Re: [v2 PATCH] arm64: mm: show direct mapping use in /proc/meminfo

[Yang Shi <yang@os.amperecomputing.com>]

On 11/19/25 1:10 PM, Catalin Marinas wrote:
> On Wed, Nov 19, 2025 at 11:37:13AM -0800, Yang Shi wrote:
>> On 11/19/25 11:12 AM, Catalin Marinas wrote:
>>> From: Catalin Marinas <catalin.marinas@arm.com>
>>>
>>> On Thu, 23 Oct 2025 14:52:10 -0700, Yang Shi wrote:
>>>> Since commit a166563e7ec3 ("arm64: mm: support large block mapping when
>>>> rodata=full"), the direct mapping may be split on some machines instead
>>>> keeping static since boot. It makes more sense to show the direct mapping
>>>> use in /proc/meminfo than before.
>>>> This patch will make /proc/meminfo show the direct mapping use like the
>>>> below (4K base page size):
>>>> DirectMap4K:	   94792 kB
>>>> DirectMap64K:	  134208 kB
>>>> DirectMap2M:	 1173504 kB
>>>> DirectMap32M:	 5636096 kB
>>>> DirectMap1G:	529530880 kB
>>>>
>>>> [...]
>>> It matches Documentation/filesystems/proc.rst, so I'm fine to align
>>> arm64 with it.
>> Yes, it does.
>>
>>> Applied to arm64 (for-next/misc), thanks!
>>>
>>> [1/1] arm64: mm: show direct mapping use in /proc/meminfo
>>>         https://git.kernel.org/arm64/c/1102778cb023
>> Thanks for taking the patch. However, Ryan noticed some over-accounting
>> problems and had some suggestions to code cleanup as well. I'm going to
>> submit v3 to the mailing list soon.
> Ah, yes, I forgot about that. If the change is small, can you submit a
> fix instead? I tend not to rebase the for-next/* branches. Otherwise I
> can revert and apply a new one.

The fix is basically simple. It just added a check about whether the 
address is linear map address or not, and the add and sub operations are 
encapsulated in helpers, however we need pass in the "addr" in split 
functions. Anyway I'm going to send the fix.

Thanks,
Yang

>
> Thanks.
>