Files
Vladislav Bolkhovitin bb4d276a54 Merged revisions 7096,7101,7104,7106-7111,7115-7118,7125,7134-7137,7139,7141,7143,7152-7154,7158,7162-7163,7168-7171,7173,7175-7179,7185,7190,7193,7203,7208 via svnmerge from
svn+ssh://svn.code.sf.net/p/scst/svn/trunk

........
  r7096 | bvassche | 2017-02-23 18:08:17 -0800 (Thu, 23 Feb 2017) | 4 lines
  
  scst_copy_mgr: Avoid that LUN removal triggers a BUG()
  
  Reported-by: Jinpu Wang <jinpu.wang@profitbricks.com>
........
  r7101 | bvassche | 2017-03-01 07:31:59 -0800 (Wed, 01 Mar 2017) | 20 lines
  
  scst_vdisk: Avoid that LUN refresh triggers a general protection fault
  
  Avoid that triggering LUN referesh concurrently with device deletion
  triggers the following:
  
  general protection fault: 0000 [#1]
  Workqueue: events vdev_inq_changed_fn [scst_vdisk]
  Call Trace:
   _raw_spin_lock_bh+0x2b/0x30
   scst_cm_update_dev+0x87/0x190 [scst]
   scst_dev_inquiry_data_changed+0xfb/0x1b0 [scst]
   vdev_inq_changed_fn+0x60/0x120 [scst_vdisk]
   process_one_work+0x14d/0x410
   worker_thread+0x66/0x460
   kthread+0xdb/0x100
   ret_from_fork+0x3f/0x70
  
  Reported-by: Jinpu Wang <jinpu.wang@profitbricks.com>
  Tested-by: Jinpu Wang <jinpu.wang@profitbricks.com>
........
  r7104 | vlnb | 2017-03-07 20:36:45 -0800 (Tue, 07 Mar 2017) | 5 lines
  
  Linux kernel v4.10 build fix.
  
  Signed-off-by: Sebastian Herbszt <herbszt@gmx.de>
........
  r7106 | bvassche | 2017-04-11 11:48:36 -0700 (Tue, 11 Apr 2017) | 4 lines
  
  ib_srpt: Ensure that the BUG_ON() argument has no side effects
  
  Reported-by: David Butterfield <dab21774@gmail.com>
........
  r7107 | vlnb | 2017-04-13 15:04:07 -0700 (Thu, 13 Apr 2017) | 7 lines
  
  The argument to sleep() would get "promoted" to an integer with value zero.
  -               sleep(0.1);
  +               usleep(100*1000);
  
  Signed-off-by: David Butterfield <dab21774@gmail.com>
........
  r7108 | vlnb | 2017-04-13 15:30:25 -0700 (Thu, 13 Apr 2017) | 6 lines
  
  create_and_open_dev() returns a (-errno), so the "if (iser_fd...)" check should
  detect *any* negative return value as a case when fd should be set to -1.
  
  Signed-off-by: David Butterfield <dab21774@gmail.com>
........
  r7109 | vlnb | 2017-04-13 15:38:38 -0700 (Thu, 13 Apr 2017) | 8 lines
  
  Change memcpy() to strncpy() because the source name string is not guaranteed
  to exist as valid addressable memory beyond the NULL byte.
  
  Signed-off-by: David Butterfield <dab21774@gmail.com>
  
  with small addition to force set last byte NULL
........
  r7110 | vlnb | 2017-04-13 16:02:18 -0700 (Thu, 13 Apr 2017) | 7 lines
  
  Thre is potential buffer overflow in iscsi_session_alloc() due to
  short computation of needed string size.  Notice the "%s@%s" in the first call
  to sprintf().
  
  Signed-off-by: David Butterfield <dab21774@gmail.com>
........
  r7111 | vlnb | 2017-04-13 16:12:46 -0700 (Thu, 13 Apr 2017) | 8 lines
  
  iscsi: avoid a crash in iscsi_extracheck_is_rd_thread()
  
  Add an extra check in iscsi_extracheck_is_rd_thread() to avoid a crash when
  conn->rd_task is NULL.
  
