After having changed the queue pair state into "error", queue an additional
work request instead of waiting for the last WQE event.
git-svn-id: http://svn.code.sf.net/p/scst/svn/trunk@5357 d57e44dd-8a1f-0410-8b47-8ef2f437770f
This is necessary to avoid that a login gets rejected due to reusing a
queue pair number that has not yet been freed by the target side.
git-svn-id: http://svn.code.sf.net/p/scst/svn/trunk@5356 d57e44dd-8a1f-0410-8b47-8ef2f437770f
Store initiator and target port ID's once per nexus instead of in each
channel data structure.
git-svn-id: http://svn.code.sf.net/p/scst/svn/trunk@5354 d57e44dd-8a1f-0410-8b47-8ef2f437770f
Although harmless for SCSI commands with SIMPLE ordering, avoid that commands
received before RTU can get reordered.
git-svn-id: http://svn.code.sf.net/p/scst/svn/trunk@5349 d57e44dd-8a1f-0410-8b47-8ef2f437770f
The mlx4_core driver generates a GID change event after a port has been
changed from IB into Ethernet mode. Avoid that this causes the following
error message to appear in the system log:
ib_srpt: ***ERROR***: received unrecognized IB event 18
git-svn-id: http://svn.code.sf.net/p/scst/svn/trunk@5346 d57e44dd-8a1f-0410-8b47-8ef2f437770f
All ioctx->state manipulations are serialized per command so it is
not necessary to use locking to protect these manipulations.
git-svn-id: http://svn.code.sf.net/p/scst/svn/trunk@5345 d57e44dd-8a1f-0410-8b47-8ef2f437770f
Make the behavior of SCST sysfs attributes more clear by adding
examples in scst/README of code for reading and writing these attributes.
Signed-off-by: Bart Van Assche <bvanassche@acm.org>
git-svn-id: http://svn.code.sf.net/p/scst/svn/trunk@5340 d57e44dd-8a1f-0410-8b47-8ef2f437770f
Ensure that COMPARE AND WRITE is executed atomically by serializing
all COMPARE AND WRITE commands per device (SCST_SERIALIZED).
Signed-off-by: Bart Van Assche <bvanassche@acm.org>
git-svn-id: http://svn.code.sf.net/p/scst/svn/trunk@5334 d57e44dd-8a1f-0410-8b47-8ef2f437770f
There is already an implementation of the VERIFY command for vdisk_fileio
devices. Add an implementation for vdisk_blockio devices.
Signed-off-by: Bart Van Assche <bvanassche@acm.org>
git-svn-id: http://svn.code.sf.net/p/scst/svn/trunk@5333 d57e44dd-8a1f-0410-8b47-8ef2f437770f
In r5316 the default vendor name for vdisk_blockio devices was changed
into "SCST_FIO". Change this back into "SCST_BIO".
Signed-off-by: Bart Van Assche <bvanassche@acm.org>
git-svn-id: http://svn.code.sf.net/p/scst/svn/trunk@5332 d57e44dd-8a1f-0410-8b47-8ef2f437770f
The "struct vdisk_cmd_params *p" parameter is neither used by vdisk_fsync(),
vdisk_fsync_blockio() nor by vdisk_fsync_fileio() so remove it.
Signed-off-by: Bart Van Assche <bvanassche@acm.org>
git-svn-id: http://svn.code.sf.net/p/scst/svn/trunk@5331 d57e44dd-8a1f-0410-8b47-8ef2f437770f
When iscsi-scstd exits for some reason, release on character device ctr_name is called.
The release function of this character device calls target_del_all().
In order to avoid a scenario where isert-scst is being removed at the same time
as iscsi-scstd is closing ctr_name device, increment reference count in portal creation
so that the module will not disappear before target_del_all() finishes using it.
This scenario is only relevant for isert-scst, as iscsi-scst will have positive reference count
at least untill target_del_all() finishes due to __fput() being called only after the release
method of the character device has returned.
