Re: Re: memory barriers (was: Yes, WaitLatch is vulnerable to weak-memory-ordering bugs)

Robert Haas <robertmhaas@gmail.com>

From: Robert Haas <robertmhaas@gmail.com>

To: Martijn van Oosterhout <kleptog@svana.org>

Cc: Greg Stark <stark@mit.edu>, Jeff Davis <pgsql@j-davis.com>, Thom Brown <thom@linux.com>, Peter Geoghegan <peter@2ndquadrant.com>, Tom Lane <tgl@sss.pgh.pa.us>, pgsql-hackers@postgresql.org

Date: 2011-09-24T16:46:48Z

Lists: pgsql-hackers

On Sat, Sep 24, 2011 at 9:45 AM, Martijn van Oosterhout
<kleptog@svana.org> wrote:
> I think memory accesses are also fantastically expensive, so it's worth
> some effort to optimise that.

This is definitely true.

> I found the Linux kernel document on this topic quite readable. I think
> the main lesson here is that processors track data dependancies (other
> than the Alpha apparently), but not control dependancies.  So in the
> example, the value of i is dependant on num_items, but not via any
> calculation.  IThat control dependancies are not tracked makes some
> sense, since branches depend on flags bit, and just about any
> calculation changes the flag bits, but most of the time these changes
> are not used.

Oh, that's interesting.  So that implies that a read-barrier would be
needed here even on non-Alpha.

-- 
Robert Haas
EnterpriseDB: http://www.enterprisedb.com
The Enterprise PostgreSQL Company