DPDK patches and discussions
 help / color / mirror / Atom feed
From: Stephen Hemminger <stephen@networkplumber.org>
To: dev@dpdk.org
Cc: Stephen Hemminger <stephen@networkplumber.org>,
	Reshma Pattan <reshma.pattan@intel.com>
Subject: [PATCH 2/2] latencystats: optimize locking on transmit
Date: Thu, 12 Jun 2025 17:34:26 -0700	[thread overview]
Message-ID: <20250613003547.39239-3-stephen@networkplumber.org> (raw)
In-Reply-To: <20250613003547.39239-1-stephen@networkplumber.org>

If transmit callback is called, and there are no packets
in the burst with timestamp set, then the expensive operations
of locking and accessing TSC can be skipped.

Signed-off-by: Stephen Hemminger <stephen@networkplumber.org>
---
 lib/latencystats/rte_latencystats.c | 17 +++++++++++++----
 1 file changed, 13 insertions(+), 4 deletions(-)

diff --git a/lib/latencystats/rte_latencystats.c b/lib/latencystats/rte_latencystats.c
index 181c53dd0e..dc415c4682 100644
--- a/lib/latencystats/rte_latencystats.c
+++ b/lib/latencystats/rte_latencystats.c
@@ -190,17 +190,26 @@ calc_latency(uint16_t pid __rte_unused,
 		void *_ __rte_unused)
 {
 	unsigned int i;
-	uint64_t now, latency;
+	uint64_t ts_flags = 0;
 	static uint64_t prev_latency;
 
-	now = rte_rdtsc();
+	for (i = 0; i < nb_pkts; i++)
+		ts_flags |= (pkts[i]->ol_flags & timestamp_dynflag);
 
+	/* no samples in this burst, skip locking */
+	if (likely(ts_flags == 0))
+		return nb_pkts;
+
+	uint64_t now = rte_rdtsc();
 	rte_spinlock_lock(&glob_stats->lock);
 	for (i = 0; i < nb_pkts; i++) {
-		if (!(pkts[i]->ol_flags & timestamp_dynflag))
+		struct rte_mbuf *m = pkts[i];
+		uint64_t latency;
+
+		if (!(m->ol_flags & timestamp_dynflag))
 			continue;
 
-		latency = now - *timestamp_dynfield(pkts[i]);
+		latency = now - *timestamp_dynfield(m);
 
 		if (glob_stats->samples++ == 0) {
 			glob_stats->min_latency = latency;
-- 
2.47.2


  parent reply	other threads:[~2025-06-13  0:36 UTC|newest]

Thread overview: 4+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2025-06-13  0:34 [PATCH 0/2] Latencystat optimization and fix Stephen Hemminger
2025-06-13  0:34 ` [PATCH 1/2] latencystats: fix receive sample MP issues Stephen Hemminger
2025-06-13  0:34 ` Stephen Hemminger [this message]
2025-06-13 11:13 ` [PATCH 0/2] Latencystat optimization and fix Varghese, Vipin

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=20250613003547.39239-3-stephen@networkplumber.org \
    --to=stephen@networkplumber.org \
    --cc=dev@dpdk.org \
    --cc=reshma.pattan@intel.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;
as well as URLs for NNTP newsgroup(s).