Re: [PATCH v3 2/5] zswap: make shrinking memcg-aware

On Wed, Oct 18, 2023 at 4:46 PM Nhat Pham nphamcs@gmail.com wrote:

On Wed, Oct 18, 2023 at 4:20 PM Yosry Ahmed yosryahmed@google.com wrote:

...

On Tue, Oct 17, 2023 at 4:21 PM Nhat Pham nphamcs@gmail.com wrote:

...

From: Domenico Cerasuolo cerasuolodomenico@gmail.com

Currently, we only have a single global LRU for zswap. This makes it impossible to perform worload-specific shrinking - an memcg cannot determine which pages in the pool it owns, and often ends up writing pages from other memcgs. This issue has been previously observed in practice and mitigated by simply disabling memcg-initiated shrinking:

https://lore.kernel.org/all/20230530232435.3097106-1-nphamcs@gmail.com/T/#u

This patch fully resolves the issue by replacing the global zswap LRU with memcg- and NUMA-specific LRUs, and modify the reclaim logic:

a) When a store attempt hits an memcg limit, it now triggers a synchronous reclaim attempt that, if successful, allows the new hotter page to be accepted by zswap. b) If the store attempt instead hits the global zswap limit, it will trigger an asynchronous reclaim attempt, in which an memcg is selected for reclaim in a round-robin-like fashion.

Could you explain the rationale behind the difference in behavior here between the global limit and the memcg limit?

The global limit hit reclaim behavior was previously asynchronous too. We just added the round-robin part because now the zswap LRU is cgroup-aware :)

For the cgroup limit hit, however, we cannot make it asynchronous, as it is a bit hairy to add a per-cgroup shrink_work. So, we just perform the reclaim synchronously.

The question is whether it makes sense to make the global limit reclaim synchronous too. That is a task of its own IMO.

(FWIW, this somewhat mirrors the direct reclaimer v.s kswapd story to me, but don't quote me too hard on this).

...

...

Signed-off-by: Domenico Cerasuolo cerasuolodomenico@gmail.com Co-developed-by: Nhat Pham nphamcs@gmail.com Signed-off-by: Nhat Pham nphamcs@gmail.com

---

include/linux/memcontrol.h | 5 ++ mm/swap.h | 3 +- mm/swap_state.c | 17 +++- mm/zswap.c | 179 ++++++++++++++++++++++++++----------- 4 files changed, 147 insertions(+), 57 deletions(-)

This is a dense patch, I haven't absorbed all of it yet, but the first round of comments below.

Regardless, thanks for the feedback, Yosry! Domenico definitely knows more than me about this, but I'll respond with what I know, and he can expand and/or fact-check me :)

...

...

diff --git a/include/linux/memcontrol.h b/include/linux/memcontrol.h index 031102ac9311..3de10fabea0f 100644 --- a/include/linux/memcontrol.h +++ b/include/linux/memcontrol.h @@ -1179,6 +1179,11 @@ static inline struct mem_cgroup *page_memcg_check(struct page *page) return NULL; }

+static inline struct mem_cgroup *get_mem_cgroup_from_objcg(struct obj_cgroup *objcg) +{

•   return NULL;
  

+}

static inline bool folio_memcg_kmem(struct folio *folio) { return false; diff --git a/mm/swap.h b/mm/swap.h index 8a3c7a0ace4f..bbd6ce661a20 100644 --- a/mm/swap.h +++ b/mm/swap.h @@ -50,7 +50,8 @@ struct page *read_swap_cache_async(swp_entry_t entry, gfp_t gfp_mask, struct page *__read_swap_cache_async(swp_entry_t entry, gfp_t gfp_mask, struct vm_area_struct *vma, unsigned long addr,

•                                bool *new_page_allocated);
  

