, but it
does not guarantee that the planner and the executor see the same
view. Since this allows pointers to old PartitionDesc entries to
survive even after a relcache rebuild, also postpone removing the old
PartitionDesc entry until we're certain no one is using it.
For the most part, it seems to be OK for the planner and executor to
have different views of the PartitionDesc, because the executor will
just ignore any concurrently added partitions which were unknown at
plan time; those partitions won't be part of the inheritance
expansion, but invalidation messages will trigger replanning at some
point. Normally, this happens by the time the very next command is
executed, but if the next command acquires no locks and executes a
prepared query, it can manage not to notice until a new transaction is
started. We might want to tighten that up, but it's material for a
separate patch. There would still be a small window where a query
that started just after an ATTACH PARTITION command committed might
fail to notice its results -- but only if the command starts before
the commit has been acknowledged to the user. All in all, the warts
here around serializability seem small enough to be worth accepting
for the considerable advantage of being able to add partitions without
a full table lock.
Although in general the consequences of new partitions showing up
between planning and execution are limited to the query not noticing
the new partitions, run-time partition pruning will get confused in
that case, so that's the third problem that this patch fixes.
Run-time partition pruning assumes that indexes into the PartitionDesc
are stable between planning and execution. So, add code so that if
new partitions are added between plan time and execution time, the
indexes stored in the subplan_map[] and subpart_map[] arrays within
the plan's PartitionedRelPruneInfo get adjusted accordingly. There
does not seem to be a simple way to generalize this scheme to cope
with partitions that are removed, mostly because they could then get
added back again with different bounds, but it works OK for added
partitions.
This code does not try to ensure that every backend participating in
a parallel query sees the same view of the PartitionDesc. That
currently doesn't matter, because we never pass PartitionDesc
indexes between backends. Each backend will ignore the concurrently
added partitions which it notices, and it doesn't matter if different
backends are ignoring different sets of concurrently added partitions.
If in the future that matters, for example because we allow writes in
parallel query and want all participants to do tuple routing to the same
set of partitions, the PartitionDirectory concept could be improved to
share PartitionDescs across backends. There is a draft patch to
serialize and restore PartitionDescs on the thread where this patch
was discussed, which may be a useful place to start.
Patch by me. Thanks to Alvaro Herrera, David Rowley, Simon Riggs,
Amit Langote, and Michael Paquier for discussion, and to Alvaro
Herrera for some review.
Discussion: http://postgr.es/m/CA+Tgmobt2upbSocvvDej3yzokd7AkiT+PvgFH+a9-5VV1oJNSQ@mail.gmail.com
Discussion: http://postgr.es/m/CA+TgmoZE0r9-cyA-aY6f8WFEROaDLLL7Vf81kZ8MtFCkxpeQSw@mail.gmail.com
Discussion: http://postgr.es/m/CA+TgmoY13KQZF-=HNTrt9UYWYx3_oYOQpu9ioNT49jGgiDpUEA@mail.gmail.com
This clause is used to indicate the passing mode of a XML document, but
we were doing it wrong: we accepted BY REF and ignored it, and rejected
BY VALUE as a syntax error. The reality, however, is that documents are
always passed BY VALUE, so rejecting that clause was silly. Change
things so that we accept BY VALUE.
BY REF continues to be accepted, and continues to be ignored.
Author: Chapman Flack
Reviewed-by: Pavel Stehule
Discussion: https://postgr.es/m/5C297BB7.9070509@anastigmatix.net
Before commit 3fa2bb31 this type appeared in the catalogs to
select which of several block storage mechanisms each relation
used.
New features under development propose to revive the concept of
different block storage managers for new kinds of data accessed
via bufmgr.c, but don't need to put references to them in the
catalogs. So, avoid useless maintenance work on this type by
dropping it. Update some regression tests that were referencing
it where any type would do.
Discussion: https://postgr.es/m/CA%2BhUKG%2BDE0mmiBZMtZyvwWtgv1sZCniSVhXYsXkvJ_Wo%2B83vvw%40mail.gmail.com
Until now the the slot to store the conflicting tuple, and the result
of the ON CONFLICT SET, where reused between partitions. That
necessitated changing slots descriptor when switching partitions.
Besides the overhead of switching descriptors on a slot (which
requires memory allocations and prevents JITing), that's importantly
also problematic for tableam. There individual partitions might belong
to different tableams, needing different kinds of slots.
In passing also fix ExecOnConflictUpdate to clear the existing slot at
exit. Otherwise that slot could continue to hold a pin till the query
ends, which could be far too long if the input data set is large, and
there's no further conflicts. While previously also problematic, it's
now more important as there will be more such slots when partitioned.
Author: Andres Freund
Reviewed-By: Robert Haas, David Rowley
Discussion: https://postgr.es/m/20180703070645.wchpu5muyto5n647@alap3.anarazel.de
In a complete brown paper bag moment, I forgot to include equalfuncs
in my previous fix of copy/out/readfuncs. Thanks Tom for noticing.
Discussion: https://postgr.es/m/1659.1551903210@sss.pgh.pa.us
This includes a catversion bump, as IntoClause is theoretically
speaking part of storable rules. In practice I don't think that can
happen, but there's no reason to be stingy here.
Per buildfarm member calliphoridae.
This introduces the concept of table access methods, i.e. CREATE
ACCESS METHOD ... TYPE TABLE and
CREATE TABLE ... USING (storage-engine).
No table access functionality is delegated to table AMs as of this
commit, that'll be done in following commits.
Subsequent commits will incrementally abstract table access
functionality to be routed through table access methods. That change
is too large to be reviewed & committed at once, so it'll be done
incrementally.
Docs will be updated at the end, as adding them incrementally would
likely make them less coherent, and definitely is a lot more work,
without a lot of benefit.
Table access methods are specified similar to index access methods,
i.e. pg_am.amhandler returns, as INTERNAL, a pointer to a struct with
callbacks. In contrast to index AMs that struct needs to live as long
as a backend, typically that's achieved by just returning a pointer to
a constant struct.
Psql's \d+ now displays a table's access method. That can be disabled
with HIDE_TABLEAM=true, which is mainly useful so regression tests can
be run against different AMs. It's quite possible that this behaviour
still needs to be fine tuned.
For now it's not allowed to set a table AM for a partitioned table, as
we've not resolved how partitions would inherit that. Disallowing
allows us to introduce, if we decide that's the way forward, such a
behaviour without a compatibility break.
Catversion bumped, to add the heap table AM and references to it.
Author: Haribabu Kommi, Andres Freund, Alvaro Herrera, Dimitri Golgov and others
Discussion:
https://postgr.es/m/20180703070645.wchpu5muyto5n647@alap3.anarazel.dehttps://postgr.es/m/20160812231527.GA690404@alvherre.pgsqlhttps://postgr.es/m/20190107235616.6lur25ph22u5u5av@alap3.anarazel.dehttps://postgr.es/m/20190304234700.w5tmhducs5wxgzls@alap3.anarazel.de
I broke/typoed this in 4da597edf1. Astonishingly this mostly
doesn't cause breakage, except when trying to change the tuple
descriptor of a slot (because TTS_FLAG_FIXED is assumed to be set).
