LockBufferForCleanup() acquires a cleanup lock unconditionally, and
ConditionalLockBufferForCleanup() acquires a cleanup lock if it is
possible to do so without waiting; this patch adds a new API,
IsBufferCleanupOK(), which tests whether an exclusive lock already
held happens to be a cleanup lock. This is possible because a cleanup
lock simply means an exclusive lock plus the assurance any other pins
on the buffer are newer than our own pin. Therefore, just as the
existing functions decide that the exclusive lock that they've just
taken is a cleanup lock if they observe the pin count to be 1, this
new function allows us to observe that the pin count is 1 on a buffer
we've already locked.
This is useful in situations where a backend definitely wishes to
modify the buffer and also wishes to perform cleanup operations if
possible. The patch to eliminate heavyweight locking by hash indexes
uses this, and it may have other applications as well.
Amit Kapila, per a suggestion from me. Some comment adjustments by me
as well.
Commit 2bd9e412f9 added these in error.
They were part of an earlier design for that patch and survived in the
committed version only by inadvertency.
Julien Rouhaud
This will write the received transaction log into a file called
pg_wal.tar(.gz) next to the other tarfiles instead of writing it to
base.tar. When using fetch mode, the transaction log is still written to
base.tar like before, and when used against a pre-10 server, the file
is named pg_xlog.tar.
To do this, implement a new concept of a "walmethod", which is
responsible for writing the WAL. Two implementations exist, one that
writes to a plain directory (which is also used by pg_receivexlog) and
one that writes to a tar file with optional compression.
Reviewed by Michael Paquier
"xlog" is not a particularly clear abbreviation for "write-ahead log",
and it sometimes confuses users into believe that the contents of the
"pg_xlog" directory are not critical data, leading to unpleasant
consequences. So, rename the directory to "pg_wal".
This patch modifies pg_upgrade and pg_basebackup to understand both
the old and new directory layouts; the former is necessary given the
purpose of the tool, while the latter merely avoids an unnecessary
backward-compatibility break.
We may wish to consider renaming other programs, switches, and
functions which still use the old "xlog" naming to also refer to
"wal". However, that's still under discussion, so let's do just this
much for now.
Discussion: CAB7nPqTeC-8+zux8_-4ZD46V7YPwooeFxgndfsq5Rg8ibLVm1A@mail.gmail.com
Michael Paquier
This reverts commit 9e083fd468. That was a
few bricks shy of a load:
* Query cancel stopped working
* Buildfarm member pademelon stopped working, because the box doesn't have
/dev/urandom nor /dev/random.
This clearly needs some more discussion, and a quite different patch, so
revert for now.
This adds a new routine, pg_strong_random() for generating random bytes,
for use in both frontend and backend. At the moment, it's only used in
the backend, but the upcoming SCRAM authentication patches need strong
random numbers in libpq as well.
pg_strong_random() is based on, and replaces, the existing implementation
in pgcrypto. It can acquire strong random numbers from a number of sources,
depending on what's available:
- OpenSSL RAND_bytes(), if built with OpenSSL
- On Windows, the native cryptographic functions are used
- /dev/urandom
- /dev/random
Original patch by Magnus Hagander, with further work by Michael Paquier
and me.
Discussion: <CAB7nPqRy3krN8quR9XujMVVHYtXJ0_60nqgVc6oUk8ygyVkZsA@mail.gmail.com>
The more efficient hashtable speeds up hash-aggregations with more than
a few hundred groups significantly. Improvements of over 120% have been
measured.
Due to the the different hash table queries that not fully
determined (e.g. GROUP BY without ORDER BY) may change their result
order.
The conversion is largely straight-forward, except that, due to the
static element types of simplehash.h type hashes, the additional data
some users store in elements (e.g. the per-group working data for hash
aggregaters) is now stored in TupleHashEntryData->additional. The
meaning of BuildTupleHashTable's entrysize (renamed to additionalsize)
has been changed to only be about the additionally stored size. That
size is only used for the initial sizing of the hash-table.
