| Commit message (Collapse) | Author | Age |
|---|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
The Self-Join Elimination (SJE) feature removes an inner join of a plain
table to itself in the query tree if it is proven that the join can be
replaced with a scan without impacting the query result. Self-join and
inner relation get replaced with the outer in query, equivalence classes,
and planner info structures. Also, the inner restrictlist moves to the
outer one with the removal of duplicated clauses. Thus, this optimization
reduces the length of the range table list (this especially makes sense for
partitioned relations), reduces the number of restriction clauses and,
in turn, selectivity estimations, and potentially improves total planner
prediction for the query.
This feature is dedicated to avoiding redundancy, which can appear after
pull-up transformations or the creation of an EquivalenceClass-derived clause
like the below.
SELECT * FROM t1 WHERE x IN (SELECT t3.x FROM t1 t3);
SELECT * FROM t1 WHERE EXISTS (SELECT t3.x FROM t1 t3 WHERE t3.x = t1.x);
SELECT * FROM t1,t2, t1 t3 WHERE t1.x = t2.x AND t2.x = t3.x;
In the future, we could also reduce redundancy caused by subquery pull-up
after unnecessary outer join removal in cases like the one below.
SELECT * FROM t1 WHERE x IN
(SELECT t3.x FROM t1 t3 LEFT JOIN t2 ON t2.x = t1.x);
Also, it can drastically help to join partitioned tables, removing entries
even before their expansion.
The SJE proof is based on innerrel_is_unique() machinery.
We can remove a self-join when for each outer row:
1. At most, one inner row matches the join clause;
2. Each matched inner row must be (physically) the same as the outer one;
3. Inner and outer rows have the same row mark.
In this patch, we use the next approach to identify a self-join:
1. Collect all merge-joinable join quals which look like a.x = b.x;
2. Add to the list above the baseretrictinfo of the inner table;
3. Check innerrel_is_unique() for the qual list. If it returns false, skip
this pair of joining tables;
4. Check uniqueness, proved by the baserestrictinfo clauses. To prove the
possibility of self-join elimination, the inner and outer clauses must
match exactly.
The relation replacement procedure is not trivial and is partly combined
with the one used to remove useless left joins. Tests covering this feature
were added to join.sql. Some of the existing regression tests changed due
to self-join removal logic.
Discussion: https://postgr.es/m/flat/64486b0b-0404-e39e-322d-0801154901f3%40postgrespro.ru
Author: Andrey Lepikhov <a.lepikhov@postgrespro.ru>
Author: Alexander Kuzmenkov <a.kuzmenkov@postgrespro.ru>
Co-authored-by: Alexander Korotkov <aekorotkov@gmail.com>
Co-authored-by: Alena Rybakina <lena.ribackina@yandex.ru>
Reviewed-by: Tom Lane <tgl@sss.pgh.pa.us>
Reviewed-by: Robert Haas <robertmhaas@gmail.com>
Reviewed-by: Andres Freund <andres@anarazel.de>
Reviewed-by: Simon Riggs <simon@2ndquadrant.com>
Reviewed-by: Jonathan S. Katz <jkatz@postgresql.org>
Reviewed-by: David Rowley <david.rowley@2ndquadrant.com>
Reviewed-by: Thomas Munro <thomas.munro@enterprisedb.com>
Reviewed-by: Konstantin Knizhnik <k.knizhnik@postgrespro.ru>
Reviewed-by: Heikki Linnakangas <hlinnaka@iki.fi>
Reviewed-by: Hywel Carver <hywel@skillerwhale.com>
Reviewed-by: Laurenz Albe <laurenz.albe@cybertec.at>
Reviewed-by: Ronan Dunklau <ronan.dunklau@aiven.io>
Reviewed-by: vignesh C <vignesh21@gmail.com>
Reviewed-by: Zhihong Yu <zyu@yugabyte.com>
Reviewed-by: Greg Stark <stark@mit.edu>
Reviewed-by: Jaime Casanova <jcasanov@systemguards.com.ec>
Reviewed-by: Michał Kłeczek <michal@kleczek.org>
Reviewed-by: Alena Rybakina <lena.ribackina@yandex.ru>
Reviewed-by: Alexander Korotkov <aekorotkov@gmail.com>
|
| |
|
|
| |
Backpatch-through: 13
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Remove the code for inserting flag columns in the inputs of a SetOp.
That was the only reason why there would be resjunk columns in a
set-operations plan tree, so we can get rid of some code that
supported that, too.
Get rid of choose_hashed_setop() in favor of building Paths for
the hashed and sorted alternatives, and letting them fight it out
within add_path().
Remove set_operation_ordered_results_useful(), which was giving wrong
answers due to examining the wrong ancestor node: we need to examine
the immediate SetOperationStmt parent not the topmost node. Instead
make each caller of recurse_set_operations() pass down the relevant
parent node. (This thinko seems to have led only to wasted planning
cycles and possibly-inferior plans, not wrong query answers. Perhaps
we should back-patch it, but I'm not doing so right now.)
Teach generate_nonunion_paths() to consider pre-sorted inputs for
sorted SetOps, rather than always generating a Sort node.
Patch by me; thanks to Richard Guo and David Rowley for review.
