Up to now, PG has assumed that any given timezone abbreviation (such as
"EDT") represents a constant GMT offset in the usage of any particular
region; we had a way to configure what that offset was, but not for it
to be changeable over time. But, as with most things horological, this
view of the world is too simplistic: there are numerous regions that have
at one time or another switched to a different GMT offset but kept using
the same timezone abbreviation. Almost the entire Russian Federation did
that a few years ago, and later this month they're going to do it again.
And there are similar examples all over the world.
To cope with this, invent the notion of a "dynamic timezone abbreviation",
which is one that is referenced to a particular underlying timezone
(as defined in the IANA timezone database) and means whatever it currently
means in that zone. For zones that use or have used daylight-savings time,
the standard and DST abbreviations continue to have the property that you
can specify standard or DST time and get that time offset whether or not
DST was theoretically in effect at the time. However, the abbreviations
mean what they meant at the time in question (or most recently before that
time) rather than being absolutely fixed.
The standard abbreviation-list files have been changed to use this behavior
for abbreviations that have actually varied in meaning since 1970. The
old simple-numeric definitions are kept for abbreviations that have not
changed, since they are a bit faster to resolve.
While this is clearly a new feature, it seems necessary to back-patch it
into all active branches, because otherwise use of Russian zone
abbreviations is going to become even more problematic than it already was.
This change supersedes the changes in commit 513d06ded et al to modify the
fixed meanings of the Russian abbreviations; since we've not shipped that
yet, this will avoid an undesirably incompatible (not to mention incorrect)
change in behavior for timestamps between 2011 and 2014.
This patch makes some cosmetic changes in ecpglib to keep its usage of
datetime lookup tables as similar as possible to the backend code, but
doesn't do anything about the increasingly obsolete set of timezone
abbreviation definitions that are hard-wired into ecpglib. Whatever we
do about that will likely not be appropriate material for back-patching.
Also, a potential free() of a garbage pointer after an out-of-memory
failure in ecpglib has been fixed.
This patch also fixes pre-existing bugs in DetermineTimeZoneOffset() that
caused it to produce unexpected results near a timezone transition, if
both the "before" and "after" states are marked as standard time. We'd
only ever thought about or tested transitions between standard and DST
time, but that's not what's happening when a zone simply redefines their
base GMT offset.
In passing, update the SGML documentation to refer to the Olson/zoneinfo/
zic timezone database as the "IANA" database, since it's now being
maintained under the auspices of IANA.
We've gotten enough push-back on that change to make it clear that it
wasn't an especially good idea to do it like that. Revert plain EXPLAIN
to its previous behavior, but keep the extra output in EXPLAIN ANALYZE.
Per discussion.
Internally, I set this up as a separate flag ExplainState.summary that
controls printing of planning time and execution time. For now it's
just copied from the ANALYZE option, but we could consider exposing it
to users.
If we expect batching at the very beginning, we size nbuckets for
"full work_mem" (see how many tuples we can get into work_mem,
while not breaking NTUP_PER_BUCKET threshold).
If we expect to be fine without batching, we start with the 'right'
nbuckets and track the optimal nbuckets as we go (without actually
resizing the hash table). Once we hit work_mem (considering the
optimal nbuckets value), we keep the value.
At the end of the first batch, we check whether (nbuckets !=
nbuckets_optimal) and resize the hash table if needed. Also, we
keep this value for all batches (it's OK because it assumes full
work_mem, and it makes the batchno evaluation trivial). So the
resize happens only once.
There could be cases where it would improve performance to allow
the NTUP_PER_BUCKET threshold to be exceeded to keep everything in
one batch rather than spilling to a second batch, but attempts to
generate such a case have so far been unsuccessful; that issue may
be addressed with a follow-on patch after further investigation.
Tomas Vondra with minor format and comment cleanup by me
Reviewed by Robert Haas, Heikki Linnakangas, and Kevin Grittner
The new header contains many prototypes for functions in ruleutils.c
that are not exposed to the SQL level.
