Re: [PATCH v2 mptcp-net] net: introduce and use lock_sock_fast_nested()

[Paolo Abeni <pabeni@redhat.com>]

On Fri, 2021-09-24 at 11:51 -0700, Mat Martineau wrote:
> On Fri, 24 Sep 2021, Paolo Abeni wrote:
> 
> > Syzkaller reported a false positive deadlock involving
> > the nl socket lock and the subflow socket lock:
> > 
> 
> ...
> 
> > the problem originates from uncorrect lock annotation in the mptcp
> > code and is only visible since commit 2dcb96bacce3 ("net: core: Correct
> > the sock::sk_lock.owned lockdep annotations"), but is present since
> > the port-based endpoint support initial implementation.
> > 
> > This patch addresses the issue introducing a nested variant of
> > lock_sock_fast() and using it in the relevant code path.
> > 
> > Fixes: 1729cf186d8a ("mptcp: create the listening socket for new port")
> > Fixes: 2dcb96bacce3 ("net: core: Correct the sock::sk_lock.owned lockdep annotations")
> > Suggested-by: Thomas Gleixner <tglx@linutronix.de>
> > Reported-and-tested-by: syzbot+1dd53f7a89b299d59eaf@syzkaller.appspotmail.com
> > Signed-off-by: Paolo Abeni <pabeni@redhat.com>
> > ---
> > v1 -> v2:
> > - fix a typo into a comment (Mat)
> 
> Thanks for the v2 and the answers about inline functions vs. exported 
> symbols.
> 
> Ok with me to merge in the export branch and/or upstream. I'm guessing you 
> might want to send this one to netdev yourself?

I'm fine doing that. I'll send it after somo more staging into the
export branch.

Thanks!

Paolo