Merge commit 'osd/linux-next'

22 files changed, 4182 insertions, 0 deletions

diff --git a/Documentation/devices.txt b/Documentation/devices.txt
index 2be08240ee80..62254d4510c6 100644
--- a/Documentation/devices.txt
+++ b/Documentation/devices.txt
@@ -3145,6 +3145,12 @@ Your cooperation is appreciated.
 		  1 = /dev/blockrom1	Second ROM card's translation layer interface
 		  ...
 
+260 char	OSD (Object-based-device) SCSI Device
+		  0 = /dev/osd0		First OSD Device
+		  1 = /dev/osd1		Second OSD Device
+		  ...
+		  255 = /dev/osd255	256th OSD Device
+
  ****	ADDITIONAL /dev DIRECTORY ENTRIES
 
 This section details additional entries that should or may exist in
diff --git a/Documentation/scsi/osd.txt b/Documentation/scsi/osd.txt
new file mode 100644
index 000000000000..a6735ae4ec04
--- /dev/null
+++ b/Documentation/scsi/osd.txt
@@ -0,0 +1,198 @@
+The OSD Standard
+================
+OSD (Object-Based Storage Device) is a T10 SCSI command set that is designed
+to provide efficient operation of input/output logical units that manage the
+allocation, placement, and accessing of variable-size data-storage containers,
+called objects. Objects are intended to contain operating system and application
+constructs. Each object has associated attributes attached to it, which are
+integral part of the object and provide metadata about the object. The standard
+defines some common obligatory attributes, but user attributes can be added as
+needed.
+
+See: http://www.t10.org/ftp/t10/drafts/osd2/ for the latest draft for OSD 2
+or search the web for "OSD SCSI"
+
+OSD in the Linux Kernel
+=======================
+osd-initiator:
+  The main component of OSD in Kernel is the osd-initiator library. Its main
+user is intended to be the pNFS-over-objects layout driver, which uses objects
+as its back-end data storage. Other clients are the other osd parts listed below.
+
+osd-uld:
+  This is a SCSI ULD that registers for OSD type devices and provides a testing
+platform, both for the in-kernel initiator as well as connected targets. It
+currently has no useful user-mode API, though it could have if need be.
+
+osdfs:
+  Is an OSD based Linux file system. It uses the osd-initiator and osd-uld,
+to export a usable file system for users.
+See Documentation/filesystems/osdfs.txt for more details
+
+osd target:
+  There are no current plans for an OSD target implementation in kernel. For all
+needs, a user-mode target that is based on the scsi tgt target framework is
+available from Ohio Supercomputer Center (OSC) at:
+http://www.open-osd.org/bin/view/Main/OscOsdProject
+There are several other target implementations. See http://open-osd.org for more
+links.
+
+Files and Folders
+=================
+This is the complete list of files included in this work:
+include/scsi/
+	osd_initiator.h   Main API for the initiator library
+	osd_types.h	  Common OSD types
+	osd_sec.h	  Security Manager API
+	osd_protocol.h	  Wire definitions of the OSD standard protocol
+	osd_attributes.h  Wire definitions of OSD attributes
+
+drivers/scsi/osd/
+	osd_initiator.c   OSD-Initiator library implementation
+	osd_uld.c	  The OSD scsi ULD
+	osd_ktest.{h,c}	  In-kernel test suite (called by osd_uld)
+	osd_debug.h	  Some printk macros
+	Makefile	  For both in-tree and out-of-tree compilation
+	Kconfig		  Enables inclusion of the different pieces
+	osd_test.c	  User-mode application to call the kernel tests
+
+The OSD-Initiator Library
+=========================
+osd_initiator is a low level implementation of an osd initiator encoder.
+But even though, it should be intuitive and easy to use. Perhaps over time an
+higher lever will form that automates some of the more common recipes.
+
+init/fini:
+- osd_dev_init() associates a scsi_device with an osd_dev structure
+  and initializes some global pools. This should be done once per scsi_device
+  (OSD LUN). The osd_dev structure is needed for calling osd_start_request().
+
+- osd_dev_fini() cleans up before a osd_dev/scsi_device destruction.
+
+OSD commands encoding, execution, and decoding of results:
+
+struct osd_request's is used to iteratively encode an OSD command and carry
+its state throughout execution. Each request goes through these stages:
+
+a. osd_start_request() allocates the request.
+
+b. Any of the osd_req_* methods is used to encode a request of the specified
+   type.
+
+c. osd_req_add_{get,set}_attr_* may be called to add get/set attributes to the
+   CDB. "List" or "Page" mode can be used exclusively. The attribute-list API
+   can be called multiple times on the same request. However, only one
+   attribute-page can be read, as mandated by the OSD standard.
+
+d. osd_finalize_request() computes offsets into the data-in and data-out buffers
+   and signs the request using the provided capability key and integrity-
+   check parameters.
+
+e. osd_execute_request() may be called to execute the request via the block
+   layer and wait for its completion.  The request can be executed
+   asynchronously by calling the block layer API directly.
+
+f. After execution, osd_req_decode_sense() can be called to decode the request's
+   sense information.
+
+g. osd_req_decode_get_attr() may be called to retrieve osd_add_get_attr_list()
+   values.
+
+h. osd_end_request() must be called to deallocate the request and any resource
+   associated with it. Note that osd_end_request cleans up the request at any
+   stage and it must always be called after a successful osd_start_request().
+
+osd_request's structure:
+
+The OSD standard defines a complex structure of IO segments pointed to by
+members in the CDB. Up to 3 segments can be deployed in the IN-Buffer and up to
+4 in the OUT-Buffer. The ASCII illustration below depicts a secure-read with
+associated get+set of attributes-lists. Other combinations very on the same
+basic theme. From no-segments-used up to all-segments-used.
+
+|________OSD-CDB__________|
+|                         |
+|read_len (offset=0)     -|---------\
+|                         |         |
+|get_attrs_list_length    |         |
+|get_attrs_list_offset   -|----\    |
+|                         |    |    |
+|retrieved_attrs_alloc_len|    |    |
+|retrieved_attrs_offset  -|----|----|-\
+|                         |    |    | |
+|set_attrs_list_length    |    |    | |
+|set_attrs_list_offset   -|-\  |    | |
+|                         | |  |    | |
+|in_data_integ_offset    -|-|--|----|-|-\
+|out_data_integ_offset   -|-|--|--\ | | |
+\_________________________/ |  |  | | | |
+                            |  |  | | | |
+|_______OUT-BUFFER________| |  |  | | | |
+|      Set attr list      |</  |  | | | |
+|                         |    |  | | | |
+|-------------------------|    |  | | | |
+|   Get attr descriptors  |<---/  | | | |
+|                         |       | | | |
+|-------------------------|       | | | |
+|    Out-data integrity   |<------/ | | |
+|                         |         | | |
+\_________________________/         | | |
+                                    | | |
+|________IN-BUFFER________|         | | |
+|      In-Data read       |<--------/ | |
+|                         |           | |
+|-------------------------|           | |
+|      Get attr list      |<----------/ |
+|                         |             |
+|-------------------------|             |
+|    In-data integrity    |<------------/
+|                         |
+\_________________________/
+
+A block device request can carry bidirectional payload by means of associating
+a bidi_read request with a main write-request. Each in/out request is described
+by a chain of BIOs associated with each request.
+The CDB is of a SCSI VARLEN CDB format, as described by OSD standard.
+The OSD standard also mandates alignment restrictions at start of each segment.
+
+In the code, in struct osd_request, there are two _osd_io_info structures to
+describe the IN/OUT buffers above, two BIOs for the data payload and up to five
+_osd_req_data_segment structures to hold the different segments allocation and
+information.
+
+Important: We have chosen to disregard the assumption that a BIO-chain (and
+the resulting sg-list) describes a linear memory buffer. Meaning only first and
+last scatter chain can be incomplete and all the middle chains are of PAGE_SIZE.
+For us, a scatter-gather-list, as its name implies and as used by the Networking
+layer, is to describe a vector of buffers that will be transferred to/from the
+wire. It works very well with current iSCSI transport. iSCSI is currently the
+only deployed OSD transport. In the future we anticipate SAS and FC attached OSD
+devices as well.
+
+The OSD Testing ULD
+===================
+TODO: More user-mode control on tests.
+
+Authors, Mailing list
+=====================
+Please communicate with us on any deployment of osd, whether using this code
+or not.
+
+Any problems, questions, bug reports, lonely OSD nights, please email:
+   OSD Dev List <osd-dev@open-osd.org>
+
+More up-to-date information can be found on:
+http://open-osd.org
+
+Boaz Harrosh <bharrosh@panasas.com>
+Benny Halevy <bhalevy@panasas.com>
+
+References
+==========
+Weber, R., "SCSI Object-Based Storage Device Commands",
+T10/1355-D ANSI/INCITS 400-2004,
+http://www.t10.org/ftp/t10/drafts/osd/osd-r10.pdf
+
+Weber, R., "SCSI Object-Based Storage Device Commands -2 (OSD-2)"
+T10/1729-D, Working Draft, rev. 3
+http://www.t10.org/ftp/t10/drafts/osd2/osd2r03.pdf
diff --git a/MAINTAINERS b/MAINTAINERS
index abb586344227..16386847a6fd 100644
--- a/MAINTAINERS
+++ b/MAINTAINERS
@@ -3237,6 +3237,16 @@ L:	orinoco-devel@lists.sourceforge.net
 W:	http://www.nongnu.org/orinoco/
 S:	Maintained
 
+OSD LIBRARY
+P:	Boaz Harrosh
+M:	bharrosh@panasas.com
+P:	Benny Halevy
+M:	bhalevy@panasas.com
+L:	osd-dev@open-osd.org
+W:	http://open-osd.org
+T:	git://git.open-osd.org/open-osd.git
+S:	Maintained
+
 P54 WIRELESS DRIVER
 P:	Michael Wu
 M:	flamingice@sourmilk.net
diff --git a/drivers/scsi/Kconfig b/drivers/scsi/Kconfig
index 403ecad48d4b..e6af21ca8bd9 100644
--- a/drivers/scsi/Kconfig
+++ b/drivers/scsi/Kconfig
@@ -1780,4 +1780,6 @@ source "drivers/scsi/pcmcia/Kconfig"
 
 source "drivers/scsi/device_handler/Kconfig"
 
+source "drivers/scsi/osd/Kconfig"
+
 endmenu
diff --git a/drivers/scsi/Makefile b/drivers/scsi/Makefile
index 72fd5043cfa1..1e49632d25c7 100644
--- a/drivers/scsi/Makefile
+++ b/drivers/scsi/Makefile
@@ -135,6 +135,8 @@ obj-$(CONFIG_CHR_DEV_SG)	+= sg.o
 obj-$(CONFIG_CHR_DEV_SCH)	+= ch.o
 obj-$(CONFIG_SCSI_ENCLOSURE)	+= ses.o
 
+obj-$(CONFIG_SCSI_OSD_INITIATOR) += osd/
+
 # This goes last, so that "real" scsi devices probe earlier
 obj-$(CONFIG_SCSI_DEBUG)	+= scsi_debug.o
 
diff --git a/drivers/scsi/osd/Kbuild b/drivers/scsi/osd/Kbuild
new file mode 100644
index 000000000000..6f96ad5b5797
--- /dev/null
+++ b/drivers/scsi/osd/Kbuild
@@ -0,0 +1,33 @@
+#
+# Kbuild for the OSD modules
+#
+# Copyright (C) 2008 Panasas Inc.  All rights reserved.
+#
+# Authors:
+#   Boaz Harrosh <bharrosh@panasas.com>
+#   Benny Halevy <bhalevy@panasas.com>
+#
+# This program is free software; you can redistribute it and/or modify
+# it under the terms of the GNU General Public License version 2
+#
+
+ifneq ($(OSD_INC),)
+# we are built out-of-tree Kconfigure everything as on
+
+CONFIG_SCSI_OSD_INITIATOR=m
+ccflags-y += -DCONFIG_SCSI_OSD_INITIATOR -DCONFIG_SCSI_OSD_INITIATOR_MODULE
+
+CONFIG_SCSI_OSD_ULD=m
+ccflags-y += -DCONFIG_SCSI_OSD_ULD -DCONFIG_SCSI_OSD_ULD_MODULE
+
+ccflags-y += -I$(OSD_INC)
+# ccflags-y += -DCONFIG_SCSI_OSD_DEBUG
+endif
+
+# libosd.ko - osd-initiator library
+libosd-y := osd_initiator.o
+obj-$(CONFIG_SCSI_OSD_INITIATOR) += libosd.o
+
+# osd.ko - SCSI ULD and char-device
+osd-y := osd_uld.o osd_ktests.o
+obj-$(CONFIG_SCSI_OSD_ULD) += osd.o
diff --git a/drivers/scsi/osd/Kconfig b/drivers/scsi/osd/Kconfig
new file mode 100644
index 000000000000..ac3be8a2ba04
--- /dev/null
+++ b/drivers/scsi/osd/Kconfig
@@ -0,0 +1,41 @@
+#
+# Kernel configuration file for the OSD scsi protocol
+#
+# Copyright (C) 2008 Panasas Inc.  All rights reserved.
+#
+# Authors:
+#   Boaz Harrosh <bharrosh@panasas.com>
+#   Benny Halevy <bhalevy@panasas.com>
+#
+# This program is free software; you can redistribute it and/or modify
+# it under the terms of the GNU General Public version 2 License as
+# published by the Free Software Foundation
+#
+# FIXME: SCSI_OSD_INITIATOR should select CONFIG (HMAC) SHA1 somehow.
+#        How is it done properly?
+#
+
+config SCSI_OSD_INITIATOR
+	tristate "OSD-Initiator library"
+	depends on SCSI
+	help
+		Enable the OSD-Initiator library (libosd.ko).
+		NOTE: You must also select CRYPTO_SHA1 + CRYPTO_HMAC and their
+		dependencies
+
+config SCSI_OSD_ULD
+	tristate "OSD Upper Level driver"
+	depends on SCSI_OSD_INITIATOR
+	help
+		Build a SCSI upper layer driver that exports /dev/osdX devices
+		to user-mode for testing and controlling OSD devices. It is also
+		needed by osdfs, for mounting an OSD based file system.
+
+config SCSI_OSD_DEBUG
+	bool "Compile All OSD modules with lots of DEBUG prints"
+	default n
+	depends on SCSI_OSD_INITIATOR
+	help
+		OSD Code is populated with lots of OSD_DEBUG(..) printouts to
+		dmesg. Enable this if you found a bug and you want to help us
+		track the problem (see also MAINTAINERS).
diff --git a/drivers/scsi/osd/Makefile b/drivers/scsi/osd/Makefile
new file mode 100755
index 000000000000..92d8b19dcef8
--- /dev/null
+++ b/drivers/scsi/osd/Makefile
@@ -0,0 +1,43 @@
+#
+# Makefile for the OSD modules (out of tree)
+#
+# Copyright (C) 2008 Panasas Inc.  All rights reserved.
+#
+# Authors:
+#   Boaz Harrosh <bharrosh@panasas.com>
+#   Benny Halevy <bhalevy@panasas.com>
+#
+# This program is free software; you can redistribute it and/or modify
+# it under the terms of the GNU General Public License version 2
+#
+# This Makefile is used to call the kernel Makefile in case of an out-of-tree
+# build.
+# $KSRC should point to a Kernel source tree otherwise host's default is
+# used. (eg. /lib/modules/`uname -r`/build)
+
+# include path for out-of-tree Headers
+OSD_INC ?= `pwd`/../../../include
+
+# allow users to override these
+# e.g. to compile for a kernel that you aren't currently running
+KSRC ?= /lib/modules/$(shell uname -r)/build
+KBUILD_OUTPUT ?=
+ARCH ?=
+V ?= 0
+
+# this is the basic Kbuild out-of-tree invocation, with the M= option
+KBUILD_BASE = +$(MAKE) -C $(KSRC) M=`pwd` KBUILD_OUTPUT=$(KBUILD_OUTPUT) ARCH=$(ARCH) V=$(V)
+
+all: libosd osd_test
+
+libosd: ;
+	$(KBUILD_BASE) OSD_INC=$(OSD_INC) modules
+
+clean: osd_test_clean
+	$(KBUILD_BASE) clean
+
+osd_test: osd_test.o
+	$(CC) -o $@ $<
+
+osd_test_clean:
+	rm -vf osd_test
diff --git a/drivers/scsi/osd/osd_debug.h b/drivers/scsi/osd/osd_debug.h
new file mode 100644
index 000000000000..579e491f11df
--- /dev/null
+++ b/drivers/scsi/osd/osd_debug.h
@@ -0,0 +1,30 @@
+/*
+ * osd_debug.h - Some kprintf macros
+ *
+ * Copyright (C) 2008 Panasas Inc.  All rights reserved.