  Signed-off-by: David Butterfield <dab21774@gmail.com>
........
  r7115 | vlnb | 2017-04-13 16:37:13 -0700 (Thu, 13 Apr 2017) | 12 lines
  
  scst: fix memory leak in scst_proc_group_add()
  
  Valgrind noticed that the "name" allocated in scst_proc_group_add() was
  leaking.  It turns out that scst_alloc_add_acg makes its own copy of the name
  passed to it from this code, making the string duplication done here redundant
  (and leaky).  The change eliminates the string duplication (along with all its
  associated error handling logic) and simply passes the (unowned) incoming
  string down for duplication below.
  
  Signed-off-by: David Butterfield <dab21774@gmail.com>
........
  r7116 | vlnb | 2017-04-13 16:39:14 -0700 (Thu, 13 Apr 2017) | 7 lines
  
  scst_vdisk: fix memory leak in vdisk_write_proc()
  
  Another leak valgrind popped out, this one in vdisk_write_proc().
  
  Signed-off-by: David Butterfield <dab21774@gmail.com>
........
  r7117 | vlnb | 2017-04-13 16:41:14 -0700 (Thu, 13 Apr 2017) | 9 lines
  
  scst: take scst_mutex before calling scst_del_free_acg() in exit_scst()
  
  scst_del_free_acg() does lockdep_assert_held(&scst_mutex), so we'd better take
  the lock before calling it.
  
  Signed-off-by: David Butterfield <dab21774@gmail.com>
........
  r7118 | vlnb | 2017-04-13 16:42:51 -0700 (Thu, 13 Apr 2017) | 8 lines
  
  scst: take scst_mutex before calling scst_del_free_acg() in scst_proc_cleanup_module()
  
  Take lock before a call that ends up at the lockdep_assert_held() in
  scst_del_free_acg() without locking.
  
  Signed-off-by: David Butterfield <dab21774@gmail.com>
........
  r7125 | vlnb | 2017-04-13 18:17:45 -0700 (Thu, 13 Apr 2017) | 6 lines
  
  extraclean does "rm tags cscope.out"
  
  Signed-off-by: David Butterfield <dab21774@gmail.com>
........
  r7134 | vlnb | 2017-04-17 20:57:12 -0700 (Mon, 17 Apr 2017) | 5 lines
  
  iscsi-scst: replace strncpy() by strlcpy()
  
  Follow up for r7109: strlcpy() is more appropriate in this place.
........
  r7135 | vlnb | 2017-04-17 21:02:44 -0700 (Mon, 17 Apr 2017) | 5 lines
  
  fcst: Linux kernel v4.10 build fix
  
  Signed-off-by: Sebastian Herbszt <herbszt@gmx.de>
........
  r7136 | vlnb | 2017-04-17 21:06:18 -0700 (Mon, 17 Apr 2017) | 5 lines
  
  scst: avoid possible side effect with WARN_ON_ONCE()
  
  Reported-By: David Butterfield <dab21774@gmail.com>
........
  r7137 | vlnb | 2017-04-18 20:44:20 -0700 (Tue, 18 Apr 2017) | 5 lines
  
  fileio_tgt: change "#if DEBUG_TM_FN_IGNORE" to "#ifdef ..."
  
  Signed-off-by: David Butterfield <dab21774@gmail.com>
........
  r7139 | vlnb | 2017-04-20 18:02:25 -0700 (Thu, 20 Apr 2017) | 7 lines
  
  iscsi-scstd: replace signal() with sigaction()
  
  Replace signal() with sigaction() for validity in a multithreaded process
  
  Signed-off-by: David Butterfield <dab21774@gmail.com>
........
  r7141 | vlnb | 2017-04-20 18:11:37 -0700 (Thu, 20 Apr 2017) | 7 lines
  
  scst: set file size for NULLIO in PROCFS build
  
  The file size wasn't getting set for NULLIO with /proc support
  
  Signed-off-by: David Butterfield <dab21774@gmail.com>
........
  r7143 | vlnb | 2017-04-20 18:32:07 -0700 (Thu, 20 Apr 2017) | 9 lines
  
  iscsi-scst: fix ENOMEM path
  
  In an error path in iscsi_threads_pool_get(), when a new pool cannot be
  allocated, if there is a pool on iscsi_thread_pools_list, it passes that back as
  an alternative, so return zero in that case.
  