Discussion: https://postgr.es/m/1850138.1731549611@sss.pgh.pa.us
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
The original design for set operations involved appending the two
input relations into one and adding a flag column that allows
distinguishing which side each row came from. Then the SetOp node
pries them apart again based on the flag. This is bizarre. The
only apparent reason to do it is that when sorting, we'd only need
one Sort node not two. But since sorting is at least O(N log N),
sorting all the data is actually worse than sorting each side
separately --- plus, we have no chance of taking advantage of
presorted input. On top of that, adding the flag column frequently
requires an additional projection step that adds cycles, and then
the Append node isn't free either. Let's get rid of all of that
and make the SetOp node have two separate children, using the
existing outerPlan/innerPlan infrastructure.
This initial patch re-implements nodeSetop.c and does a bare minimum
of work on the planner side to generate correctly-shaped plans.
In particular, I've tried not to change the cost estimates here,
so that the visible changes in the regression test results will only
involve removal of useless projection steps and not any changes in
whether to use sorted vs hashed mode.
For SORTED mode, we combine successive identical tuples from each
input into groups, and then merge-join the groups. The tuple
comparisons now use SortSupport instead of simple equality, but
the group-formation part should involve roughly the same number of
tuple comparisons as before. The cross-comparisons between left and
right groups probably add to that, but I'm not sure to quantify how
many more comparisons we might need.
For HASHED mode, nodeSetop's logic is almost the same as before,
just refactored into two separate loops instead of one loop that
has an assumption that it will see all the left-hand inputs first.
In both modes, I added early-exit logic to not bother reading the
right-hand relation if the left-hand input is empty, since neither
INTERSECT nor EXCEPT modes can produce any output if the left input
is empty. This could have been done before in the hashed mode, but
not in sorted mode. Sorted mode can also stop as soon as it exhausts
the left input; any remaining right-hand tuples cannot have matches.
Also, this patch adds some infrastructure for detecting whether
child plan nodes all output the same type of tuple table slot.
If they do, the hash table logic can use slightly more efficient
code based on assuming that that's the input slot type it will see.
We'll make use of that infrastructure in other plan node types later.
Patch by me; thanks to Richard Guo and David Rowley for review.
Discussion: https://postgr.es/m/1850138.1731549611@sss.pgh.pa.us
|
| |
|
|
|
| |
Reported-by: Richard Guo
Discussion: https://postgr.es/m/CAMbWs492vMy3XNjDZRtqtHfFTK6HVeDwhrEQH7eXGgF_h5Jnzw@mail.gmail.com
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
In the dim past we figured it was okay to ignore collations
when combining UNION set-operation nodes into a single N-way
UNION operation. I believe that was fine at the time, but
it stopped being fine when we added nondeterministic collations:
the semantics of distinct-ness are affected by those. v17 made
it even less fine by allowing per-child sorting operations to
be merged via MergeAppend, although I think we accidentally
avoided any live bug from that.
Add a check that collations match before deciding that two
UNION nodes are equivalent. I also failed to resist the
temptation to comment plan_union_children() a little better.
Back-patch to all supported branches (v13 now), since they
all have nondeterministic collations.
Discussion: https://postgr.es/m/3605568.1731970579@sss.pgh.pa.us
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Previously, when a path type was disabled by e.g. enable_seqscan=false,
we either avoided generating that path type in the first place, or
more commonly, we added a large constant, called disable_cost, to the
estimated startup cost of that path. This latter approach can distort
planning. For instance, an extremely expensive non-disabled path
could seem to be worse than a disabled path, especially if the full
cost of that path node need not be paid (e.g. due to a Limit).
Or, as in the regression test whose expected output changes with this
commit, the addition of disable_cost can make two paths that would
normally be distinguishible in cost seem to have fuzzily the same cost.
To fix that, we now count the number of disabled path nodes and
consider that a high-order component of both the startup cost and the
total cost. Hence, the path list is now sorted by disabled_nodes and
then by total_cost, instead of just by the latter, and likewise for
the partial path list. It is important that this number is a count
and not simply a Boolean; else, as soon as we're unable to respect
disabled path types in all portions of the path, we stop trying to
avoid them where we can.
Because the path list is now sorted by the number of disabled nodes,
the join prechecks must compute the count of disabled nodes during
the initial cost phase instead of postponing it to final cost time.
Counts of disabled nodes do not cross subquery levels; at present,
there is no reason for them to do so, since the we do not postpone
path selection across subquery boundaries (see make_subplan).
Reviewed by Andres Freund, Heikki Linnakangas, and David Rowley.
Discussion: http://postgr.es/m/CA+TgmoZ_+MS+o6NeGK2xyBv-xM+w1AfFVuHE4f_aq6ekHv7YSQ@mail.gmail.com
|
| |
|
|
|
|
|
|
|
|
|
|
| |
The previous logic would fail to set groupList when
grouping_is_sortable() returned false, which could cause queries
to return wrong answers when some columns were not sortable.
David Rowley, reviewed by Heikki Linnakangas and Álvaro Herrera.
Reported by Hubert "depesz" Lubaczewski.
Discussion: http://postgr.es/m/Zktzf926vslR35Fv@depesz.com
Discussion: http://postgr.es/m/CAApHDvra=c8_zZT0J-TftByWN2Y+OJfnjNJFg4Dfdi2s+QzmqA@mail.gmail.com
|
| |
|
|
|
|
|
|
|
|
|
| |
This reverts commit 7204f35919b7e021e8d1bc9f2d76fd6bfcdd2070,
thus restoring 66c0185a3 (Allow planner to use Merge Append to
efficiently implement UNION) as well as the follow-on commits
d5d2205c8, 3b1a7eb28, 7487044d6.
Per further discussion on pgsql-release, we wish to ship beta1 with
this feature, and patch the bug that was found just before wrap,
rather than shipping beta1 with the feature reverted.
