CVE-2026-72069

In the Linux kernel, the following vulnerability has been resolved:

locking/rt: Fix the incorrect RCU protection in rtspinunlock()

rtspinunlock() releases the RCU protection before unlocking the

lock. That opens the door for the following UAF scenario:

 T1 T2

 spinlock(&p->lock); rcuread_lock();

 invalidate(p); p = rcu_dereference(ptr);

 rcuassignpointer(ptr, NULL); if (!p) return;

 spinunlock(&p->lock); spinlock(&p->lock)

     lock(&lock->lock);

   rcureadlock();

 kfreercu(p); rcuread_unlock();

....

spin_unlock(&p->lock)

  rcureadunlock(); // Ends grace period

 rcudobatch()

   kfree(p);

    UAF ->   rtmutexcmpxchg_release(&lock->lock...)

Regular spinlocks keep preemption disabled accross the unlock operation,

which provides full RCU protection, but the RT substitution fails to

resemble that. Same applies for the rwlock substitution.

Move the rcureadunlock() invocation past the unlock operations to match

the non-RT semantics. This makes it asymmetric vs. rtxxxlock(), but

that's harmless as the caller needs to hold RCU read lock across the lock

operation. The migrate_enable() call stays before the unlock operation

because there is no per CPU operation in the unlock path which would

require migration to be kept disabled.