Re: [GSoC] aarch64 Optimization

From: Stephen Frost <sfrost(at)snowman(dot)net>
To: Artem Luzyanin <lisyonok85(at)yahoo(dot)ca>
Cc: "pgsql-students(at)postgresql(dot)org" <pgsql-students(at)postgresql(dot)org>
Subject: Re: [GSoC] aarch64 Optimization
Date: 2015-03-13 02:10:43
Message-ID: 20150313021043.GW29780@tamriel.snowman.net
Views: Raw Message | Whole Thread | Download mbox | Resend email
Thread:
Lists: pgsql-students

Artem,

* Artem Luzyanin (lisyonok85(at)yahoo(dot)ca) wrote:
>         My name is Artem Luzyanin, I am a student in a college, and I am currently taking a Software Portability and Optimization course. Our project for the course is to optimize a some part of a software for  aarch64 architecture, and I would really like to contribute to PostgreSQL community with that matter. I am not particularly interested to participate in GSoC program, as I am doing is specifically for the course, but I can apply for it as well, depending on the requirements. If I should be communicating through a different mailing list, please let me know!
>         In terms of my knowledge, I am proficient with both SQL and C, and I have written programs on Assembler for aarch64, so I strongly believe I am capable of trying my hands on some architecture-specific optimization. Having said that, I would really like to discuss with one of the community members about possible parts of the code that could be in need of such reworking. If no such parts are knows, I would like to see if I could find them myself, although it could take a substantial amount of time, so any input is welcome!

To the extent that this is independent from GSOC, I would suggest you
email the pgsql-hackers mailing list with your specific thoughts about
what optimizations you would add to PostgreSQL for aarch64. Note that
the PostgreSQL community is unlikely to accept optimziations which are
architecture-specific unless they are very clear, have very few
drawbacks, and are able to be maintained by the community.
Optimizations which slow down other, more common, platforms are very
unlikely to be accepted. Further, PostgreSQL absolutely must produce
the correct answer on all platforms and that may preclude the ability to
use certain platform-specific optimization as that could lead to
different results on different platforms.

Thanks!

Stephen

In response to

Browse pgsql-students by date

  From Date Subject
Next Message hitesh ramani 2015-03-13 15:03:24 Re: GSOC 2015: Student Introduction
Previous Message Stephen Frost 2015-03-13 02:06:35 Re: GSOC 2015: Student Introduction