[PATCH 0/7] mm: zswap: move writeback LRU from zpool to zswap

[PATCH 0/7] mm: zswap: move writeback LRU from zpool to zswap

[Domenico Cerasuolo]

This series aims to improve the zswap reclaim mechanism by reorganizing
the LRU management. In the current implementation, the LRU is maintained
within each zpool driver, resulting in duplicated code across the three
drivers. The proposed change consists in moving the LRU management from
the individual implementations up to the zswap layer.

The primary objective of this refactoring effort is to simplify the
codebase. By unifying the reclaim loop and consolidating LRU handling
within zswap, we can eliminate redundant code and improve
maintainability. Additionally, this change enables the reclamation of
stored pages in their actual LRU order. Presently, the zpool drivers
link backing pages in an LRU, causing compressed pages with different
LRU positions to be written back simultaneously.

The series consists of several patches. The first patch implements the
LRU and the reclaim loop in zswap, but it is not used yet because all
three driver implementations are marked as zpool_evictable.
The following three commits modify each zpool driver to be not
zpool_evictable, allowing the use of the reclaim loop in zswap.
As the drivers removed their shrink functions, the zpool interface is
then trimmed by removing zpool_evictable, zpool_ops, and zpool_shrink.
Finally, the code in zswap is further cleaned up by simplifying the
writeback function and removing the now unnecessary zswap_header.

Based on mm-stable + commit 399ab221f3ff
("mm: zswap: shrink until can accept") currently in mm-unstable.

Domenico Cerasuolo (7):
  mm: zswap: add pool shrinking mechanism
  mm: zswap: remove page reclaim logic from zbud
  mm: zswap: remove page reclaim logic from z3fold
  mm: zswap: remove page reclaim logic from zsmalloc
  mm: zswap: remove shrink from zpool interface
  mm: zswap: simplify writeback function
  mm: zswap: remove zswap_header

 include/linux/zpool.h |  19 +--
 mm/z3fold.c           | 249 +----------------------------------
 mm/zbud.c             | 167 +-----------------------
 mm/zpool.c            |  48 +------
 mm/zsmalloc.c         | 294 +-----------------------------------------
 mm/zswap.c            | 159 +++++++++++++----------
 6 files changed, 104 insertions(+), 832 deletions(-)

-- 
2.34.1

[RFC PATCH 1/7] mm: zswap: add pool shrinking mechanism

[Domenico Cerasuolo]

Each zpool driver (zbud, z3fold and zsmalloc) implements its own shrink
function, which is called from zpool_shrink. However, with this commit,
a unified shrink function is added to zswap. The ultimate goal is to
eliminate the need for zpool_shrink once all zpool implementations have
dropped their shrink code.

To ensure the functionality of each commit, this change focuses solely
on adding the mechanism itself. No modifications are made to
the backends, meaning that functionally, there are no immediate changes.
The zswap mechanism will only come into effect once the backends have
removed their shrink code. The subsequent commits will address the
modifications needed in the backends.

Signed-off-by: Domenico Cerasuolo <cerasuolodomenico@gmail.com>
---
 mm/zswap.c | 83 ++++++++++++++++++++++++++++++++++++++++++++++++++----
 1 file changed, 78 insertions(+), 5 deletions(-)

diff --git a/mm/zswap.c b/mm/zswap.c
index bcb82e09eb64..80d7780bf066 100644
--- a/mm/zswap.c
+++ b/mm/zswap.c
@@ -159,6 +159,8 @@ struct zswap_pool {
 	struct work_struct shrink_work;
 	struct hlist_node node;
 	char tfm_name[CRYPTO_MAX_ALG_NAME];
+	struct list_head lru;
+	spinlock_t lock;
 };
 
 /*
@@ -176,10 +178,12 @@ struct zswap_pool {
  *            be held while changing the refcount.  Since the lock must
  *            be held, there is no reason to also make refcount atomic.
  * length - the length in bytes of the compressed page data.  Needed during
- *          decompression. For a same value filled page length is 0.
+ *          decompression. For a same value filled page length is 0, and both
+ *          pool and lru are invalid and must be ignored.
  * pool - the zswap_pool the entry's data is in
  * handle - zpool allocation handle that stores the compressed page data
  * value - value of the same-value filled pages which have same content
+ * lru - handle to the pool's lru used to evict pages.
  */
 struct zswap_entry {
 	struct rb_node rbnode;
@@ -192,6 +196,7 @@ struct zswap_entry {
 		unsigned long value;
 	};
 	struct obj_cgroup *objcg;
+	struct list_head lru;
 };
 
 struct zswap_header {
@@ -364,6 +369,9 @@ static void zswap_free_entry(struct zswap_entry *entry)
 	if (!entry->length)
 		atomic_dec(&zswap_same_filled_pages);
 	else {
+		spin_lock(&entry->pool->lock);
+		list_del_init(&entry->lru);
+		spin_unlock(&entry->pool->lock);
 		zpool_free(entry->pool->zpool, entry->handle);
 		zswap_pool_put(entry->pool);
 	}
@@ -584,14 +592,65 @@ static struct zswap_pool *zswap_pool_find_get(char *type, char *compressor)
 	return NULL;
 }
 
+static int zswap_shrink(struct zswap_pool *pool)
+{
+	struct zswap_entry *lru_entry, *tree_entry = NULL;
+	struct zswap_header *zhdr;
+	struct zswap_tree *tree;
+	swp_entry_t swpentry;
+	int ret;
+
+	/* get a reclaimable entry from LRU */
+	spin_lock(&pool->lock);
+	if (list_empty(&pool->lru)) {
+		spin_unlock(&pool->lock);
+		return -EINVAL;
+	}
+	lru_entry = list_last_entry(&pool->lru, struct zswap_entry, lru);
+	list_del_init(&lru_entry->lru);
+	zhdr = zpool_map_handle(pool->zpool, lru_entry->handle, ZPOOL_MM_RO);
+	tree = zswap_trees[swp_type(zhdr->swpentry)];
+	zpool_unmap_handle(pool->zpool, lru_entry->handle);
+	swpentry = zhdr->swpentry;
+	spin_unlock(&pool->lock);
+
+	/* hold a reference from tree so it won't be freed during writeback */
+	spin_lock(&tree->lock);
+	tree_entry = zswap_entry_find_get(&tree->rbroot, swp_offset(swpentry));
+	if (tree_entry != lru_entry) {
+		if (tree_entry)
+			zswap_entry_put(tree, tree_entry);
+		spin_unlock(&tree->lock);
+		return -EAGAIN;
+	}
+	spin_unlock(&tree->lock);
+
+	ret = zswap_writeback_entry(pool->zpool, lru_entry->handle);
+
+	spin_lock(&tree->lock);
+	if (ret) {
+		spin_lock(&pool->lock);
+		list_move(&lru_entry->lru, &pool->lru);
+		spin_unlock(&pool->lock);
+	}
+	zswap_entry_put(tree, tree_entry);
+	spin_unlock(&tree->lock);
+
+	return ret ? -EAGAIN : 0;
+}
+
 static void shrink_worker(struct work_struct *w)
 {
 	struct zswap_pool *pool = container_of(w, typeof(*pool),
 						shrink_work);
 	int ret, failures = 0;
 
+	/* zpool_evictable will be removed once all 3 backends have migrated*/
 	do {
-		ret = zpool_shrink(pool->zpool, 1, NULL);
+		if (zpool_evictable(pool->zpool))
+			ret = zpool_shrink(pool->zpool, 1, NULL);
+		else
+			ret = zswap_shrink(pool);
 		if (ret) {
 			zswap_reject_reclaim_fail++;
 			if (ret != -EAGAIN)
@@ -655,6 +714,8 @@ static struct zswap_pool *zswap_pool_create(char *type, char *compressor)
 	 */
 	kref_init(&pool->kref);
 	INIT_LIST_HEAD(&pool->list);
+	INIT_LIST_HEAD(&pool->lru);
+	spin_lock_init(&pool->lock);
 	INIT_WORK(&pool->shrink_work, shrink_worker);
 
 	zswap_pool_debug("created", pool);
@@ -1270,7 +1331,7 @@ static int zswap_frontswap_store(unsigned type, pgoff_t offset,
 	}
 
 	/* store */
-	hlen = zpool_evictable(entry->pool->zpool) ? sizeof(zhdr) : 0;
+	hlen = sizeof(zhdr);
 	gfp = __GFP_NORETRY | __GFP_NOWARN | __GFP_KSWAPD_RECLAIM;
 	if (zpool_malloc_support_movable(entry->pool->zpool))
 		gfp |= __GFP_HIGHMEM | __GFP_MOVABLE;
@@ -1313,6 +1374,13 @@ static int zswap_frontswap_store(unsigned type, pgoff_t offset,
 			zswap_entry_put(tree, dupentry);
 		}
 	} while (ret == -EEXIST);
+	INIT_LIST_HEAD(&entry->lru);
+	/* zpool_evictable will be removed once all 3 backends have migrated*/
+	if (entry->length && !zpool_evictable(entry->pool->zpool)) {
+		spin_lock(&entry->pool->lock);
+		list_add(&entry->lru, &entry->pool->lru);
+		spin_unlock(&entry->pool->lock);
+	}
 	spin_unlock(&tree->lock);
 
 	/* update stats */
@@ -1384,8 +1452,7 @@ static int zswap_frontswap_load(unsigned type, pgoff_t offset,
 	/* decompress */
 	dlen = PAGE_SIZE;
 	src = zpool_map_handle(entry->pool->zpool, entry->handle, ZPOOL_MM_RO);
-	if (zpool_evictable(entry->pool->zpool))
-		src += sizeof(struct zswap_header);
+	src += sizeof(struct zswap_header);
 
 	if (!zpool_can_sleep_mapped(entry->pool->zpool)) {
 		memcpy(tmp, src, entry->length);
@@ -1415,6 +1482,12 @@ static int zswap_frontswap_load(unsigned type, pgoff_t offset,
 freeentry:
 	spin_lock(&tree->lock);
 	zswap_entry_put(tree, entry);
+	/* zpool_evictable will be removed once all 3 backends have migrated*/
+	if (entry->length && !zpool_evictable(entry->pool->zpool)) {
+		spin_lock(&entry->pool->lock);
+		list_move(&entry->lru, &entry->pool->lru);
+		spin_unlock(&entry->pool->lock);
+	}
 	spin_unlock(&tree->lock);
 
 	return ret;
-- 
2.34.1

[RFC PATCH 2/7] mm: zswap: remove page reclaim logic from zbud

[Domenico Cerasuolo]

With the recent enhancement to zswap enabling direct page writeback, the
need for the shrink code in zbud has become obsolete. As a result, this
commit removes the page reclaim logic from zbud entirely.

Signed-off-by: Domenico Cerasuolo <cerasuolodomenico@gmail.com>
---
 mm/zbud.c | 164 +-----------------------------------------------------
 1 file changed, 2 insertions(+), 162 deletions(-)

diff --git a/mm/zbud.c b/mm/zbud.c
index 3acd26193920..19bc662ef5e9 100644
--- a/mm/zbud.c
+++ b/mm/zbud.c
@@ -83,11 +83,7 @@ struct zbud_pool;
  *		its free region.
  * @buddied:	list tracking the zbud pages that contain two buddies;
  *		these zbud pages are full
- * @lru:	list tracking the zbud pages in LRU order by most recently
- *		added buddy.
  * @pages_nr:	number of zbud pages in the pool.
- * @zpool:	zpool driver
- * @zpool_ops:	zpool operations structure with an evict callback
  *
  * This structure is allocated at pool creation time and maintains metadata
  * pertaining to a particular zbud pool.
@@ -102,26 +98,20 @@ struct zbud_pool {
 		struct list_head buddied;
 		struct list_head unbuddied[NCHUNKS];
 	};
-	struct list_head lru;
 	u64 pages_nr;
-	struct zpool *zpool;
-	const struct zpool_ops *zpool_ops;
 };
 
 /*
  * struct zbud_header - zbud page metadata occupying the first chunk of each
  *			zbud page.
  * @buddy:	links the zbud page into the unbuddied/buddied lists in the pool
- * @lru:	links the zbud page into the lru list in the pool
  * @first_chunks:	the size of the first buddy in chunks, 0 if free
  * @last_chunks:	the size of the last buddy in chunks, 0 if free
  */
 struct zbud_header {
 	struct list_head buddy;
-	struct list_head lru;
 	unsigned int first_chunks;
 	unsigned int last_chunks;
-	bool under_reclaim;
 };
 
 /*****************
@@ -149,8 +139,6 @@ static struct zbud_header *init_zbud_page(struct page *page)
 	zhdr->first_chunks = 0;
 	zhdr->last_chunks = 0;
 	INIT_LIST_HEAD(&zhdr->buddy);
-	INIT_LIST_HEAD(&zhdr->lru);
-	zhdr->under_reclaim = false;
 	return zhdr;
 }
 
@@ -221,7 +209,6 @@ static struct zbud_pool *zbud_create_pool(gfp_t gfp)
 	for_each_unbuddied_list(i, 0)
 		INIT_LIST_HEAD(&pool->unbuddied[i]);
 	INIT_LIST_HEAD(&pool->buddied);
-	INIT_LIST_HEAD(&pool->lru);
 	pool->pages_nr = 0;
 	return pool;
 }
@@ -310,11 +297,6 @@ static int zbud_alloc(struct zbud_pool *pool, size_t size, gfp_t gfp,
 		list_add(&zhdr->buddy, &pool->buddied);
 	}
 
-	/* Add/move zbud page to beginning of LRU */
-	if (!list_empty(&zhdr->lru))
-		list_del(&zhdr->lru);
-	list_add(&zhdr->lru, &pool->lru);
-
 	*handle = encode_handle(zhdr, bud);
 	spin_unlock(&pool->lock);
 
@@ -325,11 +307,6 @@ static int zbud_alloc(struct zbud_pool *pool, size_t size, gfp_t gfp,
  * zbud_free() - frees the allocation associated with the given handle
  * @pool:	pool in which the allocation resided
  * @handle:	handle associated with the allocation returned by zbud_alloc()
- *
- * In the case that the zbud page in which the allocation resides is under
- * reclaim, as indicated by the PG_reclaim flag being set, this function
- * only sets the first|last_chunks to 0.  The page is actually freed
- * once both buddies are evicted (see zbud_reclaim_page() below).
  */
 static void zbud_free(struct zbud_pool *pool, unsigned long handle)
 {
@@ -345,18 +322,11 @@ static void zbud_free(struct zbud_pool *pool, unsigned long handle)
 	else
 		zhdr->first_chunks = 0;
 
-	if (zhdr->under_reclaim) {
-		/* zbud page is under reclaim, reclaim will free */
-		spin_unlock(&pool->lock);
-		return;
-	}
-
 	/* Remove from existing buddy list */
 	list_del(&zhdr->buddy);
 
 	if (zhdr->first_chunks == 0 && zhdr->last_chunks == 0) {
 		/* zbud page is empty, free */
-		list_del(&zhdr->lru);
 		free_zbud_page(zhdr);
 		pool->pages_nr--;
 	} else {
@@ -368,110 +338,6 @@ static void zbud_free(struct zbud_pool *pool, unsigned long handle)
 	spin_unlock(&pool->lock);
 }
 
-/**
- * zbud_reclaim_page() - evicts allocations from a pool page and frees it
- * @pool:	pool from which a page will attempt to be evicted
- * @retries:	number of pages on the LRU list for which eviction will
- *		be attempted before failing
- *
- * zbud reclaim is different from normal system reclaim in that the reclaim is
- * done from the bottom, up.  This is because only the bottom layer, zbud, has
- * information on how the allocations are organized within each zbud page. This
- * has the potential to create interesting locking situations between zbud and
- * the user, however.
- *
- * To avoid these, this is how zbud_reclaim_page() should be called:
- *
- * The user detects a page should be reclaimed and calls zbud_reclaim_page().
- * zbud_reclaim_page() will remove a zbud page from the pool LRU list and call
- * the user-defined eviction handler with the pool and handle as arguments.
- *
- * If the handle can not be evicted, the eviction handler should return
- * non-zero. zbud_reclaim_page() will add the zbud page back to the
- * appropriate list and try the next zbud page on the LRU up to
- * a user defined number of retries.
- *
- * If the handle is successfully evicted, the eviction handler should
- * return 0 _and_ should have called zbud_free() on the handle. zbud_free()
- * contains logic to delay freeing the page if the page is under reclaim,
- * as indicated by the setting of the PG_reclaim flag on the underlying page.
- *
- * If all buddies in the zbud page are successfully evicted, then the
- * zbud page can be freed.
- *
- * Returns: 0 if page is successfully freed, otherwise -EINVAL if there are
- * no pages to evict or an eviction handler is not registered, -EAGAIN if
- * the retry limit was hit.
- */
-static int zbud_reclaim_page(struct zbud_pool *pool, unsigned int retries)
-{
-	int i, ret, freechunks;
-	struct zbud_header *zhdr;
-	unsigned long first_handle = 0, last_handle = 0;
-
-	spin_lock(&pool->lock);
-	if (list_empty(&pool->lru)) {
-		spin_unlock(&pool->lock);
-		return -EINVAL;
-	}
-	for (i = 0; i < retries; i++) {
-		zhdr = list_last_entry(&pool->lru, struct zbud_header, lru);
-		list_del(&zhdr->lru);
-		list_del(&zhdr->buddy);
-		/* Protect zbud page against free */
-		zhdr->under_reclaim = true;
-		/*
-		 * We need encode the handles before unlocking, since we can
-		 * race with free that will set (first|last)_chunks to 0
-		 */
-		first_handle = 0;
-		last_handle = 0;
-		if (zhdr->first_chunks)
-			first_handle = encode_handle(zhdr, FIRST);
-		if (zhdr->last_chunks)
-			last_handle = encode_handle(zhdr, LAST);
-		spin_unlock(&pool->lock);
-
-		/* Issue the eviction callback(s) */
-		if (first_handle) {
-			ret = pool->zpool_ops->evict(pool->zpool, first_handle);
-			if (ret)
-				goto next;
-		}
-		if (last_handle) {
-			ret = pool->zpool_ops->evict(pool->zpool, last_handle);
-			if (ret)
-				goto next;
-		}
-next:
-		spin_lock(&pool->lock);
-		zhdr->under_reclaim = false;
-		if (zhdr->first_chunks == 0 && zhdr->last_chunks == 0) {
-			/*
-			 * Both buddies are now free, free the zbud page and
-			 * return success.
-			 */
-			free_zbud_page(zhdr);
-			pool->pages_nr--;
-			spin_unlock(&pool->lock);
-			return 0;
-		} else if (zhdr->first_chunks == 0 ||
-				zhdr->last_chunks == 0) {
-			/* add to unbuddied list */
-			freechunks = num_free_chunks(zhdr);
-			list_add(&zhdr->buddy, &pool->unbuddied[freechunks]);
-		} else {
-			/* add to buddied list */
-			list_add(&zhdr->buddy, &pool->buddied);
-		}
-
-		/* add to beginning of LRU */
-		list_add(&zhdr->lru, &pool->lru);
-	}
-	spin_unlock(&pool->lock);
-	return -EAGAIN;
-}
-
 /**
  * zbud_map() - maps the allocation associated with the given handle
  * @pool:	pool in which the allocation resides
@@ -518,14 +384,7 @@ static void *zbud_zpool_create(const char *name, gfp_t gfp,
 			       const struct zpool_ops *zpool_ops,
 			       struct zpool *zpool)
 {
-	struct zbud_pool *pool;
-
-	pool = zbud_create_pool(gfp);
-	if (pool) {
-		pool->zpool = zpool;
-		pool->zpool_ops = zpool_ops;
-	}
-	return pool;
+	return zbud_create_pool(gfp);
 }
 
 static void zbud_zpool_destroy(void *pool)
@@ -543,25 +402,6 @@ static void zbud_zpool_free(void *pool, unsigned long handle)
 	zbud_free(pool, handle);
 }
 