+ *
+ * Authors:
+ *   Boaz Harrosh <bharrosh@panasas.com>
+ *   Benny Halevy <bhalevy@panasas.com>
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License version 2
+ *
+ */
+#ifndef __OSD_DEBUG_H__
+#define __OSD_DEBUG_H__
+
+#define OSD_ERR(fmt, a...) printk(KERN_ERR "osd: " fmt, ##a)
+#define OSD_INFO(fmt, a...) printk(KERN_NOTICE "osd: " fmt, ##a)
+
+#ifdef CONFIG_SCSI_OSD_DEBUG
+#define OSD_DEBUG(fmt, a...) \
+	printk(KERN_NOTICE "osd @%s:%d: " fmt, __func__, __LINE__, ##a)
+#else
+#define OSD_DEBUG(fmt, a...) do {} while (0)
+#endif
+
+/* u64 has problems with printk this will cast it to unsigned long long */
+#define _LLU(x) (unsigned long long)(x)
+
+#endif /* ndef __OSD_DEBUG_H__ */
diff --git a/drivers/scsi/osd/osd_initiator.c b/drivers/scsi/osd/osd_initiator.c
new file mode 100644
index 000000000000..a79c14ded944
--- /dev/null
+++ b/drivers/scsi/osd/osd_initiator.c
@@ -0,0 +1,1457 @@
+/*
+ * osd_initiator - Main body of the osd initiator library.
+ *
+ * Note: The file does not contain the advanced security functionality which
+ * is only needed by the security_manager's initiators.
+ *
+ * Copyright (C) 2008 Panasas Inc.  All rights reserved.
+ *
+ * Authors:
+ *   Boaz Harrosh <bharrosh@panasas.com>
+ *   Benny Halevy <bhalevy@panasas.com>
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License version 2
+ *
+ * Redistribution and use in source and binary forms, with or without
+ * modification, are permitted provided that the following conditions
+ * are met:
+ *
+ *  1. Redistributions of source code must retain the above copyright
+ *     notice, this list of conditions and the following disclaimer.
+ *  2. Redistributions in binary form must reproduce the above copyright
+ *     notice, this list of conditions and the following disclaimer in the
+ *     documentation and/or other materials provided with the distribution.
+ *  3. Neither the name of the Panasas company nor the names of its
+ *     contributors may be used to endorse or promote products derived
+ *     from this software without specific prior written permission.
+ *
+ * THIS SOFTWARE IS PROVIDED ``AS IS'' AND ANY EXPRESS OR IMPLIED
+ * WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF
+ * MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE
+ * DISCLAIMED. IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE
+ * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR
+ * CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF
+ * SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR
+ * BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF
+ * LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING
+ * NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS
+ * SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
+ */
+
+#include <scsi/osd_initiator.h>
+#include <scsi/osd_sec.h>
+#include <scsi/osd_attributes.h>
+#include <scsi/scsi_device.h>
+
+#include "osd_debug.h"
+
+enum { OSD_REQ_RETRIES = 1 };
+
+MODULE_AUTHOR("Boaz Harrosh <bharrosh@panasas.com>");
+MODULE_DESCRIPTION("open-osd initiator library libosd.ko");
+MODULE_LICENSE("GPL");
+
+static inline void build_test(void)
+{
+	/* structures were not packed */
+	BUILD_BUG_ON(sizeof(struct osd_capability) != OSD_CAP_LEN);
+	BUILD_BUG_ON(sizeof(struct osdv2_cdb) != OSD_TOTAL_CDB_LEN);
+	BUILD_BUG_ON(sizeof(struct osdv1_cdb) != OSDv1_TOTAL_CDB_LEN);
+}
+
+static char *_osd_ver_desc(struct osd_request *or)
+{
+	return osd_req_is_ver1(or) ? "OSD1" : "OSD2";
+}
+
+#define ATTR_DEF_RI(id, len) ATTR_DEF(OSD_APAGE_ROOT_INFORMATION, id, len)
+
+static int _osd_print_system_info(struct osd_dev *od, void *caps)
+{
+	struct osd_request *or;
+	struct osd_attr get_attrs[] = {
+		ATTR_DEF_RI(OSD_ATTR_RI_VENDOR_IDENTIFICATION, 8),
+		ATTR_DEF_RI(OSD_ATTR_RI_PRODUCT_IDENTIFICATION, 16),
+		ATTR_DEF_RI(OSD_ATTR_RI_PRODUCT_MODEL, 32),
+		ATTR_DEF_RI(OSD_ATTR_RI_PRODUCT_REVISION_LEVEL, 4),
+		ATTR_DEF_RI(OSD_ATTR_RI_PRODUCT_SERIAL_NUMBER, 64 /*variable*/),
+		ATTR_DEF_RI(OSD_ATTR_RI_OSD_NAME, 64 /*variable*/),
+		ATTR_DEF_RI(OSD_ATTR_RI_TOTAL_CAPACITY, 8),
+		ATTR_DEF_RI(OSD_ATTR_RI_USED_CAPACITY, 8),
+		ATTR_DEF_RI(OSD_ATTR_RI_NUMBER_OF_PARTITIONS, 8),
+		ATTR_DEF_RI(OSD_ATTR_RI_CLOCK, 6),
+		/* IBM-OSD-SIM Has a bug with this one put it last */
+		ATTR_DEF_RI(OSD_ATTR_RI_OSD_SYSTEM_ID, 20),
+	};
+	void *iter = NULL, *pFirst;
+	int nelem = ARRAY_SIZE(get_attrs), a = 0;
+	int ret;
+
+	or = osd_start_request(od, GFP_KERNEL);
+	if (!or)
+		return -ENOMEM;
+
+	/* get attrs */
+	osd_req_get_attributes(or, &osd_root_object);
+	osd_req_add_get_attr_list(or, get_attrs, ARRAY_SIZE(get_attrs));
+
+	ret = osd_finalize_request(or, 0, caps, NULL);
+	if (ret)
+		goto out;
+
+	ret = osd_execute_request(or);
+	if (ret) {
+		OSD_ERR("Failed to detect %s => %d\n", _osd_ver_desc(or), ret);
+		goto out;
+	}
+
+	osd_req_decode_get_attr_list(or, get_attrs, &nelem, &iter);
+
+	OSD_INFO("Detected %s device\n",
+		_osd_ver_desc(or));
+
+	pFirst = get_attrs[a++].val_ptr;
+	OSD_INFO("OSD_ATTR_RI_VENDOR_IDENTIFICATION [%s]\n",
+		(char *)pFirst);
+
+	pFirst = get_attrs[a++].val_ptr;
+	OSD_INFO("OSD_ATTR_RI_PRODUCT_IDENTIFICATION [%s]\n",
+		(char *)pFirst);
+
+	pFirst = get_attrs[a++].val_ptr;
+	OSD_INFO("OSD_ATTR_RI_PRODUCT_MODEL [%s]\n",
+		(char *)pFirst);
+
+	pFirst = get_attrs[a++].val_ptr;
+	OSD_INFO("OSD_ATTR_RI_PRODUCT_REVISION_LEVEL [%u]\n",
+		get_unaligned_be32(pFirst));
+
+	pFirst = get_attrs[a++].val_ptr;
+	OSD_INFO("OSD_ATTR_RI_PRODUCT_SERIAL_NUMBER [%s]\n",
+		(char *)pFirst);
+
+	pFirst = get_attrs[a].val_ptr;
+	OSD_INFO("OSD_ATTR_RI_OSD_NAME [%s]\n", (char *)pFirst);
+	a++;
+
+	pFirst = get_attrs[a++].val_ptr;
+	OSD_INFO("OSD_ATTR_RI_TOTAL_CAPACITY [0x%llx]\n",
+		_LLU(get_unaligned_be64(pFirst)));
+
+	pFirst = get_attrs[a++].val_ptr;
+	OSD_INFO("OSD_ATTR_RI_USED_CAPACITY [0x%llx]\n",
+		_LLU(get_unaligned_be64(pFirst)));
+
+	pFirst = get_attrs[a++].val_ptr;
+	OSD_INFO("OSD_ATTR_RI_NUMBER_OF_PARTITIONS [%llu]\n",
+		_LLU(get_unaligned_be64(pFirst)));
+
+	/* FIXME: Where are the time utilities */
+	pFirst = get_attrs[a++].val_ptr;
+	OSD_INFO("OSD_ATTR_RI_CLOCK [0x%02x%02x%02x%02x%02x%02x]\n",
+		((char *)pFirst)[0], ((char *)pFirst)[1],
+		((char *)pFirst)[2], ((char *)pFirst)[3],
+		((char *)pFirst)[4], ((char *)pFirst)[5]);
+
+	if (a < nelem) { /* IBM-OSD-SIM bug, Might not have it */
+		int len = get_attrs[a].len;
+		u8 sid_dump[32*4 + 2]; /* 2nibbles+space+ASCII */
+
+		hex_dump_to_buffer(get_attrs[a].val_ptr, len, 32, 1,
+				   sid_dump, sizeof(sid_dump), true);
+		OSD_INFO("OSD_ATTR_RI_OSD_SYSTEM_ID(%d) [%s]\n", len, sid_dump);
+		a++;
+	}
+out:
+	osd_end_request(or);
+	return ret;
+}
+
+int osd_auto_detect_ver(struct osd_dev *od, void *caps)
+{
+	int ret;
+
+	/* Auto-detect the osd version */
+	ret = _osd_print_system_info(od, caps);
+	if (ret) {
+		osd_dev_set_ver(od, OSD_VER1);
+		OSD_DEBUG("converting to OSD1\n");
+		ret = _osd_print_system_info(od, caps);
+	}
+
+	return ret;
+}
+EXPORT_SYMBOL(osd_auto_detect_ver);
+
+static unsigned _osd_req_cdb_len(struct osd_request *or)
+{
+	return osd_req_is_ver1(or) ? OSDv1_TOTAL_CDB_LEN : OSD_TOTAL_CDB_LEN;
+}
+
+static unsigned _osd_req_alist_elem_size(struct osd_request *or, unsigned len)
+{
+	return osd_req_is_ver1(or) ?
+		osdv1_attr_list_elem_size(len) :
+		osdv2_attr_list_elem_size(len);
+}
+
+static unsigned _osd_req_alist_size(struct osd_request *or, void *list_head)
+{
+	return osd_req_is_ver1(or) ?
+		osdv1_list_size(list_head) :
+		osdv2_list_size(list_head);
+}
+
+static unsigned _osd_req_sizeof_alist_header(struct osd_request *or)
+{
+	return osd_req_is_ver1(or) ?
+		sizeof(struct osdv1_attributes_list_header) :
+		sizeof(struct osdv2_attributes_list_header);
+}
+
+static void _osd_req_set_alist_type(struct osd_request *or,
+	void *list, int list_type)
+{
+	if (osd_req_is_ver1(or)) {
+		struct osdv1_attributes_list_header *attr_list = list;
+
+		memset(attr_list, 0, sizeof(*attr_list));
+		attr_list->type = list_type;
+	} else {
+		struct osdv2_attributes_list_header *attr_list = list;
+
+		memset(attr_list, 0, sizeof(*attr_list));
+		attr_list->type = list_type;
+	}
+}
+
+static bool _osd_req_is_alist_type(struct osd_request *or,
+	void *list, int list_type)
+{
+	if (!list)
+		return false;
+
+	if (osd_req_is_ver1(or)) {
+		struct osdv1_attributes_list_header *attr_list = list;
+
+		return attr_list->type == list_type;
+	} else {
+		struct osdv2_attributes_list_header *attr_list = list;
+
+		return attr_list->type == list_type;
+	}
+}
+
+/* This is for List-objects not Attributes-Lists */
+static void _osd_req_encode_olist(struct osd_request *or,
+	struct osd_obj_id_list *list)
+{
+	struct osd_cdb_head *cdbh = osd_cdb_head(&or->cdb);
+
+	if (osd_req_is_ver1(or)) {
+		cdbh->v1.list_identifier = list->list_identifier;
+		cdbh->v1.start_address = list->continuation_id;
+	} else {
+		cdbh->v2.list_identifier = list->list_identifier;
+		cdbh->v2.start_address = list->continuation_id;
+	}
+}
+
+static osd_cdb_offset osd_req_encode_offset(struct osd_request *or,
+	u64 offset, unsigned *padding)
+{
+	return __osd_encode_offset(offset, padding,
+			osd_req_is_ver1(or) ?
+				OSDv1_OFFSET_MIN_SHIFT : OSD_OFFSET_MIN_SHIFT,
+			OSD_OFFSET_MAX_SHIFT);
+}
+
+static struct osd_security_parameters *
+_osd_req_sec_params(struct osd_request *or)
+{
+	struct osd_cdb *ocdb = &or->cdb;
+
+	if (osd_req_is_ver1(or))
+		return &ocdb->v1.sec_params;
+	else
+		return &ocdb->v2.sec_params;
+}
+
+void osd_dev_init(struct osd_dev *osdd, struct scsi_device *scsi_device)
+{
+	memset(osdd, 0, sizeof(*osdd));
+	osdd->scsi_device = scsi_device;
+	osdd->def_timeout = BLK_DEFAULT_SG_TIMEOUT;
+#ifdef OSD_VER1_SUPPORT
+	osdd->version = OSD_VER2;
+#endif
+	/* TODO: Allocate pools for osd_request attributes ... */
+}
+EXPORT_SYMBOL(osd_dev_init);
+
+void osd_dev_fini(struct osd_dev *osdd)
+{
+	/* TODO: De-allocate pools */
+
+	osdd->scsi_device = NULL;
+}
+EXPORT_SYMBOL(osd_dev_fini);
+
+static struct osd_request *_osd_request_alloc(gfp_t gfp)
+{
+	struct osd_request *or;
+
+	/* TODO: Use mempool with one saved request */
+	or = kzalloc(sizeof(*or), gfp);
+	return or;
+}
+
+static void _osd_request_free(struct osd_request *or)
+{
+	kfree(or);
+}
+
+struct osd_request *osd_start_request(struct osd_dev *dev, gfp_t gfp)
+{
+	struct osd_request *or;
+
+	or = _osd_request_alloc(gfp);
+	if (!or)
+		return NULL;
+
+	or->osd_dev = dev;
+	or->alloc_flags = gfp;
+	or->timeout = dev->def_timeout;
+	or->retries = OSD_REQ_RETRIES;
+
+	return or;
+}
+EXPORT_SYMBOL(osd_start_request);
+
+/*
+ * If osd_finalize_request() was called but the request was not executed through
+ * the block layer, then we must release BIOs.
