Re: [PATCH v1 4/4] mm: remove boolean output parameters from folio_pte_batch_ext()

[Lance Yang <lance.yang@linux.dev>]

On 2025/6/27 23:11, David Hildenbrand wrote:
> On 27.06.25 16:34, Lance Yang wrote:
>> On Fri, Jun 27, 2025 at 7:55 PM David Hildenbrand <david@redhat.com> 
>> wrote:
>>>
>>> Instead, let's just allow for specifying through flags whether we want
>>> to have bits merged into the original PTE.
>>>
>>> For the madvise() case, simplify by having only a single parameter for
>>> merging young+dirty. For madvise_cold_or_pageout_pte_range() merging the
>>> dirty bit is not required, but also not harmful. This code is not that
>>> performance critical after all to really force all micro-optimizations.
>>
>> IIRC, this work you've wanted to do for a long time - maybe even a 
>> year ago?
> 
> Heh, maybe, I don't remember.
> 
> For this user here, I realized that we might already check the existence 
> of any_dirty at runtime -- because the compiler will not necessarily 
> create two optimized functions.

Ah, I see! That's a very sharp observation about the compiler's behavior ;)

> 
> So we already have runtime checks ... instead of checking whether 
> any_dirty == NULL, we now simply do the merging (checking for 
> pte_young() instead) now.

Thanks for the lesson!
Lance