-static int zbud_zpool_shrink(void *pool, unsigned int pages,
-			unsigned int *reclaimed)
-{
-	unsigned int total = 0;
-	int ret = -EINVAL;
-
-	while (total < pages) {
-		ret = zbud_reclaim_page(pool, 8);
-		if (ret < 0)
-			break;
-		total++;
-	}
-
-	if (reclaimed)
-		*reclaimed = total;
-
-	return ret;
-}
-
 static void *zbud_zpool_map(void *pool, unsigned long handle,
 			enum zpool_mapmode mm)
 {
@@ -585,7 +425,7 @@ static struct zpool_driver zbud_zpool_driver = {
 	.destroy =	zbud_zpool_destroy,
 	.malloc =	zbud_zpool_malloc,
 	.free =		zbud_zpool_free,
-	.shrink =	zbud_zpool_shrink,
+	.shrink =	NULL,
 	.map =		zbud_zpool_map,
 	.unmap =	zbud_zpool_unmap,
 	.total_size =	zbud_zpool_total_size,
-- 
2.34.1

[RFC PATCH 3/7] mm: zswap: remove page reclaim logic from z3fold

[Domenico Cerasuolo]

With the recent enhancement to zswap enabling direct page writeback, the
need for the shrink code in z3fold has become obsolete. As a result,
this commit removes the page reclaim logic from z3fold entirely.

Signed-off-by: Domenico Cerasuolo <cerasuolodomenico@gmail.com>
---
 mm/z3fold.c | 246 +---------------------------------------------------
 1 file changed, 3 insertions(+), 243 deletions(-)

diff --git a/mm/z3fold.c b/mm/z3fold.c
index 0cef845d397b..4af8741553ac 100644
--- a/mm/z3fold.c
+++ b/mm/z3fold.c
@@ -125,13 +125,11 @@ struct z3fold_header {
 /**
  * struct z3fold_pool - stores metadata for each z3fold pool
  * @name:	pool name
- * @lock:	protects pool unbuddied/lru lists
+ * @lock:	protects pool unbuddied lists
  * @stale_lock:	protects pool stale page list
  * @unbuddied:	per-cpu array of lists tracking z3fold pages that contain 2-
  *		buddies; the list each z3fold page is added to depends on
  *		the size of its free region.
- * @lru:	list tracking the z3fold pages in LRU order by most recently
- *		added buddy.
  * @stale:	list of pages marked for freeing
  * @pages_nr:	number of z3fold pages in the pool.
  * @c_handle:	cache for z3fold_buddy_slots allocation
@@ -149,12 +147,9 @@ struct z3fold_pool {
 	spinlock_t lock;
 	spinlock_t stale_lock;
 	struct list_head *unbuddied;
-	struct list_head lru;
 	struct list_head stale;
 	atomic64_t pages_nr;
 	struct kmem_cache *c_handle;
-	struct zpool *zpool;
-	const struct zpool_ops *zpool_ops;
 	struct workqueue_struct *compact_wq;
 	struct workqueue_struct *release_wq;
 	struct work_struct work;
@@ -329,7 +324,6 @@ static struct z3fold_header *init_z3fold_page(struct page *page, bool headless,
 	struct z3fold_header *zhdr = page_address(page);
 	struct z3fold_buddy_slots *slots;
 
-	INIT_LIST_HEAD(&page->lru);
 	clear_bit(PAGE_HEADLESS, &page->private);
 	clear_bit(MIDDLE_CHUNK_MAPPED, &page->private);
 	clear_bit(NEEDS_COMPACTING, &page->private);
@@ -451,8 +445,6 @@ static void __release_z3fold_page(struct z3fold_header *zhdr, bool locked)
 	set_bit(PAGE_STALE, &page->private);
 	clear_bit(NEEDS_COMPACTING, &page->private);
 	spin_lock(&pool->lock);
-	if (!list_empty(&page->lru))
-		list_del_init(&page->lru);
 	spin_unlock(&pool->lock);
 
 	if (locked)
@@ -930,7 +922,6 @@ static struct z3fold_pool *z3fold_create_pool(const char *name, gfp_t gfp)
 		for_each_unbuddied_list(i, 0)
 			INIT_LIST_HEAD(&unbuddied[i]);
 	}
-	INIT_LIST_HEAD(&pool->lru);
 	INIT_LIST_HEAD(&pool->stale);
 	atomic64_set(&pool->pages_nr, 0);
 	pool->name = name;
@@ -1073,12 +1064,6 @@ static int z3fold_alloc(struct z3fold_pool *pool, size_t size, gfp_t gfp,
 
 headless:
 	spin_lock(&pool->lock);
-	/* Add/move z3fold page to beginning of LRU */
-	if (!list_empty(&page->lru))
-		list_del(&page->lru);
-
-	list_add(&page->lru, &pool->lru);
-
 	*handle = encode_handle(zhdr, bud);
 	spin_unlock(&pool->lock);
 	if (bud != HEADLESS)
@@ -1115,9 +1100,6 @@ static void z3fold_free(struct z3fold_pool *pool, unsigned long handle)
 		 * immediately so we don't care about its value any more.
 		 */
 		if (!page_claimed) {
-			spin_lock(&pool->lock);
-			list_del(&page->lru);
-			spin_unlock(&pool->lock);
 			put_z3fold_header(zhdr);
 			free_z3fold_page(page, true);
 			atomic64_dec(&pool->pages_nr);
@@ -1172,194 +1154,6 @@ static void z3fold_free(struct z3fold_pool *pool, unsigned long handle)
 	put_z3fold_header(zhdr);
 }
 
-/**
- * z3fold_reclaim_page() - evicts allocations from a pool page and frees it
- * @pool:	pool from which a page will attempt to be evicted
- * @retries:	number of pages on the LRU list for which eviction will
- *		be attempted before failing
- *
- * z3fold reclaim is different from normal system reclaim in that it is done
- * from the bottom, up. This is because only the bottom layer, z3fold, has
- * information on how the allocations are organized within each z3fold page.
- * This has the potential to create interesting locking situations between
- * z3fold and the user, however.
- *
- * To avoid these, this is how z3fold_reclaim_page() should be called:
- *
- * The user detects a page should be reclaimed and calls z3fold_reclaim_page().
- * z3fold_reclaim_page() will remove a z3fold page from the pool LRU list and
- * call the user-defined eviction handler with the pool and handle as
- * arguments.
- *
- * If the handle can not be evicted, the eviction handler should return
- * non-zero. z3fold_reclaim_page() will add the z3fold page back to the
- * appropriate list and try the next z3fold page on the LRU up to
- * a user defined number of retries.
- *
- * If the handle is successfully evicted, the eviction handler should
- * return 0 _and_ should have called z3fold_free() on the handle. z3fold_free()
- * contains logic to delay freeing the page if the page is under reclaim,
- * as indicated by the setting of the PG_reclaim flag on the underlying page.
- *
- * If all buddies in the z3fold page are successfully evicted, then the
- * z3fold page can be freed.
- *
- * Returns: 0 if page is successfully freed, otherwise -EINVAL if there are
- * no pages to evict or an eviction handler is not registered, -EAGAIN if
- * the retry limit was hit.
- */
-static int z3fold_reclaim_page(struct z3fold_pool *pool, unsigned int retries)
-{
-	int i, ret = -1;
-	struct z3fold_header *zhdr = NULL;
-	struct page *page = NULL;
-	struct list_head *pos;
-	unsigned long first_handle = 0, middle_handle = 0, last_handle = 0;
-	struct z3fold_buddy_slots slots __attribute__((aligned(SLOTS_ALIGN)));
-
-	rwlock_init(&slots.lock);
-	slots.pool = (unsigned long)pool | (1 << HANDLES_NOFREE);
-
-	spin_lock(&pool->lock);
-	for (i = 0; i < retries; i++) {
-		if (list_empty(&pool->lru)) {
-			spin_unlock(&pool->lock);
-			return -EINVAL;
-		}
-		list_for_each_prev(pos, &pool->lru) {
-			page = list_entry(pos, struct page, lru);
-
-			zhdr = page_address(page);
-			if (test_bit(PAGE_HEADLESS, &page->private)) {
-				/*
-				 * For non-headless pages, we wait to do this
-				 * until we have the page lock to avoid racing
-				 * with __z3fold_alloc(). Headless pages don't
-				 * have a lock (and __z3fold_alloc() will never
-				 * see them), but we still need to test and set
-				 * PAGE_CLAIMED to avoid racing with
-				 * z3fold_free(), so just do it now before
-				 * leaving the loop.
-				 */
-				if (test_and_set_bit(PAGE_CLAIMED, &page->private))
-					continue;
-
-				break;
-			}
-
-			if (!z3fold_page_trylock(zhdr)) {
-				zhdr = NULL;
-				continue; /* can't evict at this point */
-			}
-
-			/* test_and_set_bit is of course atomic, but we still
-			 * need to do it under page lock, otherwise checking
-			 * that bit in __z3fold_alloc wouldn't make sense
-			 */
-			if (zhdr->foreign_handles ||
-			    test_and_set_bit(PAGE_CLAIMED, &page->private)) {
-				z3fold_page_unlock(zhdr);
-				zhdr = NULL;
-				continue; /* can't evict such page */
-			}
-			list_del_init(&zhdr->buddy);
-			zhdr->cpu = -1;
-			/* See comment in __z3fold_alloc. */
-			kref_get(&zhdr->refcount);
-			break;
-		}
-
-		if (!zhdr)
-			break;
-
-		list_del_init(&page->lru);
-		spin_unlock(&pool->lock);
-
-		if (!test_bit(PAGE_HEADLESS, &page->private)) {
-			/*
-			 * We need encode the handles before unlocking, and
-			 * use our local slots structure because z3fold_free
-			 * can zero out zhdr->slots and we can't do much
-			 * about that
-			 */
-			first_handle = 0;
-			last_handle = 0;
-			middle_handle = 0;
-			memset(slots.slot, 0, sizeof(slots.slot));
-			if (zhdr->first_chunks)
-				first_handle = __encode_handle(zhdr, &slots,
-								FIRST);
-			if (zhdr->middle_chunks)
-				middle_handle = __encode_handle(zhdr, &slots,
-								MIDDLE);
-			if (zhdr->last_chunks)
-				last_handle = __encode_handle(zhdr, &slots,
-								LAST);
-			/*
-			 * it's safe to unlock here because we hold a
-			 * reference to this page
-			 */
-			z3fold_page_unlock(zhdr);
-		} else {
-			first_handle = encode_handle(zhdr, HEADLESS);
-			last_handle = middle_handle = 0;
-		}
-		/* Issue the eviction callback(s) */
-		if (middle_handle) {
-			ret = pool->zpool_ops->evict(pool->zpool, middle_handle);
-			if (ret)
-				goto next;
-		}
-		if (first_handle) {
-			ret = pool->zpool_ops->evict(pool->zpool, first_handle);
-			if (ret)
-				goto next;
-		}
-		if (last_handle) {
-			ret = pool->zpool_ops->evict(pool->zpool, last_handle);
-			if (ret)
-				goto next;
-		}
-next:
-		if (test_bit(PAGE_HEADLESS, &page->private)) {
-			if (ret == 0) {
-				free_z3fold_page(page, true);
-				atomic64_dec(&pool->pages_nr);
-				return 0;
-			}
-			spin_lock(&pool->lock);
-			list_add(&page->lru, &pool->lru);
-			spin_unlock(&pool->lock);
-			clear_bit(PAGE_CLAIMED, &page->private);
-		} else {
-			struct z3fold_buddy_slots *slots = zhdr->slots;
-			z3fold_page_lock(zhdr);
-			if (kref_put(&zhdr->refcount,
-					release_z3fold_page_locked)) {
-				kmem_cache_free(pool->c_handle, slots);
-				return 0;
-			}
-			/*
-			 * if we are here, the page is still not completely
-			 * free. Take the global pool lock then to be able
-			 * to add it back to the lru list
-			 */
-			spin_lock(&pool->lock);
-			list_add(&page->lru, &pool->lru);
-			spin_unlock(&pool->lock);
-			if (list_empty(&zhdr->buddy))
-				add_to_unbuddied(pool, zhdr);
-			clear_bit(PAGE_CLAIMED, &page->private);
-			z3fold_page_unlock(zhdr);
-		}
-
-		/* We started off locked to we need to lock the pool back */
-		spin_lock(&pool->lock);
-	}
-	spin_unlock(&pool->lock);
-	return -EAGAIN;
-}
-
 /**
  * z3fold_map() - maps the allocation associated with the given handle
  * @pool:	pool in which the allocation resides
@@ -1470,8 +1264,6 @@ static bool z3fold_page_isolate(struct page *page, isolate_mode_t mode)
 	spin_lock(&pool->lock);
 	if (!list_empty(&zhdr->buddy))
 		list_del_init(&zhdr->buddy);
-	if (!list_empty(&page->lru))
-		list_del_init(&page->lru);
 	spin_unlock(&pool->lock);
 
 	kref_get(&zhdr->refcount);
@@ -1531,9 +1323,6 @@ static int z3fold_page_migrate(struct page *newpage, struct page *page,
 		encode_handle(new_zhdr, MIDDLE);
 	set_bit(NEEDS_COMPACTING, &newpage->private);
 	new_zhdr->cpu = smp_processor_id();
-	spin_lock(&pool->lock);
-	list_add(&newpage->lru, &pool->lru);
-	spin_unlock(&pool->lock);
 	__SetPageMovable(newpage, &z3fold_mops);
 	z3fold_page_unlock(new_zhdr);
 
@@ -1559,9 +1348,6 @@ static void z3fold_page_putback(struct page *page)
 	INIT_LIST_HEAD(&page->lru);
 	if (kref_put(&zhdr->refcount, release_z3fold_page_locked))
 		return;
-	spin_lock(&pool->lock);
-	list_add(&page->lru, &pool->lru);
-	spin_unlock(&pool->lock);
 	if (list_empty(&zhdr->buddy))
 		add_to_unbuddied(pool, zhdr);
 	clear_bit(PAGE_CLAIMED, &page->private);
@@ -1582,14 +1368,7 @@ static void *z3fold_zpool_create(const char *name, gfp_t gfp,
 			       const struct zpool_ops *zpool_ops,
 			       struct zpool *zpool)
 {
-	struct z3fold_pool *pool;
-
-	pool = z3fold_create_pool(name, gfp);
-	if (pool) {
-		pool->zpool = zpool;
-		pool->zpool_ops = zpool_ops;
-	}
-	return pool;
+	return z3fold_create_pool(name, gfp);
 }
 
 static void z3fold_zpool_destroy(void *pool)
@@ -1607,25 +1386,6 @@ static void z3fold_zpool_free(void *pool, unsigned long handle)
 	z3fold_free(pool, handle);
 }
 
-static int z3fold_zpool_shrink(void *pool, unsigned int pages,
-			unsigned int *reclaimed)
-{
-	unsigned int total = 0;
-	int ret = -EINVAL;
-
-	while (total < pages) {
-		ret = z3fold_reclaim_page(pool, 8);
-		if (ret < 0)
-			break;
-		total++;
-	}
-
-	if (reclaimed)
-		*reclaimed = total;
-
-	return ret;
-}
-
 static void *z3fold_zpool_map(void *pool, unsigned long handle,
 			enum zpool_mapmode mm)
 {
@@ -1649,7 +1409,7 @@ static struct zpool_driver z3fold_zpool_driver = {
 	.destroy =	z3fold_zpool_destroy,
 	.malloc =	z3fold_zpool_malloc,
 	.free =		z3fold_zpool_free,
-	.shrink =	z3fold_zpool_shrink,
+	.shrink =	NULL,
 	.map =		z3fold_zpool_map,
 	.unmap =	z3fold_zpool_unmap,
 	.total_size =	z3fold_zpool_total_size,
-- 
2.34.1

[RFC PATCH 4/7] mm: zswap: remove page reclaim logic from zsmalloc

[Domenico Cerasuolo]

With the recent enhancement to zswap enabling direct page writeback, the
need for the shrink code in zsmalloc has become obsolete. As a result,
this commit removes the page reclaim logic from zsmalloc entirely.