Reviewed-By: Tomas Vondra
Discussion: <20160727004333.r3e2k2y6fvk2ntup@alap3.anarazel.de>
dynahash.c hash tables aren't quite fast enough for some
use-cases. There are several reasons for lacking performance:
- the use of chaining for collision handling makes them cache
inefficient, that's especially an issue when the tables get bigger.
- as the element sizes for dynahash are only determined at runtime,
offset computations are somewhat expensive
- hash and element comparisons are indirect function calls, causing
unnecessary pipeline stalls
- it's two level structure has some benefits (somewhat natural
partitioning), but increases the number of indirections
to fix several of these the hash tables have to be adjusted to the
individual use-case at compile-time. C unfortunately doesn't provide a
good way to do compile code generation (like e.g. c++'s templates for
all their weaknesses do). Thus the somewhat ugly approach taken here is
to allow for code generation using a macro-templatized header file,
which generates functions and types based on a prefix and other
parameters.
Later patches use this infrastructure to use such hash tables for
tidbitmap.c (bitmap scans) and execGrouping.c (hash aggregation,
...). In queries where these use up a large fraction of the time, this
has been measured to lead to performance improvements of over 100%.
There are other cases where this could be useful (e.g. catcache.c).
The hash table design chosen is a variant of linear open-addressing. The
biggest disadvantage of simple linear addressing schemes are highly
variable lookup times due to clustering, and deletions leaving a lot of
tombstones around. To address these issues a variant of "robin hood"
hashing is employed. Robin hood hashing optimizes chaining lengths by
moving elements close to their optimal bucket ("rich" elements), out of
the way if a to-be-inserted element is further away from its optimal
position (i.e. it's "poor"). While that can make insertions slower, the
average lookup performance is a lot better, and higher fill factors can
be used in a still performant manner. To avoid tombstones - which
normally solve the issue that a deleted node's presence is relevant to
determine whether a lookup needs to continue looking or is done -
buckets following a deleted element are shifted backwards, unless
they're empty or already at their optimal position.
There's further possible improvements that can be made to this
implementation. Amongst others:
- Use distance as a termination criteria during searches. This is
generally a good idea, but I've been able to see the overhead of
distance calculations in some cases.
- Consider combining the 'empty' status into the hashvalue, and enforce
storing the hashvalue. That could, in some cases, increase memory
density and remove a few instructions.
- Experiment further with the, very conservatively choosen, fillfactor.
- Make maximum size of hashtable configurable, to allow storing very
very large tables. That'd require 64bit hash values to be more common
than now, though.
- some smaller memcpy calls could be optimized to copy larger chunks
But since the new implementation is already considerably faster than
dynahash it seem sensible to start using it.
Reviewed-By: Tomas Vondra
Discussion: <20160727004333.r3e2k2y6fvk2ntup@alap3.anarazel.de>
These are useful for very hot code paths. Because it's easy to guess
wrongly about likelihood, and because such likelihoods change over time,
they should be used sparingly.
Past tests have shown it'd be a good idea to use them in some places,
e.g. in error checks around ereports that ERROR out, but that's work for
later.
Discussion: <20160727004333.r3e2k2y6fvk2ntup@alap3.anarazel.de>
This turns out not to be as harmless as I thought: MSVC will complain
if it sees an "extern" declaration without PGDLLEXPORT and then one with.
(Seems fairly silly, given that this can be changed after the fact by the
linker, but there you have it.) Therefore, contrib modules that have
extern's for V1 functions in header files are falling over in the
buildfarm, since none of those externs are marked PGDLLEXPORT.
We might or might not conclude that we're willing to plaster those
declarations with PGDLLEXPORT in HEAD, but in any case there's no way we're
going to ship this change in the back branches. Third-party authors would
not thank us for breaking their code in a minor release. Hence, revert
the addition of PGDLLEXPORT (but let's keep the extra info in the comment).
If we do the other changes we can revert this commit in HEAD.
Per buildfarm.
These functions were originally added in commit d8cedf67a to support
use of int2vector columns as catcache lookup keys. However, there are
no catcaches that use such columns. (Indeed I now think it must always
have been dead code: a catcache with such a key column would need an
underlying unique index on the column, but we've never had an int2vector
btree opclass.)
