Skip to contentSkip to conversation

ipv6: prevent in6_dev_get() from resurrecting inet6_dev

PATCH Superseded David Lee 3 participants 3 messages Jul 31 – Aug 2

Subsystems (3)

v1 of 2 Superseded Latest is v2 →

lore
Share & tools
lore

David Lee

author
#0 lore raw

PATCH 1/1 ipv6: prevent in6_dev_get() from resurrecting inet6_dev

in6_dev_get() reads dev->ip6_ptr under RCU and then unconditionally increments its refcount. Device teardown can clear the pointer and drop the last reference between these operations. The increment then resurrects an object whose RCU free has already been queued, so callers can use it after it is freed.

Use refcount_inc_not_zero() and return NULL when the object has already reached zero. RCU keeps the memory accessible through the attempted reference acquisition, and a successful increment pins the object for the caller.

Fixes 8814c4b53381 ("[IPV6] ADDRCONF: Convert addrconf_lock to RCU.")

Bug found and triaged by OpenAI Security Research and validated by Trail of Bits.

Assisted-by Codex:gpt-5.6-sol gpt-5.5-cyber

---
Trail of Bits has a reproducer for this bug that triggers a KASAN use-after-free and can share if needed.

 include/net/addrconf.h | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)
include/net/addrconf.h +2 -2
index 8ced27a82..e67642459 100644
@@ -405,8 +405,8 @@ static inline struct inet6_dev *in6_dev_get(const struct net_device *dev)
 
 	rcu_read_lock();
 	idev = rcu_dereference(dev->ip6_ptr);
-	if (idev)
-		refcount_inc(&idev->refcnt);
+	if (idev && !refcount_inc_not_zero(&idev->refcnt))
+		idev = NULL;
 	rcu_read_unlock();
 	return idev;
 }
#1 lore raw
Quoted 6 lines · On Friday, 31 July 2026 15:52:01 CEST David Lee wrote:
Fixes 8814c4b53381 ("[IPV6] ADDRCONF: Convert addrconf_lock to RCU.")

Bug found and triaged by OpenAI Security Research and validated by Trail of Bits.

Assisted-by Codex:gpt-5.6-sol gpt-5.5-cyber

1. weird text (advertisement) + newlines in the tags section which shouldn't be there
https://docs.kernel.org/process/submitting-patches.html#sign-your-work-the-developer-s-certificate-of-origin
(and all these tag/format related sections below)

2. author didn't sign off the patch (see below)

3. odd unrelated Signed-off-by without any obvious connection to the patch

- I actually would guess that Kyle Zeng <[email protected]> is the author and David Lee removed its authorship (From: ) before sending the patch. But in its current form, David is claiming to be the author. And even when it is the case, David must still add the the non-Author SoB at the end.

Regards,
Sven

maintainer
#2 lore raw
In reply to David Lee

In addition to the other process comments you got, subject prefix should have been "[PATCH net]" ("[PATCH net v2]" now). See:

https://docs.kernel.org/process/maintainer-netdev.html

Get the process right with one patch, then send the rest.

Quoted 12 lines · David Lee

in6_dev_get() reads dev->ip6_ptr under RCU and then unconditionally increments its refcount. Device teardown can clear the pointer and drop the last reference between these operations. The increment then resurrects an object whose RCU free has already been queued, so callers can use it after it is freed.

Use refcount_inc_not_zero() and return NULL when the object has already reached zero. RCU keeps the memory accessible through the attempted reference acquisition, and a successful increment pins the object for the caller.

Fixes 8814c4b53381 ("[IPV6] ADDRCONF: Convert addrconf_lock to RCU.")
Quoted 8 lines

Bug found and triaged by OpenAI Security Research and validated by Trail of Bits.

Assisted-by Codex:gpt-5.6-sol gpt-5.5-cyber

---
Trail of Bits has a reproducer for this bug that triggers a KASAN use-after-free and can share if needed.

Didn't you see an "addition on 0; use-after-free" trace before KASAN? Please include it in the commit message.

Was the bug reproduced on an unpatched kernel?

Also, what about IPv4? in_dev_get() has the same structure and at least inet_netconf_get_devconf() is calling it without holding RTNL.