On Fri, 7 Nov 2008, Thor Lancelot Simon wrote:
On Fri, Nov 07, 2008 at 09:30:08AM -0800, Paul Goyette wrote:I've had a couple people suggest that maybe we ought to include the spdmem(4) in GENERIC kernels, at least for amd64 and i386. (It's already included in macppc, IIRC.) The driver doesn't do much, and other than providing EEPROM data via a sysctl(8) variable, it does nothing once the machine is booted.It lets you see how many memory sockets are filled. That's a big deal if you are trying to plan upgrades for a remote server. Can it probe empty sockets, by the way? I assume not.
It doesn't do any probing. For each instance that's configured, it attempts to read the EEPROM and verify the checksum (or CRC). If the checksum/CRC check fails, it doesn't attach.
If a socket is empty, or non-existent, the result is the same. It will read nothing from the EEPROM, and the check will fail. There's no way to differentiate between an empty socket and a non-existent socket.
---------------------------------------------------------------------- | Paul Goyette | PGP DSS Key fingerprint: | E-mail addresses: | | Customer Service | FA29 0E3B 35AF E8AE 6651 | paul%whooppee.com@localhost | | Network Engineer | 0786 F758 55DE 53BA 7731 | pgoyette%juniper.net@localhost | ----------------------------------------------------------------------