Author: Andres Freund
The original 200 default value was set back in f425b605f4 when the cost
delay settings were first added. Hardware has improved quite a bit since
then and we've also made improvements such as sorting buffers during
checkpoints (9cd00c457e) which should result in less random writes.
This low default value was reportedly causing problems with badly
configured servers and in the absence of a native method to remove
excessive bloat from tables without incurring an AccessExclusiveLock, this
often made cleaning up the damage caused by badly configured auto-vacuums
difficult.
It seems more likely that someone will notice that auto-vacuum is running
too quickly than too slowly, so let's go all out and multiple the default
value for the setting by 10. With the default vacuum_cost_page_dirty and
autovacuum_vacuum_cost_delay (assuming a page size of 8192 bytes), this
allows autovacuum a theoretical maximum dirty write rate of around 39MB/s
instead of just 3.9MB/s.
Author: David Rowley
Discussion: https://postgr.es/m/CAKJS1f_YbXC2qTMPyCbmsPiKvZYwpuQNQMohiRXLj1r=8_rYvw@mail.gmail.com
Scanning an index in physical order is faster than walking it in logical
order, because sequential I/O is faster than random I/O. The idea and code
structure is borrowed from B-tree vacuum code.
Patch by Andrey Borodin, with changes by me. Based on early work by
Konstantin Kuznetsov, although the patch has been rewritten multiple times
since his original version.
Discussion: https://www.postgresql.org/message-id/1B9FAC6F-FA19-4A24-8C1B-F4F574844892%40yandex-team.ru
StoreIndexTuple was a loop over index_getattr, which is O(N^2)
if the index columns are variable-width, and the performance
impact is already quite visible at ten columns. The obvious
move is to replace that with a call to index_deform_tuple ...
but that's *also* a loop over index_getattr. Improve it to
be essentially a clone of heap_deform_tuple.
(There are a few other places that loop over all index columns
with index_getattr, and perhaps should be changed likewise,
but most of them don't seem performance-critical. Anyway, the
rest would mostly only be interested in the index key columns,
which there aren't likely to be so many of. Wide index tuples
are a new thing with INCLUDE.)
Konstantin Knizhnik
Discussion: https://postgr.es/m/e06b2d27-04fc-5c0e-bb8c-ecd72aa24959@postgrespro.ru
Previously, rewriteTargetListIU() generated a list of attribute
numbers from the targetlist, which were passed to rewriteValuesRTE(),
which expected them to contain the same number of entries as there are
columns in the VALUES RTE, and to be in the same order. That was fine
when the target relation was a table, but for an updatable view it
could be broken in at least three different ways ---
rewriteTargetListIU() could insert additional targetlist entries for
view columns with defaults, the view columns could be in a different
order from the columns of the underlying base relation, and targetlist
entries could be merged together when assigning to elements of an
array or composite type. As a result, when recursing to the base
relation, the list of attribute numbers generated from the rewritten
targetlist could no longer be relied upon to match the columns of the
VALUES RTE. We got away with that prior to 41531e42d3 because it used
to always be the case that rewriteValuesRTE() did nothing for the
underlying base relation, since all DEFAULTS had already been replaced
when it was initially invoked for the view, but that was incorrect
because it failed to apply defaults from the base relation.
Fix this by examining the targetlist entries more carefully and
picking out just those that are simple Vars referencing the VALUES
RTE. That's sufficient for the purposes of rewriteValuesRTE(), which
is only responsible for dealing with DEFAULT items in the VALUES
RTE. Any DEFAULT item in the VALUES RTE that doesn't have a matching
simple-Var-assignment in the targetlist is an error which we complain
about, but in theory that ought to be impossible.
Additionally, move this code into rewriteValuesRTE() to give a clearer
separation of concerns between the 2 functions. There is no need for
rewriteTargetListIU() to know about the details of the VALUES RTE.
While at it, fix the comment for rewriteValuesRTE() which claimed that
it doesn't support array element and field assignments --- that hasn't
been true since a3c7a993d5 (9.6 and later).
Back-patch to all supported versions, with minor differences for the
pre-9.6 branches, which don't support array element and field
assignments to the same column in multi-row VALUES lists.
Reviewed by Amit Langote.
Discussion: https://postgr.es/m/15623-5d67a46788ec8b7f@postgresql.org
This changes the partition functions so as tables and indexes which are
not part of partition trees are handled the same way as what is done for
undefined objects and unsupported relkinds: pg_partition_tree() returns
no rows and pg_partition_root() returns a NULL result. Hence,
partitioned tables, partitioned indexes and relations whose flag
pg_class.relispartition is set are considered as valid objects to
process.
Previously, tables and indexes not included in a partition tree were
processed the same way as a partition or a partitioned table, which
caused the functions to return inconsistent results for inherited
tables, especially when inheriting from multiple tables.
Reported-by: Álvaro Herrera
Author: Amit Langote, Michael Paquier
Reviewed-by: Tom Lane
Discussion: https://postgr.es/m/20190228193203.GA26151@alvherre.pgsql
In particular, make it possible to prove/refute "x IS NULL" and
"x IS NOT NULL" predicates from a clause involving a ScalarArrayOpExpr
even when we are unable or unwilling to deconstruct the expression
into an AND/OR tree. This avoids a former unexpected degradation of
plan quality when the size of an ARRAY[] expression or array constant
exceeded the arbitrary MAX_SAOP_ARRAY_SIZE limit. For IS-NULL proofs,
we don't really care about the values of the individual array elements;
at most, we care whether there are any, and for some common cases we
needn't even know that.
The main user-visible effect of this is to let the optimizer recognize
applicability of partial indexes with "x IS NOT NULL" predicates to
queries with "x IN (array)" clauses in some cases where it previously
failed to recognize that. The structure of predtest.c is such that a
bunch of related proofs will now also succeed, but they're probably
much less useful in the wild.
James Coleman, reviewed by David Rowley
Discussion: https://postgr.es/m/CAAaqYe8yKSvzbyu8w-dThRs9aTFMwrFxn_BkTYeXgjqe3CbNjg@mail.gmail.com
For the upcoming pluggable table access methods it's quite
inconvenient to store tuples as HeapTuples, as that'd require
converting tuples from a their native format into HeapTuples. Instead
use slots to manage epq tuples.
To fit into that scheme, change the foreign data wrapper callback
RefetchForeignRow, to store the tuple in a slot. Insist on using the
caller provided slot, so it conveniently can be stored in the
corresponding EPQ slot. As there is no in core user of
RefetchForeignRow, that change was done blindly, but we plan to test
that soon.
To avoid duplicating that work for row locks, move row locks to just
directly use the EPQ slots - it previously temporarily stored tuples
in LockRowsState.lr_curtuples, but that doesn't seem beneficial, given
we'd possibly end up with a significant number of additional slots.
The behaviour of es_epqTupleSet[rti -1] is now checked by
es_epqTupleSlot[rti -1] != NULL, as that is distinguishable from a
slot containing an empty tuple.