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
This reverts 66c0185a3 (Allow planner to use Merge Append to
efficiently implement UNION) as well as the follow-on commits
d5d2205c8, 3b1a7eb28, 7487044d6. In addition to those, 07746a8ef
had to be removed then re-applied in a different place, because
66c0185a3 moved the relevant code.
The reason for this last-minute thrashing is that depesz found a
case in which the patched code creates a completely wrong plan
that silently gives incorrect query results. It's unclear what
the cause is or how many cases are affected, but with beta1 wrap
staring us in the face, there's no time for closer investigation.
After we figure that out, we can decide whether to un-revert this
for beta2 or hold it for v18.
Discussion: https://postgr.es/m/Zktzf926vslR35Fv@depesz.com
(also some private discussion among pgsql-release)
|
| |
|
|
|
|
|
|
|
|
| |
The code change this made might well be fine to keep, but the
comment justifying it by reference to self-join removal isn't.
Let's just go back to the status quo ante, pending a more thorough
review/redesign of SJE.
(I found this by grepping to see if any references to self-join
removal remained in the tree.)
|
| |
|
|
|
|
|
|
|
|
|
|
| |
This fixes various typos, duplicated words, and tiny bits of whitespace
mainly in code comments but also in docs.
Author: Daniel Gustafsson <daniel@yesql.se>
Author: Heikki Linnakangas <hlinnaka@iki.fi>
Author: Alexander Lakhin <exclusion@gmail.com>
Author: David Rowley <dgrowleyml@gmail.com>
Author: Nazir Bilal Yavuz <byavuz81@gmail.com>
Discussion: https://postgr.es/m/3F577953-A29E-4722-98AD-2DA9EFF2CBB8@yesql.se
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
66c0185a3 adjusted the UNION planner to request that union child queries
produce Paths correctly ordered to implement the UNION by way of
MergeAppend followed by Unique. The code there made a bad assumption
that if the root->parent_root->parse had setOperations set that the
query must be the child subquery of a set operation. That's not true
when it comes to planning a non-inlined CTE which is parented by a set
operation. This causes issues as the CTE's targetlist has no
requirement to match up to the SetOperationStmt's groupClauses
Fix this by adding a new parameter to both subquery_planner() and
grouping_planner() to explicitly pass the SetOperationStmt only when
planning set operation child subqueries.
Thank you to Tom Lane for helping to rationalize the decision on the
best function signature for subquery_planner().
Reported-by: Alexander Lakhin
Discussion: https://postgr.es/m/242fc7c6-a8aa-2daf-ac4c-0a231e2619c1@gmail.com
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Until now, UNION queries have often been suboptimal as the planner has
only ever considered using an Append node and making the results unique
by either using a Hash Aggregate, or by Sorting the entire Append result
and running it through the Unique operator. Both of these methods
always require reading all rows from the union subqueries.
Here we adjust the union planner so that it can request that each subquery
produce results in target list order so that these can be Merge Appended
together and made unique with a Unique node. This can improve performance
significantly as the union child can make use of the likes of btree
indexes and/or Merge Joins to provide the top-level UNION with presorted
input. This is especially good if the top-level UNION contains a LIMIT
node that limits the output rows to a small subset of the unioned rows as
cheap startup plans can be used.
Author: David Rowley
Reviewed-by: Richard Guo, Andy Fan
Discussion: https://postgr.es/m/CAApHDvpb_63XQodmxKUF8vb9M7CxyUyT4sWvEgqeQU-GB7QFoQ@mail.gmail.com
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
as determined by include-what-you-use (IWYU)
While IWYU also suggests to *add* a bunch of #include's (which is its
main purpose), this patch does not do that. In some cases, a more
specific #include replaces another less specific one.
Some manual adjustments of the automatic result:
- IWYU currently doesn't know about includes that provide global
variable declarations (like -Wmissing-variable-declarations), so
those includes are being kept manually.
- All includes for port(ability) headers are being kept for now, to
play it safe.
- No changes of catalog/pg_foo.h to catalog/pg_foo_d.h, to keep the
patch from exploding in size.
Note that this patch touches just *.c files, so nothing declared in
header files changes in hidden ways.
As a small example, in src/backend/access/transam/rmgr.c, some IWYU
pragma annotations are added to handle a special case there.
Discussion: https://www.postgresql.org/message-id/flat/af837490-6b2f-46df-ba05-37ea6a6653fc%40eisentraut.org
|
| |
|
|
|
|
|
|
| |
Reported-by: Michael Paquier
Discussion: https://postgr.es/m/ZZKTDPxBBMt3C0J9@paquier.xyz
Backpatch-through: 12
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
recurse_set_operations() uses the parse tree for the group number estimation,
because of the "varno 0" hack. At the same time 2489d76c49 made root->parse
and corresponding parent_root->simple_rte_array[]->subquery distinct copies
of the parse tree, while d3d55ce571 introduced self-join removal replacing
relid of removed relation only in one of the copies.
The present commit fixes this bug by making recurse_set_operations() call
estimate_num_groups() with the copy of the parse tree processed by self-join
removal.
In future, we may think about maintaining just one copy of the parse tree
and/or keeping removed relids as aliases.
Reported-by: Zuming Jiang
Bug: #18170
Discussion: https://postgr.es/m/flat/18170-f1d17bf9a0d58b24%40postgresql.org
Author: Richard Guo, Alexander Korotkov
Reviewed-by: Andrei Lepikhov
|
| |
|
|
| |
Backpatch-through: 11
|
| |
|
|
|
|
|
|
|
|
|
|
| |
In a similar effort to f01592f91, here we mostly rename shadowed local
variables to remove the warnings produced when compiling with
-Wshadow=compatible-local.
