At the SCSI transport level, there is no distinction between
user data and protection information. Thus, iscsi header field
"expected data transfer length" should include protection
information.
Patch #1 introduces scsi helpers scsi_transfer_length (compute
wire transfer byte count) and scsi_prot_len (compute protection
information byte count).
Patch #2 modifies iscsi initiator to set correct wire transfer length
in iscsi header data_length field (and modifies iser accordingly).
Patch #3 modifies target core to expect protection information included
in the wire transfer length (and modifies loopback transport to do so).
I have 2 patches that convert lpfc/qla2xxx drivers to use scsi helpers
but these are completely untested at the moment. Once we get this set
to land upstream, I can queue them up as RFCs.
Changes from v0:
- Introduce scsi helpers to compute correct transfer length in the
presence of protection information (instead of having each transport
doing the same computation).
- Modify iscsi to set correct transfer length using scsi helpers
- Modify loopback transport to set correct transfer length using
scsi helpers
Sagi Grimberg (3):
scsi_cmnd: Introduce scsi_transfer_length helper
libiscsi, iser: Adjust data_length to include protection information
TARGET/sbc,loopback: Adjust command data length in case pi exists on
the wire
drivers/infiniband/ulp/iser/iser_initiator.c | 34 ++++++----------------
drivers/scsi/libiscsi.c | 18 ++++++------
drivers/target/loopback/tcm_loop.c | 15 ++++++++--
drivers/target/target_core_sbc.c | 15 ++++++++-
include/scsi/scsi_cmnd.h | 39 ++++++++++++++++++++++++++
5 files changed, 83 insertions(+), 38 deletions(-)
--
To unsubscribe from this list: send the line "unsubscribe linux-rdma" in
the body of a message to [email protected]
More majordomo info at http://vger.kernel.org/majordomo-info.html
