Horton

Re: WIP: Faster Expression Processing v4

Ants Aasma <ants.aasma@eesti.ee>

View thread View on postgr.es
From: Ants Aasma <ants.aasma@eesti.ee>
To: Andres Freund <andres@anarazel.de>
Cc: Tom Lane <tgl@sss.pgh.pa.us>, Heikki Linnakangas <hlinnaka@iki.fi>, Tomas Vondra <tomas.vondra@2ndquadrant.com>, PostgreSQL-development <pgsql-hackers@postgresql.org>
Date: 2017-03-25T23:56:11Z
Lists: pgsql-hackers 
On Sun, Mar 26, 2017 at 12:22 AM, Andres Freund <andres@anarazel.de> wrote:
>> At least with current gcc (6.3.1 on Fedora 25) at -O2,
>> what I see is multiple places jumping to the same indirect jump
>> instruction :-(.  It's not a total disaster: as best I can tell, all the
>> uses of EEO_JUMP remain distinct.  But gcc has chosen to implement about
>> 40 of the 71 uses of EEO_NEXT by jumping to the same couple of
>> instructions that increment the "op" register and then do an indirect
>> jump :-(.
>
> Yea, I see some of that too - "usually" when there's more than just the
> jump in common.  I think there's some gcc variables that influence this
> (min-crossjump-insns (5), max-goto-duplication-insns (8)).  Might be
> worthwhile experimenting with setting them locally via a pragma or such.
> I think Aants wanted to experiment with that, too.

I haven't had the time to research this properly, but initial tests
show that with GCC 6.2 adding

#pragma GCC optimize ("no-crossjumping")

fixes merging of the op tail jumps.

Some quick and dirty benchmarking suggests that the benefit for the
interpreter is about 15% (5% speedup on a workload that spends 1/3 in
ExecInterpExpr). My idea of prefetching op->resnull/resvalue to local
vars before the indirect jump is somewhere between a tiny benefit and
no effect, certainly not worth introducing extra complexity. Clang 3.8
does the correct thing out of the box and is a couple of percent
faster than GCC with the pragma.

Regards,
Ants Aasma

Commits

  1. Improve performance of ExecEvalWholeRowVar.

  2. Remove unreachable code in expression evaluation.

  3. Faster expression evaluation and targetlist projection.

  4. Avoid syntax error on platforms that have neither LOCALE_T nor ICU.

  5. Add configure test to see if the C compiler has gcc-style computed gotos.

  6. Improve regression test coverage for TID scanning.

  7. Improve expression evaluation test coverage.

  8. Fix two errors with nested CASE/WHEN constructs.