This fixes 63 warnings and leaves just 5.
Author: Justin Pryzby, David Rowley
Reviewed-by: Justin Pryzby
Discussion https://postgr.es/m/20220817145434.GC26426%40telsasoft.com
|
| |
|
|
|
|
|
|
|
|
|
|
| |
In a similar effort to f01592f91, here we're targetting fixing the
warnings where we've deemed the shadowing variable to serve a close enough
purpose to the shadowed variable just to reuse the shadowed version and
not declare the shadowing variable at all.
By my count, this takes the warning count from 106 down to 71.
Author: Justin Pryzby
Discussion: https://postgr.es/m/20220825020839.GT2342@telsasoft.com
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
| |
Commit 4f658dc8 provided the traditional BSD fls() function in
src/port/fls.c so it could be used in several places. Later we added a
bunch of similar facilities in pg_bitutils.h, based on compiler
builtins that map to hardware instructions. It's a bit confusing to
have both 1-based and 0-based variants of this operation in use in
different parts of the tree, and neither is blessed by a standard.
Let's drop fls.c and the configure probe, and reuse the newer code.
Reviewed-by: David Rowley <dgrowleyml@gmail.com>
Reviewed-by: Tom Lane <tgl@sss.pgh.pa.us>
Discussion: https://postgr.es/m/CA%2BhUKG%2B7dSX1XF8yFGmYk-%3D48dbjH2kmzZj16XvhbrWP-9BzRg%40mail.gmail.com
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
setrefs.c contains logic to discard no-op SubqueryScan nodes, that is,
ones that have no qual to check and copy the input targetlist unchanged.
(Formally it's not very nice to be applying such optimizations so late
in the planner, but there are practical reasons for it; mostly that we
can't unify relids between the subquery and the parent query until we
flatten the rangetable during setrefs.c.) This behavior falsifies our
previous cost estimates, since we would've charged cpu_tuple_cost per
row just to pass data through the node. Most of the time that's little
enough to not matter, but there are cases where this effect visibly
changes the plan compared to what you would've gotten with no
sub-select.
To improve the situation, make the callers of cost_subqueryscan tell
it whether they think the targetlist is trivial. cost_subqueryscan
already has the qual list, so it can check the other half of the
condition easily. It could make its own determination of tlist
triviality too, but doing so would be repetitive (for callers that
may call it several times) or unnecessarily expensive (for callers
that can determine this more cheaply than a general test would do).
This isn't a 100% solution, because createplan.c also does things
that can falsify any earlier estimate of whether the tlist is
trivial. However, it fixes nearly all cases in practice, if results
for the regression tests are anything to go by.
setrefs.c also contains logic to discard no-op Append and MergeAppend
nodes. We did have knowledge of that behavior at costing time, but
somebody failed to update it when a check on parallel-awareness was
added to the setrefs.c logic. Fix that while we're here.
These changes result in two minor changes in query plans shown in
our regression tests. Neither is relevant to the purposes of its
test case AFAICT.
Patch by me; thanks to Richard Guo for review.
Discussion: https://postgr.es/m/2581077.1651703520@sss.pgh.pa.us
|
| |
|
|
| |
Backpatch-through: 10
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
The point of introducing the hash_mem_multiplier GUC was to let users
reproduce the old behavior of hash aggregation, i.e. that it could use
more than work_mem at need. However, the implementation failed to get
the job done on Win64, where work_mem is clamped to 2GB to protect
various places that calculate memory sizes using "long int". As
written, the same clamp was applied to hash_mem. This resulted in
severe performance regressions for queries requiring a bit more than
2GB for hash aggregation, as they now spill to disk and there's no
way to stop that.
Getting rid of the work_mem restriction seems like a good idea, but
it's a big job and could not conceivably be back-patched. However,
there's only a fairly small number of places that are concerned with
the hash_mem value, and it turns out to be possible to remove the
restriction there without too much code churn or any ABI breaks.
So, let's do that for now to fix the regression, and leave the
larger task for another day.
This patch does introduce a bit more infrastructure that should help
with the larger task, namely pg_bitutils.h support for working with
size_t values.
Per gripe from Laurent Hasson. Back-patch to v13 where the
behavior change came in.
Discussion: https://postgr.es/m/997817.1627074924@sss.pgh.pa.us
Discussion: https://postgr.es/m/MN2PR15MB25601E80A9B6D1BA6F592B1985E39@MN2PR15MB2560.namprd15.prod.outlook.com
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Here we add a new output parameter to estimate_num_groups() to allow it to
inform the caller of additional, possibly useful information about the
estimation.
The new output parameter is a struct that currently contains just a single
field with a set of flags. This was done rather than having the flags as
an output parameter to allow future fields to be added without having to
change the signature of the function at a later date when we want to pass
back further information that might not be suitable to store in the flags
field.
It seems reasonable that one day in the future that the planner would want
to know more about the estimation. For example, how many individual sets
of statistics was the estimation generated from? The planner may want to
take that into account if we ever want to consider risks as well as costs
when generating plans.
For now, there's only 1 flag we set in the flags field. This is to
indicate if the estimation fell back on using the hard-coded constants in
any part of the estimation. Callers may like to change their behavior if
this is set, and this gives them the ability to do so. Callers may pass
the flag pointer as NULL if they have no interest in obtaining any
additional information about the estimate.
