Skip to contentSkip to conversation

net: devmem: use READ_ONCE/WRITE_ONCE on binding->dev

PATCH Applied Bobby Eshleman 2 participants 2 messages Mar 3 – Mar 5

Subsystems (2)
lore
Share & tools
lore
#0 lore raw

PATCH v2 1/1 net: devmem: use READ_ONCE/WRITE_ONCE on binding->dev

From: Bobby Eshleman <[email protected]>

binding->dev is protected on the write-side in mp_dmabuf_devmem_uninstall() against concurrent writes, but due to the concurrent bare reads in net_devmem_get_binding() and validate_xmit_unreadable_skb() it should be wrapped in a READ_ONCE/WRITE_ONCE pair to make sure no compiler optimizations play with the underlying register in unforeseen ways.

Doesn't present a critical bug because the known compiler optimizations don't result in bad behavior. There is no tearing on u64, and load omissions/invented loads would only break if additional binding->dev references were inlined together (they aren't right now).

This just more strictly follows the linux memory model (i.e., "Lock-Protected Writes With Lockless Reads" in tools/memory-model/Documentation/access-marking.txt).

Fixes bd61848900bf ("net: devmem: Implement TX path")

---
Changes in v2:
- add ASSERT_EXCLUSIVE_WRITER() to mark write-side critical section
- add READ_ONCE() in validate_xmit_unreadable_skb() too
- Link to v1: https://lore.kernel.org/r/[email protected]

---
 net/core/dev.c    | 2 +-
 net/core/devmem.c | 6 ++++--
 2 files changed, 5 insertions(+), 3 deletions(-)
net/core/dev.c +1 -1
index c1a9f7fdcffa..9036fcf9f965 100644
@@ -3987,7 +3987,7 @@ static struct sk_buff *validate_xmit_unreadable_skb(struct sk_buff *skb,
 	if (shinfo->nr_frags > 0) {
 		niov = netmem_to_net_iov(skb_frag_netmem(&shinfo->frags[0]));
 		if (net_is_devmem_iov(niov) &&
-		    net_devmem_iov_binding(niov)->dev != dev)
+		    READ_ONCE(net_devmem_iov_binding(niov)->dev) != dev)
 			goto out_free;
 	}
 
net/core/devmem.c +4 -2
index 8c9aad776bf4..69d79aee07ef 100644
@@ -396,7 +396,8 @@ struct net_devmem_dmabuf_binding *net_devmem_get_binding(struct sock *sk,
 	 * net_device.
 	 */
 	dst_dev = dst_dev_rcu(dst);
-	if (unlikely(!dst_dev) || unlikely(dst_dev != binding->dev)) {
+	if (unlikely(!dst_dev) ||
+	    unlikely(dst_dev != READ_ONCE(binding->dev))) {
 		err = -ENODEV;
 		goto out_unlock;
 	}
@@ -513,7 +514,8 @@ static void mp_dmabuf_devmem_uninstall(void *mp_priv,
 			xa_erase(&binding->bound_rxqs, xa_idx);
 			if (xa_empty(&binding->bound_rxqs)) {
 				mutex_lock(&binding->lock);
-				binding->dev = NULL;
+				ASSERT_EXCLUSIVE_WRITER(binding->dev);
+				WRITE_ONCE(binding->dev, NULL);
 				mutex_unlock(&binding->lock);
 			}
 			break;

---
base-commit: 9439a661c2e80485406ce2c90b107ca17858382d
change-id: 20260223-devmem-membar-fix-3a5cd9618f8a

Best regards,

Signature · 1 line
Bobby Eshleman <[email protected]>
#1 lore raw
This patch was applied to netdev/net.git (main) by Jakub Kicinski <[email protected]>:

Hello:

This patch was applied to netdev/net.git (main)
by Jakub Kicinski <[email protected]>:

Quoted 10 lines · On Mon, 02 Mar 2026 16:32:56 -0800 you wrote:

From: Bobby Eshleman <[email protected]>

binding->dev is protected on the write-side in mp_dmabuf_devmem_uninstall() against concurrent writes, but due to the concurrent bare reads in net_devmem_get_binding() and validate_xmit_unreadable_skb() it should be wrapped in a READ_ONCE/WRITE_ONCE pair to make sure no compiler optimizations play with the underlying register in unforeseen ways.

[...]

Here is the summary with links:
- [net,v2] net: devmem: use READ_ONCE/WRITE_ONCE on binding->dev

    https://git.kernel.org/netdev/net/c/40bf00ec2ee2

You are awesome, thank you!

Signature · 2 lines
Deet-doot-dot, I am a bot.
https://korg.docs.kernel.org/patchwork/pwbot.html