Re: [PATCH v7 1/3] mm/swap: colocate page-cluster sysctl with swap readahead

[Kairui Song]

On Sun, Jul 5, 2026 at 10:14 PM Jianyue Wu <wujianyue000@xxxxxxxxx> wrote:
>
> page_cluster and the vm.page-cluster sysctl are only used by swap-in
> readahead in swap_state.c. Move them out of swap.c together with
> swap_readahead_setup(), and make page_cluster static to that file.
>
> Rename swap_setup() while moving it as well. The helper is internal to
> MM and now only sets up swap readahead defaults and its sysctl hook, so
> the more specific name matches its reduced scope.
>
> swap_setup() previously lived in mm/swap.c, which is built
> unconditionally, so the vm.page-cluster sysctl was registered also on
> CONFIG_SWAP=n kernels. swap_readahead_setup() is now a no-op stub when
> CONFIG_SWAP is disabled, so vm.page-cluster is no longer registered
> there. The knob only tunes swap-in readahead and had no effect without
> swap.
>
> Suggested-by: Baoquan He <bhe@xxxxxxxxxx>
> Suggested-by: Barry Song <baohua@xxxxxxxxxx>
> Suggested-by: Johannes Weiner <hannes@xxxxxxxxxxx>
> Acked-by: Johannes Weiner <hannes@xxxxxxxxxxx>
> Acked-by: David Hildenbrand (Arm) <david@xxxxxxxxxx>
> Signed-off-by: Jianyue Wu <wujianyue000@xxxxxxxxx>
> ---
>  include/linux/swap.h |  1 -
>  mm/swap.c            | 36 ------------------------------------
>  mm/swap.h            |  9 +++++++--
>  mm/swap_state.c      | 37 +++++++++++++++++++++++++++++++++++++
>  mm/vmscan.c          |  2 +-
>  5 files changed, 45 insertions(+), 40 deletions(-)
>
> diff --git a/include/linux/swap.h b/include/linux/swap.h
> index 3f31b6a56788..45656ca9792d 100644
> --- a/include/linux/swap.h
> +++ b/include/linux/swap.h
> @@ -340,7 +340,6 @@ extern void lru_add_drain_cpu_zone(struct zone *zone);
>  extern void lru_add_drain_all(void);
>  void folio_deactivate(struct folio *folio);
>  void folio_mark_lazyfree(struct folio *folio);
> -extern void swap_setup(void);
>
>  /* linux/mm/vmscan.c */
>  extern unsigned long zone_reclaimable_pages(struct zone *zone);
> diff --git a/mm/swap.c b/mm/swap.c
> index 58e4eff698cc..b3ea30edbaad 100644
> --- a/mm/swap.c
> +++ b/mm/swap.c
> @@ -43,10 +43,6 @@
>  #define CREATE_TRACE_POINTS
>  #include <trace/events/pagemap.h>
>
> -/* How many pages do we try to swap or page in/out together? As a power of 2 */
> -int page_cluster;
> -static const int page_cluster_max = 31;
> -
>  struct cpu_fbatches {
>         /*
>          * The following folio batches are grouped together because they are protected
> @@ -1176,35 +1172,3 @@ void lru_reparent_memcg(struct mem_cgroup *memcg, struct mem_cgroup *parent, int
>                 lruvec_reparent_lru(child_lruvec, parent_lruvec, lru, nid);
>  }
>  #endif
> -
> -static const struct ctl_table swap_sysctl_table[] = {
> -       {
> -               .procname       = "page-cluster",
> -               .data           = &page_cluster,
> -               .maxlen         = sizeof(int),
> -               .mode           = 0644,
> -               .proc_handler   = proc_dointvec_minmax,
> -               .extra1         = SYSCTL_ZERO,
> -               .extra2         = (void *)&page_cluster_max,
> -       }
> -};
> -
> -/*
> - * Perform any setup for the swap system
> - */
> -void __init swap_setup(void)
> -{
> -       unsigned long megs = PAGES_TO_MB(totalram_pages());
> -
> -       /* Use a smaller cluster for small-memory machines */
> -       if (megs < 16)
> -               page_cluster = 2;
> -       else
> -               page_cluster = 3;
> -       /*
> -        * Right now other parts of the system means that we