+ */
+static void _abort_unexecuted_bios(struct request *rq)
+{
+	struct bio *bio;
+
+	while ((bio = rq->bio) != NULL) {
+		rq->bio = bio->bi_next;
+		bio_endio(bio, 0);
+	}
+}
+
+static void _osd_free_seg(struct osd_request *or,
+	struct _osd_req_data_segment *seg)
+{
+	if (!seg->buff || !seg->alloc_size)
+		return;
+
+	kfree(seg->buff);
+	seg->buff = NULL;
+	seg->alloc_size = 0;
+}
+
+void osd_end_request(struct osd_request *or)
+{
+	struct request *rq = or->request;
+
+	_osd_free_seg(or, &or->set_attr);
+	_osd_free_seg(or, &or->enc_get_attr);
+	_osd_free_seg(or, &or->get_attr);
+
+	if (rq) {
+		if (rq->next_rq) {
+			_abort_unexecuted_bios(rq->next_rq);
+			blk_put_request(rq->next_rq);
+		}
+
+		_abort_unexecuted_bios(rq);
+		blk_put_request(rq);
+	}
+	_osd_request_free(or);
+}
+EXPORT_SYMBOL(osd_end_request);
+
+int osd_execute_request(struct osd_request *or)
+{
+	return blk_execute_rq(or->request->q, NULL, or->request, 0);
+}
+EXPORT_SYMBOL(osd_execute_request);
+
+static void osd_request_async_done(struct request *req, int error)
+{
+	struct osd_request *or = req->end_io_data;
+
+	or->async_error = error;
+
+	if (error)
+		OSD_DEBUG("osd_request_async_done error recieved %d\n", error);
+
+	if (or->async_done)
+		or->async_done(or, or->async_private);
+	else
+		osd_end_request(or);
+}
+
+int osd_execute_request_async(struct osd_request *or,
+	osd_req_done_fn *done, void *private)
+{
+	or->request->end_io_data = or;
+	or->async_private = private;
+	or->async_done = done;
+
+	blk_execute_rq_nowait(or->request->q, NULL, or->request, 0,
+			      osd_request_async_done);
+	return 0;
+}
+EXPORT_SYMBOL(osd_execute_request_async);
+
+u8 sg_out_pad_buffer[1 << OSDv1_OFFSET_MIN_SHIFT];
+u8 sg_in_pad_buffer[1 << OSDv1_OFFSET_MIN_SHIFT];
+
+static int _osd_realloc_seg(struct osd_request *or,
+	struct _osd_req_data_segment *seg, unsigned max_bytes)
+{
+	void *buff;
+
+	if (seg->alloc_size >= max_bytes)
+		return 0;
+
+	buff = krealloc(seg->buff, max_bytes, or->alloc_flags);
+	if (!buff) {
+		OSD_ERR("Failed to Realloc %d-bytes was-%d\n", max_bytes,
+			seg->alloc_size);
+		return -ENOMEM;
+	}
+
+	memset(buff + seg->alloc_size, 0, max_bytes - seg->alloc_size);
+	seg->buff = buff;
+	seg->alloc_size = max_bytes;
+	return 0;
+}
+
+static int _alloc_set_attr_list(struct osd_request *or,
+	const struct osd_attr *oa, unsigned nelem, unsigned add_bytes)
+{
+	unsigned total_bytes = add_bytes;
+
+	for (; nelem; --nelem, ++oa)
+		total_bytes += _osd_req_alist_elem_size(or, oa->len);
+
+	OSD_DEBUG("total_bytes=%d\n", total_bytes);
+	return _osd_realloc_seg(or, &or->set_attr, total_bytes);
+}
+
+static int _alloc_get_attr_desc(struct osd_request *or, unsigned max_bytes)
+{
+	OSD_DEBUG("total_bytes=%d\n", max_bytes);
+	return _osd_realloc_seg(or, &or->enc_get_attr, max_bytes);
+}
+
+static int _alloc_get_attr_list(struct osd_request *or)
+{
+	OSD_DEBUG("total_bytes=%d\n", or->get_attr.total_bytes);
+	return _osd_realloc_seg(or, &or->get_attr, or->get_attr.total_bytes);
+}
+
+/*
+ * Common to all OSD commands
+ */
+
+static void _osdv1_req_encode_common(struct osd_request *or,
+	__be16 act, const struct osd_obj_id *obj, u64 offset, u64 len)
+{
+	struct osdv1_cdb *ocdb = &or->cdb.v1;
+
+	/*
+	 * For speed, the commands
+	 *	OSD_ACT_PERFORM_SCSI_COMMAND	, V1 0x8F7E, V2 0x8F7C
+	 *	OSD_ACT_SCSI_TASK_MANAGEMENT	, V1 0x8F7F, V2 0x8F7D
+	 * are not supported here. Should pass zero and set after the call
+	 */
+	act &= __constant_cpu_to_be16(~0x0080); /* V1 action code */
+
+	OSD_DEBUG("OSDv1 execute opcode 0x%x\n", be16_to_cpu(act));
+
+	ocdb->h.varlen_cdb.opcode = VARIABLE_LENGTH_CMD;
+	ocdb->h.varlen_cdb.additional_cdb_length = OSD_ADDITIONAL_CDB_LENGTH;
+	ocdb->h.varlen_cdb.service_action = act;
+
+	ocdb->h.partition = cpu_to_be64(obj->partition);
+	ocdb->h.object = cpu_to_be64(obj->id);
+	ocdb->h.v1.length = cpu_to_be64(len);
+	ocdb->h.v1.start_address = cpu_to_be64(offset);
+}
+
+static void _osdv2_req_encode_common(struct osd_request *or,
+	 __be16 act, const struct osd_obj_id *obj, u64 offset, u64 len)
+{
+	struct osdv2_cdb *ocdb = &or->cdb.v2;
+
+	OSD_DEBUG("OSDv2 execute opcode 0x%x\n", be16_to_cpu(act));
+
+	ocdb->h.varlen_cdb.opcode = VARIABLE_LENGTH_CMD;
+	ocdb->h.varlen_cdb.additional_cdb_length = OSD_ADDITIONAL_CDB_LENGTH;
+	ocdb->h.varlen_cdb.service_action = act;
+
+	ocdb->h.partition = cpu_to_be64(obj->partition);
+	ocdb->h.object = cpu_to_be64(obj->id);
+	ocdb->h.v2.length = cpu_to_be64(len);
+	ocdb->h.v2.start_address = cpu_to_be64(offset);
+}
+
+static void _osd_req_encode_common(struct osd_request *or,
+	__be16 act, const struct osd_obj_id *obj, u64 offset, u64 len)
+{
+	if (osd_req_is_ver1(or))
+		_osdv1_req_encode_common(or, act, obj, offset, len);
+	else
+		_osdv2_req_encode_common(or, act, obj, offset, len);
+}
+
+/*
+ * Device commands
+ */
+/*TODO: void osd_req_set_master_seed_xchg(struct osd_request *, ...); */
+/*TODO: void osd_req_set_master_key(struct osd_request *, ...); */
+
+void osd_req_format(struct osd_request *or, u64 tot_capacity)
+{
+	_osd_req_encode_common(or, OSD_ACT_FORMAT_OSD, &osd_root_object, 0,
+				tot_capacity);
+}
+EXPORT_SYMBOL(osd_req_format);
+
+int osd_req_list_dev_partitions(struct osd_request *or,
+	osd_id initial_id, struct osd_obj_id_list *list, unsigned nelem)
+{
+	return osd_req_list_partition_objects(or, 0, initial_id, list, nelem);
+}
+EXPORT_SYMBOL(osd_req_list_dev_partitions);
+
+static void _osd_req_encode_flush(struct osd_request *or,
+	enum osd_options_flush_scope_values op)
+{
+	struct osd_cdb_head *ocdb = osd_cdb_head(&or->cdb);
+
+	ocdb->command_specific_options = op;
+}
+
+void osd_req_flush_obsd(struct osd_request *or,
+	enum osd_options_flush_scope_values op)
+{
+	_osd_req_encode_common(or, OSD_ACT_FLUSH_OSD, &osd_root_object, 0, 0);
+	_osd_req_encode_flush(or, op);
+}
+EXPORT_SYMBOL(osd_req_flush_obsd);
+
+/*TODO: void osd_req_perform_scsi_command(struct osd_request *,
+	const u8 *cdb, ...); */
+/*TODO: void osd_req_task_management(struct osd_request *, ...); */
+
+/*
+ * Partition commands
+ */
+static void _osd_req_encode_partition(struct osd_request *or,
+	__be16 act, osd_id partition)
+{
+	struct osd_obj_id par = {
+		.partition = partition,
+		.id = 0,
+	};
+
+	_osd_req_encode_common(or, act, &par, 0, 0);
+}
+
+void osd_req_create_partition(struct osd_request *or, osd_id partition)
+{
+	_osd_req_encode_partition(or, OSD_ACT_CREATE_PARTITION, partition);
+}
+EXPORT_SYMBOL(osd_req_create_partition);
+
+void osd_req_remove_partition(struct osd_request *or, osd_id partition)
+{
+	_osd_req_encode_partition(or, OSD_ACT_REMOVE_PARTITION, partition);
+}
+EXPORT_SYMBOL(osd_req_remove_partition);
+
+/*TODO: void osd_req_set_partition_key(struct osd_request *,
+	osd_id partition, u8 new_key_id[OSD_CRYPTO_KEYID_SIZE],
+	u8 seed[OSD_CRYPTO_SEED_SIZE]); */
+
+static int _osd_req_list_objects(struct osd_request *or,
+	__be16 action, const struct osd_obj_id *obj, osd_id initial_id,
+	struct osd_obj_id_list *list, unsigned nelem)
+{
+	struct request_queue *q = or->osd_dev->scsi_device->request_queue;
+	u64 len = nelem * sizeof(osd_id) + sizeof(*list);
+	struct bio *bio;
+
+	_osd_req_encode_common(or, action, obj, (u64)initial_id, len);
+
+	if (list->list_identifier)
+		_osd_req_encode_olist(or, list);
+
+	WARN_ON(or->in.bio);
+	bio = bio_map_kern(q, list, len, or->alloc_flags);
+	if (!bio) {
+		OSD_ERR("!!! Failed to allocate list_objects BIO\n");
+		return -ENOMEM;
+	}
+
+	bio->bi_rw &= ~(1 << BIO_RW);
+	or->in.bio = bio;
+	or->in.total_bytes = bio->bi_size;
+	return 0;
+}
+
+int osd_req_list_partition_collections(struct osd_request *or,
+	osd_id partition, osd_id initial_id, struct osd_obj_id_list *list,
+	unsigned nelem)
+{
+	struct osd_obj_id par = {
+		.partition = partition,
+		.id = 0,
+	};
+
+	return osd_req_list_collection_objects(or, &par, initial_id, list,
+					       nelem);
+}
+EXPORT_SYMBOL(osd_req_list_partition_collections);
+
+int osd_req_list_partition_objects(struct osd_request *or,
+	osd_id partition, osd_id initial_id, struct osd_obj_id_list *list,
+	unsigned nelem)
+{
+	struct osd_obj_id par = {
+		.partition = partition,
+		.id = 0,
+	};
+
+	return _osd_req_list_objects(or, OSD_ACT_LIST, &par, initial_id, list,
+				     nelem);
+}
+EXPORT_SYMBOL(osd_req_list_partition_objects);
+
+void osd_req_flush_partition(struct osd_request *or,
+	osd_id partition, enum osd_options_flush_scope_values op)
+{
+	_osd_req_encode_partition(or, OSD_ACT_FLUSH_PARTITION, partition);
+	_osd_req_encode_flush(or, op);
+}
+EXPORT_SYMBOL(osd_req_flush_partition);
+
+/*
+ * Collection commands
+ */
+/*TODO: void osd_req_create_collection(struct osd_request *,
+	const struct osd_obj_id *); */
+/*TODO: void osd_req_remove_collection(struct osd_request *,
+	const struct osd_obj_id *); */
+
+int osd_req_list_collection_objects(struct osd_request *or,
+	const struct osd_obj_id *obj, osd_id initial_id,
+	struct osd_obj_id_list *list, unsigned nelem)
+{
+	return _osd_req_list_objects(or, OSD_ACT_LIST_COLLECTION, obj,
+				     initial_id, list, nelem);
+}
+EXPORT_SYMBOL(osd_req_list_collection_objects);
+
+/*TODO: void query(struct osd_request *, ...); V2 */
+
+void osd_req_flush_collection(struct osd_request *or,
+	const struct osd_obj_id *obj, enum osd_options_flush_scope_values op)
+{
+	_osd_req_encode_common(or, OSD_ACT_FLUSH_PARTITION, obj, 0, 0);
+	_osd_req_encode_flush(or, op);
+}
+EXPORT_SYMBOL(osd_req_flush_collection);
+
+/*TODO: void get_member_attrs(struct osd_request *, ...); V2 */
+/*TODO: void set_member_attrs(struct osd_request *, ...); V2 */
+
+/*
+ * Object commands
+ */
+void osd_req_create_object(struct osd_request *or, struct osd_obj_id *obj)
+{
+	_osd_req_encode_common(or, OSD_ACT_CREATE, obj, 0, 0);
+}
+EXPORT_SYMBOL(osd_req_create_object);
+
+void osd_req_remove_object(struct osd_request *or, struct osd_obj_id *obj)
+{
+	_osd_req_encode_common(or, OSD_ACT_REMOVE, obj, 0, 0);
+}
+EXPORT_SYMBOL(osd_req_remove_object);
+
+
+/*TODO: void osd_req_create_multi(struct osd_request *or,
+	struct osd_obj_id *first, struct osd_obj_id_list *list, unsigned nelem);
+*/
+
+void osd_req_write(struct osd_request *or,
+	const struct osd_obj_id *obj, struct bio *bio, u64 offset)
+{
+	_osd_req_encode_common(or, OSD_ACT_WRITE, obj, offset, bio->bi_size);
+	WARN_ON(or->out.bio || or->out.total_bytes);
+	bio->bi_rw |= (1 << BIO_RW);
+	or->out.bio = bio;
+	or->out.total_bytes = bio->bi_size;
+}
+EXPORT_SYMBOL(osd_req_write);
+
+/*TODO: void osd_req_append(struct osd_request *,
+	const struct osd_obj_id *, struct bio *data_out); */
+/*TODO: void osd_req_create_write(struct osd_request *,
+	const struct osd_obj_id *, struct bio *data_out, u64 offset); */
+/*TODO: void osd_req_clear(struct osd_request *,
+	const struct osd_obj_id *, u64 offset, u64 len); */
+/*TODO: void osd_req_punch(struct osd_request *,
+	const struct osd_obj_id *, u64 offset, u64 len); V2 */
+
+void osd_req_flush_object(struct osd_request *or,
+	const struct osd_obj_id *obj, enum osd_options_flush_scope_values op,
+	/*V2*/ u64 offset, /*V2*/ u64 len)
+{
+	if (unlikely(osd_req_is_ver1(or) && (offset || len))) {
+		OSD_DEBUG("OSD Ver1 flush on specific range ignored\n");
+		offset = 0;
+		len = 0;
+	}
+
+	_osd_req_encode_common(or, OSD_ACT_FLUSH, obj, offset, len);
+	_osd_req_encode_flush(or, op);
+}
+EXPORT_SYMBOL(osd_req_flush_object);
+
+void osd_req_read(struct osd_request *or,
+	const struct osd_obj_id *obj, struct bio *bio, u64 offset)
+{
+	_osd_req_encode_common(or, OSD_ACT_READ, obj, offset, bio->bi_size);
+	WARN_ON(or->in.bio || or->in.total_bytes);
+	bio->bi_rw &= ~(1 << BIO_RW);
+	or->in.bio = bio;
+	or->in.total_bytes = bio->bi_size;
+}
+EXPORT_SYMBOL(osd_req_read);
+
+void osd_req_get_attributes(struct osd_request *or,
+	const struct osd_obj_id *obj)
+{
+	_osd_req_encode_common(or, OSD_ACT_GET_ATTRIBUTES, obj, 0, 0);
+}
+EXPORT_SYMBOL(osd_req_get_attributes);
+
+void osd_req_set_attributes(struct osd_request *or,
+	const struct osd_obj_id *obj)
+{
+	_osd_req_encode_common(or, OSD_ACT_SET_ATTRIBUTES, obj, 0, 0);
+}
+EXPORT_SYMBOL(osd_req_set_attributes);
+
+/*
+ * Attributes List-mode
+ */
+
+int osd_req_add_set_attr_list(struct osd_request *or,
+	const struct osd_attr *oa, unsigned nelem)
+{
+	unsigned total_bytes = or->set_attr.total_bytes;
+	void *attr_last;
+	int ret;
+
+	if (or->attributes_mode &&
+	    or->attributes_mode != OSD_CDB_GET_SET_ATTR_LISTS) {
+		WARN_ON(1);
+		return -EINVAL;
+	}
+	or->attributes_mode = OSD_CDB_GET_SET_ATTR_LISTS;
+
+	if (!total_bytes) { /* first-time: allocate and put list header */
+		total_bytes = _osd_req_sizeof_alist_header(or);
+		ret = _alloc_set_attr_list(or, oa, nelem, total_bytes);
+		if (ret)
+			return ret;
+		_osd_req_set_alist_type(or, or->set_attr.buff,
+					OSD_ATTR_LIST_SET_RETRIEVE);
+	}
+	attr_last = or->set_attr.buff + total_bytes;
+
+	for (; nelem; --nelem) {
+		struct osd_attributes_list_element *attr;
+		unsigned elem_size = _osd_req_alist_elem_size(or, oa->len);
+
+		total_bytes += elem_size;
+		if (unlikely(or->set_attr.alloc_size < total_bytes)) {
+			or->set_attr.total_bytes = total_bytes - elem_size;
+			ret = _alloc_set_attr_list(or, oa, nelem, total_bytes);
+			if (ret)
+				return ret;
+			attr_last =
+				or->set_attr.buff + or->set_attr.total_bytes;
+		}
+
+		attr = attr_last;
+		attr->page = cpu_to_be32(oa->page);
+		attr->attr_id = cpu_to_be32(oa->attr_id);
+		attr->attr_bytes = cpu_to_be16(oa->len);
+		memcpy(attr->attr_val, oa->val_ptr, oa->len);
+
+		attr_last += elem_size;
+		++oa;
+	}
+
+	or->set_attr.total_bytes = total_bytes;
+	return 0;
+}
+EXPORT_SYMBOL(osd_req_add_set_attr_list);
+
+static int _append_map_kern(struct request *req,
+	void *buff, unsigned len, gfp_t flags)
+{
+	struct bio *bio;
+	int ret;
+
+	bio = bio_map_kern(req->q, buff, len, flags);
+	if (IS_ERR(bio)) {
+		OSD_ERR("Failed bio_map_kern(%p, %d) => %ld\n", buff, len,
+			PTR_ERR(bio));
+		return PTR_ERR(bio);
+	}
