This commit includes various optimizations to improve the performance of
tuple deformation.
We now precalculate CompactAttribute's attcacheoff, which allows us to
remove the code from the deform routines which was setting the
attcacheoff. Setting the attcacheoff is now handled by
TupleDescFinalize(), which must be called before the TupleDesc is used for
anything. Having TupleDescFinalize() means we can store the first
attribute in the TupleDesc which does not have an offset cached. That
allows us to add a dedicated deforming loop to deform all attributes up
to the final one with an attcacheoff set, or up to the first NULL
attribute, whichever comes first.
Here we also improve tuple deformation performance of tuples with NULLs.
Previously, if the HEAP_HASNULL bit was set in the tuple's t_infomask,
deforming would, one-by-one, check each and every bit in the NULL bitmap
to see if it was zero. Now, we process the NULL bitmap 1 byte at a time
rather than 1 bit at a time to find the attnum with the first NULL. We
can now deform the tuple without checking for NULLs up to just before that
attribute.
We also record the maximum attribute number which is guaranteed to exist
in the tuple, that is, has a NOT NULL constraint and isn't an
atthasmissing attribute. When deforming only attributes prior to the
guaranteed attnum, we've no need to access the tuple's natt count. As an
additional optimization, we only count fixed-width columns when
calculating the maximum guaranteed column, as this eliminates the need to
emit code to fetch byref types in the deformation loop for guaranteed
attributes.
Some locations in the code deform tuples that have yet to go through NOT
NULL constraint validation. We're unable to perform the guaranteed
attribute optimization when that's the case. This optimization is opt-in
via the TupleTableSlot using the TTS_FLAG_OBEYS_NOT_NULL_CONSTRAINTS
flag.
This commit also adds a more efficient way of populating the isnull
array by using a bit-wise SWAR trick which performs multiplication on the
inverse of the tuple's bitmap byte and masking out all but the lower bit
of each of the boolean's byte. This results in much more optimal code
when compared to determining the NULLness via att_isnull(). 8 isnull
elements are processed at once using this method, which means we need to
round the tts_isnull array size up to the next 8 bytes. The palloc code
does this anyway, but the round-up needed to be formalized so as not to
overwrite the sentinel byte in MEMORY_CONTEXT_CHECKING builds. Doing
this also allows the NULL-checking deforming loop to more efficiently
check the isnull array, rather than doing the bit-wise processing for each
attribute that att_isnull() does.
The level of performance improvement from these changes seems to vary
depending on the CPU architecture. Apple's M chips seem particularly
fond of the changes, with some of the tested deform-heavy queries going
over twice as fast as before. With x86-64, the speedups aren't quite as
large. With tables containing only a small number of columns, the
speedups will be less.
Author: David Rowley <dgrowleyml@gmail.com>
Reviewed-by: Chao Li <li.evan.chao@gmail.com>
Reviewed-by: Andres Freund <andres@anarazel.de>
Reviewed-by: John Naylor <johncnaylorls@gmail.com>
Reviewed-by: Amit Langote <amitlangote09@gmail.com>
Reviewed-by: Zsolt Parragi <zsolt.parragi@percona.com>
Reviewed-by: Álvaro Herrera <alvherre@kurilemu.de>
Reviewed-by: Junwang Zhao <zhjwpku@gmail.com>
Discussion: https://postgr.es/m/CAApHDvpoFjaj3%2Bw_jD5uPnGazaw41A71tVJokLDJg2zfcigpMQ%40mail.gmail.com
Instead of using comments to mark fallthrough switch cases, use the
fallthrough attribute. This will (in the future, not here) allow
supporting other compilers besides gcc. The commenting convention is
only supported by gcc, the attribute is supported by clang, and in the
fullness of time the C23 standard attribute would allow supporting
other compilers as well.
Right now, we package the attribute into a macro called
pg_fallthrough. This commit defines that macro and replaces the
existing comments with that macro invocation.
We also raise the level of the gcc -Wimplicit-fallthrough= option from
3 to 5 to enforce the use of the attribute.
Reviewed-by: Jelte Fennema-Nio <postgres@jeltef.nl>
Discussion: https://www.postgresql.org/message-id/flat/76a8efcd-925a-4eaf-bdd1-d972cd1a32ff%40eisentraut.org
The idea is to encourage more the use of these new routines across the
tree, as these offer stronger type safety guarantees than palloc().
This batch of changes includes most of the trivial changes suggested by
the author for src/backend/.
A total of 334 files are updated here. Among these files, 48 of them
have their build change slightly; these are caused by line number
changes as the new allocation formulas are simpler, shaving around 100
lines of code in total.
Similar work has been done in 0c3c5c3b06 and 31d3847a37.
Author: David Geier <geidav.pg@gmail.com>
Discussion: https://postgr.es/m/ad0748d4-3080-436e-b0bc-ac8f86a3466a@gmail.com
In v14, bb437f995 added support for scanning for ranges of TIDs using a
dedicated executor node for the purpose. Here, we allow these scans to
be parallelized. The range of blocks to scan is divvied up similarly to
how a Parallel Seq Scans does that, where 'chunks' of blocks are
allocated to each worker and the size of those chunks is slowly reduced
down to 1 block per worker by the time we're nearing the end of the
scan. Doing that means workers finish at roughly the same time.
