Re: general protection fault in tomoyo_socket_sendmsg_permission

[Shuah Khan]

On 2/10/21 12:07 PM, Tetsuo Handa wrote:
> On 2021/02/11 3:17, Shuah Khan wrote:
> > I am looking to understand the syzbot configuration and a reproducer
> > to be able to debug and fix the problem. How is syzbot triggering the
> > vhci_hcd attach and detach sequence?
>
> I don't know. I'm waiting for syzbot to reproduce the problem on linux-next
> with https://git.kernel.org/pub/scm/linux/kernel/git/next/linux-next.git/commit/drivers/usb/usbip?id=f1bdf414e7dd0cbc26460425719fc3ea479947a2 .
>
> > This helps me determine all these fix suggestions that are coming in
> > are fixes or papering over a real problem.
>
> What are these fix suggestions?
>
> "general protection fault in tomoyo_socket_sendmsg_permission" is a NULL pointer
> dereference which can happen if vhci_device_reset() and/or vhci_device_init()
> (which does vdev->ud.tcp_socket = NULL;) were unexpectedly called. There is no
> reproducer, and (as far as I know) no fix suggestion.
Right. I would like to get a clear understanding of how this condition
is triggered. I am not saying this isn't a problem. Understanding how
it is triggered helps find the best fix.

> "KASAN: null-ptr-deref Write in vhci_shutdown_connection" is an ERR_PTR(-EINTR)
> pointer dereference which can happen if kthread_create() was SIGKILLed. There is
> a reproducer, and https://lkml.kernel.org/r/20210205135707.4574-1-penguin-kernel@xxxxxxxxxxxxxxxxxxx
> is a fix suggestion.

This is a good find. I already replied to the thread to send a complete
fix.

thanks,
-- Shuah