On Mon, Mar 7, 2011 at 4:43 PM, Carl-Daniel Hailfinger <c-d.hailfinger.devel.2006@gmx.net> wrote:
Hi David,

Auf 08.03.2011 00:18, David Hendricks schrieb:
> tested read and write operations on a Cr48.
>

While you have that Cr-48 nearby, could you please run
lspci -nnvvvxxx
superiotool -deV
dmidecode

Attached.

The SMBIOS stuff is a bit tricky. I ran into dmidecode issues due to some EFI weirdness (we discussed this on IRC a while back), so instead I dumped the raw SMBIOS tables from memory and faked the entry point structure which is prepended in smbios_raw.bin. Newer versions of dmidecode (I'm using 2.11) have a --from-dump option you may use to get the fully decoded output from the binary file. As you might expect, about the only useful info is the type 1 table's manufacturer (IEC) and product name (Mario). The BIOS on this particular machine is non-production, so the serial numbers and stuff will look weird.

On Mon, Mar 7, 2011 at 4:43 PM, Carl-Daniel Hailfinger <c-d.hailfinger.devel.2006@gmx.net> wrote: 
Is there an EFI equivalent of dmidecode? If yes, please run that as well
in verbose mode.

dmidecode works with EFI, however there is an interesting constraint -- the efi_vars kernel module will explode if the BIOS is 64-bit and the kernel is 32-bit (more details here: http://code.google.com/p/chromium-os/issues/detail?id=3146)

--
David Hendricks (dhendrix)
Systems Software Engineer, Google Inc.