Re: different engine for JIT

From: Adrien Nayrat <adrien(dot)nayrat(at)anayrat(dot)info>
To: Pavel Stehule <pavel(dot)stehule(at)gmail(dot)com>, PostgreSQL Hackers <pgsql-hackers(at)postgresql(dot)org>
Subject: Re: different engine for JIT
Date: 2024-05-03 09:21:19
Message-ID: edf80448-2c12-42c5-9d30-62fad91f0167@anayrat.info
Views: Raw Message | Whole Thread | Download mbox | Resend email
Thread:
Lists: pgsql-hackers

On 5/3/24 10:07 AM, Pavel Stehule wrote:
> Hi
>
> I found https://github.com/vnmakarov/mir?tab=readme-ov-file
> <https://github.com/vnmakarov/mir?tab=readme-ov-file>
>
> It should be significantly faster than llvm (compilation).
>
> Regards
>
> Pavel

Hello,

I can't tell about JIT, it is too complicated for me ^^

Fwiw, Pierre Ducroquet wrote an article about another (faster)
implementation:
https://www.pinaraf.info/2024/03/look-ma-i-wrote-a-new-jit-compiler-for-postgresql/

Xing Guo replied to his article by mentioning another two prototypes :

https://github.com/higuoxing/pg_slowjit
https://github.com/higuoxing/pg_asmjit

It is good to see interest in this area. Current implementation is quite
expensive and costing model risky. If often recommend disabling it and
enabling it when it is really needed. (It is disabled by default on AWS).

Regards,

--
Adrien NAYRAT

In response to

Browse pgsql-hackers by date

  From Date Subject
Next Message Ashutosh Bapat 2024-05-03 09:32:52 Re: tablecmds.c/MergeAttributes() cleanup
Previous Message Peter Eisentraut 2024-05-03 09:17:35 Re: tablecmds.c/MergeAttributes() cleanup