Signed-off-by: Domenico Cerasuolo <cerasuolodomenico@gmail.com>
---
 mm/zsmalloc.c | 291 +-------------------------------------------------
 1 file changed, 2 insertions(+), 289 deletions(-)

diff --git a/mm/zsmalloc.c b/mm/zsmalloc.c
index 02f7f414aade..c87a60514f21 100644
--- a/mm/zsmalloc.c
+++ b/mm/zsmalloc.c
@@ -250,13 +250,6 @@ struct zs_pool {
 	/* Compact classes */
 	struct shrinker shrinker;
 
-#ifdef CONFIG_ZPOOL
-	/* List tracking the zspages in LRU order by most recently added object */
-	struct list_head lru;
-	struct zpool *zpool;
-	const struct zpool_ops *zpool_ops;
-#endif
-
 #ifdef CONFIG_ZSMALLOC_STAT
 	struct dentry *stat_dentry;
 #endif
@@ -279,13 +272,6 @@ struct zspage {
 	unsigned int freeobj;
 	struct page *first_page;
 	struct list_head list; /* fullness list */
-
-#ifdef CONFIG_ZPOOL
-	/* links the zspage to the lru list in the pool */
-	struct list_head lru;
-	bool under_reclaim;
-#endif
-
 	struct zs_pool *pool;
 	rwlock_t lock;
 };
@@ -393,14 +379,7 @@ static void *zs_zpool_create(const char *name, gfp_t gfp,
 	 * different contexts and its caller must provide a valid
 	 * gfp mask.
 	 */
-	struct zs_pool *pool = zs_create_pool(name);
-
-	if (pool) {
-		pool->zpool = zpool;
-		pool->zpool_ops = zpool_ops;
-	}
-
-	return pool;
+	return zs_create_pool(name);
 }
 
 static void zs_zpool_destroy(void *pool)
@@ -422,27 +401,6 @@ static void zs_zpool_free(void *pool, unsigned long handle)
 	zs_free(pool, handle);
 }
 
-static int zs_reclaim_page(struct zs_pool *pool, unsigned int retries);
-
-static int zs_zpool_shrink(void *pool, unsigned int pages,
-			unsigned int *reclaimed)
-{
-	unsigned int total = 0;
-	int ret = -EINVAL;
-
-	while (total < pages) {
-		ret = zs_reclaim_page(pool, 8);
-		if (ret < 0)
-			break;
-		total++;
-	}
-
-	if (reclaimed)
-		*reclaimed = total;
-
-	return ret;
-}
-
 static void *zs_zpool_map(void *pool, unsigned long handle,
 			enum zpool_mapmode mm)
 {
@@ -481,7 +439,7 @@ static struct zpool_driver zs_zpool_driver = {
 	.malloc_support_movable = true,
 	.malloc =		  zs_zpool_malloc,
 	.free =			  zs_zpool_free,
-	.shrink =		  zs_zpool_shrink,
+	.shrink =		  NULL,
 	.map =			  zs_zpool_map,
 	.unmap =		  zs_zpool_unmap,
 	.total_size =		  zs_zpool_total_size,
@@ -884,14 +842,6 @@ static inline bool obj_allocated(struct page *page, void *obj, unsigned long *ph
 	return obj_tagged(page, obj, phandle, OBJ_ALLOCATED_TAG);
 }
 
-#ifdef CONFIG_ZPOOL
-static bool obj_stores_deferred_handle(struct page *page, void *obj,
-		unsigned long *phandle)
-{
-	return obj_tagged(page, obj, phandle, OBJ_DEFERRED_HANDLE_TAG);
-}
-#endif
-
 static void reset_page(struct page *page)
 {
 	__ClearPageMovable(page);
@@ -1006,9 +956,6 @@ static void free_zspage(struct zs_pool *pool, struct size_class *class,
 	}
 
 	remove_zspage(class, zspage, ZS_INUSE_RATIO_0);
-#ifdef CONFIG_ZPOOL
-	list_del(&zspage->lru);
-#endif
 	__free_zspage(pool, class, zspage);
 }
 
@@ -1054,11 +1001,6 @@ static void init_zspage(struct size_class *class, struct zspage *zspage)
 		off %= PAGE_SIZE;
 	}
 
-#ifdef CONFIG_ZPOOL
-	INIT_LIST_HEAD(&zspage->lru);
-	zspage->under_reclaim = false;
-#endif
-
 	set_freeobj(zspage, 0);
 }
 
@@ -1525,13 +1467,6 @@ unsigned long zs_malloc(struct zs_pool *pool, size_t size, gfp_t gfp)
 	/* We completely set up zspage so mark them as movable */
 	SetZsPageMovable(pool, zspage);
 out:
-#ifdef CONFIG_ZPOOL
-	/* Add/move zspage to beginning of LRU */
-	if (!list_empty(&zspage->lru))
-		list_del(&zspage->lru);
-	list_add(&zspage->lru, &pool->lru);
-#endif
-
 	spin_unlock(&pool->lock);
 
 	return handle;
@@ -1600,20 +1535,6 @@ void zs_free(struct zs_pool *pool, unsigned long handle)
 	class = zspage_class(pool, zspage);
 
 	class_stat_dec(class, ZS_OBJS_INUSE, 1);
-
-#ifdef CONFIG_ZPOOL
-	if (zspage->under_reclaim) {
-		/*
-		 * Reclaim needs the handles during writeback. It'll free
-		 * them along with the zspage when it's done with them.
-		 *
-		 * Record current deferred handle in the object's header.
-		 */
-		obj_free(class->size, obj, &handle);
-		spin_unlock(&pool->lock);
-		return;
-	}
-#endif
 	obj_free(class->size, obj, NULL);
 
 	fullness = fix_fullness_group(class, zspage);
@@ -1890,23 +1811,6 @@ static void lock_zspage(struct zspage *zspage)
 }
 #endif /* defined(CONFIG_ZPOOL) || defined(CONFIG_COMPACTION) */
 
-#ifdef CONFIG_ZPOOL
-/*
- * Unlocks all the pages of the zspage.
- *
- * pool->lock must be held before this function is called
- * to prevent the underlying pages from migrating.
- */
-static void unlock_zspage(struct zspage *zspage)
-{
-	struct page *page = get_first_page(zspage);
-
-	do {
-		unlock_page(page);
-	} while ((page = get_next_page(page)) != NULL);
-}
-#endif /* CONFIG_ZPOOL */
-
 static void migrate_lock_init(struct zspage *zspage)
 {
 	rwlock_init(&zspage->lock);
@@ -2126,9 +2030,6 @@ static void async_free_zspage(struct work_struct *work)
 		VM_BUG_ON(fullness != ZS_INUSE_RATIO_0);
 		class = pool->size_class[class_idx];
 		spin_lock(&pool->lock);
-#ifdef CONFIG_ZPOOL
-		list_del(&zspage->lru);
-#endif
 		__free_zspage(pool, class, zspage);
 		spin_unlock(&pool->lock);
 	}
@@ -2474,10 +2375,6 @@ struct zs_pool *zs_create_pool(const char *name)
 	 */
 	zs_register_shrinker(pool);
 
-#ifdef CONFIG_ZPOOL
-	INIT_LIST_HEAD(&pool->lru);
-#endif
-
 	return pool;
 
 err:
@@ -2520,190 +2417,6 @@ void zs_destroy_pool(struct zs_pool *pool)
 }
 EXPORT_SYMBOL_GPL(zs_destroy_pool);
 
-#ifdef CONFIG_ZPOOL
-static void restore_freelist(struct zs_pool *pool, struct size_class *class,
-		struct zspage *zspage)
-{
-	unsigned int obj_idx = 0;
-	unsigned long handle, off = 0; /* off is within-page offset */
-	struct page *page = get_first_page(zspage);
-	struct link_free *prev_free = NULL;
-	void *prev_page_vaddr = NULL;
-
-	/* in case no free object found */
-	set_freeobj(zspage, (unsigned int)(-1UL));
-
-	while (page) {
-		void *vaddr = kmap_atomic(page);
-		struct page *next_page;
-
-		while (off < PAGE_SIZE) {
-			void *obj_addr = vaddr + off;
-
-			/* skip allocated object */
-			if (obj_allocated(page, obj_addr, &handle)) {
-				obj_idx++;
-				off += class->size;
-				continue;
-			}
-
-			/* free deferred handle from reclaim attempt */
-			if (obj_stores_deferred_handle(page, obj_addr, &handle))
-				cache_free_handle(pool, handle);
-
-			if (prev_free)
-				prev_free->next = obj_idx << OBJ_TAG_BITS;
-			else /* first free object found */
-				set_freeobj(zspage, obj_idx);
-
-			prev_free = (struct link_free *)vaddr + off / sizeof(*prev_free);
-			/* if last free object in a previous page, need to unmap */
-			if (prev_page_vaddr) {
-				kunmap_atomic(prev_page_vaddr);
-				prev_page_vaddr = NULL;
-			}
-
-			obj_idx++;
-			off += class->size;
-		}
-
-		/*
-		 * Handle the last (full or partial) object on this page.
-		 */
-		next_page = get_next_page(page);
-		if (next_page) {
-			if (!prev_free || prev_page_vaddr) {
-				/*
-				 * There is no free object in this page, so we can safely
-				 * unmap it.
-				 */
-				kunmap_atomic(vaddr);
-			} else {
-				/* update prev_page_vaddr since prev_free is on this page */
-				prev_page_vaddr = vaddr;
-			}
-		} else { /* this is the last page */
-			if (prev_free) {
-				/*
-				 * Reset OBJ_TAG_BITS bit to last link to tell
-				 * whether it's allocated object or not.
-				 */
-				prev_free->next = -1UL << OBJ_TAG_BITS;
-			}
-
-			/* unmap previous page (if not done yet) */
-			if (prev_page_vaddr) {
-				kunmap_atomic(prev_page_vaddr);
-				prev_page_vaddr = NULL;
-			}
-
-			kunmap_atomic(vaddr);
-		}
-
-		page = next_page;
-		off %= PAGE_SIZE;
-	}
-}
-
-static int zs_reclaim_page(struct zs_pool *pool, unsigned int retries)
-{
-	int i, obj_idx, ret = 0;
-	unsigned long handle;
-	struct zspage *zspage;
-	struct page *page;
-	int fullness;
-
-	/* Lock LRU and fullness list */
-	spin_lock(&pool->lock);
-	if (list_empty(&pool->lru)) {
-		spin_unlock(&pool->lock);
-		return -EINVAL;
-	}
-
-	for (i = 0; i < retries; i++) {
-		struct size_class *class;
-
-		zspage = list_last_entry(&pool->lru, struct zspage, lru);
-		list_del(&zspage->lru);
-
-		/* zs_free may free objects, but not the zspage and handles */
-		zspage->under_reclaim = true;
-
-		class = zspage_class(pool, zspage);
-		fullness = get_fullness_group(class, zspage);
-
-		/* Lock out object allocations and object compaction */
-		remove_zspage(class, zspage, fullness);
-
-		spin_unlock(&pool->lock);
-		cond_resched();
-
-		/* Lock backing pages into place */
-		lock_zspage(zspage);
-
-		obj_idx = 0;
-		page = get_first_page(zspage);
-		while (1) {
-			handle = find_alloced_obj(class, page, &obj_idx);
-			if (!handle) {
-				page = get_next_page(page);
-				if (!page)
-					break;
-				obj_idx = 0;
-				continue;
-			}
-
-			/*
-			 * This will write the object and call zs_free.
-			 *
-			 * zs_free will free the object, but the
-			 * under_reclaim flag prevents it from freeing
-			 * the zspage altogether. This is necessary so
-			 * that we can continue working with the
-			 * zspage potentially after the last object
-			 * has been freed.
-			 */
-			ret = pool->zpool_ops->evict(pool->zpool, handle);
-			if (ret)
-				goto next;
-
-			obj_idx++;
-		}
-
-next:
-		/* For freeing the zspage, or putting it back in the pool and LRU list. */
-		spin_lock(&pool->lock);
-		zspage->under_reclaim = false;
-
-		if (!get_zspage_inuse(zspage)) {
-			/*
-			 * Fullness went stale as zs_free() won't touch it
-			 * while the page is removed from the pool. Fix it
-			 * up for the check in __free_zspage().
-			 */
-			zspage->fullness = ZS_INUSE_RATIO_0;
-
-			__free_zspage(pool, class, zspage);
-			spin_unlock(&pool->lock);
-			return 0;
-		}
-
-		/*
-		 * Eviction fails on one of the handles, so we need to restore zspage.
-		 * We need to rebuild its freelist (and free stored deferred handles),
-		 * put it back to the correct size class, and add it to the LRU list.
-		 */
-		restore_freelist(pool, class, zspage);
-		putback_zspage(class, zspage);
-		list_add(&zspage->lru, &pool->lru);
-		unlock_zspage(zspage);
-	}
-
-	spin_unlock(&pool->lock);
-	return -EAGAIN;
-}
-#endif /* CONFIG_ZPOOL */
-
 static int __init zs_init(void)
 {
 	int ret;
-- 
2.34.1

[RFC PATCH 5/7] mm: zswap: remove shrink from zpool interface

[Domenico Cerasuolo]

Now that all three zswap backends have removed their shrink code, it is
no longer necessary for the zpool interface to include shrink/writeback
endpoints.

Signed-off-by: Domenico Cerasuolo <cerasuolodomenico@gmail.com>
---
 include/linux/zpool.h | 19 ++---------------
 mm/z3fold.c           |  5 +----
 mm/zbud.c             |  5 +----
 mm/zpool.c            | 48 ++-----------------------------------------
 mm/zsmalloc.c         |  5 +----
 mm/zswap.c            | 18 ++++------------
 6 files changed, 11 insertions(+), 89 deletions(-)

diff --git a/include/linux/zpool.h b/include/linux/zpool.h
index e8997010612a..6b15a4213de5 100644
--- a/include/linux/zpool.h
+++ b/include/linux/zpool.h
@@ -14,10 +14,6 @@
 
 struct zpool;
 
-struct zpool_ops {
-	int (*evict)(struct zpool *pool, unsigned long handle);
-};
-
 /*
  * Control how a handle is mapped.  It will be ignored if the
  * implementation does not support it.  Its use is optional.
@@ -40,7 +36,7 @@ enum zpool_mapmode {
 bool zpool_has_pool(char *type);
 
 struct zpool *zpool_create_pool(const char *type, const char *name,
-			gfp_t gfp, const struct zpool_ops *ops);
+			gfp_t gfp);
 
 const char *zpool_get_type(struct zpool *pool);
 
@@ -53,9 +49,6 @@ int zpool_malloc(struct zpool *pool, size_t size, gfp_t gfp,
 
 void zpool_free(struct zpool *pool, unsigned long handle);
 
-int zpool_shrink(struct zpool *pool, unsigned int pages,
-			unsigned int *reclaimed);
-
 void *zpool_map_handle(struct zpool *pool, unsigned long handle,
 			enum zpool_mapmode mm);
 
@@ -72,7 +65,6 @@ u64 zpool_get_total_size(struct zpool *pool);
  * @destroy:	destroy a pool.
  * @malloc:	allocate mem from a pool.
  * @free:	free mem from a pool.
- * @shrink:	shrink the pool.
  * @sleep_mapped: whether zpool driver can sleep during map.
  * @map:	map a handle.
  * @unmap:	unmap a handle.
@@ -87,10 +79,7 @@ struct zpool_driver {
 	atomic_t refcount;
 	struct list_head list;
 
-	void *(*create)(const char *name,
-			gfp_t gfp,
-			const struct zpool_ops *ops,
-			struct zpool *zpool);
+	void *(*create)(const char *name, gfp_t gfp);
 	void (*destroy)(void *pool);
 
 	bool malloc_support_movable;
@@ -98,9 +87,6 @@ struct zpool_driver {
 				unsigned long *handle);
 	void (*free)(void *pool, unsigned long handle);
 
-	int (*shrink)(void *pool, unsigned int pages,
-				unsigned int *reclaimed);
-
 	bool sleep_mapped;
 	void *(*map)(void *pool, unsigned long handle,
 				enum zpool_mapmode mm);
@@ -113,7 +99,6 @@ void zpool_register_driver(struct zpool_driver *driver);
 
 int zpool_unregister_driver(struct zpool_driver *driver);
 
-bool zpool_evictable(struct zpool *pool);
 bool zpool_can_sleep_mapped(struct zpool *pool);
 
 #endif
diff --git a/mm/z3fold.c b/mm/z3fold.c
index 4af8741553ac..e84de91ecccb 100644
--- a/mm/z3fold.c
+++ b/mm/z3fold.c
@@ -1364,9 +1364,7 @@ static const struct movable_operations z3fold_mops = {
  * zpool
  ****************/
 
-static void *z3fold_zpool_create(const char *name, gfp_t gfp,
-			       const struct zpool_ops *zpool_ops,
-			       struct zpool *zpool)
+static void *z3fold_zpool_create(const char *name, gfp_t gfp)
 {
 	return z3fold_create_pool(name, gfp);
 }
@@ -1409,7 +1407,6 @@ static struct zpool_driver z3fold_zpool_driver = {
 	.destroy =	z3fold_zpool_destroy,
 	.malloc =	z3fold_zpool_malloc,
 	.free =		z3fold_zpool_free,
-	.shrink =	NULL,
 	.map =		z3fold_zpool_map,
 	.unmap =	z3fold_zpool_unmap,
 	.total_size =	z3fold_zpool_total_size,
diff --git a/mm/zbud.c b/mm/zbud.c
index 19bc662ef5e9..2190cc1f37b3 100644
--- a/mm/zbud.c
+++ b/mm/zbud.c
@@ -380,9 +380,7 @@ static u64 zbud_get_pool_size(struct zbud_pool *pool)
  * zpool
  ****************/
 
