Subject: Re: PostgreSQL
To: Ignatios Souvatzis <>
From: Johnny Billquist <>
List: tech-net
Date: 02/02/2006 13:02:37
Um? Not entirely true.
You don't need locked bus cycles for atomic operations which actually 
are atomic on the bus.
A write is a write is a write. It can never be split. It's atomic. Can't 
be anything else.
However, a read-modify-write operation needs to be interlocked if it is 
to appear to be atomic.


Ignatios Souvatzis wrote:
> On Wed, Feb 01, 2006 at 06:58:04PM -0800, Alfred Perlstein wrote:
>>Also, I'm pretty sure that synchronization between threads doesn't
>>require locked bus cycles, only atomic ops, again another saving.
> If this was the case, I'd be very interested in your definition of
> "locked bus cycle" vs.  "atomic ops".
> In a multi-CPU (with non-shared cache) system, you have to implement
> atomic operations via some sort of locked bus cycles.
> This is independent of the thread vs. processes-with-a-shared-variable
> question.
> Regards,
> 	-is
> P.S.: Why, exactly, is this discussion on tech-net, among others?

Johnny Billquist                  || "I'm on a bus
                                   ||  on a psychedelic trip
email:           ||  Reading murder books
pdp is alive!                     ||  tryin' to stay hip" - B. Idol