mm: zswap: multiple zpools support

Support using multiple zpools of the same type in zswap, for concurrency
purposes.  A fixed number of 32 zpools is suggested by this commit, which
was determined empirically.  It can be later changed or made into a config
option if needed.

On a setup with zswap and zsmalloc, comparing a single zpool to 32 zpools
shows improvements in the zsmalloc lock contention, especially on the swap
out path.

The following shows the perf analysis of the swapout path when 10
workloads are simultaneously reclaiming and refaulting tmpfs pages.  There
are some improvements on the swap in path as well, but less significant.

1 zpool:

 |--28.99%--zswap_frontswap_store
       |
       <snip>
       |
       |--8.98%--zpool_map_handle
       |     |
       |      --8.98%--zs_zpool_map
       |           |
       |            --8.95%--zs_map_object
       |                 |
       |                  --8.38%--_raw_spin_lock
       |                       |
       |                        --7.39%--queued_spin_lock_slowpath
       |
       |--8.82%--zpool_malloc
       |     |
       |      --8.82%--zs_zpool_malloc
       |           |
       |            --8.80%--zs_malloc
       |                 |
       |                 |--7.21%--_raw_spin_lock
       |                 |     |
       |                 |      --6.81%--queued_spin_lock_slowpath
       <snip>

32 zpools:

 |--16.73%--zswap_frontswap_store
       |
       <snip>
       |
       |--1.81%--zpool_malloc
       |     |
       |      --1.81%--zs_zpool_malloc
       |           |
       |            --1.79%--zs_malloc
       |                 |
       |                  --0.73%--obj_malloc
       |
       |--1.06%--zswap_update_total_size
       |
       |--0.59%--zpool_map_handle
       |     |
       |      --0.59%--zs_zpool_map
       |           |
       |            --0.57%--zs_map_object
       |                 |
       |                  --0.51%--_raw_spin_lock
       <snip>

Link: https://lkml.kernel.org/r/20230620194644.3142384-1-yosryahmed@google.comSigned-off-by: Yosry Ahmed <yosryahmed@google.com>
Suggested-by: Yu Zhao <yuzhao@google.com>
Acked-by: Chris Li (Google) <chrisl@kernel.org>
Reviewed-by: Nhat Pham <nphamcs@gmail.com>
Tested-by: Nhat Pham <nphamcs@gmail.com>
Cc: Dan Streetman <ddstreet@ieee.org>
Cc: Domenico Cerasuolo <cerasuolodomenico@gmail.com>
Cc: Johannes Weiner <hannes@cmpxchg.org>
Cc: Konrad Rzeszutek Wilk <konrad.wilk@oracle.com>
Cc: Seth Jennings <sjenning@redhat.com>
Cc: Vitaly Wool <vitaly.wool@konsulko.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

mm/zswap.c

@@ -142,6 +142,9 @@ static bool zswap_exclusive_loads_enabled = IS_ENABLED(
 		CONFIG_ZSWAP_EXCLUSIVE_LOADS_DEFAULT_ON);
 module_param_named(exclusive_loads, zswap_exclusive_loads_enabled, bool, 0644);
 
+/* Number of zpools in zswap_pool (empirically determined for scalability) */
+#define ZSWAP_NR_ZPOOLS 32
+
 /*********************************
 * data structures
 **********************************/
@@ -161,7 +164,7 @@ struct crypto_acomp_ctx {
  * needs to be verified that it's still valid in the tree.
  */
 struct zswap_pool {
-	struct zpool *zpool;
+	struct zpool *zpools[ZSWAP_NR_ZPOOLS];
 	struct crypto_acomp_ctx __percpu *acomp_ctx;
 	struct kref kref;
 	struct list_head list;
@@ -248,7 +251,7 @@ static bool zswap_has_pool;
 
 #define zswap_pool_debug(msg, p)				\
 	pr_debug("%s pool %s/%s\n", msg, (p)->tfm_name,		\
-		 zpool_get_type((p)->zpool))
+		 zpool_get_type((p)->zpools[0]))
 
 static int zswap_writeback_entry(struct zswap_entry *entry,
 				 struct zswap_tree *tree);
@@ -272,11 +275,13 @@ static void zswap_update_total_size(void)
 {
 	struct zswap_pool *pool;
 	u64 total = 0;
+	int i;
 
 	rcu_read_lock();
 
 	list_for_each_entry_rcu(pool, &zswap_pools, list)
-		total += zpool_get_total_size(pool->zpool);
+		for (i = 0; i < ZSWAP_NR_ZPOOLS; i++)
+			total += zpool_get_total_size(pool->zpools[i]);
 
 	rcu_read_unlock();
 
