svn+ssh://svn.code.sf.net/p/scst/svn/trunk
........
r7247 | bvassche | 2017-10-06 08:08:00 -0700 (Fri, 06 Oct 2017) | 4 lines
fcst: Fix in-tree build
See also r7243.
........
r7248 | bvassche | 2017-10-09 11:10:48 -0700 (Mon, 09 Oct 2017) | 4 lines
scst_vdisk: Fix a bug in the Linux kernel v4.13 port
See also trunk r7240.
........
r7249 | bvassche | 2017-10-09 11:15:00 -0700 (Mon, 09 Oct 2017) | 4 lines
scst: Fix SCSI pass-through for kernel version v4.11 and later
See also trunk r7153.
........
r7250 | bvassche | 2017-10-09 12:58:15 -0700 (Mon, 09 Oct 2017) | 1 line
scst/include/backport.h: Follow-up for r7248
........
r7251 | vlnb | 2017-10-09 19:06:13 -0700 (Mon, 09 Oct 2017) | 5 lines
qla2x00t: update FW version information as suggested by Jens-U. Mozdzen <jmozdzen@nde.ag>
See https://sourceforge.net/p/scst/mailman/message/36068827/
........
git-svn-id: http://svn.code.sf.net/p/scst/svn/branches/3.3.x@7252 d57e44dd-8a1f-0410-8b47-8ef2f437770f
There is no need for a detect() method - code that is present in such
a method can be moved to module_init(). Hence declare this method
obsolete. Leave the detect pointer member in the target template for
now for out-of-tree SCST drivers.
Signed-off-by: Bart Van Assche <bart.vanassche@sandisk.com>
with some improvements
git-svn-id: http://svn.code.sf.net/p/scst/svn/trunk@6113 d57e44dd-8a1f-0410-8b47-8ef2f437770f
Retry sending XFER_RDY, data and response frames if the network
driver reports that sending failed (-ENOMEM) instead of reporting
a kernel warning (WARN_ON(1)). If sending XFER_RDY or data frames
failed for another reason, report this to the initiator as a
write error (ASC = 03; ASCQ = 00 which stands for PERIPHERAL
DEVICE WRITE FAULT). If sending a response frame failed with
another error code than -ENOMEM, do not send a response.
git-svn-id: http://svn.code.sf.net/p/scst/svn/trunk@5607 d57e44dd-8a1f-0410-8b47-8ef2f437770f
The biggest difference is that libfc no longer allocates a sequence
and exchange for incoming requests, and that eliminates a parameter
on the receive function. This was done to help performance and
might help on short reads sometime in the future.
This also fixes an error in ft_send_read_data() that exceeds the
skb's SG list length limit on long reads when the NIC
supports large send offload.
I also made other minor changes like adding debugging.
git-svn-id: http://svn.code.sf.net/p/scst/svn/trunk@2823 d57e44dd-8a1f-0410-8b47-8ef2f437770f
Add a method for template get_initiator_port_transport_id()
to allow persistent reservations to work.
Initialize struct ft_sess more fully before scst_register_session()
so that ft_get_transport_id() can use it.
git-svn-id: http://svn.code.sf.net/p/scst/svn/trunk@1759 d57e44dd-8a1f-0410-8b47-8ef2f437770f
The target template max_hw_pending_time was zero, causing
premature timeouts and a fast retry in scst_hw_pending_work_fn().
Set it to 20 seconds, like qla2x00t does, for now.
It could be the wrong thing for tape, but should be
plenty long enough for disks.
Signed-off-by: Joe Eykholt <jeykholt@cisco.com>
git-svn-id: http://svn.code.sf.net/p/scst/svn/trunk@1522 d57e44dd-8a1f-0410-8b47-8ef2f437770f
that accepts FCP requests from libfc HBAs running Fibre Channel
over Ethernet (FCoE) and passes them to SCST.
Signed-off-by: Joe Eykholt <jeykholt@cisco.com>
git-svn-id: http://svn.code.sf.net/p/scst/svn/trunk@1514 d57e44dd-8a1f-0410-8b47-8ef2f437770f