•                                bool *new_page_allocated,
  

•                                bool fail_if_exists);
  

struct page *swap_cluster_readahead(swp_entry_t entry, gfp_t flag, struct vm_fault *vmf); struct page *swapin_readahead(swp_entry_t entry, gfp_t flag, diff --git a/mm/swap_state.c b/mm/swap_state.c index b3b14bd0dd64..0356df52b06a 100644 --- a/mm/swap_state.c +++ b/mm/swap_state.c @@ -411,7 +411,7 @@ struct folio *filemap_get_incore_folio(struct address_space *mapping,

struct page *__read_swap_cache_async(swp_entry_t entry, gfp_t gfp_mask, struct vm_area_struct *vma, unsigned long addr,

•                   bool *new_page_allocated)
  

•                   bool *new_page_allocated, bool fail_if_exists)
  

nit: I don't feel like "fail" is the correct word here. Perhaps "skip"?

...

{ struct swap_info_struct *si; struct folio *folio; @@ -468,6 +468,15 @@ struct page *__read_swap_cache_async(swp_entry_t entry, gfp_t gfp_mask, if (err != -EEXIST) goto fail_put_swap;

•           /*
  

•            * This check guards against a state that happens if a call
  

•            * to __read_swap_cache_async triggers a reclaim, if the
  

•            * reclaimer (zswap's writeback as of now) then decides to
  

•            * reclaim that same entry, then the subsequent call to
  

•            * __read_swap_cache_async would get stuck in this loop.
  

I think this comment needs to first state that it is protecting against a recursive call in general, not necessarily in reclaim, as __read_swap_cache_async() is not usually called in the context of reclaim so this can be confusing. Then it can give the exact example we have today. Perhaps something like:

Protect against a recursive call to __read_swap_cache_async() on the same entry waiting forever here because SWAP_HAS_CACHE is set but the folio is not the swap cache yet. This can happen today if mem_cgroup_swapin_charge_folio() below triggers reclaim through zswap, which may call __read_swap_cache_async() in the writeback path.

...

•            */
  

•           if (fail_if_exists && err == -EEXIST)
  

We already made sure in the preceding condition that err is -EEXIST.

...

•                   goto fail_put_swap;
            /*
             * We might race against __delete_from_swap_cache(), and
             * stumble across a swap_map entry whose SWAP_HAS_CACHE
  

@@ -530,7 +539,7 @@ struct page *read_swap_cache_async(swp_entry_t entry, gfp_t gfp_mask, { bool page_was_allocated; struct page *retpage = __read_swap_cache_async(entry, gfp_mask,

•                   vma, addr, &page_was_allocated);
  

•                   vma, addr, &page_was_allocated, false);
  
    if (page_was_allocated)
            swap_readpage(retpage, false, plug);
  

@@ -649,7 +658,7 @@ struct page *swap_cluster_readahead(swp_entry_t entry, gfp_t gfp_mask, /* Ok, do the async read-ahead now */ page = __read_swap_cache_async( swp_entry(swp_type(entry), offset),

•                   gfp_mask, vma, addr, &page_allocated);
  

•                   gfp_mask, vma, addr, &page_allocated, false);
            if (!page)
                    continue;
            if (page_allocated) {
  

@@ -815,7 +824,7 @@ static struct page *swap_vma_readahead(swp_entry_t fentry, gfp_t gfp_mask, pte_unmap(pte); pte = NULL; page = __read_swap_cache_async(entry, gfp_mask, vma,

•                                          addr, &page_allocated);
  

•                                          addr, &page_allocated, false);
            if (!page)
                    continue;
            if (page_allocated) {
  

diff --git a/mm/zswap.c b/mm/zswap.c index 083c693602b8..d2989ad11814 100644 --- a/mm/zswap.c +++ b/mm/zswap.c @@ -34,6 +34,7 @@ #include <linux/writeback.h> #include <linux/pagemap.h> #include <linux/workqueue.h> +#include <linux/list_lru.h>

#include "swap.h" #include "internal.h" @@ -171,8 +172,8 @@ 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 lru_lock;
  

