On Mon, Feb 8, 2021 at 6:17 AM Heikki Linnakangas <
hlinnaka@iki.fi> wrote:
>
> I also tested the fallback implementation from the simdjson library
> (included in the patch, if you uncomment it in simdjson-glue.c):
>
> mixed | ascii
> -------+-------
> 447 | 46
> (1 row)
>
> I think we should at least try to adopt that. At a high level, it looks
> pretty similar your patch: you load the data 8 bytes at a time, check if
> there are all ASCII. If there are any non-ASCII chars, you check the
> bytes one by one, otherwise you load the next 8 bytes. Your patch should
> be able to achieve the same performance, if done right. I don't think
> the simdjson code forbids \0 bytes, so that will add a few cycles, but
> still.
Attached is a patch that does roughly what simdjson fallback did, except I use straight tests on the bytes and only calculate code points in assertion builds. In the course of doing this, I found that my earlier concerns about putting the ascii check in a static inline function were due to my suboptimal loop implementation. I had assumed that if the chunked ascii check failed, it had to check all those bytes one at a time. As it turns out, that's a waste of the branch predictor. In the v2 patch, we do the chunked ascii check every time we loop. With that, I can also confirm the claim in the Lemire paper that it's better to do the check on 16-byte chunks:
(MacOS, Clang 10)
master:
chinese | mixed | ascii
---------+-------+-------
1081 | 761 | 366
v2 patch, with 16-byte stride:
chinese | mixed | ascii
---------+-------+-------
806 | 474 | 83
patch but with 8-byte stride:
chinese | mixed | ascii
---------+-------+-------
792 | 490 | 105
I also included the fast path in all other multibyte encodings, and that is also pretty good performance-wise. It regresses from master on pure multibyte input, but that case is still faster than PG13, which I simulated by reverting 6c5576075b0f9 and b80e10638e3:
~PG13:
chinese | mixed | ascii
---------+-------+-------
1565 | 848 | 365
ascii fast-path plus pg_*_verifychar():
chinese | mixed | ascii
---------+-------+-------
1279 | 656 | 94
v2 has a rough start to having multiple implementations in src/backend/port. Next steps are:
1. Add more tests for utf-8 coverage (in addition to the ones to be added by the noError argument patch)
2. Add SSE4 validator -- it turns out the demo I referred to earlier doesn't match the algorithm in the paper. I plan to only copy the lookup tables from simdjson verbatim, but the code will basically be written from scratch, using simdjson as a hint.
3. Adjust
configure.ac--
John Naylor
EDB:
http://www.enterprisedb.com