We're not adding any actual usages of these flags here. Some follow-up
commits will make use of this feature. Additionally, we're also not
making any changes to add support for clauselist_selectivity() and
clauselist_selectivity_ext(). However, if this is required in the future
then the same struct being added here should be fine to use as a new
output argument for those functions too.
Author: David Rowley
Discussion: https://postgr.es/m/CAApHDvqQqpk=1W-G_ds7A9CsXX3BggWj_7okinzkLVhDubQzjA@mail.gmail.com
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
It turns out that the calculation of [Merge]AppendPath.partitioned_rels
in allpaths.c is faulty and sometimes omits relevant non-leaf partitions,
allowing an assertion added by commit a929e17e5a8 to trigger. Rather
than fix that, it seems better to get rid of those fields altogether.
We don't really need the info until create_plan time, and calculating
it once for the selected plan should be cheaper than calculating it
for each append path we consider.
The preceding two commits did away with all use of the partitioned_rels
values; this commit just mechanically removes the fields and the code
that calculated them.
Discussion: https://postgr.es/m/87sg8tqhsl.fsf@aurora.ydns.eu
Discussion: https://postgr.es/m/CAJKUy5gCXDSmFs2c=R+VGgn7FiYcLCsEFEuDNNLGfoha=pBE_g@mail.gmail.com
|
| |
|
|
| |
Backpatch-through: 9.5
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Commit a477bfc1d fixed eval_const_expressions() to ensure that it
didn't generate unnecessary RelabelType nodes, but I failed to notice
that some other places in the planner had the same issue. Really
noplace in the planner should be using plain makeRelabelType(), for
fear of generating expressions that should be equal() to semantically
equivalent trees, but aren't.
An example is that because canonicalize_ec_expression() failed
to be careful about this, we could end up with an equivalence class
containing both a plain Const, and a Const-with-RelabelType
representing exactly the same value. So far as I can tell this led to
no visible misbehavior, but we did waste a bunch of cycles generating
and evaluating "Const = Const-with-RelabelType" to prove such entries
are redundant.
Hence, move the support function added by a477bfc1d to where it can
be more generally useful, and use it in the places where planner code
previously used makeRelabelType.
Back-patch to v12, like the previous patch. While I have no concrete
evidence of any real misbehavior here, it's certainly possible that
I overlooked a case where equivalent expressions that aren't equal()
could cause a user-visible problem. In any case carrying extra
RelabelType nodes through planning to execution isn't very desirable.
Discussion: https://postgr.es/m/1311836.1597781384@sss.pgh.pa.us
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Add a GUC that acts as a multiplier on work_mem. It gets applied when
sizing executor node hash tables that were previously size constrained
using work_mem alone.
The new GUC can be used to preferentially give hash-based nodes more
memory than the generic work_mem limit. It is intended to enable admin
tuning of the executor's memory usage. Overall system throughput and
system responsiveness can be improved by giving hash-based executor
nodes more memory (especially over sort-based alternatives, which are
often much less sensitive to being memory constrained).
The default value for hash_mem_multiplier is 1.0, which is also the
minimum valid value. This means that hash-based nodes continue to apply
work_mem in the traditional way by default.
hash_mem_multiplier is generally useful. However, it is being added now
due to concerns about hash aggregate performance stability for users
that upgrade to Postgres 13 (which added disk-based hash aggregation in
commit 1f39bce0). While the old hash aggregate behavior risked
out-of-memory errors, it is nevertheless likely that many users actually
benefited. Hash agg's previous indifference to work_mem during query
execution was not just faster; it also accidentally made aggregation
resilient to grouping estimate problems (at least in cases where this
didn't create destabilizing memory pressure).
hash_mem_multiplier can provide a certain kind of continuity with the
behavior of Postgres 12 hash aggregates in cases where the planner
incorrectly estimates that all groups (plus related allocations) will
fit in work_mem/hash_mem. This seems necessary because hash-based
aggregation is usually much slower when only a small fraction of all
groups can fit. Even when it isn't possible to totally avoid hash
aggregates that spill, giving hash aggregation more memory will reliably
improve performance (the same cannot be said for external sort
operations, which appear to be almost unaffected by memory availability
provided it's at least possible to get a single merge pass).
The PostgreSQL 13 release notes should advise users that increasing
hash_mem_multiplier can help with performance regressions associated
with hash aggregation. That can be taken care of by a later commit.
Author: Peter Geoghegan
Reviewed-By: Álvaro Herrera, Jeff Davis
Discussion: https://postgr.es/m/20200625203629.7m6yvut7eqblgmfo@alap3.anarazel.de
Discussion: https://postgr.es/m/CAH2-WzmD%2Bi1pG6rc1%2BCjc4V6EaFJ_qSuKCCHVnH%3DoruqD-zqow%40mail.gmail.com
Backpatch: 13-, where disk-based hash aggregation was introduced.
|
| |
|
|
|
|
| |
Oversight in commit 1f39bce0, which added disk-based hash aggregation.
Backpatch: 13-, where disk-based hash aggregation was introduced.
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
While performing hash aggregation, track memory usage when adding new
groups to a hash table. If the memory usage exceeds work_mem, enter
"spill mode".
In spill mode, new groups are not created in the hash table(s), but
existing groups continue to be advanced if input tuples match. Tuples
that would cause a new group to be created are instead spilled to a
logical tape to be processed later.
The tuples are spilled in a partitioned fashion. When all tuples from
the outer plan are processed (either by advancing the group or
spilling the tuple), finalize and emit the groups from the hash
table. Then, create new batches of work from the spilled partitions,
and select one of the saved batches and process it (possibly spilling
recursively).
