Linux-ext4 Archive on lore.kernel.org
 help / color / Atom feed
* [PATCH v2] ext4: use percpu_counters for extent_status cache hits/misses
@ 2019-08-28  9:19 Shaokun Zhang
  2019-08-28 15:20 ` Theodore Y. Ts'o
  0 siblings, 1 reply; 2+ messages in thread
From: Shaokun Zhang @ 2019-08-28  9:19 UTC (permalink / raw)
  To: linux-ext4, linux-kernel
  Cc: Yang Guo, Theodore Ts'o, Andreas Dilger, Eric Biggers, Shaokun Zhang

From: Yang Guo <guoyang2@huawei.com>

@es_stats_cache_hits and @es_stats_cache_misses are accessed frequently in
ext4_es_lookup_extent function, it would influence the ext4 read/write
performance in NUMA system. Let's optimize it using percpu_counter,
it is profitable for the performance.

The test command is as below:
fio -name=randwrite -numjobs=8 -filename=/mnt/test1 -rw=randwrite
-ioengine=libaio -direct=1 -iodepth=64 -sync=0 -norandommap
-group_reporting -runtime=120 -time_based -bs=4k -size=5G

And the result is better 10% than the initial implement:
without the patch,IOPS=197k, BW=770MiB/s (808MB/s)(90.3GiB/120002msec)
with the patch,  IOPS=218k, BW=852MiB/s (894MB/s)(99.9GiB/120002msec)

Cc: "Theodore Ts'o" <tytso@mit.edu>
Cc: Andreas Dilger <adilger.kernel@dilger.ca>
Cc: Eric Biggers <ebiggers@kernel.org>
Signed-off-by: Yang Guo <guoyang2@huawei.com>
Signed-off-by: Shaokun Zhang <zhangshaokun@hisilicon.com>
---
ChangeLog:
    Fix the issue that there is no percpu_counter_destroy() for the new
 percpu counters.

 fs/ext4/extents_status.c | 37 ++++++++++++++++++++++++-------------
 fs/ext4/extents_status.h |  4 ++--
 2 files changed, 26 insertions(+), 15 deletions(-)

diff --git a/fs/ext4/extents_status.c b/fs/ext4/extents_status.c
index 7521de2dcf3a..3c03062a8d6a 100644
--- a/fs/ext4/extents_status.c
+++ b/fs/ext4/extents_status.c
@@ -947,9 +947,9 @@ int ext4_es_lookup_extent(struct inode *inode, ext4_lblk_t lblk,
 		es->es_pblk = es1->es_pblk;
 		if (!ext4_es_is_referenced(es1))
 			ext4_es_set_referenced(es1);
-		stats->es_stats_cache_hits++;
+		percpu_counter_inc(&stats->es_stats_cache_hits);
 	} else {
-		stats->es_stats_cache_misses++;
+		percpu_counter_inc(&stats->es_stats_cache_misses);
 	}
 
 	read_unlock(&EXT4_I(inode)->i_es_lock);
@@ -1235,9 +1235,9 @@ int ext4_seq_es_shrinker_info_show(struct seq_file *seq, void *v)
 	seq_printf(seq, "stats:\n  %lld objects\n  %lld reclaimable objects\n",
 		   percpu_counter_sum_positive(&es_stats->es_stats_all_cnt),
 		   percpu_counter_sum_positive(&es_stats->es_stats_shk_cnt));
-	seq_printf(seq, "  %lu/%lu cache hits/misses\n",
-		   es_stats->es_stats_cache_hits,
-		   es_stats->es_stats_cache_misses);
+	seq_printf(seq, "  %lld/%lld cache hits/misses\n",
+		   percpu_counter_sum_positive(&es_stats->es_stats_cache_hits),
+		   percpu_counter_sum_positive(&es_stats->es_stats_cache_misses));
 	if (inode_cnt)
 		seq_printf(seq, "  %d inodes on list\n", inode_cnt);
 
@@ -1264,35 +1264,46 @@ int ext4_es_register_shrinker(struct ext4_sb_info *sbi)
 	sbi->s_es_nr_inode = 0;
 	spin_lock_init(&sbi->s_es_lock);
 	sbi->s_es_stats.es_stats_shrunk = 0;
-	sbi->s_es_stats.es_stats_cache_hits = 0;
-	sbi->s_es_stats.es_stats_cache_misses = 0;
+	err = percpu_counter_init(&sbi->s_es_stats.es_stats_cache_hits, 0,
+				  GFP_KERNEL);
+	if (err)
+		return err;
+	err = percpu_counter_init(&sbi->s_es_stats.es_stats_cache_misses, 0,
+				  GFP_KERNEL);
+	if (err)
+		goto err1;
 	sbi->s_es_stats.es_stats_scan_time = 0;
 	sbi->s_es_stats.es_stats_max_scan_time = 0;
 	err = percpu_counter_init(&sbi->s_es_stats.es_stats_all_cnt, 0, GFP_KERNEL);
 	if (err)
-		return err;
+		goto err2;
 	err = percpu_counter_init(&sbi->s_es_stats.es_stats_shk_cnt, 0, GFP_KERNEL);
 	if (err)
-		goto err1;
+		goto err3;
 