-static void *zbud_zpool_create(const char *name, gfp_t gfp,
-			       const struct zpool_ops *zpool_ops,
-			       struct zpool *zpool)
+static void *zbud_zpool_create(const char *name, gfp_t gfp)
 {
 	return zbud_create_pool(gfp);
 }
@@ -425,7 +423,6 @@ static struct zpool_driver zbud_zpool_driver = {
 	.destroy =	zbud_zpool_destroy,
 	.malloc =	zbud_zpool_malloc,
 	.free =		zbud_zpool_free,
-	.shrink =	NULL,
 	.map =		zbud_zpool_map,
 	.unmap =	zbud_zpool_unmap,
 	.total_size =	zbud_zpool_total_size,
diff --git a/mm/zpool.c b/mm/zpool.c
index 6a19c4a58f77..846410479c2f 100644
--- a/mm/zpool.c
+++ b/mm/zpool.c
@@ -133,7 +133,6 @@ EXPORT_SYMBOL(zpool_has_pool);
  * @type:	The type of the zpool to create (e.g. zbud, zsmalloc)
  * @name:	The name of the zpool (e.g. zram0, zswap)
  * @gfp:	The GFP flags to use when allocating the pool.
- * @ops:	The optional ops callback.
  *
  * This creates a new zpool of the specified type.  The gfp flags will be
  * used when allocating memory, if the implementation supports it.  If the
@@ -145,8 +144,7 @@ EXPORT_SYMBOL(zpool_has_pool);
  *
  * Returns: New zpool on success, NULL on failure.
  */
-struct zpool *zpool_create_pool(const char *type, const char *name, gfp_t gfp,
-		const struct zpool_ops *ops)
+struct zpool *zpool_create_pool(const char *type, const char *name, gfp_t gfp)
 {
 	struct zpool_driver *driver;
 	struct zpool *zpool;
@@ -173,7 +171,7 @@ struct zpool *zpool_create_pool(const char *type, const char *name, gfp_t gfp,
 	}
 
 	zpool->driver = driver;
-	zpool->pool = driver->create(name, gfp, ops, zpool);
+	zpool->pool = driver->create(name, gfp);
 
 	if (!zpool->pool) {
 		pr_err("couldn't create %s pool\n", type);
@@ -279,30 +277,6 @@ void zpool_free(struct zpool *zpool, unsigned long handle)
 	zpool->driver->free(zpool->pool, handle);
 }
 
-/**
- * zpool_shrink() - Shrink the pool size
- * @zpool:	The zpool to shrink.
- * @pages:	The number of pages to shrink the pool.
- * @reclaimed:	The number of pages successfully evicted.
- *
- * This attempts to shrink the actual memory size of the pool
- * by evicting currently used handle(s).  If the pool was
- * created with no zpool_ops, or the evict call fails for any
- * of the handles, this will fail.  If non-NULL, the @reclaimed
- * parameter will be set to the number of pages reclaimed,
- * which may be more than the number of pages requested.
- *
- * Implementations must guarantee this to be thread-safe.
- *
- * Returns: 0 on success, negative value on error/failure.
- */
-int zpool_shrink(struct zpool *zpool, unsigned int pages,
-			unsigned int *reclaimed)
-{
-	return zpool->driver->shrink ?
-	       zpool->driver->shrink(zpool->pool, pages, reclaimed) : -EINVAL;
-}
-
 /**
  * zpool_map_handle() - Map a previously allocated handle into memory
  * @zpool:	The zpool that the handle was allocated from
@@ -359,24 +333,6 @@ u64 zpool_get_total_size(struct zpool *zpool)
 	return zpool->driver->total_size(zpool->pool);
 }
 
-/**
- * zpool_evictable() - Test if zpool is potentially evictable
- * @zpool:	The zpool to test
- *
- * Zpool is only potentially evictable when it's created with struct
- * zpool_ops.evict and its driver implements struct zpool_driver.shrink.
- *
- * However, it doesn't necessarily mean driver will use zpool_ops.evict
- * in its implementation of zpool_driver.shrink. It could do internal
- * defragmentation instead.
- *
- * Returns: true if potentially evictable; false otherwise.
- */
-bool zpool_evictable(struct zpool *zpool)
-{
-	return zpool->driver->shrink;
-}
-
 /**
  * zpool_can_sleep_mapped - Test if zpool can sleep when do mapped.
  * @zpool:	The zpool to test
diff --git a/mm/zsmalloc.c b/mm/zsmalloc.c
index c87a60514f21..316eb2784643 100644
--- a/mm/zsmalloc.c
+++ b/mm/zsmalloc.c
@@ -370,9 +370,7 @@ static void record_obj(unsigned long handle, unsigned long obj)
 
 #ifdef CONFIG_ZPOOL
 
-static void *zs_zpool_create(const char *name, gfp_t gfp,
-			     const struct zpool_ops *zpool_ops,
-			     struct zpool *zpool)
+static void *zs_zpool_create(const char *name, gfp_t gfp)
 {
 	/*
 	 * Ignore global gfp flags: zs_malloc() may be invoked from
@@ -439,7 +437,6 @@ static struct zpool_driver zs_zpool_driver = {
 	.malloc_support_movable = true,
 	.malloc =		  zs_zpool_malloc,
 	.free =			  zs_zpool_free,
-	.shrink =		  NULL,
 	.map =			  zs_zpool_map,
 	.unmap =		  zs_zpool_unmap,
 	.total_size =		  zs_zpool_total_size,
diff --git a/mm/zswap.c b/mm/zswap.c
index 80d7780bf066..f5fcb07a181d 100644
--- a/mm/zswap.c
+++ b/mm/zswap.c
@@ -248,10 +248,6 @@ static int zswap_writeback_entry(struct zpool *pool, unsigned long handle);
 static int zswap_pool_get(struct zswap_pool *pool);
 static void zswap_pool_put(struct zswap_pool *pool);
 
-static const struct zpool_ops zswap_zpool_ops = {
-	.evict = zswap_writeback_entry
-};
-
 static bool zswap_is_full(void)
 {
 	return totalram_pages() * zswap_max_pool_percent / 100 <
@@ -645,12 +641,8 @@ static void shrink_worker(struct work_struct *w)
 						shrink_work);
 	int ret, failures = 0;
 
-	/* zpool_evictable will be removed once all 3 backends have migrated*/
 	do {
-		if (zpool_evictable(pool->zpool))
-			ret = zpool_shrink(pool->zpool, 1, NULL);
-		else
-			ret = zswap_shrink(pool);
+		ret = zswap_shrink(pool);
 		if (ret) {
 			zswap_reject_reclaim_fail++;
 			if (ret != -EAGAIN)
@@ -688,7 +680,7 @@ static struct zswap_pool *zswap_pool_create(char *type, char *compressor)
 	/* unique name for each pool specifically required by zsmalloc */
 	snprintf(name, 38, "zswap%x", atomic_inc_return(&zswap_pools_count));
 
-	pool->zpool = zpool_create_pool(type, name, gfp, &zswap_zpool_ops);
+	pool->zpool = zpool_create_pool(type, name, gfp);
 	if (!pool->zpool) {
 		pr_err("%s zpool not available\n", type);
 		goto error;
@@ -1375,8 +1367,7 @@ static int zswap_frontswap_store(unsigned type, pgoff_t offset,
 		}
 	} while (ret == -EEXIST);
 	INIT_LIST_HEAD(&entry->lru);
-	/* zpool_evictable will be removed once all 3 backends have migrated*/
-	if (entry->length && !zpool_evictable(entry->pool->zpool)) {
+	if (entry->length) {
 		spin_lock(&entry->pool->lock);
 		list_add(&entry->lru, &entry->pool->lru);
 		spin_unlock(&entry->pool->lock);
@@ -1482,8 +1473,7 @@ static int zswap_frontswap_load(unsigned type, pgoff_t offset,
 freeentry:
 	spin_lock(&tree->lock);
 	zswap_entry_put(tree, entry);
-	/* zpool_evictable will be removed once all 3 backends have migrated*/
-	if (entry->length && !zpool_evictable(entry->pool->zpool)) {
+	if (entry->length) {
 		spin_lock(&entry->pool->lock);
 		list_move(&entry->lru, &entry->pool->lru);
 		spin_unlock(&entry->pool->lock);
-- 
2.34.1

[RFC PATCH 6/7] mm: zswap: simplify writeback function

[Domenico Cerasuolo]

Previously, in zswap, the writeback function was passed to zpool drivers
for their usage in calling the writeback operation. However, since the
drivers did not possess references to entries and the function was
specifically designed to work with handles, the writeback process has
been modified to occur directly within zswap. Consequently, this change
allows for some simplification of the writeback function, taking into
account the updated workflow.

Signed-off-by: Domenico Cerasuolo <cerasuolodomenico@gmail.com>
---
 mm/zswap.c | 50 +++++++++++---------------------------------------
 1 file changed, 11 insertions(+), 39 deletions(-)

diff --git a/mm/zswap.c b/mm/zswap.c
index f5fcb07a181d..8ee30a989508 100644
--- a/mm/zswap.c
+++ b/mm/zswap.c
@@ -244,7 +244,8 @@ static bool zswap_has_pool;
 	pr_debug("%s pool %s/%s\n", msg, (p)->tfm_name,		\
 		 zpool_get_type((p)->zpool))
 
-static int zswap_writeback_entry(struct zpool *pool, unsigned long handle);
+static int zswap_writeback_entry(struct zswap_entry *entry, struct zswap_header *zhdr,
+				 struct zswap_tree *tree);
 static int zswap_pool_get(struct zswap_pool *pool);
 static void zswap_pool_put(struct zswap_pool *pool);
 
@@ -621,7 +622,7 @@ static int zswap_shrink(struct zswap_pool *pool)
 	}
 	spin_unlock(&tree->lock);
 
-	ret = zswap_writeback_entry(pool->zpool, lru_entry->handle);
+	ret = zswap_writeback_entry(lru_entry, zhdr, tree);
 
 	spin_lock(&tree->lock);
 	if (ret) {
@@ -1028,16 +1029,14 @@ static int zswap_get_swap_cache_page(swp_entry_t entry,
  * the swap cache, the compressed version stored by zswap can be
  * freed.
  */
-static int zswap_writeback_entry(struct zpool *pool, unsigned long handle)
+static int zswap_writeback_entry(struct zswap_entry *entry, struct zswap_header *zhdr,
+				 struct zswap_tree *tree)
 {
-	struct zswap_header *zhdr;
-	swp_entry_t swpentry;
-	struct zswap_tree *tree;
-	pgoff_t offset;
-	struct zswap_entry *entry;
+	swp_entry_t swpentry = zhdr->swpentry;
 	struct page *page;
 	struct scatterlist input, output;
 	struct crypto_acomp_ctx *acomp_ctx;
+	struct zpool *pool = entry->pool->zpool;
 
 	u8 *src, *tmp = NULL;
 	unsigned int dlen;
@@ -1052,25 +1051,6 @@ static int zswap_writeback_entry(struct zpool *pool, unsigned long handle)
 			return -ENOMEM;
 	}
 
-	/* extract swpentry from data */
-	zhdr = zpool_map_handle(pool, handle, ZPOOL_MM_RO);
-	swpentry = zhdr->swpentry; /* here */
-	tree = zswap_trees[swp_type(swpentry)];
-	offset = swp_offset(swpentry);
-	zpool_unmap_handle(pool, handle);
-
-	/* find and ref zswap entry */
-	spin_lock(&tree->lock);
-	entry = zswap_entry_find_get(&tree->rbroot, offset);
-	if (!entry) {
-		/* entry was invalidated */
-		spin_unlock(&tree->lock);
-		kfree(tmp);
-		return 0;
-	}
-	spin_unlock(&tree->lock);
-	BUG_ON(offset != entry->offset);
-
 	/* try to allocate swap cache page */
 	switch (zswap_get_swap_cache_page(swpentry, &page)) {
 	case ZSWAP_SWAPCACHE_FAIL: /* no memory or invalidate happened */
@@ -1104,12 +1084,12 @@ static int zswap_writeback_entry(struct zpool *pool, unsigned long handle)
 		acomp_ctx = raw_cpu_ptr(entry->pool->acomp_ctx);
 		dlen = PAGE_SIZE;
 
-		zhdr = zpool_map_handle(pool, handle, ZPOOL_MM_RO);
+		zhdr = zpool_map_handle(pool, entry->handle, ZPOOL_MM_RO);
 		src = (u8 *)zhdr + sizeof(struct zswap_header);
 		if (!zpool_can_sleep_mapped(pool)) {
 			memcpy(tmp, src, entry->length);
 			src = tmp;
-			zpool_unmap_handle(pool, handle);
+			zpool_unmap_handle(pool, entry->handle);
 		}
 
 		mutex_lock(acomp_ctx->mutex);
@@ -1124,7 +1104,7 @@ static int zswap_writeback_entry(struct zpool *pool, unsigned long handle)
 		if (!zpool_can_sleep_mapped(pool))
 			kfree(tmp);
 		else
-			zpool_unmap_handle(pool, handle);
+			zpool_unmap_handle(pool, entry->handle);
 
 		BUG_ON(ret);
 		BUG_ON(dlen != PAGE_SIZE);
@@ -1142,9 +1122,6 @@ static int zswap_writeback_entry(struct zpool *pool, unsigned long handle)
 	zswap_written_back_pages++;
 
 	spin_lock(&tree->lock);
-	/* drop local reference */
-	zswap_entry_put(tree, entry);
-
 	/*
 	* There are two possible situations for entry here:
 	* (1) refcount is 1(normal case),  entry is valid and on the tree
@@ -1152,7 +1129,7 @@ static int zswap_writeback_entry(struct zpool *pool, unsigned long handle)
 	*     because invalidate happened during writeback
 	*  search the tree and free the entry if find entry
 	*/
-	if (entry == zswap_rb_search(&tree->rbroot, offset))
+	if (entry == zswap_rb_search(&tree->rbroot, swp_offset(swpentry)))
 		zswap_entry_put(tree, entry);
 	spin_unlock(&tree->lock);
 
@@ -1166,13 +1143,8 @@ static int zswap_writeback_entry(struct zpool *pool, unsigned long handle)
 	* if we get here due to ZSWAP_SWAPCACHE_EXIST
 	* a load may be happening concurrently.
 	* it is safe and okay to not free the entry.
-	* if we free the entry in the following put
 	* it is also okay to return !0
 	*/
-	spin_lock(&tree->lock);
-	zswap_entry_put(tree, entry);
-	spin_unlock(&tree->lock);
-
 	return ret;
 }
 
-- 
2.34.1

[RFC PATCH 7/7] mm: zswap: remove zswap_header

[Domenico Cerasuolo]

Previously, zswap_header served the purpose of storing the swpentry
within zpool pages. This allowed zpool implementations to pass relevant
information to the writeback function. However, with the current
implementation, writeback is directly handled within zswap.
Consequently, there is no longer a necessity for zswap_header, as the
swp_entry_t can be stored directly in zswap_entry.

Suggested-by: Yosry Ahmed <yosryahmed@google.com>
Signed-off-by: Domenico Cerasuolo <cerasuolodomenico@gmail.com>
---
 mm/zswap.c | 52 ++++++++++++++++++++++------------------------------
 1 file changed, 22 insertions(+), 30 deletions(-)

diff --git a/mm/zswap.c b/mm/zswap.c
index 8ee30a989508..093beaeaaebc 100644
--- a/mm/zswap.c
+++ b/mm/zswap.c
@@ -187,7 +187,7 @@ struct zswap_pool {
  */
 struct zswap_entry {
 	struct rb_node rbnode;
-	pgoff_t offset;
+	swp_entry_t swpentry;
 	int refcount;
 	unsigned int length;
 	struct zswap_pool *pool;
@@ -199,10 +199,6 @@ struct zswap_entry {
 	struct list_head lru;
 };
 
-struct zswap_header {
-	swp_entry_t swpentry;
-};
-
 /*
  * The tree lock in the zswap_tree struct protects a few things:
  * - the rbtree
@@ -244,7 +240,7 @@ static bool zswap_has_pool;
 	pr_debug("%s pool %s/%s\n", msg, (p)->tfm_name,		\
 		 zpool_get_type((p)->zpool))
 
-static int zswap_writeback_entry(struct zswap_entry *entry, struct zswap_header *zhdr,
+static int zswap_writeback_entry(struct zswap_entry *entry,
 				 struct zswap_tree *tree);
 static int zswap_pool_get(struct zswap_pool *pool);
 static void zswap_pool_put(struct zswap_pool *pool);
@@ -305,12 +301,14 @@ static struct zswap_entry *zswap_rb_search(struct rb_root *root, pgoff_t offset)
 {
 	struct rb_node *node = root->rb_node;
 	struct zswap_entry *entry;
+	pgoff_t entry_offset;
 
 	while (node) {
 		entry = rb_entry(node, struct zswap_entry, rbnode);
-		if (entry->offset > offset)
+		entry_offset = swp_offset(entry->swpentry);
+		if (entry_offset > offset)
 			node = node->rb_left;
-		else if (entry->offset < offset)
+		else if (entry_offset < offset)
 			node = node->rb_right;
 		else
 			return entry;
@@ -327,13 +325,15 @@ static int zswap_rb_insert(struct rb_root *root, struct zswap_entry *entry,
 {
 	struct rb_node **link = &root->rb_node, *parent = NULL;
 	struct zswap_entry *myentry;
+	pgoff_t myentry_offset, entry_offset = swp_offset(entry->swpentry);
 
 	while (*link) {
 		parent = *link;
 		myentry = rb_entry(parent, struct zswap_entry, rbnode);
-		if (myentry->offset > entry->offset)
+		myentry_offset = swp_offset(myentry->swpentry);
+		if (myentry_offset > entry_offset)
 			link = &(*link)->rb_left;
-		else if (myentry->offset < entry->offset)
+		else if (myentry_offset < entry_offset)
 			link = &(*link)->rb_right;
 		else {
 			*dupentry = myentry;
@@ -592,7 +592,6 @@ static struct zswap_pool *zswap_pool_find_get(char *type, char *compressor)
 static int zswap_shrink(struct zswap_pool *pool)
 {
 	struct zswap_entry *lru_entry, *tree_entry = NULL;
-	struct zswap_header *zhdr;
 	struct zswap_tree *tree;
 	swp_entry_t swpentry;
 	int ret;
@@ -605,10 +604,8 @@ static int zswap_shrink(struct zswap_pool *pool)
 	}
 	lru_entry = list_last_entry(&pool->lru, struct zswap_entry, lru);
 	list_del_init(&lru_entry->lru);
-	zhdr = zpool_map_handle(pool->zpool, lru_entry->handle, ZPOOL_MM_RO);
-	tree = zswap_trees[swp_type(zhdr->swpentry)];
-	zpool_unmap_handle(pool->zpool, lru_entry->handle);
-	swpentry = zhdr->swpentry;
+	swpentry = lru_entry->swpentry;
+	tree = zswap_trees[swp_type(lru_entry->swpentry)];
 	spin_unlock(&pool->lock);
 
