Received-SPF: pass (google.com: best guess record for domain of linux-kernel-owner@vger.kernel.org designates 209.132.180.67 as permitted sender) client-ip=209.132.180.67;
Date:   Mon, 20 Jan 2020 12:23:59 -0800
From:   "Paul E. McKenney" <paulmck@kernel.org>
To:     Peter Zijlstra <peterz@infradead.org>
Cc:     Marco Elver <elver@google.com>, andreyknvl@google.com,
        glider@google.com, dvyukov@google.com, kasan-dev@googlegroups.com,
        linux-kernel@vger.kernel.org, mark.rutland@arm.com,
        will@kernel.org, boqun.feng@gmail.com, arnd@arndb.de,
        viro@zeniv.linux.org.uk, christophe.leroy@c-s.fr, dja@axtens.net,
        mpe@ellerman.id.au, rostedt@goodmis.org, mhiramat@kernel.org,
        mingo@kernel.org, christian.brauner@ubuntu.com,
        daniel@iogearbox.net, cyphar@cyphar.com, keescook@chromium.org,
        linux-arch@vger.kernel.org
Subject: Re: [PATCH 3/5] asm-generic, kcsan: Add KCSAN instrumentation for
 bitops
Message-ID: <20200120202359.GF2935@paulmck-ThinkPad-P72>
Reply-To: paulmck@kernel.org
References: <20200120141927.114373-1-elver@google.com>
 <20200120141927.114373-3-elver@google.com>
 <20200120144048.GB14914@hirez.programming.kicks-ass.net>
 <20200120162725.GE2935@paulmck-ThinkPad-P72>
 <20200120165223.GC14914@hirez.programming.kicks-ass.net>
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
In-Reply-To: <20200120165223.GC14914@hirez.programming.kicks-ass.net>
User-Agent: Mutt/1.9.4 (2018-02-28)
Sender: linux-kernel-owner@vger.kernel.org
Precedence: bulk

On Mon, Jan 20, 2020 at 05:52:23PM +0100, Peter Zijlstra wrote:
> On Mon, Jan 20, 2020 at 08:27:25AM -0800, Paul E. McKenney wrote:
> > On Mon, Jan 20, 2020 at 03:40:48PM +0100, Peter Zijlstra wrote:
> > > On Mon, Jan 20, 2020 at 03:19:25PM +0100, Marco Elver wrote:
> > > > Add explicit KCSAN checks for bitops.
> > > > 
> > > > Note that test_bit() is an atomic bitop, and we instrument it as such,
> > > 
> > > Well, it is 'atomic' in the same way that atomic_read() is. Both are
> > > very much not atomic ops, but are part of an interface that facilitates
> > > atomic operations.
> > 
> > True, but they all are either inline assembly or have either an
> > implicit or explicit cast to volatile, so they could be treated
> > the same as atomic_read(), correct?  If not, what am I missing?
> 
> Sure, but that is due to instrumentation requirements, not anything
> else.
> 
> Also note the distinct lack of __test_bit(), to mirror the non-atomic
> __set_bit() and __clear_bit().

OK, I will bite.  ;-)

We also don't have __atomic_read() and __atomic_set(), yet atomic_read()
and atomic_set() are considered to be non-racy, right?

							Thanx, Paul