Reviewed by Andres Freund and Michael Paquier.
shm_mq_sendv sends a message to the queue assembled from multiple
locations. This is expected to be used by forthcoming patches to
allow frontend/backend protocol messages to be sent via shm_mq, but
might be useful for other purposes as well.
shm_mq_set_handle associates a BackgroundWorkerHandle with an
already-existing shm_mq_handle. This solves a timing problem when
creating a shm_mq to communicate with a newly-launched background
worker: if you attach to the queue first, and the background worker
fails to start, you might block forever trying to do I/O on the queue;
but if you start the background worker first, but then die before
attaching to the queue, the background worrker might block forever
trying to do I/O on the queue. This lets you attach before starting
the worker (so that the worker is protected) and then associate the
BackgroundWorkerHandle later (so that you are also protected).
Patch by me, reviewed by Stephen Frost.
This clause changes the behavior of SELECT locking clauses in the
presence of locked rows: instead of causing a process to block waiting
for the locks held by other processes (or raise an error, with NOWAIT),
SKIP LOCKED makes the new reader skip over such rows. While this is not
appropriate behavior for general purposes, there are some cases in which
it is useful, such as queue-like tables.
Catalog version bumped because this patch changes the representation of
stored rules.
Reviewed by Craig Ringer (based on a previous attempt at an
implementation by Simon Riggs, who also provided input on the syntax
used in the current patch), David Rowley, and Álvaro Herrera.
Author: Thomas Munro
Testing by Amit Kapila, Andres Freund, and myself, with and without
other patches that also aim to improve scalability, seems to indicate
that this change is a significant win over the current value and over
smaller values such as 64. It's not clear how high we can push this
value before it starts to have negative side-effects elsewhere, but
going this far looks OK.
As of commit a87c72915 (which later got backpatched as far as 9.1),
we're explicitly supporting the notion that append relations can be
nested; this can occur when UNION ALL constructs are nested, or when
a UNION ALL contains a table with inheritance children.
Bug #11457 from Nelson Page, as well as an earlier report from Elvis
Pranskevichus, showed that there were still nasty bugs associated with such
cases: in particular the EquivalenceClass mechanism could try to generate
"join" clauses connecting an appendrel child to some grandparent appendrel,
which would result in assertion failures or bogus plans.
Upon investigation I concluded that all current callers of
find_childrel_appendrelinfo() need to be fixed to explicitly consider
multiple levels of parent appendrels. The most complex fix was in
processing of "broken" EquivalenceClasses, which are ECs for which we have
been unable to generate all the derived equality clauses we would like to
because of missing cross-type equality operators in the underlying btree
operator family. That code path is more or less entirely untested by
the regression tests to date, because no standard opfamilies have such
holes in them. So I wrote a new regression test script to try to exercise
it a bit, which turned out to be quite a worthwhile activity as it exposed
existing bugs in all supported branches.
The present patch is essentially the same as far back as 9.2, which is
where parameterized paths were introduced. In 9.0 and 9.1, we only need
to back-patch a small fragment of commit 5b7b5518d, which fixes failure to
propagate out the original WHERE clauses when a broken EC contains constant
members. (The regression test case results show that these older branches
are noticeably stupider than 9.2+ in terms of the quality of the plans
generated; but we don't really care about plan quality in such cases,
only that the plan not be outright wrong. A more invasive fix in the
older branches would not be a good idea anyway from a plan-stability
standpoint.)
I left the GUC in place for the beta period, so that people could experiment
with different values. No-one's come up with any data that a different value
would be better under some circumstances, so rather than try to document to
users what the GUC, let's just hard-code the current value, 8.
As noted in http://bugs.debian.org/763098 there is a conflict between
postgres' definition of CACHE_LINE_SIZE and the definition by various
*bsd platforms. It's debatable who has the right to define such a
name, but postgres' use was only introduced in 375d8526f2 (9.4), so
it seems like a good idea to rename it.
Discussion: 20140930195756.GC27407@msg.df7cb.de
Per complaint of Christoph Berg in the above email, although he's not
the original bug reporter.
Backpatch to 9.4 where the define was introduced.
Per discussion, revert the commit which added 'ignore_nulls' to
row_to_json. This capability would be better added as an independent
function rather than being bolted on to row_to_json. Additionally,
the implementation didn't address complex JSON objects, and so was
incomplete anyway.
Pointed out by Tom and discussed with Andrew and Robert.
