Re: [RFC PATCH 1/7] Alchemy: remove cpu_table.

[Sergei Shtylyov <sshtylyov@ru.mvista.com>]

Hello.

Manuel Lauss wrote:

>>> diff --git a/arch/mips/mm/c-r4k.c b/arch/mips/mm/c-r4k.c
>>> index 2709675..bb22649 100644
>>> --- a/arch/mips/mm/c-r4k.c
>>> +++ b/arch/mips/mm/c-r4k.c
>>>   
>>>       
>> [...]
>>
>>     
>>> @@ -1287,20 +1262,18 @@ static void __cpuinit coherency_setup(void)
>>>  	case CPU_R4400MC:
>>>  		clear_c0_config(CONF_CU);
>>>  		break;
>>> -	/*
>>> -	 * We need to catch the early Alchemy SOCs with
>>> -	 * the write-only co_config.od bit and set it back to one...
>>> -	 */
>>> -	case CPU_AU1000: /* rev. DA, HA, HB */
>>> -	case CPU_AU1100: /* rev. AB, BA, BC ?? */
>>> -	case CPU_AU1500: /* rev. AB */
>>> -		au1x00_fixup_config_od();
>>> -		break;
>>>   	case PRID_IMP_PR4450:
>>>  		nxp_pr4450_fixup_config();
>>>  		break;
>>>  	}
>>> +
>>> +#ifdef CONFIG_MACH_ALCHEMY
>>> +	{
>>> +		extern void au1x00_fixup_config_od(void);
>>> +		au1x00_fixup_config_od();
>>> +	}
>>> +#endif
>>>  }
>>>   
>>>       
>>   I've been thru this before. Ralf will hardly accept #ifdef'ery and extern 
>> in this file. That's why we have what we have now. :-)
>>     
>
> I believe it can be removed entirely from c-r4k.c, the Alchemy
> plat_mem_setup and resume code run it too (and earlier).
>   

   You cleraly didn't get it. ;-)
   It's here because what is written earlier was *discarded* while doing 
the coherency setup (remember, the OD bit is write only, so you can't 
just RMW the Config reg).
That lead to some crap with USB on Au1500 in particular, IIRC.

> Thanks for having a look!
> 	Manuel Lauss
>   

WBR, Sergei