CVE-2021-47069

In the Linux kernel, the following vulnerability has been resolved: ipc/mqueue, msg, sem: avoid relying on a stack reference past its expiry do_mq_timedreceive calls wq_sleep with a stack local address. The sender (do_mq_timedsend) uses this address to later call pipelined_send. This leads to a very hard to trigger race where a do_mq_timedreceive call might return and leave do_mq_timedsend to rely on an invalid address, causing the following crash: RIP: 0010:wake_q_add_safe+0x13/0x60 Call Trace: __x64_sys_mq_timedsend+0x2a9/0x490 do_syscall_64+0x80/0x680 entry_SYSCALL_64_after_hwframe+0x44/0xa9 RIP: 0033:0x7f5928e40343 The race occurs as: 1. do_mq_timedreceive calls wq_sleep with the address of `struct ext_wait_queue` on function stack (aliased as `ewq_addr` here) - it holds a valid `struct ext_wait_queue *` as long as the stack has not been overwritten. 2. `ewq_addr` gets added to info->e_wait_q[RECV].list in wq_add, and do_mq_timedsend receives it via wq_get_first_waiter(info, RECV) to call __pipelined_op. 3. Sender calls __pipelined_op::smp_store_release(&this->state, STATE_READY). Here is where the race window begins. (`this` is `ewq_addr`.) 4. If the receiver wakes up now in do_mq_timedreceive::wq_sleep, it will see `state == STATE_READY` and break. 5. do_mq_timedreceive returns, and `ewq_addr` is no longer guaranteed to be a `struct ext_wait_queue *` since it was on do_mq_timedreceive's stack. (Although the address may not get overwritten until another function happens to touch it, which means it can persist around for an indefinite time.) 6. do_mq_timedsend::__pipelined_op() still believes `ewq_addr` is a `struct ext_wait_queue *`, and uses it to find a task_struct to pass to the wake_q_add_safe call. In the lucky case where nothing has overwritten `ewq_addr` yet, `ewq_addr->task` is the right task_struct. In the unlucky case, __pipelined_op::wake_q_add_safe gets handed a bogus address as the receiver's task_struct causing the crash. do_mq_timedsend::__pipelined_op() should not dereference `this` after setting STATE_READY, as the receiver counterpart is now free to return. Change __pipelined_op to call wake_q_add_safe on the receiver's task_struct returned by get_task_struct, instead of dereferencing `this` which sits on the receiver's stack. As Manfred pointed out, the race potentially also exists in ipc/msg.c::expunge_all and ipc/sem.c::wake_up_sem_queue_prepare. Fix those in the same way.

In the Linux kernel, the following vulnerability has been resolved: ipc/mqueue, msg, sem: avoid relying on a stack reference past its expiry do_mq_timedreceive calls wq_sleep with a stack local address. The sender (do_mq_timedsend) uses this address to later call pipelined_send. This leads to a very hard to trigger race where a do_mq_timedreceive call might return and leave do_mq_timedsend to rely on an invalid address, causing the following crash: RIP: 0010:wake_q_add_safe+0x13/0x60 Call Trace: __x64_sys_mq_timedsend+0x2a9/0x490 do_syscall_64+0x80/0x680 entry_SYSCALL_64_after_hwframe+0x44/0xa9 RIP: 0033:0x7f5928e40343 The race occurs as: 1. do_mq_timedreceive calls wq_sleep with the address of `struct ext_wait_queue` on function stack (aliased as `ewq_addr` here) - it holds a valid `struct ext_wait_queue *` as long as the stack has not been overwritten. 2. `ewq_addr` gets added to info->e_wait_q[RECV].list in wq_add, and do_mq_timedsend receives it via wq_get_first_waiter(info, RECV) to call __pipelined_op. 3. Sender calls __pipelined_op::smp_store_release(&this->state, STATE_READY). Here is where the race window begins. (`this` is `ewq_addr`.) 4. If the receiver wakes up now in do_mq_timedreceive::wq_sleep, it will see `state == STATE_READY` and break. 5. do_mq_timedreceive returns, and `ewq_addr` is no longer guaranteed to be a `struct ext_wait_queue *` since it was on do_mq_timedreceive's stack. (Although the address may not get overwritten until another function happens to touch it, which means it can persist around for an indefinite time.) 6. do_mq_timedsend::__pipelined_op() still believes `ewq_addr` is a `struct ext_wait_queue *`, and uses it to find a task_struct to pass to the wake_q_add_safe call. In the lucky case where nothing has overwritten `ewq_addr` yet, `ewq_addr->task` is the right task_struct. In the unlucky case, __pipelined_op::wake_q_add_safe gets handed a bogus address as the receiver's task_struct causing the crash. do_mq_timedsend::__pipelined_op() should not dereference `this` after setting STATE_READY, as the receiver counterpart is now free to return. Change __pipelined_op to call wake_q_add_safe on the receiver's task_struct returned by get_task_struct, instead of dereferencing `this` which sits on the receiver's stack. As Manfred pointed out, the race potentially also exists in ipc/msg.c::expunge_all and ipc/sem.c::wake_up_sem_queue_prepare. Fix those in the same way.

