In the Linux kernel, the following vulnerability has been resolved: net: Remove RTNL dance for SIOCBRADDIF and SIOCBRDELIF. SIOCBRDELIF is passed to dev_ioctl() first and later forwarded to br_ioctl_call(), which causes unnecessary RTNL dance and the splat below [0] under RTNL pressure. Let's say Thread A is trying to detach a device from a bridge and Thread B is trying to remove the bridge. In dev_ioctl(), Thread A bumps the bridge device's refcnt by netdev_hold() and releases RTNL because the following br_ioctl_call() also re-acquires RTNL. In the race window, Thread B could acquire RTNL and try to remove the bridge device. Then, rtnl_unlock() by Thread B will release RTNL and wait for netdev_put() by Thread A. Thread A, however, must hold RTNL after the unlock in dev_ifsioc(), which may take long under RTNL pressure, resulting in the splat by Thread B. Thread A (SIOCBRDELIF) Thread B (SIOCBRDELBR) ---------------------- ---------------------- sock_ioctl sock_ioctl `- sock_do_ioctl `- br_ioctl_call `- dev_ioctl `- br_ioctl_stub |- rtnl_lock | |- dev_ifsioc ' ' |- dev = __dev_get_by_name(...) |- netdev_hold(dev, ...) . / |- rtnl_unlock ------. | | |- br_ioctl_call `---> |- rtnl_lock Race | | `- br_ioctl_stub |- br_del_bridge Window | | | |- dev = __dev_get_by_name(...) | | | May take long | `- br_dev_delete(dev, ...) | | | under RTNL pressure | `- unregister_netdevice_queue(dev, ...) | | | | `- rtnl_unlock \ | |- rtnl_lock <-' `- netdev_run_todo | |- ... `- netdev_run_todo | `- rtnl_unlock |- __rtnl_unlock | |- netdev_wait_allrefs_any |- netdev_put(dev, ...) <----------------' Wait refcnt decrement and log splat below To avoid blocking SIOCBRDELBR unnecessarily, let's not call dev_ioctl() for SIOCBRADDIF and SIOCBRDELIF. In the dev_ioctl() path, we do the following: 1. Copy struct ifreq by get_user_ifreq in sock_do_ioctl() 2. Check CAP_NET_ADMIN in dev_ioctl() 3. Call dev_load() in dev_ioctl() 4. Fetch the master dev from ifr.ifr_name in dev_ifsioc() 3. can be done by request_module() in br_ioctl_call(), so we move 1., 2., and 4. to br_ioctl_stub(). Note that 2. is also checked later in add_del_if(), but it's better performed before RTNL. SIOCBRADDIF and SIOCBRDELIF have been processed in dev_ioctl() since the pre-git era, and there seems to be no specific reason to process them there. [0]: unregister_netdevice: waiting for wpan3 to become free. Usage count = 2 ref_tracker: wpan3@ffff8880662d8608 has 1/1 users at __netdev_tracker_alloc include/linux/netdevice.h:4282 [inline] netdev_hold include/linux/netdevice.h:4311 [inline] dev_ifsioc+0xc6a/0x1160 net/core/dev_ioctl.c:624 dev_ioctl+0x255/0x10c0 net/core/dev_ioctl.c:826 sock_do_ioctl+0x1ca/0x260 net/socket.c:1213 sock_ioctl+0x23a/0x6c0 net/socket.c:1318 vfs_ioctl fs/ioctl.c:51 [inline] __do_sys_ioctl fs/ioctl.c:906 [inline] __se_sys_ioctl fs/ioctl.c:892 [inline] __x64_sys_ioctl+0x1a4/0x210 fs/ioctl.c:892 do_syscall_x64 arch/x86/entry/common.c:52 [inline] do_syscall_64+0xcb/0x250 arch/x86/entry/common.c:83 entry_SYSCALL_64_after_hwframe+0x77/0x7f
En el kernel de Linux, se ha resuelto la siguiente vulnerabilidad: net: Eliminación del movimiento RTNL para SIOCBRADDIF y SIOCBRDELIF. SIOCBRDELIF se pasa primero a dev_ioctl() y luego a br_ioctl_call(), lo que provoca un movimiento RTNL innecesario y el splat debajo de [0] bajo presión RTNL. Supongamos que el subproceso A intenta desconectar un dispositivo de un puente y el subproceso B intenta eliminar el puente. En dev_ioctl(), el subproceso A aumenta el refcnt del dispositivo puente mediante netdev_hold() y libera RTNL porque el subproceso br_ioctl_call() también vuelve a adquirir RTNL. En la ventana de ejecución, el subproceso B podría adquirir RTNL e intentar eliminar el dispositivo puente. Luego, rtnl_unlock() del subproceso B liberará RTNL y esperará a netdev_put() del subproceso A. Sin embargo, el subproceso A debe mantener RTNL después del desbloqueo en dev_ifsioc(), lo que puede tardar mucho bajo la presión de RTNL, lo que resulta en el splat del subproceso B. Subproceso A (SIOCBRDELIF) Subproceso B (SIOCBRDELBR) ---------------------- ---------------------- sock_ioctl sock_ioctl `- sock_do_ioctl `- br_ioctl_call `- dev_ioctl `- br_ioctl_stub |- rtnl_lock | |- dev_ifsioc ' ' |- dev = __dev_get_by_name(...) |- netdev_hold(dev, ...) . / |- rtnl_unlock ------. | | |- br_ioctl_call `---> |- rtnl_lock Race | | `- br_ioctl_stub |- br_del_bridge Ventana | | | |- dev = __dev_get_by_name(...) | | | Puede tomar mucho tiempo | `- br_dev_delete(dev, ...) | | | bajo presión RTNL | `- unregister_netdevice_queue(dev, ...) | | | | `- rtnl_unlock \ | |- rtnl_lock <-' `- netdev_run_todo | |- ... `- netdev_run_todo | `- rtnl_unlock |- __rtnl_unlock | |- netdev_wait_allrefs_any |- netdev_put(dev, ...) <----------------' Espere la disminución de refcnt y registre splat a continuación Para evitar bloquear SIOCBRDELBR innecesariamente, no llamemos a dev_ioctl() para SIOCBRADDIF y SIOCBRDELIF. En la ruta dev_ioctl(), hacemos lo siguiente: 1. Copiar struct ifreq por get_user_ifreq en sock_do_ioctl() 2. Verificar CAP_NET_ADMIN en dev_ioctl() 3. Llamar a dev_load() en dev_ioctl() 4. Obtener el dev maestro de ifr.ifr_name en dev_ifsioc() 3. se puede hacer por request_module() en br_ioctl_call(), por lo que movemos 1., 2. y 4. a br_ioctl_stub(). Tenga en cuenta que 2. también se verifica más adelante en add_del_if(), pero se realiza mejor antes de RTNL. SIOCBRADDIF y SIOCBRDELIF se han procesado en dev_ioctl() desde la era anterior a Git, y no parece haber una razón específica para procesarlos allí. [0]: unregister_netdevice: esperando a que wpan3 quede libre. Recuento de uso = 2 ref_tracker: wpan3@ffff8880662d8608 tiene 1/1 usuarios en __netdev_tracker_alloc include/linux/netdevice.h:4282 [en línea] netdev_hold include/linux/netdevice.h:4311 [en línea] dev_ifsioc+0xc6a/0x1160 net/core/dev_ioctl.c:624 dev_ioctl+0x255/0x10c0 net/core/dev_ioctl.c:826 sock_do_ioctl+0x1ca/0x260 net/socket.c:1213 sock_ioctl+0x23a/0x6c0 net/socket.c:1318 vfs_ioctl fs/ioctl.c:51 [en línea] __do_sys_ioctl fs/ioctl.c:906 [en línea] __se_sys_ioctl fs/ioctl.c:892 [en línea] __x64_sys_ioctl+0x1a4/0x210 fs/ioctl.c:892 do_syscall_x64 arch/x86/entry/common.c:52 [en línea] do_syscall_64+0xcb/0x250 arch/x86/entry/common.c:83 entry_SYSCALL_64_after_hwframe+0x77/0x7f
CVSS:3.1/AV:L/AC:L/PR:L/UI:N/S:U/C:N/I:N/A:H
| Attack Vector | LOCAL |
|---|---|
| Attack Complexity | LOW |
| Privileges Required | LOW |
| User Interaction | NONE |
| Scope | UNCHANGED |
| Confidentiality Impact | NONE |
| Integrity Impact | NONE |
| Availability Impact | HIGH |
| Source | Type | Description |
|---|---|---|
| [email protected] | Primary |
en
NVD-CWE-noinfo
|
| Vendor | Product | Version | Update | Type |
|---|---|---|---|---|
| linux | linux_kernel | * | <built-in method update of dict object at 0x702bdd67d6c0> | Operating System |
| Vulnerable | CPE |
|---|---|
| Yes | cpe:2.3:o:linux:linux_kernel:*:*:*:*:*:*:*:* |