Re: Improving spin-lock implementation on ARM. - Mailing list pgsql-hackers

From Alexander Korotkov
Subject Re: Improving spin-lock implementation on ARM.
Date
Msg-id CAPpHfdtOHKWZC5U3J5tV-Gs6g7LFZuj1POiOhg8Csj818HQX2g@mail.gmail.com
Whole thread Raw
In response to Re: Improving spin-lock implementation on ARM.  (Krunal Bauskar <krunalbauskar@gmail.com>)
List pgsql-hackers
On Mon, Nov 30, 2020 at 9:20 AM Krunal Bauskar <krunalbauskar@gmail.com> wrote:
> Some of us may be surprised by the fact that enabling lse is causing regression (1816 -> 892 or 714 -> 610) with HEAD
itself.
> While lse is meant to improve the performance. This, unfortunately, is not always the case at-least based on my
previousexperience with LSE.too.
 

I doubt this is correct.  As Tom shown upthread, Apple clang have LSE
enabled by default [1].  It might happen that --CFLAGS="-O2
-march=armv8-a+lse" disables some other optimizations, which are
enabled by default in Apple clang...

Links
1. https://www.postgresql.org/message-id/663376.1606432828%40sss.pgh.pa.us

------
Regards,
Alexander Korotkov



pgsql-hackers by date:

Previous
From: Yulin PEI
Date:
Subject: [PATCH] BUG FIX: Core dump could happen when VACUUM FULL in standalone mode
Next
From: Pavel Stehule
Date:
Subject: Re: [patch] [doc] Minor variable related cleanup and rewording of plpgsql docs