From: Mike Christie <michael.chris...@oracle.com>

[ Upstream commit f36199355c64a39fe82cfddc7623d827c7e050da ]

Maurizio 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.

Link: 
https://lore.kernel.org/r/1605318378-9269-1-git-send-email-michael.chris...@oracle.com
Reported-by: Maurizio Lombardi <mlomb...@redhat.com>
Reviewed-by: Maurizio Lombardi <mlomb...@redhat.com>
Signed-off-by: Mike Christie <michael.chris...@oracle.com>
Signed-off-by: Martin K. Petersen <martin.peter...@oracle.com>
Signed-off-by: Sasha Levin <sas...@kernel.org>
---
 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 cbb4414edd71b..c48aca1360c89 100644
--- a/drivers/target/iscsi/iscsi_target.c
+++ b/drivers/target/iscsi/iscsi_target.c
@@ -493,8 +493,7 @@ static void iscsit_aborted_task(struct iscsi_conn *conn, 
struct iscsi_cmd *cmd)
        bool scsi_cmd = (cmd->iscsi_opcode == ISCSI_OP_SCSI_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);
 
@@ -4228,12 +4227,22 @@ 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);
                }
        }
-- 
2.27.0



Reply via email to