diff mbox

[v2] IB/hfi1: Protect the global dev_cntr_names and port_cntr_names

Message ID 20170523194407.28059.59929.stgit@phlsvslse11.ph.intel.com (mailing list archive)
State Not Applicable
Headers show

Commit Message

Marciniszyn, Mike May 23, 2017, 7:44 p.m. UTC
From: Tadeusz Struk <tadeusz.struk@intel.com>

commit 62eed66e98b4c2286fef2ce5911d8d75b7515f7b upstream.

Protect the global dev_cntr_names and port_cntr_names with the global
mutex as they are allocated and freed in a function called per device.
Otherwise there is a danger of double free and memory leaks.

Cc: stable <stable@vger.kernel.org> # v4.10
Fixes: Commit b7481944b06e ("IB/hfi1: Show statistics counters under IB stats interface")
Reviewed-by: Dennis Dalessandro <dennis.dalessandro@intel.com>
Reviewed-by: Easwar Hariharan <easwar.hariharan@intel.com>
Signed-off-by: Tadeusz Struk <tadeusz.struk@intel.com>
Signed-off-by: Dennis Dalessandro <dennis.dalessandro@intel.com>
Signed-off-by: Mike Marciniszyn <mike.marciniszyn@intel.com>
Signed-off-by: Doug Ledford <dledford@redhat.com>
---
v2 - fix upstream commit id.
 drivers/infiniband/hw/hfi1/verbs.c |   12 +++++++++++-
 1 file changed, 11 insertions(+), 1 deletion(-)


--
To unsubscribe from this list: send the line "unsubscribe linux-rdma" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Comments

Greg KH May 23, 2017, 7:55 p.m. UTC | #1
On Tue, May 23, 2017 at 03:44:29PM -0400, Mike Marciniszyn wrote:
> From: Tadeusz Struk <tadeusz.struk@intel.com>
> 
> commit 62eed66e98b4c2286fef2ce5911d8d75b7515f7b upstream.
> 
> Protect the global dev_cntr_names and port_cntr_names with the global
> mutex as they are allocated and freed in a function called per device.
> Otherwise there is a danger of double free and memory leaks.
> 
> Cc: stable <stable@vger.kernel.org> # v4.10

Why resend this???
--
To unsubscribe from this list: send the line "unsubscribe linux-rdma" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Marciniszyn, Mike May 23, 2017, 7:59 p.m. UTC | #2
> > Cc: stable <stable@vger.kernel.org> # v4.10
> 
> Why resend this???

There are two patches, one for 4.10 and one for 4.11.

Each of the v2's should have had a different version on the stable cc line in the signoff block.

Mike
--
To unsubscribe from this list: send the line "unsubscribe linux-rdma" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Greg Kroah-Hartman May 23, 2017, 8:06 p.m. UTC | #3
On Tue, May 23, 2017 at 07:59:21PM +0000, Marciniszyn, Mike wrote:
> > > Cc: stable <stable@vger.kernel.org> # v4.10
> > 
> > Why resend this???
> 
> There are two patches, one for 4.10 and one for 4.11.
> 
> Each of the v2's should have had a different version on the stable cc line in the signoff block.

But I already emailed you saying, 4.10 is end-of-life, no one is doing
anything with it, so why resend it?

greg k-h
--
To unsubscribe from this list: send the line "unsubscribe linux-rdma" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
diff mbox

Patch

diff --git a/drivers/infiniband/hw/hfi1/verbs.c b/drivers/infiniband/hw/hfi1/verbs.c
index 6002aa9..5105b95 100644
--- a/drivers/infiniband/hw/hfi1/verbs.c
+++ b/drivers/infiniband/hw/hfi1/verbs.c
@@ -1605,6 +1605,7 @@  static void hfi1_get_dev_fw_str(struct ib_device *ibdev, char *str,
 	"DRIVER_EgrHdrFull"
 };
 
+static DEFINE_MUTEX(cntr_names_lock); /* protects the *_cntr_names bufers */
 static const char **dev_cntr_names;
 static const char **port_cntr_names;
 static int num_driver_cntrs = ARRAY_SIZE(driver_cntr_names);
@@ -1659,6 +1660,7 @@  static struct rdma_hw_stats *alloc_hw_stats(struct ib_device *ibdev,
 {
 	int i, err;
 
+	mutex_lock(&cntr_names_lock);
 	if (!cntr_names_initialized) {
 		struct hfi1_devdata *dd = dd_from_ibdev(ibdev);
 
@@ -1667,8 +1669,10 @@  static struct rdma_hw_stats *alloc_hw_stats(struct ib_device *ibdev,
 				      num_driver_cntrs,
 				      &num_dev_cntrs,
 				      &dev_cntr_names);
-		if (err)
+		if (err) {
+			mutex_unlock(&cntr_names_lock);
 			return NULL;
+		}
 
 		for (i = 0; i < num_driver_cntrs; i++)
 			dev_cntr_names[num_dev_cntrs + i] =
@@ -1682,10 +1686,12 @@  static struct rdma_hw_stats *alloc_hw_stats(struct ib_device *ibdev,
 		if (err) {
 			kfree(dev_cntr_names);
 			dev_cntr_names = NULL;
+			mutex_unlock(&cntr_names_lock);
 			return NULL;
 		}
 		cntr_names_initialized = 1;
 	}
+	mutex_unlock(&cntr_names_lock);
 
 	if (!port_num)
 		return rdma_alloc_hw_stats_struct(
@@ -1903,9 +1909,13 @@  void hfi1_unregister_ib_device(struct hfi1_devdata *dd)
 	del_timer_sync(&dev->mem_timer);
 	verbs_txreq_exit(dev);
 
+	mutex_lock(&cntr_names_lock);
 	kfree(dev_cntr_names);
 	kfree(port_cntr_names);
+	dev_cntr_names = NULL;
+	port_cntr_names = NULL;
 	cntr_names_initialized = 0;
+	mutex_unlock(&cntr_names_lock);
 }
 
 void hfi1_cnp_rcv(struct hfi1_packet *packet)