 	/* hold a reference from tree so it won't be freed during writeback */
@@ -622,7 +619,7 @@ static int zswap_shrink(struct zswap_pool *pool)
 	}
 	spin_unlock(&tree->lock);
 
-	ret = zswap_writeback_entry(lru_entry, zhdr, tree);
+	ret = zswap_writeback_entry(lru_entry, tree);
 
 	spin_lock(&tree->lock);
 	if (ret) {
@@ -1029,10 +1026,10 @@ static int zswap_get_swap_cache_page(swp_entry_t entry,
  * the swap cache, the compressed version stored by zswap can be
  * freed.
  */
-static int zswap_writeback_entry(struct zswap_entry *entry, struct zswap_header *zhdr,
+static int zswap_writeback_entry(struct zswap_entry *entry,
 				 struct zswap_tree *tree)
 {
-	swp_entry_t swpentry = zhdr->swpentry;
+	swp_entry_t swpentry = entry->swpentry;
 	struct page *page;
 	struct scatterlist input, output;
 	struct crypto_acomp_ctx *acomp_ctx;
@@ -1072,7 +1069,7 @@ static int zswap_writeback_entry(struct zswap_entry *entry, struct zswap_header
 		 * writing.
 		 */
 		spin_lock(&tree->lock);
-		if (zswap_rb_search(&tree->rbroot, entry->offset) != entry) {
+		if (zswap_rb_search(&tree->rbroot, swp_offset(entry->swpentry)) != entry) {
 			spin_unlock(&tree->lock);
 			delete_from_swap_cache(page_folio(page));
 			ret = -ENOMEM;
@@ -1084,8 +1081,7 @@ static int zswap_writeback_entry(struct zswap_entry *entry, struct zswap_header
 		acomp_ctx = raw_cpu_ptr(entry->pool->acomp_ctx);
 		dlen = PAGE_SIZE;
 
-		zhdr = zpool_map_handle(pool, entry->handle, ZPOOL_MM_RO);
-		src = (u8 *)zhdr + sizeof(struct zswap_header);
+		src = zpool_map_handle(pool, entry->handle, ZPOOL_MM_RO);
 		if (!zpool_can_sleep_mapped(pool)) {
 			memcpy(tmp, src, entry->length);
 			src = tmp;
@@ -1192,11 +1188,10 @@ static int zswap_frontswap_store(unsigned type, pgoff_t offset,
 	struct obj_cgroup *objcg = NULL;
 	struct zswap_pool *pool;
 	int ret;
-	unsigned int hlen, dlen = PAGE_SIZE;
+	unsigned int dlen = PAGE_SIZE;
 	unsigned long handle, value;
 	char *buf;
 	u8 *src, *dst;
-	struct zswap_header zhdr = { .swpentry = swp_entry(type, offset) };
 	gfp_t gfp;
 
 	/* THP isn't supported */
@@ -1241,7 +1236,7 @@ static int zswap_frontswap_store(unsigned type, pgoff_t offset,
 		src = kmap_atomic(page);
 		if (zswap_is_page_same_filled(src, &value)) {
 			kunmap_atomic(src);
-			entry->offset = offset;
+			entry->swpentry = swp_entry(type, offset);
 			entry->length = 0;
 			entry->value = value;
 			atomic_inc(&zswap_same_filled_pages);
@@ -1295,11 +1290,10 @@ static int zswap_frontswap_store(unsigned type, pgoff_t offset,
 	}
 
 	/* store */
-	hlen = sizeof(zhdr);
 	gfp = __GFP_NORETRY | __GFP_NOWARN | __GFP_KSWAPD_RECLAIM;
 	if (zpool_malloc_support_movable(entry->pool->zpool))
 		gfp |= __GFP_HIGHMEM | __GFP_MOVABLE;
-	ret = zpool_malloc(entry->pool->zpool, hlen + dlen, gfp, &handle);
+	ret = zpool_malloc(entry->pool->zpool, dlen, gfp, &handle);
 	if (ret == -ENOSPC) {
 		zswap_reject_compress_poor++;
 		goto put_dstmem;
@@ -1309,13 +1303,12 @@ static int zswap_frontswap_store(unsigned type, pgoff_t offset,
 		goto put_dstmem;
 	}
 	buf = zpool_map_handle(entry->pool->zpool, handle, ZPOOL_MM_WO);
-	memcpy(buf, &zhdr, hlen);
-	memcpy(buf + hlen, dst, dlen);
+	memcpy(buf, dst, dlen);
 	zpool_unmap_handle(entry->pool->zpool, handle);
 	mutex_unlock(acomp_ctx->mutex);
 
 	/* populate entry */
-	entry->offset = offset;
+	entry->swpentry = swp_entry(type, offset);
 	entry->handle = handle;
 	entry->length = dlen;
 
@@ -1415,7 +1408,6 @@ static int zswap_frontswap_load(unsigned type, pgoff_t offset,
 	/* decompress */
 	dlen = PAGE_SIZE;
 	src = zpool_map_handle(entry->pool->zpool, entry->handle, ZPOOL_MM_RO);
-	src += sizeof(struct zswap_header);
 
 	if (!zpool_can_sleep_mapped(entry->pool->zpool)) {
 		memcpy(tmp, src, entry->length);
-- 
2.34.1

Re: [RFC PATCH 1/7] mm: zswap: add pool shrinking mechanism

[Johannes Weiner]

Hi Domenico,

On Mon, Jun 05, 2023 at 10:54:13AM +0200, Domenico Cerasuolo wrote:
> @@ -364,6 +369,9 @@ static void zswap_free_entry(struct zswap_entry *entry)
>  	if (!entry->length)
>  		atomic_dec(&zswap_same_filled_pages);
>  	else {
> +		spin_lock(&entry->pool->lock);
> +		list_del_init(&entry->lru);
> +		spin_unlock(&entry->pool->lock);

This should be list_del(), as the entry is freed right after this and
the list isn't reused anymore.

The slab memory is recycled, but the allocation site (commented on
below) doesn't need the list initialized.

However, I think it also needs to check !zpool_evictable(). If
alloc/store doesn't do the list_add(), this would be a double delete.

> @@ -584,14 +592,65 @@ static struct zswap_pool *zswap_pool_find_get(char *type, char *compressor)
>  	return NULL;
>  }
>  
> +static int zswap_shrink(struct zswap_pool *pool)
> +{
> +	struct zswap_entry *lru_entry, *tree_entry = NULL;
> +	struct zswap_header *zhdr;
> +	struct zswap_tree *tree;
> +	swp_entry_t swpentry;
> +	int ret;
> +
> +	/* get a reclaimable entry from LRU */
> +	spin_lock(&pool->lock);
> +	if (list_empty(&pool->lru)) {
> +		spin_unlock(&pool->lock);
> +		return -EINVAL;
> +	}
> +	lru_entry = list_last_entry(&pool->lru, struct zswap_entry, lru);
> +	list_del_init(&lru_entry->lru);
> +	zhdr = zpool_map_handle(pool->zpool, lru_entry->handle, ZPOOL_MM_RO);
> +	tree = zswap_trees[swp_type(zhdr->swpentry)];
> +	zpool_unmap_handle(pool->zpool, lru_entry->handle);
> +	swpentry = zhdr->swpentry;
> +	spin_unlock(&pool->lock);

Once the pool lock is dropped, the lru_entry might get freed. But the
swpentry is copied to the stack, and lru_entry isn't deref'd again
until the entry is verified to still be alive in the tree.

This could use a comment.

> +	/* hold a reference from tree so it won't be freed during writeback */
> +	spin_lock(&tree->lock);
> +	tree_entry = zswap_entry_find_get(&tree->rbroot, swp_offset(swpentry));
> +	if (tree_entry != lru_entry) {
> +		if (tree_entry)
> +			zswap_entry_put(tree, tree_entry);
> +		spin_unlock(&tree->lock);
> +		return -EAGAIN;
> +	}
> +	spin_unlock(&tree->lock);

It's pretty outrageous how much simpler this is compared to the
<backend>_reclaim_page() functions! The backends have to jump through
a lot of hoops to serialize against freeing, whereas zswap can simply
hold a reference. This is clearly a much better design.

> +	ret = zswap_writeback_entry(pool->zpool, lru_entry->handle);
> +
> +	spin_lock(&tree->lock);
> +	if (ret) {
> +		spin_lock(&pool->lock);
> +		list_move(&lru_entry->lru, &pool->lru);
> +		spin_unlock(&pool->lock);
> +	}
> +	zswap_entry_put(tree, tree_entry);
> +	spin_unlock(&tree->lock);
> +
> +	return ret ? -EAGAIN : 0;
> +}
> +
>  static void shrink_worker(struct work_struct *w)
>  {
>  	struct zswap_pool *pool = container_of(w, typeof(*pool),
>  						shrink_work);
>  	int ret, failures = 0;
>  
> +	/* zpool_evictable will be removed once all 3 backends have migrated*/

Missing space between text and */

>  	do {
> -		ret = zpool_shrink(pool->zpool, 1, NULL);
> +		if (zpool_evictable(pool->zpool))
> +			ret = zpool_shrink(pool->zpool, 1, NULL);
> +		else
> +			ret = zswap_shrink(pool);
>  		if (ret) {
>  			zswap_reject_reclaim_fail++;
>  			if (ret != -EAGAIN)
> @@ -655,6 +714,8 @@ static struct zswap_pool *zswap_pool_create(char *type, char *compressor)
>  	 */
>  	kref_init(&pool->kref);
>  	INIT_LIST_HEAD(&pool->list);
> +	INIT_LIST_HEAD(&pool->lru);
> +	spin_lock_init(&pool->lock);
>  	INIT_WORK(&pool->shrink_work, shrink_worker);
>  
>  	zswap_pool_debug("created", pool);
> @@ -1270,7 +1331,7 @@ static int zswap_frontswap_store(unsigned type, pgoff_t offset,
>  	}
>  
>  	/* store */
> -	hlen = zpool_evictable(entry->pool->zpool) ? sizeof(zhdr) : 0;
> +	hlen = sizeof(zhdr);
>  	gfp = __GFP_NORETRY | __GFP_NOWARN | __GFP_KSWAPD_RECLAIM;
>  	if (zpool_malloc_support_movable(entry->pool->zpool))
>  		gfp |= __GFP_HIGHMEM | __GFP_MOVABLE;
> @@ -1313,6 +1374,13 @@ static int zswap_frontswap_store(unsigned type, pgoff_t offset,
>  			zswap_entry_put(tree, dupentry);
>  		}
>  	} while (ret == -EEXIST);
> +	INIT_LIST_HEAD(&entry->lru);

The list_add() below initializes the entry, so this shouldn't be
needed.

> +	/* zpool_evictable will be removed once all 3 backends have migrated*/
> +	if (entry->length && !zpool_evictable(entry->pool->zpool)) {
> +		spin_lock(&entry->pool->lock);
> +		list_add(&entry->lru, &entry->pool->lru);
> +		spin_unlock(&entry->pool->lock);
> +	}
>  	spin_unlock(&tree->lock);
>  
>  	/* update stats */

Re: [RFC PATCH 4/7] mm: zswap: remove page reclaim logic from zsmalloc

[Johannes Weiner]

On Mon, Jun 05, 2023 at 10:54:16AM +0200, Domenico Cerasuolo wrote:
> @@ -884,14 +842,6 @@ static inline bool obj_allocated(struct page *page, void *obj, unsigned long *ph
>  	return obj_tagged(page, obj, phandle, OBJ_ALLOCATED_TAG);
>  }
>  
> -#ifdef CONFIG_ZPOOL
> -static bool obj_stores_deferred_handle(struct page *page, void *obj,
> -		unsigned long *phandle)
> -{
> -	return obj_tagged(page, obj, phandle, OBJ_DEFERRED_HANDLE_TAG);
> -}
> -#endif

You can actually remove even more here.

The entire concept of deferred_handle is about serializing free with
reclaim. It can all go: OBJ_DEFERRED_HANDLE_TAG, the member in struct
link_free, this function here, find_deferred_handle_obj() (declaration
and implementation), free_handles(), and the deferred handle bits in
obj_free() including the handle parameter itself.

Re: [RFC PATCH 6/7] mm: zswap: simplify writeback function

[Johannes Weiner]

On Mon, Jun 05, 2023 at 10:54:18AM +0200, Domenico Cerasuolo wrote:
> @@ -1142,9 +1122,6 @@ static int zswap_writeback_entry(struct zpool *pool, unsigned long handle)
>  	zswap_written_back_pages++;
>  
>  	spin_lock(&tree->lock);
> -	/* drop local reference */
> -	zswap_entry_put(tree, entry);
> -
>  	/*
>  	* There are two possible situations for entry here:
>  	* (1) refcount is 1(normal case),  entry is valid and on the tree
> @@ -1152,7 +1129,7 @@ static int zswap_writeback_entry(struct zpool *pool, unsigned long handle)
>  	*     because invalidate happened during writeback
>  	*  search the tree and free the entry if find entry
>  	*/
> -	if (entry == zswap_rb_search(&tree->rbroot, offset))
> +	if (entry == zswap_rb_search(&tree->rbroot, swp_offset(swpentry)))
>  		zswap_entry_put(tree, entry);
>  	spin_unlock(&tree->lock);

This can be moved to zswap_shrink() as well. It already has a
post-writeback tree->lock section for lru putback and dropping its
local reference, it should do this as well.

Writeback is then is done after it bumped zswap_written_pages.

Re: [RFC PATCH 4/7] mm: zswap: remove page reclaim logic from zsmalloc

[Nhat Pham]

On Mon, Jun 5, 2023 at 8:37 AM Johannes Weiner <hannes@cmpxchg.org> wrote:
>
> On Mon, Jun 05, 2023 at 10:54:16AM +0200, Domenico Cerasuolo wrote:
> > @@ -884,14 +842,6 @@ static inline bool obj_allocated(struct page *page, void *obj, unsigned long *ph
> >       return obj_tagged(page, obj, phandle, OBJ_ALLOCATED_TAG);
> >  }
> >
> > -#ifdef CONFIG_ZPOOL
> > -static bool obj_stores_deferred_handle(struct page *page, void *obj,
> > -             unsigned long *phandle)
> > -{
> > -     return obj_tagged(page, obj, phandle, OBJ_DEFERRED_HANDLE_TAG);
> > -}
> > -#endif
>
> You can actually remove even more here.
>
> The entire concept of deferred_handle is about serializing free with
> reclaim. It can all go: OBJ_DEFERRED_HANDLE_TAG, the member in struct
> link_free, this function here, find_deferred_handle_obj() (declaration
> and implementation), free_handles(), and the deferred handle bits in
> obj_free() including the handle parameter itself.