Author: Andres Freund, Haribabu Kommi, Ashutosh Bapat
Discussion: https://postgr.es/m/20180703070645.wchpu5muyto5n647@alap3.anarazel.de
After commit b0eaa4c51b, we use a local map of pages to find the required
space for small relations. We do clear this map when we have found a block
with enough free space, when we extend the relation, or on transaction
abort so that it can be used next time. However, we miss to clear it when
we didn't find any pages to try from the map which leads to an assertion
failure when we later tried to use it after relation extension.
In the passing, I have improved some comments in this area.
Reported-by: Tom Lane based on buildfarm results
Author: Amit Kapila
Reviewed-by: John Naylor
Tested-by: Kuntal Ghosh
Discussion: https://postgr.es/m/32368.1551114120@sss.pgh.pa.us
The function was tweaked so as it returned one row full of NULLs when
working on an unsupported relkind or an undefined object as of cc53123,
and after discussion with Amit and Álvaro it looks more natural to make
it return no rows.
Author: Michael Paquier
Reviewed-by: Álvaro Herrera, Amit Langote
Discussion: https://postgr.es/m/20190227184808.GA17357@alvherre.pgsql
After 5408e233f0 it's not necessary to force materializing the
target slot when copying from one buffer slot to another. Previously
that was required because the HeapTupleData portion of the source slot
wasn't guaranteed to stay valid long enough, but now we can simply
copy that part into the destination slot's tupdata.
Author: Andres Freund
We have forboth() and forthree() macros that simplify iterating
through several parallel lists, but not everyplace that could
reasonably use those was doing so. Also invent forfour() and
forfive() macros to do the same for four or five parallel lists,
and use those where applicable.
The immediate motivation for doing this is to reduce the number
of ad-hoc lnext() calls, to reduce the footprint of a WIP patch.
However, it seems like good cleanup and error-proofing anyway;
the places that were combining forthree() with a manually iterated
loop seem particularly illegible and bug-prone.
There was some speculation about restructuring related parsetree
representations to reduce the need for parallel list chasing of
this sort. Perhaps that's a win, or perhaps not, but in any case
it would be considerably more invasive than this patch; and it's
not particularly related to my immediate goal of improving the
List infrastructure. So I'll leave that question for another day.
Patch by me; thanks to David Rowley for review.
Discussion: https://postgr.es/m/11587.1550975080@sss.pgh.pa.us
There was a mix of old_slot/oldslot, new_slot/newslot. Since we've
changed everything from row to slot, we might as well take this
opportunity to clean this up.
Also update some more comments for the slot change.
ri_triggers.c spends a lot of space catering to a not-yet-implemented
MATCH PARTIAL option. An actual implementation would probably not use
the existing code structure anyway, so let's just simplify this for
now.
First, have ri_FetchConstraintInfo() check that riinfo->confmatchtype
is valid. Then we don't have to repeat that everywhere.
In the various referential action functions, we don't need to pay
attention to the match type at all right now, so remove all that code.
A future MATCH PARTIAL implementation would probably have some
conditions added to the present code, but it won't need an entirely
separate switch branch in each case.
In RI_FKey_fk_upd_check_required(), reorganize the code to make it
much simpler.
Reviewed-by: Corey Huinker <corey.huinker@gmail.com>
Discussion: https://www.postgresql.org/message-id/flat/0ccdd3e1-10b0-dd05-d8a7-183507c11eb1%402ndquadrant.com
Reflecting an updated parameter value requires a server restart, which
was not mentioned in the documentation and in postgresql.conf.sample.
Reported-by: Thomas Poty
Discussion: https://postgr.es/m/15659-0cd812f13027a2d8@postgresql.org
After ff11e7f4b9 Tom's compiler warns about accessing a potentially
uninitialized rInfo. That's not actually possible, but it's
understandable the compiler would get this wrong. NULL initialize too.
Reported-By: Tom Lane
Discussion: https://postgr.es/m/11199.1551285318@sss.pgh.pa.us
By default, the fallback_application_name for a physical walreceiver
is "walreceiver". This means that multiple standbys cannot be
distinguished easily on a primary, for example in pg_stat_activity or
synchronous_standby_names.
If cluster_name is set, use that for fallback_application_name in the
walreceiver. (If it's not set, it remains "walreceiver".) If someone
set cluster_name to identify their instance, we might as well use that
by default to identify the node remotely as well. It's still possible
to specify another application_name in primary_conninfo explicitly.
Reviewed-by: Euler Taveira <euler@timbira.com.br>
Discussion: https://www.postgresql.org/message-id/flat/1257eaee-4874-e791-e83a-46720c72cac7@2ndquadrant.com
In preparation for abstracting table storage, convert trigger.c to
track tuples in slots. Which also happens to make code calling
triggers simpler.
As the calling interface for triggers themselves is not changed in
this patch, HeapTuples still are extracted from the slot at that
time. But that's handled solely inside trigger.c, not visible to
callers. It's quite likely that we'll want to revise the external
trigger interface, but that's a separate large project.
As part of this work the slots used for old/new/return tuples are
moved from EState into ResultRelInfo, as different updated tables
might need different slots. The slots are now also now created
on-demand, which is good both from an efficiency POV, but also makes
the modifying code simpler.
Author: Andres Freund, Amit Khandekar and Ashutosh Bapat
Discussion: https://postgr.es/m/20180703070645.wchpu5muyto5n647@alap3.anarazel.de
After the introduction of tuple table slots all table AMs need to
support returning the table oid of the tuple stored in a slot created
by said AM. It does not make sense to re-implement that in every AM,
therefore move handling of table OIDs into the TupleTableSlot
structure itself. It's possible that we, at a later date, might want
to get rid of HeapTupleData.t_tableOid entirely, but doing so before
the abstractions for table AMs are integrated turns out to be too
hard, so delay that for now.
Similarly, every AM needs to support the concept of a tuple
identifier (tid / item pointer) for its tuples. It's quite possible
that we'll generalize the exact form of a tid at a future point (to
allow for things like index organized tables), but for now many parts
of the code know about tids, so there's not much point in abstracting
tids away. Therefore also move into slot (rather than providing API to
set/get the tid associated with the tuple in a slot).
Once table AM includes insert/updating/deleting tuples, the
responsibility to set the correct tid after such an action will move
into that. After that change, code doing such modifications, should
not have to deal with HeapTuples directly anymore.
Author: Andres Freund, Haribabu Kommi and Ashutosh Bapat
Discussion: https://postgr.es/m/20180703070645.wchpu5muyto5n647@alap3.anarazel.de
This allows to avoid an unnecessary pin/unpin cycle when storing a
tuple in an already pinned buffer into a slot, when the pin isn't
further needed at the call site.
Only a single caller for now (to ensure coverage), but upcoming
patches will increase use of the new function.
Author: Andres Freund
Discussion: https://postgr.es/m/20180703070645.wchpu5muyto5n647@alap3.anarazel.de
Previously, this function acquired locks in the order using
find_all_inheritors(), which locks the children of each table that it
processes in ascending OID order, and which processes the inheritance
hierarchy as a whole in a breadth-first fashion. Now, it processes
the inheritance hierarchy in a depth-first fashion, and at each level
it proceeds in the order in which tables appear in the PartitionDesc.
If table inheritance rather than table partitioning is used, the old
order is preserved.
