From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: X-Spam-Checker-Version: SpamAssassin 3.4.0 (2014-02-07) on aws-us-west-2-korg-lkml-1.web.codeaurora.org Received: from vger.kernel.org (vger.kernel.org [23.128.96.18]) by smtp.lore.kernel.org (Postfix) with ESMTP id AEF44C4167B for ; Sun, 18 Dec 2022 16:49:33 +0000 (UTC) Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S232302AbiLRQta (ORCPT ); Sun, 18 Dec 2022 11:49:30 -0500 Received: from lindbergh.monkeyblade.net ([23.128.96.19]:52044 "EHLO lindbergh.monkeyblade.net" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S232449AbiLRQr1 (ORCPT ); Sun, 18 Dec 2022 11:47:27 -0500 Received: from ams.source.kernel.org (ams.source.kernel.org [IPv6:2604:1380:4601:e00::1]) by lindbergh.monkeyblade.net (Postfix) with ESMTPS id 7D5F511A35; Sun, 18 Dec 2022 08:17:43 -0800 (PST) Received: from smtp.kernel.org (relay.kernel.org [52.25.139.140]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by ams.source.kernel.org (Postfix) with ESMTPS id 25ACEB80BA7; Sun, 18 Dec 2022 16:17:42 +0000 (UTC) Received: by smtp.kernel.org (Postfix) with ESMTPSA id ABB58C433A8; Sun, 18 Dec 2022 16:17:39 +0000 (UTC) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/simple; d=kernel.org; s=k20201202; t=1671380260; bh=TXTvjtfPLfcHfjhTIu0kSJEnGFEchVY+QHMRRyjNRbg=; h=From:To:Cc:Subject:Date:In-Reply-To:References:From; b=EBnIJ2m6GM+EVltBnkG3jLmqw12L+M+fXKFokO7ahdFUOS2EHn4JKC/LTDF9/vYrG 7OKgx7ItRtefYgJ7m/XK7fJ4yL/1HIo+E74fVPxTxdGfDe9OeXd2Kwzu6WBsFwtbH2 1164WOD/dNwSX9RC+6YF8IpddwWnmTjSpWKKrLSEBhcZFfB1HGlDFBLjcvWjb6iHX9 Qq1JXK3TpNsSvbY6owtQF9QAukAoR+ChZn+5dLbEl1thhA45XA+lo6wvR1IA0pkOo4 yYci5xOy+/aO0BvtZ9JOIbGQ5zNg+o4bCjzz4AsW9L5YlqSXe4aInYv+9oeTRu5Yvj Hhj+9Dyk2gyzQ== From: Sasha Levin To: linux-kernel@vger.kernel.org, stable@vger.kernel.org Cc: Eric Dumazet , "David S . Miller" , Sasha Levin , kuba@kernel.org, pabeni@redhat.com, imagedong@tencent.com, bigeasy@linutronix.de, kuniyu@amazon.com, petrm@nvidia.com, netdev@vger.kernel.org Subject: [PATCH AUTOSEL 5.10 25/39] net: add atomic_long_t to net_device_stats fields Date: Sun, 18 Dec 2022 11:15:45 -0500 Message-Id: <20221218161559.932604-25-sashal@kernel.org> X-Mailer: git-send-email 2.35.1 In-Reply-To: <20221218161559.932604-1-sashal@kernel.org> References: <20221218161559.932604-1-sashal@kernel.org> MIME-Version: 1.0 X-stable: review X-Patchwork-Hint: Ignore Content-Transfer-Encoding: 8bit Precedence: bulk List-ID: X-Mailing-List: netdev@vger.kernel.org From: Eric Dumazet [ Upstream commit 6c1c5097781f563b70a81683ea6fdac21637573b ] Long standing KCSAN issues are caused by data-race around some dev->stats changes. Most performance critical paths already use per-cpu variables, or per-queue ones. It is reasonable (and more correct) to use atomic operations for the slow paths. This patch adds an union for each field of net_device_stats, so that we can convert paths that are not yet protected by a spinlock or a mutex. netdev_stats_to_stats64() no longer has an #if BITS_PER_LONG==64 Note that the memcpy() we were using on 64bit arches had no provision to avoid load-tearing, while atomic_long_read() is providing the needed protection at no cost. Signed-off-by: Eric Dumazet Signed-off-by: David S. Miller Signed-off-by: Sasha Levin --- include/linux/netdevice.h | 58 +++++++++++++++++++++++---------------- include/net/dst.h | 5 ++-- net/core/dev.c | 14 ++-------- 3 files changed, 40 insertions(+), 37 deletions(-) diff --git a/include/linux/netdevice.h b/include/linux/netdevice.h index ef75567efd27..b478a16ef284 100644 --- a/include/linux/netdevice.h +++ b/include/linux/netdevice.h @@ -166,31 +166,38 @@ static inline bool dev_xmit_complete(int rc) * (unsigned long) so they can be