The original design used an array of offsets into the variable-length
portion of a JSONB container. However, such an array is basically
uncompressible by simple compression techniques such as TOAST's LZ
compressor. That's bad enough, but because the offset array is at the
front, it tended to trigger the give-up-after-1KB heuristic in the TOAST
code, so that the entire JSONB object was stored uncompressed; which was
the root cause of bug #11109 from Larry White.
To fix without losing the ability to extract a random array element in O(1)
time, change this scheme so that most of the JEntry array elements hold
lengths rather than offsets. With data that's compressible at all, there
tend to be fewer distinct element lengths, so that there is scope for
compression of the JEntry array. Every N'th entry is still an offset.
To determine the length or offset of any specific element, we might have
to examine up to N preceding JEntrys, but that's still O(1) so far as the
total container size is concerned. Testing shows that this cost is
negligible compared to other costs of accessing a JSONB field, and that
the method does largely fix the incompressible-data problem.
While at it, rearrange the order of elements in a JSONB object so that
it's "all the keys, then all the values" not alternating keys and values.
This doesn't really make much difference right at the moment, but it will
allow providing a fast path for extracting individual object fields from
large JSONB values stored EXTERNAL (ie, uncompressed), analogously to the
existing optimization for substring extraction from large EXTERNAL text
values.
Bump catversion to denote the incompatibility in on-disk format.
We will need to fix pg_upgrade to disallow upgrading jsonb data stored
with 9.4 betas 1 and 2.
Heikki Linnakangas and Tom Lane
* Don't play tricks for a more efficient pg_atomic_clear_flag() in the
generic gcc implementation. The old version was broken on gcc < 4.7
on !x86 platforms. Per buildfarm member chipmunk.
* Make usage of __atomic() fences depend on HAVE_GCC__ATOMIC_INT32_CAS
instead of HAVE_GCC__ATOMIC_INT64_CAS - there's platforms with 32bit
support that don't support 64bit atomics.
* Blindly fix two superflous #endif in generic-xlc.h
* Check for --disable-atomics in platforms but x86.
Some x86 32bit versions of gcc apparently generate references to the
nonexistant %sil register when using when using the r input
constraint, but not with the =q constraint. The latter restricts
allocations to a/b/c/d which should all work.
Several upcoming performance/scalability improvements require atomic
operations. This new API avoids the need to splatter compiler and
architecture dependent code over all the locations employing atomic
ops.
For several of the potential usages it'd be problematic to maintain
both, a atomics using implementation and one using spinlocks or
similar. In all likelihood one of the implementations would not get
tested regularly under concurrency. To avoid that scenario the new API
provides a automatic fallback of atomic operations to spinlocks. All
properties of atomic operations are maintained. This fallback -
obviously - isn't as fast as just using atomic ops, but it's not bad
either. For one of the future users the atomics ontop spinlocks
implementation was actually slightly faster than the old purely
spinlock using implementation. That's important because it reduces the
fear of regressing older platforms when improving the scalability for
new ones.
The API, loosely modeled after the C11 atomics support, currently
provides 'atomic flags' and 32 bit unsigned integers. If the platform
efficiently supports atomic 64 bit unsigned integers those are also
provided.
To implement atomics support for a platform/architecture/compiler for
a type of atomics 32bit compare and exchange needs to be
implemented. If available and more efficient native support for flags,
32 bit atomic addition, and corresponding 64 bit operations may also
be provided. Additional useful atomic operations are implemented
generically ontop of these.
The implementation for various versions of gcc, msvc and sun studio have
been tested. Additional existing stub implementations for
* Intel icc
* HUPX acc
* IBM xlc
are included but have never been tested. These will likely require
fixes based on buildfarm and user feedback.
As atomic operations also require barriers for some operations the
existing barrier support has been moved into the atomics code.
Author: Andres Freund with contributions from Oskari Saarenmaa
Reviewed-By: Amit Kapila, Robert Haas, Heikki Linnakangas and Álvaro Herrera
Discussion: CA+TgmoYBW+ux5-8Ja=Mcyuy8=VXAnVRHp3Kess6Pn3DMXAPAEA@mail.gmail.com,
20131015123303.GH5300@awork2.anarazel.de,
20131028205522.GI20248@awork2.anarazel.de
Previously, we used an lwlock that was held from the time we began
seeking a candidate buffer until the time when we found and pinned
one, which is disastrous for concurrency. Instead, use a spinlock
which is held just long enough to pop the freelist or advance the
clock sweep hand, and then released. If we need to advance the clock
sweep further, we reacquire the spinlock once per buffer.