This change moves the locking of any given partition much closer to
the code that actually expands that partition. This seems essential
if we ever want to allow concurrent DDL to add or remove partitions,
because if the set of partitions can change, we must use the same data
to decide which partitions to lock as we do to decide which partitions
to expand; otherwise, we might expand a partition that we haven't
locked. It should hopefully also facilitate efforts to postpone
inheritance expansion or locking for performance reasons, because
there's really no way to postpone locking some partitions if
we're blindly locking them all using find_all_inheritors().
The only downside of this change which is known to me is that it
further deviates from the principle that we should always lock the
inheritance hierarchy in find_all_inheritors() order to avoid deadlock
risk. However, we've already crossed that bridge in commit
9eefba181f and there are futher patches
pending that make similar changes, so this isn't really giving up
anything that we haven't surrendered already -- and it seems entirely
worth it, given the performance benefits some of those changes seem
likely to bring.
Patch by me; thanks to David Rowley for discussion of these issues.
Discussion: http://postgr.es/m/CAKJS1f_eEYVEq5tM8sm1k-HOwG0AyCPwX54XG9x4w0zy_N4Q_Q@mail.gmail.com
Discussion: http://postgr.es/m/CA+TgmoZUwPf_uanjF==gTGBMJrn8uCq52XYvAEorNkLrUdoawg@mail.gmail.com
_bt_getstackbuf() is called at exactly two points following commit
efada2b8e9 (one call site is concerned with page splits, while the
other is concerned with page deletion). The parent buffer returned by
_bt_getstackbuf() is write-locked in both cases. Remove the 'access'
argument and make _bt_getstackbuf() assume that callers require a
write-lock.
Commit efada2b8e9, which made the nbtree page deletion algorithm more
robust, removed _bt_getstackbuf() calls from _bt_pagedel(). It failed
to update a comment that referenced the earlier approach. Update the
comment to explain that the _bt_getstackbuf() page deletion call site
mirrors the only other remaining _bt_getstackbuf() call site, which is
reached during page splits.
Remove some unnecessary, legacy-looking use of the PROCEDURAL keyword
before LANGUAGE. We mostly don't use this anymore, so some of these
look a bit old.
There is still some use in pg_dump, which is harder to remove because
it's baked into the archive format, so I'm not touching that.
Discussion: https://www.postgresql.org/message-id/2330919b-62d9-29ac-8de3-58c024fdcb96@2ndquadrant.com
When preparing a transaction in two-phase commit, a dummy PGPROC entry
holding the GID used for the transaction is registered, which gets
released once COMMIT PREPARED is run. Prior releasing its shared memory
state, all the locks taken in the prepared transaction are released
using a dedicated set of callbacks (pgstat and multixact having similar
callbacks), which may cause the locks to be released before the GID is
set free.
Hence, there is a small window where lock conflicts could happen, for
example:
- Transaction A releases its locks, still holding its GID in shared
memory.
- Transaction B held a lock which conflicted with locks of transaction
A.
- Transaction B continues its processing, reusing the same GID as
transaction A.
- Transaction B fails because of a conflicting GID, already in use by
transaction A.
This commit changes the shared memory state release so as post-commit
callbacks and predicate lock cleanup happen consistently with the shared
memory state cleanup for the dummy PGPROC entry. The race window is
small and 2PC had this issue from the start, so no backpatch is done.
On top if that fixes discussed involved ABI breakages, which are not
welcome in stable branches.
Reported-by: Oleksii Kliukin, Ildar Musin
Diagnosed-by: Oleksii Kliukin, Ildar Musin
Author: Michael Paquier
Reviewed-by: Masahiko Sawada, Oleksii Kliukin
Discussion: https://postgr.es/m/BF9B38A4-2BFF-46E8-BA87-A2D00A8047A6@hintbits.com
Commit 16be2fd1 introduced the flag DSA_ALLOC_NO_OOM to control whether
the DSA allocator would raise an error or return InvalidDsaPointer on
failure to allocate. One edge case was not handled correctly: if we
fail to allocate an internal "span" object for a large allocation, we
would always return InvalidDsaPointer regardless of the flag; a caller
not expecting that could then dereference a null pointer.
This is a plausible explanation for a one-off report of a segfault.
Remove a redundant pair of braces so that all three stanzas that handle
DSA_ALLOC_NO_OOM match in style, for visual consistency.
While fixing inconsistencies, if FreePageManagerGet() can't supply the
pages that our book-keeping says it should be able to supply, then we
should always report a FATAL error. Previously we treated that as a
regular allocation failure in one code path, but as a FATAL condition
in another.
Back-patch to 10, where dsa.c landed.
Author: Thomas Munro
Reported-by: Jakub Glapa
Discussion: https://postgr.es/m/CAEepm=2oPqXxyWQ-1o60tpOLrwkw=VpgNXqqF1VN2EyO9zKGQw@mail.gmail.com
The Bison documentation clearly states that a semicolon is required
after every grammar rule, and our scripts that generate ecpg's
grammar from the backend's implicitly assumed this is true. But it
turns out that only ancient versions of Bison actually enforce that.
There have been a couple of rules without trailing semicolons in
gram.y for some time, and as a consequence, ecpg's grammar was faulty
and produced wrong output for the affected statements.
To fix, add the missing semis, and add some cross-checks to ecpg's
scripts so that they'll bleat if we mess this up again.
The cases that were broken were:
* "SET variable = DEFAULT" (but not "SET variable TO DEFAULT"),
as well as allied syntaxes such as ALTER SYSTEM SET ... DEFAULT.
These produced syntactically invalid output that the server
would reject.
* Multiple type names in DROP TYPE/DOMAIN commands. Only the
first type name would be listed in the emitted command.
Per report from Daisuke Higuchi. Back-patch to all supported versions.
Discussion: https://postgr.es/m/1803D792815FC24D871C00D17AE95905DB51CE@g01jpexmbkw24
Previously, we tolerated EBADF as a way for the operating system to
indicate that it doesn't support fsync() on a directory. Tolerate
EINVAL too, for older versions of Linux CIFS.
Bug #15636. Back-patch all the way.
Reported-by: John Klann
Discussion: https://postgr.es/m/15636-d380890dafd78fc6@postgresql.org
One unintended consequence of commit 9ccdd7f6 was that Windows WSL
users started getting a panic whenever we tried to initiate data
flushing with sync_file_range(), because WSL does not implement that
system call. Previously, they got a stream of periodic warnings,
which was also undesirable but at least ignorable.
Prevent the panic by handling ENOSYS specially and skipping the panic
promotion with data_sync_elevel(). Also suppress future attempts
after the first such failure so that the pre-existing problem of
noisy warnings is improved.
Back-patch to 9.6 (older branches were not affected in this way by
9ccdd7f6).
Author: Thomas Munro and James Sewell
Tested-by: James Sewell
Reported-by: Bruce Klein
Discussion: https://postgr.es/m/CA+mCpegfOUph2U4ZADtQT16dfbkjjYNJL1bSTWErsazaFjQW9A@mail.gmail.com
In the case where inheritance_planner() finds that every table has
been excluded by constraints, it thought it could get away with
making a plan consisting of just a dummy Result node. While certainly
there's no updating or deleting to be done, this had two user-visible
problems: the plan did not report the correct set of output columns
when a RETURNING clause was present, and if there were any
statement-level triggers that should be fired, it didn't fire them.
