The patch titled
Subject: mm: zswap: optimize zswap pool size tracking
has been added to the -mm mm-unstable branch. Its filename is
mm-zswap-optimize-zswap-pool-size-tracking.patch
This patch will shortly appear at
https://git.kernel.org/pub/scm/linux/kernel/git/akpm/25-new.git/tree/patches/mm-zswap-optimize-zswap-pool-size-tracking.patch
This patch will later appear in the mm-unstable branch at
git://git.kernel.org/pub/scm/linux/kernel/git/akpm/mm
Before you just go and hit "reply", please:
a) Consider who else should be cc'ed
b) Prefer to cc a suitable mailing list as well
c) Ideally: find the original patch on the mailing list and do a
reply-to-all to that, adding suitable additional cc's
*** Remember to use Documentation/process/submit-checklist.rst when testing your code ***
The -mm tree is included into linux-next via the mm-everything
branch at git://git.kernel.org/pub/scm/linux/kernel/git/akpm/mm
and is updated there every 2-3 working days
------------------------------------------------------
From: Johannes Weiner <hannes@xxxxxxxxxxx>
Subject: mm: zswap: optimize zswap pool size tracking
Date: Tue, 12 Mar 2024 11:34:11 -0400
Profiling the munmap() of a zswapped memory region shows 60% of the total
cycles currently going into updating the zswap_pool_total_size.
There are three consumers of this counter:
- store, to enforce the globally configured pool limit
- meminfo & debugfs, to report the size to the user
- shrink, to determine the batch size for each cycle
Instead of aggregating everytime an entry enters or exits the zswap
pool, aggregate the value from the zpools on-demand:
- Stores aggregate the counter anyway upon success. Aggregating to
check the limit instead is the same amount of work.
- Meminfo & debugfs might benefit somewhat from a pre-aggregated
counter, but aren't exactly hotpaths.
- Shrinking can aggregate once for every cycle instead of doing it for
every freed entry. As the shrinker might work on tens or hundreds of
objects per scan cycle, this is a large reduction in aggregations.
The paths that benefit dramatically are swapin, swapoff, and unmaps.
There could be millions of pages being processed until somebody asks for
the pool size again. This eliminates the pool size updates from those
paths entirely.
Top profile entries for a 24G range munmap(), before:
38.54% zswap-unmap [kernel.kallsyms] [k] zs_zpool_total_size
12.51% zswap-unmap [kernel.kallsyms] [k] zpool_get_total_size
9.10% zswap-unmap [kernel.kallsyms] [k] zswap_update_total_size
2.95% zswap-unmap [kernel.kallsyms] [k] obj_cgroup_uncharge_zswap
2.88% zswap-unmap [kernel.kallsyms] [k] __slab_free
2.86% zswap-unmap [kernel.kallsyms] [k] xas_store
and after:
7.70% zswap-unmap [kernel.kallsyms] [k] __slab_free
7.16% zswap-unmap [kernel.kallsyms] [k] obj_cgroup_uncharge_zswap
6.74% zswap-unmap [kernel.kallsyms] [k] xas_store
It was also briefly considered to move to a single atomic in zswap
that is updated by the backends, since zswap only cares about the sum
of all pools anyway. However, zram directly needs per-pool information
out of zsmalloc. To keep the backend from having to update two atomics
every time, I opted for the lazy aggregation instead for now.