This represents a significant increase in atomic operations around
buffer eviction, but it still wins on many workloads. On others, it
may result in no gain, or even cause a regression, unless the number
of buffer mapping locks is also increased. However, that seems like
material for a separate commit. We may also need to consider other
methods of mitigating contention on this spinlock, such as splitting
it into multiple locks or jumping the clock sweep hand more than one
buffer at a time, but those, too, seem like separate improvements.
Patch by me, inspired by a much larger patch from Amit Kapila.
Reviewed by Andres Freund.
Buildfarm member tick identified an issue where the policies in the
relcache for a relation were were being replaced underneath a running
query, leading to segfaults while processing the policies to be added
to a query. Similar to how TupleDesc RuleLocks are handled, add in a
equalRSDesc() function to check if the policies have actually changed
and, if not, swap back the rsdesc field (using the original instead of
the temporairly built one; the whole structure is swapped and then
specific fields swapped back). This now passes a CLOBBER_CACHE_ALWAYS
for me and should resolve the buildfarm error.
In addition to addressing this, add a new chapter in Data Definition
under Privileges which explains row security and provides examples of
its usage, change \d to always list policies (even if row security is
disabled- but note that it is disabled, or enabled with no policies),
rework check_role_for_policy (it really didn't need the entire policy,
but it did need to be using has_privs_of_role()), and change the field
in pg_class to relrowsecurity from relhasrowsecurity, based on
Heikki's suggestion. Also from Heikki, only issue SET ROW_SECURITY in
pg_restore when talking to a 9.5+ server, list Bypass RLS in \du, and
document --enable-row-security options for pg_dump and pg_restore.
Lastly, fix a number of minor whitespace and typo issues from Heikki,
Dimitri, add a missing #include, per Peter E, fix a few minor
variable-assigned-but-not-used and resource leak issues from Coverity
and add tab completion for role attribute bypassrls as well.
There are four weaknesses in728f152e07f998d2cb4fe5f24ec8da2c3bda98f2:
* append_init() in heapdesc.c was ugly and required that rm_identify
return values are only valid till the next call. Instead just add a
couple more switch() cases for the INIT_PAGE cases. Now the returned
value will always be valid.
* a couple rm_identify() callbacks missed masking xl_info with
~XLR_INFO_MASK.
* pg_xlogdump didn't map a NULL rm_identify to UNKNOWN or a similar
string.
* append_init() was called when id=NULL - which should never actually
happen. But it's better to be careful.
Building on the updatable security-barrier views work, add the
ability to define policies on tables to limit the set of rows
which are returned from a query and which are allowed to be added
to a table. Expressions defined by the policy for filtering are
added to the security barrier quals of the query, while expressions
defined to check records being added to a table are added to the
with-check options of the query.
New top-level commands are CREATE/ALTER/DROP POLICY and are
controlled by the table owner. Row Security is able to be enabled
and disabled by the owner on a per-table basis using
ALTER TABLE .. ENABLE/DISABLE ROW SECURITY.
Per discussion, ROW SECURITY is disabled on tables by default and
must be enabled for policies on the table to be used. If no
policies exist on a table with ROW SECURITY enabled, a default-deny
policy is used and no records will be visible.
By default, row security is applied at all times except for the
table owner and the superuser. A new GUC, row_security, is added
which can be set to ON, OFF, or FORCE. When set to FORCE, row
security will be applied even for the table owner and superusers.
When set to OFF, row security will be disabled when allowed and an
error will be thrown if the user does not have rights to bypass row
security.
Per discussion, pg_dump sets row_security = OFF by default to ensure
that exports and backups will have all data in the table or will
error if there are insufficient privileges to bypass row security.
A new option has been added to pg_dump, --enable-row-security, to
ask pg_dump to export with row security enabled.