  Signed-off-by: David Butterfield <dab21774@gmail.com>
........
  r7152 | bvassche | 2017-04-26 16:53:11 -0700 (Wed, 26 Apr 2017) | 5 lines
  
  scst: Introduce scst_scsi_execute()
  
  This patch does not change any functionality but makes it easier
  to port SCST to Linux kernel v4.11.
........
  r7153 | bvassche | 2017-04-26 17:17:22 -0700 (Wed, 26 Apr 2017) | 1 line
  
  scst: Port to Linux kernel v4.11
........
  r7154 | vlnb | 2017-04-28 17:58:55 -0700 (Fri, 28 Apr 2017) | 5 lines
  
  scst: create proc/scst_threads with mode S_IRUGO, not 0
  
  Signed-off-by: David Butterfield <dab21774@gmail.com>
........
  r7158 | bvassche | 2017-05-01 14:01:33 -0700 (Mon, 01 May 2017) | 1 line
  
  scst: Kernel v4.12 build fixes
........
  r7162 | bvassche | 2017-05-02 07:13:15 -0700 (Tue, 02 May 2017) | 1 line
  
  scst_lib: Fix kernel 2.6.30 build
........
  r7163 | bvassche | 2017-05-02 07:23:00 -0700 (Tue, 02 May 2017) | 1 line
  
  scst: Fix build for kernels before v2.6.39
........
  r7168 | bvassche | 2017-05-03 19:56:27 -0700 (Wed, 03 May 2017) | 1 line
  
  scst: Fix build for kernels before v2.6.39
........
  r7169 | vlnb | 2017-05-05 18:31:59 -0700 (Fri, 05 May 2017) | 3 lines
  
  scst: nolockdep patch for kernel 4.9
........
  r7170 | vlnb | 2017-05-10 20:51:32 -0700 (Wed, 10 May 2017) | 3 lines
  
  qla2x00t: update to kernel 4.10
........
  r7171 | vlnb | 2017-05-10 20:52:51 -0700 (Wed, 10 May 2017) | 3 lines
  
  scst: nolockdep patch for kernel 4.10
........
  r7173 | vlnb | 2017-05-10 21:00:29 -0700 (Wed, 10 May 2017) | 7 lines
  
  fcst: Linux kernel v4.11 build fix
  
  Linux kernel v4.11 build fix.
  
  Signed-off-by: Sebastian Herbszt <herbszt@gmx.de>
........
  r7175 | bvassche | 2017-05-13 17:42:24 -0700 (Sat, 13 May 2017) | 4 lines
  
  scst/include/backport.h: Remove duplicate definition of kthread_create_on_node()
  
  This patch reverts most of r7168.
........
  r7176 | bvassche | 2017-05-13 17:55:39 -0700 (Sat, 13 May 2017) | 1 line
  
  scst/include/backport.h: Add a comment
........
  r7177 | bvassche | 2017-05-13 19:55:24 -0700 (Sat, 13 May 2017) | 1 line
  
  scst/include/backport.h: Fix kthread_create_on_node() definition
........
  r7178 | bvassche | 2017-05-13 20:06:54 -0700 (Sat, 13 May 2017) | 4 lines
  
  scst/include/backport.h: Add a kref_read() backport
  
  This patch does not change any functionality.
........
  r7179 | bvassche | 2017-05-13 20:13:56 -0700 (Sat, 13 May 2017) | 4 lines
  
  scst/include/backport.h: Add a backport of rcu_dereference_protected()
  
  This patch does not change any functionality.
........
  r7185 | bvassche | 2017-05-14 11:56:09 -0700 (Sun, 14 May 2017) | 14 lines
  
  scst_local: Fix a race condition
  
  Avoid that the following crash can occur:
  
  general protection fault: 0000 [#1] PREEMPT SMP
  RIP: 0010:scsi_is_host_device+0x7/0x20 [scsi_mod]
  Call Trace:
   scst_process_aens+0x95/0x1d0 [scst_local]
   scst_aen_work_fn+0x6f/0x120 [scst_local]
   process_one_work+0x20b/0x6c0
   worker_thread+0x4e/0x4a0
   kthread+0x113/0x150
   ret_from_fork+0x31/0x40
........
  r7190 | vlnb | 2017-05-19 20:00:28 -0700 (Fri, 19 May 2017) | 8 lines
  
  scst: fix possible NULL dereference in TM code
  
  TM command accessing a non-existing LUN might lead NULL dereference in
  scst_call_dev_task_mgmt_fn_done().
  
