Change the data type of iscsi_conn.read_iov into struct kvec * and
switch from iscsi_conn.read_size to iscsi_conn.read_msg.count for
kernel version 3.19 and later. Cast the return type of
get_zeroed_page() to void * since with this patch the read_iov
data type depends on the kernel version. Remove the (void __force
__user *) casts when assigning to iov_base since for a kvec iov_base
is a kernel pointer instead of a user space pointer. Switch to
iov_iter_kvec() to initialize conn->read_msg.msg_iter.
Signed-off-by: Bart Van Assche <bart.vanassche@sandisk.com>
Rework the iSCSI target driver code such that it becomes easier
to port this code to kernel 3.19.
Signed-off-by: Bart Van Assche <bart.vanassche@sandisk.com>
The put_page_callback patch only improves performance on single-socket
systems but not on systems with multiple CPU sockets. Hence do not
print a warning if the put_page_callback patch has not been applied.
Signed-off-by: Bart Van Assche <bart.vanassche@sandisk.com>
The T10-PI patch (r6162) increases the time needed to run mkfs.ext4
over iSCSI from less than a second to several minutes. This is a
regression so revert the T10-PI patch until there is sufficient time
to find the root cause of this regression and to fix this regression.
git-svn-id: http://svn.code.sf.net/p/scst/svn/trunk@6171 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
Avoid that checkpatch reports the following warning:
WARNING: type 'long long unsigned' should be specified in 'unsigned long long' order.
git-svn-id: http://svn.code.sf.net/p/scst/svn/trunk@5885 d57e44dd-8a1f-0410-8b47-8ef2f437770f
For SCST devices with a short name determining which SCST command
threads serve a given LUN requires an (expensive) walk of the
process table. For SCST devices with a long name it is not possible
to determine unambiguously which command threads serve a given
LUN. Hence add a thread_pid sysfs attribute that makes it easy to
figure out which command threads serve a LUN.
Signed-off-by: Bart Van Assche <bvanassche@acm.org>
git-svn-id: http://svn.code.sf.net/p/scst/svn/trunk@5851 d57e44dd-8a1f-0410-8b47-8ef2f437770f
Instead of adding sg[idx].offset to addr and immediately subtracting
sg[idx].offset again, leave out both arithmetic operations.
Signed-off-by: Bart Van Assche <bvanassche@acm.org>
git-svn-id: http://svn.code.sf.net/p/scst/svn/trunk@5735 d57e44dd-8a1f-0410-8b47-8ef2f437770f
Since scst.h contains a backport of the definition of set_cpus_allowed_ptr(),
the set_cpus_allowed() invocations are superfluous. Hence remove these.
Signed-off-by: Bart Van Assche <bvanassche@acm.org>
git-svn-id: http://svn.code.sf.net/p/scst/svn/trunk@5734 d57e44dd-8a1f-0410-8b47-8ef2f437770f
Avoid that the checkpatch tool included in Linux kernel v3.16 reports the
following warning:
macros should not use a trailing semicolon
This patch does not change any functionality.
git-svn-id: http://svn.code.sf.net/p/scst/svn/trunk@5724 d57e44dd-8a1f-0410-8b47-8ef2f437770f
Start at the proper offset in the receive buffer if sg[0].offset != 0.
Return the proper data to the initiator if sg[0].offset != 0.
This patch reworks trunk r5281, "iscsi-scst: fix offset
calculation", February 13, 2014.
Signed-off-by: Bart Van Assche <bvanassche@acm.org>
git-svn-id: http://svn.code.sf.net/p/scst/svn/trunk@5683 d57e44dd-8a1f-0410-8b47-8ef2f437770f
With CONFIG_BUG=y both BUG() and BUG_ON(1) halt the system. However,
with CONFIG_BUG=n BUG() halts the system but BUG_ON(1) not. To avoid
such subtleties, change BUG_ON(1) into BUG().
See also patch Josh Triplett, "bug: Make BUG() always stop the machine",
7 April 2014 (commit ID a4b5d580e07875f9be29f62a57c67fbbdbb40ba2).
Signed-off-by: Bart Van Assche <bvanassche@acm.org>
git-svn-id: http://svn.code.sf.net/p/scst/svn/trunk@5618 d57e44dd-8a1f-0410-8b47-8ef2f437770f
The cpus_*() functions were deprecated via patch "cpumask:
introduce new API, without changing anything" (November 2008,
commit ID 2d3854a37e8b). Hence switch from the cpus_*() API to
the cpumask_*() API.
This patch has the intended side effect of not adding the "[key]"
property to cpumask sysfs attributes that contain the default
cpumask. The current code namely reads uninitialized bits on
systems where nr_cpu_ids < NR_CPUS because cpus_equal() compares
more bits than those that were set by cpumask_copy().
Signed-off-by: Bart Van Assche <bvanassche@acm.org>
git-svn-id: http://svn.code.sf.net/p/scst/svn/trunk@5596 d57e44dd-8a1f-0410-8b47-8ef2f437770f
The lockdep_assert_held() macro is a convenient debugging tool.
However, it is inconvenient to surround each invocation of that
macro by an #ifdef/#endif pair. Hence make it easier to use this
macro with older kernel versions.
Signed-off-by: Bart Van Assche <bvanassche@acm.org>
git-svn-id: http://svn.code.sf.net/p/scst/svn/trunk@5573 d57e44dd-8a1f-0410-8b47-8ef2f437770f
Assign one major number per thread pool instead of as many major
numbers as there are threads in a thread pool. Do not increment
'major' if thread pool allocation fails. Micro-optimize
iscsi_threads_pool_get() by eliminating the assignment to 'fn'
and the write via snprintf() into name[].
Signed-off-by: Bart Van Assche <bvanassche@acm.org>
git-svn-id: http://svn.code.sf.net/p/scst/svn/trunk@5520 d57e44dd-8a1f-0410-8b47-8ef2f437770f
Fix the following checkpatch 3.14 warning:
Unnecessary parentheses - maybe == should be = ?
git-svn-id: http://svn.code.sf.net/p/scst/svn/trunk@5435 d57e44dd-8a1f-0410-8b47-8ef2f437770f
May be not completed, because this kernel has too many internal issues
reported after enabling "Kernel hacking" config options
git-svn-id: http://svn.code.sf.net/p/scst/svn/trunk@5163 d57e44dd-8a1f-0410-8b47-8ef2f437770f