> -        * _really_ don't want to cluster much more
> -        */
> -
> -       register_sysctl_init("vm", swap_sysctl_table);
> -}
> diff --git a/mm/swap.h b/mm/swap.h
> index 44ab8e1e595b..624ea8ff3d0d 100644
> --- a/mm/swap.h
> +++ b/mm/swap.h
> @@ -8,8 +8,6 @@ struct mempolicy;
>  struct swap_iocb;
>  struct swap_memcg_table;
>
> -extern int page_cluster;
> -
>  #if defined(MAX_POSSIBLE_PHYSMEM_BITS)
>  #define SWAP_CACHE_PFN_BITS (MAX_POSSIBLE_PHYSMEM_BITS - PAGE_SHIFT)
>  #elif defined(MAX_PHYSMEM_BITS)
> @@ -79,6 +77,8 @@ enum swap_cluster_flags {
>  };
>
>  #ifdef CONFIG_SWAP
> +void swap_readahead_setup(void);
> +
>  #include <linux/swapops.h> /* for swp_offset */
>  #include <linux/blk_types.h> /* for bio_end_io_t */
>
> @@ -336,6 +336,11 @@ static inline unsigned int folio_swap_flags(struct folio *folio)
>
>  #else /* CONFIG_SWAP */
>  struct swap_iocb;
> +
> +static inline void swap_readahead_setup(void)
> +{
> +}
> +
>  static inline struct swap_cluster_info *swap_cluster_lock(
>         struct swap_info_struct *si, pgoff_t offset, bool irq)
>  {
> diff --git a/mm/swap_state.c b/mm/swap_state.c
> index 6fd6e3415b71..a9f5b944172b 100644
> --- a/mm/swap_state.c
> +++ b/mm/swap_state.c
> @@ -22,10 +22,15 @@
>  #include <linux/vmalloc.h>
>  #include <linux/huge_mm.h>
>  #include <linux/shmem_fs.h>
> +#include <linux/sysctl.h>
>  #include "internal.h"
>  #include "swap_table.h"
>  #include "swap.h"
>
> +/* Swap readahead cluster size, as a power of 2 pages. */
> +static int page_cluster;
> +static const int page_cluster_max = 31;
> +
>  /*
>   * swapper_space is a fiction, retained to simplify the path through
>   * vmscan's shrink_folio_list.
> @@ -985,6 +990,38 @@ struct folio *swapin_readahead(swp_entry_t entry, gfp_t gfp_mask,
>         return folio;
>  }
>
> +static const struct ctl_table swap_readahead_sysctl_table[] = {
> +       {
> +               .procname       = "page-cluster",
> +               .data           = &page_cluster,
> +               .maxlen         = sizeof(int),
> +               .mode           = 0644,
> +               .proc_handler   = proc_dointvec_minmax,
> +               .extra1         = SYSCTL_ZERO,
> +               .extra2         = (void *)&page_cluster_max,
> +       }
> +};
> +
> +/**
> + * swap_readahead_setup - defaults and sysctl for swap cache readahead clustering
> + */
> +void __init swap_readahead_setup(void)
> +{
> +       unsigned long megs = PAGES_TO_MB(totalram_pages());
> +
> +       /* Use a smaller cluster for small-memory machines */
> +       if (megs < 16)
> +               page_cluster = 2;
> +       else
> +               page_cluster = 3;
> +       /*
> +        * Right now other parts of the system means that we
> +        * _really_ don't want to cluster much more
> +        */
> +
> +       register_sysctl_init("vm", swap_readahead_sysctl_table);
> +}
> +

Thanks, I also think they belong here.

>  #ifdef CONFIG_SYSFS
>  static ssize_t vma_ra_enabled_show(struct kobject *kobj,
>                                      struct kobj_attribute *attr, char *buf)
> diff --git a/mm/vmscan.c b/mm/vmscan.c
> index f40cfe9d703b..ae710027d546 100644
> --- a/mm/vmscan.c
> +++ b/mm/vmscan.c
> @@ -7646,7 +7646,7 @@ static int __init kswapd_init(void)
>  {
>         int nid;
>
> -       swap_setup();
> +       swap_readahead_setup();

There is already a subsys_initcall(swap_init); in swap_state.c so
perhaps calling this in that init function is even cleaner. And that
init hook runs earlier than kswapd init so totally fine I think.

Reviewed-by: Kairui Song <kasong@xxxxxxxxxxx>