[coreboot] flashrom issues??

Carl-Daniel Hailfinger c-d.hailfinger.devel.2006 at gmx.net
Sat Mar 1 16:56:15 CET 2008


On 01.03.2008 16:50, Peter Stuge wrote:
> On Sat, Mar 01, 2008 at 08:04:54AM -0500, joe at smittys.pointclark.net wrote:
>   
>> Little confused about SMM? In the nothbridge datasheet it talks
>> about SMM (System Management Space), I do not have any of these
>> registers set, just defaults. In the southbridge datasheet it talks
>> about SMM (special mask mode), I do not have any of these registers
>> set eithor, just defaults. Are these related, or two different
>> things? Do I need to set these registers up?
>>     
>
> They are different things.
>
> SMM when talking about flashing refers to System Management Mode
> which is a special task context supported by mobile-only x86 since
> mobile 386SL and later also non-mobile x86.
>
> http://en.wikipedia.org/wiki/System_Management_Mode
>
> SMM is entered by triggering an SMI (System Management Interrupt)
> which can be done either by electrical signal to a dedicated pin on
> the CPU, or by software traps that are defined when initializing SMM.
>
> Software traps can be set on IO and memory access.
>
> This is the technology used by VSA to emulate legacy hardware on
> Geodes.
>
> When it comes to flashing, accessing the flash chip as usual causes a
> software trap and the code in SMM (put in place by the factory BIOS
> during boot) will determine whether to allow this particular access
> to continue onto real hardware or just fail and return 0xff.
>
> There have been some stories on the list about what the SMM code does
> to make it's decision, checksums in reserved memory area and whatnot.
>   

That's why I want to know if this also happens under coreboot. If so, we 
can rule out that sort of SMM because coreboot does not use SMM yet.

Regards,
Carl-Daniel

-- 
http://www.hailfinger.org/





More information about the coreboot mailing list