From: | Tom Lane <tgl(at)sss(dot)pgh(dot)pa(dot)us> |
---|---|
To: | Robert Haas <robertmhaas(at)gmail(dot)com> |
Cc: | pgsql-committers(at)postgresql(dot)org |
Subject: | Re: pgsql: Fix a few problems in barrier.h. |
Date: | 2013-07-18 03:30:50 |
Message-ID: | 27977.1374118250@sss.pgh.pa.us |
Views: | Raw Message | Whole Thread | Download mbox | Resend email |
Thread: | |
Lists: | pgsql-committers |
Robert Haas <robertmhaas(at)gmail(dot)com> writes:
> According to http://en.wikipedia.org/wiki/Memory_ordering, PA-RISC has
> weak memory ordering.
BTW, so far as I can see, their reference for that claim is the two
articles by Paul McKenney; but what McKenney actually says is
Although the PA-RISC architecture permits full reordering of
loads and stores, actual CPUs run fully ordered. This means the
Linux kernel's memory-ordering primitives generate no code; they
do, however, use the GCC memory attribute to disable compiler
optimizations that would reorder code across the memory barrier.
So the committed patch duplicates what the kernel does (unless someone
cares to dig into the kernel and show otherwise).
regards, tom lane
From | Date | Subject | |
---|---|---|---|
Next Message | Hitoshi Harada | 2013-07-18 04:39:37 | Re: [COMMITTERS] pgsql: Add support for REFRESH MATERIALIZED VIEW CONCURRENTLY. |
Previous Message | Tom Lane | 2013-07-18 02:43:10 | Re: pgsql: Fix a few problems in barrier.h. |