From: Jason Gunthorpe <jgunthorpe@obsidianresearch.com>
To: Haggai Eran <haggaie@mellanox.com>
Cc: Doug Ledford <dledford@redhat.com>,
linux-rdma@vger.kernel.org, netdev@vger.kernel.org,
Liran Liss <liranl@mellanox.com>,
Guy Shapiro <guysh@mellanox.com>,
Shachar Raindel <raindel@mellanox.com>,
Yotam Kenneth <yotamke@mellanox.com>
Subject: Re: [PATCH v4 for-next 05/12] IB/cm: Share listening CM IDs
Date: Tue, 19 May 2015 12:35:45 -0600 [thread overview]
Message-ID: <20150519183545.GH18675@obsidianresearch.com> (raw)
In-Reply-To: <1431841868-28063-6-git-send-email-haggaie@mellanox.com>
On Sun, May 17, 2015 at 08:51:01AM +0300, Haggai Eran wrote:
> @@ -212,6 +212,8 @@ struct cm_id_private {
> spinlock_t lock; /* Do not acquire inside cm.lock */
> struct completion comp;
> atomic_t refcount;
> + /* Number of clients sharing this ib_cm_id. Only valid for listeners. */
> + atomic_t sharecount;
No need for this atomic, hold the lock
The use of the atomic looks racy:
> + if (!atomic_dec_and_test(&cm_id_priv->sharecount)) {
> + /* The id is still shared. */
> + return;
> + }
Might race with this:
> + if (atomic_inc_return(&cm_id_priv->sharecount) == 1) {
> + /* This ID is already being destroyed */
> + atomic_dec(&cm_id_priv->sharecount);
> + goto new_id;
> + }
> +
Resulting in use-after-free of cm_id_priv->sharecount
Don't try and be clever with atomics, it is almost always wrong.
The share count should be 'listen_sharecount' because it *only* works
for listen.
The above test in cm_destroy_id should only be in the listen branch of
the if.
> + * Create a new listening ib_cm_id and listen on the given service ID.
> + *
> + * If there's an existing ID listening on that same device and service ID,
> + * return it.
> + *
.. Callers should call cm_destroy_id when done with the listen ..
Jason
next prev parent reply other threads:[~2015-05-19 18:35 UTC|newest]
Thread overview: 68+ messages / expand[flat|nested] mbox.gz Atom feed top
2015-05-17 5:50 [PATCH v4 for-next 00/12] Add network namespace support in the RDMA-CM Haggai Eran
2015-05-17 5:50 ` [PATCH v4 for-next 02/12] IB/addr: Pass network namespace as a parameter Haggai Eran
2015-05-17 5:50 ` [PATCH v4 for-next 03/12] IB/core: Find the network namespace matching connection parameters Haggai Eran
[not found] ` <1431841868-28063-4-git-send-email-haggaie-VPRAkNaXOzVWk0Htik3J/w@public.gmane.org>
2015-05-19 18:26 ` Jason Gunthorpe
[not found] ` <20150519182616.GF18675-ePGOBjL8dl3ta4EC/59zMFaTQe2KTcn/@public.gmane.org>
2015-05-20 14:48 ` Haggai Eran
2015-05-17 5:51 ` [PATCH v4 for-next 04/12] IB/ipoib: Return IPoIB devices " Haggai Eran
[not found] ` <1431841868-28063-5-git-send-email-haggaie-VPRAkNaXOzVWk0Htik3J/w@public.gmane.org>
2015-05-19 18:28 ` Jason Gunthorpe
[not found] ` <20150519182810.GG18675-ePGOBjL8dl3ta4EC/59zMFaTQe2KTcn/@public.gmane.org>