@@ -365,6 +370,16 @@ static bool zswap_rb_erase(struct rb_root *root, struct zswap_entry *entry)
 	return false;
 }
 
+static struct zpool *zswap_find_zpool(struct zswap_entry *entry)
+{
+	int i = 0;
+
+	if (ZSWAP_NR_ZPOOLS > 1)
+		i = hash_ptr(entry, ilog2(ZSWAP_NR_ZPOOLS));
+
+	return entry->pool->zpools[i];
+}
+
 /*
  * Carries out the common pattern of freeing and entry's zpool allocation,
  * freeing the entry itself, and decrementing the number of stored pages.
@@ -381,7 +396,7 @@ static void zswap_free_entry(struct zswap_entry *entry)
 		spin_lock(&entry->pool->lru_lock);
 		list_del(&entry->lru);
 		spin_unlock(&entry->pool->lru_lock);
-		zpool_free(entry->pool->zpool, entry->handle);
+		zpool_free(zswap_find_zpool(entry), entry->handle);
 		zswap_pool_put(entry->pool);
 	}
 	zswap_entry_cache_free(entry);
@@ -590,7 +605,8 @@ static struct zswap_pool *zswap_pool_find_get(char *type, char *compressor)
 	list_for_each_entry_rcu(pool, &zswap_pools, list) {
 		if (strcmp(pool->tfm_name, compressor))
 			continue;
-		if (strcmp(zpool_get_type(pool->zpool), type))
+		/* all zpools share the same type */
+		if (strcmp(zpool_get_type(pool->zpools[0]), type))
 			continue;
 		/* if we can't get it, it's about to be destroyed */
 		if (!zswap_pool_get(pool))
@@ -695,6 +711,7 @@ static void shrink_worker(struct work_struct *w)
 
 static struct zswap_pool *zswap_pool_create(char *type, char *compressor)
 {
+	int i;
 	struct zswap_pool *pool;
 	char name[38]; /* 'zswap' + 32 char (max) num + \0 */
 	gfp_t gfp = __GFP_NORETRY | __GFP_NOWARN | __GFP_KSWAPD_RECLAIM;
@@ -715,15 +732,18 @@ static struct zswap_pool *zswap_pool_create(char *type, char *compressor)
 	if (!pool)
 		return NULL;
 
-	/* unique name for each pool specifically required by zsmalloc */
-	snprintf(name, 38, "zswap%x", atomic_inc_return(&zswap_pools_count));
+	for (i = 0; i < ZSWAP_NR_ZPOOLS; i++) {
+		/* 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);
-	if (!pool->zpool) {
-		pr_err("%s zpool not available\n", type);
-		goto error;
+		pool->zpools[i] = zpool_create_pool(type, name, gfp);
+		if (!pool->zpools[i]) {
+			pr_err("%s zpool not available\n", type);
+			goto error;
+		}
 	}
-	pr_debug("using %s zpool\n", zpool_get_type(pool->zpool));
+	pr_debug("using %s zpool\n", zpool_get_type(pool->zpools[0]));
 
 	strscpy(pool->tfm_name, compressor, sizeof(pool->tfm_name));
 
@@ -755,8 +775,8 @@ static struct zswap_pool *zswap_pool_create(char *type, char *compressor)
 error:
 	if (pool->acomp_ctx)
 		free_percpu(pool->acomp_ctx);
-	if (pool->zpool)
-		zpool_destroy_pool(pool->zpool);
+	while (i--)
+		zpool_destroy_pool(pool->zpools[i]);
 	kfree(pool);
 	return NULL;
 }
@@ -805,11 +825,14 @@ static struct zswap_pool *__zswap_pool_create_fallback(void)
 
 static void zswap_pool_destroy(struct zswap_pool *pool)
 {
+	int i;
+
 	zswap_pool_debug("destroying", pool);
 
 	cpuhp_state_remove_instance(CPUHP_MM_ZSWP_POOL_PREPARE, &pool->node);
 	free_percpu(pool->acomp_ctx);
-	zpool_destroy_pool(pool->zpool);
+	for (i = 0; i < ZSWAP_NR_ZPOOLS; i++)
+		zpool_destroy_pool(pool->zpools[i]);
 	kfree(pool);
 }
 
