From: "Amrani, Ram" <Ram.Amrani@cavium.com>
To: Jia-Ju Bai <baijiaju1990@163.com>,
        "monis@mellanox.com" <monis@mellanox.com>,
        "sean.hefty@intel.com" <sean.hefty@intel.com>,
        "dledford@redhat.com" <dledford@redhat.com>,
        "hal.rosenstock@gmail.com" <hal.rosenstock@gmail.com>,
        "leon@kernel.org" <leon@kernel.org>
CC: "linux-rdma@vger.kernel.org" <linux-rdma@vger.kernel.org>,
        "linux-kernel@vger.kernel.org" <linux-kernel@vger.kernel.org>
Subject: RE: [PATCH V2] rxe: Fix a sleep-in-atomic bug in post_one_send
Thread-Topic: [PATCH V2] rxe: Fix a sleep-in-atomic bug in post_one_send
Thread-Index: AQHS2qrXKsehwmu4okacQzxsMYR4cqIPpyMQ
Date: Thu, 1 Jun 2017 08:11:12 +0000
Message-ID: <BN3PR07MB25781AF1049545F0555D3E83F8F60@BN3PR07MB2578.namprd07.prod.outlook.com>
References: <1496303141-14925-1-git-send-email-baijiaju1990@163.com>
In-Reply-To: <1496303141-14925-1-git-send-email-baijiaju1990@163.com>
Accept-Language: en-US
Content-Language: en-US
spamdiagnosticoutput: 1:99
spamdiagnosticmetadata: NSPM
Content-Type: text/plain; charset="us-ascii"
MIME-Version: 1.0
X-MS-Exchange-CrossTenant-originalarrivaltime: 01 Jun 2017 08:11:12.3459
 (UTC)
X-MS-Exchange-CrossTenant-fromentityheader: Hosted
X-MS-Exchange-CrossTenant-id: 711e4ccf-2e9b-4bcf-a551-4094005b6194
X-MS-Exchange-Transport-CrossTenantHeadersStamped: BN3PR07MB2579
Sender: linux-kernel-owner@vger.kernel.org
List-ID: <linux-kernel.vger.kernel.org>
X-Mailing-List: linux-kernel@vger.kernel.org
Content-Transfer-Encoding: 8bit
X-MIME-Autoconverted: from quoted-printable to 8bit by mail.home.local id v518BNph002380
Content-Length: 1259
Lines: 35

> The driver may sleep under a spin lock, and the function call path is:
> post_one_send (acquire the lock by spin_lock_irqsave)
>   init_send_wqe
>     copy_from_user --> may sleep
> 
> To fix it, the lock is released before copy_from_user, and the lock is
> acquired again after this function. The parameter "flags" is used to
> restore and save the irq status.
> Thank Leon for good advice.
> 

...

>  	init_send_wr(qp, &wqe->wr, ibwr);
> @@ -742,7 +742,12 @@ static int init_send_wqe(struct rxe_qp *qp, struct ib_send_wr *ibwr,
>  		for (i = 0; i < num_sge; i++, sge++) {
>  			if (qp->is_user && copy_from_user(p, (__user void *)
>  					    (uintptr_t)sge->addr, sge->length))
> -				return -EFAULT;
> +				spin_unlock_irqrestore(&qp->sq.sq_lock, *flags);
> +				err = copy_from_user(p, (__user void *)
> +					    (uintptr_t)sge->addr, sge->length);
> +				spin_lock_irqsave(&qp->sq.sq_lock, *flags);
> +				if (qp->is_user && err)
> +					return -EFAULT;
> 
>  			else if (!qp->is_user)
>  				memcpy(p, (void *)(uintptr_t)sge->addr,

This isn't my area of expertise. Still something seems weird.
You are still calling 'copy_from_user' unprotected in the 'if'.
Also, did you mean to use curly brackets on the indented part after the first if?!

Ram