Signed-off-by: Yan Burman <yanb@mellanox.com>
git-svn-id: http://svn.code.sf.net/p/scst/svn/branches/iser@5327 d57e44dd-8a1f-0410-8b47-8ef2f437770f
Keep the default size of vdisk_nullio devices at VDISK_NULLIO_SIZE. Add
a sysfs attribute 'size' which is the size of a vdisk device in
bytes. Make the size of vdisk_nullio devices configurable. Accept
"size" and "size_mb" as creation parameters for vdisk_nullio
devices. Generate a CAPACITY DATA HAS CHANGED unit attention after size
changes. Refuse any attempt to change the size into a number that is
not a multiple of the block size.
Signed-off-by: Bart Van Assche <bvanassche@acm.org>
git-svn-id: http://svn.code.sf.net/p/scst/svn/trunk@5322 d57e44dd-8a1f-0410-8b47-8ef2f437770f
svn+ssh://yanb123@svn.code.sf.net/p/scst/svn/trunk
........
r5246 | vlnb | 2014-01-29 05:30:24 +0200 (Wed, 29 Jan 2014) | 3 lines
Put CDB control byte parsing in one place
........
r5247 | vlnb | 2014-01-29 06:16:58 +0200 (Wed, 29 Jan 2014) | 3 lines
Better version of the previous patch
........
r5248 | vlnb | 2014-01-30 03:40:48 +0200 (Thu, 30 Jan 2014) | 7 lines
[PATCH 1/2] scst_sysfs: Make it easier to add new target sysfs attributes
This patch does not change any functionality.
Signed-off-by: Bart Van Assche <bvanassche@acm.org>
........
r5249 | vlnb | 2014-01-30 03:41:54 +0200 (Thu, 30 Jan 2014) | 10 lines
[PATCH 2/2] scst_sysfs: Add I/O statistics per target
Although it is possible to obtain these statistics by iterating over
all sessions and by computing the sum of the per-target statistics,
make per-target statistics directly available such that these can be
retrieved easily.
Signed-off-by: Bart Van Assche <bvanassche@acm.org>
........
r5250 | vlnb | 2014-01-30 04:32:44 +0200 (Thu, 30 Jan 2014) | 3 lines
Update for 3.13 kernels
........
r5251 | bvassche | 2014-01-30 11:16:27 +0200 (Thu, 30 Jan 2014) | 1 line
nightly build: Add kernel 3.13 build infrastructure
........
r5252 | bvassche | 2014-01-30 11:30:18 +0200 (Thu, 30 Jan 2014) | 1 line
scripts/kernel-functions: Add a bug fix for the kernel 3.13 series that is not yet present in the kernel 3.13 stable series
........
r5253 | bvassche | 2014-01-30 11:31:32 +0200 (Thu, 30 Jan 2014) | 1 line
nightly build: Add kernel version 3.13.1
........
r5254 | vlnb | 2014-01-31 04:32:02 +0200 (Fri, 31 Jan 2014) | 10 lines
scst_pres: Simplify PR locking
Since the time during which a PR read or write lock is held is short,
use a mutex to implement PR read and write locking. So although this
patch excludes multiple simultaneous readers that shouldn't affect the
time needed to process a PR operation measurably.
Signed-off-by: Bart Van Assche <bvanassche@acm.org>
........
r5255 | vlnb | 2014-01-31 04:33:11 +0200 (Fri, 31 Jan 2014) | 5 lines
scst_vdisk: Check that "filename" is specified at most once
Signed-off-by: Bart Van Assche <bvanassche@acm.org>
........
r5256 | vlnb | 2014-01-31 04:35:20 +0200 (Fri, 31 Jan 2014) | 5 lines
scst_vdisk: Sort "add_device_parameters" alphabetically
Signed-off-by: Bart Van Assche <bvanassche@acm.org>
........
r5260 | bvassche | 2014-02-03 11:03:03 +0200 (Mon, 03 Feb 2014) | 1 line
scripts/list-source-files: Handle Mercurial subdirectories properly
........
r5264 | bvassche | 2014-02-06 14:46:51 +0200 (Thu, 06 Feb 2014) | 9 lines
scst_local: Fix a kernel oops for kernel versions < 2.6.37
Avoid that scst_local triggers "BUG: unable to handle kernel NULL
pointer dereference" on kernel versions before 2.6.37. This patch
fixes a regression introduced via patch "scst_local: Avoid
deadlock during module removal with kernel 3.6" (trunk r4566).