read and written atomically. */ +#define NET_DEV_STAT(FIELD) \ + union { \ + unsigned long FIELD; \ + atomic_long_t __##FIELD; \ + } + struct net_device_stats { - unsigned long rx_packets; - unsigned long tx_packets; - unsigned long rx_bytes; - unsigned long tx_bytes; - unsigned long rx_errors; - unsigned long tx_errors; - unsigned long rx_dropped; - unsigned long tx_dropped; - unsigned long multicast; - unsigned long collisions; - unsigned long rx_length_errors; - unsigned long rx_over_errors; - unsigned long rx_crc_errors; - unsigned long rx_frame_errors; - unsigned long rx_fifo_errors; - unsigned long rx_missed_errors; - unsigned long tx_aborted_errors; - unsigned long tx_carrier_errors; - unsigned long tx_fifo_errors; - unsigned long tx_heartbeat_errors; - unsigned long tx_window_errors; - unsigned long rx_compressed; - unsigned long tx_compressed; + NET_DEV_STAT(rx_packets); + NET_DEV_STAT(tx_packets); + NET_DEV_STAT(rx_bytes); + NET_DEV_STAT(tx_bytes); + NET_DEV_STAT(rx_errors); + NET_DEV_STAT(tx_errors); + NET_DEV_STAT(rx_dropped); + NET_DEV_STAT(tx_dropped); + NET_DEV_STAT(multicast); + NET_DEV_STAT(collisions); + NET_DEV_STAT(rx_length_errors); + NET_DEV_STAT(rx_over_errors); + NET_DEV_STAT(rx_crc_errors); + NET_DEV_STAT(rx_frame_errors); + NET_DEV_STAT(rx_fifo_errors); + NET_DEV_STAT(rx_missed_errors); + NET_DEV_STAT(tx_aborted_errors); + NET_DEV_STAT(tx_carrier_errors); + NET_DEV_STAT(tx_fifo_errors); + NET_DEV_STAT(tx_heartbeat_errors); + NET_DEV_STAT(tx_window_errors); + NET_DEV_STAT(rx_compressed); + NET_DEV_STAT(tx_compressed); }; +#undef NET_DEV_STAT #include @@ -5256,4 +5263,9 @@ do { \ extern struct net_device *blackhole_netdev; +/* Note: Avoid these macros in fast path, prefer per-cpu or per-queue counters. */ +#define DEV_STATS_INC(DEV, FIELD) atomic_long_inc(&(DEV)->stats.__##FIELD) +#define DEV_STATS_ADD(DEV, FIELD, VAL) \ + atomic_long_add((VAL), &(DEV)->stats.__##FIELD) + #endif /* _LINUX_NETDEVICE_H */ diff --git a/include/net/dst.h b/include/net/dst.h index acd15c544cf3..ae2cf57d796b 100644 --- a/include/net/dst.h +++ b/include/net/dst.h @@ -356,9 +356,8 @@ static inline void __skb_tunnel_rx(struct sk_buff *skb, struct net_device *dev, static inline void skb_tunnel_rx(struct sk_buff *skb, struct net_device *dev, struct net *net) { - /* TODO : stats should be SMP safe */ - dev->stats.rx_packets++; - dev->stats.rx_bytes += skb->len; + DEV_STATS_INC(dev, rx_packets); + DEV_STATS_ADD(dev, rx_bytes, skb->len); __skb_tunnel_rx(skb, dev, net); } diff --git a/net/core/dev.c b/net/core/dev.c index 34b5aab42b91..5afe56e8e763 100644 --- a/net/core/dev.c +++ b/net/core/dev.c @@ -10320,24 +10320,16 @@ void netdev_run_todo(void) void netdev_stats_to_stats64(struct rtnl_link_stats64 *stats64, const struct net_device_stats *netdev_stats) { -#if BITS_PER_LONG == 64 - BUILD_BUG_ON(sizeof(*stats64) < sizeof(*netdev_stats)); - memcpy(stats64, netdev_stats, sizeof(*netdev_stats)); - /* zero out counters that only exist in rtnl_link_stats64 */ - memset((char *)stats64 + sizeof(*netdev_stats), 0, - sizeof(*stats64) - sizeof(*netdev_stats)); -#else - size_t i, n = sizeof(*netdev_stats) / sizeof(unsigned long); - const unsigned long *src = (const unsigned long *)netdev_stats; + size_t i, n = sizeof(*netdev_stats) / sizeof(atomic_long_t); + const atomic_long_t *src = (atomic_long_t *)netdev_stats; u64 *dst = (u64 *)stats64; BUILD_BUG_ON(n > sizeof(*stats64) / sizeof(u64)); for (i = 0; i < n; i++) - dst[i] = src[i]; + dst[i] = atomic_long_read(&src[i]); /* zero out counters that only exist in rtnl_link_stats64 */ memset((char *)stats64 + n * sizeof(u64), 0, sizeof(*stats64) - n * sizeof(u64)); -#endif } EXPORT_SYMBOL(netdev_stats_to_stats64); -- 2.35.1