A new role capability, BYPASSRLS, which can only be set by the
superuser, is added to allow other users to be able to bypass row
security using row_security = OFF.
Many thanks to the various individuals who have helped with the
design, particularly Robert Haas for his feedback.
Authors include Craig Ringer, KaiGai Kohei, Adam Brightwell, Dean
Rasheed, with additional changes and rework by me.
Reviewers have included all of the above, Greg Smith,
Jeff McCormick, and Robert Haas.
x86's memory barrier assembly was marked as clobbering "memory" but
not "cc" even though 'addl' sets various flags. As it turns out gcc on
x86 implicitly assumes "cc" on every inline assembler statement, so
it's not a bug. But as that's poorly documented and might get copied
to architectures or compilers where that's not the case, it seems
better to be precise.
Discussion: 20140919100016.GH4277@alap3.anarazel.de
To keep the code common, backpatch to 9.2 where explicit memory
barriers were introduced.
This is primarily useful for the upcoming pg_xlogdump --stats feature,
but also allows to remove some duplicated code in the rmgr_desc
routines.
Due to the separation and harmonization, the output of dipsplayed
records changes somewhat. But since this isn't enduser oriented
content that's ok.
It's potentially desirable to further change pg_xlogdump's display of
records. It previously wasn't possible to show the record type
separately from the description forcing it to be in the last
column. But that's better done in a separate commit.
Author: Abhijit Menon-Sen, slightly editorialized by me
Reviewed-By: Álvaro Herrera, Andres Freund, and Heikki Linnakangas
Discussion: 20140604104716.GA3989@toroid.org
They were marked to return a boolean, but they actually return a
GinTernaryValue, which is more like a "char". It makes no practical
difference, as the triConsistent functions cannot be called directly from
SQL because they have "internal" arguments, but this nevertheless seems
more correct.
Also fix the GinTernaryValue name in the documentation. I renamed the enum
earlier, but neglected the docs.
Alexander Korotkov. This is new in 9.4, so backpatch there.
This new GUC context option allows GUC parameters to have the combined
properties of PGC_BACKEND and PGC_SUSET, ie, they don't change after
session start and non-superusers can't change them. This is a more
appropriate choice for log_connections and log_disconnections than their
previous context of PGC_BACKEND, because we don't want non-superusers
to be able to affect whether their sessions get logged.
Note: the behavior for log_connections is still a bit odd, in that when
a superuser attempts to set it from PGOPTIONS, the setting takes effect
but it's too late to enable or suppress connection startup logging.
It's debatable whether that's worth fixing, and in any case there is
a reasonable argument for PGC_SU_BACKEND to exist.
In passing, re-pgindent the files touched by this commit.
Fujii Masao, reviewed by Joe Conway and Amit Kapila
Previously replication commands like IDENTIFY_COMMAND were not logged
even when log_statements is set to all. Some users who want to audit
all types of statements were not satisfied with this situation. To
address the problem, this commit adds new GUC log_replication_commands.
If it's enabled, all replication commands are logged in the server log.
There are many ways to allow us to enable that logging. For example,
we can extend log_statement so that replication commands are logged
when it's set to all. But per discussion in the community, we reached
the consensus to add separate GUC for that.
Reviewed by Ian Barwick, Robert Haas and Heikki Linnakangas.
Provide an option to skip NULL values in a row when generating a JSON
object from that row with row_to_json. This can reduce the size of the
JSON object in cases where columns are NULL without really reducing the
information in the JSON object.
This also makes row_to_json into a single function with default values,
rather than having multiple functions. In passing, change array_to_json
to also be a single function with default values (we don't add an
'ignore_nulls' option yet- it's not clear that there is a sensible
use-case there, and it hasn't been asked for in any case).
Pavel Stehule
Instead of palloc'ing each HashJoinTuple individually, allocate 32kB chunks
and pack the tuples densely in the chunks. This avoids the AllocChunk
header overhead, and the space wasted by standard allocator's habit of
rounding sizes up to the nearest power of two.
This doesn't contain any planner changes, because the planner's estimate of
memory usage ignores the palloc overhead. Now that the overhead is smaller,
the planner's estimates are in fact more accurate.
Tomas Vondra, reviewed by Robert Haas.