Link: https://lkml.kernel.org/r/20240312153901.3441-1-hannes@xxxxxxxxxxx
Signed-off-by: Johannes Weiner <hannes@xxxxxxxxxxx>
Acked-by: Yosry Ahmed <yosryahmed@xxxxxxxxxx>
Reviewed-by: Chengming Zhou <chengming.zhou@xxxxxxxxx>
Reviewed-by: Nhat Pham <nphamcs@xxxxxxxxx>
Signed-off-by: Andrew Morton <akpm@xxxxxxxxxxxxxxxxxxxx>
---
fs/proc/meminfo.c | 3 -
include/linux/zswap.h | 2
mm/zswap.c | 101 +++++++++++++++++++---------------------
3 files changed, 52 insertions(+), 54 deletions(-)
--- a/fs/proc/meminfo.c~mm-zswap-optimize-zswap-pool-size-tracking
+++ a/fs/proc/meminfo.c
@@ -89,8 +89,7 @@ static int meminfo_proc_show(struct seq_
show_val_kb(m, "SwapTotal: ", i.totalswap);
show_val_kb(m, "SwapFree: ", i.freeswap);
#ifdef CONFIG_ZSWAP
- seq_printf(m, "Zswap: %8lu kB\n",
- (unsigned long)(zswap_pool_total_size >> 10));
+ show_val_kb(m, "Zswap: ", zswap_total_pages());
seq_printf(m, "Zswapped: %8lu kB\n",
(unsigned long)atomic_read(&zswap_stored_pages) <<
(PAGE_SHIFT - 10));
--- a/include/linux/zswap.h~mm-zswap-optimize-zswap-pool-size-tracking
+++ a/include/linux/zswap.h
@@ -7,7 +7,6 @@
struct lruvec;
-extern u64 zswap_pool_total_size;
extern atomic_t zswap_stored_pages;
#ifdef CONFIG_ZSWAP
@@ -27,6 +26,7 @@ struct zswap_lruvec_state {
atomic_long_t nr_zswap_protected;
};
+unsigned long zswap_total_pages(void);
bool zswap_store(struct folio *folio);
bool zswap_load(struct folio *folio);
void zswap_invalidate(swp_entry_t swp);
--- a/mm/zswap.c~mm-zswap-optimize-zswap-pool-size-tracking
+++ a/mm/zswap.c
@@ -43,8 +43,6 @@
/*********************************
* statistics
**********************************/
-/* Total bytes used by the compressed storage */
-u64 zswap_pool_total_size;
/* The number of compressed pages currently stored in zswap */
atomic_t zswap_stored_pages = ATOMIC_INIT(0);
/* The number of same-value filled pages currently stored in zswap */
@@ -265,45 +263,6 @@ static inline struct zswap_tree *swap_zs
pr_debug("%s pool %s/%s\n", msg, (p)->tfm_name, \
zpool_get_type((p)->zpools[0]))
-static bool zswap_is_full(void)
-{
- return totalram_pages() * zswap_max_pool_percent / 100 <
- DIV_ROUND_UP(zswap_pool_total_size, PAGE_SIZE);
-}
-
-static bool zswap_can_accept(void)
-{
- return totalram_pages() * zswap_accept_thr_percent / 100 *
- zswap_max_pool_percent / 100 >
- DIV_ROUND_UP(zswap_pool_total_size, PAGE_SIZE);
-}
-
-static u64 get_zswap_pool_size(struct zswap_pool *pool)
-{
- u64 pool_size = 0;
- int i;
-
- for (i = 0; i < ZSWAP_NR_ZPOOLS; i++)
- pool_size += zpool_get_total_size(pool->zpools[i]);
-
- return pool_size;
-}
-
-static void zswap_update_total_size(void)
-{
- struct zswap_pool *pool;
- u64 total = 0;
-
- rcu_read_lock();
-
- list_for_each_entry_rcu(pool, &zswap_pools, list)
- total += get_zswap_pool_size(pool);
-
- rcu_read_unlock();
-
- zswap_pool_total_size = total;
-}
-
/*********************************
* pool functions
**********************************/
@@ -541,6 +500,33 @@ static struct zswap_pool *zswap_pool_fin
return NULL;
}
+static unsigned long zswap_max_pages(void)
+{
+ return totalram_pages() * zswap_max_pool_percent / 100;
+}
+
+static unsigned long zswap_accept_thr_pages(void)
+{
+ return zswap_max_pages() * zswap_accept_thr_percent / 100;
+}
+
+unsigned long zswap_total_pages(void)
+{
+ struct zswap_pool *pool;
+ u64 total = 0;
+
+ rcu_read_lock();
+ list_for_each_entry_rcu(pool, &zswap_pools, list) {
+ int i;
+
+ for (i = 0; i < ZSWAP_NR_ZPOOLS; i++)
+ total += zpool_get_total_size(pool->zpools[i]);