+	ret = blk_rq_append_bio(req->q, req, bio);
+	if (ret) {
+		OSD_ERR("Failed blk_rq_append_bio(%p) => %d\n", bio, ret);
+		bio_put(bio);
+	}
+	return ret;
+}
+
+static int _req_append_segment(struct osd_request *or,
+	int padding, struct _osd_req_data_segment *seg,
+	struct _osd_req_data_segment *last_seg, struct _osd_io_info *io)
+{
+	void *pad_buff;
+	int ret;
+
+	if (padding) {
+		/* check if we can just add it to last buffer */
+		if (last_seg &&
+		    (padding <= last_seg->alloc_size - last_seg->total_bytes))
+			pad_buff = last_seg->buff + last_seg->total_bytes;
+		else
+			pad_buff = io->pad_buff;
+
+		ret = _append_map_kern(io->req, pad_buff, padding,
+				       or->alloc_flags);
+		if (ret)
+			return ret;
+		io->total_bytes += padding;
+	}
+
+	ret = _append_map_kern(io->req, seg->buff, seg->total_bytes,
+			       or->alloc_flags);
+	if (ret)
+		return ret;
+
+	io->total_bytes += seg->total_bytes;
+	OSD_DEBUG("padding=%d buff=%p total_bytes=%d\n", padding, seg->buff,
+		  seg->total_bytes);
+	return 0;
+}
+
+static int _osd_req_finalize_set_attr_list(struct osd_request *or)
+{
+	struct osd_cdb_head *cdbh = osd_cdb_head(&or->cdb);
+	int padding;
+	int ret;
+
+	if (!or->set_attr.total_bytes) {
+		cdbh->attrs_list.set_attr_offset = OSD_OFFSET_UNUSED;
+		return 0;
+	}
+
+	cdbh->attrs_list.set_attr_bytes = cpu_to_be32(or->set_attr.total_bytes);
+	cdbh->attrs_list.set_attr_offset =
+		osd_req_encode_offset(or, or->out.total_bytes, &padding);
+
+	ret = _req_append_segment(or, padding, &or->set_attr,
+				  or->out.last_seg, &or->out);
+	if (ret)
+		return ret;
+
+	or->out.last_seg = &or->set_attr;
+	return 0;
+}
+
+int osd_req_add_get_attr_list(struct osd_request *or,
+	const struct osd_attr *oa, unsigned nelem)
+{
+	unsigned total_bytes = or->enc_get_attr.total_bytes;
+	void *attr_last;
+	int ret;
+
+	if (or->attributes_mode &&
+	    or->attributes_mode != OSD_CDB_GET_SET_ATTR_LISTS) {
+		WARN_ON(1);
+		return -EINVAL;
+	}
+	or->attributes_mode = OSD_CDB_GET_SET_ATTR_LISTS;
+
+	/* first time calc data-in list header size */
+	if (!or->get_attr.total_bytes)
+		or->get_attr.total_bytes = _osd_req_sizeof_alist_header(or);
+
+	/* calc data-out info */
+	if (!total_bytes) { /* first-time: allocate and put list header */
+		unsigned max_bytes;
+
+		total_bytes = _osd_req_sizeof_alist_header(or);
+		max_bytes = total_bytes +
+			nelem * sizeof(struct osd_attributes_list_attrid);
+		ret = _alloc_get_attr_desc(or, max_bytes);
+		if (ret)
+			return ret;
+
+		_osd_req_set_alist_type(or, or->enc_get_attr.buff,
+					OSD_ATTR_LIST_GET);
+	}
+	attr_last = or->enc_get_attr.buff + total_bytes;
+
+	for (; nelem; --nelem) {
+		struct osd_attributes_list_attrid *attrid;
+		const unsigned cur_size = sizeof(*attrid);
+
+		total_bytes += cur_size;
+		if (unlikely(or->enc_get_attr.alloc_size < total_bytes)) {
+			or->enc_get_attr.total_bytes = total_bytes - cur_size;
+			ret = _alloc_get_attr_desc(or,
+					total_bytes + nelem * sizeof(*attrid));
+			if (ret)
+				return ret;
+			attr_last = or->enc_get_attr.buff +
+				or->enc_get_attr.total_bytes;
+		}
+
+		attrid = attr_last;
+		attrid->page = cpu_to_be32(oa->page);
+		attrid->attr_id = cpu_to_be32(oa->attr_id);
+
+		attr_last += cur_size;
+
+		/* calc data-in size */
+		or->get_attr.total_bytes +=
+			_osd_req_alist_elem_size(or, oa->len);
+		++oa;
+	}
+
+	or->enc_get_attr.total_bytes = total_bytes;
+
+	OSD_DEBUG(
+	       "get_attr.total_bytes=%u(%u) enc_get_attr.total_bytes=%u(%Zu)\n",
+	       or->get_attr.total_bytes,
+	       or->get_attr.total_bytes - _osd_req_sizeof_alist_header(or),
+	       or->enc_get_attr.total_bytes,
+	       (or->enc_get_attr.total_bytes - _osd_req_sizeof_alist_header(or))
+			/ sizeof(struct osd_attributes_list_attrid));
+
+	return 0;
+}
+EXPORT_SYMBOL(osd_req_add_get_attr_list);
+
+static int _osd_req_finalize_get_attr_list(struct osd_request *or)
+{
+	struct osd_cdb_head *cdbh = osd_cdb_head(&or->cdb);
+	unsigned out_padding;
+	unsigned in_padding;
+	int ret;
+
+	if (!or->enc_get_attr.total_bytes) {
+		cdbh->attrs_list.get_attr_desc_offset = OSD_OFFSET_UNUSED;
+		cdbh->attrs_list.get_attr_offset = OSD_OFFSET_UNUSED;
+		return 0;
+	}
+
+	ret = _alloc_get_attr_list(or);
+	if (ret)
+		return ret;
+
+	/* The out-going buffer info update */
+	OSD_DEBUG("out-going\n");
+	cdbh->attrs_list.get_attr_desc_bytes =
+		cpu_to_be32(or->enc_get_attr.total_bytes);
+
+	cdbh->attrs_list.get_attr_desc_offset =
+		osd_req_encode_offset(or, or->out.total_bytes, &out_padding);
+
+	ret = _req_append_segment(or, out_padding, &or->enc_get_attr,
+				  or->out.last_seg, &or->out);
+	if (ret)
+		return ret;
+	or->out.last_seg = &or->enc_get_attr;
+
+	/* The incoming buffer info update */
+	OSD_DEBUG("in-coming\n");
+	cdbh->attrs_list.get_attr_alloc_length =
+		cpu_to_be32(or->get_attr.total_bytes);
+
+	cdbh->attrs_list.get_attr_offset =
+		osd_req_encode_offset(or, or->in.total_bytes, &in_padding);
+
+	ret = _req_append_segment(or, in_padding, &or->get_attr, NULL,
+				  &or->in);
+	if (ret)
+		return ret;
+	or->in.last_seg = &or->get_attr;
+
+	return 0;
+}
+
+int osd_req_decode_get_attr_list(struct osd_request *or,
+	struct osd_attr *oa, int *nelem, void **iterator)
+{
+	unsigned cur_bytes, returned_bytes, n;
+	const unsigned sizeof_attr_list = _osd_req_sizeof_alist_header(or);
+	void *cur_p;
+
+	if (!_osd_req_is_alist_type(or, or->get_attr.buff,
+				    OSD_ATTR_LIST_SET_RETRIEVE)) {
+		oa->page = 0;
+		oa->attr_id = 0;
+		oa->val_ptr = NULL;
+		oa->len = 0;
+		*iterator = NULL;
+		return 0;
+	}
+
+	if (*iterator) {
+		BUG_ON((*iterator < or->get_attr.buff) ||
+		     (or->get_attr.buff + or->get_attr.alloc_size < *iterator));
+		cur_p = *iterator;
+		cur_bytes = (*iterator - or->get_attr.buff) - sizeof_attr_list;
+		returned_bytes = or->get_attr.total_bytes;
+	} else { /* first time decode the list header */
+		cur_bytes = sizeof_attr_list;
+		returned_bytes = _osd_req_alist_size(or, or->get_attr.buff) +
+					sizeof_attr_list;
+
+		cur_p = or->get_attr.buff + sizeof_attr_list;
+
+		if (returned_bytes > or->get_attr.alloc_size) {
+			OSD_DEBUG("target report: space was not big enough! "
+				  "Allocate=%u Needed=%u\n",
+				  or->get_attr.alloc_size,
+				  returned_bytes + sizeof_attr_list);
+
+			returned_bytes =
+				or->get_attr.alloc_size - sizeof_attr_list;
+		}
+		or->get_attr.total_bytes = returned_bytes;
+	}
+
+	for (n = 0; (n < *nelem) && (cur_bytes < returned_bytes); ++n) {
+		struct osd_attributes_list_element *attr = cur_p;
+		unsigned inc;
+
+		oa->len = be16_to_cpu(attr->attr_bytes);
+		inc = _osd_req_alist_elem_size(or, oa->len);
+		OSD_DEBUG("oa->len=%d inc=%d cur_bytes=%d\n",
+			  oa->len, inc, cur_bytes);
+		cur_bytes += inc;
+		if (cur_bytes > returned_bytes) {
+			OSD_ERR("BAD FOOD from target. list not valid!"
+				"c=%d r=%d n=%d\n",
+				cur_bytes, returned_bytes, n);
+			oa->val_ptr = NULL;
+			break;
+		}
+
+		oa->page = be32_to_cpu(attr->page);
+		oa->attr_id = be32_to_cpu(attr->attr_id);
+		oa->val_ptr = attr->attr_val;
+
+		cur_p += inc;
+		++oa;
+	}
+
+	*iterator = (returned_bytes - cur_bytes) ? cur_p : NULL;
+	*nelem = n;
+	return returned_bytes - cur_bytes;
+}
+EXPORT_SYMBOL(osd_req_decode_get_attr_list);
+
+/*
+ * Attributes Page-mode
+ */
+
+int osd_req_add_get_attr_page(struct osd_request *or,
+	u32 page_id, void *attar_page, unsigned max_page_len,
+	const struct osd_attr *set_one_attr)
+{
+	struct osd_cdb_head *cdbh = osd_cdb_head(&or->cdb);
+
+	if (or->attributes_mode &&
+	    or->attributes_mode != OSD_CDB_GET_ATTR_PAGE_SET_ONE) {
+		WARN_ON(1);
+		return -EINVAL;
+	}
+	or->attributes_mode = OSD_CDB_GET_ATTR_PAGE_SET_ONE;
+
+	or->get_attr.buff = attar_page;
+	or->get_attr.total_bytes = max_page_len;
+
+	or->set_attr.buff = set_one_attr->val_ptr;
+	or->set_attr.total_bytes = set_one_attr->len;
+
+	cdbh->attrs_page.get_attr_page = cpu_to_be32(page_id);
+	cdbh->attrs_page.get_attr_alloc_length = cpu_to_be32(max_page_len);
+	/* ocdb->attrs_page.get_attr_offset; */
+
+	cdbh->attrs_page.set_attr_page = cpu_to_be32(set_one_attr->page);
+	cdbh->attrs_page.set_attr_id = cpu_to_be32(set_one_attr->attr_id);
+	cdbh->attrs_page.set_attr_length = cpu_to_be32(set_one_attr->len);
+	/* ocdb->attrs_page.set_attr_offset; */
+	return 0;
+}
+EXPORT_SYMBOL(osd_req_add_get_attr_page);
+
+static int _osd_req_finalize_attr_page(struct osd_request *or)
+{
+	struct osd_cdb_head *cdbh = osd_cdb_head(&or->cdb);
+	int in_padding, out_padding;
+	int ret;
+
+	/* returned page */
+	cdbh->attrs_page.get_attr_offset =
+		osd_req_encode_offset(or, or->in.total_bytes, &in_padding);
+
+	ret = _req_append_segment(or, in_padding, &or->get_attr, NULL,
+				  &or->in);
+	if (ret)
+		return ret;
+
+	/* set one value */
+	cdbh->attrs_page.set_attr_offset =
+		osd_req_encode_offset(or, or->out.total_bytes, &out_padding);
+
+	ret = _req_append_segment(or, out_padding, &or->enc_get_attr, NULL,
+				  &or->out);
+	return ret;
+}
+
+static int _osd_req_finalize_data_integrity(struct osd_request *or,
+	bool has_in, bool has_out, const u8 *cap_key)
+{
+	struct osd_security_parameters *sec_parms = _osd_req_sec_params(or);
+	int ret;
+
+	if (!osd_is_sec_alldata(sec_parms))
+		return 0;
+
+	if (has_out) {
+		struct _osd_req_data_segment seg = {
+			.buff = &or->out_data_integ,
+			.total_bytes = sizeof(or->out_data_integ),
+		};
+		int pad;
+
+		or->out_data_integ.data_bytes = cpu_to_be64(
+			or->out.bio ? or->out.bio->bi_size : 0);
+		or->out_data_integ.set_attributes_bytes = cpu_to_be64(
+			or->set_attr.total_bytes);
+		or->out_data_integ.get_attributes_bytes = cpu_to_be64(
+			or->enc_get_attr.total_bytes);
+
+		sec_parms->data_out_integrity_check_offset =
+			osd_req_encode_offset(or, or->out.total_bytes, &pad);
+
+		ret = _req_append_segment(or, pad, &seg, or->out.last_seg,
+					  &or->out);
+		if (ret)
+			return ret;
+		or->out.last_seg = NULL;
+
+		/* they are now all chained to request sign them all together */
+		osd_sec_sign_data(&or->out_data_integ, or->out.req->bio,
+				  cap_key);
+	}
+
+	if (has_in) {
+		struct _osd_req_data_segment seg = {
+			.buff = &or->in_data_integ,
+			.total_bytes = sizeof(or->in_data_integ),
+		};
+		int pad;
+
+		sec_parms->data_in_integrity_check_offset =
+			osd_req_encode_offset(or, or->in.total_bytes, &pad);
+
+		ret = _req_append_segment(or, pad, &seg, or->in.last_seg,
+					  &or->in);
+		if (ret)
+			return ret;
+
+		or->in.last_seg = NULL;
+	}
+
+	return 0;
+}
+
+/*
+ * osd_finalize_request and helpers
+ */
+
+static int _init_blk_request(struct osd_request *or,
+	bool has_in, bool has_out)
+{
+	gfp_t flags = or->alloc_flags;
+	struct scsi_device *scsi_device = or->osd_dev->scsi_device;
+	struct request_queue *q = scsi_device->request_queue;
+	struct request *req;
+	int ret = -ENOMEM;
+
+	req = blk_get_request(q, has_out, flags);
+	if (!req)
+		goto out;
+
+	or->request = req;
+	req->cmd_type = REQ_TYPE_BLOCK_PC;
+	req->timeout = or->timeout;
+	req->retries = or->retries;
+	req->sense = or->sense;
+	req->sense_len = 0;
+
+	if (has_out) {
+		or->out.req = req;
+		if (has_in) {
+			/* allocate bidi request */
+			req = blk_get_request(q, READ, flags);
+			if (!req) {
+				OSD_DEBUG("blk_get_request for bidi failed\n");
+				goto out;
+			}
+			req->cmd_type = REQ_TYPE_BLOCK_PC;
+			or->in.req = or->request->next_rq = req;
+		}
+	} else if (has_in)
+		or->in.req = req;
+
+	ret = 0;
+out:
+	OSD_DEBUG("or=%p has_in=%d has_out=%d => %d, %p\n",
+			or, has_in, has_out, ret, or->request);
+	return ret;
+}
+
+int osd_finalize_request(struct osd_request *or,
+	u8 options, const void *cap, const u8 *cap_key)
+{
+	struct osd_cdb_head *cdbh = osd_cdb_head(&or->cdb);
+	bool has_in, has_out;
+	int ret;
+
+	if (options & OSD_REQ_FUA)
+		cdbh->options |= OSD_CDB_FUA;
+
+	if (options & OSD_REQ_DPO)
+		cdbh->options |= OSD_CDB_DPO;
+
+	if (options & OSD_REQ_BYPASS_TIMESTAMPS)
+		cdbh->timestamp_control = OSD_CDB_BYPASS_TIMESTAMPS;
+
+	osd_set_caps(&or->cdb, cap);
+
+	has_in = or->in.bio || or->get_attr.total_bytes;
+	has_out = or->out.bio || or->set_attr.total_bytes ||
+		or->enc_get_attr.total_bytes;
+
+	ret = _init_blk_request(or, has_in, has_out);
+	if (ret) {
+		OSD_DEBUG("_init_blk_request failed\n");
+		return ret;
+	}
+
+	if (or->out.bio) {
+		ret = blk_rq_append_bio(or->request->q, or->out.req,
+					or->out.bio);
+		if (ret) {
+			OSD_DEBUG("blk_rq_append_bio out failed\n");
+			return ret;
+		}
+		OSD_DEBUG("out bytes=%llu (bytes_req=%u)\n",
+			_LLU(or->out.total_bytes), or->out.req->data_len);
+	}
+	if (or->in.bio) {
+		ret = blk_rq_append_bio(or->request->q, or->in.req, or->in.bio);
+		if (ret) {
+			OSD_DEBUG("blk_rq_append_bio in failed\n");
+			return ret;
+		}
+		OSD_DEBUG("in bytes=%llu (bytes_req=%u)\n",
+			_LLU(or->in.total_bytes), or->in.req->data_len);
+	}
+
+	or->out.pad_buff = sg_out_pad_buffer;
+	or->in.pad_buff = sg_in_pad_buffer;
+
+	if (!or->attributes_mode)
+		or->attributes_mode = OSD_CDB_GET_SET_ATTR_LISTS;
+	cdbh->command_specific_options |= or->attributes_mode;
+	if (or->attributes_mode == OSD_CDB_GET_ATTR_PAGE_SET_ONE) {
+		ret = _osd_req_finalize_attr_page(or);
+	} else {
+		/* TODO: I think that for the GET_ATTR command these 2 should
+		 * be reversed to keep them in execution order (for embeded
+		 * targets with low memory footprint)
+		 */
+		ret = _osd_req_finalize_set_attr_list(or);
+		if (ret) {
+			OSD_DEBUG("_osd_req_finalize_set_attr_list failed\n");
+			return ret;
+		}
+
+		ret = _osd_req_finalize_get_attr_list(or);
+		if (ret) {
+			OSD_DEBUG("_osd_req_finalize_get_attr_list failed\n");
+			return ret;
+		}
+	}
+
+	ret = _osd_req_finalize_data_integrity(or, has_in, has_out, cap_key);
+	if (ret)
+		return ret;
+
+	osd_sec_sign_cdb(&or->cdb, cap_key);
+
+	or->request->cmd = or->cdb.buff;
+	or->request->cmd_len = _osd_req_cdb_len(or);
+
+	return 0;
+}
+EXPORT_SYMBOL(osd_finalize_request);
+
+/*
+ * Implementation of osd_sec.h API
+ * TODO: Move to a separate osd_sec.c file at a later stage.
