stable.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
To: stable@vger.kernel.org
Cc: Greg Kroah-Hartman <gregkh@linuxfoundation.org>,
	patches@lists.linux.dev, Eric Dumazet <edumazet@google.com>,
	Simon Horman <simon.horman@corigine.com>,
	Kuniyuki Iwashima <kuniyu@amazon.com>,
	"David S. Miller" <davem@davemloft.net>,
	Sasha Levin <sashal@kernel.org>
Subject: [PATCH 4.19 07/23] rfs: annotate lockless accesses to sk->sk_rxhash
Date: Mon, 12 Jun 2023 12:26:08 +0200	[thread overview]
Message-ID: <20230612101651.407477221@linuxfoundation.org> (raw)
In-Reply-To: <20230612101651.138592130@linuxfoundation.org>

From: Eric Dumazet <edumazet@google.com>

[ Upstream commit 1e5c647c3f6d4f8497dedcd226204e1880e0ffb3 ]

Add READ_ONCE()/WRITE_ONCE() on accesses to sk->sk_rxhash.

This also prevents a (smart ?) compiler to remove the condition in:

if (sk->sk_rxhash != newval)
	sk->sk_rxhash = newval;

We need the condition to avoid dirtying a shared cache line.

Fixes: fec5e652e58f ("rfs: Receive Flow Steering")
Signed-off-by: Eric Dumazet <edumazet@google.com>
Reviewed-by: Simon Horman <simon.horman@corigine.com>
Reviewed-by: Kuniyuki Iwashima <kuniyu@amazon.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Sasha Levin <sashal@kernel.org>
---
 include/net/sock.h | 18 +++++++++++++-----
 1 file changed, 13 insertions(+), 5 deletions(-)

diff --git a/include/net/sock.h b/include/net/sock.h
index c140c6f86e4b1..616e84d1670df 100644
--- a/include/net/sock.h
+++ b/include/net/sock.h
@@ -988,8 +988,12 @@ static inline void sock_rps_record_flow(const struct sock *sk)
 		 * OR	an additional socket flag
 		 * [1] : sk_state and sk_prot are in the same cache line.
 		 */
-		if (sk->sk_state == TCP_ESTABLISHED)
-			sock_rps_record_flow_hash(sk->sk_rxhash);
+		if (sk->sk_state == TCP_ESTABLISHED) {
+			/* This READ_ONCE() is paired with the WRITE_ONCE()
+			 * from sock_rps_save_rxhash() and sock_rps_reset_rxhash().
+			 */
+			sock_rps_record_flow_hash(READ_ONCE(sk->sk_rxhash));
+		}
 	}
 #endif
 }
@@ -998,15 +1002,19 @@ static inline void sock_rps_save_rxhash(struct sock *sk,
 					const struct sk_buff *skb)
 {
 #ifdef CONFIG_RPS
-	if (unlikely(sk->sk_rxhash != skb->hash))
-		sk->sk_rxhash = skb->hash;
+	/* The following WRITE_ONCE() is paired with the READ_ONCE()
+	 * here, and another one in sock_rps_record_flow().
+	 */
+	if (unlikely(READ_ONCE(sk->sk_rxhash) != skb->hash))
+		WRITE_ONCE(sk->sk_rxhash, skb->hash);
 #endif
 }
 
 static inline void sock_rps_reset_rxhash(struct sock *sk)
 {
 #ifdef CONFIG_RPS
-	sk->sk_rxhash = 0;
+	/* Paired with READ_ONCE() in sock_rps_record_flow() */
+	WRITE_ONCE(sk->sk_rxhash, 0);
 #endif
 }
 
-- 
2.39.2




  parent reply	other threads:[~2023-06-12 10:46 UTC|newest]