•   struct list_lru list_lru;
  

•   struct mem_cgroup *next_shrink;
  

};

/* @@ -288,15 +289,25 @@ static void zswap_update_total_size(void) zswap_pool_total_size = total; }

+static inline struct mem_cgroup *get_mem_cgroup_from_entry(struct zswap_entry *entry) +{

•   return entry->objcg ? get_mem_cgroup_from_objcg(entry->objcg) : NULL;
  

+}

+static inline int entry_to_nid(struct zswap_entry *entry) +{

•   return page_to_nid(virt_to_page(entry));
  

+}

/*********************************

• zswap entry functions

**********************************/ static struct kmem_cache *zswap_entry_cache;

-static struct zswap_entry *zswap_entry_cache_alloc(gfp_t gfp) +static struct zswap_entry *zswap_entry_cache_alloc(gfp_t gfp, int nid) { struct zswap_entry *entry;

•   entry = kmem_cache_alloc(zswap_entry_cache, gfp);
  

•   entry = kmem_cache_alloc_node(zswap_entry_cache, gfp, nid);
    if (!entry)
            return NULL;
    entry->refcount = 1;
  

@@ -309,6 +320,27 @@ static void zswap_entry_cache_free(struct zswap_entry *entry) kmem_cache_free(zswap_entry_cache, entry); }

+/********************************* +* lru functions +**********************************/ +static bool zswap_lru_add(struct list_lru *list_lru, struct zswap_entry *entry) +{

•   struct mem_cgroup *memcg = get_mem_cgroup_from_entry(entry);
  

Could we avoid the need for get/put with an rcu_read_lock() instead?

...

•   bool added = __list_lru_add(list_lru, &entry->lru, entry_to_nid(entry), memcg);
  

•   mem_cgroup_put(memcg);
  

•   return added;
  

+}

+static bool zswap_lru_del(struct list_lru *list_lru, struct zswap_entry *entry) +{

•   struct mem_cgroup *memcg = get_mem_cgroup_from_entry(entry);
  

•   bool removed = __list_lru_del(list_lru, &entry->lru, entry_to_nid(entry), memcg);
  

•   mem_cgroup_put(memcg);
  

•   return removed;
  

+}

/*********************************

• rbtree functions

**********************************/ @@ -393,9 +425,7 @@ static void zswap_free_entry(struct zswap_entry *entry) if (!entry->length) atomic_dec(&zswap_same_filled_pages); else {

•           spin_lock(&entry->pool->lru_lock);
  

•           list_del(&entry->lru);
  

•           spin_unlock(&entry->pool->lru_lock);
  

•           zswap_lru_del(&entry->pool->list_lru, entry);
            zpool_free(zswap_find_zpool(entry), entry->handle);
            zswap_pool_put(entry->pool);
    }
  

@@ -629,21 +659,16 @@ static void zswap_invalidate_entry(struct zswap_tree *tree, zswap_entry_put(tree, entry); }

-static int zswap_reclaim_entry(struct zswap_pool *pool) +static enum lru_status shrink_memcg_cb(struct list_head *item, struct list_lru_one *l,

•                                  spinlock_t *lock, void *arg)
  

{

•   struct zswap_entry *entry;
  

•   struct zswap_entry *entry = container_of(item, struct zswap_entry, lru);
  

•   struct mem_cgroup *memcg;
    struct zswap_tree *tree;
    pgoff_t swpoffset;
  

•   int ret;
  

•   enum lru_status ret = LRU_REMOVED_RETRY;
  

•   int writeback_result;
  
  

•   /* Get an entry off the LRU */
  

•   spin_lock(&pool->lru_lock);
  

•   if (list_empty(&pool->lru)) {
  

•           spin_unlock(&pool->lru_lock);
  

•           return -EINVAL;
  

•   }
  

•   entry = list_last_entry(&pool->lru, struct zswap_entry, lru);
  