Reported-by: Sebastian Herbszt <herbszt@gmx.de>
........
r5266 | bvassche | 2014-02-06 15:30:06 +0200 (Thu, 06 Feb 2014) | 14 lines
Hush Coverity warning of scst_ws_push_single_write() uninitialized pointer
Coverity warns that sgv may be used uninitialized. The warning
applies to WRITE SAME commands with LBDATA == PBDATA == 0 (replicate
a single block of user data into the specified LBA range).
The warning appears to be spurious - when LBDATA == PBDATA == 0,
scst_ws_write_cmd_finished() will not use the uninitialized value
saved by scst_ws_push_single_write().
Move initialization of sgv earlier in the function to quiesce the warning.
Signed-off-by: Steven J. Magnani <steve@digidescorp.com>
........
r5267 | bvassche | 2014-02-06 15:38:28 +0200 (Thu, 06 Feb 2014) | 7 lines
qla2x00t: Re-sync help text with the code
The ql2xfdmienable module parameter defaults to 1, but the help text
claims it defaults to zero.
Signed-off-by: Steven J. Magnani <steve@digidescorp.com>
........
r5268 | bvassche | 2014-02-06 16:17:49 +0200 (Thu, 06 Feb 2014) | 6 lines
ib_srpt: Avoid that disabling a target triggers a race condition
Avoid that disabling a target triggers a race condition with
SRP relogin. At least in theory this race condition could result
in a kernel crash.
........
r5269 | bvassche | 2014-02-07 09:31:38 +0200 (Fri, 07 Feb 2014) | 9 lines
scst_sysfs: Fix a build failure on kernels 2.6.2[678]
The sysfs API is supported from kernel 2.6.26 on and uses the swap()
macro while the swap() macro was introduced in kernel 2.6.29. Hence
provide a definition of the swap() macro for kernels before 2.6.29.
Signed-off-by: Sebastian Herbszt <herbszt@gmx.de>
[bvanassche: Moved swap() definition a few lines down and added #ifndef/#endif]
........
r5270 | bvassche | 2014-02-07 09:45:15 +0200 (Fri, 07 Feb 2014) | 2 lines
regression tests: Run the 2.6.26..2.6.32 tests on the sysfs code instead of procfs
........
r5271 | bvassche | 2014-02-07 10:11:28 +0200 (Fri, 07 Feb 2014) | 1 line
nightly build: Update kernel versions
........
r5272 | bvassche | 2014-02-07 14:43:25 +0200 (Fri, 07 Feb 2014) | 16 lines
scst_user, rt: Wake command processing thread when needed
In a fully-preemptible realtime kernel (CONFIG_PREEMPT_RT_FULL=y),
SCSI commands from an initiator time out because the userland target
application is never woken to process them.
This is because in a fully-preemptible realtime kernel, soft-IRQ
(tasklet) execution always occurs in a ksoftirqd thread and
preempt_count is not manipulated on soft-IRQ processing entry/exit.
This makes in_interrupt() useless for determining whether soft-IRQ
processing is occurring; instead, in_serving_softirq() should be
used for that purpose.
Signed-off-by: Steven J. Magnani <steve@digidescorp.com>
[bvanassche: Elaborated source code comment]
........
r5273 | bvassche | 2014-02-07 14:46:39 +0200 (Fri, 07 Feb 2014) | 9 lines
scst_vdisk: Build fix for kernels 2.6.27..2.6.30
add_to_page_cache_lru and __lock_page_killable are exported since
kernel version 2.6.30. See also patch "Staging: pohmelfs: kconfig/makefile
and vfs changes" (commit 18bc0bbd162e3eb3e7ea2953c315ad4113a57164;
included in kernel v2.6.30).
