CVE-2021-46970

En el kernel de Linux, se ha resuelto la siguiente vulnerabilidad: bus: mhi: pci_generic: Eliminar el indicador WQ_MEM_RECLAIM de la cola de trabajo de estado Un cambio reciente creó una cola de trabajo dedicada para el trabajo de cambio de estado con los indicadores WQ_HIGHPRI (sin ninguna razón importante para ello) y WQ_MEM_RECLAIM. pero el trabajo de cambio de estado (mhi_pm_st_worker) no garantiza el progreso hacia adelante bajo presión de la memoria, e incluso esperará varias asignaciones de memoria cuando, por ejemplo, se crean dispositivos, se carga firmware, etc... El trabajo entonces no forma parte de una ruta de recuperación de memoria. .. Además, esto provoca una advertencia en check_flush_dependency() ya que terminamos en un código que vacía una cola de trabajo que no es de recuperación: [ 40.969601] cola de trabajo: WQ_MEM_RECLAIM mhi_hiprio_wq:mhi_pm_st_worker [mhi] está descargando !WQ_MEM_RECLAIM events_highpri:flush_backlog [ 40.969612] ADVERTENCIA : CPU: 4 PID: 158 en kernel/workqueue.c:2607 check_flush_dependency+0x11c/0x140 [40.969733] Seguimiento de llamadas: [40.969740] __flush_work+0x97/0x1d0 [40.969745]? proceso_despertador+0x15/0x20 [40.969749]? insertar_trabajo+0x70/0x80 [40.969750]? __queue_work+0x14a/0x3e0 [ 40.969753] Flush_work+0x10/0x20 [ 40.969756] rollback_registered_many+0x1c9/0x510 [ 40.969759] unregister_netdevice_queue+0x94/0x120 [ 40.969761] anular el registro _netdev+0x1d/0x30 [ 40.969765] mhi_net_remove+0x1a/0x40 [mhi_net] [ 40.969770 ] mhi_driver_remove+0x124/0x250 [mhi] [ 40.969776] dispositivo_release_driver_internal+0xf0/0x1d0 [ 40.969778] dispositivo_release_driver+0x12/0x20 [ 40.969782] bus_remove_device+0xe1/0x150 [ 40.9 69786] dispositivo_del+0x17b/0x3e0 [ 40.969791] mhi_destroy_device+0x9a/0x100 [ mhi] [40.969796]? mhi_unmap_single_use_bb+0x50/0x50 [mhi] [ 40.969799] dispositivo_para_cada_niño+0x5e/0xa0 [ 40.969804] mhi_pm_st_worker+0x921/0xf50 [mhi]

In the Linux kernel, the following vulnerability has been resolved: bus: mhi: pci_generic: Remove WQ_MEM_RECLAIM flag from state workqueue A recent change created a dedicated workqueue for the state-change work with WQ_HIGHPRI (no strong reason for that) and WQ_MEM_RECLAIM flags, but the state-change work (mhi_pm_st_worker) does not guarantee forward progress under memory pressure, and will even wait on various memory allocations when e.g. creating devices, loading firmware, etc... The work is then not part of a memory reclaim path... Moreover, this causes a warning in check_flush_dependency() since we end up in code that flushes a non-reclaim workqueue: [ 40.969601] workqueue: WQ_MEM_RECLAIM mhi_hiprio_wq:mhi_pm_st_worker [mhi] is flushing !WQ_MEM_RECLAIM events_highpri:flush_backlog [ 40.969612] WARNING: CPU: 4 PID: 158 at kernel/workqueue.c:2607 check_flush_dependency+0x11c/0x140 [ 40.969733] Call Trace: [ 40.969740] __flush_work+0x97/0x1d0 [ 40.969745] ? wake_up_process+0x15/0x20 [ 40.969749] ? insert_work+0x70/0x80 [ 40.969750] ? __queue_work+0x14a/0x3e0 [ 40.969753] flush_work+0x10/0x20 [ 40.969756] rollback_registered_many+0x1c9/0x510 [ 40.969759] unregister_netdevice_queue+0x94/0x120 [ 40.969761] unregister_netdev+0x1d/0x30 [ 40.969765] mhi_net_remove+0x1a/0x40 [mhi_net] [ 40.969770] mhi_driver_remove+0x124/0x250 [mhi] [ 40.969776] device_release_driver_internal+0xf0/0x1d0 [ 40.969778] device_release_driver+0x12/0x20 [ 40.969782] bus_remove_device+0xe1/0x150 [ 40.969786] device_del+0x17b/0x3e0 [ 40.969791] mhi_destroy_device+0x9a/0x100 [mhi] [ 40.969796] ? mhi_unmap_single_use_bb+0x50/0x50 [mhi] [ 40.969799] device_for_each_child+0x5e/0xa0 [ 40.969804] mhi_pm_st_worker+0x921/0xf50 [mhi]