diff options
| author | Mike Christie <michael.christie@oracle.com> | 2021-05-25 21:18:10 +0300 |
|---|---|---|
| committer | Martin K. Petersen <martin.petersen@oracle.com> | 2021-06-02 08:28:22 +0300 |
| commit | 99b0603313eecb9b0b99fd2e60bae1e03e6b038b (patch) | |
| tree | 5fcba3b60c010b8b7d5f7757b35a2c8d2505465e /drivers/scsi/libiscsi.c | |
| parent | 7ce9fc5ecde0d8bd64c29baee6c5e3ce7074ec9a (diff) | |
| download | linux-99b0603313eecb9b0b99fd2e60bae1e03e6b038b.tar.xz | |
scsi: iscsi: Hold task ref during TMF timeout handling
For aborts, qedi needs to cleanup the FW then send the TMF from a worker
thread. While it's doing these the cmd could complete normally and the TMF
could time out. libiscsi would then complete the iscsi_task which will call
into the driver to cleanup the driver level resources while it still might
be accessing them for the cleanup/abort.
This has iscsi_eh_abort keep the iscsi_task ref if the TMF times out, so
qedi does not have to worry about if the task is being freed while in use
and does not need to get its own ref.
Link: https://lore.kernel.org/r/20210525181821.7617-18-michael.christie@oracle.com
Reviewed-by: Lee Duncan <lduncan@suse.com>
Signed-off-by: Mike Christie <michael.christie@oracle.com>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
Diffstat (limited to 'drivers/scsi/libiscsi.c')
| -rw-r--r-- | drivers/scsi/libiscsi.c | 15 |
1 files changed, 14 insertions, 1 deletions
diff --git a/drivers/scsi/libiscsi.c b/drivers/scsi/libiscsi.c index 8222db4f8fef..e57d6355e7c7 100644 --- a/drivers/scsi/libiscsi.c +++ b/drivers/scsi/libiscsi.c @@ -573,6 +573,11 @@ static bool cleanup_queued_task(struct iscsi_task *task) __iscsi_put_task(task); } + if (conn->session->running_aborted_task == task) { + conn->session->running_aborted_task = NULL; + __iscsi_put_task(task); + } + if (conn->task == task) { conn->task = NULL; __iscsi_put_task(task); @@ -2334,6 +2339,7 @@ int iscsi_eh_abort(struct scsi_cmnd *sc) iscsi_start_tx(conn); goto success_unlocked; case TMF_TIMEDOUT: + session->running_aborted_task = task; spin_unlock_bh(&session->frwd_lock); iscsi_conn_failure(conn, ISCSI_ERR_SCSI_EH_SESSION_RST); goto failed_unlocked; @@ -2367,7 +2373,14 @@ failed: failed_unlocked: ISCSI_DBG_EH(session, "abort failed [sc %p itt 0x%x]\n", sc, task ? task->itt : 0); - iscsi_put_task(task); + /* + * The driver might be accessing the task so hold the ref. The conn + * stop cleanup will drop the ref after ep_disconnect so we know the + * driver's no longer touching the task. + */ + if (!session->running_aborted_task) + iscsi_put_task(task); + iscsi_put_conn(conn->cls_conn); mutex_unlock(&session->eh_mutex); return FAILED; |