aio: Fix null ptr deref in aio_complete() wakeup

commit caeb4b0a11b3393e43f7fa8e0a5a18462acc66bd upstream.

list_del_init_careful() needs to be the last access to the wait queue
entry - it effectively unlocks access.

Previously, finish_wait() would see the empty list head and skip taking
the lock, and then we'd return - but the completion path would still
attempt to do the wakeup after the task_struct pointer had been
overwritten.

Fixes: 71eb6b6b0ba9 ("fs/aio: obey min_nr when doing wakeups")
Cc: stable@vger.kernel.org
Link: https://lore.kernel.org/linux-fsdevel/CAHTA-ubfwwB51A5Wg5M6H_rPEQK9pNf8FkAGH=vr=FEkyRrtqw@mail.gmail.com/
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
Link: https://lore.kernel.org/stable/20240331215212.522544-1-kent.overstreet%40linux.dev
Link: https://lore.kernel.org/r/20240331215212.522544-1-kent.overstreet@linux.dev
Signed-off-by: Christian Brauner <brauner@kernel.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

diff --git a/fs/aio.c b/fs/aio.c
index 9cdaa2faa5363333627e0cba54a4efe75b45b144..0f4f531c97800c648437fb2eb7409ccc2b198536 100644 (file)
--- a/fs/aio.c
+++ b/fs/aio.c
@@ -1202,8 +1202,8 @@ static void aio_complete(struct aio_kiocb *iocb)
                spin_lock_irqsave(&ctx->wait.lock, flags);
                list_for_each_entry_safe(curr, next, &ctx->wait.head, w.entry)
                        if (avail >= curr->min_nr) {
-                               list_del_init_careful(&curr->w.entry);
                                wake_up_process(curr->w.private);
+                               list_del_init_careful(&curr->w.entry);
                        }
                spin_unlock_irqrestore(&ctx->wait.lock, flags);
        }