Getting rid of the int2vector-specific operator and function does not
lose any functionality, because operations on int2vectors will now fall
back to the generic anyarray support. This avoids a wart that a btree
index on an int2vector column (made using anyarray_ops) would fail to
match equality searches, because int2vectoreq wasn't a member of the
opclass. We don't really care much about that, since int2vector is not
meant as a type for users to use, but it's silly to have extra code and
less functionality.
If we ever do want a catcache to be indexed by an int2vector column,
we'd need to put back full btree and hash opclasses for int2vector,
comparable to the support for oidvector. (The anyarray code can't be
used at such a low level, because it needs to do catcache lookups.)
But we'll deal with that if/when the need arises.
Also worth noting is that removal of the hash int2vector_ops opclass will
break any user-created hash indexes on int2vector columns. While hash
anyarray_ops would serve the same purpose, it would probably not compute
the same hash values and thus wouldn't be on-disk-compatible. Given that
int2vector isn't a user-facing type and we're planning other incompatible
changes in hash indexes for v10 anyway, this doesn't seem like something
to worry about, but it's probably worth mentioning here.
Amit Langote
Discussion: <d9bb74f8-b194-7307-9ebd-90645d377e45@lab.ntt.co.jp>
This isn't really necessary for our own code, because we use a .DEF file
in MSVC builds (see gendef.pl), or --export-all-symbols in MinGW and
Cygwin builds, to ensure that all global symbols in loadable modules
will be exported on Windows. However, third-party authors might use
different build processes that need this marker, and it's harmless
enough for our own builds.
To some extent, this is an oversight in commit e7128e8db, so back-patch
to 9.4 where that was added.
Laurenz Albe
Discussion: <A737B7A37273E048B164557ADEF4A58B539300BD@ntex2010a.host.magwien.gv.at>
Pass the buffer size as argument to LogicalTapeRewindForRead, rather than
setting it earlier with the separate LogicTapeAssignReadBufferSize call.
This way, the buffer size is set closer to where it's actually used, which
makes the code easier to understand.
This makes the calculation for how much memory to use for the buffers less
precise. We now use the same amount of memory for every tape, rounded down
to the nearest BLCKSZ boundary, instead of using one more block for some
tapes, to get the total up to exact amount of memory available. That should
be OK, merging isn't too sensitive to the exact amount of memory used.
Reviewed by Peter Geoghegan
Discussion: <0f607c4b-df23-353e-bf56-c0389d28495f@iki.fi>
While this isn't a lot of code, it's been essentially untestable for
a very long time, because libpq doesn't support anything older than
protocol 2.0, and has not since release 6.3. There's no reason to
believe any other client-side code still uses that protocol, either.
Discussion: <2661.1475849167@sss.pgh.pa.us>
SCO OpenServer and SCO UnixWare are more or less dead platforms.
We have never had a buildfarm member testing the "sco" port, and
the last "unixware" member was last heard from in 2012, so it's
fair to doubt that the code even compiles anymore on either one.
Remove both ports. We can always undo this if someone shows up
with an interest in maintaining and testing these platforms.
Discussion: <17177.1476136994@sss.pgh.pa.us>
I somehow had assumed that in the spinlock (in turn possibly using
semaphores) based fallback atomics implementation 32 bit writes could be
done without a lock. As far as the write goes that's correct, since
postgres supports only platforms with single-copy atomicity for aligned
32bit writes. But writing without holding the spinlock breaks
read-modify-write operations like pg_atomic_compare_exchange_u32(),
since they'll potentially "miss" a concurrent write, which can't happen
in actual hardware implementations.
In 9.6+ when using the fallback atomics implementation this could lead
to buffer header locks not being properly marked as released, and
potentially some related state corruption. I don't see a related danger
in 9.5 (earliest release with the API), because pg_atomic_write_u32()
wasn't used in a concurrent manner there.
The state variable of local buffers, before this change, were
manipulated using pg_atomic_write_u32(), to avoid unnecessary
synchronization overhead. As that'd not be the case anymore, introduce
and use pg_atomic_unlocked_write_u32(), which does not correctly
interact with RMW operations.
This bug only caused issues when postgres is compiled on platforms
without atomics support (i.e. no common new platform), or when compiled
with --disable-atomics, which explains why this wasn't noticed in
testing.