2015-05-20 15:17 ` Haggai Eran
2015-05-19 23:55 ` Jason Gunthorpe
[not found] ` <20150519235502.GB26634-ePGOBjL8dl3ta4EC/59zMFaTQe2KTcn/@public.gmane.org>
2015-05-21 5:33 ` Haggai Eran
[not found] ` <555D6E41.10606-VPRAkNaXOzVWk0Htik3J/w@public.gmane.org>
2015-05-21 5:48 ` Or Gerlitz
[not found] ` <CAJ3xEMjN+o=vC4abAeG5EuOo3Y1gSyh1qPDseA_aaYmoLWAunw-JsoAwUIsXosN+BqQ9rBEUg@public.gmane.org>
2015-05-21 6:33 ` Haggai Eran
[not found] ` <555D7C4A.2060708-VPRAkNaXOzVWk0Htik3J/w@public.gmane.org>
2015-05-21 10:31 ` Or Gerlitz
2015-05-21 17:43 ` Jason Gunthorpe
[not found] ` <20150521174336.GA6771-ePGOBjL8dl3ta4EC/59zMFaTQe2KTcn/@public.gmane.org>
2015-05-28 11:51 ` Haggai Eran
2015-05-28 15:45 ` Jason Gunthorpe
2015-05-21 5:48 ` Haggai Eran
2015-05-17 5:51 ` [PATCH v4 for-next 05/12] IB/cm: Share listening CM IDs Haggai Eran
2015-05-19 18:35 ` Jason Gunthorpe [this message]
[not found] ` <20150519183545.GH18675-ePGOBjL8dl3ta4EC/59zMFaTQe2KTcn/@public.gmane.org>
2015-05-19 22:35 ` Jason Gunthorpe
[not found] ` <20150519223502.GA26324-ePGOBjL8dl3ta4EC/59zMFaTQe2KTcn/@public.gmane.org>
2015-05-21 8:08 ` Haggai Eran
2015-05-21 17:54 ` Jason Gunthorpe
2015-05-21 7:07 ` Haggai Eran
[not found] ` <1431841868-28063-1-git-send-email-haggaie-VPRAkNaXOzVWk0Htik3J/w@public.gmane.org>
2015-05-17 5:50 ` [PATCH v4 for-next 01/12] IB/core: Add rwsem to allow reading device list or client list Haggai Eran
2015-05-17 5:51 ` [PATCH v4 for-next 06/12] IB/cm: Expose service ID in request events Haggai Eran
2015-05-17 5:51 ` [PATCH v4 for-next 07/12] IB/cma: Refactor RDMA IP CM private-data parsing code Haggai Eran
2015-05-17 5:51 ` [PATCH v4 for-next 08/12] IB/cma: Add compare_data checks to the RDMA CM module Haggai Eran
2015-05-17 5:51 ` [PATCH v4 for-next 12/12] IB/ucma: Take the network namespace from the process Haggai Eran
2015-05-17 5:51 ` [PATCH v4 for-next 09/12] IB/cma: Separate port allocation to network namespaces Haggai Eran
2015-05-17 5:51 ` [PATCH v4 for-next 10/12] IB/cma: Share CM IDs between namespaces Haggai Eran
2015-05-17 5:51 ` [PATCH v4 for-next 11/12] IB/cma: Add support for network namespaces Haggai Eran
2015-05-19 14:30 ` [PATCH v4 for-next 00/12] Add network namespace support in the RDMA-CM Yann Droneaud
2015-05-19 14:54 ` Haggai Eran
[not found] ` <555B4EBE.7010900-VPRAkNaXOzVWk0Htik3J/w@public.gmane.org>
2015-05-19 16:39 ` Parav Pandit
2015-05-19 18:01 ` Haggai Eran
[not found] ` <1432058488417.98688-VPRAkNaXOzVWk0Htik3J/w@public.gmane.org>
2015-05-19 18:42 ` Parav Pandit
2015-05-19 18:38 ` Jason Gunthorpe
[not found] ` <20150519183843.GI18675-ePGOBjL8dl3ta4EC/59zMFaTQe2KTcn/@public.gmane.org>
2015-05-19 18:44 ` Parav Pandit
[not found] ` <CAGgvQNTXAWkQWzBBrQfk39GaCQ2ck63AhgURpYFFBPTbkpx4kg-JsoAwUIsXosN+BqQ9rBEUg@public.gmane.org>
