From: Nicholas Bellinger <nab@xxxxxxxxxxxxxxx>
Hi Roland and Madhu,
This series converts qla2xxx + tcm_qla2xxx to hold qla_hw_data->hardware_lock
when checking all shutdown related state bits for individual qla_tgt_cmd
descriptors, instead of trying to use fancy barrier() + atomic operations that
ends up being problematic and error prone for a number of reasons.
This will likely have some effect on overall performance, but makes the
shutdown logic alot more simple than the original approach. So far this has
been running overnight with active I/O shutdown of WRITEs + READs and has not
run into problems with full kernel debugging enabled.
Please have a look and comment as necessary.
Thanks,
--nab
Nicholas Bellinger (2):
qla2xxx: Run qla_tgt_wait_for_cmds state checks with hardware_lock
held
tcm_qla2xxx: Run ->check_stop_free and ->release_cmd with
hardware_lock held
drivers/scsi/qla2xxx/qla_target.c | 52 +++++++----------
drivers/scsi/qla2xxx/qla_target.h | 1 +
drivers/target/tcm_qla2xxx/tcm_qla2xxx_fabric.c | 68 +++++++++++++----------
3 files changed, 62 insertions(+), 59 deletions(-)
--
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
