MIME-Version: 1.0
In-Reply-To: <20170424083250.h2wv2exbi4ytigac@hirez.programming.kicks-ass.net>
References: <20170421220939.GA65363@beast> <20170424083250.h2wv2exbi4ytigac@hirez.programming.kicks-ass.net>
From: Kees Cook <keescook@chromium.org>
Date: Mon, 24 Apr 2017 13:16:24 -0700
Message-ID: <CAGXu5jLY0DpTWhL-CXOF15BvFvQ_4+_pg+vMbuvKqLu9tU_jQw@mail.gmail.com>
Subject: Re: [PATCH] x86/refcount: Implement fast refcount_t handling
To: Peter Zijlstra <peterz@infradead.org>
Cc: LKML <linux-kernel@vger.kernel.org>,
        Eric Biggers <ebiggers3@gmail.com>,
        Christoph Hellwig <hch@infradead.org>,
        "axboe@kernel.dk" <axboe@kernel.dk>,
        James Bottomley <James.Bottomley@hansenpartnership.com>,
        Elena Reshetova <elena.reshetova@intel.com>,
        Hans Liljestrand <ishkamiel@gmail.com>,
        David Windsor <dwindsor@gmail.com>, "x86@kernel.org" <x86@kernel.org>,
        Ingo Molnar <mingo@kernel.org>, Arnd Bergmann <arnd@arndb.de>,
        Greg Kroah-Hartman <gregkh@linuxfoundation.org>,
        Jann Horn <jann@thejh.net>, "David S. Miller" <davem@davemloft.net>,
        linux-arch <linux-arch@vger.kernel.org>,
        "kernel-hardening@lists.openwall.com" 
        <kernel-hardening@lists.openwall.com>,
        PaX Team <pageexec@freemail.hu>
Content-Type: text/plain; charset=UTF-8
Sender: linux-kernel-owner@vger.kernel.org
Content-Length: 1161
Lines: 30

On Mon, Apr 24, 2017 at 1:32 AM, Peter Zijlstra <peterz@infradead.org> wrote:
> On Fri, Apr 21, 2017 at 03:09:39PM -0700, Kees Cook wrote:
>> This patch ports the x86-specific atomic overflow handling from PaX's
>> PAX_REFCOUNT to the upstream refcount_t API. This is an updated version
>> from PaX that eliminates the saturation race condition by resetting the
>> atomic counter back to the INT_MAX saturation value on both overflow and
>> underflow. To win a race, a system would have to have INT_MAX threads
>> simultaneously overflow before the saturation handler runs.
>
> And is this impossible? Highly unlikely I'll grant you, but absolutely
> impossible?

I'll adjust the language. "Highly unlikely" is still better than
"trivially doable with a single thread". :)

> Also, you forgot nr_cpus in your bound. Afaict the worst case here is
> O(nr_tasks + 3*nr_cpus).
>
> Because PaX does it, is not a correctness argument. And this really
> wants one.

Sure, I didn't mean to imply anything other than a demonstration of
what PaX is doing (and that it's better than not having it). If we can
improve it, that's great.

-Kees

-- 
Kees Cook
Pixel Security