Re: [PATCH v2 3/3] mm: rmap: support batched unmapping for file large folios

[Lorenzo Stoakes <lorenzo.stoakes@oracle.com>]

On Thu, Dec 11, 2025 at 04:16:56PM +0800, Baolin Wang wrote:
> Similar to folio_referenced_one(), we can apply batched unmapping for file
> large folios to optimize the performance of file folios reclamation.
>
> Performance testing:
> Allocate 10G clean file-backed folios by mmap() in a memory cgroup, and try to
> reclaim 8G file-backed folios via the memory.reclaim interface. I can observe
> 75% performance improvement on my Arm64 32-core server.

Again, you must test on non-arm64 architectures and report the numbers for this
also.

>
> W/o patch:
> real    0m1.018s
> user    0m0.000s
> sys     0m1.018s
>
> W/ patch:
> real	0m0.249s
> user	0m0.000s
> sys	0m0.249s
>
> Signed-off-by: Baolin Wang <baolin.wang@linux.alibaba.com>
> ---
>  mm/rmap.c | 7 ++++---
>  1 file changed, 4 insertions(+), 3 deletions(-)
>
> diff --git a/mm/rmap.c b/mm/rmap.c
> index ec232165c47d..4c9d5777c8da 100644
> --- a/mm/rmap.c
> +++ b/mm/rmap.c
> @@ -1855,9 +1855,10 @@ static inline unsigned int folio_unmap_pte_batch(struct folio *folio,
>  	end_addr = pmd_addr_end(addr, vma->vm_end);
>  	max_nr = (end_addr - addr) >> PAGE_SHIFT;
>
> -	/* We only support lazyfree batching for now ... */
> -	if (!folio_test_anon(folio) || folio_test_swapbacked(folio))
> +	/* We only support lazyfree or file folios batching for now ... */
> +	if (folio_test_anon(folio) && folio_test_swapbacked(folio))

Why is it now ok to support file-backed batched unmapping when it wasn't in
Barry's series (see [0])? You don't seem to be justifying this?

[0]:https://lore.kernel.org/all/20250214093015.51024-4-21cnbao@gmail.com/T/#u


>  		return 1;
> +
>  	if (pte_unused(pte))
>  		return 1;
>
> @@ -2223,7 +2224,7 @@ static bool try_to_unmap_one(struct folio *folio, struct vm_area_struct *vma,
>  			 *
>  			 * See Documentation/mm/mmu_notifier.rst
>  			 */
> -			dec_mm_counter(mm, mm_counter_file(folio));
> +			add_mm_counter(mm, mm_counter_file(folio), -nr_pages);

Was this just a bug before?

>  		}
>  discard:
>  		if (unlikely(folio_test_hugetlb(folio))) {
> --
> 2.47.3
>

Thanks, Lorenzo