Well! I'll be DA**ED! :/
I too have a crippled AMI BIOS on my beloved MSI P6N-SLI-FI.
I just rebuilt my 2.6.26 kernel with a built-in (feature of newer kernels-- no need for initrd) modified DSDT, and I got quite a pleasant surprise during boot!
Dmesg shows a somewhat different story than it did before I included the new DSDT, and though things are surely not PERFECT, I can see already that things are a bit better
For one, my LILO setting of ACPI_OSI=Linux is now reported to be 'honoured by BIOS..' and then all my motherboard/platform DMI information being produced properly rather than all being blank/missing.
Also, I get a number of positive ACPI messages telling me to "Report the above information to
acpi-whatever@vger.org", though I doubt that is really valid considering the firmware is not stock now (technically).
EDIT - As well, I always (ALWAYS) got a forboding message telling me that my system or BIOS was "..not E820 reserved", and that is gone now too.
As much as I (???) love my MSI stuff, I am now rather miffed.
I read the thread given by the poster above, as well as the info on the lesswatts.org site (the linux acpi site) and it all looks to be true. We need more testers, and more complainers!!
I WAS going to make a new post somewhere here with comparable Dmesg outputs from PRE-dsdt-mod boot and POST-dsdt-mod boot, so we can see the differences. However I have decided not to consume the space, as probably anyone can do the same, and see for them selves.
Board: MSI P6N-SLI-FI (MS7350-010)
BIOS: AMI v2.5 (Feb/Mar 2008)
Sasha
EDIT: RE the next post-- I won't go so far as to call it sabotage, however I do feel that MSI, Foxconn, and AMI, individually or as a group, are too experienced to be incompetent. Plus, why go to the trouble of setting up resources in a BIOS (for linux to use) and then writing the BIOS code in such a way that the resources go UNused by that OS? Weird...