07c8651dd9 currently causes compilation errors on mscv (and
probably some other) compilers because our getopt_long()
implementation doesn't have support for optional_argument.
Thus implement optional_argument in our fallback implemenation. It's
quite possibly also useful in other cases.
Arguably this needs a configure check for optional_argument, but it
has existed pretty much since getopt_long() was introduced and thus
doesn't seem worth the configure runtime.
Normally I'd would not push a patch this fast, but this allows msvc to
build again and has low risk as only optional_argument behaviour has
changed.
Author: Michael Paquier and Andres Freund
Discussion: CAB7nPqS5VeedSCxrK=QouokbawgGKLpyc1Q++RRFCa_sjcSVrg@mail.gmail.com
Previously, they functioned as barriers against CPU reordering but not
compiler reordering, an odd API that required extensive use of volatile
everywhere that spinlocks are used. That's error-prone and has negative
implications for performance, so change it.
In theory, this makes it safe to remove many of the uses of volatile
that we currently have in our code base, but we may find that there are
some bugs in this effort when we do. In the long run, though, this
should make for much more maintainable code.
Patch by me. Review by Andres Freund.
This provides a convenient method of classifying input values into buckets
that are not necessarily equal-width. It works on any sortable data type.
The choice of function name is a bit debatable, perhaps, but showing that
there's a relationship to the SQL standard's width_bucket() function seems
more attractive than the other proposals.
Petr Jelinek, reviewed by Pavel Stehule
Some Sparc CPUs can be run in various coherence models, ranging from
RMO (relaxed) over PSO (partial) to TSO (total). Solaris has always
run CPUs in TSO mode while in userland, but linux didn't use to and
the various *BSDs still don't. Unfortunately the sparc TAS/S_UNLOCK
were only correct under TSO. Fix that by adding the necessary memory
barrier instructions. On sparcv8+, which should be all relevant CPUs,
these are treated as NOPs if the current consistency model doesn't
require the barriers.
Discussion: 20140630222854.GW26930@awork2.anarazel.de
Will be backpatched to all released branches once a few buildfarm
cycles haven't shown up problems. As I've no access to sparc, this is
blindly written.
Every redo routine uses the same idiom to determine what to do to a page:
check if there's a backup block for it, and if not read, the buffer if the
block exists, and check its LSN. Refactor that into a common function,
XLogReadBufferForRedo, making all the redo routines shorter and more
readable.
This has no user-visible effect, and makes no changes to the WAL format.
Reviewed by Andres Freund, Alvaro Herrera, Michael Paquier.
Since the dawn of time (aka Postgres95) multiple pins of the same
buffer by one backend have been optimized not to modify the shared
refcount more than once. This optimization has always used a NBuffer
sized array in each backend keeping track of a backend's pins.
That array (PrivateRefCount) was one of the biggest per-backend memory
allocations, depending on the shared_buffers setting. Besides the
waste of memory it also has proven to be a performance bottleneck when
assertions are enabled as we make sure that there's no remaining pins
left at the end of transactions. Also, on servers with lots of memory
and a correspondingly high shared_buffers setting the amount of random
memory accesses can also lead to poor cpu cache efficiency.
Because of these reasons a backend's buffers pins are now kept track
of in a small statically sized array that overflows into a hash table
when necessary. Benchmarks have shown neutral to positive performance
results with considerably lower memory usage.
Patch by me, review by Robert Haas.
Discussion: 20140321182231.GA17111@alap3.anarazel.de
This reverts commit e23014f3d4.
As the side effect of the reverted commit, when the unit is
specified, the reloption was stored in the catalog with the unit.
This broke pg_dump (specifically, it prevented pg_dump from
outputting restorable backup regarding the reloption) and
turned the buildfarm red. Revert the commit until the fixed
version is ready.
This introduces an infrastructure which allows us to specify the units
like ms (milliseconds) in integer relation option, like GUC parameter.
Currently only autovacuum_vacuum_cost_delay reloption can accept
the units.
Reviewed by Michael Paquier
Tom Lane
Kevin Grittner
Alvaro Herrera
Robert Haas
Andres Freund
Heikki Linnakangas
Stephen Frost
Peter Eisentraut
Andrew Dunstan
Fujii Masao
Bruce Momjian