  Reported-By: <Ilan Steinberg <ilan.steinberg@kaminario.com>>
........
  r7193 | vlnb | 2017-05-22 19:23:38 -0700 (Mon, 22 May 2017) | 5 lines
  
  qla2x00t: fix broken 4.9 kernels build
  
  Reported-By: Marc Smith <marc.smith@parodyne.com>
........
  r7203 | vlnb | 2017-06-02 19:38:51 -0700 (Fri, 02 Jun 2017) | 3 lines
  
  Update to 4.11 kernels
........
  r7208 | vlnb | 2017-06-12 20:58:26 -0700 (Mon, 12 Jun 2017) | 5 lines
  
  scst: Linux kernel v4.12 warning fix.
  
  Signed-off-by: Sebastian Herbszt <herbszt@gmx.de>
........


git-svn-id: http://svn.code.sf.net/p/scst/svn/branches/3.2.x@7209 d57e44dd-8a1f-0410-8b47-8ef2f437770f
2017-07-14 00:17:31 +00:00
..
2011-04-06 12:28:40 +00:00
2011-03-29 19:14:02 +00:00
2015-11-06 03:40:37 +00:00
2016-01-22 04:04:43 +00:00
2016-12-16 02:51:54 +00:00
2016-12-16 03:00:15 +00:00

User space FILEIO handler
=========================

Version 3.2.0
-------------

User space program fileio_tgt uses interface of SCST's scst_user dev
handler and allows to see how it works in various modes. Fileio_tgt
provides mostly the same functionality as the kernel space SCST's
scst_vdisk handler with the only exceptions that it supports O_DIRECT
mode and doesn't support BLOCKIO one. O_DIRECT mode is basically the
same as BLOCKIO, but also supports files, so for some loads it could be
significantly faster, than the regular FILEIO access. All the words
about BLOCKIO mode from SCST's README file apply to O_DIRECT mode as
well.

Installation
------------

Make sure you have installed SCST core, if not, see its README file for
details how to do it.

Then run "make all install"

For the best performance make sure you don't have any debug options
(i.e. lines like "CFLAGS += -DDEBUG" containing "-DDEBUG*" string)
enabled in the Makefile.

Usage
-----

Load scst_user module using modprobe or insmod, then start fileio_tgt
program. It can be used as the following:

fileio_tgt [OPTION] name path

Where:

 - OPTION - one or several not required options, see below.
 
 - name - name of the virtual device as it will be seen by SCST and used
   in the subsequent access management using the corresponding SCST interface.
   
 - path - path to the device file.

The following options are supported:

 -b or --block=size: block size, must be power of 2 and >=512

 -e or --threads=count: number of threads

 -t or --write_through: write through mode

 -r or --read_only: read only

 -o or --direct: O_DIRECT mode, see above for details

 -n or --nullio: NULLIO mode, see SCST's README file for details

 -c or --nv_cache: NV_CACHE mode, see SCST's README file for details

 -p or --parse=type: parse type, one of "std" (default), "call" or "excpt"

 -f or --on_free=type: on free call type, one of "ignore" (default) or "call"

 -m or --mem_reuse=type: Memory reuse type, one of "all" (default), "read",
  "write" or "none"

 -s or --prio_thread: Use separate thread for mgmt (prio) commands

 -l or --non_blocking: Use non-blocking operations

Also in the debug builds the following options are supported:

 -d or --debug=level: debug tracing level

 -g or --debug_tm_ignore: turn on DEBUG_TM_IGNORE, one of the task management
  debugging features

If you don't understand some these options, don't use them, default
values provide the best performance.

Vladislav Bolkhovitin <vst@vlnb.net>, http://scst.sourceforge.net