cvs commit: src/sys/amd64/conf GENERIC

[David O'Brien]

On Wed, Dec 13, 2006 at 08:13:24PM -0800, mjacob at freebsd.org wrote:
> Secondly, I would *like* to have SMP on as a default in RELENG_6 for 
> amd64, as it would avoid doing what I've done twice already- fresh 
> sysinstall may or may not have installed an SMP kernel but going off and 
> building GENERIC then lost me my SMP. It's sort a violation of POLA.

Rather close what happened to me and what lead to the commit.


On Wed, Dec 13, 2006 at 10:17:02PM -0700, Scott Long wrote:
> mjacob at freebsd.org wrote:
> > Scott Long wrote:
> > > There wasn't a full switchover to SMP at 6.0 because an SMP kernel
> > > on a UP system incurs a measurable runtime overhead, and we wanted
> > > to present a system that showed the best of FreeBSD to people who
> > > wanted to run it

Then why is sparc64's GENERIC kernel SMP in RELENG_6?


> > But David's point is that most AMD64 boxes *are* SMP, not UP.  Is
> > that wrong?
> 
> 1. There are plenty of single core Opterons and Athlon64 chips still
> in service.

A few points:

* FreeBSD is predominately a server OS, not home desktop OS.

* Practically all single core Opteron's are used in multi-socket systems.
There are very, very few Opteron single socket, single core Opteron's in
service as there are only a very small handful of single socket 940-pin
motherboards.

* Most Athlon64 single-core CPU's are run in 32-bit "i386" mode.
FreeBSD/i386 isn't the platform I (and presumably Matt) are talking
about.

> Maybe AMD sells more SMP systems now than UP systems, but their prior
> sales of UP systems didn't magically disappear overnight.

This is not an AMD-only situation.

Intel has been selling dual-core processors since 18 April 2005, and now
sells quad-core processors.  Also, Intel kept Long Model (i.e., EM64T)
to the server/workstation segment until its dual-core processors.  For
instance, most Intel laptops cannot run FreeBSD/AMD64.

-- 
-- David  (obrien at FreeBSD.org)