Userspace could still be accessing the ring and might even still have commands it is trying to complete (if cmd_time_out != 0, so the kernel has dropped the cmd/ref count and made the device deletable) when a deletion is executed. This moves the userspace deletion to the beginning of the function so we know once it is done userspace is not accessing the kernel. Signed-off-by: Mike Christie <mchristi@xxxxxxxxxx> --- drivers/target/target_core_user.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/drivers/target/target_core_user.c b/drivers/target/target_core_user.c index 7d1da43..db129e9 100644 --- a/drivers/target/target_core_user.c +++ b/drivers/target/target_core_user.c @@ -1653,6 +1653,8 @@ static void tcmu_destroy_device(struct se_device *dev) bool all_expired = true; int i; + tcmu_netlink_event(udev, TCMU_CMD_REMOVED_DEVICE, 0, NULL); + del_timer_sync(&udev->timeout); mutex_lock(&root_udev_mutex); @@ -1680,8 +1682,6 @@ static void tcmu_destroy_device(struct se_device *dev) tcmu_blocks_release(udev); - tcmu_netlink_event(udev, TCMU_CMD_REMOVED_DEVICE, 0, NULL); - uio_unregister_device(&udev->uio_info); /* release ref from configure */ -- 2.7.2 -- To unsubscribe from this list: send the line "unsubscribe target-devel" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