 	sbi->s_es_shrinker.scan_objects = ext4_es_scan;
 	sbi->s_es_shrinker.count_objects = ext4_es_count;
 	sbi->s_es_shrinker.seeks = DEFAULT_SEEKS;
 	err = register_shrinker(&sbi->s_es_shrinker);
 	if (err)
-		goto err2;
+		goto err4;
 
 	return 0;
-
-err2:
+err4:
 	percpu_counter_destroy(&sbi->s_es_stats.es_stats_shk_cnt);
-err1:
+err3:
 	percpu_counter_destroy(&sbi->s_es_stats.es_stats_all_cnt);
+err2:
+	percpu_counter_destroy(&sbi->s_es_stats.es_stats_cache_misses);
+err1:
+	percpu_counter_destroy(&sbi->s_es_stats.es_stats_cache_hits);
 	return err;
 }
 
 void ext4_es_unregister_shrinker(struct ext4_sb_info *sbi)
 {
+	percpu_counter_destroy(&sbi->s_es_stats.es_stats_cache_hits);
+	percpu_counter_destroy(&sbi->s_es_stats.es_stats_cache_misses);
 	percpu_counter_destroy(&sbi->s_es_stats.es_stats_all_cnt);
 	percpu_counter_destroy(&sbi->s_es_stats.es_stats_shk_cnt);
 	unregister_shrinker(&sbi->s_es_shrinker);
diff --git a/fs/ext4/extents_status.h b/fs/ext4/extents_status.h
index 131a8b7df265..e722dd9bd06e 100644
--- a/fs/ext4/extents_status.h
+++ b/fs/ext4/extents_status.h
@@ -70,8 +70,8 @@ struct ext4_es_tree {
 
 struct ext4_es_stats {
 	unsigned long es_stats_shrunk;
-	unsigned long es_stats_cache_hits;
-	unsigned long es_stats_cache_misses;
+	struct percpu_counter es_stats_cache_hits;
+	struct percpu_counter es_stats_cache_misses;
 	u64 es_stats_scan_time;
 	u64 es_stats_max_scan_time;
 	struct percpu_counter es_stats_all_cnt;
-- 
2.7.4


^ permalink raw reply	[flat|nested] 2+ messages in thread

* Re: [PATCH v2] ext4: use percpu_counters for extent_status cache hits/misses
  2019-08-28  9:19 [PATCH v2] ext4: use percpu_counters for extent_status cache hits/misses Shaokun Zhang
@ 2019-08-28 15:20 ` Theodore Y. Ts'o
  0 siblings, 0 replies; 2+ messages in thread
From: Theodore Y. Ts'o @ 2019-08-28 15:20 UTC (permalink / raw)
  To: Shaokun Zhang
  Cc: linux-ext4, linux-kernel, Yang Guo, Andreas Dilger, Eric Biggers

On Wed, Aug 28, 2019 at 05:19:17PM +0800, Shaokun Zhang wrote:
> From: Yang Guo <guoyang2@huawei.com>
> 
> @es_stats_cache_hits and @es_stats_cache_misses are accessed frequently in
> ext4_es_lookup_extent function, it would influence the ext4 read/write
> performance in NUMA system. Let's optimize it using percpu_counter,
> it is profitable for the performance.
> 
> The test command is as below:
> fio -name=randwrite -numjobs=8 -filename=/mnt/test1 -rw=randwrite
> -ioengine=libaio -direct=1 -iodepth=64 -sync=0 -norandommap
> -group_reporting -runtime=120 -time_based -bs=4k -size=5G
> 
> And the result is better 10% than the initial implement:
> without the patch,IOPS=197k, BW=770MiB/s (808MB/s)(90.3GiB/120002msec)
> with the patch,  IOPS=218k, BW=852MiB/s (894MB/s)(99.9GiB/120002msec)
> 
> Cc: "Theodore Ts'o" <tytso@mit.edu>
> Cc: Andreas Dilger <adilger.kernel@dilger.ca>
> Cc: Eric Biggers <ebiggers@kernel.org>
> Signed-off-by: Yang Guo <guoyang2@huawei.com>
> Signed-off-by: Shaokun Zhang <zhangshaokun@hisilicon.com>

Thanks, applied.

						- Ted

^ permalink raw reply	[flat|nested] 2+ messages in thread

end of thread, back to index

Thread overview: 2+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2019-08-28  9:19 [PATCH v2] ext4: use percpu_counters for extent_status cache hits/misses Shaokun Zhang
2019-08-28 15:20 ` Theodore Y. Ts'o

Linux-ext4 Archive on lore.kernel.org

Archives are clonable:
	git clone --mirror https://lore.kernel.org/linux-ext4/0 linux-ext4/git/0.git

	# If you have public-inbox 1.1+ installed, you may
	# initialize and index your mirror using the following commands:
	public-inbox-init -V2 linux-ext4 linux-ext4/ https://lore.kernel.org/linux-ext4 \
		linux-ext4@vger.kernel.org linux-ext4@archiver.kernel.org
	public-inbox-index linux-ext4


Newsgroup available over NNTP:
	nntp://nntp.lore.kernel.org/org.kernel.vger.linux-ext4


AGPL code for this site: git clone https://public-inbox.org/ public-inbox