For more context on this:
https://lore.kernel.org/all/20230110231701.326724-1-nphamcs@gmail.com/T/#u

Re: [RFC PATCH 4/7] mm: zswap: remove page reclaim logic from zsmalloc

[Nhat Pham]

On Mon, Jun 5, 2023 at 1:54 AM Domenico Cerasuolo
<cerasuolodomenico@gmail.com> wrote:
>
> With the recent enhancement to zswap enabling direct page writeback, the
> need for the shrink code in zsmalloc has become obsolete. As a result,
> this commit removes the page reclaim logic from zsmalloc entirely.
>
> Signed-off-by: Domenico Cerasuolo <cerasuolodomenico@gmail.com>
> ---
>  mm/zsmalloc.c | 291 +-------------------------------------------------
>  1 file changed, 2 insertions(+), 289 deletions(-)
>
> diff --git a/mm/zsmalloc.c b/mm/zsmalloc.c
> index 02f7f414aade..c87a60514f21 100644
> --- a/mm/zsmalloc.c
> +++ b/mm/zsmalloc.c
> @@ -250,13 +250,6 @@ struct zs_pool {
>         /* Compact classes */
>         struct shrinker shrinker;
>
> -#ifdef CONFIG_ZPOOL
> -       /* List tracking the zspages in LRU order by most recently added object */
> -       struct list_head lru;
> -       struct zpool *zpool;
> -       const struct zpool_ops *zpool_ops;
> -#endif
> -
>  #ifdef CONFIG_ZSMALLOC_STAT
>         struct dentry *stat_dentry;
>  #endif
> @@ -279,13 +272,6 @@ struct zspage {
>         unsigned int freeobj;
>         struct page *first_page;
>         struct list_head list; /* fullness list */
> -
> -#ifdef CONFIG_ZPOOL
> -       /* links the zspage to the lru list in the pool */
> -       struct list_head lru;
> -       bool under_reclaim;
> -#endif
> -
>         struct zs_pool *pool;
>         rwlock_t lock;
>  };
> @@ -393,14 +379,7 @@ static void *zs_zpool_create(const char *name, gfp_t gfp,
>          * different contexts and its caller must provide a valid
>          * gfp mask.
>          */
> -       struct zs_pool *pool = zs_create_pool(name);
> -
> -       if (pool) {
> -               pool->zpool = zpool;
> -               pool->zpool_ops = zpool_ops;
> -       }
> -
> -       return pool;
> +       return zs_create_pool(name);
>  }
>
>  static void zs_zpool_destroy(void *pool)
> @@ -422,27 +401,6 @@ static void zs_zpool_free(void *pool, unsigned long handle)
>         zs_free(pool, handle);
>  }
>
> -static int zs_reclaim_page(struct zs_pool *pool, unsigned int retries);
> -
> -static int zs_zpool_shrink(void *pool, unsigned int pages,
> -                       unsigned int *reclaimed)
> -{
> -       unsigned int total = 0;
> -       int ret = -EINVAL;
> -
> -       while (total < pages) {
> -               ret = zs_reclaim_page(pool, 8);
> -               if (ret < 0)
> -                       break;
> -               total++;
> -       }
> -
> -       if (reclaimed)
> -               *reclaimed = total;
> -
> -       return ret;
> -}
> -
>  static void *zs_zpool_map(void *pool, unsigned long handle,
>                         enum zpool_mapmode mm)
>  {
> @@ -481,7 +439,7 @@ static struct zpool_driver zs_zpool_driver = {
>         .malloc_support_movable = true,
>         .malloc =                 zs_zpool_malloc,
>         .free =                   zs_zpool_free,
> -       .shrink =                 zs_zpool_shrink,
> +       .shrink =                 NULL,
>         .map =                    zs_zpool_map,
>         .unmap =                  zs_zpool_unmap,
>         .total_size =             zs_zpool_total_size,
> @@ -884,14 +842,6 @@ static inline bool obj_allocated(struct page *page, void *obj, unsigned long *ph
>         return obj_tagged(page, obj, phandle, OBJ_ALLOCATED_TAG);
>  }
>
> -#ifdef CONFIG_ZPOOL
> -static bool obj_stores_deferred_handle(struct page *page, void *obj,
> -               unsigned long *phandle)
> -{
> -       return obj_tagged(page, obj, phandle, OBJ_DEFERRED_HANDLE_TAG);
> -}
> -#endif
> -
>  static void reset_page(struct page *page)
>  {
>         __ClearPageMovable(page);
> @@ -1006,9 +956,6 @@ static void free_zspage(struct zs_pool *pool, struct size_class *class,
>         }
>
>         remove_zspage(class, zspage, ZS_INUSE_RATIO_0);
> -#ifdef CONFIG_ZPOOL
> -       list_del(&zspage->lru);
> -#endif
>         __free_zspage(pool, class, zspage);
>  }
>
> @@ -1054,11 +1001,6 @@ static void init_zspage(struct size_class *class, struct zspage *zspage)
>                 off %= PAGE_SIZE;
>         }
>
> -#ifdef CONFIG_ZPOOL
> -       INIT_LIST_HEAD(&zspage->lru);
> -       zspage->under_reclaim = false;
> -#endif
> -
>         set_freeobj(zspage, 0);
>  }
>
> @@ -1525,13 +1467,6 @@ unsigned long zs_malloc(struct zs_pool *pool, size_t size, gfp_t gfp)
>         /* We completely set up zspage so mark them as movable */
>         SetZsPageMovable(pool, zspage);
>  out:
> -#ifdef CONFIG_ZPOOL
> -       /* Add/move zspage to beginning of LRU */
> -       if (!list_empty(&zspage->lru))
> -               list_del(&zspage->lru);
> -       list_add(&zspage->lru, &pool->lru);
> -#endif
> -
>         spin_unlock(&pool->lock);
>
>         return handle;
> @@ -1600,20 +1535,6 @@ void zs_free(struct zs_pool *pool, unsigned long handle)
>         class = zspage_class(pool, zspage);
>
>         class_stat_dec(class, ZS_OBJS_INUSE, 1);
> -
> -#ifdef CONFIG_ZPOOL
> -       if (zspage->under_reclaim) {
> -               /*
> -                * Reclaim needs the handles during writeback. It'll free
> -                * them along with the zspage when it's done with them.
> -                *
> -                * Record current deferred handle in the object's header.
> -                */
> -               obj_free(class->size, obj, &handle);
> -               spin_unlock(&pool->lock);
> -               return;
> -       }
> -#endif
>         obj_free(class->size, obj, NULL);
>
>         fullness = fix_fullness_group(class, zspage);
> @@ -1890,23 +1811,6 @@ static void lock_zspage(struct zspage *zspage)
>  }
>  #endif /* defined(CONFIG_ZPOOL) || defined(CONFIG_COMPACTION) */

If I recall correctly, the defined(CONFIG_ZPOOL) condition is
only needed for the lock_zspage() call in zs_reclaim_page().

You might be able to get away with just
#ifdef CONFIG_COMPACTION if you're removing writeback.

Do fact-check me of course - try to build with these CONFIGs turned
off and see. I'm surprised kernel test robot has not complained about
unused static function lock_zspage() in the case
CONFIG_ZPOOL && !CONFIG_COMPACTION

>
> -#ifdef CONFIG_ZPOOL
> -/*
> - * Unlocks all the pages of the zspage.
> - *
> - * pool->lock must be held before this function is called
> - * to prevent the underlying pages from migrating.
> - */
> -static void unlock_zspage(struct zspage *zspage)
> -{
> -       struct page *page = get_first_page(zspage);
> -
> -       do {
> -               unlock_page(page);
> -       } while ((page = get_next_page(page)) != NULL);
> -}
> -#endif /* CONFIG_ZPOOL */
> -
>  static void migrate_lock_init(struct zspage *zspage)
>  {
>         rwlock_init(&zspage->lock);
> @@ -2126,9 +2030,6 @@ static void async_free_zspage(struct work_struct *work)
>                 VM_BUG_ON(fullness != ZS_INUSE_RATIO_0);
>                 class = pool->size_class[class_idx];
>                 spin_lock(&pool->lock);
> -#ifdef CONFIG_ZPOOL
> -               list_del(&zspage->lru);
> -#endif
>                 __free_zspage(pool, class, zspage);
>                 spin_unlock(&pool->lock);
>         }
> @@ -2474,10 +2375,6 @@ struct zs_pool *zs_create_pool(const char *name)
>          */
>         zs_register_shrinker(pool);
>
> -#ifdef CONFIG_ZPOOL
> -       INIT_LIST_HEAD(&pool->lru);
> -#endif
> -
>         return pool;
>
>  err:
> @@ -2520,190 +2417,6 @@ void zs_destroy_pool(struct zs_pool *pool)
>  }
>  EXPORT_SYMBOL_GPL(zs_destroy_pool);
>
> -#ifdef CONFIG_ZPOOL
> -static void restore_freelist(struct zs_pool *pool, struct size_class *class,
> -               struct zspage *zspage)
> -{
> -       unsigned int obj_idx = 0;
> -       unsigned long handle, off = 0; /* off is within-page offset */
> -       struct page *page = get_first_page(zspage);
> -       struct link_free *prev_free = NULL;
> -       void *prev_page_vaddr = NULL;
> -
> -       /* in case no free object found */
> -       set_freeobj(zspage, (unsigned int)(-1UL));
> -
> -       while (page) {
> -               void *vaddr = kmap_atomic(page);
> -               struct page *next_page;
> -
> -               while (off < PAGE_SIZE) {
> -                       void *obj_addr = vaddr + off;
> -
> -                       /* skip allocated object */
> -                       if (obj_allocated(page, obj_addr, &handle)) {
> -                               obj_idx++;
> -                               off += class->size;
> -                               continue;
> -                       }
> -
> -                       /* free deferred handle from reclaim attempt */
> -                       if (obj_stores_deferred_handle(page, obj_addr, &handle))
> -                               cache_free_handle(pool, handle);
> -
> -                       if (prev_free)
> -                               prev_free->next = obj_idx << OBJ_TAG_BITS;
> -                       else /* first free object found */
> -                               set_freeobj(zspage, obj_idx);
> -
> -                       prev_free = (struct link_free *)vaddr + off / sizeof(*prev_free);
> -                       /* if last free object in a previous page, need to unmap */
> -                       if (prev_page_vaddr) {
> -                               kunmap_atomic(prev_page_vaddr);
> -                               prev_page_vaddr = NULL;
> -                       }
> -
> -                       obj_idx++;
> -                       off += class->size;
> -               }
> -
> -               /*
> -                * Handle the last (full or partial) object on this page.
> -                */
> -               next_page = get_next_page(page);
> -               if (next_page) {
> -                       if (!prev_free || prev_page_vaddr) {
> -                               /*
> -                                * There is no free object in this page, so we can safely
> -                                * unmap it.
> -                                */
> -                               kunmap_atomic(vaddr);
> -                       } else {
> -                               /* update prev_page_vaddr since prev_free is on this page */
> -                               prev_page_vaddr = vaddr;
> -                       }
> -               } else { /* this is the last page */
> -                       if (prev_free) {
> -                               /*
> -                                * Reset OBJ_TAG_BITS bit to last link to tell
> -                                * whether it's allocated object or not.
> -                                */
> -                               prev_free->next = -1UL << OBJ_TAG_BITS;
> -                       }
> -
> -                       /* unmap previous page (if not done yet) */
> -                       if (prev_page_vaddr) {
> -                               kunmap_atomic(prev_page_vaddr);
> -                               prev_page_vaddr = NULL;
> -                       }
> -
> -                       kunmap_atomic(vaddr);
> -               }
> -
> -               page = next_page;
> -               off %= PAGE_SIZE;
> -       }
> -}
> -
> -static int zs_reclaim_page(struct zs_pool *pool, unsigned int retries)
> -{
> -       int i, obj_idx, ret = 0;
> -       unsigned long handle;
> -       struct zspage *zspage;
> -       struct page *page;
> -       int fullness;
> -
> -       /* Lock LRU and fullness list */
> -       spin_lock(&pool->lock);
> -       if (list_empty(&pool->lru)) {
> -               spin_unlock(&pool->lock);
> -               return -EINVAL;
> -       }
> -
> -       for (i = 0; i < retries; i++) {
> -               struct size_class *class;
> -
> -               zspage = list_last_entry(&pool->lru, struct zspage, lru);
> -               list_del(&zspage->lru);
> -
> -               /* zs_free may free objects, but not the zspage and handles */
> -               zspage->under_reclaim = true;
> -
> -               class = zspage_class(pool, zspage);
> -               fullness = get_fullness_group(class, zspage);
> -
> -               /* Lock out object allocations and object compaction */
> -               remove_zspage(class, zspage, fullness);
> -
> -               spin_unlock(&pool->lock);
> -               cond_resched();
> -
> -               /* Lock backing pages into place */
> -               lock_zspage(zspage);
> -
> -               obj_idx = 0;
> -               page = get_first_page(zspage);
> -               while (1) {
> -                       handle = find_alloced_obj(class, page, &obj_idx);
> -                       if (!handle) {
> -                               page = get_next_page(page);
> -                               if (!page)
> -                                       break;
> -                               obj_idx = 0;
> -                               continue;
> -                       }
> -
> -                       /*
> -                        * This will write the object and call zs_free.
> -                        *
> -                        * zs_free will free the object, but the
> -                        * under_reclaim flag prevents it from freeing
> -                        * the zspage altogether. This is necessary so
> -                        * that we can continue working with the
> -                        * zspage potentially after the last object
> -                        * has been freed.
> -                        */
> -                       ret = pool->zpool_ops->evict(pool->zpool, handle);
> -                       if (ret)
> -                               goto next;
> -
> -                       obj_idx++;
> -               }
> -
> -next:
> -               /* For freeing the zspage, or putting it back in the pool and LRU list. */
> -               spin_lock(&pool->lock);
> -               zspage->under_reclaim = false;
> -
> -               if (!get_zspage_inuse(zspage)) {
> -                       /*
> -                        * Fullness went stale as zs_free() won't touch it
> -                        * while the page is removed from the pool. Fix it
> -                        * up for the check in __free_zspage().
> -                        */
> -                       zspage->fullness = ZS_INUSE_RATIO_0;
> -
> -                       __free_zspage(pool, class, zspage);
> -                       spin_unlock(&pool->lock);
> -                       return 0;
> -               }
> -
> -               /*
> -                * Eviction fails on one of the handles, so we need to restore zspage.
> -                * We need to rebuild its freelist (and free stored deferred handles),
> -                * put it back to the correct size class, and add it to the LRU list.
> -                */
> -               restore_freelist(pool, class, zspage);
> -               putback_zspage(class, zspage);
> -               list_add(&zspage->lru, &pool->lru);
> -               unlock_zspage(zspage);
> -       }
> -
> -       spin_unlock(&pool->lock);
> -       return -EAGAIN;
> -}
> -#endif /* CONFIG_ZPOOL */
> -
>  static int __init zs_init(void)
>  {
>         int ret;
> --
> 2.34.1
>

Re: [RFC PATCH 1/7] mm: zswap: add pool shrinking mechanism

[Yosry Ahmed]

Hi Domenico,

On Mon, Jun 5, 2023 at 1:54 AM Domenico Cerasuolo
<cerasuolodomenico@gmail.com> wrote:
>
> Each zpool driver (zbud, z3fold and zsmalloc) implements its own shrink
> function, which is called from zpool_shrink. However, with this commit,
> a unified shrink function is added to zswap. The ultimate goal is to
> eliminate the need for zpool_shrink once all zpool implementations have
> dropped their shrink code.
>
> To ensure the functionality of each commit, this change focuses solely
> on adding the mechanism itself. No modifications are made to
> the backends, meaning that functionally, there are no immediate changes.
> The zswap mechanism will only come into effect once the backends have
> removed their shrink code. The subsequent commits will address the
> modifications needed in the backends.
>
> Signed-off-by: Domenico Cerasuolo <cerasuolodomenico@gmail.com>
> ---
>  mm/zswap.c | 83 ++++++++++++++++++++++++++++++++++++++++++++++++++----
>  1 file changed, 78 insertions(+), 5 deletions(-)
>
> diff --git a/mm/zswap.c b/mm/zswap.c
> index bcb82e09eb64..80d7780bf066 100644
> --- a/mm/zswap.c
> +++ b/mm/zswap.c
> @@ -159,6 +159,8 @@ struct zswap_pool {
>         struct work_struct shrink_work;
>         struct hlist_node node;
>         char tfm_name[CRYPTO_MAX_ALG_NAME];
> +       struct list_head lru;
> +       spinlock_t lock;

If this lock is only protecting the lru then I believe it's better to
call in lru_lock to make it explicit.

>  };
>
>  /*
> @@ -176,10 +178,12 @@ struct zswap_pool {
>   *            be held while changing the refcount.  Since the lock must
>   *            be held, there is no reason to also make refcount atomic.
>   * length - the length in bytes of the compressed page data.  Needed during
> - *          decompression. For a same value filled page length is 0.
> + *          decompression. For a same value filled page length is 0, and both
> + *          pool and lru are invalid and must be ignored.
>   * pool - the zswap_pool the entry's data is in
>   * handle - zpool allocation handle that stores the compressed page data
>   * value - value of the same-value filled pages which have same content
> + * lru - handle to the pool's lru used to evict pages.
>   */
>  struct zswap_entry {
>         struct rb_node rbnode;
> @@ -192,6 +196,7 @@ struct zswap_entry {
>                 unsigned long value;
>         };
>         struct obj_cgroup *objcg;
> +       struct list_head lru;
>  };
>
>  struct zswap_header {
> @@ -364,6 +369,9 @@ static void zswap_free_entry(struct zswap_entry *entry)
>         if (!entry->length)
>                 atomic_dec(&zswap_same_filled_pages);
>         else {
> +               spin_lock(&entry->pool->lock);
> +               list_del_init(&entry->lru);
> +               spin_unlock(&entry->pool->lock);

I think we should document the lock ordering somewhere (tree lock ->
lru lock), otherwise we may run into an ABBA deadlock down the road.