@@ -1073,7 +1096,7 @@ static int zswap_writeback_entry(struct zswap_entry *entry,
 	struct page *page;
 	struct scatterlist input, output;
 	struct crypto_acomp_ctx *acomp_ctx;
-	struct zpool *pool = entry->pool->zpool;
+	struct zpool *pool = zswap_find_zpool(entry);
 
 	u8 *src, *tmp = NULL;
 	unsigned int dlen;
@@ -1214,6 +1237,7 @@ static int zswap_frontswap_store(unsigned type, pgoff_t offset,
 	struct crypto_acomp_ctx *acomp_ctx;
 	struct obj_cgroup *objcg = NULL;
 	struct zswap_pool *pool;
+	struct zpool *zpool;
 	int ret;
 	unsigned int dlen = PAGE_SIZE;
 	unsigned long handle, value;
@@ -1324,10 +1348,11 @@ static int zswap_frontswap_store(unsigned type, pgoff_t offset,
 	}
 
 	/* store */
+	zpool = zswap_find_zpool(entry);
 	gfp = __GFP_NORETRY | __GFP_NOWARN | __GFP_KSWAPD_RECLAIM;
-	if (zpool_malloc_support_movable(entry->pool->zpool))
+	if (zpool_malloc_support_movable(zpool))
 		gfp |= __GFP_HIGHMEM | __GFP_MOVABLE;
-	ret = zpool_malloc(entry->pool->zpool, dlen, gfp, &handle);
+	ret = zpool_malloc(zpool, dlen, gfp, &handle);
 	if (ret == -ENOSPC) {
 		zswap_reject_compress_poor++;
 		goto put_dstmem;
@@ -1336,9 +1361,9 @@ static int zswap_frontswap_store(unsigned type, pgoff_t offset,
 		zswap_reject_alloc_fail++;
 		goto put_dstmem;
 	}
-	buf = zpool_map_handle(entry->pool->zpool, handle, ZPOOL_MM_WO);
+	buf = zpool_map_handle(zpool, handle, ZPOOL_MM_WO);
 	memcpy(buf, dst, dlen);
-	zpool_unmap_handle(entry->pool->zpool, handle);
+	zpool_unmap_handle(zpool, handle);
 	mutex_unlock(acomp_ctx->mutex);
 
 	/* populate entry */
@@ -1409,6 +1434,7 @@ static int zswap_frontswap_load(unsigned type, pgoff_t offset,
 	struct scatterlist input, output;
 	struct crypto_acomp_ctx *acomp_ctx;
 	u8 *src, *dst, *tmp;
+	struct zpool *zpool;
 	unsigned int dlen;
 	int ret;
 
@@ -1430,7 +1456,8 @@ static int zswap_frontswap_load(unsigned type, pgoff_t offset,
 		goto stats;
 	}
 
-	if (!zpool_can_sleep_mapped(entry->pool->zpool)) {
+	zpool = zswap_find_zpool(entry);
+	if (!zpool_can_sleep_mapped(zpool)) {
 		tmp = kmalloc(entry->length, GFP_KERNEL);
 		if (!tmp) {
 			ret = -ENOMEM;
@@ -1440,12 +1467,12 @@ 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 = zpool_map_handle(zpool, entry->handle, ZPOOL_MM_RO);
 
-	if (!zpool_can_sleep_mapped(entry->pool->zpool)) {
+	if (!zpool_can_sleep_mapped(zpool)) {
 		memcpy(tmp, src, entry->length);
 		src = tmp;
-		zpool_unmap_handle(entry->pool->zpool, entry->handle);
+		zpool_unmap_handle(zpool, entry->handle);
 	}
 
 	acomp_ctx = raw_cpu_ptr(entry->pool->acomp_ctx);
@@ -1457,8 +1484,8 @@ static int zswap_frontswap_load(unsigned type, pgoff_t offset,
 	ret = crypto_wait_req(crypto_acomp_decompress(acomp_ctx->req), &acomp_ctx->wait);
 	mutex_unlock(acomp_ctx->mutex);
 
-	if (zpool_can_sleep_mapped(entry->pool->zpool))
-		zpool_unmap_handle(entry->pool->zpool, entry->handle);
+	if (zpool_can_sleep_mapped(zpool))
+		zpool_unmap_handle(zpool, entry->handle);
 	else
 		kfree(tmp);
 
@@ -1619,7 +1646,7 @@ static int zswap_setup(void)
 	pool = __zswap_pool_create_fallback();
 	if (pool) {
 		pr_info("loaded using pool %s/%s\n", pool->tfm_name,
-			zpool_get_type(pool->zpool));
+			zpool_get_type(pool->zpools[0]));
 		list_add(&pool->list, &zswap_pools);
 		zswap_has_pool = true;
 	} else {