Reported-By: Tom Lane
Discussion: <14947.1475690465@sss.pgh.pa.us>
Backpatch: 9.5-, where the atomic operations API was introduced.
Windows apparently has a constant named WAIT_TIMEOUT, and some of these
other names are pretty generic, too. Insert "PG_" at the front of each
name in order to disambiguate.
Michael Paquier
WaitLatch, WaitLatchOrSocket, and WaitEventSetWait now taken an
additional wait_event_info parameter; legal values are defined in
pgstat.h. This makes it possible to uniquely identify every point in
the core code where we are waiting for a latch; extensions can pass
WAIT_EXTENSION.
Because latches were the major wait primitive not previously covered
by this patch, it is now possible to see information in
pg_stat_activity on a large number of important wait events not
previously addressed, such as ClientRead, ClientWrite, and SyncRep.
Unfortunately, many of the wait events added by this patch will fail
to appear in pg_stat_activity because they're only used in background
processes which don't currently appear in pg_stat_activity. We should
fix this either by creating a separate view for such information, or
else by deciding to include them in pg_stat_activity after all.
Michael Paquier and Robert Haas, reviewed by Alexander Korotkov and
Thomas Munro.
Commit 88e982302 invented GUC_UNIT_XSEGS for min_wal_size and max_wal_size,
but neglected to make it display sensibly in pg_settings.unit (by adding a
case to the switch in GetConfigOptionByNum). Fix that, and adjust said
switch to throw a run-time error the next time somebody forgets.
In passing, avoid using a static buffer for the output string --- the rest
of this function pstrdup's from a local buffer, and I see no very good
reason why the units code should do it differently and less safely.
Per report from Otar Shavadze. Back-patch to 9.5 where the new unit type
was added.
Report: <CAG-jOyA=iNFhN+yB4vfvqh688B7Tr5SArbYcFUAjZi=0Exp-Lg@mail.gmail.com>
Don't pre-read tuples into SortTuple slots during merge. Instead, use the
memory for larger read buffers in logtape.c. We're doing the same number
of READTUP() calls either way, but managing the pre-read SortTuple slots
is much more complicated. Also, the on-tape representation is more compact
than SortTuples, so we can fit more pre-read tuples into the same amount
of memory this way. And we have better cache-locality, when we use just a
small number of SortTuple slots.
Now that we only hold one tuple from each tape in the SortTuple slots, we
can greatly simplify the "batch memory" management. We now maintain a
small set of fixed-sized slots, to hold the tuples, and fall back to
palloc() for larger tuples. We use this method during all merge phases,
not just the final merge, and also when randomAccess is requested, and
also in the TSS_SORTEDONTAPE case. In other words, it's used whenever we
do an external sort.
Reviewed by Peter Geoghegan and Claudio Freire.
Discussion: <CAM3SWZTpaORV=yQGVCG8Q4axcZ3MvF-05xe39ZvORdU9JcD6hQ@mail.gmail.com>
Several places weren't careful about fsyncing in the way. See 1d4a0ab1
and 606e0f98 for details about required fsyncs.
This adds a couple of functions in src/common/ that have an equivalent
in the backend: durable_rename(), fsync_parent_path()
From: Michael Paquier <michael.paquier@gmail.com>
The previous patch broke this by returning NULL for a failed CRC check,
which pg_controldata would then try to read. Fix by returning the
result of the CRC check in a separate argument.
Michael Paquier and myself
This makes the parameter easier to extend, to support other password-based
authentication protocols than MD5. (SCRAM is being worked on.)
The GUC still accepts on/off as aliases for "md5" and "plain", although
we may want to remove those once we actually add support for another
password hash type.
Michael Paquier, reviewed by David Steele, with some further edits by me.
Discussion: <CAB7nPqSMXU35g=W9X74HVeQp0uvgJxvYOuA4A-A3M+0wfEBv-w@mail.gmail.com>
We had thirty different GIN array opclasses sharing the same operators and
support functions. That still didn't cover all the built-in types, nor
did it cover arrays of extension-added types. What we want is a single
polymorphic opclass for "anyarray". There were two missing features needed
to make this possible:
1. We have to be able to declare the index storage type as ANYELEMENT
when the opclass is declared to index ANYARRAY. This just takes a few
more lines in index_create(). Although this currently seems of use only
for GIN, there's no reason to make index_create() restrict it to that.
