Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 23.128.96.18 as permitted sender) client-ip=23.128.96.18;
MIME-Version: 1.0
References: <20200511204150.27858-1-will@kernel.org> <20200512081826.GE2978@hirez.programming.kicks-ass.net>
 <CANpmjNNo3rhwqG=xEbpP9JiSd8-Faw8fkoUhYJjesHK5S5_KQQ@mail.gmail.com> <20200512190755.GL2957@hirez.programming.kicks-ass.net>
In-Reply-To: <20200512190755.GL2957@hirez.programming.kicks-ass.net>
From:   Marco Elver <elver@google.com>
Date:   Tue, 12 May 2020 22:31:44 +0200
Message-ID: <CANpmjNNeSnrAgfkskE5Y0NNu3-DS6hk+SwjkBunrr8FRxwwT-Q@mail.gmail.com>
Subject: Re: [PATCH v5 00/18] Rework READ_ONCE() to improve codegen
To:     Peter Zijlstra <peterz@infradead.org>
Cc:     Will Deacon <will@kernel.org>, LKML <linux-kernel@vger.kernel.org>,
        Thomas Gleixner <tglx@linutronix.de>,
        "Paul E. McKenney" <paulmck@kernel.org>,
        Ingo Molnar <mingo@kernel.org>
Content-Type: text/plain; charset="UTF-8"
Sender: linux-kernel-owner@vger.kernel.org
Precedence: bulk

On Tue, 12 May 2020 at 21:08, Peter Zijlstra <peterz@infradead.org> wrote:
>
> On Tue, May 12, 2020 at 07:53:00PM +0200, Marco Elver wrote:
> > I just ran a bunch of KCSAN tests. While this series alone would have
> > passed the tests, there appears to be a problem with
> > __READ_ONCE/__WRITE_ONCE. I think they should already be using
> > 'data_race()', as otherwise we will get lots of false positives in
> > future.
> >
> > I noticed this when testing -tip/locking/kcsan, which breaks
> > unfortunately, because I see a bunch of spurious data races with
> > arch_atomic_{read,set} because "locking/atomics: Flip fallbacks and
> > instrumentation" changed them to use __READ_ONCE()/__WRITE_ONCE().
> > From what I see, the intent was to not double-instrument,
> > unfortunately they are still double-instrumented because
> > __READ_ONCE/__WRITE_ONCE doesn't hide the access from KCSAN (nor KASAN
> > actually). I don't think we can use __no_sanitize_or_inline for the
> > arch_ functions, because we really want them to be __always_inline
> > (also to avoid calls to these functions in uaccess regions, which
> > objtool would notice).
> >
> > I think the easiest way to resolve this is to wrap the accesses in
> > __*_ONCE with data_race().
>
> But we can't... because I need arch_atomic_*() and __READ_ONCE() to not
> call out to _ANYTHING_.
>
> Sadly, because the compilers are 'broken' that whole __no_sanitize thing
> didn't work, but I'll be moving a whole bunch of code into .c files with
> all the sanitizers killed dead. And we'll be validating it'll not be
> calling out to anything.
>
> data_race() will include active calls to kcsan_{dis,en}able_current(),
> and this must not happen.

Only if instrumentation is enabled for the compilation unit. If you
have KCSAN_SANITIZE_foo.c := n, no calls are emitted not even to
kcsan_{dis,en}able_current(). Does that help?

By default, right now __READ_ONCE() will still generate a call due to
instrumentation (call to __tsan_readX).

Thanks,
-- Marco