•   list_del_init(&entry->lru);
    /*
     * Once the lru lock is dropped, the entry might get freed. The
     * swpoffset is copied to the stack, and entry isn't deref'd again
  

@@ -651,28 +676,33 @@ static int zswap_reclaim_entry(struct zswap_pool *pool) */ swpoffset = swp_offset(entry->swpentry); tree = zswap_trees[swp_type(entry->swpentry)];

•   spin_unlock(&pool->lru_lock);
  

•   list_lru_isolate(l, item);
  

•   spin_unlock(lock);
  

Perhaps a comment somewhere stating that we only return either LRU_REMOVED_RETRY or LRU_RETRY, so it's fine to drop and reacquire the lock.

...

    /* Check for invalidate() race */
    spin_lock(&tree->lock);
    if (entry != zswap_rb_search(&tree->rbroot, swpoffset)) {

•           ret = -EAGAIN;
            goto unlock;
    }
  

nit: braces no longer needed?

Ah, for some reason checkpatch did not pick up on this. Weird.

...

...

    /* Hold a reference to prevent a free during writeback */
    zswap_entry_get(entry);
    spin_unlock(&tree->lock);

•   ret = zswap_writeback_entry(entry, tree);
  

•   writeback_result = zswap_writeback_entry(entry, tree);
  
    spin_lock(&tree->lock);
  

•   if (ret) {
  

•           /* Writeback failed, put entry back on LRU */
  

•           spin_lock(&pool->lru_lock);
  

•           list_move(&entry->lru, &pool->lru);
  

•           spin_unlock(&pool->lru_lock);
  

•   if (writeback_result) {
  

•           zswap_reject_reclaim_fail++;
  

•           memcg = get_mem_cgroup_from_entry(entry);
  

•           spin_lock(lock);
  

•           /* we cannot use zswap_lru_add here, because it increments node's lru count */
  

•           list_lru_putback(&entry->pool->list_lru, item, entry_to_nid(entry), memcg);
  

•           spin_unlock(lock);
  

•           mem_cgroup_put(memcg);
  

•           ret = LRU_RETRY;
            goto put_unlock;
    }
  

•   zswap_written_back_pages++;
  

Why is this moved here from zswap_writeback_entry()? Also why is zswap_reject_reclaim_fail incremented here instead of inside zswap_writeback_entry()?

Domenico should know this better than me, but my understanding is that moving it here protects concurrent modifications of zswap_written_back_pages with the tree lock.

Is writeback single-threaded in the past? This counter is non-atomic, and doesn't seem to be protected by any locks...

There definitely can be concurrent stores now though - with

/s/stores/writebacks

a synchronous reclaim from cgroup-limit hit and another from the old shrink worker.

(and with the new zswap shrinker, concurrent reclaim is the expectation!)

zswap_reject_reclaim_fail was previously incremented in shrink_worker I think. We need it to be incremented for the shrinker as well, so might as well move it here.

...

...

    /*
     * Writeback started successfully, the page now belongs to the

@@ -686,7 +716,36 @@ static int zswap_reclaim_entry(struct zswap_pool *pool) zswap_entry_put(tree, entry); unlock: spin_unlock(&tree->lock);

•   return ret ? -EAGAIN : 0;
  

•   spin_lock(lock);
  

•   return ret;
  

+}

+static int shrink_memcg(struct mem_cgroup *memcg) +{

•   struct zswap_pool *pool;
  

•   int nid, shrunk = 0;
  

•   pool = zswap_pool_current_get();
  

•   if (!pool)
  

•           return -EINVAL;
  

•   /*
  

•    * Skip zombies because their LRUs are reparented and we would be
  

•    * reclaiming from the parent instead of the dead memcgroup.
  

nit: s/memcgroup/memcg.

...

