From: Nicholas Bellinger <nab@xxxxxxxxxxxxxxx>
This patch removes the original usage of __transport_execute_tasks() ahead
of every transport_get_cmd_from_queue() call in transport_processing_thread().
This helps reduce se_device->execute_task_lock contention between qla2xxx wq
with target_submit_cmd() for READs and transport_processing_thread()
context servicing WRITEs with full payloads for I/O submission.
It also adds a __transport_execute_tasks() to kick the task queue again
without a *se_cmd descriptor with existing queue full logic, but this may
end up not being necessary.
Cc: Christoph Hellwig <hch@xxxxxx>
Cc: Roland Dreier <roland@xxxxxxxxxxxxxxx>
Cc: Joern Engel <joern@xxxxxxxxx>
Signed-off-by: Nicholas Bellinger <nab@xxxxxxxxxxxxxxx>
---
drivers/target/target_core_transport.c | 4 ++--
1 files changed, 2 insertions(+), 2 deletions(-)
diff --git a/drivers/target/target_core_transport.c b/drivers/target/target_core_transport.c
index 7f867d2..823bfd2 100644
--- a/drivers/target/target_core_transport.c
+++ b/drivers/target/target_core_transport.c
@@ -4718,8 +4718,6 @@ static int transport_processing_thread(void *param)
goto out;
get_cmd:
- __transport_execute_tasks(dev);
-
cmd = transport_get_cmd_from_queue(&dev->dev_queue_obj);
if (!cmd)
continue;
@@ -4753,9 +4751,11 @@ get_cmd:
break;
case TRANSPORT_COMPLETE_QF_WP:
transport_write_pending_qf(cmd);
+ __transport_execute_tasks(dev, NULL);
break;
case TRANSPORT_COMPLETE_QF_OK:
transport_complete_qf(cmd);
+ __transport_execute_tasks(dev, NULL);
break;
default:
pr_err("Unknown t_state: %d for ITT: 0x%08x "
--
1.7.2.5
--
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
