Wind River Support Network

Description

In the Linux kernel, the following vulnerability has been resolved:[EOL][EOL]io_uring: wait interruptibly for request completions on exit[EOL][EOL]WHen the ring exits, cleanup is done and the final cancelation and[EOL]waiting on completions is done by io_ring_exit_work. That function is[EOL]invoked by kworker, which doesn't take any signals. Because of that, it[EOL]doesn't really matter if we wait for completions in TASK_INTERRUPTIBLE[EOL]or TASK_UNINTERRUPTIBLE state. However, it does matter to the hung task[EOL]detection checker![EOL][EOL]Normally we expect cancelations and completions to happen rather[EOL]quickly. Some test cases, however, will exit the ring and park the[EOL]owning task stopped (eg via SIGSTOP). If the owning task needs to run[EOL]task_work to complete requests, then io_ring_exit_work won't make any[EOL]progress until the task is runnable again. Hence io_ring_exit_work can[EOL]trigger the hung task detection, which is particularly problematic if[EOL]panic-on-hung-task is enabled.[EOL][EOL]As the ring exit doesn't take signals to begin with, have it wait[EOL]interruptibly rather than uninterruptibly. io_uring has a separate[EOL]stuck-exit warning that triggers independently anyway, so we're not[EOL]really missing anything by making this switch.