+ */
+
+enum { OSD_SEC_CAP_V1_ALL_CAPS =
+	OSD_SEC_CAP_APPEND | OSD_SEC_CAP_OBJ_MGMT | OSD_SEC_CAP_REMOVE   |
+	OSD_SEC_CAP_CREATE | OSD_SEC_CAP_SET_ATTR | OSD_SEC_CAP_GET_ATTR |
+	OSD_SEC_CAP_WRITE  | OSD_SEC_CAP_READ     | OSD_SEC_CAP_POL_SEC  |
+	OSD_SEC_CAP_GLOBAL | OSD_SEC_CAP_DEV_MGMT
+};
+
+enum { OSD_SEC_CAP_V2_ALL_CAPS =
+	OSD_SEC_CAP_V1_ALL_CAPS | OSD_SEC_CAP_QUERY | OSD_SEC_CAP_M_OBJECT
+};
+
+void osd_sec_init_nosec_doall_caps(void *caps,
+	const struct osd_obj_id *obj, bool is_collection, const bool is_v1)
+{
+	struct osd_capability *cap = caps;
+	u8 type;
+	u8 descriptor_type;
+
+	if (likely(obj->id)) {
+		if (unlikely(is_collection)) {
+			type = OSD_SEC_OBJ_COLLECTION;
+			descriptor_type = is_v1 ? OSD_SEC_OBJ_DESC_OBJ :
+						  OSD_SEC_OBJ_DESC_COL;
+		} else {
+			type = OSD_SEC_OBJ_USER;
+			descriptor_type = OSD_SEC_OBJ_DESC_OBJ;
+		}
+		WARN_ON(!obj->partition);
+	} else {
+		type = obj->partition ? OSD_SEC_OBJ_PARTITION :
+					OSD_SEC_OBJ_ROOT;
+		descriptor_type = OSD_SEC_OBJ_DESC_PAR;
+	}
+
+	memset(cap, 0, sizeof(*cap));
+
+	cap->h.format = OSD_SEC_CAP_FORMAT_VER1;
+	cap->h.integrity_algorithm__key_version = 0; /* MAKE_BYTE(0, 0); */
+	cap->h.security_method = OSD_SEC_NOSEC;
+/*	cap->expiration_time;
+	cap->AUDIT[30-10];
+	cap->discriminator[42-30];
+	cap->object_created_time; */
+	cap->h.object_type = type;
+	osd_sec_set_caps(&cap->h, OSD_SEC_CAP_V1_ALL_CAPS);
+	cap->h.object_descriptor_type = descriptor_type;
+	cap->od.obj_desc.policy_access_tag = 0;
+	cap->od.obj_desc.allowed_partition_id = cpu_to_be64(obj->partition);
+	cap->od.obj_desc.allowed_object_id = cpu_to_be64(obj->id);
+}
+EXPORT_SYMBOL(osd_sec_init_nosec_doall_caps);
+
+/* FIXME: Extract version from caps pointer.
+ *        Also Pete's target only supports caps from OSDv1 for now
+ */
+void osd_set_caps(struct osd_cdb *cdb, const void *caps)
+{
+	bool is_ver1 = true;
+	/* NOTE: They start at same address */
+	memcpy(&cdb->v1.caps, caps, is_ver1 ? OSDv1_CAP_LEN : OSD_CAP_LEN);
+}
+
+bool osd_is_sec_alldata(struct osd_security_parameters *sec_parms)
+{
+	return false;
+}
+
+void osd_sec_sign_cdb(struct osd_cdb *ocdb, const u8 *cap_key)
+{
+}
+
+void osd_sec_sign_data(void *data_integ, struct bio *bio, const u8 *cap_key)
+{
+}
+
+/*
+ * Declared in osd_protocol.h
+ * 4.12.5 Data-In and Data-Out buffer offsets
+ * byte offset = mantissa * (2^(exponent+8))
+ * Returns the smallest allowed encoded offset that contains given @offset
+ * The actual encoded offset returned is @offset + *@padding.
+ */
+osd_cdb_offset __osd_encode_offset(
+	u64 offset, unsigned *padding, int min_shift, int max_shift)
+{
+	u64 try_offset = -1, mod, align;
+	osd_cdb_offset be32_offset;
+	int shift;
+
+	*padding = 0;
+	if (!offset)
+		return 0;
+
+	for (shift = min_shift; shift < max_shift; ++shift) {
+		try_offset = offset >> shift;
+		if (try_offset < (1 << OSD_OFFSET_MAX_BITS))
+			break;
+	}
+
+	BUG_ON(shift == max_shift);
+
+	align = 1 << shift;
+	mod = offset & (align - 1);
+	if (mod) {
+		*padding = align - mod;
+		try_offset += 1;
+	}
+
+	try_offset |= ((shift - 8) & 0xf) << 28;
+	be32_offset = cpu_to_be32((u32)try_offset);
+
+	OSD_DEBUG("offset=%llu mantissa=%llu exp=%d encoded=%x pad=%d\n",
+		 _LLU(offset), _LLU(try_offset & 0x0FFFFFFF), shift,
+		 be32_offset, *padding);
+	return be32_offset;
+}
diff --git a/drivers/scsi/osd/osd_ktests.c b/drivers/scsi/osd/osd_ktests.c
new file mode 100644
index 000000000000..d09832358096
--- /dev/null
+++ b/drivers/scsi/osd/osd_ktests.c
@@ -0,0 +1,441 @@
+/*
+ * osd_ktests.c - An osd_initiator library in-kernel test suite
+ *              called by the osd_uld module
+ *
+ * Copyright (C) 2008 Panasas Inc.  All rights reserved.
+ *
+ * Authors:
+ *   Boaz Harrosh <bharrosh@panasas.com>
+ *   Benny Halevy <bhalevy@panasas.com>
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License version 2
+ *
+ * Redistribution and use in source and binary forms, with or without
+ * modification, are permitted provided that the following conditions
+ * are met:
+ *
+ *  1. Redistributions of source code must retain the above copyright
+ *     notice, this list of conditions and the following disclaimer.
+ *  2. Redistributions in binary form must reproduce the above copyright
+ *     notice, this list of conditions and the following disclaimer in the
+ *     documentation and/or other materials provided with the distribution.
+ *  3. Neither the name of the Panasas company nor the names of its
+ *     contributors may be used to endorse or promote products derived
+ *     from this software without specific prior written permission.
+ *
+ * THIS SOFTWARE IS PROVIDED ``AS IS'' AND ANY EXPRESS OR IMPLIED
+ * WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF
+ * MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE
+ * DISCLAIMED. IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE
+ * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR
+ * CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF
+ * SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR
+ * BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF
+ * LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING
+ * NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS
+ * SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
+ *
+ */
+#include <asm/unaligned.h>
+#include <linux/vmalloc.h>
+#include <scsi/scsi_device.h>
+
+#include <scsi/osd_initiator.h>
+#include <scsi/osd_sec.h>
+#include <scsi/osd_attributes.h>
+
+#include "osd_ktests.h"
+#include "osd_debug.h"
+
+enum {
+	K = 1024,
+	M = 1024 * K,
+	G = 1024 * M,
+};
+
+const u64 format_total_capacity = 128 * M;
+const osd_id first_par_id = 0x17171717L;
+const osd_id first_obj_id = 0x18181818L;
+const unsigned BUFF_SIZE = PAGE_SIZE;
+
+const int num_partitions = 1;
+const int num_objects = 2; /* per partition */
+
+int test_exec(struct osd_request *or, void *caps, const struct osd_obj_id *obj)
+{
+	int ret;
+
+	osd_sec_init_nosec_doall_caps(caps, obj, false, true);
+	ret = osd_finalize_request(or, 0, caps, NULL);
+	if (ret)
+		return ret;
+
+	ret = osd_execute_request(or);
+	/* osd_req_decode_sense(or, ret); */
+	return ret;
+}
+
+#define KTEST_START_REQ(osd_dev, or) do { \
+	or = osd_start_request(osd_dev, GFP_KERNEL); \
+	if (!or) { \
+		OSD_ERR("Error @%s:%d: osd_start_request", __func__,\
+			__LINE__); \
+		return -ENOMEM; \
+	} \
+} while (0)
+
+#define KTEST_EXEC_END(or, obj, g_caps, msg) do { \
+	ret = test_exec(or, g_caps, obj); \
+	osd_end_request(or); \
+	if (ret) { \
+		OSD_ERR("Error executing "msg" => %d\n", ret); \
+		return ret; \
+	} \
+	OSD_DEBUG(msg "\n"); \
+} while (0)
+
+int ktest_format(struct osd_dev *osd_dev)
+{
+	struct osd_request *or;
+	u8 g_caps[OSD_CAP_LEN];
+	int ret;
+
+	KTEST_START_REQ(osd_dev, or);
+	osd_req_format(or, format_total_capacity);
+	KTEST_EXEC_END(or, &osd_root_object, g_caps, "format");
+	return 0;
+}
+
+int ktest_creat_par(struct osd_dev *osd_dev)
+{
+	struct osd_request *or;
+	u8 g_caps[OSD_CAP_LEN];
+	int ret;
+	int p;
+
+	for (p = 0; p < num_partitions; p++) {
+		struct osd_obj_id par = {
+			.partition = first_par_id + p,
+			.id = 0
+		};
+
+		KTEST_START_REQ(osd_dev, or);
+		osd_req_create_partition(or, par.partition);
+		KTEST_EXEC_END(or, &par, g_caps, "create_partition");
+	}
+
+	return 0;
+}
+
+int ktest_creat_obj(struct osd_dev *osd_dev)
+{
+	struct osd_request *or;
+	u8 g_caps[OSD_CAP_LEN];
+	int ret;
+	int p, o;
+
+	for (p = 0; p < num_partitions; p++)
+		for (o = 0; o < num_objects; o++) {
+			struct osd_obj_id obj = {
+				.partition = first_par_id + p,
+				.id = first_obj_id + o
+			};
+
+			KTEST_START_REQ(osd_dev, or);
+			osd_req_create_object(or, &obj);
+			KTEST_EXEC_END(or, &obj, g_caps, "create_object");
+		}
+
+	return 0;
+}
+
+int ktest_write_obj(struct osd_dev *osd_dev, void *write_buff)
+{
+	struct request_queue *req_q = osd_dev->scsi_device->request_queue;
+	struct osd_request *or;
+	u8 g_caps[OSD_CAP_LEN];
+	int ret;
+	int p, o; u64 offset = 0;
+	struct bio *write_bio;
+
+	for (p = 0; p < num_partitions; p++)
+		for (o = 0; o < num_objects; o++) {
+			struct osd_obj_id obj = {
+				.partition = first_par_id + p,
+				.id = first_obj_id + o
+			};
+
+			KTEST_START_REQ(osd_dev, or);
+			write_bio = bio_map_kern(req_q, write_buff,
+						 BUFF_SIZE, GFP_KERNEL);
+			if (!write_bio) {
+				OSD_ERR("!!! Failed to allocate write BIO\n");
+				return -ENOMEM;
+			}
+
+			osd_req_write(or, &obj, write_bio, offset);
+			KTEST_EXEC_END(or, &obj, g_caps, "write");
+			write_bio = NULL; /* released by scsi_midlayer */
+			offset += BUFF_SIZE;
+		}
+
+	return 0;
+}
+
+int ktest_read_obj(struct osd_dev *osd_dev, void *write_buff, void *read_buff)
+{
+	struct request_queue *req_q = osd_dev->scsi_device->request_queue;
+	struct osd_request *or;
+	u8 g_caps[OSD_CAP_LEN];
+	int ret;
+	int p, o; u64 offset = 0;
+	struct bio *read_bio;
+
+	for (p = 0; p < num_partitions; p++)
+		for (o = 0; o < num_objects; o++) {
+			struct osd_obj_id obj = {
+				.partition = first_par_id + p,
+				.id = first_obj_id + o
+			};
+
+			KTEST_START_REQ(osd_dev, or);
+			read_bio = bio_map_kern(req_q, read_buff,
+						BUFF_SIZE, GFP_KERNEL);
+			if (!read_bio) {
+				OSD_ERR("!!! Failed to allocate read BIO\n");
+				return -ENOMEM;
+			}
+
+			osd_req_read(or, &obj, read_bio, offset);
+			KTEST_EXEC_END(or, &obj, g_caps, "read");
+			read_bio = NULL;
+			if (memcmp(read_buff, write_buff, BUFF_SIZE))
+				OSD_ERR("!!! Read did not compare");
+			offset += BUFF_SIZE;
+		}
+
+	return 0;
+}
+
+int ktest_remove_obj(struct osd_dev *osd_dev)
+{
+	struct osd_request *or;
+	u8 g_caps[OSD_CAP_LEN];
+	int ret;
+	int p, o;
+
+	for (p = 0; p < num_partitions; p++)
+		for (o = 0; o < num_objects; o++) {
+			struct osd_obj_id obj = {
+				.partition = first_par_id + p,
+				.id = first_obj_id + o
+			};
+
+			KTEST_START_REQ(osd_dev, or);
+			osd_req_remove_object(or, &obj);
+			KTEST_EXEC_END(or, &obj, g_caps, "remove_object");
+		}
+
+	return 0;
+}
+
+int ktest_remove_par(struct osd_dev *osd_dev)
+{
+	struct osd_request *or;
+	u8 g_caps[OSD_CAP_LEN];
+	int ret;
+	int p;
+
+	for (p = 0; p < num_partitions; p++) {
+		struct osd_obj_id par = {
+			.partition = first_par_id + p,
+			.id = 0
+		};
+
+		KTEST_START_REQ(osd_dev, or);
+		osd_req_remove_partition(or, par.partition);
+		KTEST_EXEC_END(or, &par, g_caps, "remove_partition");
+	}
+
+	return 0;
+}
+
+int ktest_write_read_attr(struct osd_dev *osd_dev, void *buff,
+	bool doread, bool doset, bool doget)
+{
+	struct request_queue *req_q = osd_dev->scsi_device->request_queue;
+	struct osd_request *or;
+	char g_caps[OSD_CAP_LEN];
+	int ret;
+	struct bio *bio;
+	char *domsg;
+	/* set attrs */
+	static char name[] = "ktest_write_read_attr";
+	__be64 max_len = cpu_to_be64(0x80000000L);
+	struct osd_obj_id obj = {
+		.partition = first_par_id,
+		.id = first_obj_id,
+	};
+	struct osd_attr set_attrs[] = {
+		ATTR_SET(OSD_APAGE_OBJECT_QUOTAS, OSD_ATTR_OQ_MAXIMUM_LENGTH,
+			sizeof(max_len), &max_len),
+		ATTR_SET(OSD_APAGE_OBJECT_INFORMATION, OSD_ATTR_OI_USERNAME,
+			sizeof(name), name),
+	};
+	struct osd_attr get_attrs[] = {
+		ATTR_DEF(OSD_APAGE_OBJECT_INFORMATION,
+			OSD_ATTR_OI_USED_CAPACITY, sizeof(__be64)),
+		ATTR_DEF(OSD_APAGE_OBJECT_INFORMATION,
+			OSD_ATTR_OI_LOGICAL_LENGTH, sizeof(__be64)),
+	};
+
+	KTEST_START_REQ(osd_dev, or);
+	bio = bio_map_kern(req_q, buff, BUFF_SIZE, GFP_KERNEL);
+	if (!bio) {
+		OSD_ERR("!!! Failed to allocate BIO\n");
+		return -ENOMEM;
+	}
+
+	if (doread) {
+		osd_req_read(or, &obj, bio, 0);
+		domsg = "Read-with-attr";
+	} else {
+		osd_req_write(or, &obj, bio, 0);
+		domsg = "Write-with-attr";
+	}
+
+	if (doset)
+		osd_req_add_set_attr_list(or, set_attrs, 2);
+	if (doget)
+		osd_req_add_get_attr_list(or, get_attrs, 2);
+
+/*	KTEST_EXEC_END(or, &obj, "");*/
+	ret = test_exec(or, g_caps, &obj);
+	if (!ret && doget) {
+		void *iter = NULL, *pFirst, *pSec;
+		int nelem = 2;
+		u64 capacity_len = ~0;
+		u64 logical_len = ~0;
+
+		osd_req_decode_get_attr_list(or, get_attrs, &nelem, &iter);
+
+		/*FIXME: Std does not guaranty order of return attrs */
+		pFirst = get_attrs[0].val_ptr;
+		if (pFirst)
+			capacity_len = get_unaligned_be64(pFirst);
+		else
+			OSD_ERR("failed to read capacity_used\n");
+		pSec = get_attrs[1].val_ptr;
+		if (pSec)
+			logical_len = get_unaligned_be64(pSec);
+		else
+			OSD_ERR("failed to read logical_length\n");
+		OSD_INFO("%s capacity=%llu len=%llu\n",
+			domsg, _LLU(capacity_len), _LLU(logical_len));
+	}
+
+	osd_end_request(or);
+	if (ret) {
+		OSD_ERR("!!! Error executing %s => %d doset=%d doget=%d\n",
+			domsg, ret, doset, doget);
+		return ret;
+	}
+	OSD_DEBUG("%s\n", domsg);
+
+	return 0;
+}
+
+int do_test_17(struct osd_dev *od)
+{
+	void *write_buff = NULL;
+	void *read_buff = NULL;
+	int ret = -ENOMEM, i;
+
+/* osd_format */
+	if (ktest_format(od))
+		goto dev_fini;
+
+/* create some partition */
+	if (ktest_creat_par(od))
+		goto dev_fini;
+/* list partition see if they're all there */
+/* create some objects on some partitions */
+	if (ktest_creat_obj(od))
+		goto dev_fini;
+
+/* Alloc some buffers and bios */
+/*	write_buff = kmalloc(BUFF_SIZE, or->alloc_flags);*/
+/*	read_buff = kmalloc(BUFF_SIZE, or->alloc_flags);*/
+	write_buff = (void *)__get_free_page(GFP_KERNEL);
+	read_buff = (void *)__get_free_page(GFP_KERNEL);
+	if (!write_buff || !read_buff) {
+		OSD_ERR("!!! Failed to allocate memory for test\n");
+		goto dev_fini;
+	}
+	for (i = 0; i < BUFF_SIZE / 4; i++)
+		((int *)write_buff)[i] = i;
+	OSD_DEBUG("allocate buffers\n");
+
+/* write to objects */
+	ret = ktest_write_obj(od, write_buff);
+	if (ret)
+		goto dev_fini;
+
+/* read from objects and compare to write */
+	ret = ktest_read_obj(od, write_buff, read_buff);
+	if (ret)
+		goto dev_fini;
+
+/* List all objects */
+
+/* Write with get_attr */
+	ret = ktest_write_read_attr(od, write_buff, false, false, true);
+	if (ret)
+		goto dev_fini;
+
+/* Write with set_attr */
+	ret = ktest_write_read_attr(od, write_buff, false, true, false);
+	if (ret)
+		goto dev_fini;
+
+/* Write with set_attr + get_attr */
+	ret = ktest_write_read_attr(od, write_buff, false, true, true);
+	if (ret)
+		goto dev_fini;
+
+/* Read with set_attr */
+	ret = ktest_write_read_attr(od, write_buff, true, true, false);
+	if (ret)
+		goto dev_fini;
+
+/* Read with get_attr */
+	ret = ktest_write_read_attr(od, write_buff, true, false, true);
+	if (ret)
+		goto dev_fini;
+
+/* Read with get_attr + set_attr */
+	ret = ktest_write_read_attr(od, write_buff, true, true, true);
+	if (ret)
+		goto dev_fini;
+
+/* remove objects */
+	ret = ktest_remove_obj(od);
+	if (ret)
+		goto dev_fini;
+
+/* remove partitions */
+	ret = ktest_remove_par(od);
+	if (ret)
+		goto dev_fini;
+
+/* good and done */
+	OSD_INFO("test17: All good and done\n");
+dev_fini:
+	if (read_buff)
+		free_page((ulong)read_buff);
+	if (write_buff)
+		free_page((ulong)write_buff);
+
+	return ret;
+}
diff --git a/drivers/scsi/osd/osd_ktests.h b/drivers/scsi/osd/osd_ktests.h
new file mode 100644
index 000000000000..b625ee5d2f7f
--- /dev/null
+++ b/drivers/scsi/osd/osd_ktests.h
@@ -0,0 +1,31 @@
+/*
+ * osd_ktests.h - Define the ktests.c API
+ *
+ * Copyright (C) 2008 Panasas Inc.  All rights reserved.