2015-05-19 19:20 ` Jason Gunthorpe
2015-05-26 13:34 ` Doug Ledford
[not found] ` <1432647280.28905.107.camel-H+wXaHxf7aLQT0dZR+AlfA@public.gmane.org>
2015-05-26 16:59 ` Jason Gunthorpe
2015-05-26 17:46 ` Doug Ledford
[not found] ` <1432662396.28905.157.camel-H+wXaHxf7aLQT0dZR+AlfA@public.gmane.org>
2015-05-26 18:47 ` Jason Gunthorpe
2015-05-28 13:22 ` Haggai Eran
2015-05-28 15:46 ` Jason Gunthorpe
[not found] ` <20150528154633.GB2962-ePGOBjL8dl3ta4EC/59zMFaTQe2KTcn/@public.gmane.org>
2015-06-03 10:07 ` Haggai Eran
2015-05-28 13:15 ` Haggai Eran
2015-05-26 17:55 ` Christian Benvenuti (benve)
2015-05-28 13:07 ` Haggai Eran
[not found] ` <55671309.6080303-VPRAkNaXOzVWk0Htik3J/w@public.gmane.org>
2015-05-28 14:07 ` Doug Ledford
[not found] ` <1432822057.114391.26.camel-H+wXaHxf7aLQT0dZR+AlfA@public.gmane.org>
2015-05-28 16:21 ` Or Gerlitz
[not found] ` <55674077.5040707-VPRAkNaXOzVWk0Htik3J/w@public.gmane.org>
2015-05-28 17:43 ` Jason Gunthorpe
[not found] ` <20150528174337.GA10448-ePGOBjL8dl3ta4EC/59zMFaTQe2KTcn/@public.gmane.org>
2015-05-28 18:22 ` Doug Ledford
[not found] ` <1432837360.114391.35.camel-H+wXaHxf7aLQT0dZR+AlfA@public.gmane.org>
2015-05-28 19:05 ` Or Gerlitz
[not found] ` <CAJ3xEMh2T5-56rFxWVdct2uAZYW1ZrKivWfS45V-mvhAfwyGaA-JsoAwUIsXosN+BqQ9rBEUg@public.gmane.org>
2015-05-28 21:55 ` Doug Ledford
[not found] ` <1432850150.114391.56.camel-H+wXaHxf7aLQT0dZR+AlfA@public.gmane.org>
2015-06-03 10:03 ` Haggai Eran
2015-06-03 16:14 ` Jason Gunthorpe
[not found] ` <20150603161447.GC12073-ePGOBjL8dl3ta4EC/59zMFaTQe2KTcn/@public.gmane.org>
2015-06-03 19:05 ` Or Gerlitz
2015-06-03 19:53 ` Jason Gunthorpe
[not found] ` <20150603195325.GC7902-ePGOBjL8dl3ta4EC/59zMFaTQe2KTcn/@public.gmane.org>
2015-06-03 20:07 ` Or Gerlitz
[not found] ` <CAJ3xEMiO+hEzOJ2oJ5G-mmBeaX4ZHvUyhNSAzsrRDui6dFjvCg-JsoAwUIsXosN+BqQ9rBEUg@public.gmane.org>
2015-06-03 21:45 ` Jason Gunthorpe
2015-06-04 9:41 ` Haggai Eran
2015-06-04 16:06 ` Jason Gunthorpe
2015-06-03 23:48 ` Jason Gunthorpe
[not found] ` <20150603234811.GA15128-ePGOBjL8dl3ta4EC/59zMFaTQe2KTcn/@public.gmane.org>
2015-06-04 6:24 ` Haggai Eran
[not found] ` <556FEF25.80409-VPRAkNaXOzVWk0Htik3J/w@public.gmane.org>
2015-06-04 16:40 ` Jason Gunthorpe
[not found] ` <20150604164058.GB27699-ePGOBjL8dl3ta4EC/59zMFaTQe2KTcn/@public.gmane.org>
2015-06-08 7:52 ` Haggai Eran
2015-06-08 16:53 ` Jason Gunthorpe
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=20150519183545.GH18675@obsidianresearch.com \
--to=jgunthorpe@obsidianresearch.com \
--cc=dledford@redhat.com \
--cc=guysh@mellanox.com \
--cc=haggaie@mellanox.com \
--cc=linux-rdma@vger.kernel.org \
--cc=liranl@mellanox.com \
--cc=netdev@vger.kernel.org \
--cc=raindel@mellanox.com \
--cc=yotamke@mellanox.com \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox