From: Eric Dumazet <edumazet@google.com>
Date: Tue, 8 Apr 2025 20:27:42 +0000 (+0000)
Subject: net: remove cpu stall in txq_trans_update()
X-Git-Tag: v6.16-rc1~132^2~348
X-Git-Url: http://git.ipfire.org/?a=commitdiff_plain;h=229671ac60e298b85c2644f52d7e487e9f487d06;p=thirdparty%2Fkernel%2Flinux.git

net: remove cpu stall in txq_trans_update()

txq_trans_update() currently uses txq->xmit_lock_owner
to conditionally update txq->trans_start.

For regular devices, txq->xmit_lock_owner is updated
from HARD_TX_LOCK() and HARD_TX_UNLOCK(), and this apparently
causes cpu stalls.

Using dev->lltx, which sits in a read-mostly cache-line,
and already used in HARD_TX_LOCK() and HARD_TX_UNLOCK()
helps cpu prediction.

On an AMD EPYC 7B12 dual socket server, tcp_rr with 128 threads
and 30,000 flows gets a 5 % increase in throughput.

As explained in commit 95ecba62e2fd ("net: fix races in
netdev_tx_sent_queue()/dev_watchdog()") I am planning
to no longer update txq->trans_start in the fast path
in a followup patch.

Signed-off-by: Eric Dumazet <edumazet@google.com>
Link: https://patch.msgid.link/20250408202742.2145516-1-edumazet@google.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
---

diff --git a/drivers/net/ethernet/ti/am65-cpsw-nuss.c b/drivers/net/ethernet/ti/am65-cpsw-nuss.c
index c9fd34787c998..e78de79a5d78c 100644
--- a/drivers/net/ethernet/ti/am65-cpsw-nuss.c
+++ b/drivers/net/ethernet/ti/am65-cpsw-nuss.c
@@ -427,7 +427,7 @@ static void am65_cpsw_nuss_ndo_host_tx_timeout(struct net_device *ndev,
 
 	if (netif_tx_queue_stopped(netif_txq)) {
 		/* try recover if stopped by us */
-		txq_trans_update(netif_txq);
+		txq_trans_update(ndev, netif_txq);
 		netif_tx_wake_queue(netif_txq);
 	}
 }
diff --git a/include/linux/netdevice.h b/include/linux/netdevice.h
index dece2ae396a1e..a28a08046615f 100644
--- a/include/linux/netdevice.h
+++ b/include/linux/netdevice.h
@@ -4693,9 +4693,10 @@ static inline void __netif_tx_unlock_bh(struct netdev_queue *txq)
 /*
  * txq->trans_start can be read locklessly from dev_watchdog()
  */
-static inline void txq_trans_update(struct netdev_queue *txq)
+static inline void txq_trans_update(const struct net_device *dev,
+				    struct netdev_queue *txq)
 {
-	if (txq->xmit_lock_owner != -1)
+	if (!dev->lltx)
 		WRITE_ONCE(txq->trans_start, jiffies);
 }
 
@@ -5214,7 +5215,7 @@ static inline netdev_tx_t netdev_start_xmit(struct sk_buff *skb, struct net_devi
 
 	rc = __netdev_start_xmit(ops, skb, dev, more);
 	if (rc == NETDEV_TX_OK)
-		txq_trans_update(txq);
+		txq_trans_update(dev, txq);
 
 	return rc;
 }