Feedback-ID: iad51458e:Fastmail
Date:   Mon, 13 Feb 2023 18:09:16 -0800
From:   Boqun Feng <boqun.feng@gmail.com>
To:     Alan Stern <stern@rowland.harvard.edu>
Cc:     Peter Zijlstra <peterz@infradead.org>,
        Kent Overstreet <kent.overstreet@linux.dev>,
        Kent Overstreet <kent.overstreet@gmail.com>,
        Linus Torvalds <torvalds@linux-foundation.org>,
        Coly Li <colyli@suse.de>,
        Tetsuo Handa <penguin-kernel@i-love.sakura.ne.jp>,
        syzkaller <syzkaller@googlegroups.com>,
        Dmitry Vyukov <dvyukov@google.com>,
        Greg Kroah-Hartman <gregkh@linuxfoundation.org>,
        "Rafael J. Wysocki" <rafael@kernel.org>,
        Ingo Molnar <mingo@redhat.com>,
        LKML <linux-kernel@vger.kernel.org>,
        USB list <linux-usb@vger.kernel.org>,
        Hillf Danton <hdanton@sina.com>
Subject: Re: [PATCH RFC] drivers/core: Replace lockdep_set_novalidate_class()
 with unique class keys
Message-ID: <Y+rtTLD9UtPI5uGM@boqun-archlinux>
References: <Y+hRurRwm//1+IcK@rowland.harvard.edu>
 <Y+hTEtCKPuO0zGIt@moria.home.lan>
 <Y+hW74TAVzCpSv7c@rowland.harvard.edu>
 <Y+hYn6uzIUBaxDdV@moria.home.lan>
 <Y+kEgDLSRwdODRdD@rowland.harvard.edu>
 <Y+oBveWO2z6xdTW/@hirez.programming.kicks-ass.net>
 <Y+pWhyFJeE93nlWd@rowland.harvard.edu>
 <Y+plfZnEqw6mG+XH@hirez.programming.kicks-ass.net>
 <Y+rpD7QPheQQ8Lxj@boqun-archlinux>
 <Y+rr4p6njVOTSYxs@rowland.harvard.edu>
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
In-Reply-To: <Y+rr4p6njVOTSYxs@rowland.harvard.edu>
Precedence: bulk

On Mon, Feb 13, 2023 at 09:03:14PM -0500, Alan Stern wrote:
> On Mon, Feb 13, 2023 at 05:51:11PM -0800, Boqun Feng wrote:
> > Basically if you have two lock instances A and B with the same class,
> > and you know that locking ordering is always A -> B, then you can do
> > 
> > 	mutex_lock(A);
> > 	mutex_lock_nest_lock(B, A); // lock B.
> > 
> > to tell the lockdep this is not deadlock, plus lockdep will treat the
> > acquisition of A and the precondition of acquisition B, so the following
> > is not a deadlock as well:
> > 
> > T1:
> > 	mutex_lock(A);
> > 	mutex_lock(C);
> > 	mutex_lock_nest_lock(B, A);
> > 
> > T2:
> > 	mutex_lock(A);
> > 	mutex_lock_nest_lock(B, A);
> > 	mutex_lock(C);
> 
> Why isn't this treated as a deadlock?  It sure looks like a deadlock to 
> me.  Is this an example where lockdep just doesn't get the right answer?
> 

Because A serializes B and C, so that particular piece of code doesn't
cause deadlock. Note that you can still use you normal mutex_lock() for
B, so if there is more code:

T3:
	mutex_lock(C);
	mutex_lock(B);

lockdep will report deadlock.

Regards,
Boqun

> Alan Stern