diff mbox series

iscsi target: fix cmd abort fabric stop race

Message ID 1605318378-9269-1-git-send-email-michael.christie@oracle.com (mailing list archive)
State Accepted
Commit f36199355c64a39fe82cfddc7623d827c7e050da
Headers show
Series iscsi target: fix cmd abort fabric stop race | expand

Commit Message

Mike Christie Nov. 14, 2020, 1:46 a.m. UTC
Maurizio Lombardi <mlombard@redhat.com> found a race where the abort
and cmd stop paths can race as follows:

1. thread1 runs iscsit_release_commands_from_conn and sets
CMD_T_FABRIC_STOP.
2. thread2 runs iscsit_aborted_task and then does __iscsit_free_cmd. It
then returns from the aborted_task callout and we finish
target_handle_abort and do:

target_handle_abort -> transport_cmd_check_stop_to_fabric ->
lio_check_stop_free -> target_put_sess_cmd

The cmd is now freed.
3. thread1 now finishes iscsit_release_commands_from_conn and runs
iscsit_free_cmd while accessing a command we just released.

In __target_check_io_state we check for CMD_T_FABRIC_STOP and set the
CMD_T_ABORTED if the driver is not cleaning up the cmd because of
a session shutdown. However, iscsit_release_commands_from_conn only
sets the CMD_T_FABRIC_STOP and does not check to see if the abort path
has claimed completion ownership of the command.

This adds a check in iscsit_release_commands_from_conn so only the
abort or fabric stop path cleanup the command.

Signed-off-by: Mike Christie <michael.christie@oracle.com>

---
 drivers/target/iscsi/iscsi_target.c | 17 +++++++++++++----
 1 file changed, 13 insertions(+), 4 deletions(-)

Comments

Maurizio Lombardi Nov. 16, 2020, 1:38 p.m. UTC | #1
Dne 14. 11. 20 v 2:46 Mike Christie napsal(a):
> Maurizio Lombardi <mlombard@redhat.com> found a race where the abort
> and cmd stop paths can race as follows:
> 
> 1. thread1 runs iscsit_release_commands_from_conn and sets
> CMD_T_FABRIC_STOP.
> 2. thread2 runs iscsit_aborted_task and then does __iscsit_free_cmd. It
> then returns from the aborted_task callout and we finish
> target_handle_abort and do:
> 
> target_handle_abort -> transport_cmd_check_stop_to_fabric ->
> lio_check_stop_free -> target_put_sess_cmd
> 
> The cmd is now freed.
> 3. thread1 now finishes iscsit_release_commands_from_conn and runs
> iscsit_free_cmd while accessing a command we just released.
> 
> In __target_check_io_state we check for CMD_T_FABRIC_STOP and set the
> CMD_T_ABORTED if the driver is not cleaning up the cmd because of
> a session shutdown. However, iscsit_release_commands_from_conn only
> sets the CMD_T_FABRIC_STOP and does not check to see if the abort path
> has claimed completion ownership of the command.
> 
> This adds a check in iscsit_release_commands_from_conn so only the
> abort or fabric stop path cleanup the command.
> 
> Signed-off-by: Mike Christie <michael.christie@oracle.com>
> 
> ---
>  drivers/target/iscsi/iscsi_target.c | 17 +++++++++++++----
>  1 file changed, 13 insertions(+), 4 deletions(-)
> 
> diff --git a/drivers/target/iscsi/iscsi_target.c b/drivers/target/iscsi/iscsi_target.c
> index f77e5ee..518fac4 100644
> --- a/drivers/target/iscsi/iscsi_target.c
> +++ b/drivers/target/iscsi/iscsi_target.c
> @@ -483,8 +483,7 @@ int iscsit_queue_rsp(struct iscsi_conn *conn, struct iscsi_cmd *cmd)
>  void iscsit_aborted_task(struct iscsi_conn *conn, struct iscsi_cmd *cmd)
>  {
>  	spin_lock_bh(&conn->cmd_lock);
> -	if (!list_empty(&cmd->i_conn_node) &&
> -	    !(cmd->se_cmd.transport_state & CMD_T_FABRIC_STOP))
> +	if (!list_empty(&cmd->i_conn_node))
>  		list_del_init(&cmd->i_conn_node);
>  	spin_unlock_bh(&conn->cmd_lock);
>  
> @@ -4083,12 +4082,22 @@ static void iscsit_release_commands_from_conn(struct iscsi_conn *conn)
>  	spin_lock_bh(&conn->cmd_lock);
>  	list_splice_init(&conn->conn_cmd_list, &tmp_list);
>  
> -	list_for_each_entry(cmd, &tmp_list, i_conn_node) {
> +	list_for_each_entry_safe(cmd, cmd_tmp, &tmp_list, i_conn_node) {
>  		struct se_cmd *se_cmd = &cmd->se_cmd;
>  
>  		if (se_cmd->se_tfo != NULL) {
>  			spin_lock_irq(&se_cmd->t_state_lock);
> -			se_cmd->transport_state |= CMD_T_FABRIC_STOP;
> +			if (se_cmd->transport_state & CMD_T_ABORTED) {
> +				/*
> +				 * LIO's abort path owns the cleanup for this,
> +				 * so put it back on the list and let
> +				 * aborted_task handle it.
> +				 */
> +				list_move_tail(&cmd->i_conn_node,
> +					       &conn->conn_cmd_list);
> +			} else {
> +				se_cmd->transport_state |= CMD_T_FABRIC_STOP;
> +			}
>  			spin_unlock_irq(&se_cmd->t_state_lock);
>  		}
>  	}
> 