•    */
  

•   if (memcg && !mem_cgroup_online(memcg))
  

•           goto out;
  

If we move this above zswap_pool_current_get(), we can return directly and remove the label. I noticed we will return -EAGAIN if memcg is offline. IIUC -EAGAIN for the caller will move on to the next memcg, but I am wondering if a different errno would be clearer here.

...

•   for_each_node_state(nid, N_NORMAL_MEMORY) {
  

•           unsigned long nr_to_walk = 1;
  

•           if (list_lru_walk_one(&pool->list_lru, nid, memcg, &shrink_memcg_cb,
  

•                                 NULL, &nr_to_walk))
  

•                   shrunk++;
  

nit: shrunk += list_lru_walk_one(..);

yeah might be a tad cleaner.

...

...

•   }
  

+out:

•   zswap_pool_put(pool);
  

•   return shrunk ? 0 : -EAGAIN;
  

}

static void shrink_worker(struct work_struct *w) @@ -695,10 +754,13 @@ static void shrink_worker(struct work_struct *w) shrink_work); int ret, failures = 0;

•   /* global reclaim will select cgroup in a round-robin fashion. */
    do {
  

•           ret = zswap_reclaim_entry(pool);
  

•           pool->next_shrink = mem_cgroup_iter(NULL, pool->next_shrink, NULL);
  

Perhaps next_shrink_memcg is a better name here?

...

•           ret = shrink_memcg(pool->next_shrink);
  

•           if (ret) {
  

•                   zswap_reject_reclaim_fail++;
                    if (ret != -EAGAIN)
                            break;
                    if (++failures == MAX_RECLAIM_RETRIES)
  

@@ -764,8 +826,7 @@ 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->lru_lock);
  

•   list_lru_init_memcg(&pool->list_lru, NULL);
    INIT_WORK(&pool->shrink_work, shrink_worker);
  
    zswap_pool_debug("created", pool);
  

@@ -831,6 +892,9 @@ static void zswap_pool_destroy(struct zswap_pool *pool)

    cpuhp_state_remove_instance(CPUHP_MM_ZSWP_POOL_PREPARE, &pool->node);
    free_percpu(pool->acomp_ctx);

•   list_lru_destroy(&pool->list_lru);
  

•   if (pool->next_shrink)
  

•           mem_cgroup_put(pool->next_shrink);
    for (i = 0; i < ZSWAP_NR_ZPOOLS; i++)
            zpool_destroy_pool(pool->zpools[i]);
    kfree(pool);
  

@@ -1076,7 +1140,7 @@ static int zswap_writeback_entry(struct zswap_entry *entry,

    /* try to allocate swap cache page */
    page = __read_swap_cache_async(swpentry, GFP_KERNEL, NULL, 0,

•                                  &page_was_allocated);
  

•                                  &page_was_allocated, true);
    if (!page) {
            ret = -ENOMEM;
            goto fail;
  

@@ -1142,7 +1206,6 @@ static int zswap_writeback_entry(struct zswap_entry *entry, /* start writeback */ __swap_writepage(page, &wbc); put_page(page);

•   zswap_written_back_pages++;
  
    return ret;
  
  

@@ -1199,8 +1262,10 @@ bool zswap_store(struct folio *folio) struct scatterlist input, output; struct crypto_acomp_ctx *acomp_ctx; struct obj_cgroup *objcg = NULL;

•   struct mem_cgroup *memcg = NULL;
    struct zswap_pool *pool;
    struct zpool *zpool;
  

•   int lru_alloc_ret;
    unsigned int dlen = PAGE_SIZE;
    unsigned long handle, value;
    char *buf;
  

@@ -1230,15 +1295,15 @@ bool zswap_store(struct folio *folio) zswap_invalidate_entry(tree, dupentry); } spin_unlock(&tree->lock);

•   /*
  

•    * XXX: zswap reclaim does not work with cgroups yet. Without a
  

•    * cgroup-aware entry LRU, we will push out entries system-wide based on
  

•    * local cgroup limits.
  