>                 zpool_free(entry->pool->zpool, entry->handle);
>                 zswap_pool_put(entry->pool);
>         }
> @@ -584,14 +592,65 @@ static struct zswap_pool *zswap_pool_find_get(char *type, char *compressor)
>         return NULL;
>  }
>
> +static int zswap_shrink(struct zswap_pool *pool)
> +{
> +       struct zswap_entry *lru_entry, *tree_entry = NULL;
> +       struct zswap_header *zhdr;
> +       struct zswap_tree *tree;
> +       swp_entry_t swpentry;
> +       int ret;
> +
> +       /* get a reclaimable entry from LRU */
> +       spin_lock(&pool->lock);
> +       if (list_empty(&pool->lru)) {
> +               spin_unlock(&pool->lock);
> +               return -EINVAL;
> +       }
> +       lru_entry = list_last_entry(&pool->lru, struct zswap_entry, lru);
> +       list_del_init(&lru_entry->lru);
> +       zhdr = zpool_map_handle(pool->zpool, lru_entry->handle, ZPOOL_MM_RO);
> +       tree = zswap_trees[swp_type(zhdr->swpentry)];
> +       zpool_unmap_handle(pool->zpool, lru_entry->handle);
> +       swpentry = zhdr->swpentry;
> +       spin_unlock(&pool->lock);
> +
> +       /* hold a reference from tree so it won't be freed during writeback */
> +       spin_lock(&tree->lock);
> +       tree_entry = zswap_entry_find_get(&tree->rbroot, swp_offset(swpentry));
> +       if (tree_entry != lru_entry) {
> +               if (tree_entry)
> +                       zswap_entry_put(tree, tree_entry);
> +               spin_unlock(&tree->lock);
> +               return -EAGAIN;
> +       }
> +       spin_unlock(&tree->lock);
> +
> +       ret = zswap_writeback_entry(pool->zpool, lru_entry->handle);
> +
> +       spin_lock(&tree->lock);
> +       if (ret) {
> +               spin_lock(&pool->lock);
> +               list_move(&lru_entry->lru, &pool->lru);
> +               spin_unlock(&pool->lock);
> +       }
> +       zswap_entry_put(tree, tree_entry);
> +       spin_unlock(&tree->lock);
> +
> +       return ret ? -EAGAIN : 0;
> +}
> +
>  static void shrink_worker(struct work_struct *w)
>  {
>         struct zswap_pool *pool = container_of(w, typeof(*pool),
>                                                 shrink_work);
>         int ret, failures = 0;
>
> +       /* zpool_evictable will be removed once all 3 backends have migrated*/
>         do {
> -               ret = zpool_shrink(pool->zpool, 1, NULL);
> +               if (zpool_evictable(pool->zpool))
> +                       ret = zpool_shrink(pool->zpool, 1, NULL);
> +               else
> +                       ret = zswap_shrink(pool);
>                 if (ret) {
>                         zswap_reject_reclaim_fail++;
>                         if (ret != -EAGAIN)
> @@ -655,6 +714,8 @@ static struct zswap_pool *zswap_pool_create(char *type, char *compressor)
>          */
>         kref_init(&pool->kref);
>         INIT_LIST_HEAD(&pool->list);
> +       INIT_LIST_HEAD(&pool->lru);
> +       spin_lock_init(&pool->lock);
>         INIT_WORK(&pool->shrink_work, shrink_worker);
>
>         zswap_pool_debug("created", pool);
> @@ -1270,7 +1331,7 @@ static int zswap_frontswap_store(unsigned type, pgoff_t offset,
>         }
>
>         /* store */
> -       hlen = zpool_evictable(entry->pool->zpool) ? sizeof(zhdr) : 0;
> +       hlen = sizeof(zhdr);
>         gfp = __GFP_NORETRY | __GFP_NOWARN | __GFP_KSWAPD_RECLAIM;
>         if (zpool_malloc_support_movable(entry->pool->zpool))
>                 gfp |= __GFP_HIGHMEM | __GFP_MOVABLE;
> @@ -1313,6 +1374,13 @@ static int zswap_frontswap_store(unsigned type, pgoff_t offset,
>                         zswap_entry_put(tree, dupentry);
>                 }
>         } while (ret == -EEXIST);
> +       INIT_LIST_HEAD(&entry->lru);
> +       /* zpool_evictable will be removed once all 3 backends have migrated*/
> +       if (entry->length && !zpool_evictable(entry->pool->zpool)) {
> +               spin_lock(&entry->pool->lock);
> +               list_add(&entry->lru, &entry->pool->lru);
> +               spin_unlock(&entry->pool->lock);
> +       }
>         spin_unlock(&tree->lock);
>
>         /* update stats */
> @@ -1384,8 +1452,7 @@ static int zswap_frontswap_load(unsigned type, pgoff_t offset,
>         /* decompress */
>         dlen = PAGE_SIZE;
>         src = zpool_map_handle(entry->pool->zpool, entry->handle, ZPOOL_MM_RO);
> -       if (zpool_evictable(entry->pool->zpool))
> -               src += sizeof(struct zswap_header);
> +       src += sizeof(struct zswap_header);
>
>         if (!zpool_can_sleep_mapped(entry->pool->zpool)) {
>                 memcpy(tmp, src, entry->length);
> @@ -1415,6 +1482,12 @@ static int zswap_frontswap_load(unsigned type, pgoff_t offset,
>  freeentry:
>         spin_lock(&tree->lock);
>         zswap_entry_put(tree, entry);
> +       /* zpool_evictable will be removed once all 3 backends have migrated*/
> +       if (entry->length && !zpool_evictable(entry->pool->zpool)) {
> +               spin_lock(&entry->pool->lock);
> +               list_move(&entry->lru, &entry->pool->lru);
> +               spin_unlock(&entry->pool->lock);
> +       }
>         spin_unlock(&tree->lock);
>
>         return ret;
> --
> 2.34.1
>

Re: [RFC PATCH 1/7] mm: zswap: add pool shrinking mechanism

[Domenico Cerasuolo]

On Mon, Jun 5, 2023 at 5:29 PM Johannes Weiner <hannes@cmpxchg.org> wrote:
>
> Hi Domenico,
>
> On Mon, Jun 05, 2023 at 10:54:13AM +0200, Domenico Cerasuolo wrote:
> > @@ -364,6 +369,9 @@ static void zswap_free_entry(struct zswap_entry *entry)
> >       if (!entry->length)
> >               atomic_dec(&zswap_same_filled_pages);
> >       else {
> > +             spin_lock(&entry->pool->lock);
> > +             list_del_init(&entry->lru);
> > +             spin_unlock(&entry->pool->lock);
>
> This should be list_del(), as the entry is freed right after this and
> the list isn't reused anymore.
>
> The slab memory is recycled, but the allocation site (commented on
> below) doesn't need the list initialized.
>
> However, I think it also needs to check !zpool_evictable(). If
> alloc/store doesn't do the list_add(), this would be a double delete.
>

Thanks Johannes, I'm addressing all of your comments in the series in a V2.

> > @@ -584,14 +592,65 @@ static struct zswap_pool *zswap_pool_find_get(char *type, char *compressor)
> >       return NULL;
> >  }
> >
> > +static int zswap_shrink(struct zswap_pool *pool)
> > +{
> > +     struct zswap_entry *lru_entry, *tree_entry = NULL;
> > +     struct zswap_header *zhdr;
> > +     struct zswap_tree *tree;
> > +     swp_entry_t swpentry;
> > +     int ret;
> > +
> > +     /* get a reclaimable entry from LRU */
> > +     spin_lock(&pool->lock);
> > +     if (list_empty(&pool->lru)) {
> > +             spin_unlock(&pool->lock);
> > +             return -EINVAL;
> > +     }
> > +     lru_entry = list_last_entry(&pool->lru, struct zswap_entry, lru);
> > +     list_del_init(&lru_entry->lru);
> > +     zhdr = zpool_map_handle(pool->zpool, lru_entry->handle, ZPOOL_MM_RO);
> > +     tree = zswap_trees[swp_type(zhdr->swpentry)];
> > +     zpool_unmap_handle(pool->zpool, lru_entry->handle);
> > +     swpentry = zhdr->swpentry;
> > +     spin_unlock(&pool->lock);
>
> Once the pool lock is dropped, the lru_entry might get freed. But the
> swpentry is copied to the stack, and lru_entry isn't deref'd again
> until the entry is verified to still be alive in the tree.
>
> This could use a comment.
>
> > +     /* hold a reference from tree so it won't be freed during writeback */
> > +     spin_lock(&tree->lock);
> > +     tree_entry = zswap_entry_find_get(&tree->rbroot, swp_offset(swpentry));
> > +     if (tree_entry != lru_entry) {
> > +             if (tree_entry)
> > +                     zswap_entry_put(tree, tree_entry);
> > +             spin_unlock(&tree->lock);
> > +             return -EAGAIN;
> > +     }
> > +     spin_unlock(&tree->lock);
>
> It's pretty outrageous how much simpler this is compared to the
> <backend>_reclaim_page() functions! The backends have to jump through
> a lot of hoops to serialize against freeing, whereas zswap can simply
> hold a reference. This is clearly a much better design.
>
> > +     ret = zswap_writeback_entry(pool->zpool, lru_entry->handle);
> > +
> > +     spin_lock(&tree->lock);
> > +     if (ret) {
> > +             spin_lock(&pool->lock);
> > +             list_move(&lru_entry->lru, &pool->lru);
> > +             spin_unlock(&pool->lock);
> > +     }
> > +     zswap_entry_put(tree, tree_entry);
> > +     spin_unlock(&tree->lock);
> > +
> > +     return ret ? -EAGAIN : 0;
> > +}
> > +
> >  static void shrink_worker(struct work_struct *w)
> >  {
> >       struct zswap_pool *pool = container_of(w, typeof(*pool),
> >                                               shrink_work);
> >       int ret, failures = 0;
> >
> > +     /* zpool_evictable will be removed once all 3 backends have migrated*/
>
> Missing space between text and */
>
> >       do {
> > -             ret = zpool_shrink(pool->zpool, 1, NULL);
> > +             if (zpool_evictable(pool->zpool))
> > +                     ret = zpool_shrink(pool->zpool, 1, NULL);
> > +             else
> > +                     ret = zswap_shrink(pool);
> >               if (ret) {
> >                       zswap_reject_reclaim_fail++;
> >                       if (ret != -EAGAIN)
> > @@ -655,6 +714,8 @@ static struct zswap_pool *zswap_pool_create(char *type, char *compressor)
> >        */
> >       kref_init(&pool->kref);
> >       INIT_LIST_HEAD(&pool->list);
> > +     INIT_LIST_HEAD(&pool->lru);
> > +     spin_lock_init(&pool->lock);
> >       INIT_WORK(&pool->shrink_work, shrink_worker);
> >
> >       zswap_pool_debug("created", pool);
> > @@ -1270,7 +1331,7 @@ static int zswap_frontswap_store(unsigned type, pgoff_t offset,
> >       }
> >
> >       /* store */
> > -     hlen = zpool_evictable(entry->pool->zpool) ? sizeof(zhdr) : 0;
> > +     hlen = sizeof(zhdr);
> >       gfp = __GFP_NORETRY | __GFP_NOWARN | __GFP_KSWAPD_RECLAIM;
> >       if (zpool_malloc_support_movable(entry->pool->zpool))
> >               gfp |= __GFP_HIGHMEM | __GFP_MOVABLE;
> > @@ -1313,6 +1374,13 @@ static int zswap_frontswap_store(unsigned type, pgoff_t offset,
> >                       zswap_entry_put(tree, dupentry);
> >               }
> >       } while (ret == -EEXIST);
> > +     INIT_LIST_HEAD(&entry->lru);
>
> The list_add() below initializes the entry, so this shouldn't be
> needed.
>
> > +     /* zpool_evictable will be removed once all 3 backends have migrated*/
> > +     if (entry->length && !zpool_evictable(entry->pool->zpool)) {
> > +             spin_lock(&entry->pool->lock);
> > +             list_add(&entry->lru, &entry->pool->lru);
> > +             spin_unlock(&entry->pool->lock);
> > +     }
> >       spin_unlock(&tree->lock);
> >
> >       /* update stats */

Re: [RFC PATCH 1/7] mm: zswap: add pool shrinking mechanism

[Domenico Cerasuolo]

On Tue, Jun 6, 2023 at 4:19 AM Yosry Ahmed <yosryahmed@google.com> wrote:
>
> Hi Domenico,
>
> On Mon, Jun 5, 2023 at 1:54 AM Domenico Cerasuolo
> <cerasuolodomenico@gmail.com> wrote:
> >
> > Each zpool driver (zbud, z3fold and zsmalloc) implements its own shrink
> > function, which is called from zpool_shrink. However, with this commit,
> > a unified shrink function is added to zswap. The ultimate goal is to
> > eliminate the need for zpool_shrink once all zpool implementations have
> > dropped their shrink code.
> >
> > To ensure the functionality of each commit, this change focuses solely
> > on adding the mechanism itself. No modifications are made to
> > the backends, meaning that functionally, there are no immediate changes.
> > The zswap mechanism will only come into effect once the backends have
> > removed their shrink code. The subsequent commits will address the
> > modifications needed in the backends.
> >
> > Signed-off-by: Domenico Cerasuolo <cerasuolodomenico@gmail.com>
> > ---
> >  mm/zswap.c | 83 ++++++++++++++++++++++++++++++++++++++++++++++++++----
> >  1 file changed, 78 insertions(+), 5 deletions(-)
> >
> > diff --git a/mm/zswap.c b/mm/zswap.c
> > index bcb82e09eb64..80d7780bf066 100644
> > --- a/mm/zswap.c
> > +++ b/mm/zswap.c
> > @@ -159,6 +159,8 @@ struct zswap_pool {
> >         struct work_struct shrink_work;
> >         struct hlist_node node;
> >         char tfm_name[CRYPTO_MAX_ALG_NAME];
> > +       struct list_head lru;
> > +       spinlock_t lock;
>
> If this lock is only protecting the lru then I believe it's better to
> call in lru_lock to make it explicit.

Hi Yosry,

thanks for the input, it makes sense to call it lru_lock, will update.

>
> >  };
> >
> >  /*
> > @@ -176,10 +178,12 @@ struct zswap_pool {
> >   *            be held while changing the refcount.  Since the lock must
> >   *            be held, there is no reason to also make refcount atomic.
> >   * length - the length in bytes of the compressed page data.  Needed during
> > - *          decompression. For a same value filled page length is 0.
> > + *          decompression. For a same value filled page length is 0, and both
> > + *          pool and lru are invalid and must be ignored.
> >   * pool - the zswap_pool the entry's data is in
> >   * handle - zpool allocation handle that stores the compressed page data
> >   * value - value of the same-value filled pages which have same content
> > + * lru - handle to the pool's lru used to evict pages.
> >   */
> >  struct zswap_entry {
> >         struct rb_node rbnode;
> > @@ -192,6 +196,7 @@ struct zswap_entry {
> >                 unsigned long value;
> >         };
> >         struct obj_cgroup *objcg;
> > +       struct list_head lru;
> >  };
> >
> >  struct zswap_header {
> > @@ -364,6 +369,9 @@ static void zswap_free_entry(struct zswap_entry *entry)
> >         if (!entry->length)
> >                 atomic_dec(&zswap_same_filled_pages);
> >         else {
> > +               spin_lock(&entry->pool->lock);
> > +               list_del_init(&entry->lru);
> > +               spin_unlock(&entry->pool->lock);
>
> I think we should document the lock ordering somewhere (tree lock ->
> lru lock), otherwise we may run into an ABBA deadlock down the road.

Will update in the next iteration.

>
> >                 zpool_free(entry->pool->zpool, entry->handle);
> >                 zswap_pool_put(entry->pool);
> >         }
> > @@ -584,14 +592,65 @@ static struct zswap_pool *zswap_pool_find_get(char *type, char *compressor)
> >         return NULL;
> >  }
> >
> > +static int zswap_shrink(struct zswap_pool *pool)
> > +{
> > +       struct zswap_entry *lru_entry, *tree_entry = NULL;
> > +       struct zswap_header *zhdr;
> > +       struct zswap_tree *tree;
> > +       swp_entry_t swpentry;
> > +       int ret;
> > +
> > +       /* get a reclaimable entry from LRU */
> > +       spin_lock(&pool->lock);
> > +       if (list_empty(&pool->lru)) {
> > +               spin_unlock(&pool->lock);
> > +               return -EINVAL;
> > +       }
> > +       lru_entry = list_last_entry(&pool->lru, struct zswap_entry, lru);
> > +       list_del_init(&lru_entry->lru);
> > +       zhdr = zpool_map_handle(pool->zpool, lru_entry->handle, ZPOOL_MM_RO);
> > +       tree = zswap_trees[swp_type(zhdr->swpentry)];
> > +       zpool_unmap_handle(pool->zpool, lru_entry->handle);
> > +       swpentry = zhdr->swpentry;
> > +       spin_unlock(&pool->lock);
> > +
> > +       /* hold a reference from tree so it won't be freed during writeback */
> > +       spin_lock(&tree->lock);
> > +       tree_entry = zswap_entry_find_get(&tree->rbroot, swp_offset(swpentry));
> > +       if (tree_entry != lru_entry) {
> > +               if (tree_entry)
> > +                       zswap_entry_put(tree, tree_entry);
> > +               spin_unlock(&tree->lock);
> > +               return -EAGAIN;
> > +       }
> > +       spin_unlock(&tree->lock);
> > +
> > +       ret = zswap_writeback_entry(pool->zpool, lru_entry->handle);
> > +
> > +       spin_lock(&tree->lock);
> > +       if (ret) {
> > +               spin_lock(&pool->lock);
> > +               list_move(&lru_entry->lru, &pool->lru);
> > +               spin_unlock(&pool->lock);
> > +       }
> > +       zswap_entry_put(tree, tree_entry);
> > +       spin_unlock(&tree->lock);
> > +
> > +       return ret ? -EAGAIN : 0;
> > +}
> > +
> >  static void shrink_worker(struct work_struct *w)
> >  {
> >         struct zswap_pool *pool = container_of(w, typeof(*pool),
> >                                                 shrink_work);
> >         int ret, failures = 0;
> >
> > +       /* zpool_evictable will be removed once all 3 backends have migrated*/
> >         do {
> > -               ret = zpool_shrink(pool->zpool, 1, NULL);
> > +               if (zpool_evictable(pool->zpool))
> > +                       ret = zpool_shrink(pool->zpool, 1, NULL);
> > +               else
> > +                       ret = zswap_shrink(pool);
> >                 if (ret) {
> >                         zswap_reject_reclaim_fail++;
> >                         if (ret != -EAGAIN)
> > @@ -655,6 +714,8 @@ static struct zswap_pool *zswap_pool_create(char *type, char *compressor)
> >          */
> >         kref_init(&pool->kref);
> >         INIT_LIST_HEAD(&pool->list);
> > +       INIT_LIST_HEAD(&pool->lru);
> > +       spin_lock_init(&pool->lock);
> >         INIT_WORK(&pool->shrink_work, shrink_worker);
> >
> >         zswap_pool_debug("created", pool);
> > @@ -1270,7 +1331,7 @@ static int zswap_frontswap_store(unsigned type, pgoff_t offset,
> >         }
> >
> >         /* store */
> > -       hlen = zpool_evictable(entry->pool->zpool) ? sizeof(zhdr) : 0;
> > +       hlen = sizeof(zhdr);
> >         gfp = __GFP_NORETRY | __GFP_NOWARN | __GFP_KSWAPD_RECLAIM;
> >         if (zpool_malloc_support_movable(entry->pool->zpool))
> >                 gfp |= __GFP_HIGHMEM | __GFP_MOVABLE;
> > @@ -1313,6 +1374,13 @@ static int zswap_frontswap_store(unsigned type, pgoff_t offset,
> >                         zswap_entry_put(tree, dupentry);
> >                 }
> >         } while (ret == -EEXIST);
> > +       INIT_LIST_HEAD(&entry->lru);
> > +       /* zpool_evictable will be removed once all 3 backends have migrated*/
> > +       if (entry->length && !zpool_evictable(entry->pool->zpool)) {
> > +               spin_lock(&entry->pool->lock);
> > +               list_add(&entry->lru, &entry->pool->lru);
> > +               spin_unlock(&entry->pool->lock);
> > +       }
> >         spin_unlock(&tree->lock);
> >
> >         /* update stats */
> > @@ -1384,8 +1452,7 @@ static int zswap_frontswap_load(unsigned type, pgoff_t offset,
> >         /* decompress */
> >         dlen = PAGE_SIZE;
> >         src = zpool_map_handle(entry->pool->zpool, entry->handle, ZPOOL_MM_RO);
> > -       if (zpool_evictable(entry->pool->zpool))
> > -               src += sizeof(struct zswap_header);
> > +       src += sizeof(struct zswap_header);
> >
> >         if (!zpool_can_sleep_mapped(entry->pool->zpool)) {
> >                 memcpy(tmp, src, entry->length);
> > @@ -1415,6 +1482,12 @@ static int zswap_frontswap_load(unsigned type, pgoff_t offset,
> >  freeentry:
> >         spin_lock(&tree->lock);
> >         zswap_entry_put(tree, entry);
> > +       /* zpool_evictable will be removed once all 3 backends have migrated*/
> > +       if (entry->length && !zpool_evictable(entry->pool->zpool)) {
> > +               spin_lock(&entry->pool->lock);
> > +               list_move(&entry->lru, &entry->pool->lru);
> > +               spin_unlock(&entry->pool->lock);
> > +       }
> >         spin_unlock(&tree->lock);
> >
> >         return ret;
> > --
> > 2.34.1
> >

