linux-kernel.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
To: linux-kernel@vger.kernel.org, stable@vger.kernel.org
Cc: Greg Kroah-Hartman <gregkh@linuxfoundation.org>,
	alan@lxorguk.ukuu.org.uk, Patrick McHardy <kaber@trash.net>,
	Roland Dreier <roland@purestorage.com>
Subject: [ 26/33] IPoIB: Fix use-after-free of multicast object
Date: Thu,  4 Oct 2012 14:26:48 -0700	[thread overview]
Message-ID: <20121004210601.734344224@linuxfoundation.org> (raw)
In-Reply-To: <20121004210558.383865383@linuxfoundation.org>

3.0-stable review patch.  If anyone has any objections, please let me know.

------------------

From: Patrick McHardy <kaber@trash.net>

commit bea1e22df494a729978e7f2c54f7bda328f74bc3 upstream.

Fix a crash in ipoib_mcast_join_task().  (with help from Or Gerlitz)

Commit c8c2afe360b7 ("IPoIB: Use rtnl lock/unlock when changing device
flags") added a call to rtnl_lock() in ipoib_mcast_join_task(), which
is run from the ipoib_workqueue, and hence the workqueue can't be
flushed from the context of ipoib_stop().

In the current code, ipoib_stop() (which doesn't flush the workqueue)
calls ipoib_mcast_dev_flush(), which goes and deletes all the
multicast entries.  This takes place without any synchronization with
a possible running instance of ipoib_mcast_join_task() for the same
ipoib device, leading to a crash due to NULL pointer dereference.

Fix this by making sure that the workqueue is flushed before
ipoib_mcast_dev_flush() is called.  To make that possible, we move the
RTNL-lock wrapped code to ipoib_mcast_join_finish().

Signed-off-by: Patrick McHardy <kaber@trash.net>
Signed-off-by: Roland Dreier <roland@purestorage.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

---
 drivers/infiniband/ulp/ipoib/ipoib_main.c      |    2 +-
 drivers/infiniband/ulp/ipoib/ipoib_multicast.c |   19 ++++++++++---------
 2 files changed, 11 insertions(+), 10 deletions(-)

--- a/drivers/infiniband/ulp/ipoib/ipoib_main.c
+++ b/drivers/infiniband/ulp/ipoib/ipoib_main.c
@@ -148,7 +148,7 @@ static int ipoib_stop(struct net_device
 
 	netif_stop_queue(dev);
 
-	ipoib_ib_dev_down(dev, 0);
+	ipoib_ib_dev_down(dev, 1);
 	ipoib_ib_dev_stop(dev, 0);
 
 	if (!test_bit(IPOIB_FLAG_SUBINTERFACE, &priv->flags)) {
--- a/drivers/infiniband/ulp/ipoib/ipoib_multicast.c
+++ b/drivers/infiniband/ulp/ipoib/ipoib_multicast.c
@@ -189,7 +189,9 @@ static int ipoib_mcast_join_finish(struc
 
 	mcast->mcmember = *mcmember;
 
-	/* Set the cached Q_Key before we attach if it's the broadcast group */
+	/* Set the multicast MTU and cached Q_Key before we attach if it's
+	 * the broadcast group.
+	 */
 	if (!memcmp(mcast->mcmember.mgid.raw, priv->dev->broadcast + 4,
 		    sizeof (union ib_gid))) {
 		spin_lock_irq(&priv->lock);
@@ -197,10 +199,17 @@ static int ipoib_mcast_join_finish(struc
 			spin_unlock_irq(&priv->lock);
 			return -EAGAIN;
 		}
+		priv->mcast_mtu = IPOIB_UD_MTU(ib_mtu_enum_to_int(priv->broadcast->mcmember.mtu));
 		priv->qkey = be32_to_cpu(priv->broadcast->mcmember.qkey);
 		spin_unlock_irq(&priv->lock);
 		priv->tx_wr.wr.ud.remote_qkey = priv->qkey;
 		set_qkey = 1;
+
+		if (!ipoib_cm_admin_enabled(dev)) {
+			rtnl_lock();
+			dev_set_mtu(dev, min(priv->mcast_mtu, priv->admin_mtu));
+			rtnl_unlock();
+		}
 	}
 
 	if (!test_bit(IPOIB_MCAST_FLAG_SENDONLY, &mcast->flags)) {
@@ -586,14 +595,6 @@ void ipoib_mcast_join_task(struct work_s
 		return;
 	}
 
-	priv->mcast_mtu = IPOIB_UD_MTU(ib_mtu_enum_to_int(priv->broadcast->mcmember.mtu));
-
-	if (!ipoib_cm_admin_enabled(dev)) {
-		rtnl_lock();
-		dev_set_mtu(dev, min(priv->mcast_mtu, priv->admin_mtu));
-		rtnl_unlock();
-	}
-
 	ipoib_dbg_mcast(priv, "successfully joined all multicast groups\n");
 
 	clear_bit(IPOIB_MCAST_RUN, &priv->flags);



  parent reply	other threads:[~2012-10-04 21:30 UTC|newest]