Allowing TID Range Scans to be parallelized removes the dilemma from the
planner as to whether a Parallel Seq Scan will cost less than a
non-parallel TID Range Scan due to the CPU concurrency of the Seq Scan
(disk costs are not divided by the number of workers). It was possible
the planner could choose the Parallel Seq Scan which would result in
reading additional blocks during execution than the TID Scan would have.
Allowing Parallel TID Range Scans removes the trade-off the planner
makes when choosing between reduced CPU costs due to parallelism vs
additional I/O from the Parallel Seq Scan due to it scanning blocks from
outside of the required TID range. There is also, of course, the
traditional parallelism performance benefits to be gained as well, which
likely doesn't need to be explained here.
Author: Cary Huang <cary.huang@highgo.ca>
Author: David Rowley <dgrowleyml@gmail.com>
Reviewed-by: Junwang Zhao <zhjwpku@gmail.com>
Reviewed-by: Rafia Sabih <rafia.pghackers@gmail.com>
Reviewed-by: Steven Niu <niushiji@gmail.com>
Discussion: https://postgr.es/m/18f2c002a24.11bc2ab825151706.3749144144619388582@highgo.ca
The EvalPlanQual recheck for TID Range Scan wasn't rechecking the TID qual
still passed after following update chains. This could result in tuples
being updated or deleted by plans using TID Range Scans where the ctid of
the new (updated) tuple no longer matches the clause of the scan. This
isn't desired behavior, and isn't consistent with what would happen if the
chosen plan had used an Index or Seq Scan, and that could lead to hard to
predict behavior for scans that contain TID quals and other quals as the
planner has freedom to choose TID Range or some other non-TID scan method
for such queries, and the chosen plan could change at any moment.
Here we fix this by properly implementing the recheck function for TID
Range Scans.
Backpatch to 14, where TID Range Scans were added
Reported-by: Sophie Alpert <pg@sophiebits.com>
Author: Sophie Alpert <pg@sophiebits.com>
Author: David Rowley <dgrowleyml@gmail.com>
Reviewed-by: David Rowley <dgrowleyml@gmail.com>
Reviewed-by: Chao Li <li.evan.chao@gmail.com>
Discussion: https://postgr.es/m/4a6268ff-3340-453a-9bf5-c98d51a6f729@app.fastmail.com
Backpatch-through: 14
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
Since C99, there can be a trailing comma after the last value in an
enum definition. A lot of new code has been introducing this style on
the fly. Some new patches are now taking an inconsistent approach to
this. Some add the last comma on the fly if they add a new last
value, some are trying to preserve the existing style in each place,
some are even dropping the last comma if there was one. We could
nudge this all in a consistent direction if we just add the trailing
commas everywhere once.
I omitted a few places where there was a fixed "last" value that will
always stay last. I also skipped the header files of libpq and ecpg,
in case people want to use those with older compilers. There were
also a small number of cases where the enum type wasn't used anywhere
(but the enum values were), which ended up confusing pgindent a bit,
so I left those alone.
Discussion: https://www.postgresql.org/message-id/flat/386f8c45-c8ac-4681-8add-e3b0852c1620%40eisentraut.org
This commit removes unnecessary ExecExprFreeContext() calls in
ExecEnd* routines because the actual cleanup is managed by
FreeExecutorState(). With no callers remaining for
ExecExprFreeContext(), this commit also removes the function.
This commit also drops redundant ExecClearTuple() calls, because
ExecResetTupleTable() in ExecEndPlan() already takes care of
resetting and dropping all TupleTableSlots initialized with
ExecInitScanTupleSlot() and ExecInitExtraTupleSlot().
After these modifications, the ExecEnd*() routines for ValuesScan,
NamedTuplestoreScan, and WorkTableScan became redundant. So, this
commit removes them.
Reviewed-by: Robert Haas
Discussion: https://postgr.es/m/CA+HiwqFGkMSge6TgC9KQzde0ohpAycLQuV7ooitEEpbKB0O_mg@mail.gmail.com
The IsCTIDVar() tests in nodeTidscan.c and nodeTidrangescan.c
look buggy at first sight: they aren't checking that the varno
matches the table to be scanned. Actually they're safe because
any Var in a scan-level qual must be for the correct table ...
but if we're depending on that, it's pretty pointless to verify
varlevelsup. (Besides which, varlevelsup is *always* zero at
execution, since we've flattened the rangetable long since.)
Remove the useless varlevelsup check, and instead add some
commentary explaining why we don't need to check varno.
Noted while fooling with a planner change that causes the order
of "t1.ctid = t2.ctid" to change in some tidscan.sql tests;
I was briefly fooled into thinking there was a live bug here.
This adds a new executor node named TID Range Scan. The query planner
will generate paths for TID Range scans when quals are discovered on base
relations which search for ranges on the table's ctid column. These
ranges may be open at either end. For example, WHERE ctid >= '(10,0)';
will return all tuples on page 10 and over.
To support this, two new optional callback functions have been added to
table AM. scan_set_tidrange is used to set the scan range to just the
given range of TIDs. scan_getnextslot_tidrange fetches the next tuple
in the given range.
For AMs were scanning ranges of TIDs would not make sense, these functions
can be set to NULL in the TableAmRoutine. The query planner won't
generate TID Range Scan Paths in that case.
Author: Edmund Horner, David Rowley
Reviewed-by: David Rowley, Tomas Vondra, Tom Lane, Andres Freund, Zhihong Yu
Discussion: https://postgr.es/m/CAMyN-kB-nFTkF=VA_JPwFNo08S0d-Yk0F741S2B7LDmYAi8eyA@mail.gmail.com