En el kernel de Linux, se resolvió la siguiente vulnerabilidad: ipc/mqueue, msg, sem: evite confiar en una referencia de pila después de su vencimiento do_mq_timedreceive llama a wq_sleep con una dirección local de pila. El remitente (do_mq_timedsend) usa esta dirección para luego llamar a pipelined_send. Esto conduce a una ejecución muy difícil de desencadenar en la que una llamada do_mq_timedreceive puede regresar y dejar que do_mq_timedsend dependa de una dirección no válida, lo que provoca el siguiente bloqueo: RIP: 0010:wake_q_add_safe+0x13/0x60 Seguimiento de llamadas: __x64_sys_mq_timedsend+0x2a9/0x490 do_syscall_64+0x80 /0x680 Entry_SYSCALL_64_after_hwframe+0x44/0xa9 RIP: 0033:0x7f5928e40343 La ejecución ocurre como: 1. do_mq_timedreceive llama a wq_sleep con la dirección de `struct ext_wait_queue` en la pila de funciones (alias `ewq_addr` aquí): contiene una `struct ext_wait_queue * válida ` siempre y cuando la pila no haya sido sobrescrita. 2. `ewq_addr` se agrega a info->e_wait_q[RECV].list en wq_add, y do_mq_timedsend lo recibe a través de wq_get_first_waiter(info, RECV) para llamar a __pipelined_op. 3. El remitente llama a __pipelined_op::smp_store_release(&this->state, STATE_READY). Aquí es donde comienza la ventana de ejecución. (`esto` es `ewq_addr`.) 4. Si el receptor se activa ahora en do_mq_timedreceive::wq_sleep, verá `state == STATE_READY` y se interrumpirá. 5. do_mq_timedreceive regresa y ya no se garantiza que `ewq_addr` sea una `struct ext_wait_queue *` ya que estaba en la pila de do_mq_timedreceive. (Aunque es posible que la dirección no se sobrescriba hasta que otra función la toque, lo que significa que puede persistir por un tiempo indefinido). 6. do_mq_timedsend::__pipelined_op() todavía cree que `ewq_addr` es una `struct ext_wait_queue *`, y lo usa para encontrar una task_struct para pasar a la llamada wake_q_add_safe. En el afortunado caso de que nada haya sobrescrito `ewq_addr` todavía, `ewq_addr->task` es la estructura de tarea correcta. En el desafortunado caso, __pipelined_op::wake_q_add_safe recibe una dirección falsa como la task_struct del receptor que causa el bloqueo. do_mq_timedsend::__pipelined_op() no debe eliminar la referencia a "esto" después de configurar STATE_READY, ya que la contraparte del receptor ahora puede regresar. Cambie __pipelined_op para llamar a wake_q_add_safe en el task_struct del receptor devuelto por get_task_struct, en lugar de desreferenciar "this" que se encuentra en la pila del receptor. Como señaló Manfred, la ejecución también existe potencialmente en ipc/msg.c::expunge_all e ipc/sem.c::wake_up_sem_queue_prepare. Arréglelos de la misma manera.