Author: Jeff Davis
Reviewed-by: Tomas Vondra, Adam Lee, Justin Pryzby, Taylor Vesely, Melanie Plageman
Discussion: https://postgr.es/m/507ac540ec7c20136364b5272acbcd4574aa76ef.camel@j-davis.com
|
| |
|
|
| |
Backpatch-through: update all files in master, backpatch legal files through 9.4
|
| |
|
|
|
|
|
|
|
|
|
|
|
| |
Merge setup_append_rel_array into setup_simple_rel_arrays. There's no
particularly good reason to keep them separate, and it's inconsistent
with the lack of separation in expand_planner_arrays. The only apparent
benefit was that the fast path for trivial queries in query_planner()
doesn't need to set up the append_rel_array; but all we're saving there
is an if-test and NULL assignment, which surely ought to be negligible.
Also improve some obsolete comments.
Discussion: https://postgr.es/m/17220.1565301350@sss.pgh.pa.us
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Previously in order to determine which ECs a relation had members in, we
had to loop over all ECs stored in PlannerInfo's eq_classes and check if
ec_relids mentioned the relation. For the most part, this was fine, as
generally, unless queries were fairly complex, the overhead of performing
the lookup would have not been that significant. However, when queries
contained large numbers of joins and ECs, the overhead to find the set of
classes matching a given set of relations could become a significant
portion of the overall planning effort.
Here we allow a much more efficient method to access the ECs which match a
given relation or set of relations. A new Bitmapset field in RelOptInfo
now exists to store the indexes into PlannerInfo's eq_classes list which
each relation is mentioned in. This allows very fast lookups to find all
ECs belonging to a single relation. When we need to lookup ECs belonging
to a given pair of relations, we can simply bitwise-AND the Bitmapsets from
each relation and use the result to perform the lookup.
We also take the opportunity to write a new implementation of
generate_join_implied_equalities which makes use of the new indexes.
generate_join_implied_equalities_for_ecs must remain as is as it can be
given a custom list of ECs, which we can't easily determine the indexes of.
This was originally intended to fix the performance penalty of looking up
foreign keys matching a join condition which was introduced by 100340e2d.
However, we're speeding up much more than just that here.
Author: David Rowley, Tom Lane
Reviewed-by: Tom Lane, Tomas Vondra
Discussion: https://postgr.es/m/6970.1545327857@sss.pgh.pa.us
|
| |
|
|
|
|
|
|
|
|
|
| |
This is numbered take 7, and addresses a set of issues around:
- Fixes for typos and incorrect reference names.
- Removal of unneeded comments.
- Removal of unreferenced functions and structures.
- Fixes regarding variable name consistency.
Author: Alexander Lakhin
Discussion: https://postgr.es/m/10bfd4ac-3e7c-40ab-2b2e-355ed15495e8@gmail.com
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Originally, Postgres Lists were a more or less exact reimplementation of
Lisp lists, which consist of chains of separately-allocated cons cells,
each having a value and a next-cell link. We'd hacked that once before
(commit d0b4399d8) to add a separate List header, but the data was still
in cons cells. That makes some operations -- notably list_nth() -- O(N),
and it's bulky because of the next-cell pointers and per-cell palloc
overhead, and it's very cache-unfriendly if the cons cells end up
scattered around rather than being adjacent.
In this rewrite, we still have List headers, but the data is in a
resizable array of values, with no next-cell links. Now we need at
most two palloc's per List, and often only one, since we can allocate
some values in the same palloc call as the List header. (Of course,
extending an existing List may require repalloc's to enlarge the array.
But this involves just O(log N) allocations not O(N).)
Of course this is not without downsides. The key difficulty is that
addition or deletion of a list entry may now cause other entries to
move, which it did not before.
For example, that breaks foreach() and sister macros, which historically
used a pointer to the current cons-cell as loop state. We can repair
those macros transparently by making their actual loop state be an
integer list index; the exposed "ListCell *" pointer is no longer state
carried across loop iterations, but is just a derived value. (In
practice, modern compilers can optimize things back to having just one
loop state value, at least for simple cases with inline loop bodies.)
In principle, this is a semantics change for cases where the loop body
inserts or deletes list entries ahead of the current loop index; but
I found no such cases in the Postgres code.
The change is not at all transparent for code that doesn't use foreach()
but chases lists "by hand" using lnext(). The largest share of such
code in the backend is in loops that were maintaining "prev" and "next"
variables in addition to the current-cell pointer, in order to delete
list cells efficiently using list_delete_cell(). However, we no longer
need a previous-cell pointer to delete a list cell efficiently. Keeping
a next-cell pointer doesn't work, as explained above, but we can improve
matters by changing such code to use a regular foreach() loop and then
using the new macro foreach_delete_current() to delete the current cell.
(This macro knows how to update the associated foreach loop's state so
that no cells will be missed in the traversal.)
There remains a nontrivial risk of code assuming that a ListCell *
pointer will remain good over an operation that could now move the list
contents. To help catch such errors, list.c can be compiled with a new
define symbol DEBUG_LIST_MEMORY_USAGE that forcibly moves list contents
whenever that could possibly happen. This makes list operations
significantly more expensive so it's not normally turned on (though it
is on by default if USE_VALGRIND is on).
There are two notable API differences from the previous code:
* lnext() now requires the List's header pointer in addition to the
current cell's address.
* list_delete_cell() no longer requires a previous-cell argument.
