[PATCH][RFT] please use this one
kumar.gala at freescale.com
Sat Oct 22 04:34:01 EST 2005
On Oct 21, 2005, at 1:43 PM, Rupert Eibauer wrote:
> On Friday 21 October 2005 17:23, Kumar Gala wrote:
>> On Oct 21, 2005, at 10:11 AM, Rupert Eibauer wrote:
>>> On Friday 21 October 2005 16:28, Kumar Gala wrote:
>>>> The concept of PPC_HIGH_BATS is already some what handled with
>>>> CPU_FTR_HAS_HIGH_BATS. I recommend you use this feature instead of
>>>> introducing another one as well as a compile time option.
>>> CPU_FTR_HAS_HIGH_BATS only clears the high bats, but does
>>> not attempt to use them. My patch will use them.
>> I understand, I'm suggesting you expand CPU_FTR_HAS_HIGH_BATS to
>> include your additional functionality.
> How do you mean that exactly?
> Do you mean making the feature always-on (not selectable in Kconfig)?
> Having CONFIG_PPC_LARGE_BATS and CONFIG_HIGH_PATS always-on would
> be ok for me, but for PHYS_64BIT should be left up to the platform
> to decide if it should be presented to the user or not.
Yes, I think just making support for BATS 4-7 always on and use
CONFIG_FTR_HAS_HIGH_BATS for any runtime fixup.
> Do you think the other two features, CPU_FTR_BIG_PHYS and
> CPU_FTR_LARGE_BATS should be merged into one bit? They could for now,
> because there is no processor supporting only one of them.
Yes, this should still be a config option. Merging
CPU_FTR_LARGE_BATS into CPU_FTR_BIG_PHYS should be ok.
More information about the Linuxppc-dev