Reviewed-by: Maurizio Lombardi <mlombard@redhat.com>
Martin K. Petersen Nov. 17, 2020, 6:06 a.m. UTC | #2
On Fri, 13 Nov 2020 19:46:18 -0600, Mike Christie wrote:

> Maurizio Lombardi <mlombard@redhat.com> found a race where the abort
> and cmd stop paths can race as follows:
> 
> 1. thread1 runs iscsit_release_commands_from_conn and sets
> CMD_T_FABRIC_STOP.
> 2. thread2 runs iscsit_aborted_task and then does __iscsit_free_cmd. It
> then returns from the aborted_task callout and we finish
> target_handle_abort and do:
> 
> [...]

Applied to 5.10/scsi-fixes, thanks!

[1/1] scsi: target: iscsi: Fix cmd abort fabric stop race
      https://git.kernel.org/mkp/scsi/c/f36199355c64
diff mbox series

Patch

diff --git a/drivers/target/iscsi/iscsi_target.c b/drivers/target/iscsi/iscsi_target.c
index f77e5ee..518fac4 100644
--- a/drivers/target/iscsi/iscsi_target.c
+++ b/drivers/target/iscsi/iscsi_target.c
@@ -483,8 +483,7 @@  int iscsit_queue_rsp(struct iscsi_conn *conn, struct iscsi_cmd *cmd)
 void iscsit_aborted_task(struct iscsi_conn *conn, struct iscsi_cmd *cmd)
 {
 	spin_lock_bh(&conn->cmd_lock);
-	if (!list_empty(&cmd->i_conn_node) &&
-	    !(cmd->se_cmd.transport_state & CMD_T_FABRIC_STOP))
+	if (!list_empty(&cmd->i_conn_node))
 		list_del_init(&cmd->i_conn_node);
 	spin_unlock_bh(&conn->cmd_lock);
 
@@ -4083,12 +4082,22 @@  static void iscsit_release_commands_from_conn(struct iscsi_conn *conn)
 	spin_lock_bh(&conn->cmd_lock);
 	list_splice_init(&conn->conn_cmd_list, &tmp_list);
 
-	list_for_each_entry(cmd, &tmp_list, i_conn_node) {
+	list_for_each_entry_safe(cmd, cmd_tmp, &tmp_list, i_conn_node) {
 		struct se_cmd *se_cmd = &cmd->se_cmd;
 
 		if (se_cmd->se_tfo != NULL) {
 			spin_lock_irq(&se_cmd->t_state_lock);
-			se_cmd->transport_state |= CMD_T_FABRIC_STOP;
+			if (se_cmd->transport_state & CMD_T_ABORTED) {
+				/*
+				 * LIO's abort path owns the cleanup for this,
+				 * so put it back on the list and let
+				 * aborted_task handle it.
+				 */
+				list_move_tail(&cmd->i_conn_node,
+					       &conn->conn_cmd_list);
+			} else {
+				se_cmd->transport_state |= CMD_T_FABRIC_STOP;
+			}
 			spin_unlock_irq(&se_cmd->t_state_lock);
 		}
 	}