Date: Tue, 31 Mar 2026 11:30:54 -0500
From: Nathan Bossart <nathandbossart@gmail.com>
To: KAZAR Ayoub <ma_kazar@esi.dz>
Cc: Andres Freund <andres@anarazel.de>,
	Pg Hackers <pgsql-hackers@postgresql.org>,
	Neil Conway <neil.conway@gmail.com>,
	Manni Wood <manni.wood@enterprisedb.com>,
	Andrew Dunstan <andrew@dunslane.net>,
	Shinya Kato <shinya11.kato@gmail.com>,
	Mark Wong <markwkm@gmail.com>,
	Nazir Bilal Yavuz <byavuz81@gmail.com>
Subject: Re: Speed up COPY TO text/CSV parsing using SIMD
Message-ID: <acv2vu8miagnHG1B@nathan>
References: 
 <CA+K2Runi_H2CBL0yMm3De2KqcR9RMA0HK5cLJjEhoNszC7myeg@mail.gmail.com>
 <aY5C6Xa5im72NF_Y@alap3.anarazel.de>
 <CA+K2Rum_QTZqTUrdMOL5hr-OOpCwGR_9Nj1z15BFObjktMOY6A@mail.gmail.com>
 <abBuKalOno33MQFw@nathan>
 <CA+K2Rum7+Jm2rm65K5msxaiAM8QTkhSNAYarPBP9O7nBXYo12Q@mail.gmail.com>
 <abmiNPQOqBrRlf_m@nathan>
 <CA+K2Rum-TB_iNzDWoXOJspf=jq0gd-wees8+9tBTJNyhy9cK5g@mail.gmail.com>
 <CA+K2Ru=PdZuXQbcfvqKysTkebTyXNd9j7dp+mTFQEYpLdGw1eA@mail.gmail.com>
 <acWj5FntidHJ9nVP@nathan>
 <CA+K2Runq+1gy8p6a-DsxpT2OkEkEu3cUGsZ9tdiGNrg_=P39gg@mail.gmail.com>
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
In-Reply-To: 
 <CA+K2Runq+1gy8p6a-DsxpT2OkEkEu3cUGsZ9tdiGNrg_=P39gg@mail.gmail.com>
Archived-At: <https://www.postgresql.org/message-id/acv2vu8miagnHG1B%40nathan>
Precedence: bulk

On Fri, Mar 27, 2026 at 07:48:38PM +0100, KAZAR Ayoub wrote:
> I added a prescan loop inside the simd helpers trying to catch special
> chars in sizeof(Vector8) characters, i measured how good is this at
> reducing the overhead of starting simd and exiting at first vector:
> the scalar loop is better than SIMD for one vector if it finds a special
> character before 6th character, worst case is not a clean vector, where the
> scalar loop needs 20 more cycles compared to SIMD.
> This helps mitigate the case of JSON(B) in CSV format, this is why I only
> added this for CSV case only.

Interesting.

> In a benchmark with 10M early SIMD exit like the JSONB case, the previous
> 3% regression is gone.

While these are nice results, I think it's best that we target v20 for this
patch so that we have more time to benchmark and explore edge cases.

-- 
nathan