Hence, rather than only generating the dummy Result, we need to
stick a valid ModifyTable node on top, which requires a tad more
effort here.
It's been broken this way for as long as inheritance_planner() has
known about deleting excluded subplans at all (cf commit 635d42e9c),
so back-patch to all supported branches.
Amit Langote and Tom Lane, per a report from Petr Fedorov.
Discussion: https://postgr.es/m/5da6f0f0-1364-1876-6978-907678f89a3e@phystech.edu
It seems to make more sense for this to be in selfuncs.c, since it's
largely a statistical-estimation thing, and it's related to other
functions like estimate_hash_bucket_stats that are there.
While at it, change the result type from Size to double. Perhaps at one
point it was impossible for the result to overflow an integer, but
I've got no confidence in that proposition anymore. Nothing's actually
done with the result except to compare it to a work_mem-based limit,
so as long as we don't get an overflow on the way to that comparison,
things should be fine even with very large dNumGroups.
Code movement proposed by Antonin Houska, type change by me
Discussion: https://postgr.es/m/25767.1549359615@localhost
Change pg_stat_ssl so that an unprivileged user can only see their own
rows; other rows will be all null. This makes the behavior consistent
with pg_stat_activity, where information about where the connection
came from is also restricted.
Reviewed-by: Michael Paquier <michael@paquier.xyz>
Discussion: https://www.postgresql.org/message-id/flat/63117976-d02c-c8e2-3aef-caa31a5ab8d3%402ndquadrant.com
This is similar in spirit to the existing partbounds.c file in the
same directory, except that there's a lot less code in the new file
created by this commit. Pending work in this area proposes to add a
bunch more code related to PartitionDescs, though, and this will give
us a good place to put it.
Discussion: http://postgr.es/m/CA+TgmoZUwPf_uanjF==gTGBMJrn8uCq52XYvAEorNkLrUdoawg@mail.gmail.com
Instead of locking all partitions to which we might route a tuple at
executor startup, just lock them as we use them. In some cases such a
partition might get locked at executor startup anyway because it
appears in the query's range table for some other reason, but in other
cases this is a bit savings.
This changes the order in which partitions are locked in some cases,
which might conceivably create deadlock hazards that don't exist
today, but per discussion, it seems like such cases should be rare
enough that we can neglect them in favor of improving performance.
David Rowley, reviewed and tested by Tomas Vondra, Sho Kato, John
Naylor, Tom Lane, and me.
Discussion: http://postgr.es/m/CAKJS1f-=FnMqmQP6qitkD+xEddxw22ySLP-0xFk3JAqUX2yfMw@mail.gmail.com
Check ec_relids before bothering to iterate through the EC members.
On a perhaps extreme, but still real-world, query in which
match_eclasses_to_foreign_key_col() accounts for the bulk of the
planner's runtime, this saves nearly 40% of the runtime. It's a bit
of a stopgap fix, but it's simple enough to be back-patched to 9.6
where this code came in; so let's do that.
David Rowley
Discussion: https://postgr.es/m/6970.1545327857@sss.pgh.pa.us
PG type coercions are generally strict, ie a NULL input must produce
a NULL output (or, in domain cases, possibly an error). The planner's
understanding of that was a bit incomplete though, so improve it:
* Teach contain_nonstrict_functions() that CoerceViaIO can always be
considered strict. Previously it believed that only if the underlying
I/O functions were marked strict, which is often but not always true.
* Teach clause_is_strict_for() that CoerceViaIO, ArrayCoerceExpr,
ConvertRowtypeExpr, CoerceToDomain can all be considered strict.
Previously it knew nothing about any of them.
The main user-visible impact of this is that IS NOT NULL predicates
can be proven to hold from expressions involving casts in more cases
than before, allowing partial indexes with such predicates to be used
without extra pushups. This reduces the surprise factor for users,
who may well be used to ordinary (function-call-based) casts being
known to be strict.
Per a gripe from Samuel Williams. This doesn't rise to the level of
a bug, IMO, so no back-patch.
Discussion: https://postgr.es/m/27571.1550617881@sss.pgh.pa.us
The recursion in contain_nonstrict_functions_walker() was done wrong,
causing the strictness check to be bypassed for a parse node that
is the immediate input of an ArrayCoerceExpr node. This could allow,
for example, incorrect decisions about whether a strict SQL function
can be inlined.
I didn't add a regression test, because (a) the bug is so narrow
and (b) I couldn't think of a test case that wasn't dependent on a
large number of other behaviors, to the point where it would likely
soon rot to the point of not testing what it was intended to.
I broke this in commit c12d570fa, so back-patch to v11.
Discussion: https://postgr.es/m/27571.1550617881@sss.pgh.pa.us
pg_identify_object_as_address crashes when passed certain tuples from
inconsistent system catalogs. Make it more defensive.
Author: Álvaro Herrera
Reviewed-by: Michaël Paquier
Discussion: https://postgr.es/m/20190218202743.GA12392@alvherre.pgsql
INSERT ... VALUES for a single VALUES row is implemented differently
from a multi-row VALUES list, which causes inconsistent behaviour in
the way that DEFAULT items are handled. In particular, when inserting
into an auto-updatable view on top of a table with a column default, a
DEFAULT item in a single VALUES row gets correctly replaced with the
table column's default, but for a multi-row VALUES list it is replaced
with NULL.
Fix this by allowing rewriteValuesRTE() to leave DEFAULT items in the
VALUES list untouched if the target relation is an auto-updatable view
and has no column default, deferring DEFAULT-expansion until the query
against the base relation is rewritten. For all other types of target
relation, including tables and trigger- and rule-updatable views, we
must continue to replace DEFAULT items with NULL in the absence of a
column default.
This is somewhat complicated by the fact that if an auto-updatable
view has DO ALSO rules attached, the VALUES lists for the product
queries need to be handled differently from the original query, since
the product queries need to act like rule-updatable views whereas the
original query has auto-updatable view semantics.
Back-patch to all supported versions.
Reported by Roger Curley (bug #15623). Patch by Amit Langote and me.
Discussion: https://postgr.es/m/15623-5d67a46788ec8b7f@postgresql.org
When building an initial slot snapshot, snapshots are marked with
historic MVCC snapshots as type with the marker field being set in
SnapBuildBuildSnapshot() but not overriden in SnapBuildInitialSnapshot().
Existing callers of SnapBuildBuildSnapshot() do not care about the type
of snapshot used, but extensions calling it actually may, as reported.
While on it, mark correctly the snapshot type when importing one. This
is cosmetic as the field is enforced to 0.
Author: Antonin Houska
Reviewed-by: Álvaro Herrera, Michael Paquier
Discussion: https://postgr.es/m/23215.1527665193@localhost
Backpatch-through: 9.4
For the convenience of extensions, we previously only saved PathTargets
for grouped, window, and final relations in root->upper_targets[] in
grouping_planner(). To improve the convenience, save PathTargets for
distinct and ordered relations as well.