+ *
+ * Authors:
+ *   Boaz Harrosh <bharrosh@panasas.com>
+ *   Benny Halevy <bhalevy@panasas.com>
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License version 2
+ *
+ */
+#ifndef __OSD_KTESTS_H__
+#define __OSD_KTESTS_H__
+
+/* Tests from osd_ktests.c */
+/* TODO: Only one simple test for now. Later I will add a test definition
+ *	structure that will define what tests to preform and with some
+ *	parametrization, so concurrent tests could be run on same OSD lun
+ *	without stepping on each other. (E.g. Format called when other tests
+ *	are in progress)
+ */
+
+enum { OSD_TEST_ALL = 17 };
+
+#ifdef __KERNEL__
+extern int do_test_17(struct osd_dev *od);
+#endif /* __KERNEL__ */
+
+#endif /*ndef __OSD_KTESTS_H__*/
diff --git a/drivers/scsi/osd/osd_test.c b/drivers/scsi/osd/osd_test.c
new file mode 100644
index 000000000000..dc748657e59c
--- /dev/null
+++ b/drivers/scsi/osd/osd_test.c
@@ -0,0 +1,74 @@
+/*
+ * osd_test.c - A user-mode program that calls into the osd ULD
+ *
+ * Copyright (C) 2008 Panasas Inc.  All rights reserved.
+ *
+ * Authors:
+ *   Boaz Harrosh <bharrosh@panasas.com>
+ *   Benny Halevy <bhalevy@panasas.com>
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License version 2
+ *
+ * Redistribution and use in source and binary forms, with or without
+ * modification, are permitted provided that the following conditions
+ * are met:
+ *
+ *  1. Redistributions of source code must retain the above copyright
+ *     notice, this list of conditions and the following disclaimer.
+ *  2. Redistributions in binary form must reproduce the above copyright
+ *     notice, this list of conditions and the following disclaimer in the
+ *     documentation and/or other materials provided with the distribution.
+ *  3. Neither the name of the Panasas company nor the names of its
+ *     contributors may be used to endorse or promote products derived
+ *     from this software without specific prior written permission.
+ *
+ * THIS SOFTWARE IS PROVIDED ``AS IS'' AND ANY EXPRESS OR IMPLIED
+ * WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF
+ * MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE
+ * DISCLAIMED. IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE
+ * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR
+ * CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF
+ * SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR
+ * BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF
+ * LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING
+ * NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS
+ * SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
+ */
+
+#include <stdio.h>
+#include <stdlib.h>
+#include <sys/ioctl.h>
+#include <fcntl.h>
+
+#include "osd_ktests.h"
+
+void usage(void)
+{
+	printf("usage: osd_test /dev/osdX testNo\n");
+}
+
+int main(int argc, char *argv[])
+{
+	int osd_file, ret;
+
+	if (argc <= 1) {
+		usage();
+		return -2;
+	}
+
+	osd_file = open(argv[1], O_RDWR);
+	if (osd_file < 0) {
+		printf("Error opening <%s>\n", argv[1]);
+		return -3;
+	}
+
+	ret = ioctl(osd_file, OSD_TEST_ALL, 0);
+	if (ret) {
+		printf("ioctl 17 returned %d\n", ret);
+		return ret;
+	}
+
+	close(osd_file);
+	return 0;
+}
diff --git a/drivers/scsi/osd/osd_uld.c b/drivers/scsi/osd/osd_uld.c
new file mode 100644
index 000000000000..168ae9745bf0
--- /dev/null
+++ b/drivers/scsi/osd/osd_uld.c
@@ -0,0 +1,445 @@
+/*
+ * osd_uld.c - OSD Upper Layer Driver
+ *
+ * A Linux driver module that registers as a SCSI ULD and probes
+ * for OSD type SCSI devices.
+ * It's main function is to export osd devices to in-kernel users like
+ * osdfs and pNFS-objects-LD. It also provides one ioctl for running
+ * in Kernel tests.
+ *
+ * Copyright (C) 2008 Panasas Inc.  All rights reserved.
+ *
+ * Authors:
+ *   Boaz Harrosh <bharrosh@panasas.com>
+ *   Benny Halevy <bhalevy@panasas.com>
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License version 2
+ *
+ * Redistribution and use in source and binary forms, with or without
+ * modification, are permitted provided that the following conditions
+ * are met:
+ *
+ *  1. Redistributions of source code must retain the above copyright
+ *     notice, this list of conditions and the following disclaimer.
+ *  2. Redistributions in binary form must reproduce the above copyright
+ *     notice, this list of conditions and the following disclaimer in the
+ *     documentation and/or other materials provided with the distribution.
+ *  3. Neither the name of the Panasas company nor the names of its
+ *     contributors may be used to endorse or promote products derived
+ *     from this software without specific prior written permission.
+ *
+ * THIS SOFTWARE IS PROVIDED ``AS IS'' AND ANY EXPRESS OR IMPLIED
+ * WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF
+ * MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE
+ * DISCLAIMED. IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE
+ * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR
+ * CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF
+ * SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR
+ * BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF
+ * LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING
+ * NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS
+ * SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
+ */
+
+#include <linux/namei.h>
+#include <linux/cdev.h>
+#include <linux/fs.h>
+#include <linux/module.h>
+#include <linux/device.h>
+#include <linux/idr.h>
+#include <linux/major.h>
+
+#include <scsi/scsi.h>
+#include <scsi/scsi_driver.h>
+#include <scsi/scsi_device.h>
+#include <scsi/scsi_ioctl.h>
+
+#include <scsi/osd_initiator.h>
+#include <scsi/osd_sec.h>
+
+#include "osd_ktests.h"
+#include "osd_debug.h"
+
+#ifndef TYPE_OSD
+#  define TYPE_OSD 0x11
+#endif
+
+#ifndef SCSI_OSD_MAJOR
+#  define SCSI_OSD_MAJOR 260
+#endif
+#define SCSI_OSD_MAX_MINOR 64
+
+static const char osd_name[] = "osd";
+static const char *osd_version_string = "open-osd 0.1.0";
+const char osd_symlink[] = "scsi_osd";
+
+MODULE_AUTHOR("Boaz Harrosh <bharrosh@panasas.com>");
+MODULE_DESCRIPTION("open-osd Upper-Layer-Driver osd.ko");
+MODULE_LICENSE("GPL");
+MODULE_ALIAS_CHARDEV_MAJOR(SCSI_OSD_MAJOR);
+MODULE_ALIAS_SCSI_DEVICE(TYPE_OSD);
+
+struct osd_uld_device {
+	int minor;
+	struct kref kref;
+	struct cdev cdev;
+	struct osd_dev od;
+	struct gendisk *disk;
+	struct device *class_member;
+};
+
+static void __uld_get(struct osd_uld_device *oud);
+static void __uld_put(struct osd_uld_device *oud);
+
+/*
+ * Char Device operations
+ */
+
+static int osd_uld_open(struct inode *inode, struct file *file)
+{
+	struct osd_uld_device *oud = container_of(inode->i_cdev,
+					struct osd_uld_device, cdev);
+
+	__uld_get(oud);
+	/* cache osd_uld_device on file handle */
+	file->private_data = oud;
+	OSD_DEBUG("osd_uld_open %p\n", oud);
+	return 0;
+}
+
+static int osd_uld_release(struct inode *inode, struct file *file)
+{
+	struct osd_uld_device *oud = file->private_data;
+
+	OSD_DEBUG("osd_uld_release %p\n", file->private_data);
+	file->private_data = NULL;
+	__uld_put(oud);
+	return 0;
+}
+
+static long osd_uld_ioctl(struct file *file, unsigned int cmd,
+	unsigned long arg)
+{
+	struct osd_uld_device *oud = file->private_data;
+	int ret;
+
+	switch (cmd) {
+	case OSD_TEST_ALL:
+		OSD_DEBUG("Kernel test %d: osd_uld_device=%p\n", cmd, oud);
+		ret = do_test_17(&oud->od);
+		break;
+	default:
+		OSD_ERR("Unknown osd_uld_ioctl %d\n", cmd);
+		ret = -ENOIOCTLCMD;
+	}
+	return ret;
+}
+
+static const struct file_operations osd_fops = {
+	.owner          = THIS_MODULE,
+	.open           = osd_uld_open,
+	.release        = osd_uld_release,
+	.unlocked_ioctl = osd_uld_ioctl,
+};
+
+struct osd_dev *osduld_path_lookup(const char *path)
+{
+	struct nameidata nd;
+	struct inode *inode;
+	struct cdev *cdev;
+	struct osd_uld_device *oud;
+	int error;
+
+	if (!path || !*path)
+		return ERR_PTR(-EINVAL);
+
+	error = path_lookup(path, LOOKUP_FOLLOW, &nd);
+	if (error)
+		return ERR_PTR(error);
+
+	inode = nd.path.dentry->d_inode;
+	error = -EINVAL; /* Not the right device e.g osd_uld_device */
+	if (!S_ISCHR(inode->i_mode))
+		goto out;
+
+	cdev = inode->i_cdev;
+	if (!cdev)
+		goto out;
+
+	/* The Magic wand. Is it our char-dev */
+	/* TODO: Support sg devices */
+	if (cdev->owner != THIS_MODULE)
+		goto out;
+
+	oud = container_of(cdev, struct osd_uld_device, cdev);
+
+	__uld_get(oud);
+	error = 0;
+
+out:
+	path_put(&nd.path);
+	return error ? ERR_PTR(error) : &oud->od;
+}
+EXPORT_SYMBOL(osduld_path_lookup);
+
+void osduld_put_device(struct osd_dev *od)
+{
+	if (od) {
+		struct osd_uld_device *oud = container_of(od,
+						struct osd_uld_device, od);
+
+		__uld_put(oud);
+	}
+}
+EXPORT_SYMBOL(osduld_put_device);
+
+/*
+ * Scsi Device operations
+ */
+
+static int __detect_osd(struct osd_uld_device *oud)
+{
+	struct scsi_device *scsi_device = oud->od.scsi_device;
+	char caps[OSD_CAP_LEN];
+	int error;
+
+	/* sending a test_unit_ready as first command seems to be needed
+	 * by some targets
+	 */
+	OSD_DEBUG("start scsi_test_unit_ready %p %p %p\n",
+			oud, scsi_device, scsi_device->request_queue);
+	error = scsi_test_unit_ready(scsi_device, 10*HZ, 5, NULL);
+	if (error)
+		OSD_ERR("warning: scsi_test_unit_ready failed\n");
+
+	osd_sec_init_nosec_doall_caps(caps, &osd_root_object, false, true);
+	if (osd_auto_detect_ver(&oud->od, caps))
+		return -ENODEV;
+
+	return 0;
+}
+
+static struct class *osd_sysfs_class;
+static DEFINE_IDA(osd_minor_ida);
+
+static int osd_probe(struct device *dev)
+{
+	struct scsi_device *scsi_device = to_scsi_device(dev);
+	struct gendisk *disk;
+	struct osd_uld_device *oud;
+	int minor;
+	int error;
+
+	if (scsi_device->type != TYPE_OSD)
+		return -ENODEV;
+
+	do {
+		if (!ida_pre_get(&osd_minor_ida, GFP_KERNEL))
+			return -ENODEV;
+
+		error = ida_get_new(&osd_minor_ida, &minor);
+	} while (error == -EAGAIN);
+
+	if (error)
+		return error;
+	if (minor >= SCSI_OSD_MAX_MINOR) {
+		error = -EBUSY;
+		goto err_retract_minor;
+	}
+
+	error = -ENOMEM;
+	oud = kzalloc(sizeof(*oud), GFP_KERNEL);
+	if (NULL == oud)
+		goto err_retract_minor;
+
+	kref_init(&oud->kref);
+	dev_set_drvdata(dev, oud);
+	oud->minor = minor;
+
+	/* allocate a disk and set it up */
+	/* FIXME: do we need this since sg has already done that */
+	disk = alloc_disk(1);
+	if (!disk) {
+		OSD_ERR("alloc_disk failed\n");
+		goto err_free_osd;
+	}
+	disk->major = SCSI_OSD_MAJOR;
+	disk->first_minor = oud->minor;
+	sprintf(disk->disk_name, "osd%d", oud->minor);
+	oud->disk = disk;
+
+	/* hold one more reference to the scsi_device that will get released
+	 * in __release, in case a logout is happening while fs is mounted
+	 */
+	scsi_device_get(scsi_device);
+	osd_dev_init(&oud->od, scsi_device);
+
+	/* Detect the OSD Version */
+	error = __detect_osd(oud);
+	if (error) {
+		OSD_ERR("osd detection failed, non-compatible OSD device\n");
+		goto err_put_disk;
+	}
+
+	/* init the char-device for communication with user-mode */
+	cdev_init(&oud->cdev, &osd_fops);
+	oud->cdev.owner = THIS_MODULE;
+	error = cdev_add(&oud->cdev,
+			 MKDEV(SCSI_OSD_MAJOR, oud->minor), 1);
+	if (error) {
+		OSD_ERR("cdev_add failed\n");
+		goto err_put_disk;
+	}
+	kobject_get(&oud->cdev.kobj); /* 2nd ref see osd_remove() */
+
+	/* class_member */
+	oud->class_member = device_create(osd_sysfs_class, dev,
+		MKDEV(SCSI_OSD_MAJOR, oud->minor), "%s", disk->disk_name);
+	if (IS_ERR(oud->class_member)) {
+		OSD_ERR("class_device_create failed\n");
+		error = PTR_ERR(oud->class_member);
+		goto err_put_cdev;
+	}
+
+	dev_set_drvdata(oud->class_member, oud);
+	error = sysfs_create_link(&scsi_device->sdev_gendev.kobj,
+				  &oud->class_member->kobj, osd_symlink);
+	if (error)
+		OSD_ERR("warning: unable to make symlink\n");
+
+	OSD_INFO("osd_probe %s\n", disk->disk_name);
+	return 0;
+
+err_put_cdev:
+	cdev_del(&oud->cdev);
+err_put_disk:
+	scsi_device_put(scsi_device);
+	put_disk(disk);
+err_free_osd:
+	dev_set_drvdata(dev, NULL);
+	kfree(oud);
+err_retract_minor:
+	ida_remove(&osd_minor_ida, minor);
+	return error;
+}
+
+static int osd_remove(struct device *dev)
+{
+	struct scsi_device *scsi_device = to_scsi_device(dev);
+	struct osd_uld_device *oud = dev_get_drvdata(dev);
+
+	if (!oud || (oud->od.scsi_device != scsi_device)) {
+		OSD_ERR("Half cooked osd-device %p,%p || %p!=%p",
+			dev, oud, oud ? oud->od.scsi_device : NULL,
+			scsi_device);
+	}
+
+	sysfs_remove_link(&oud->od.scsi_device->sdev_gendev.kobj, osd_symlink);
+
+	if (oud->class_member)
+		device_destroy(osd_sysfs_class,
+			       MKDEV(SCSI_OSD_MAJOR, oud->minor));
+
+	/* We have 2 references to the cdev. One is released here
+	 * and also takes down the /dev/osdX mapping. The second
+	 * Will be released in __remove() after all users have released
+	 * the osd_uld_device.
