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;
Subject: Re: [PATCH 4.14 17/89] futex: Prevent overflow by strengthen input
 validation
To:     Greg Kroah-Hartman <gregkh@linuxfoundation.org>,
        linux-kernel@vger.kernel.org
Cc:     stable@vger.kernel.org, Li Jinyue <lijinyue@huawei.com>,
        Thomas Gleixner <tglx@linutronix.de>, peterz@infradead.org,
        dvhart@infradead.org
References: <20180122083954.683903493@linuxfoundation.org>
 <20180122083956.427607580@linuxfoundation.org>
From:   Jiri Slaby <jslaby@suse.cz>
Message-ID: <20567b49-e0a3-e9e0-d520-74bbfc20d414@suse.cz>
Date:   Thu, 25 Jan 2018 14:45:21 +0100
User-Agent: Mozilla/5.0 (X11; Linux x86_64; rv:52.0) Gecko/20100101
 Thunderbird/52.5.2
MIME-Version: 1.0
In-Reply-To: <20180122083956.427607580@linuxfoundation.org>
Content-Type: text/plain; charset=utf-8
Content-Language: en-GB
Content-Transfer-Encoding: 7bit
Sender: linux-kernel-owner@vger.kernel.org
Precedence: bulk

On 01/22/2018, 09:44 AM, Greg Kroah-Hartman wrote:
> 4.14-stable review patch.  If anyone has any objections, please let me know.
> 
> ------------------
> 
> From: Li Jinyue <lijinyue@huawei.com>
> 
> commit fbe0e839d1e22d88810f3ee3e2f1479be4c0aa4a upstream.
> 
> UBSAN reports signed integer overflow in kernel/futex.c:
> 
>  UBSAN: Undefined behaviour in kernel/futex.c:2041:18
>  signed integer overflow:
>  0 - -2147483648 cannot be represented in type 'int'
> 
> Add a sanity check to catch negative values of nr_wake and nr_requeue.
> 
> Signed-off-by: Li Jinyue <lijinyue@huawei.com>
> Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
> Cc: peterz@infradead.org
> Cc: dvhart@infradead.org
> Link: https://lkml.kernel.org/r/1513242294-31786-1-git-send-email-lijinyue@huawei.com
> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
> 
> ---
>  kernel/futex.c |    3 +++
>  1 file changed, 3 insertions(+)
> 
> --- a/kernel/futex.c
> +++ b/kernel/futex.c
> @@ -1878,6 +1878,9 @@ static int futex_requeue(u32 __user *uad
>  	struct futex_q *this, *next;
>  	DEFINE_WAKE_Q(wake_q);
>  
> +	if (nr_wake < 0 || nr_requeue < 0)
> +		return -EINVAL;

This breaks strace's test suite on 4.14.15 (and is present in upstream
obviously too):
    futex(0x7ff568b44ffc, 0x3, 0xfacefeed, 0xbadda7a0ca7b100d,
0x7ff568b44ffc, 0x9caffee1) = -1: Invalid argument

strace uses weird values in the testkit to pass down to futex as can be
seen. I think like in:
commit e78c38f6bdd900b2ad9ac9df8eff58b745dc5b3c
Author: Jiri Slaby <jslaby@suse.cz>
Date:   Mon Oct 23 13:41:51 2017 +0200

    futex: futex_wake_op, do not fail on invalid op

something similar should be done here too. Maybe:
if (nr_wake < 0)
  nr_wake = 0;
if (nr_requeue < 0)
  nr_requeue = 0;
?

Maybe also with some pr_info_ratelimited like in the above commit?

thanks,
-- 
js
suse labs