The attachments to his entry is a patch to enable C3 power state on a Dell Inspiron 8200, dmesg and dmidecode output, as well as acpidump from the same machine.
Created attachment 11918 [details] patch to enable C3
Created attachment 11919 [details] dmesg
Created attachment 11920 [details] dmidecode
Created attachment 11921 [details] acpidump
BIOS description (all revisions) for the machine in question: http://www.bay-wolf.com/bios/I8200A11.TXT Patch written by Arjan van de Ven. Tested-by: Dag Bakke <dag@bakke.com>
Please submit future patches via email,thanks.
Who at Dell has stated that they agree this is a Dell BIOS bug? How do we know that Windows sees C3 on this box -- perfmon? If this is a known BIOS bug, why has Dell not fixed it?
I am sorry for not editing out the relevant comment from the patch before submitting it. I never meant to expressly state that it is a bug in the BIOS. I can only offer my observation, limited to a sample of one machine with a specific BIOS version and CPU model. It WFM. I do have a vague recollection that you (Len) or someone else with the ACPI-hat on has said something about that other OS accepting stuff which violates the ACPI spec.
Created attachment 14492 [details] updated patch for 2.6.24-rc8 updated patch for 2.6.24-rc8
Dell replied off-line that they disabled C3 on this system on purpose because of its chipset. If you are able to measure a power savings by 2nd guessing Dell, without any stability problems, then please document that here and re-open this sighting.