+	 */
+	if (oud->cdev.owner)
+		cdev_del(&oud->cdev);
+
+	__uld_put(oud);
+	return 0;
+}
+
+static void __remove(struct kref *kref)
+{
+	struct osd_uld_device *oud = container_of(kref,
+					struct osd_uld_device, kref);
+	struct scsi_device *scsi_device = oud->od.scsi_device;
+
+	/* now let delete the char_dev */
+	kobject_put(&oud->cdev.kobj);
+
+	osd_dev_fini(&oud->od);
+	scsi_device_put(scsi_device);
+
+	OSD_INFO("osd_remove %s\n",
+		 oud->disk ? oud->disk->disk_name : NULL);
+
+	if (oud->disk)
+		put_disk(oud->disk);
+
+	ida_remove(&osd_minor_ida, oud->minor);
+	kfree(oud);
+}
+
+static void __uld_get(struct osd_uld_device *oud)
+{
+	kref_get(&oud->kref);
+}
+
+static void __uld_put(struct osd_uld_device *oud)
+{
+	kref_put(&oud->kref, __remove);
+}
+
+/*
+ * Global driver and scsi registration
+ */
+
+static struct scsi_driver osd_driver = {
+	.owner			= THIS_MODULE,
+	.gendrv = {
+		.name		= osd_name,
+		.probe		= osd_probe,
+		.remove		= osd_remove,
+	}
+};
+
+static int __init osd_uld_init(void)
+{
+	int err;
+
+	osd_sysfs_class = class_create(THIS_MODULE, osd_symlink);
+	if (IS_ERR(osd_sysfs_class)) {
+		OSD_ERR("Unable to register sysfs class => %ld\n",
+			PTR_ERR(osd_sysfs_class));
+		return PTR_ERR(osd_sysfs_class);
+	}
+
+	err = register_chrdev_region(MKDEV(SCSI_OSD_MAJOR, 0),
+				     SCSI_OSD_MAX_MINOR, osd_name);
+	if (err) {
+		OSD_ERR("Unable to register major %d for osd ULD => %d\n",
+			SCSI_OSD_MAJOR, err);
+		goto err_out;
+	}
+
+	err = scsi_register_driver(&osd_driver.gendrv);
+	if (err) {
+		OSD_ERR("scsi_register_driver failed => %d\n", err);
+		goto err_out_chrdev;
+	}
+
+	OSD_INFO("LOADED %s\n", osd_version_string);
+	return 0;
+
+err_out_chrdev:
+	unregister_chrdev_region(MKDEV(SCSI_OSD_MAJOR, 0), SCSI_OSD_MAX_MINOR);
+err_out:
+	class_destroy(osd_sysfs_class);
+	return err;
+}
+
+static void __exit osd_uld_exit(void)
+{
+	scsi_unregister_driver(&osd_driver.gendrv);
+	unregister_chrdev_region(MKDEV(SCSI_OSD_MAJOR, 0), SCSI_OSD_MAX_MINOR);
+	class_destroy(osd_sysfs_class);
+	OSD_INFO("UNLOADED %s\n", osd_version_string);
+}
+
+module_init(osd_uld_init);
+module_exit(osd_uld_exit);
diff --git a/drivers/scsi/scsi_scan.c b/drivers/scsi/scsi_scan.c
index b14dc02c3ded..633cb873135c 100644
--- a/drivers/scsi/scsi_scan.c
+++ b/drivers/scsi/scsi_scan.c
@@ -783,6 +783,7 @@ static int scsi_add_lun(struct scsi_device *sdev, unsigned char *inq_result,
 	case TYPE_ENCLOSURE:
 	case TYPE_COMM:
 	case TYPE_RAID:
+	case TYPE_OSD:
 		sdev->writeable = 1;
 		break;
 	case TYPE_ROM:
diff --git a/include/linux/major.h b/include/linux/major.h
index 88249452b935..058ec15dd060 100644
--- a/include/linux/major.h
+++ b/include/linux/major.h
@@ -171,5 +171,6 @@
 #define VIOTAPE_MAJOR		230
 
 #define BLOCK_EXT_MAJOR		259
+#define SCSI_OSD_MAJOR		260	/* open-osd's OSD scsi device */
 
 #endif
diff --git a/include/scsi/osd_attributes.h b/include/scsi/osd_attributes.h
new file mode 100644
index 000000000000..c53f49b2b16d
--- /dev/null
+++ b/include/scsi/osd_attributes.h
@@ -0,0 +1,327 @@
+#ifndef __OSD_ATTRIBUTES_H__
+#define __OSD_ATTRIBUTES_H__
+
+#include "osd_protocol.h"
+
+/*
+ * Contains types and constants that define attribute pages and attribute
+ * numbers and their data types.
+ */
+
+#define ATTR_SET(pg, id, l, ptr) \
+	{ .page = pg, .attr_id = id, .len = l, .val_ptr = ptr }
+
+#define ATTR_DEF(pg, id, l) ATTR_SET(pg, id, l, NULL)
+
+/* osd-r10 4.7.3 Attributes pages */
+enum {
+	OSD_APAGE_OBJECT_FIRST		= 0x0,
+	OSD_APAGE_OBJECT_DIRECTORY	= 0,
+	OSD_APAGE_OBJECT_INFORMATION	= 1,
+	OSD_APAGE_OBJECT_QUOTAS		= 2,
+	OSD_APAGE_OBJECT_TIMESTAMP	= 3,
+	OSD_APAGE_OBJECT_COLLECTIONS	= 4,
+	OSD_APAGE_OBJECT_SECURITY	= 5,
+	OSD_APAGE_OBJECT_LAST		= 0x2fffffff,
+
+	OSD_APAGE_PARTITION_FIRST	= 0x30000000,
+	OSD_APAGE_PARTITION_DIRECTORY	= OSD_APAGE_PARTITION_FIRST + 0,
+	OSD_APAGE_PARTITION_INFORMATION = OSD_APAGE_PARTITION_FIRST + 1,
+	OSD_APAGE_PARTITION_QUOTAS	= OSD_APAGE_PARTITION_FIRST + 2,
+	OSD_APAGE_PARTITION_TIMESTAMP	= OSD_APAGE_PARTITION_FIRST + 3,
+	OSD_APAGE_PARTITION_SECURITY	= OSD_APAGE_PARTITION_FIRST + 5,
+	OSD_APAGE_PARTITION_LAST	= 0x5FFFFFFF,
+
+	OSD_APAGE_COLLECTION_FIRST	= 0x60000000,
+	OSD_APAGE_COLLECTION_DIRECTORY	= OSD_APAGE_COLLECTION_FIRST + 0,
+	OSD_APAGE_COLLECTION_INFORMATION = OSD_APAGE_COLLECTION_FIRST + 1,
+	OSD_APAGE_COLLECTION_TIMESTAMP	= OSD_APAGE_COLLECTION_FIRST + 3,
+	OSD_APAGE_COLLECTION_SECURITY	= OSD_APAGE_COLLECTION_FIRST + 5,
+	OSD_APAGE_COLLECTION_LAST	= 0x8FFFFFFF,
+
+	OSD_APAGE_ROOT_FIRST		= 0x90000000,
+	OSD_APAGE_ROOT_DIRECTORY	= OSD_APAGE_ROOT_FIRST + 0,
+	OSD_APAGE_ROOT_INFORMATION	= OSD_APAGE_ROOT_FIRST + 1,
+	OSD_APAGE_ROOT_QUOTAS		= OSD_APAGE_ROOT_FIRST + 2,
+	OSD_APAGE_ROOT_TIMESTAMP	= OSD_APAGE_ROOT_FIRST + 3,
+	OSD_APAGE_ROOT_SECURITY		= OSD_APAGE_ROOT_FIRST + 5,
+	OSD_APAGE_ROOT_LAST		= 0xBFFFFFFF,
+
+	OSD_APAGE_RESERVED_TYPE_FIRST	= 0xC0000000,
+	OSD_APAGE_RESERVED_TYPE_LAST	= 0xEFFFFFFF,
+
+	OSD_APAGE_COMMON_FIRST		= 0xF0000000,
+	OSD_APAGE_COMMON_LAST		= 0xFFFFFFFE,
+
+	OSD_APAGE_REQUEST_ALL		= 0xFFFFFFFF,
+};
+
+/* subcategories of attr pages within each range above */
+enum {
+	OSD_APAGE_STD_FIRST		= 0x0,
+	OSD_APAGE_STD_DIRECTORY		= 0,
+	OSD_APAGE_STD_INFORMATION	= 1,
+	OSD_APAGE_STD_QUOTAS		= 2,
+	OSD_APAGE_STD_TIMESTAMP		= 3,
+	OSD_APAGE_STD_COLLECTIONS	= 4,
+	OSD_APAGE_STD_POLICY_SECURITY	= 5,
+	OSD_APAGE_STD_LAST		= 0x0000007F,
+
+	OSD_APAGE_RESERVED_FIRST	= 0x00000080,
+	OSD_APAGE_RESERVED_LAST		= 0x00007FFF,
+
+	OSD_APAGE_OTHER_STD_FIRST	= 0x00008000,
+	OSD_APAGE_OTHER_STD_LAST	= 0x0000EFFF,
+
+	OSD_APAGE_PUBLIC_FIRST		= 0x0000F000,
+	OSD_APAGE_PUBLIC_LAST		= 0x0000FFFF,
+
+	OSD_APAGE_APP_DEFINED_FIRST	= 0x00010000,
+	OSD_APAGE_APP_DEFINED_LAST	= 0x1FFFFFFF,
+
+	OSD_APAGE_VENDOR_SPECIFIC_FIRST	= 0x20000000,
+	OSD_APAGE_VENDOR_SPECIFIC_LAST	= 0x2FFFFFFF,
+};
+
+enum {
+	OSD_ATTR_PAGE_IDENTIFICATION = 0, /* in all pages 40 bytes */
+};
+
+struct page_identification {
+	u8 vendor_identification[8];
+	u8 page_identification[32];
+}  __packed;
+
+struct osd_attr_page_header {
+	__be32 page_number;
+	__be32 page_length;
+} __packed;
+
+/* 7.1.2.8 Root Information attributes page (OSD_APAGE_ROOT_INFORMATION) */
+enum {
+	OSD_ATTR_RI_OSD_SYSTEM_ID            = 0x3,   /* 20       */
+	OSD_ATTR_RI_VENDOR_IDENTIFICATION    = 0x4,   /* 8        */
+	OSD_ATTR_RI_PRODUCT_IDENTIFICATION   = 0x5,   /* 16       */
+	OSD_ATTR_RI_PRODUCT_MODEL            = 0x6,   /* 32       */
+	OSD_ATTR_RI_PRODUCT_REVISION_LEVEL   = 0x7,   /* 4        */
+	OSD_ATTR_RI_PRODUCT_SERIAL_NUMBER    = 0x8,   /* variable */
+	OSD_ATTR_RI_OSD_NAME                 = 0x9,   /* variable */
+	OSD_ATTR_RI_TOTAL_CAPACITY           = 0x80,  /* 8        */
+	OSD_ATTR_RI_USED_CAPACITY            = 0x81,  /* 8        */
+	OSD_ATTR_RI_NUMBER_OF_PARTITIONS     = 0xC0,  /* 8        */
+	OSD_ATTR_RI_CLOCK                    = 0x100, /* 6        */
+};
+/* Root_Information_attributes_page does not have a get_page structure */
+
+/* 7.1.2.9 Partition Information attributes page
+ * (OSD_APAGE_PARTITION_INFORMATION)
+ */
+enum {
+	OSD_ATTR_PI_PARTITION_ID            = 0x1,     /* 8        */
+	OSD_ATTR_PI_USERNAME                = 0x9,     /* variable */
+	OSD_ATTR_PI_USED_CAPACITY           = 0x81,    /* 8        */
+	OSD_ATTR_PI_NUMBER_OF_OBJECTS       = 0xC1,    /* 8        */
+};
+/* Partition Information attributes page does not have a get_page structure */
+
+/* 7.1.2.10 Collection Information attributes page
+ * (OSD_APAGE_COLLECTION_INFORMATION)
+ */
+enum {
+	OSD_ATTR_CI_PARTITION_ID           = 0x1,       /* 8        */
+	OSD_ATTR_CI_COLLECTION_OBJECT_ID   = 0x2,       /* 8        */
+	OSD_ATTR_CI_USERNAME               = 0x9,       /* variable */
+	OSD_ATTR_CI_USED_CAPACITY          = 0x81,      /* 8        */
+};
+/* Collection Information attributes page does not have a get_page structure */
+
+/* 7.1.2.11 User Object Information attributes page
+ * (OSD_APAGE_OBJECT_INFORMATION)
+ */
+enum {
+	OSD_ATTR_OI_PARTITION_ID         = 0x1,       /* 8        */
+	OSD_ATTR_OI_OBJECT_ID            = 0x2,       /* 8        */
+	OSD_ATTR_OI_USERNAME             = 0x9,       /* variable */
+	OSD_ATTR_OI_USED_CAPACITY        = 0x81,      /* 8        */
+	OSD_ATTR_OI_LOGICAL_LENGTH       = 0x82,      /* 8        */
+};
+/* Object Information attributes page does not have a get_page structure */
+
+/* 7.1.2.12 Root Quotas attributes page (OSD_APAGE_ROOT_QUOTAS) */
+enum {
+	OSD_ATTR_RQ_DEFAULT_MAXIMUM_USER_OBJECT_LENGTH     = 0x1,      /* 8  */
+	OSD_ATTR_RQ_PARTITION_CAPACITY_QUOTA               = 0x10001,  /* 8  */
+	OSD_ATTR_RQ_PARTITION_OBJECT_COUNT                 = 0x10002,  /* 8  */
+	OSD_ATTR_RQ_PARTITION_COLLECTIONS_PER_USER_OBJECT  = 0x10081,  /* 4  */
+	OSD_ATTR_RQ_PARTITION_COUNT                        = 0x20002,  /* 8  */
+};
+
+struct Root_Quotas_attributes_page {
+	struct osd_attr_page_header hdr; /* id=R+2, size=0x24 */
+	__be64 default_maximum_user_object_length;
+	__be64 partition_capacity_quota;
+	__be64 partition_object_count;
+	__be64 partition_collections_per_user_object;
+	__be64 partition_count;
+}  __packed;
+
+/* 7.1.2.13 Partition Quotas attributes page (OSD_APAGE_PARTITION_QUOTAS)*/
+enum {
+	OSD_ATTR_PQ_DEFAULT_MAXIMUM_USER_OBJECT_LENGTH  = 0x1,        /* 8 */
+	OSD_ATTR_PQ_CAPACITY_QUOTA                      = 0x10001,    /* 8 */
+	OSD_ATTR_PQ_OBJECT_COUNT                        = 0x10002,    /* 8 */
+	OSD_ATTR_PQ_COLLECTIONS_PER_USER_OBJECT         = 0x10081,    /* 4 */
+};
+
+struct Partition_Quotas_attributes_page {
+	struct osd_attr_page_header hdr; /* id=P+2, size=0x1C */
+	__be64 default_maximum_user_object_length;
+	__be64 capacity_quota;
+	__be64 object_count;
+	__be64 collections_per_user_object;
+}  __packed;
+
+/* 7.1.2.14 User Object Quotas attributes page (OSD_APAGE_OBJECT_QUOTAS) */
+enum {
+	OSD_ATTR_OQ_MAXIMUM_LENGTH  = 0x1,        /* 8 */
+};
+
+struct Object_Quotas_attributes_page {
+	struct osd_attr_page_header hdr; /* id=U+2, size=0x8 */
+	__be64 maximum_length;
+}  __packed;
+
+/* 7.1.2.15 Root Timestamps attributes page (OSD_APAGE_ROOT_TIMESTAMP) */
+enum {
+	OSD_ATTR_RT_ATTRIBUTES_ACCESSED_TIME  = 0x2,        /* 6 */
+	OSD_ATTR_RT_ATTRIBUTES_MODIFIED_TIME  = 0x3,        /* 6 */
+	OSD_ATTR_RT_TIMESTAMP_BYPASS          = 0xFFFFFFFE, /* 1 */
+};
+
+struct root_timestamps_attributes_page {
+	struct osd_attr_page_header hdr; /* id=R+3, size=0xD */
+	struct osd_timestamp attributes_accessed_time;
+	struct osd_timestamp attributes_modified_time;
+	u8 timestamp_bypass;
+}  __packed;
+
+/* 7.1.2.16 Partition Timestamps attributes page
+ * (OSD_APAGE_PARTITION_TIMESTAMP)
+ */
+enum {
+	OSD_ATTR_PT_CREATED_TIME              = 0x1,        /* 6 */
+	OSD_ATTR_PT_ATTRIBUTES_ACCESSED_TIME  = 0x2,        /* 6 */
+	OSD_ATTR_PT_ATTRIBUTES_MODIFIED_TIME  = 0x3,        /* 6 */
+	OSD_ATTR_PT_DATA_ACCESSED_TIME        = 0x4,        /* 6 */
+	OSD_ATTR_PT_DATA_MODIFIED_TIME        = 0x5,        /* 6 */
+	OSD_ATTR_PT_TIMESTAMP_BYPASS          = 0xFFFFFFFE, /* 1 */
+};
+
+struct partition_timestamps_attributes_page {
+	struct osd_attr_page_header hdr; /* id=P+3, size=0x1F */
+	struct osd_timestamp created_time;