Signed-off-by: Sebastian Herbszt <herbszt@gmx.de>
........
r5274 | vlnb | 2014-02-08 03:04:27 +0200 (Sat, 08 Feb 2014) | 9 lines
scst_user: Convert sgv_purge_interval to jiffies before use
The sgv_purge_interval from userland is passed down without conversion to
jiffies. Yet, if it is zero, the default value is (60 * HZ).
Convert to jiffies before passing down.
Signed-off-by: Steven J. Magnani <steve@digidescorp.com>
........
r5275 | vlnb | 2014-02-08 03:52:03 +0200 (Sat, 08 Feb 2014) | 19 lines
Fix spurious BUG when parse_type != SCST_USER_PARSE_STANDARD
Changeset 4224 introduced EXTRACHECKS for valid lba/data_len and state
at the end of the parsing phase of command processing.
However, the checks do not account for deferral of parsing to userland,
as occurs when SCST_USER_PARSE_CALL or SCST_USER_PARSE_EXCEPTION are specified.
In such cases the checks report errors on commands that userland has not yet
had an opportunity to parse.
NOTE: this includes a refactoring of the EXTRACHECKS to improve clarity.
The rework is not exactly equivalent to the original code, but does
conform to the comments describing the original code.
Specifically, the original code would not trap an illegal command state
unless there was also an illegal lba or data_len.
Signed-off-by: Steven J. Magnani <steve@digidescorp.com>
with some improvements
........
r5276 | bvassche | 2014-02-08 10:24:28 +0200 (Sat, 08 Feb 2014) | 1 line
scst: Build fix for kernel versions before 2.6.37
........
r5277 | bvassche | 2014-02-09 18:50:10 +0200 (Sun, 09 Feb 2014) | 1 line
scst_debug.h: Avoid that the sBUG() and sBUG_ON() definitions confuse the smatch static code checker
........
r5281 | vlnb | 2014-02-13 06:02:56 +0200 (Thu, 13 Feb 2014) | 8 lines
iscsi-scst: fix offset calculation
Fixed a subtle bug in iSCSI-SCST with incorrectly calculated offsets
for non-page aligned transfers. Originally discovered, investigated and
fix suggested by Кирилл Тюшев, then Shahar Salzman tested and proved it.
See http://sourceforge.net/mailarchive/message.php?msg_id=31924078
........
r5282 | vlnb | 2014-02-13 06:15:31 +0200 (Thu, 13 Feb 2014) | 3 lines
Web update
........
r5283 | bvassche | 2014-02-14 15:05:55 +0200 (Fri, 14 Feb 2014) | 7 lines
Makefiles: remove redundant 'depmod' invocations
Running 'make modules_install' already triggers invocation of depmod,
hence leave it out from those Makefiles that use 'make modules_install'.
Signed-off-by: Steven J. Magnani <steve@digidescorp.com>
........
r5284 | bvassche | 2014-02-14 15:48:54 +0200 (Fri, 14 Feb 2014) | 2 lines
Makefiles: Convert from "install" to "make modules_install"
........
r5285 | bvassche | 2014-02-14 16:46:11 +0200 (Fri, 14 Feb 2014) | 1 line
mvsas_tgt/Makefile: Remove trailing whitespace
........
r5286 | bvassche | 2014-02-14 17:52:10 +0200 (Fri, 14 Feb 2014) | 18 lines
Makefiles: calculate KVER properly
When deriving the kernel version (KVER) from KDIR, the file
$(KDIR)/include/config/kernel.release should be preferred over
'make kernelversion'.
For example, the Ubuntu 3.2.0-23-generic kernel has a kernel.release
file containing '3.2.0-23-generic', but 'make kernelversion' returns
3.2.14. Since the modules are stored under /lib/modules/3.2.0-23-generic,
the value in kernel.release is the correct one to use.
