[coreboot] latest cumulative kontron patch

[Stefan Reinauer]

Peter Stuge wrote:
>> +++ southbridge/intel/i82801gx/lpc.c	(working copy)
>> @@ -128,6 +128,9 @@
>>  {
>>  	u8 reg8;
>>  	u16 reg16;
>> +#ifndef MAINBOARD_POWER_ON_AFTER_POWER_FAIL
>> +#define MAINBOARD_POWER_ON_AFTER_POWER_FAIL 1
>> +#endif
>>  
>>  	int pwr_on=MAINBOARD_POWER_ON_AFTER_POWER_FAIL;
>>  	int nmi_option;
>>     
>
> Could this go somewhere outside the function? Eventually I'd like
> this particular setting to go into Kconfig and of course be
> changeable at runtime. :)
>   
As you probably have seen before writing objections, it _is_ changeable
at runtime. ;-)

>> +static const u32 dq2030[] = {
>> +	0x08070706, 0x0a090908, 0x0d0c0b0a, 0x12100f0e, 
>> +	0x1a181614, 0x22201e1c, 0x2a282624, 0x3934302d,
>> +	0x0a090908, 0x0c0b0b0a, 0x0e0d0d0c, 0x1211100f, 
>> +	0x19171513, 0x211f1d1b, 0x2d292623, 0x3f393531
>> +};
>>     
>
> These tables.. So ugly. sigh..
>   
I don't understand? Don't you like tables? It's a common way to store
information in programming languages. We could pack that information in
a lar file, but it would not make the overall code flow significantly
easier to follow or more solid.