Author: Antonin Houska, with an additional change by me
Discussion: https://postgr.es/m/10994.1549559088@localhost
When reading a new page internally and depending on the way the WAL
reader facility gets used by plugins, the current implementation of the
WAL reader may finish by reading a block multiple times while it is not
actually necessary as the requested data length may be equal to what has
been already read. This can happen for any size, but is more likely to
happen at the end of a page. This can cause performance penalties in
plugins which rely on the block reads to be purely sequential, zlib not
liking backward reads for example. The new behavior also shaves some
cycles when doing recovery.
Author: Arthur Zakirov
Reviewed-by: Andrey Lepikhov, Michael Paquier, Grigory Smolkin
Discussion: https://postgr.es/m/2ddf4a32-517e-d6f4-d992-4a63b6035bfd@postgrespro.ru
Teach dsm_unpin_segment() to skip segments that are in the process
of being destroyed by another backend, when searching for a handle.
Such a segment cannot possibly be the one we are looking for, even
if its handle matches. Another slot might hold a recently created
segment that has the same handle value by coincidence, and we need
to keep searching for that one.
The bug caused rare "cannot unpin a segment that is not pinned"
errors on 10 and 11. Similar to commit 6c0fb941 for dsm_attach().
Back-patch to 10, where dsm_unpin_segment() landed.
Author: Thomas Munro
Reported-by: Justin Pryzby
Tested-by: Justin Pryzby (along with other recent DSA/DSM fixes)
Discussion: https://postgr.es/m/20190216023854.GF30291@telsasoft.com
We should logically have allowed this case when we allowed zero-column
tables, but it was overlooked.
Although this might be thought a feature addition, it's really a bug
fix, because it was possible to create a zero-column view via
the convert-table-to-view code path, and then you'd have a situation
where dump/reload would fail. Hence, back-patch to all supported
branches.
Arrange the added test cases to provide coverage of the related
pg_dump code paths (since these views will be dumped and reloaded
during the pg_upgrade regression test). I also made them test
the case where pg_dump has to postpone the view rule into post-data,
which disturbingly had no regression coverage before.
Report and patch by Ashutosh Sharma (test case by me)
Discussion: https://postgr.es/m/CAE9k0PkmHdeSaeZt2ujnb_cKucmK3sDDceDzw7+d5UZoNJPYOg@mail.gmail.com
Historically we've always materialized the full output of a CTE query,
treating WITH as an optimization fence (so that, for example, restrictions
from the outer query cannot be pushed into it). This is appropriate when
the CTE query is INSERT/UPDATE/DELETE, or is recursive; but when the CTE
query is non-recursive and side-effect-free, there's no hazard of changing
the query results by pushing restrictions down.
Another argument for materialization is that it can avoid duplicate
computation of an expensive WITH query --- but that only applies if
the WITH query is called more than once in the outer query. Even then
it could still be a net loss, if each call has restrictions that
would allow just a small part of the WITH query to be computed.
Hence, let's change the behavior for WITH queries that are non-recursive
and side-effect-free. By default, we will inline them into the outer
query (removing the optimization fence) if they are called just once.
If they are called more than once, we will keep the old behavior by
default, but the user can override this and force inlining by specifying
NOT MATERIALIZED. Lastly, the user can force the old behavior by
specifying MATERIALIZED; this would mainly be useful when the query had
deliberately been employing WITH as an optimization fence to prevent a
poor choice of plan.
Andreas Karlsson, Andrew Gierth, David Fetter
Discussion: https://postgr.es/m/87sh48ffhb.fsf@news-spur.riddles.org.uk
Test for the compiler builtins __builtin_clz, __builtin_ctz, and
__builtin_popcount, and make use of these in preference to
handwritten C code if they're available. Create src/port
infrastructure for "leftmost one", "rightmost one", and "popcount"
so as to centralize these decisions.
On x86_64, __builtin_popcount generally won't make use of the POPCNT
opcode because that's not universally supported yet. Provide code
that checks CPUID and then calls POPCNT via asm() if available.
This requires indirecting through a function pointer, which is
an annoying amount of overhead for a one-instruction operation,
but it's probably not worth working harder than this for our
current use-cases.
I'm not sure we've found all the existing places that could profit
from this new infrastructure; but we at least touched all the
ones that used copied-and-pasted versions of the bitmapset.c code,
and got rid of multiple copies of the associated constant arrays.
While at it, replace c-compiler.m4's one-per-builtin-function
macros with a single one that can handle all the cases we need
to worry about so far. Also, because I'm paranoid, make those
checks into AC_LINK checks rather than just AC_COMPILE; the
former coding failed to verify that libgcc has support for the
builtin, in cases where it's not inline code.
David Rowley, Thomas Munro, Alvaro Herrera, Tom Lane
Discussion: https://postgr.es/m/CAKJS1f9WTAGG1tPeJnD18hiQW5gAk59fQ6WK-vfdAKEHyRg2RA@mail.gmail.com
This reverts commits fc6c72747a, 109de05cbb, d0b4663c23 and
711bab1e4d.
Somebody will have to try harder before submitting this patch again.
I've spent entirely too much time on it already, and the #ifdef maze yet
to be written in order for it to build at all got on my nerves. The
amount of work needed to get a platform-specific performance improvement
that's barely above the noise level is not worth it.
Get rid of deconstruct_indexquals() in favor of just iterating over the
IndexClause list directly. The extra services that that function used to
provide, such as hiding clause commutation and associating the right index
column with each clause, are no longer useful given the new data structure.
I'd originally thought that it'd provide a useful amount of abstraction
by freeing callers from paying attention to the exact clause type of each
indexqual, but that hope proves to have been vain, because few callers can
ignore the semantic differences between different clause types. Indeed,
removing it results in a net code savings, and probably some cycles shaved
by not having to build an extra list-of-structs data structure.
Also, export a few formerly-static support functions, with the goal
of allowing extension AMs to write functionality equivalent to
genericcostestimate() without pointless code duplication.
Discussion: https://postgr.es/m/24586.1550106354@sss.pgh.pa.us
The grammar IF NOT EXISTS for CTAS is supported since 9.5 and documented
as such, however the case of using EXECUTE as query has never been
covered as EXECUTE CTAS statements and normal CTAS statements are parsed
separately.
Author: Andreas Karlsson
Discussion: https://postgr.es/m/2ddcc188-e37c-a0be-32bf-a56b07c3559e@proxel.se
Backpatch-through: 9.5
DSM handle values can be reused as soon as the underlying shared memory
object has been destroyed. That means that for a brief moment we
might have two DSM slots with the same handle. While trying to attach,
if we encounter a slot with refcnt == 1, meaning that it is currently
being destroyed, we should continue our search in case the same handle
exists in another slot.
The race manifested as a rare "dsa_area could not attach to segment"
error, and was more likely in 10 and 11 due to the lack of distinct
seed for random() in parallel workers. It was made very unlikely in
in master by commit 197e4af9, and older releases don't usually create
new DSM segments in background workers so it was also unlikely there.
This fixes the root cause of bug report #15585, in which the error
could also sometimes result in a self-deadlock in the error path.
It's not yet clear if further changes are needed to avoid that failure
mode.
Back-patch to 9.4, where dsm.c arrived.
Author: Thomas Munro
Reported-by: Justin Pryzby, Sergei Kornilov
Discussion: https://postgr.es/m/20190207014719.GJ29720@telsasoft.com
Discussion: https://postgr.es/m/15585-324ff6a93a18da46@postgresql.org
In commit 1a8d5afb0, I thought it'd be a good idea to define
IndexClause.indexquals as NIL in the most common case where the given
clause (IndexClause.rinfo) is usable exactly as-is. It'd be more
consistent to define the indexquals in that case as being a one-element
list containing IndexClause.rinfo, but I thought saving the palloc
overhead for making such a list would be worthwhile.
In hindsight, that was a great example of "premature optimization is the
root of all evil": it's complicated everyplace that needs to deal with
the indexquals, requiring duplicative code to handle both the simple
case and the not-simple case. I'd initially found that tolerable but
it's getting less so as I mop up some areas that I'd not touched in
1a8d5afb0. In any case, two more pallocs during a planner run are
surely at the noise level (a conclusion confirmed by a bit of
microbenchmarking). So let's change this decision before it becomes
set in stone, and insist that IndexClause.indexquals always be a valid
list of the actual index quals for the clause.
Discussion: https://postgr.es/m/24586.1550106354@sss.pgh.pa.us
While at it, refactor patternsel() a bit so that it can be used from
the LIKE/regex planner support functions as well. This makes the
planner able to deal equally well with either operator or function
syntax for these operations. I'm not excited about that as a feature
in itself, but it provides a nice model for extensions to follow if
they want such behavior for their operations.
This change localizes the use of pattern_fixed_prefix() and
make_greater_string() so that they no longer need be exported.
(We might get pushback from extensions about that, perhaps,
in which case I'd be inclined to re-export them in a new header
file like_support.h.)
This reduces the bulk of selfuncs.c a fair amount, removing ~1370
lines or about one-sixth of that file; it's still too big, but this
is progress.
Discussion: https://postgr.es/m/24537.1550093915@sss.pgh.pa.us
These opcodes have been around in the AMD world since 2007, and 2008 in
the case of intel. They're supported in GCC and Clang via some __builtin
macros. The opcodes may be unavailable during runtime, in which case we
fall back on a C-based implementation of the code. In order to get the
POPCNT instruction we must pass the -mpopcnt option to the compiler. We
do this only for the pg_bitutils.c file.
David Rowley (with fragments taken from a patch by Thomas Munro)
Discussion: https://postgr.es/m/CAKJS1f9WTAGG1tPeJnD18hiQW5gAk59fQ6WK-vfdAKEHyRg2RA@mail.gmail.com
Previously, floating-point output was done by rounding to a specific
decimal precision; by default, to 6 or 15 decimal digits (losing
information) or as requested using extra_float_digits. Drivers that
wanted exact float values, and applications like pg_dump that must
preserve values exactly, set extra_float_digits=3 (or sometimes 2 for
historical reasons, though this isn't enough for float4).
Unfortunately, decimal rounded output is slow enough to become a
noticable bottleneck when dealing with large result sets or COPY of
large tables when many floating-point values are involved.
Floating-point output can be done much faster when the output is not
rounded to a specific decimal length, but rather is chosen as the
shortest decimal representation that is closer to the original float
value than to any other value representable in the same precision. The
recently published Ryu algorithm by Ulf Adams is both relatively
simple and remarkably fast.
Accordingly, change float4out/float8out to output shortest decimal
representations if extra_float_digits is greater than 0, and make that
the new default. Applications that need rounded output can set
extra_float_digits back to 0 or below, and take the resulting
performance hit.
We make one concession to portability for systems with buggy
floating-point input: we do not output decimal values that fall
exactly halfway between adjacent representable binary values (which
would rely on the reader doing round-to-nearest-even correctly). This
is known to be a problem at least for VS2013 on Windows.
Our version of the Ryu code originates from
https://github.com/ulfjack/ryu/ at commit c9c3fb1979, but with the
following (significant) modifications:
- Output format is changed to use fixed-point notation for small
exponents, as printf would, and also to use lowercase 'e', a
minimum of 2 exponent digits, and a mandatory sign on the exponent,
to keep the formatting as close as possible to previous output.
- The output of exact midpoint values is disabled as noted above.
- The integer fast-path code is changed somewhat (since we have
fixed-point output and the upstream did not).
- Our project style has been largely applied to the code with the
exception of C99 declaration-after-statement, which has been
retained as an exception to our present policy.
- Most of upstream's debugging and conditionals are removed, and we
use our own configure tests to determine things like uint128
availability.
Changing the float output format obviously affects a number of
regression tests. This patch uses an explicit setting of
extra_float_digits=0 for test output that is not expected to be
exactly reproducible (e.g. due to numerical instability or differing
algorithms for transcendental functions).
Conversions from floats to numeric are unchanged by this patch. These
may appear in index expressions and it is not yet clear whether any
change should be made, so that can be left for another day.
This patch assumes that the only supported floating point format is
now IEEE format, and the documentation is updated to reflect that.
Code by me, adapting the work of Ulf Adams and other contributors.
References:
https://dl.acm.org/citation.cfm?id=3192369
Reviewed-by: Tom Lane, Andres Freund, Donald Dong
Discussion: https://postgr.es/m/87r2el1bx6.fsf@news-spur.riddles.org.uk
Using strtod() creates a double-rounding problem; the input decimal
value is first rounded to the nearest double; rounding that to the
nearest float may then give an incorrect result.
An example is that 7.038531e-26 when input via strtod and then rounded
to float4 gives 0xAE43FEp-107 instead of the correct 0xAE43FDp-107.
Values output by earlier PG versions with extra_float_digits=3 should
all be read in with the same values as previously. However, values
supplied by other software using shortest representations could be
mis-read.
On platforms that lack a strtof() entirely, we fall back to the old
incorrect rounding behavior. (As strtof() is required by C99, such
platforms are considered of primarily historical interest.) On VS2013,
some workarounds are used to get correct error handling.
The regression tests now test for the correct input values, so
platforms that lack strtof() will need resultmap entries. An entry for
HP-UX 10 is included (more may be needed).
Reviewed-By: Tom Lane
Discussion: https://postgr.es/m/871s5emitx.fsf@news-spur.riddles.org.uk
Discussion: https://postgr.es/m/87d0owlqpv.fsf@news-spur.riddles.org.uk
In a corner case, a btree page was allocated during a clean-up operation
that could cause the tracking of the largest contiguous span of free
space to get out of whack. That was supposed to be prevented by the use
of the "soft" flag to avoid allocating internal pages during incidental
clean-up work, but the flag was ignored in the case where the FPM was
promoted from singleton format to btree format. Repair.
Remove an obsolete comment in passing.
Back-patch to 10, where freepage.c arrived (as support for dsa.c).
Author: Robert Haas
Diagnosed-by: Thomas Munro and Robert Haas
Reported-by: Justin Pryzby, Rick Otten, Sand Stone, Arne Roland and others
Discussion: https://postgr.es/m/CAMAYy4%2Bw3NTBM5JLWFi8twhWK4%3Dk_5L4nV5%2BbYDSPu8r4b97Zg%40mail.gmail.com
We're only going to consider key columns when creating indexquals,
so there is no point in having the outer loops in indxpath.c iterate
further than nkeycolumns.