Thread overview: 33+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2023-06-12 10:26 [PATCH 4.19 00/23] 4.19.286-rc1 review Greg Kroah-Hartman
2023-06-12 10:26 ` [PATCH 4.19 01/23] i40iw: fix build warning in i40iw_manage_apbvt() Greg Kroah-Hartman
2023-06-12 10:26 ` [PATCH 4.19 02/23] i40e: fix build warnings in i40e_alloc.h Greg Kroah-Hartman
2023-06-14  2:05   ` Ben Hutchings
2023-06-14  8:56     ` Greg Kroah-Hartman
2023-06-12 10:26 ` [PATCH 4.19 03/23] spi: qup: Request DMA before enabling clocks Greg Kroah-Hartman
2023-06-12 10:26 ` [PATCH 4.19 04/23] net: dsa: lan9303: allow vid != 0 in port_fdb_{add|del} methods Greg Kroah-Hartman
2023-06-12 10:26 ` [PATCH 4.19 05/23] Bluetooth: Fix l2cap_disconnect_req deadlock Greg Kroah-Hartman
2023-06-12 10:26 ` [PATCH 4.19 06/23] Bluetooth: L2CAP: Add missing checks for invalid DCID Greg Kroah-Hartman
2023-06-12 10:26 ` Greg Kroah-Hartman [this message]
2023-06-12 10:26 ` [PATCH 4.19 08/23] rfs: annotate lockless accesses to RFS sock flow table Greg Kroah-Hartman
2023-06-12 10:26 ` [PATCH 4.19 09/23] net: sched: move rtm_tca_policy declaration to include file Greg Kroah-Hartman
2023-06-12 10:26 ` [PATCH 4.19 10/23] net: sched: fix possible refcount leak in tc_chain_tmplt_add() Greg Kroah-Hartman
2023-06-12 10:26 ` [PATCH 4.19 11/23] lib: cpu_rmap: Fix potential use-after-free in irq_cpu_rmap_release() Greg Kroah-Hartman
2023-06-12 10:26 ` [PATCH 4.19 12/23] batman-adv: Broken sync while rescheduling delayed work Greg Kroah-Hartman
2023-06-12 10:26 ` [PATCH 4.19 13/23] Input: xpad - delete a Razer DeathAdder mouse VID/PID entry Greg Kroah-Hartman
2023-06-12 10:26 ` [PATCH 4.19 14/23] Input: psmouse - fix OOB access in Elantech protocol Greg Kroah-Hartman
2023-06-12 10:26 ` [PATCH 4.19 15/23] drm/amdgpu: fix xclk freq on CHIP_STONEY Greg Kroah-Hartman
2023-06-12 10:26 ` [PATCH 4.19 16/23] ceph: fix use-after-free bug for inodes when flushing capsnaps Greg Kroah-Hartman
2023-06-12 10:26 ` [PATCH 4.19 17/23] Bluetooth: Fix use-after-free in hci_remove_ltk/hci_remove_irk Greg Kroah-Hartman
2023-06-12 10:26 ` [PATCH 4.19 18/23] pinctrl: meson-axg: add missing GPIOA_18 gpio group Greg Kroah-Hartman
2023-06-12 10:26 ` [PATCH 4.19 19/23] i2c: sprd: Delete i2c adapter in .removes error path Greg Kroah-Hartman
2023-06-12 10:26 ` [PATCH 4.19 20/23] ext4: only check dquot_initialize_needed() when debugging Greg Kroah-Hartman
2023-06-12 10:26 ` [PATCH 4.19 21/23] btrfs: check return value of btrfs_commit_transaction in relocation Greg Kroah-Hartman
2023-06-14  2:44   ` Ben Hutchings
2023-06-12 10:26 ` [PATCH 4.19 22/23] btrfs: unset reloc control if transaction commit fails in prepare_to_relocate() Greg Kroah-Hartman
2023-06-12 10:26 ` [PATCH 4.19 23/23] Revert "staging: rtl8192e: Replace macro RTL_PCI_DEVICE with PCI_DEVICE" Greg Kroah-Hartman
2023-06-12 21:51 ` [PATCH 4.19 00/23] 4.19.286-rc1 review Chris Paterson
2023-06-12 22:25 ` Shuah Khan
2023-06-13  8:38 ` Jon Hunter
2023-06-13  9:37 ` Naresh Kamboju
2023-06-13 12:11 ` Sudip Mukherjee (Codethink)
2023-06-13 23:08 ` Guenter Roeck

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=20230612101651.407477221@linuxfoundation.org \
    --to=gregkh@linuxfoundation.org \
    --cc=davem@davemloft.net \
    --cc=edumazet@google.com \
    --cc=kuniyu@amazon.com \
    --cc=patches@lists.linux.dev \
    --cc=sashal@kernel.org \
    --cc=simon.horman@corigine.com \
    --cc=stable@vger.kernel.org \
    /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).