2. We have to be able to identify the proper GIN compare function for
the index storage type. This patch proceeds by making the compare function
optional in GIN opclass definitions, and specifying that the default btree
comparison function for the index storage type will be looked up when the
opclass omits it. Again, that seems pretty generically useful.
Since the comparison function lookup is done in initGinState(), making
use of the second feature adds an additional cache lookup to GIN index
access setup. It seems unlikely that that would be very noticeable given
the other costs involved, but maybe at some point we should consider
making GinState data persist longer than it now does --- we could keep it
in the index relcache entry, perhaps.
Rather fortuitously, we don't seem to need to do anything to get this
change to play nice with dump/reload or pg_upgrade scenarios: the new
opclass definition is automatically selected to replace existing index
definitions, and the on-disk data remains compatible. Also, if a user has
created a custom opclass definition for a non-builtin type, this doesn't
break that, since CREATE INDEX will prefer an exact match to opcintype
over a match to ANYARRAY. However, if there's anyone out there with
handwritten DDL that explicitly specifies _bool_ops or one of the other
replaced opclass names, they'll need to adjust that.
Tom Lane, reviewed by Enrique Meneses
Discussion: <14436.1470940379@sss.pgh.pa.us>
We weren't terribly consistent about whether to call Apple's OS "OS X"
or "Mac OS X", and the former is probably confusing to people who aren't
Apple users. Now that Apple has rebranded it "macOS", follow their lead
to establish a consistent naming pattern. Also, avoid the use of the
ancient project name "Darwin", except as the port code name which does not
seem desirable to change. (In short, this patch touches documentation and
comments, but no actual code.)
I didn't touch contrib/start-scripts/osx/, either. I suspect those are
obsolete and due for a rewrite, anyway.
I dithered about whether to apply this edit to old release notes, but
those were responsible for quite a lot of the inconsistencies, so I ended
up changing them too. Anyway, Apple's being ahistorical about this,
so why shouldn't we be?
Commits 470d886c3 et al intended to fix the problem that the postmaster
selected the same "random" DSM control segment ID on every start. But
using PostmasterRandom() for that destroys the intended property that the
delay between random_start_time and random_stop_time will be unpredictable.
(Said delay is probably already more predictable than we could wish, but
that doesn't mean that reducing it by a couple orders of magnitude is OK.)
Revert the previous patch and add a comment warning against misuse of
PostmasterRandom. Fix the original problem by calling srandom() early in
PostmasterMain, using a low-security seed that will later be overwritten
by PostmasterRandom.
Discussion: <20789.1474390434@sss.pgh.pa.us>
Past refactorings have removed all but one reference to SizeOfIptrData
(and that one place was in a pretty noncritical spot). Since nobody's
complained, it seems probable that there are no supported compilers
that don't think sizeof(ItemPointerData) is 6. If there are, we're
wasting MAXALIGN per heap tuple anyway, so it's rather silly to worry
about whether we can shave space in places like WAL records.
Pavan Deolasee
Discussion: <CABOikdOOawDda4hwLOT6zdA6MFfPLu3Z2YBZkX0JdayNS6JOeQ@mail.gmail.com>
pg_ctl used to determine whether a server was in standby mode by looking
for a recovery.conf file. With this change, it instead looks into
pg_control, which is potentially more accurate. There are also
occasional discussions about removing recovery.conf, so this removes one
dependency.
Reviewed-by: Michael Paquier <michael.paquier@gmail.com>
The GiST search queue is implemented as a pairing heap rather than as
Red-Black Tree, since 9.5 (commit e7032610). I neglected these comments
in that commit.
This function has no direct callers at present, but it's convenient for
manual use in a debugger, rather than having to inspect memory and do
bit-counting in your head.
In passing, get rid of useless outBitmapset() wrapper around
_outBitmapset(); let's just export the function that does the work.
Likewise for outToken().