•    */
    objcg = get_obj_cgroup_from_folio(folio);
  

•   if (objcg && !obj_cgroup_may_zswap(objcg))
  

•           goto reject;
  

•   if (objcg && !obj_cgroup_may_zswap(objcg)) {
  

•           memcg = get_mem_cgroup_from_objcg(objcg);
  

•           if (shrink_memcg(memcg)) {
  

•                   mem_cgroup_put(memcg);
  

•                   goto reject;
  

•           }
  

•           mem_cgroup_put(memcg);
  

•   }
  
    /* reclaim space if needed */
    if (zswap_is_full()) {
  

@@ -1254,10 +1319,15 @@ bool zswap_store(struct folio *folio) zswap_pool_reached_full = false; }

•   pool = zswap_pool_current_get();
  

•   if (!pool)
  

•           goto reject;
  

Why do we need to move zswap_pool_current_get() up here?

...

    /* allocate entry */

•   entry = zswap_entry_cache_alloc(GFP_KERNEL);
  

•   entry = zswap_entry_cache_alloc(GFP_KERNEL, page_to_nid(page));
    if (!entry) {
            zswap_reject_kmemcache_fail++;
  

•           zswap_pool_put(pool);
            goto reject;
    }
  
  

@@ -1269,6 +1339,7 @@ bool zswap_store(struct folio *folio) entry->length = 0; entry->value = value; atomic_inc(&zswap_same_filled_pages);

•                   zswap_pool_put(pool);
                    goto insert_entry;
            }
            kunmap_atomic(src);
  

@@ -1278,9 +1349,15 @@ bool zswap_store(struct folio *folio) goto freepage;

    /* if entry is successfully added, it keeps the reference */

•   entry->pool = zswap_pool_current_get();
  

•   if (!entry->pool)
  

•           goto freepage;
  

•   entry->pool = pool;
  

•   if (objcg) {
  

•           memcg = get_mem_cgroup_from_objcg(objcg);
  

•           lru_alloc_ret = memcg_list_lru_alloc(memcg, &pool->list_lru, GFP_KERNEL);
  

•           mem_cgroup_put(memcg);
  

•           if (lru_alloc_ret)
  

•                   goto freepage;
  

•   }
  
    /* compress */
    acomp_ctx = raw_cpu_ptr(entry->pool->acomp_ctx);
  

@@ -1358,9 +1435,8 @@ bool zswap_store(struct folio *folio) zswap_invalidate_entry(tree, dupentry); } if (entry->length) {

•           spin_lock(&entry->pool->lru_lock);
  

•           list_add(&entry->lru, &entry->pool->lru);
  

•           spin_unlock(&entry->pool->lru_lock);
  

•           INIT_LIST_HEAD(&entry->lru);
  

•           zswap_lru_add(&pool->list_lru, entry);
    }
    spin_unlock(&tree->lock);
  
  

@@ -1373,8 +1449,8 @@ bool zswap_store(struct folio *folio)

put_dstmem: mutex_unlock(acomp_ctx->mutex);

•   zswap_pool_put(entry->pool);
  

freepage:

•   zswap_pool_put(entry->pool);
    zswap_entry_cache_free(entry);
  

reject: if (objcg) @@ -1467,9 +1543,8 @@ bool zswap_load(struct folio *folio) zswap_invalidate_entry(tree, entry); folio_mark_dirty(folio); } else if (entry->length) {

•           spin_lock(&entry->pool->lru_lock);
  

•           list_move(&entry->lru, &entry->pool->lru);
  

•           spin_unlock(&entry->pool->lru_lock);
  

•           zswap_lru_del(&entry->pool->list_lru, entry);
  

•           zswap_lru_add(&entry->pool->list_lru, entry);
    }
    zswap_entry_put(tree, entry);
    spin_unlock(&tree->lock);
  

-- 2.34.1

I don't have (strong) opinions or (educated) guesses on the rest.