16.11.2012 6.35 "Stefan Tauner" stefan.tauner@student.tuwien.ac.at kirjoitti:
On Thu, 15 Nov 2012 20:31:18 +0200 Ville Herva vherva@gmail.com wrote:
Ripping the battery off for a good while helped, and I was able to
upgrade
the BIOS at POST.
usually there is also a jumper (CMOS or NVRAM reset) that does the same.
Usually, yes, but according to the manual, the jumper only clears the bios passwords. There was a faq entry at the Intel site recommending ripping the battery off to clear the bios settings. Given that, i decided to rip the battery as the first measure as it was just as easy as fiddling with the jumper.
Anyway, it would be cool to be able to flash BIOS from Linux.
that is (or was) the goal of course, but intel (and microsoft) try to block access to the flash chip more and more to have more control over what the user can run. they try to sell this as a security feature, but that is only a part of it. if it would be about security only, then the user would be in control. in the secure boot scheme on x86 machines the user IS in control about the boot loader certificates, but this is only out of fear of competition laws IMHO.
ATM it IS possible to flash the BIOS region, even if other regions are locked. we just do not know if it is safe in all cases or if the management engine interferes (as this is not documented anywhere publicly) and hence give the warnings. in the future (i.e. starting with the haswell family of processors) flashrom will most probably be crippled even more on these platforms according to recently leaked information.
so... nag intel please, not us ;) see also
http://blogs.coreboot.org/blog/2011/08/17/gsoc-2011-flashrom-part-5-%E2%80%9...
-- Kind regards/Mit freundlichen Grüßen, Stefan Tauner
Yep, interesting. Sadly not all development is going towards openness. And probably never will as long as there's money involved.