Thread overview: 34+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2012-10-04 21:26 [ 00/33] 3.0.45-stable review Greg Kroah-Hartman
2012-10-04 21:26 ` [ 01/33] vfs: dcache: fix deadlock in tree traversal Greg Kroah-Hartman
2012-10-04 21:26 ` [ 02/33] dm: handle requests beyond end of device instead of using BUG_ON Greg Kroah-Hartman
2012-10-04 21:26 ` [ 03/33] USB: option: blacklist QMI interface on ZTE MF683 Greg Kroah-Hartman
2012-10-04 21:26 ` [ 04/33] USB: ftdi_sio: add TIAO USB Multi-Protocol Adapter (TUMPA) support Greg Kroah-Hartman
2012-10-04 21:26 ` [ 05/33] USB: qcaux: add Pantech vendor class match Greg Kroah-Hartman
2012-10-04 21:26 ` [ 06/33] staging: speakup_soft: Fix reading of init string Greg Kroah-Hartman
2012-10-04 21:26 ` [ 07/33] staging: comedi: s626: dont dereference insn->data Greg Kroah-Hartman
2012-10-04 21:26 ` [ 08/33] staging: comedi: jr3_pci: fix iomem dereference Greg Kroah-Hartman
2012-10-04 21:26 ` [ 09/33] staging: comedi: dont dereference user memory for INSN_INTTRIG Greg Kroah-Hartman
2012-10-04 21:26 ` [ 10/33] staging: comedi: fix memory leak for saved channel list Greg Kroah-Hartman
2012-10-04 21:26 ` [ 11/33] Remove BUG_ON from n_tty_read() Greg Kroah-Hartman
2012-10-04 21:26 ` [ 12/33] TTY: ttyprintk, dont touch behind tty->write_buf Greg Kroah-Hartman
2012-10-04 21:26 ` [ 13/33] serial: pl011: handle corruption at high clock speeds Greg Kroah-Hartman
2012-10-04 21:26 ` [ 14/33] serial: set correct baud_base for EXSYS EX-41092 Dual 16950 Greg Kroah-Hartman
2012-10-04 21:26 ` [ 15/33] b43legacy: Fix crash on unload when firmware not available Greg Kroah-Hartman
2012-10-04 21:26 ` [ 16/33] firmware: Add missing attributes to EFI variable attribute print out from sysfs Greg Kroah-Hartman
2012-10-04 21:26 ` [ 17/33] xhci: Intel Panther Point BEI quirk Greg Kroah-Hartman
2012-10-04 21:26 ` [ 18/33] n_gsm: added interlocking for gsm_data_lock for certain code paths Greg Kroah-Hartman
2012-10-04 21:26 ` [ 19/33] coredump: prevent double-free on an error path in core dumper Greg Kroah-Hartman
2012-10-04 21:26 ` [ 20/33] Increase XHCI suspend timeout to 16ms Greg Kroah-Hartman
2012-10-04 21:26 ` [ 21/33] n_gsm: memory leak in uplink error path Greg Kroah-Hartman
2012-10-04 21:26 ` [ 22/33] UBI: fix autoresize handling in R/O mode Greg Kroah-Hartman
2012-10-04 21:26 ` [ 23/33] SCSI: ibmvscsi: Fix host config length field overflow Greg Kroah-Hartman
2012-10-04 21:26 ` [ 24/33] SCSI: hpsa: Use LUN reset instead of target reset Greg Kroah-Hartman
2012-10-04 21:26 ` [ 25/33] can: mscan-mpc5xxx: fix return value check in mpc512x_can_get_clock() Greg Kroah-Hartman
2012-10-04 21:26 ` Greg Kroah-Hartman [this message]
2012-10-04 21:26 ` [ 27/33] IB/srp: Fix use-after-free in srp_reset_req() Greg Kroah-Hartman
2012-10-04 21:26 ` [ 28/33] IB/srp: Avoid having aborted requests hang Greg Kroah-Hartman
2012-10-04 21:26 ` [ 29/33] isci: fix isci_pci_probe() generates warning on efi failure path Greg Kroah-Hartman
2012-10-04 21:26 ` [ 30/33] x86/alternatives: Fix p6 nops on non-modular kernels Greg Kroah-Hartman
2012-10-04 21:26 ` [ 31/33] PCI: honor child buses add_size in hot plug configuration Greg Kroah-Hartman
2012-10-04 21:26 ` [ 32/33] SCSI: scsi_remove_target: fix softlockup regression on hot remove Greg Kroah-Hartman
2012-10-04 21:26 ` [ 33/33] SCSI: scsi_dh_alua: Enable STPG for unavailable ports Greg Kroah-Hartman

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=20121004210601.734344224@linuxfoundation.org \
    --to=gregkh@linuxfoundation.org \
    --cc=alan@lxorguk.ukuu.org.uk \
    --cc=kaber@trash.net \
    --cc=linux-kernel@vger.kernel.org \
    --cc=roland@purestorage.com \
    --cc=stable@vger.kernel.org \
    /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;
as well as URLs for NNTP newsgroup(s).