These changes are somewhat unfortunate, but on the other hand code using
either function needs inspection to see if it is assuming anything
it shouldn't, so it's not all bad.
Programmers should be aware of these significant performance changes:
* list_nth() and related functions are now O(1); so there's no
major access-speed difference between a list and an array.
* Inserting or deleting a list element now takes time proportional to
the distance to the end of the list, due to moving the array elements.
(However, it typically *doesn't* require palloc or pfree, so except in
long lists it's probably still faster than before.) Notably, lcons()
used to be about the same cost as lappend(), but that's no longer true
if the list is long. Code that uses lcons() and list_delete_first()
to maintain a stack might usefully be rewritten to push and pop at the
end of the list rather than the beginning.
* There are now list_insert_nth...() and list_delete_nth...() functions
that add or remove a list cell identified by index. These have the
data-movement penalty explained above, but there's no search penalty.
* list_concat() and variants now copy the second list's data into
storage belonging to the first list, so there is no longer any
sharing of cells between the input lists. The second argument is
now declared "const List *" to reflect that it isn't changed.
This patch just does the minimum needed to get the new implementation
in place and fix bugs exposed by the regression tests. As suggested
by the foregoing, there's a fair amount of followup work remaining to
do.
Also, the ENABLE_LIST_COMPAT macros are finally removed in this
commit. Code using those should have been gone a dozen years ago.
Patch by me; thanks to David Rowley, Jesper Pedersen, and others
for review.
Discussion: https://postgr.es/m/11587.1550975080@sss.pgh.pa.us
|
| |
|
|
|
|
|
|
|
| |
Switch to 2.1 version of pg_bsd_indent. This formats
multiline function declarations "correctly", that is with
additional lines of parameter declarations indented to match
where the first line's left parenthesis is.
Discussion: https://postgr.es/m/CAEepm=0P3FeTXRcU5B2W3jv3PgRVZ-kGUXLGfd42FFhUROO3ug@mail.gmail.com
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
If we need ordered output from a scan of a partitioned table, but
the ordering matches the partition ordering, then we don't need to
use a MergeAppend to combine the pre-ordered per-partition scan
results: a plain Append will produce the same results. This
both saves useless comparison work inside the MergeAppend proper,
and allows us to start returning tuples after istarting up just
the first child node not all of them.
However, all is not peaches and cream, because if some of the
child nodes have high startup costs then there will be big
discontinuities in the tuples-returned-versus-elapsed-time curve.
The planner's cost model cannot handle that (yet, anyway).
If we model the Append's startup cost as being just the first
child's startup cost, we may drastically underestimate the cost
of fetching slightly more tuples than are available from the first
child. Since we've had bad experiences with over-optimistic choices
of "fast start" plans for ORDER BY LIMIT queries, that seems scary.
As a klugy workaround, set the startup cost estimate for an ordered
Append to be the sum of its children's startup costs (as MergeAppend
would). This doesn't really describe reality, but it's less likely
to cause a bad plan choice than an underestimated startup cost would.
In practice, the cases where we really care about this optimization
will have child plans that are IndexScans with zero startup cost,
so that the overly conservative estimate is still just zero.
David Rowley, reviewed by Julien Rouhaud and Antonin Houska
Discussion: https://postgr.es/m/CAKJS1f-hAqhPLRk_RaSFTgYxd=Tz5hA7kQ2h4-DhJufQk8TGuw@mail.gmail.com
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
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
|
| |
|
|
|
|
| |
Noticed this while working on another patch.
Author: Andres Freund
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
This commit moves expand_inherited_tables and underlings from
optimizer/prep/prepunionc.c to optimizer/utils/inherit.c.
Also, all of the AppendRelInfo-based expression manipulation routines
are moved to optimizer/utils/appendinfo.c.
No functional code changes. One exception is the introduction of
make_append_rel_info, but that's still just moving around code.
Also, stop including <limits.h> in prepunion.c, which no longer needs
it since 3fc6e2d7f5b6. I (Álvaro) noticed this because Amit was copying
that to inherit.c, which likewise doesn't need it.
Author: Amit Langote
Discussion: https://postgr.es/m/3be67028-a00a-502c-199a-da00eec8fb6e@lab.ntt.co.jp
|
| |
|
|
| |
Backpatch-through: certain files through 9.4
|
| |
|
|
|
|
|
|
| |
Reported-by: David Rowley
Author: David Rowley
Reviewed-by: Amit Kapila
Backpatch-through: 10
Discussion: https://postgr.es/m/CAKJS1f8EneeYyzzvdjahVZ6gbAHFkHbSFB5m_C0Y6TUJs9Dgdg@mail.gmail.com
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
This speeds up write operations (INSERT, UPDATE, DELETE, COPY, as well
as the future MERGE) on partitioned tables.
This changes the setup for tuple routing so that it does far less work
during the initial setup and pushes more work out to when partitions
receive tuples. PartitionDispatchData structs for sub-partitioned
tables are only created when a tuple gets routed through it. The
possibly large arrays in the PartitionTupleRouting struct have largely
been removed. The partitions[] array remains but now never contains any
NULL gaps. Previously the NULLs had to be skipped during
ExecCleanupTupleRouting(), which could add a large overhead to the
cleanup when the number of partitions was large. The partitions[] array
is allocated small to start with and only enlarged when we route tuples
to enough partitions that it runs out of space. This allows us to keep
simple single-row partition INSERTs running quickly. Redesign
The arrays in PartitionTupleRouting which stored the tuple translation maps
have now been removed. These have been moved out into a
PartitionRoutingInfo struct which is an additional field in ResultRelInfo.