Doing so in match_pathkeys_to_index() is actually wrong, and would have
caused crashes by now, except that we have no index AMs supporting both
amcanorderbyop and amcaninclude.
It's also wrong in relation_has_unique_index_for(). The effect there is
to fail to prove uniqueness even when the index does prove it, if there
are extra columns.
Also future-proof examine_variable() for the day when extra columns can
be expressions, and fix what's either a thinko or just an oversight in
btcostestimate(): we should consider the number of key columns, not the
total, when deciding whether to derate correlation.
None of these things seemed important enough to risk changing in a
just-before-wrap patch, but since we're past the release wrap window,
time to fix 'em.
Discussion: https://postgr.es/m/25526.1549847928@sss.pgh.pa.us
Ever since its birth, ReorderBufferBuildTupleCidHash() has contained an
assertion that a catalog tuple cannot change Cmax after acquiring one. But
that's wrong: if a subtransaction executes DDL that affects that catalog
tuple, and later aborts and another DDL affects the same tuple, it will
change Cmax. Relax the assertion to merely verify that the Cmax remains
valid and monotonically increasing, instead.
Add a test that tickles the relevant code.
Diagnosed by, and initial patch submitted by: Arseny Sher
Co-authored-by: Arseny Sher
Discussion: https://postgr.es/m/874l9p8hyw.fsf@ars-thinkpad
It's pretty unhelpful to report the wrong file name in a complaint
about syscall failure, but SnapBuildSerialize managed to do that twice
in a span of 50 lines. Also fix half a dozen missing or poorly-chosen
errcode assignments; that's mostly cosmetic, but still wrong.
Noted while studying recent failures on buildfarm member nightjar.
I'm not sure whether those reports are actually giving the wrong
filename, because there are two places here with identically
spelled error messages. The other one is specifically coded not
to report ENOENT, but if it's this one, how could we be getting
ENOENT from open() with O_CREAT? Need to sit back and await results.
However, these ereports are clearly broken from birth, so back-patch.
For a long time, indxpath.c has had the ability to extract derived (lossy)
index conditions from certain operators such as LIKE. For just as long,
it's been obvious that we really ought to make that capability available
to extensions. This commit finally accomplishes that, by adding another
API for planner support functions that lets them create derived index
conditions for their functions. As proof of concept, the hardwired
"special index operator" code formerly present in indxpath.c is pushed
out to planner support functions attached to LIKE and other relevant
operators.
A weak spot in this design is that an extension needs to know OIDs for
the operators, datatypes, and opfamilies involved in the transformation
it wants to make. The core-code prototypes use hard-wired OID references
but extensions don't have that option for their own operators etc. It's
usually possible to look up the required info, but that may be slow and
inconvenient. However, improving that situation is a separate task.
I want to do some additional refactorization around selfuncs.c, but
that also seems like a separate task.
Discussion: https://postgr.es/m/15193.1548028093@sss.pgh.pa.us
Since its introduction, max_wal_senders is counted as part of
max_connections when it comes to define how many connection slots can be
used for replication connections with a WAL sender context. This can
lead to confusion for some users, as it could be possible to block a
base backup or replication from happening because other backend sessions
are already taken for other purposes by an application, and
superuser-only connection slots are not a correct solution to handle
that case.
This commit makes max_wal_senders independent of max_connections for its
handling of PGPROC entries in ProcGlobal, meaning that connection slots
for WAL senders are handled using their own free queue, like autovacuum
workers and bgworkers.
One compatibility issue that this change creates is that a standby now
requires to have a value of max_wal_senders at least equal to its
primary. So, if a standby created enforces the value of
max_wal_senders to be lower than that, then this could break failovers.
Normally this should not be an issue though, as any settings of a
standby are inherited from its primary as postgresql.conf gets normally
copied as part of a base backup, so parameters would be consistent.
Author: Alexander Kukushkin
Reviewed-by: Kyotaro Horiguchi, Petr Jelínek, Masahiko Sawada, Oleksii
Kliukin
Discussion: https://postgr.es/m/CAFh8B=nBzHQeYAu0b8fjK-AF1X4+_p6GRtwG+cCgs6Vci2uRuQ@mail.gmail.com
The original setup for dependencies of partitioned objects had
serious problems:
1. It did not verify that a drop cascading to a partition-child object
also cascaded to at least one of the object's partition parents. Now,
normally a child object would share all its dependencies with one or
another parent (e.g. a child index's opclass dependencies would be shared
with the parent index), so that this oversight is usually harmless.
But if some dependency failed to fit this pattern, the child could be
dropped while all its parents remain, creating a logically broken
situation. (It's easy to construct artificial cases that break it,
such as attaching an unrelated extension dependency to the child object
and then dropping the extension. I'm not sure if any less-artificial
cases exist.)
2. Management of partition dependencies during ATTACH/DETACH PARTITION
was complicated and buggy; for example, after detaching a partition
table it was possible to create cases where a formerly-child index
should be dropped and was not, because the correct set of dependencies
had not been reconstructed.
Less seriously, because multiple partition relationships were
represented identically in pg_depend, there was an order-of-traversal
dependency on which partition parent was cited in error messages.
We also had some pre-existing order-of-traversal hazards for error
messages related to internal and extension dependencies. This is
cosmetic to users but causes testing problems.
To fix#1, add a check at the end of the partition tree traversal
to ensure that at least one partition parent got deleted. To fix#2,
establish a new policy that partition dependencies are in addition to,
not instead of, a child object's usual dependencies; in this way
ATTACH/DETACH PARTITION need not cope with adding or removing the
usual dependencies.
To fix the cosmetic problem, distinguish between primary and secondary
partition dependency entries in pg_depend, by giving them different
deptypes. (They behave identically except for having different
priorities for being cited in error messages.) This means that the
former 'I' dependency type is replaced with new 'P' and 'S' types.
This also fixes a longstanding bug that after handling an internal
dependency by recursing to the owning object, findDependentObjects
did not verify that the current target was now scheduled for deletion,
and did not apply the current recursion level's objflags to it.
Perhaps that should be back-patched; but in the back branches it
would only matter if some concurrent transaction had removed the
internal-linkage pg_depend entry before the recursive call found it,
or the recursive call somehow failed to find it, both of which seem
unlikely.
Catversion bump because the contents of pg_depend change for
partitioning relationships.
Patch HEAD only. It's annoying that we're not fixing #2 in v11,
but there seems no practical way to do so given that the problem
is exactly a poor choice of what entries to put in pg_depend.
We can't really fix that while staying compatible with what's
in pg_depend in existing v11 installations.
Discussion: https://postgr.es/m/CAH2-Wzkypv1R+teZrr71U23J578NnTBt2X8+Y=Odr4pOdW1rXg@mail.gmail.com
Andres Freund
Peter Geoghegan
Tom Lane
Michael Paquier
Tomas Vondra
Peter Eisentraut
Robert Haas
Thomas Munro
Alvaro Herrera
Alexander Korotkov
Heikki Linnakangas
Andrew Gierth
Amit Kapila
Etsuro Fujita
Magnus Hagander
Noah Misch
Andrew Dunstan
Dean Rasheed