Re: [PATCh] x86: overmapped fix when 4K pages on tail - 64bit

[Yinghai Lu]

On Sun, Jul 13, 2008 at 7:57 AM, Andi Kleen <andi@xxxxxxxxxxxxxx> wrote:
> Arjan van de Ven <arjan@xxxxxxxxxxxxx> writes:
>
>> On Wed, 9 Jul 2008 20:15:02 -0700
>> Yinghai Lu <yhlu.kernel@xxxxxxxxx> wrote:
>>
>>>
>>>
>>> fix phys_pmd_init to make sure not to return big value than end.
>>>
>>> also print out range split:1G/2M/4K
>>>
>>> Signed-off-by: Yinghai Lu <yhlu.kernel@xxxxxxxxx>
>>
>>
>> Making an accurate mapping solves a lot of potentially nasty/tricky
>> corner cases, so I like the approach
>
> It also means that e.g. on a 1GB system the direct mapping will
> never use 1GB pages.
>
> And the CPU has to handle this anyways because all the old
> kernels overmap and no x86 CPU can drop support for all old
> kernels.
>
> In the end it means only large systems will benefit from 1GB
> pages, which seems wrong to me.

with this patch, my 256g system still use gbpages for 1g-3g, 4g-256g

YH
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/