Linux NFS development
 help / color / mirror / Atom feed
From: Chuck Lever <cel@kernel.org>
To: NeilBrown <neil@brown.name>, Jeff Layton <jlayton@kernel.org>,
	Olga Kornievskaia <okorniev@redhat.com>,
	Dai Ngo <dai.ngo@oracle.com>, Tom Talpey <tom@talpey.com>
Cc: <linux-nfs@vger.kernel.org>
Subject: [PATCH 2/5] NFSD: Eliminate percpu counter contention in DRC memory accounting
Date: Thu, 16 Jul 2026 20:12:29 -0400	[thread overview]
Message-ID: <20260717001232.438792-3-cel@kernel.org> (raw)
In-Reply-To: <20260717001232.438792-1-cel@kernel.org>

The DRC memory usage counter (NFSD_STATS_DRC_MEM_USAGE) tracks
bytes, but percpu_counter_add() uses the global percpu_counter_batch
threshold of max(32, 2*nr_cpus). Each DRC entry add or removal
updates the counter by sizeof(struct nfsd_cacherep) (~144 bytes),
which always exceeds the batch threshold. percpu_counter_add()
then acquires the counter's global spinlock on every update,
serializing all nfsd threads.

On a 10-CPU NFS server handling a high rate of non-idempotent
NFSv3 operations, this lock accounts for a measurable fraction
of total spin lock overhead because nfsd_cache_lookup() both
inserts a new entry and prunes up to three old entries per RPC,
producing 4-7 global lock acquisitions per operation.

Switch to percpu_counter_add_local() and percpu_counter_sub_local(),
which batch with INT_MAX so that updates always remain on the per-CPU
fast path regardless of the amount. The only reader of this counter uses
percpu_counter_sum_positive(), which sums the per-CPU deltas under the
global lock, so read accuracy is unaffected.

Signed-off-by: Chuck Lever <cel@kernel.org>
---
 fs/nfsd/stats.h | 22 ++++++++++++++++++++--
 1 file changed, 20 insertions(+), 2 deletions(-)

diff --git a/fs/nfsd/stats.h b/fs/nfsd/stats.h
index 87736b7fbf28..15d30c045dc3 100644
--- a/fs/nfsd/stats.h
+++ b/fs/nfsd/stats.h
@@ -60,14 +60,32 @@ static inline void nfsd_stats_payload_misses_inc(struct nfsd_net *nn)
 	percpu_counter_inc(&nn->counter[NFSD_STATS_PAYLOAD_MISSES]);
 }
 
+/**
+ * nfsd_stats_drc_mem_usage_add - Add memory used by a cache item
+ * @nn: target network namespace
+ * @amount: byte count
+ *
+ * percpu_counter_add_local() keeps updates on the per-CPU fast
+ * path. The sole reader, percpu_counter_sum_positive(), sums the
+ * per-CPU deltas, so batching locally does not lose accuracy.
+ */
 static inline void nfsd_stats_drc_mem_usage_add(struct nfsd_net *nn, s64 amount)
 {
-	percpu_counter_add(&nn->counter[NFSD_STATS_DRC_MEM_USAGE], amount);
+	percpu_counter_add_local(&nn->counter[NFSD_STATS_DRC_MEM_USAGE],
+				 amount);
 }
 
+/**
+ * nfsd_stats_drc_mem_usage_sub - Subtract memory used by a cache item
+ * @nn: target network namespace
+ * @amount: byte count
+ *
+ * See nfsd_stats_drc_mem_usage_add() for batching rationale.
+ */
 static inline void nfsd_stats_drc_mem_usage_sub(struct nfsd_net *nn, s64 amount)
 {
-	percpu_counter_sub(&nn->counter[NFSD_STATS_DRC_MEM_USAGE], amount);
+	percpu_counter_sub_local(&nn->counter[NFSD_STATS_DRC_MEM_USAGE],
+				 amount);
 }
 
 #ifdef CONFIG_NFSD_V4
-- 
2.54.0


  parent reply	other threads:[~2026-07-17  0:12 UTC|newest]

Thread overview: 7+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2026-07-17  0:12 [PATCH 0/5] Minor NFSD global DRC clean-ups Chuck Lever
2026-07-17  0:12 ` [PATCH 1/5] NFSD: Fix off-by-one in DRC bucket pruning limit Chuck Lever
2026-07-17  0:12 ` Chuck Lever [this message]
2026-07-17  0:12 ` [PATCH 3/5] NFSD: Eliminate percpu counter contention in reply cache statistics Chuck Lever
2026-07-17  0:12 ` [PATCH 4/5] NFSD: Eliminate percpu counter contention in IO byte accounting Chuck Lever
2026-07-17  0:12 ` [PATCH 5/5] NFSD: Document reply_cache_stats ABI Chuck Lever
2026-07-17  4:44 ` [PATCH 0/5] Minor NFSD global DRC clean-ups NeilBrown

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=20260717001232.438792-3-cel@kernel.org \
    --to=cel@kernel.org \
    --cc=dai.ngo@oracle.com \
    --cc=jlayton@kernel.org \
    --cc=linux-nfs@vger.kernel.org \
    --cc=neil@brown.name \
    --cc=okorniev@redhat.com \
    --cc=tom@talpey.com \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox