Re: [PATCH] audo-detect and use -moutline-atomics compilation flag for aarch64 - Mailing list pgsql-hackers

From Heikki Linnakangas
Subject Re: [PATCH] audo-detect and use -moutline-atomics compilation flag for aarch64
Date
Msg-id 5f048a5a-c471-50ae-30ad-385dcafc616b@iki.fi
Whole thread Raw
In response to Re: [PATCH] audo-detect and use -moutline-atomics compilation flag for aarch64  ("Zidenberg, Tsahi" <tsahee@amazon.com>)
List pgsql-hackers
On 30/09/2020 19:04, Zidenberg, Tsahi wrote:
> Ubuntu 20.04 even turned it off by default for gcc-10, which seems
> like a compatibility step with the main gcc-9 compiler.
Ok, I definitely don't want to override that decision.

>> I'm marking this as Rejected in the commitfest. But thanks for the
>> benchmarking, that is valuable information nevertheless.
> 
> Could additional data change your mind?

I doubt it. IMO we shouldn't second-guess decisions made by compiler and 
distribution vendors, and more performance data won't change that 
principle. For comparison, we also don't set -O or -funroll-loops or any 
other flags to enable/disable specific optimizations. (Except for a few 
specific source files, like checksum.c, but those are exceptions the rule.)

If some other committer thinks differently, I won't object, but I'm not 
going to commit this. Maybe you should speak to the distribution vendors 
or the folk packaging PostgreSQL for those distributions, instead.

- Heikki



pgsql-hackers by date:

Previous
From: Bruce Momjian
Date:
Subject: Re: BUG #16419: wrong parsing BC year in to_date() function
Next
From: Fujii Masao
Date:
Subject: Re: Retry Cached Remote Connections for postgres_fdw in case remote backend gets killed/goes away