mm: swap: add a adaptive full cluster cache reclaim

Link all full cluster with one full list, and reclaim from it when the
allocation have ran out of all usable clusters.

There are many reason a folio can end up being in the swap cache while
having no swap count reference.  So the best way to search for such slots
is still by iterating the swap clusters.

With the list as an LRU, iterating from the oldest cluster and keep them
rotating is a very doable and clean way to free up potentially not inuse
clusters.

When any allocation failure, try reclaim and rotate only one cluster.
This is adaptive for high order allocations they can tolerate fallback.
So this avoids latency, and give the full cluster list an fair chance to
get reclaimed.  It release the usage stress for the fallback order 0
allocation or following up high order allocation.

If the swap device is getting very full, reclaim more aggresively to
ensure no OOM will happen.  This ensures order 0 heavy workload won't go
OOM as order 0 won't fail if any cluster still have any space.

[ryncsn@gmail.com: fix discard of full cluster]
Link: https://lkml.kernel.org/r/CAMgjq7CWwK75_2Zi5P40K08pk9iqOcuWKL6khu=x4Yg_nXaQag@mail.gmail.com
Link: https://lkml.kernel.org/r/20240730-swap-allocator-v5-9-cb9c148b9297@kernel.org
Signed-off-by: Kairui Song <kasong@tencent.com>
Reported-by: Barry Song <21cnbao@gmail.com>
Cc: Chris Li <chrisl@kernel.org>
Cc: "Huang, Ying" <ying.huang@intel.com>
Cc: Hugh Dickins <hughd@google.com>
Cc: Kalesh Singh <kaleshsingh@google.com>
Cc: Ryan Roberts <ryan.roberts@arm.com>
Cc: David Hildenbrand <david@redhat.com>
Cc: Kairui Song <ryncsn@gmail.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

diff --git a/include/linux/swap.h b/include/linux/swap.h
index e372122e4f4599c53ec1c6a936e58f92a3a3692f..1c8f844a9f0f268bdb14ee79d4e9c93caeebf1e4 100644 (file)
--- a/include/linux/swap.h
+++ b/include/linux/swap.h
@@ -260,6 +260,7 @@ struct swap_cluster_info {
 #define CLUSTER_FLAG_FREE 1 /* This cluster is free */
 #define CLUSTER_FLAG_NONFULL 2 /* This cluster is on nonfull list */
 #define CLUSTER_FLAG_FRAG 4 /* This cluster is on nonfull list */
+#define CLUSTER_FLAG_FULL 8 /* This cluster is on full list */
 
 /*
  * The first page in the swap file is the swap header, which is always marked
@@ -297,6 +298,7 @@ struct swap_info_struct {
        unsigned char *swap_map;        /* vmalloc'ed array of usage counts */
        struct swap_cluster_info *cluster_info; /* cluster info. Only for SSD */
        struct list_head free_clusters; /* free clusters list */
+       struct list_head full_clusters; /* full clusters list */
        struct list_head nonfull_clusters[SWAP_NR_ORDERS];
                                        /* list of cluster that contains at least one free slot */
        struct list_head frag_clusters[SWAP_NR_ORDERS];

diff --git a/mm/swapfile.c b/mm/swapfile.c
index ed5f85504f37a22c583c3e0e475c87ef2ffb1d93..b36a16131696aed1526a791e1868403f15a69bdd 100644 (file)
--- a/mm/swapfile.c
+++ b/mm/swapfile.c
@@ -440,10 +440,7 @@ static void swap_cluster_schedule_discard(struct swap_info_struct *si,
                        SWAP_MAP_BAD, SWAPFILE_CLUSTER);
 
        VM_BUG_ON(ci->flags & CLUSTER_FLAG_FREE);
-       if (ci->flags & CLUSTER_FLAG_NONFULL)
-               list_move_tail(&ci->list, &si->discard_clusters);
-       else
-               list_add_tail(&ci->list, &si->discard_clusters);
+       list_move_tail(&ci->list, &si->discard_clusters);
        ci->flags = 0;
        schedule_work(&si->discard_work);
 }
@@ -453,7 +450,7 @@ static void __free_cluster(struct swap_info_struct *si, struct swap_cluster_info
        lockdep_assert_held(&si->lock);
        lockdep_assert_held(&ci->lock);
 
