CVE-2026-80678: i2c: imx: Fix slave registration race and error handling
In the Linux kernel, the following vulnerability has been resolved:
i2c: imx: Fix slave registration race and error handling
In i2cimxregslave(), the slave pointer was assigned before pmruntimeresumeandget(). If pmruntimeresumeandget() failed, the error path returned without clearing i2cimx->slave, leaving it non-NULL and causing all subsequent registration attempts to fail with -EBUSY.
Additionally, because this driver uses a shared IRQ, the interrupt handler i2cimxisr() can execute concurrently and, after acquiring slavelock, dereference i2cimx->slave. The previous fix attempt added a lockless i2cimx->slave = NULL on the error path, but that could race with the ISR under the lock and still cause a NULL pointer dereference.
Fix both issues by deferring the assignment of i2cimx->slave and i2cimx->lastslaveevent to after a successful resume, and by performing the assignment inside the slavelock critical section. This guarantees that the slave pointer is never left stale on the error path and is always valid when observed by the interrupt handler.
Affected Software
Event History
Frequently Asked Questions
Which systems are exposed to this issue?
Systems using the Linux Kernel i2c-imx driver and registering an I2C slave are affected by the faulty registration handling. The race is relevant because the driver uses a shared IRQ, allowing its interrupt handler to run concurrently with slave registration.
What conditions cause the registration failure or crash risk?
If pm_runtime_resume_and_get() fails after the driver records the slave pointer, that pointer can remain stale and later registration attempts can fail with -EBUSY. A concurrent interrupt handler can also observe an improperly updated slave pointer, creating a NULL pointer dereference risk.
How can an administrator recognize the registration-handling failure?
After a runtime-resume failure during slave registration, subsequent attempts to register a slave may consistently fail with -EBUSY. This behavior indicates that the driver may have retained a non-NULL slave pointer from the failed attempt.