+ }
+ rcu_read_unlock();
+
+ return total >> PAGE_SHIFT;
+}
+
/*********************************
* param callbacks
**********************************/
@@ -913,7 +899,6 @@ static void zswap_entry_free(struct zswa
}
zswap_entry_cache_free(entry);
atomic_dec(&zswap_stored_pages);
- zswap_update_total_size();
}
/*
@@ -1329,7 +1314,7 @@ static unsigned long zswap_shrinker_coun
nr_stored = memcg_page_state(memcg, MEMCG_ZSWAPPED);
#else
/* use pool stats instead of memcg stats */
- nr_backing = zswap_pool_total_size >> PAGE_SHIFT;
+ nr_backing = zswap_total_pages();
nr_stored = atomic_read(&zswap_nr_stored);
#endif
@@ -1397,6 +1382,10 @@ static void shrink_worker(struct work_st
{
struct mem_cgroup *memcg;
int ret, failures = 0;
+ unsigned long thr;
+
+ /* Reclaim down to the accept threshold */
+ thr = zswap_accept_thr_pages();
/* global reclaim will select cgroup in a round-robin fashion. */
do {
@@ -1444,10 +1433,9 @@ static void shrink_worker(struct work_st
break;
if (ret && ++failures == MAX_RECLAIM_RETRIES)
break;
-
resched:
cond_resched();
- } while (!zswap_can_accept());
+ } while (zswap_total_pages() > thr);
}
static int zswap_is_page_same_filled(void *ptr, unsigned long *value)
@@ -1488,6 +1476,7 @@ bool zswap_store(struct folio *folio)
struct zswap_entry *entry, *dupentry;
struct obj_cgroup *objcg = NULL;
struct mem_cgroup *memcg = NULL;
+ unsigned long max_pages, cur_pages;
VM_WARN_ON_ONCE(!folio_test_locked(folio));
VM_WARN_ON_ONCE(!folio_test_swapcache(folio));
@@ -1499,6 +1488,7 @@ bool zswap_store(struct folio *folio)
if (!zswap_enabled)
goto check_old;
+ /* Check cgroup limits */
objcg = get_obj_cgroup_from_folio(folio);
if (objcg && !obj_cgroup_may_zswap(objcg)) {
memcg = get_mem_cgroup_from_objcg(objcg);
@@ -1509,15 +1499,18 @@ bool zswap_store(struct folio *folio)
mem_cgroup_put(memcg);
}
- /* reclaim space if needed */
- if (zswap_is_full()) {
+ /* Check global limits */
+ cur_pages = zswap_total_pages();
+ max_pages = zswap_max_pages();
+
+ if (cur_pages >= max_pages) {
zswap_pool_limit_hit++;
zswap_pool_reached_full = true;
goto shrink;
}
if (zswap_pool_reached_full) {
- if (!zswap_can_accept())
+ if (cur_pages > zswap_accept_thr_pages())
goto shrink;
else
zswap_pool_reached_full = false;
@@ -1593,7 +1586,6 @@ insert_entry:
/* update stats */
atomic_inc(&zswap_stored_pages);
- zswap_update_total_size();
count_vm_event(ZSWPOUT);
return true;
@@ -1722,6 +1714,13 @@ void zswap_swapoff(int type)
static struct dentry *zswap_debugfs_root;
+static int debugfs_get_total_size(void *data, u64 *val)
+{
+ *val = zswap_total_pages() * PAGE_SIZE;
+ return 0;
+}
+DEFINE_DEBUGFS_ATTRIBUTE(total_size_fops, debugfs_get_total_size, NULL, "%llu\n");
+
static int zswap_debugfs_init(void)
{
if (!debugfs_initialized())
@@ -1743,8 +1742,8 @@ static int zswap_debugfs_init(void)
zswap_debugfs_root, &zswap_reject_compress_poor);
debugfs_create_u64("written_back_pages", 0444,
zswap_debugfs_root, &zswap_written_back_pages);
- debugfs_create_u64("pool_total_size", 0444,
- zswap_debugfs_root, &zswap_pool_total_size);
+ debugfs_create_file("pool_total_size", 0444,
+ zswap_debugfs_root, NULL, &total_size_fops);
debugfs_create_atomic_t("stored_pages", 0444,
zswap_debugfs_root, &zswap_stored_pages);
debugfs_create_atomic_t("same_filled_pages", 0444,
_
Patches currently in -mm which might be from hannes@xxxxxxxxxxx are
mm-cachestat-fix-two-shmem-bugs.patch
mm-zswap-optimize-zswap-pool-size-tracking.patch
mm-zpool-return-pool-size-in-pages.patch
