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

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

From: Parav Pandit <pa...@mellanox.com>

commit 2fb4f4eadd180a50112618dd9c5fef7fc50d4f08 upstream.

During IB device registration process, if query_device() fails or if
ib_core fails to registers sysfs entries, rdma cgroup cleanup is
skipped.

Cc: <sta...@vger.kernel.org> # v4.2+
Fixes: 4be3a4fa51f4 ("IB/core: Fix kernel crash during fail to initialize 
device")
Reviewed-by: Daniel Jurgens <dani...@mellanox.com>
Signed-off-by: Parav Pandit <pa...@mellanox.com>
Signed-off-by: Leon Romanovsky <l...@kernel.org>
Signed-off-by: Jason Gunthorpe <j...@mellanox.com>
Signed-off-by: Greg Kroah-Hartman <gre...@linuxfoundation.org>

---
 drivers/infiniband/core/device.c |    6 ++++--
 1 file changed, 4 insertions(+), 2 deletions(-)

--- a/drivers/infiniband/core/device.c
+++ b/drivers/infiniband/core/device.c
@@ -534,14 +534,14 @@ int ib_register_device(struct ib_device
        ret = device->query_device(device, &device->attrs, &uhw);
        if (ret) {
                pr_warn("Couldn't query the device attributes\n");
-               goto cache_cleanup;
+               goto cg_cleanup;
        }
 
        ret = ib_device_register_sysfs(device, port_callback);
        if (ret) {
                pr_warn("Couldn't register device %s with driver model\n",
                        device->name);
-               goto cache_cleanup;
+               goto cg_cleanup;
        }
 
        device->reg_state = IB_DEV_REGISTERED;
@@ -557,6 +557,8 @@ int ib_register_device(struct ib_device
        mutex_unlock(&device_mutex);
        return 0;
 
+cg_cleanup:
+       ib_device_unregister_rdmacg(device);
 cache_cleanup:
        ib_cache_cleanup_one(device);
        ib_cache_release_one(device);


Reply via email to