Ashutosh Bapat, tweaked a bit by me
Discussion: <CAFjFpRdiht8e1HTVirbubr4YzaON5iZTzFJjq909y4sU8M_6eA@mail.gmail.com>
LibreSSL defines OPENSSL_VERSION_NUMBER to claim that it is version 2.0.0,
but it doesn't have the functions added in OpenSSL 1.1.0. Add autoconf
checks for the individual functions we need, and stop relying on
OPENSSL_VERSION_NUMBER.
Backport to 9.5 and 9.6, like the patch that broke this. In the
back-branches, there are still a few OPENSSL_VERSION_NUMBER checks left,
to check for OpenSSL 0.9.8 or 0.9.7. I left them as they were - LibreSSL
has all those functions, so they work as intended.
Per buildfarm member curculio.
Discussion: <2442.1473957669@sss.pgh.pa.us>
We were misapplying NameGetDatum() to plain C strings in some places.
This worked, because it was just a pointer cast anyway, but it's a type
cheat in some sense. Use CStringGetDatum instead, and modify the
NameGetDatum macro so it won't compile if applied to something that's
not a pointer to NameData. This should result in no changes to
generated code, but it is logically cleaner.
Mark Dilger, tweaked a bit by me
Discussion: <EFD8AC94-4C1F-40C1-A5EA-304080089C1B@gmail.com>
Teach the parser to reject misplaced set-returning functions during parse
analysis using p_expr_kind, in much the same way as we do for aggregates
and window functions (cf commit eaccfded9). While this isn't complete
(it misses nesting-based restrictions), it's much better than the previous
error reporting for such cases, and it allows elimination of assorted
ad-hoc expression_returns_set() error checks. We could add nesting checks
later if it seems important to catch all cases at parse time.
There is one case the parser will now throw error for although previous
versions allowed it, which is SRFs in the tlist of an UPDATE. That never
behaved sensibly (since it's ill-defined which generated row should be
used to perform the update) and it's hard to see why it should not be
treated as an error. It's a release-note-worthy change though.
Also, add a new Query field hasTargetSRFs reporting whether there are
any SRFs in the targetlist (including GROUP BY/ORDER BY expressions).
The parser can now set that basically for free during parse analysis,
and we can use it in a number of places to avoid expression_returns_set
searches. (There will be more such checks soon.) In some places, this
allows decontorting the logic since it's no longer expensive to check for
SRFs in the tlist --- so I made the checks parallel to the handling of
hasAggs/hasWindowFuncs wherever it seemed appropriate.
catversion bump because adding a Query field changes stored rules.
Andres Freund and Tom Lane
Discussion: <24639.1473782855@sss.pgh.pa.us>
lockdefs.h was only split from lock.h relatively recently, and
represents a minimal subset of the old lock.h. heapam.h only needs
that smaller subset, so adjust it to include only that. This requires
some corresponding adjustments elsewhere.
Peter Geoghegan
Some experimentation with an older version of gcc showed that it is able
to determine whether "if (elevel_ >= ERROR)" is compile-time constant
if elevel_ is declared "const", but otherwise not so much. We had
accounted for that in ereport() but were too miserly with braces to
make it so in elog(). I don't know how many currently-interesting
compilers have the same quirk, but in case it will save some code
space, let's make sure that elog() is on the same footing as ereport()
for this purpose.
Back-patch to 9.3 where we introduced pg_unreachable() calls into
elog/ereport.
The full generality of deleting an arbitrary number of tuples is no longer
needed, so let's save some code and cycles by replacing the original coding
with an implementation based on PageIndexTupleDelete.
We can always get back the old code from git if we need it again for new
callers (though I don't care for its willingness to mess with line pointers
it wasn't told to mess with).
Discussion: <552.1473445163@sss.pgh.pa.us>
Nowadays this is just a backwards-compatibility wrapper around
PageAddItemExtended, so let's avoid the extra level of function call.
In addition, because pretty much all callers are passing constants
for the two bool arguments, compilers will be able to constant-fold
the conversion to a flags bitmask.
Discussion: <552.1473445163@sss.pgh.pa.us>
Robert Haas
Peter Eisentraut
Heikki Linnakangas
Magnus Hagander
Andres Freund
Tom Lane