The find_all_inheritors() call still remains by far the slowest part of
ExecSetupPartitionTupleRouting(). This commit just removes the other slow
parts.
In passing also rename the tuple translation maps from being ParentToChild
and ChildToParent to being RootToPartition and PartitionToRoot. The old
names mislead you into thinking that a partition of some sub-partitioned
table would translate to the rowtype of the sub-partitioned table rather
than the root partitioned table.
Authors: David Rowley and Amit Langote, heavily revised by Álvaro Herrera
Testing help from Jesper Pedersen and Kato Sho.
Discussion: https://postgr.es/m/CAKJS1f_1RJyFquuCKRFHTdcXqoPX-PYqAd7nz=GVBwvGh4a6xA@mail.gmail.com
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Instead of recomputing the required lock levels in all these places,
just use what commit fdba460a2 made the parser store in the RTE fields.
This already simplifies the code measurably in these places, and
follow-on changes will remove a bunch of no-longer-needed infrastructure.
In a few cases, this change causes us to acquire a higher lock level
than we did before. This is OK primarily because said higher lock level
should've been acquired already at query parse time; thus, we're saving
a useless extra trip through the shared lock manager to acquire a lesser
lock alongside the original lock. The only known exception to this is
that re-execution of a previously planned SELECT FOR UPDATE/SHARE query,
for a table that uses ROW_MARK_REFERENCE or ROW_MARK_COPY methods, might
have gotten only AccessShareLock before. Now it will get RowShareLock
like the first execution did, which seems fine.
While there's more to do, push it in this state anyway, to let the
buildfarm help verify that nothing bad happened.
Amit Langote, reviewed by David Rowley and Jesper Pedersen,
and whacked around a bit more by me
Discussion: https://postgr.es/m/468c85d9-540e-66a2-1dde-fec2b741e688@lab.ntt.co.jp
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Previously convert_tuples_by_name_map naively performed a search of each
outdesc column starting at the first column in indesc and searched each
indesc column until a match was found. When partitioned tables had many
columns this could result in slow generation of the tuple conversion maps.
For INSERT and UPDATE statements that touched few rows, this could mean a
very large overhead indeed.
We can do a bit better with this loop. It's quite likely that the columns
in partitioned tables and their partitions are in the same order, so it
makes sense to start searching for each column outer column at the inner
column position 1 after where the previous match was found (per idea from
Alexander Kuzmenkov). This makes the best case search O(N) instead of
O(N^2). The worst case is still O(N^2), but it seems unlikely that would
happen.
Likewise, in the planner, make_inh_translation_list's search for the
matching column could often end up falling back on an O(N^2) type search.
This commit also improves that by first checking the column that follows
the previous match, instead of the column with the same attnum. If we
fail to match here we fallback on the syscache's hashtable lookup.
Author: David Rowley
Reviewed-by: Alexander Kuzmenkov
Discussion: https://www.postgresql.org/message-id/CAKJS1f9-wijVgMdRp6_qDMEQDJJ%2BA_n%3DxzZuTmLx5Fz6cwf%2B8A%40mail.gmail.com
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Since recent commit 1c7c317c, temporary relations cannot be mixed with
permanent relations within the same partition tree, and the same counts
for temporary relations created by other sessions, which the planner
simply discarded. Instead be paranoid and issue an error, as those
should be blocked at definition time, at least for now.
At the same time, a test case is added to stress what has been moved
when expand_partitioned_rtentry gets called recursively but bumps on a
partitioned relation with no partitions which should be handled the same
way as the non-inheritance case. This code may be reworked in a close
future, and covering this code path will limit surprises.
Reported-by: David Rowley
Author: David Rowley
Reviewed-by: Amit Langote, Robert Haas, Michael Paquier
Discussion: https://postgr.es/m/CAKJS1f_HyV1txn_4XSdH5EOhBMYaCwsXyAj6bHXk9gOu4JKsbw@mail.gmail.com
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
find_appinfos_by_relids had quite a large overhead when the number of
items in the append_rel_list was high, as it had to trawl through the
append_rel_list looking for AppendRelInfos belonging to the given
childrelids. Since there can only be a single AppendRelInfo for each
child rel, it seems much better to store an array in PlannerInfo which
indexes these by child relid, making the function O(1) rather than O(N).
This function was only called once inside the planner, so just replace
that call with a lookup to the new array. find_childrel_appendrelinfo
is now unused and thus removed.
This fixes a planner performance regression new to v11 reported by
Thomas Reiss.
Author: David Rowley
Reported-by: Thomas Reiss
Reviewed-by: Ashutosh Bapat
Reviewed-by: Álvaro Herrera
Discussion: https://postgr.es/m/94dd7a4b-5e50-0712-911d-2278e055c622@dalibo.com
|
| |
|
|
| |
Discussion: https://postgr.es/m/15719.1523984266@sss.pgh.pa.us
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Commit 0927d2f46ddd4cf7d6bf2cc84b3be923e0aedc52 didn't check that
consider_parallel was set for the target relation or account for
the possibility that required_outer might be non-empty.
To prevent future bugs of this ilk, add some assertions to
add_partial_path and do a bit of future-proofing of the code
recently added to recurse_set_operations.
Report by Andreas Seltenreich. Patch by Jeevan Chalke. Review
by Amit Kapila and by me.
Discussion: http://postgr.es/m/CAM2+6=U+9otsyF2fYB8x_2TBeHTR90itarqW=qAEjN-kHaC7kw@mail.gmail.com
|
