Re: Re: [COMMITTERS] pgsql: Introduce group locking to prevent parallel processes from deadl - Mailing list pgsql-hackers

From Andres Freund
Subject Re: Re: [COMMITTERS] pgsql: Introduce group locking to prevent parallel processes from deadl
Date
Msg-id 20160217083255.GJ25464@awork2.anarazel.de
Whole thread Raw
In response to Re: Re: [COMMITTERS] pgsql: Introduce group locking to prevent parallel processes from deadl  (Tom Lane <tgl@sss.pgh.pa.us>)
List pgsql-hackers
On 2016-02-16 23:33:47 -0500, Tom Lane wrote:
> Yes, exactly.  I'm not certain if there are any real platforms where
> a pointer-sized write wouldn't be atomic (it sure sounds inefficient
> for that to be true), but we have not assumed that to date and I'd
> just as soon not start here.

FWIW, there's no sizeof(void*) == 8 platform supported by PG where
aligned 8 byte writes aren't atomic. There are a number of supported
platforms with sizeof(void*) == 4 without atomic 8 byte writes though.



pgsql-hackers by date:

Previous
From: Robert Haas
Date:
Subject: Re: PostgreSQL Audit Extension
Next
From: Kyotaro HORIGUCHI
Date:
Subject: Re: Identifying a message in emit_log_hook.