-       if (ci->flags & CLUSTER_FLAG_NONFULL)
+       if (ci->flags)
                list_move_tail(&ci->list, &si->free_clusters);
        else
                list_add_tail(&ci->list, &si->free_clusters);
@@ -480,7 +477,6 @@ static void swap_do_scheduled_discard(struct swap_info_struct *si)
                                SWAPFILE_CLUSTER);
 
                spin_lock(&si->lock);
-
                spin_lock(&ci->lock);
                __free_cluster(si, ci);
                memset(si->swap_map + idx * SWAPFILE_CLUSTER,
@@ -576,12 +572,9 @@ static void dec_cluster_info_page(struct swap_info_struct *p,
 
        if (!(ci->flags & CLUSTER_FLAG_NONFULL)) {
                VM_BUG_ON(ci->flags & CLUSTER_FLAG_FREE);
-               if (ci->flags & CLUSTER_FLAG_FRAG) {
+               if (ci->flags & CLUSTER_FLAG_FRAG)
                        p->frag_cluster_nr[ci->order]--;
-                       list_move_tail(&ci->list, &p->nonfull_clusters[ci->order]);
-               } else {
-                       list_add_tail(&ci->list, &p->nonfull_clusters[ci->order]);
-               }
+               list_move_tail(&ci->list, &p->nonfull_clusters[ci->order]);
                ci->flags = CLUSTER_FLAG_NONFULL;
        }
 }
@@ -674,8 +667,8 @@ static void cluster_alloc_range(struct swap_info_struct *si, struct swap_cluster
                          (CLUSTER_FLAG_FREE | CLUSTER_FLAG_NONFULL | CLUSTER_FLAG_FRAG)));
                if (ci->flags & CLUSTER_FLAG_FRAG)
                        si->frag_cluster_nr[ci->order]--;
-               list_del(&ci->list);
-               ci->flags = 0;
+               list_move_tail(&ci->list, &si->full_clusters);
+               ci->flags = CLUSTER_FLAG_FULL;
        }
 }
 
@@ -718,6 +711,46 @@ done:
        return offset;
 }
 
+static void swap_reclaim_full_clusters(struct swap_info_struct *si)
+{
+       long to_scan = 1;
+       unsigned long offset, end;
+       struct swap_cluster_info *ci;
+       unsigned char *map = si->swap_map;
+       int nr_reclaim, total_reclaimed = 0;
+
+       if (atomic_long_read(&nr_swap_pages) <= SWAPFILE_CLUSTER)
+               to_scan = si->inuse_pages / SWAPFILE_CLUSTER;
+
+       while (!list_empty(&si->full_clusters)) {
+               ci = list_first_entry(&si->full_clusters, struct swap_cluster_info, list);
+               list_move_tail(&ci->list, &si->full_clusters);
+               offset = cluster_offset(si, ci);
+               end = min(si->max, offset + SWAPFILE_CLUSTER);
+               to_scan--;
+
+               while (offset < end) {
+                       if (READ_ONCE(map[offset]) == SWAP_HAS_CACHE) {
+                               spin_unlock(&si->lock);
+                               nr_reclaim = __try_to_reclaim_swap(si, offset,
+                                                                  TTRS_ANYWAY | TTRS_DIRECT);
+                               spin_lock(&si->lock);
+                               if (nr_reclaim > 0) {
+                                       offset += nr_reclaim;
+                                       total_reclaimed += nr_reclaim;
+                                       continue;
+                               } else if (nr_reclaim < 0) {
+                                       offset += -nr_reclaim;
+                                       continue;
+                               }
+                       }
+                       offset++;
+               }
+               if (to_scan <= 0 || total_reclaimed)
+                       break;
+       }
+}
+
 /*
  * Try to get swap entries with specified order from current cpu's swap entry
  * pool (a cluster). This might involve allocating a new cluster for current CPU
@@ -826,7 +859,15 @@ new_cluster:
                                goto done;
                }
        }
+
 done:
+       /* Try reclaim from full clusters if device is nearfull */
+       if (vm_swap_full() && (!found || (si->pages - si->inuse_pages) < SWAPFILE_CLUSTER)) {
+               swap_reclaim_full_clusters(si);
+               if (!found && !order && si->pages != si->inuse_pages)
+                       goto new_cluster;
+       }
+
        cluster->next[order] = offset;
        return found;
 }
@@ -3115,6 +3156,7 @@ static int setup_swap_map_and_extents(struct swap_info_struct *p,
        nr_good_pages = maxpages - 1;   /* omit header page */
 
        INIT_LIST_HEAD(&p->free_clusters);
+       INIT_LIST_HEAD(&p->full_clusters);
        INIT_LIST_HEAD(&p->discard_clusters);
 
        for (i = 0; i < SWAP_NR_ORDERS; i++) {