Also:
- Evaluate KVER only once
- All depmod commands must include KVER
Signed-off-by: Steven J. Magnani <steve@digidescorp.com>
[bvanassche: Split long lines / removed trailing whitespace]
........
r5287 | bvassche | 2014-02-14 21:27:09 +0200 (Fri, 14 Feb 2014) | 1 line
nightly build: Update kernel versions
........
r5288 | bvassche | 2014-02-18 10:31:44 +0200 (Tue, 18 Feb 2014) | 22 lines
scst, qla2x00t: Prevent inappropriate sleeping with a real-time kernel
With a realtime kernel with full preemption (CONFIG_PREEMPT_RT_FULL),
spinlocks can sleep, interrupt handlers run in thread context, and
the standard local_irq functions manipulate preemptibility, not HW
interruptibility. Under these conditions, most calls to local_irq
functions should be replaced by no-ops. The CONFIG_PREEMPT_RT patch
defines _nort versions of local_irq functions that compile away
under CONFIG_PREEMPT_RT_FULL and compile to their "normal"
equivalents otherwise.
Define _nort equivalents to support compilation against both
"normal" and RT-patched kernels, and use the _nort local_irq
functons in cases where spinlocks are taken within a
local_irq_save() or local_irq_disable() block. Without these
changes, runtime warnings about "sleeping function called from
invalid context" occur.
Signed-off-by: Steven J. Magnani <steve@digidescorp.com>
[bvanassche: Edited patch description and comment in scst_priv.h]
........
r5289 | bvassche | 2014-02-18 10:40:36 +0200 (Tue, 18 Feb 2014) | 13 lines
Makefiles: respect DESTDIR when specified
Not all SCST components handle DESTDIR properly, or at all.
In particular:
* INSTALL_MOD_PATH should account for DESTDIR when 'make modules_install'
is invoked, so the kernel make infrastructure deploys the modules
and runs depmod against the proper directory tree.
* depmods must include a '-b' option to reference the proper directory tree.
* Drop special ISCSI_DESTDIR.
Signed-off-by: Steven J. Magnani <steve@digidescorp.com>
........
r5290 | bvassche | 2014-02-18 10:41:30 +0200 (Tue, 18 Feb 2014) | 7 lines
Makefiles: 'uninstall' target fixes
Some components don't have 'uninstall' targets although the top-level
Makefile references them. Some others don't remove the proper file.
Signed-off-by: Steven J. Magnani <steve@digidescorp.com>
........
r5291 | vlnb | 2014-02-19 05:45:48 +0200 (Wed, 19 Feb 2014) | 8 lines
Fix incorrect start and length calculation for issuing block discard requests
Block layer always expects start and length in 512 byte blocks, so they
should be corrected for non-512b SCST devices.
Original patch from Ken Raeburn <raeburn@permabit.com>
........
r5292 | vlnb | 2014-02-19 06:06:10 +0200 (Wed, 19 Feb 2014) | 3 lines
Cleanups
........
r5293 | vlnb | 2014-02-19 06:21:00 +0200 (Wed, 19 Feb 2014) | 12 lines
scst_user: Complete "Preparing" / "finished" symmetry
Add some TRACE statements so events sent to userland are bracketed by
"Preparing" and "finished". This makes it a little easier to find the
boundaries between the various stages of command processing in trace output.
Note, this patch does not implement a 'finished' message for TM events;
there is already a "TM reply" message that can serve that purpose.
Signed-off-by: Steven J. Magnani <steve@digidescorp.com>
........
r5294 | bvassche | 2014-02-19 09:38:57 +0200 (Wed, 19 Feb 2014) | 1 line
nightly build: Update kernel versions
........
r5295 | bvassche | 2014-02-19 10:51:35 +0200 (Wed, 19 Feb 2014) | 1 line
scripts/blockdev-perftest: Fix bashisms
........
r5296 | vlnb | 2014-02-20 07:54:49 +0200 (Thu, 20 Feb 2014) | 3 lines
put_page_callback patch for 3.13.3+ kernels
........
r5300 | vlnb | 2014-02-21 04:08:05 +0200 (Fri, 21 Feb 2014) | 3 lines
Docs update
........
r5301 | bvassche | 2014-02-21 09:44:55 +0200 (Fri, 21 Feb 2014) | 1 line
nightly build: Add support for the put_page_callback-3.13.3 patch
........
