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;
Date:   Wed, 15 Sep 2021 10:18:45 -0300
From:   Jason Gunthorpe <jgg@nvidia.com>
To:     Mark Zhang <markzhang@nvidia.com>
Cc:     Tao Liu <thomas.liu@ucloud.cn>, dledford@redhat.com,
        leon@kernel.org, haakon.bugge@oracle.com, shayd@nvidia.com,
        avihaih@nvidia.com, linux-rdma@vger.kernel.org,
        linux-kernel@vger.kernel.org, thomas.liu@ucloud.com
Subject: Re: [PATCH] RDMA/cma: Fix listener leak in rdma_cma_listen_on_all()
 failure
Message-ID: <20210915131845.GK4065468@nvidia.com>
References: <20210913093344.17230-1-thomas.liu@ucloud.cn>
 <20210914195444.GA156389@nvidia.com>
 <4b711688-9a82-03f1-844c-cf1557892679@nvidia.com>
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
In-Reply-To: <4b711688-9a82-03f1-844c-cf1557892679@nvidia.com>
MIME-Version: 1.0
X-MS-Exchange-AntiSpam-MessageData-ChunkCount: 1
X-MS-Exchange-AntiSpam-MessageData-0: =?us-ascii?Q?nzvjPK34iT0a70+vpDtr4DqeSI+3ynqJtVbgz3T0nJdTLpW0R/U/Md8OVIV4?=
 =?us-ascii?Q?QbparT4misaUhcZa17rkvBKIgNnpxHABKRdwpaPFchr1s8wEsfianc6ZD3Oj?=
 =?us-ascii?Q?ffhKkCB/fAGEueCiwNmrhpj0hgToxwHciM0VPPAwQjDX07dUgRX7rDNzhjJp?=
 =?us-ascii?Q?LQyfCRvB8xg14JWa4qxcUZ+o1ZAo2r1/7I9JyjP64xGR/z1DaPV4wRFAuWCI?=
 =?us-ascii?Q?LakAIVF30DMdPpQEpynlr2WSuZXbKxlsU8qZxLWe6PsESZAC9UR8x4Ru8v4q?=
 =?us-ascii?Q?jwlWERbvu/P55QrnBd+Lj6Y/7QezVUgtf/goj63HyR9p951n3vynZ65SEPKm?=
 =?us-ascii?Q?+BOcx3itziqAbPoUMa6dMekOACbo2kfA78DS+8a5rIofBy6FM++Om2BlC8k8?=
 =?us-ascii?Q?kqUUBa7SZXH2SHvwrqMRYsdlfVMiA+Avxt3P2Lycy0BAX080ryhdlTvzZQuX?=
 =?us-ascii?Q?er0qnzeH6NFr4NM8mxlGt/dWAnpYdNNxur2JGL+9/OgsE9a2MvOWNjnnxhj7?=
 =?us-ascii?Q?Gd0zHaa68sYLH0PQZ3KifikwubTwkSni27Hj4cm8jmqx8X8GP12QLLbiMS6O?=
 =?us-ascii?Q?MkSVYMlYuOsLuzB5J+7M3ZtqBDRC87sSjpwyKCIRu2gMFUtkJG86Ade4xY1a?=
 =?us-ascii?Q?fjWhg8VgD8i8WbIcZ5Td3U9ryTThyd5wqykZqn1fWxfAXyR/t50kwUrbZ62p?=
 =?us-ascii?Q?YWBQQLPoSnYplCtwWK8McBFqq5giGm+7I2ObgB4Yk62xAij2il17zWV4B4YS?=
 =?us-ascii?Q?22R7wDpknbLQqJGsSu3/mTJrobANb9Hq7ieb6P7XxKDSXLQ/53VRBpK1WEBK?=
 =?us-ascii?Q?0x3MKKT+kb79goEHE8JiWmY+WVGvhJAy4qDSF2YYZ1SdWWRFicHQeVmhyTwZ?=
 =?us-ascii?Q?mGZ0Xc4B/IQdoh1DSE/DQTxsanUmGL8UiFXuccFDPgwOLv1+nq761rPI1fGQ?=
 =?us-ascii?Q?F29JP4r34lZBmCMszInFE+sPANKK3fExo8NMthnOwDZfUZGX4RlGGVdOHTOh?=
 =?us-ascii?Q?vaIkgZxA6gKzB0syIvMhoFEq3KHm6WGPPTcI+N42X4flHiEj40sW2haao3uB?=
 =?us-ascii?Q?6Yw0FwB6iKYyRFTODH9UzMIJuwkH1/ZtNKWrl/UgymkXEtW8Bb3R0lTZCWTH?=
 =?us-ascii?Q?JsV5SwAODXdVg3FGndGQ6jmYWAHhntmcLv6K9W7k/7MPet/dJdsYZgEtuDgT?=
 =?us-ascii?Q?d7ZL2GSY91oX+sunmBd1UzgjJ9kAbRbxRGN4hp/FVVB/f1idQ1fnsT655MYr?=
 =?us-ascii?Q?AxI5/xHnr4jIofe7vk84CmOv6Sqaa/x8O4tFhxxjczmLCX8uER1+rfNUj03n?=
 =?us-ascii?Q?GrA4uZ1B1TEdVZNET0tWCd6G?=
X-OriginatorOrg: Nvidia.com
X-MS-Exchange-CrossTenant-Network-Message-Id: b7237844-a4bd-40a7-6658-08d9784b5999
X-MS-Exchange-CrossTenant-AuthSource: BL0PR12MB5506.namprd12.prod.outlook.com
X-MS-Exchange-CrossTenant-AuthAs: Internal
X-MS-Exchange-CrossTenant-OriginalArrivalTime: 15 Sep 2021 13:18:47.2503
 (UTC)
X-MS-Exchange-CrossTenant-FromEntityHeader: Hosted
X-MS-Exchange-CrossTenant-Id: 43083d15-7273-40c1-b7db-39efd9ccc17a
X-MS-Exchange-CrossTenant-MailboxType: HOSTED
X-MS-Exchange-CrossTenant-UserPrincipalName: oNEoy+8ton74fBqEeVQafDk1WF7tE+T+4yGncTfmA5jI+1UBmB//7YsFPe7b7h43
X-MS-Exchange-Transport-CrossTenantHeadersStamped: BL1PR12MB5206
Precedence: bulk
List-ID: <linux-kernel.vger.kernel.org>
X-Mailing-List: linux-kernel@vger.kernel.org

On Wed, Sep 15, 2021 at 02:34:19PM +0800, Mark Zhang wrote:
> > -       case RDMA_CM_LISTEN:
> > -               if (cma_any_addr(cma_src_addr(id_priv)) && !id_priv->cma_dev)
> > -                       cma_cancel_listens(id_priv);
> > -               break;
> 
> If this case is removed, is this code path still good?
> 
> cma_add_one/cma_remove_one
>   cma_process_remove
>     cma_send_device_removal_put
>       cma_cancel_operation

Yes, it is complicated, but the cma_process_remove() won't proceed
past cma_dev_put() until all the ULPs have gone through _destroy_id()

So all this does is move the reaping of the listens to the destroy_id
time

Jason