Re: [RFC PATCH 4/7] mm: zswap: remove page reclaim logic from zsmalloc

[Domenico Cerasuolo]

On Tue, Jun 6, 2023 at 1:37 AM Nhat Pham <nphamcs@gmail.com> wrote:
>
> On Mon, Jun 5, 2023 at 1:54 AM Domenico Cerasuolo
> <cerasuolodomenico@gmail.com> wrote:
> >
> > With the recent enhancement to zswap enabling direct page writeback, the
> > need for the shrink code in zsmalloc has become obsolete. As a result,
> > this commit removes the page reclaim logic from zsmalloc entirely.
> >
> > Signed-off-by: Domenico Cerasuolo <cerasuolodomenico@gmail.com>
> > ---
> >  mm/zsmalloc.c | 291 +-------------------------------------------------
> >  1 file changed, 2 insertions(+), 289 deletions(-)
> >
> > diff --git a/mm/zsmalloc.c b/mm/zsmalloc.c
> > index 02f7f414aade..c87a60514f21 100644
> > --- a/mm/zsmalloc.c
> > +++ b/mm/zsmalloc.c
> > @@ -250,13 +250,6 @@ struct zs_pool {
> >         /* Compact classes */
> >         struct shrinker shrinker;
> >
> > -#ifdef CONFIG_ZPOOL
> > -       /* List tracking the zspages in LRU order by most recently added object */
> > -       struct list_head lru;
> > -       struct zpool *zpool;
> > -       const struct zpool_ops *zpool_ops;
> > -#endif
> > -
> >  #ifdef CONFIG_ZSMALLOC_STAT
> >         struct dentry *stat_dentry;
> >  #endif
> > @@ -279,13 +272,6 @@ struct zspage {
> >         unsigned int freeobj;
> >         struct page *first_page;
> >         struct list_head list; /* fullness list */
> > -
> > -#ifdef CONFIG_ZPOOL
> > -       /* links the zspage to the lru list in the pool */
> > -       struct list_head lru;
> > -       bool under_reclaim;
> > -#endif
> > -
> >         struct zs_pool *pool;
> >         rwlock_t lock;
> >  };
> > @@ -393,14 +379,7 @@ static void *zs_zpool_create(const char *name, gfp_t gfp,
> >          * different contexts and its caller must provide a valid
> >          * gfp mask.
> >          */
> > -       struct zs_pool *pool = zs_create_pool(name);
> > -
> > -       if (pool) {
> > -               pool->zpool = zpool;
> > -               pool->zpool_ops = zpool_ops;
> > -       }
> > -
> > -       return pool;
> > +       return zs_create_pool(name);
> >  }
> >
> >  static void zs_zpool_destroy(void *pool)
> > @@ -422,27 +401,6 @@ static void zs_zpool_free(void *pool, unsigned long handle)
> >         zs_free(pool, handle);
> >  }
> >
> > -static int zs_reclaim_page(struct zs_pool *pool, unsigned int retries);
> > -
> > -static int zs_zpool_shrink(void *pool, unsigned int pages,
> > -                       unsigned int *reclaimed)
> > -{
> > -       unsigned int total = 0;
> > -       int ret = -EINVAL;
> > -
> > -       while (total < pages) {
> > -               ret = zs_reclaim_page(pool, 8);
> > -               if (ret < 0)
> > -                       break;
> > -               total++;
> > -       }
> > -
> > -       if (reclaimed)
> > -               *reclaimed = total;
> > -
> > -       return ret;
> > -}
> > -
> >  static void *zs_zpool_map(void *pool, unsigned long handle,
> >                         enum zpool_mapmode mm)
> >  {
> > @@ -481,7 +439,7 @@ static struct zpool_driver zs_zpool_driver = {
> >         .malloc_support_movable = true,
> >         .malloc =                 zs_zpool_malloc,
> >         .free =                   zs_zpool_free,
> > -       .shrink =                 zs_zpool_shrink,
> > +       .shrink =                 NULL,
> >         .map =                    zs_zpool_map,
> >         .unmap =                  zs_zpool_unmap,
> >         .total_size =             zs_zpool_total_size,
> > @@ -884,14 +842,6 @@ static inline bool obj_allocated(struct page *page, void *obj, unsigned long *ph
> >         return obj_tagged(page, obj, phandle, OBJ_ALLOCATED_TAG);
> >  }
> >
> > -#ifdef CONFIG_ZPOOL
> > -static bool obj_stores_deferred_handle(struct page *page, void *obj,
> > -               unsigned long *phandle)
> > -{
> > -       return obj_tagged(page, obj, phandle, OBJ_DEFERRED_HANDLE_TAG);
> > -}
> > -#endif
> > -
> >  static void reset_page(struct page *page)
> >  {
> >         __ClearPageMovable(page);
> > @@ -1006,9 +956,6 @@ static void free_zspage(struct zs_pool *pool, struct size_class *class,
> >         }
> >
> >         remove_zspage(class, zspage, ZS_INUSE_RATIO_0);
> > -#ifdef CONFIG_ZPOOL
> > -       list_del(&zspage->lru);
> > -#endif
> >         __free_zspage(pool, class, zspage);
> >  }
> >
> > @@ -1054,11 +1001,6 @@ static void init_zspage(struct size_class *class, struct zspage *zspage)
> >                 off %= PAGE_SIZE;
> >         }
> >
> > -#ifdef CONFIG_ZPOOL
> > -       INIT_LIST_HEAD(&zspage->lru);
> > -       zspage->under_reclaim = false;
> > -#endif
> > -
> >         set_freeobj(zspage, 0);
> >  }
> >
> > @@ -1525,13 +1467,6 @@ unsigned long zs_malloc(struct zs_pool *pool, size_t size, gfp_t gfp)
> >         /* We completely set up zspage so mark them as movable */
> >         SetZsPageMovable(pool, zspage);
> >  out:
> > -#ifdef CONFIG_ZPOOL
> > -       /* Add/move zspage to beginning of LRU */
> > -       if (!list_empty(&zspage->lru))
> > -               list_del(&zspage->lru);
> > -       list_add(&zspage->lru, &pool->lru);
> > -#endif
> > -
> >         spin_unlock(&pool->lock);
> >
> >         return handle;
> > @@ -1600,20 +1535,6 @@ void zs_free(struct zs_pool *pool, unsigned long handle)
> >         class = zspage_class(pool, zspage);
> >
> >         class_stat_dec(class, ZS_OBJS_INUSE, 1);
> > -
> > -#ifdef CONFIG_ZPOOL
> > -       if (zspage->under_reclaim) {
> > -               /*
> > -                * Reclaim needs the handles during writeback. It'll free
> > -                * them along with the zspage when it's done with them.
> > -                *
> > -                * Record current deferred handle in the object's header.
> > -                */
> > -               obj_free(class->size, obj, &handle);
> > -               spin_unlock(&pool->lock);
> > -               return;
> > -       }
> > -#endif
> >         obj_free(class->size, obj, NULL);
> >
> >         fullness = fix_fullness_group(class, zspage);
> > @@ -1890,23 +1811,6 @@ static void lock_zspage(struct zspage *zspage)
> >  }
> >  #endif /* defined(CONFIG_ZPOOL) || defined(CONFIG_COMPACTION) */
>
> If I recall correctly, the defined(CONFIG_ZPOOL) condition is
> only needed for the lock_zspage() call in zs_reclaim_page().
>
> You might be able to get away with just
> #ifdef CONFIG_COMPACTION if you're removing writeback.
>
> Do fact-check me of course - try to build with these CONFIGs turned
> off and see. I'm surprised kernel test robot has not complained about
> unused static function lock_zspage() in the case
> CONFIG_ZPOOL && !CONFIG_COMPACTION
>

Thanks! It is indeed the case, with CONFIG_ZPOOL && !CONFIG_COMPACTION
lock_zspage becomes unused -> will go with #ifdef CONFIG_COMPACTION alone.

> >
> > -#ifdef CONFIG_ZPOOL
> > -/*
> > - * Unlocks all the pages of the zspage.
> > - *
> > - * pool->lock must be held before this function is called
> > - * to prevent the underlying pages from migrating.
> > - */
> > -static void unlock_zspage(struct zspage *zspage)
> > -{
> > -       struct page *page = get_first_page(zspage);
> > -
> > -       do {
> > -               unlock_page(page);
> > -       } while ((page = get_next_page(page)) != NULL);
> > -}
> > -#endif /* CONFIG_ZPOOL */
> > -
> >  static void migrate_lock_init(struct zspage *zspage)
> >  {
> >         rwlock_init(&zspage->lock);
> > @@ -2126,9 +2030,6 @@ static void async_free_zspage(struct work_struct *work)
> >                 VM_BUG_ON(fullness != ZS_INUSE_RATIO_0);
> >                 class = pool->size_class[class_idx];
> >                 spin_lock(&pool->lock);
> > -#ifdef CONFIG_ZPOOL
> > -               list_del(&zspage->lru);
> > -#endif
> >                 __free_zspage(pool, class, zspage);
> >                 spin_unlock(&pool->lock);
> >         }
> > @@ -2474,10 +2375,6 @@ struct zs_pool *zs_create_pool(const char *name)
> >          */
> >         zs_register_shrinker(pool);
> >
> > -#ifdef CONFIG_ZPOOL
> > -       INIT_LIST_HEAD(&pool->lru);
> > -#endif
> > -
> >         return pool;
> >
> >  err:
> > @@ -2520,190 +2417,6 @@ void zs_destroy_pool(struct zs_pool *pool)
> >  }
> >  EXPORT_SYMBOL_GPL(zs_destroy_pool);
> >
> > -#ifdef CONFIG_ZPOOL
> > -static void restore_freelist(struct zs_pool *pool, struct size_class *class,
> > -               struct zspage *zspage)
> > -{
> > -       unsigned int obj_idx = 0;
> > -       unsigned long handle, off = 0; /* off is within-page offset */
> > -       struct page *page = get_first_page(zspage);
> > -       struct link_free *prev_free = NULL;
> > -       void *prev_page_vaddr = NULL;
> > -
> > -       /* in case no free object found */
> > -       set_freeobj(zspage, (unsigned int)(-1UL));
> > -
> > -       while (page) {
> > -               void *vaddr = kmap_atomic(page);
> > -               struct page *next_page;
> > -
> > -               while (off < PAGE_SIZE) {
> > -                       void *obj_addr = vaddr + off;
> > -
> > -                       /* skip allocated object */
> > -                       if (obj_allocated(page, obj_addr, &handle)) {
> > -                               obj_idx++;
> > -                               off += class->size;
> > -                               continue;
> > -                       }
> > -
> > -                       /* free deferred handle from reclaim attempt */
> > -                       if (obj_stores_deferred_handle(page, obj_addr, &handle))
> > -                               cache_free_handle(pool, handle);
> > -
> > -                       if (prev_free)
> > -                               prev_free->next = obj_idx << OBJ_TAG_BITS;
> > -                       else /* first free object found */
> > -                               set_freeobj(zspage, obj_idx);
> > -
> > -                       prev_free = (struct link_free *)vaddr + off / sizeof(*prev_free);
> > -                       /* if last free object in a previous page, need to unmap */
> > -                       if (prev_page_vaddr) {
> > -                               kunmap_atomic(prev_page_vaddr);
> > -                               prev_page_vaddr = NULL;
> > -                       }
> > -
> > -                       obj_idx++;
> > -                       off += class->size;
> > -               }
> > -
> > -               /*
> > -                * Handle the last (full or partial) object on this page.
> > -                */
> > -               next_page = get_next_page(page);
> > -               if (next_page) {
> > -                       if (!prev_free || prev_page_vaddr) {
> > -                               /*
> > -                                * There is no free object in this page, so we can safely
> > -                                * unmap it.
> > -                                */
> > -                               kunmap_atomic(vaddr);
> > -                       } else {
> > -                               /* update prev_page_vaddr since prev_free is on this page */
> > -                               prev_page_vaddr = vaddr;
> > -                       }
> > -               } else { /* this is the last page */
> > -                       if (prev_free) {
> > -                               /*
> > -                                * Reset OBJ_TAG_BITS bit to last link to tell
> > -                                * whether it's allocated object or not.
> > -                                */
> > -                               prev_free->next = -1UL << OBJ_TAG_BITS;
> > -                       }
> > -
> > -                       /* unmap previous page (if not done yet) */
> > -                       if (prev_page_vaddr) {
> > -                               kunmap_atomic(prev_page_vaddr);
> > -                               prev_page_vaddr = NULL;
> > -                       }
> > -
> > -                       kunmap_atomic(vaddr);
> > -               }
> > -
> > -               page = next_page;
> > -               off %= PAGE_SIZE;
> > -       }
> > -}
> > -
> > -static int zs_reclaim_page(struct zs_pool *pool, unsigned int retries)
> > -{
> > -       int i, obj_idx, ret = 0;
> > -       unsigned long handle;
> > -       struct zspage *zspage;
> > -       struct page *page;
> > -       int fullness;
> > -
> > -       /* Lock LRU and fullness list */
> > -       spin_lock(&pool->lock);
> > -       if (list_empty(&pool->lru)) {
> > -               spin_unlock(&pool->lock);
> > -               return -EINVAL;
> > -       }
> > -
> > -       for (i = 0; i < retries; i++) {
> > -               struct size_class *class;
> > -
> > -               zspage = list_last_entry(&pool->lru, struct zspage, lru);
> > -               list_del(&zspage->lru);
> > -
> > -               /* zs_free may free objects, but not the zspage and handles */
> > -               zspage->under_reclaim = true;
> > -
> > -               class = zspage_class(pool, zspage);
> > -               fullness = get_fullness_group(class, zspage);
> > -
> > -               /* Lock out object allocations and object compaction */
> > -               remove_zspage(class, zspage, fullness);
> > -
> > -               spin_unlock(&pool->lock);
> > -               cond_resched();
> > -
> > -               /* Lock backing pages into place */
> > -               lock_zspage(zspage);
> > -
> > -               obj_idx = 0;
> > -               page = get_first_page(zspage);
> > -               while (1) {
> > -                       handle = find_alloced_obj(class, page, &obj_idx);
> > -                       if (!handle) {
> > -                               page = get_next_page(page);
> > -                               if (!page)
> > -                                       break;
> > -                               obj_idx = 0;
> > -                               continue;
> > -                       }
> > -
> > -                       /*
> > -                        * This will write the object and call zs_free.
> > -                        *
> > -                        * zs_free will free the object, but the
> > -                        * under_reclaim flag prevents it from freeing
> > -                        * the zspage altogether. This is necessary so
> > -                        * that we can continue working with the
> > -                        * zspage potentially after the last object
> > -                        * has been freed.
> > -                        */
> > -                       ret = pool->zpool_ops->evict(pool->zpool, handle);
> > -                       if (ret)
> > -                               goto next;
> > -
> > -                       obj_idx++;
> > -               }
> > -
> > -next:
> > -               /* For freeing the zspage, or putting it back in the pool and LRU list. */
> > -               spin_lock(&pool->lock);
> > -               zspage->under_reclaim = false;
> > -
> > -               if (!get_zspage_inuse(zspage)) {
> > -                       /*
> > -                        * Fullness went stale as zs_free() won't touch it
> > -                        * while the page is removed from the pool. Fix it
> > -                        * up for the check in __free_zspage().
> > -                        */
> > -                       zspage->fullness = ZS_INUSE_RATIO_0;
> > -
> > -                       __free_zspage(pool, class, zspage);
> > -                       spin_unlock(&pool->lock);
> > -                       return 0;
> > -               }
> > -
> > -               /*
> > -                * Eviction fails on one of the handles, so we need to restore zspage.
> > -                * We need to rebuild its freelist (and free stored deferred handles),
> > -                * put it back to the correct size class, and add it to the LRU list.
> > -                */
> > -               restore_freelist(pool, class, zspage);
> > -               putback_zspage(class, zspage);
> > -               list_add(&zspage->lru, &pool->lru);
> > -               unlock_zspage(zspage);
> > -       }
> > -
> > -       spin_unlock(&pool->lock);
> > -       return -EAGAIN;
> > -}
> > -#endif /* CONFIG_ZPOOL */
> > -
> >  static int __init zs_init(void)
> >  {
> >         int ret;
> > --
> > 2.34.1
> >