[U-Boot] [PATCH V2 03/10] 83xx/85xx/86xx: LBC register cleanup

[Becky Bruce]

On Jun 8, 2010, at 1:43 PM, Kim Phillips wrote:

> On Tue, 8 Jun 2010 12:28:48 -0500
> Becky Bruce <beckyb at kernel.crashing.org> wrote:
>
>>
>> On Jun 8, 2010, at 12:09 PM, Kim Phillips wrote:
>>
>>> On Tue, 8 Jun 2010 11:15:07 -0500
>>> Becky Bruce <beckyb at kernel.crashing.org> wrote:
>>>
>>>> This isn't *ELBC*, it's the existence of any sort of LBC.  ELBC is
>>>> determined per-board in the fine grained configs.  There are 2
>>>> configs.
>>>
>>> that's what I was suggesting to change; to group and centralize the
>>> definition of the two seemingly related configs - also because the
>>> ELBC
>>> should be done at the SoC level.
>>>
>>> See Scott's response though - I don't know how long the ELBC ifdef
>>> list
>>> is, but introducing a header file per soc sounds like a bit out of
>>> scope for this patchseries - it already cleans up a bunch of stuff.
>>
>> I think it is out of scope here - I think this is fine the way it is,
>> at least for now.  I'd like to get these cleanups in as soon as
>> possible before someone else replicates more code.
>>
>> We can post a followon later to change how we deal with ELBC/LBC;   
>> and
>> there are also more cleanups to be done - for example, there's still
>> code that accesses lbc space not using IO accessors.  I've cleaned up
>> the BR/OR cases, but I want to eventually fix it all.  Unfortunately,
>> I'm out of time at this point.  This patch series is a significant
>> step forward, but there's still more work to be done.
>
> ok but e.g., patch 1 in this series omits the MPC837XERDB, which needs
> to be fixed.
>
> I was curious so I compiled the ELBC CONFIG_ list, in case you want to
> change patch 1 in the series to _delete_ the entries in the board
> configs and add them as pointed out in this patch:
>
> MPC831x
> MPC837x
> MPC8536
> MPC8569
> MPC8572
>
> and, of course, the Pxxxx's.

OK, thanks, I'll try to look at it probably tonight.

-Becky
>
>> Cheers,
>> Becky
>
> Kim
>
>>
>>>
>>>> -B
>>>
>>> Kim
>>>
>>> p.s. please don't top-post :).
>>>
>>>> On Jun 3, 2010, at 7:37 PM, Kim Phillips wrote:
>>>>
>>>>> On Wed, 2 Jun 2010 13:12:02 -0500
>>>>> Becky Bruce <beckyb at kernel.crashing.org> wrote:
>>>>>
>>>>>> diff --git a/arch/powerpc/include/asm/config.h b/arch/powerpc/
>>>>>> include/asm/config.h
>>>>>> index fc3facb..01036f3 100644
>>>>>> --- a/arch/powerpc/include/asm/config.h
>>>>>> +++ b/arch/powerpc/include/asm/config.h
>>>>>> @@ -76,4 +76,10 @@
>>>>>> /* Relocation to SDRAM works on all PPC boards */
>>>>>> #define CONFIG_RELOC_FIXUP_WORKS
>>>>>>
>>>>>> +/* Most PPC SOCs have a LBC so define this here */
>>>>>> +#if defined(CONFIG_MPC85xx) || defined(CONFIG_MPC86xx) || \
>>>>>> +	defined(CONFIG_MPC83xx)
>>>>>> +#define CONFIG_FSL_LBC
>>>>>
>>>>> Instead of in the configs (such as what patch 1 in this series
>>>>> continues to do), can we do ELBC determination here, based on
>>>>> finer-grained configs such as 831x, 837x, etc?
>>>>>
>>>>>> +#endif
>>>>>
>>>>> Thanks,
>>>>>
>>>>> Kim
>>>>
>>>>
>>
>>