r5302 | bvassche | 2014-02-21 09:48:21 +0200 (Fri, 21 Feb 2014) | 1 line
nightly build: Update kernel versions
........
r5303 | bvassche | 2014-02-21 12:02:11 +0200 (Fri, 21 Feb 2014) | 1 line
nightly build: Update kernel versions
........
r5304 | bvassche | 2014-02-21 12:09:45 +0200 (Fri, 21 Feb 2014) | 1 line
nightly build: Update kernel versions
........
r5305 | bvassche | 2014-02-24 08:56:05 +0200 (Mon, 24 Feb 2014) | 1 line
nightly build: Update kernel versions
........
r5306 | bvassche | 2014-02-24 08:56:44 +0200 (Mon, 24 Feb 2014) | 1 line
Spelling fix: initator -> initiator
........
r5307 | bvassche | 2014-02-24 09:30:50 +0200 (Mon, 24 Feb 2014) | 1 line
make rpm: Do not remove rpmbuilddir
........
r5308 | bvassche | 2014-02-24 09:39:45 +0200 (Mon, 24 Feb 2014) | 5 lines
scst_local: Add newline to sysfs output
Signed-off-by: Sebastian Herbszt <herbszt@gmx.de>
[bvanassche: Reduced source code line length to 80 columns]
........
r5309 | bvassche | 2014-02-25 12:55:36 +0200 (Tue, 25 Feb 2014) | 1 line
put_page_callback-3.12.11.patch: Add
........
r5310 | bvassche | 2014-02-25 12:57:27 +0200 (Tue, 25 Feb 2014) | 1 line
put_page_callback-3.10.30.patch: Add
........
r5311 | bvassche | 2014-02-25 12:58:08 +0200 (Tue, 25 Feb 2014) | 1 line
nightly build: Add support for kernels >= 3.10.30 and >= 3.12.11
........
r5312 | bvassche | 2014-02-25 12:59:54 +0200 (Tue, 25 Feb 2014) | 1 line
nightly build: Update kernel versions
........
r5315 | vlnb | 2014-02-26 04:32:39 +0200 (Wed, 26 Feb 2014) | 3 lines
Make internal memory layout more cache friendly
........
r5316 | vlnb | 2014-02-26 04:49:38 +0200 (Wed, 26 Feb 2014) | 5 lines
scst_vdisk: Make vendor, product ID and related fields configurable via sysfs
Signed-off-by: Bart Van Assche <bvanassche@acm.org>
........
r5320 | bvassche | 2014-03-02 10:49:50 +0200 (Sun, 02 Mar 2014) | 1 line
Documentation spelling fix: change INQUERY into INQUIRY
........
git-svn-id: http://svn.code.sf.net/p/scst/svn/branches/iser@5321 d57e44dd-8a1f-0410-8b47-8ef2f437770f
In some cases, iscsi-scstd chooses to close the connection device and abort the connection.
This can cause invalid device state due to order of isert_conn_dev cleanup and disconnect handling.
Make sure we release isert_conn_dev only after we received disconnect event, or we passed the connection
to the kernel. This also fixes an issue if iscsi-scstd is run on very CPU intensive load and it does not receive
CPU time to serve the login requests. This may get to the extreme of initiator disconnecting before iscsi-scstd
had the chance to handle the login request.
Signed-off-by: Yan Burman <yanb@mellanox.com>
git-svn-id: http://svn.code.sf.net/p/scst/svn/branches/iser@5318 d57e44dd-8a1f-0410-8b47-8ef2f437770f
Instead of waiting for TIMEDWAIT event when disconnecting and blocking,
start the disconnect process when TIMEDWAIT event is received
and only perform the actual rdma_disconnect upon disconnect request.
Note that rdma_disconnect can not be called from atomic context, so need
to execute it from workqueue.
Signed-off-by: Yan Burman <yanb@mellanox.com>
git-svn-id: http://svn.code.sf.net/p/scst/svn/branches/iser@5317 d57e44dd-8a1f-0410-8b47-8ef2f437770f
Bart Van Assche
Vladislav Bolkhovitin
Yan Burman