+	struct osd_timestamp attributes_accessed_time;
+	struct osd_timestamp attributes_modified_time;
+	struct osd_timestamp data_accessed_time;
+	struct osd_timestamp data_modified_time;
+	u8 timestamp_bypass;
+}  __packed;
+
+/* 7.1.2.17/18 Collection/Object Timestamps attributes page
+ * (OSD_APAGE_COLLECTION_TIMESTAMP/OSD_APAGE_OBJECT_TIMESTAMP)
+ */
+enum {
+	OSD_ATTR_OT_CREATED_TIME              = 0x1,        /* 6 */
+	OSD_ATTR_OT_ATTRIBUTES_ACCESSED_TIME  = 0x2,        /* 6 */
+	OSD_ATTR_OT_ATTRIBUTES_MODIFIED_TIME  = 0x3,        /* 6 */
+	OSD_ATTR_OT_DATA_ACCESSED_TIME        = 0x4,        /* 6 */
+	OSD_ATTR_OT_DATA_MODIFIED_TIME        = 0x5,        /* 6 */
+};
+
+/* same for collection */
+struct object_timestamps_attributes_page {
+	struct osd_attr_page_header hdr; /* id=C+3/3, size=0x1E */
+	struct osd_timestamp created_time;
+	struct osd_timestamp attributes_accessed_time;
+	struct osd_timestamp attributes_modified_time;
+	struct osd_timestamp data_accessed_time;
+	struct osd_timestamp data_modified_time;
+}  __packed;
+
+/* 7.1.2.19 Collections attributes page */
+/* TBD */
+
+/* 7.1.2.20 Root Policy/Security attributes page (OSD_APAGE_ROOT_SECURITY) */
+enum {
+	OSD_ATTR_RS_DEFAULT_SECURITY_METHOD           = 0x1,       /* 1      */
+	OSD_ATTR_RS_OLDEST_VALID_NONCE_LIMIT          = 0x2,       /* 6      */
+	OSD_ATTR_RS_NEWEST_VALID_NONCE_LIMIT          = 0x3,       /* 6      */
+	OSD_ATTR_RS_PARTITION_DEFAULT_SECURITY_METHOD = 0x6,       /* 1      */
+	OSD_ATTR_RS_SUPPORTED_SECURITY_METHODS        = 0x7,       /* 2      */
+	OSD_ATTR_RS_ADJUSTABLE_CLOCK                  = 0x9,       /* 6      */
+	OSD_ATTR_RS_MASTER_KEY_IDENTIFIER             = 0x7FFD,    /* 0 or 7 */
+	OSD_ATTR_RS_ROOT_KEY_IDENTIFIER               = 0x7FFE,    /* 0 or 7 */
+	OSD_ATTR_RS_SUPPORTED_INTEGRITY_ALGORITHM_0   = 0x80000000,/* 1,(x16)*/
+	OSD_ATTR_RS_SUPPORTED_DH_GROUP_0              = 0x80000010,/* 1,(x16)*/
+};
+
+struct root_security_attributes_page {
+	struct osd_attr_page_header hdr; /* id=R+5, size=0x3F */
+	u8 default_security_method;
+	u8 partition_default_security_method;
+	__be16 supported_security_methods;
+	u8 mki_valid_rki_valid;
+	struct osd_timestamp oldest_valid_nonce_limit;
+	struct osd_timestamp newest_valid_nonce_limit;
+	struct osd_timestamp adjustable_clock;
+	u8 master_key_identifier[32-25];
+	u8 root_key_identifier[39-32];
+	u8 supported_integrity_algorithm[16];
+	u8 supported_dh_group[16];
+}  __packed;
+
+/* 7.1.2.21 Partition Policy/Security attributes page
+ * (OSD_APAGE_PARTITION_SECURITY)
+ */
+enum {
+	OSD_ATTR_PS_DEFAULT_SECURITY_METHOD        = 0x1,        /* 1      */
+	OSD_ATTR_PS_OLDEST_VALID_NONCE             = 0x2,        /* 6      */
+	OSD_ATTR_PS_NEWEST_VALID_NONCE             = 0x3,        /* 6      */
+	OSD_ATTR_PS_REQUEST_NONCE_LIST_DEPTH       = 0x4,        /* 2      */
+	OSD_ATTR_PS_FROZEN_WORKING_KEY_BIT_MASK    = 0x5,        /* 2      */
+	OSD_ATTR_PS_PARTITION_KEY_IDENTIFIER       = 0x7FFF,     /* 0 or 7 */
+	OSD_ATTR_PS_WORKING_KEY_IDENTIFIER_FIRST   = 0x8000,     /* 0 or 7 */
+	OSD_ATTR_PS_WORKING_KEY_IDENTIFIER_LAST    = 0x800F,     /* 0 or 7 */
+	OSD_ATTR_PS_POLICY_ACCESS_TAG              = 0x40000001, /* 4      */
+	OSD_ATTR_PS_USER_OBJECT_POLICY_ACCESS_TAG  = 0x40000002, /* 4      */
+};
+
+struct partition_security_attributes_page {
+	struct osd_attr_page_header hdr; /* id=p+5, size=0x8f */
+	u8 reserved[3];
+	u8 default_security_method;
+	struct osd_timestamp oldest_valid_nonce;
+	struct osd_timestamp newest_valid_nonce;
+	__be16 request_nonce_list_depth;
+	__be16 frozen_working_key_bit_mask;
+	__be32 policy_access_tag;
+	__be32 user_object_policy_access_tag;
+	u8 pki_valid;
+	__be16 wki_00_0f_vld;
+	struct osd_key_identifier partition_key_identifier;
+	struct osd_key_identifier working_key_identifiers[16];
+}  __packed;
+
+/* 7.1.2.22/23 Collection/Object Policy-Security attributes page
+ * (OSD_APAGE_COLLECTION_SECURITY/OSD_APAGE_OBJECT_SECURITY)
+ */
+enum {
+	OSD_ATTR_OS_POLICY_ACCESS_TAG              = 0x40000001, /* 4      */
+};
+
+struct object_security_attributes_page {
+	struct osd_attr_page_header hdr; /* id=C+5/5, size=4 */
+	__be32 policy_access_tag;
+}  __packed;
+
+#endif /*ndef __OSD_ATTRIBUTES_H__*/
diff --git a/include/scsi/osd_initiator.h b/include/scsi/osd_initiator.h
new file mode 100644
index 000000000000..f791b39f5597
--- /dev/null
+++ b/include/scsi/osd_initiator.h
@@ -0,0 +1,379 @@
+/*
+ * osd_initiator.h - OSD initiator API definition
+ *
+ * Copyright (C) 2008 Panasas Inc.  All rights reserved.
+ *
+ * Authors:
+ *   Boaz Harrosh <bharrosh@panasas.com>
+ *   Benny Halevy <bhalevy@panasas.com>
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License version 2
+ *
+ */
+#ifndef __OSD_INITIATOR_H__
+#define __OSD_INITIATOR_H__
+
+#include "osd_protocol.h"
+#include "osd_types.h"
+
+#include <linux/blkdev.h>
+
+/* Note: "NI" in comments below means "Not Implemented yet" */
+
+/* Configure of code:
+ * #undef if you *don't* want OSD v1 support in runtime.
+ * If #defined the initiator will dynamically configure to encode OSD v1
+ * CDB's if the target is detected to be OSD v1 only.
+ * OSD v2 only commands, options, and attributes will be ignored if target
+ * is v1 only.
+ * If #defined will result in bigger/slower code (OK Slower maybe not)
+ * Q: Should this be CONFIG_SCSI_OSD_VER1_SUPPORT and set from Kconfig?
+ */
+#define OSD_VER1_SUPPORT y
+
+enum osd_std_version {
+	OSD_VER_NONE = 0,
+	OSD_VER1 = 1,
+	OSD_VER2 = 2,
+};
+
+/*
+ * Object-based Storage Device.
+ * This object represents an OSD device.
+ * It is not a full linux device in any way. It is only
+ * a place to hang resources associated with a Linux
+ * request Q and some default properties.
+ */
+struct osd_dev {
+	struct scsi_device *scsi_device;
+	unsigned def_timeout;
+
+#ifdef OSD_VER1_SUPPORT
+	enum osd_std_version version;
+#endif
+};
+
+/* Retrieve/return osd_dev(s) for use by Kernel clients */
+struct osd_dev *osduld_path_lookup(const char *dev_name); /*Use IS_ERR/ERR_PTR*/
+void osduld_put_device(struct osd_dev *od);
+
+/* These are called by uld at probe time */
+void osd_dev_init(struct osd_dev *, struct scsi_device *scsi_dev);
+void osd_dev_fini(struct osd_dev *);
+
+/* some hi level device operations */
+int osd_auto_detect_ver(struct osd_dev *, void *caps);    /* GFP_KERNEL */
+
+/* we might want to use function vector in the future */
+static inline void osd_dev_set_ver(struct osd_dev *dev, enum osd_std_version v)
+{
+#ifdef OSD_VER1_SUPPORT
+	dev->version = v;
+#endif
+}
+
+struct osd_request;
+typedef void (osd_req_done_fn)(struct osd_request *, void *);
+
+struct osd_request {
+	struct osd_cdb cdb;
+	struct osd_data_out_integrity_info out_data_integ;
+	struct osd_data_in_integrity_info in_data_integ;
+
+	struct osd_dev *osd_dev;
+	struct request *request;
+
+	struct _osd_req_data_segment {
+		void *buff;
+		unsigned alloc_size; /* 0 here means not allocated by us */
+		unsigned total_bytes;
+	} set_attr, enc_get_attr, get_attr;
+
+	struct _osd_io_info {
+		struct bio *bio;
+		u64 total_bytes;
+		struct request *req;
+		struct _osd_req_data_segment *last_seg;
+		u8 *pad_buff;
+	} out, in;
+
+	gfp_t alloc_flags;
+	unsigned timeout;
+	unsigned retries;
+	u8 sense[OSD_MAX_SENSE_LEN];
+	enum osd_attributes_mode attributes_mode;
+
+	osd_req_done_fn *async_done;
+	void *async_private;
+	int async_error;
+};
+
+/* OSD Version control */
+static inline bool osd_req_is_ver1(struct osd_request *or)
+{
+#ifdef OSD_VER1_SUPPORT
+	return or->osd_dev->version == OSD_VER1;
+#else
+	return false;
+#endif
+}
+
+/**
+ * How to use the osd library:
+ *
+ * osd_start_request
+ *	Allocates a request.
+ *
+ * osd_req_*
+ *	Call one of, to encode the desired operation.
+ *
+ * osd_add_{get,set}_attr
+ *	Optionally add attributes to the CDB, list or page mode.
+ *
+ * osd_finalize_request
+ *	Computes final data out/in offsets and signs the request,
+ *	making it ready for execution.
+ *
+ * osd_execute_request
+ *	May be called to execute it through the block layer. Other wise submit
+ *	the associated block request in some other way.
+ *
+ * After execution:
+ * osd_req_decode_sense
+ *	Decodes sense information to verify execution results.
+ *
+ * osd_req_decode_get_attr
+ *	Retrieve osd_add_get_attr_list() values if used.
+ *
+ * osd_end_request
+ *	Must be called to deallocate the request.
+ */
+
+/**
+ * osd_start_request - Allocate and initialize an osd_request
+ *
+ * @osd_dev:    OSD device that holds the scsi-device and default values
+ *              that the request is associated with.
+ * @gfp:        The allocation flags to use for request allocation, and all
+ *              subsequent allocations. This will be stored at
+ *              osd_request->alloc_flags, can be changed by user later
+ *
+ * Allocate osd_request and initialize all members to the
+ * default/initial state.
+ */
+struct osd_request *osd_start_request(struct osd_dev *, gfp_t gfp);
+
+enum osd_req_options {
+	OSD_REQ_FUA = 0x08,	/* Force Unit Access */
+	OSD_REQ_DPO = 0x10,	/* Disable Page Out */
+
+	OSD_REQ_BYPASS_TIMESTAMPS = 0x80,
+};
+
+/**
+ * osd_finalize_request - Sign request and prepare request for execution
+ *
+ * @or:		osd_request to prepare
+ * @options:	combination of osd_req_options bit flags or 0.
+ * @cap	A Pointer to an OSD_CAP_LEN bytes buffer that is received from
+ *              The security manager as capabilities for this cdb.
+ * @cap_key	The cryptographic key used to sign the cdb/data. Can be null
+ *              if NOSEC is used.
+ *
+ * The actual request and bios are only allocated here, so are the get_attr
+ * buffers that will receive the returned attributes. Copy's @cap to cdb.
+ * Sign the cdb/data with @cap_key.
+ */
+int osd_finalize_request(struct osd_request *or,
+	u8 options, const void *cap, const u8 *cap_key);
+
+/**
+ * osd_execute_request - Execute the request synchronously through
+ *                       the block-layer
+ * @or:		osd_request to Executed
+ *
+ * Calls blk_execute_rq to q the command and waits for completion.
+ */
+int osd_execute_request(struct osd_request *or);
+
+/**
+ * osd_execute_request_async - submits the request for execution through
+ *                             the block-layer without waitting.
+ * @or:                      - osd_request to Executed
+ * @done: (Optional)         - Called at end of execution
+ * @private:                 - Will be passes to @done function
+ *
+ * Calls blk_execute_rq_nowait to q the command. When execution is done
+ * Optionally calles @done with @private as parameter. or->async_error has the
+ * Return code
+ */
+int osd_execute_request_async(struct osd_request *or,
+	osd_req_done_fn *done, void *private);
+
+/**
+ * osd_end_request - return osd_request to free store
+ *
+ * @or:		osd_request to free
+ *
+ * Deallocate all osd_request resources (struct req's, BIOs, buffers, etc.)
+ */
+void osd_end_request(struct osd_request *or);
+
+/*
+ * CDB Encoding
+ *
+ * Note: call only one of the following methods.
+ */
+
+/*
+ * Device commands
+ */
+void osd_req_set_master_seed_xchg(struct osd_request *, ...);/* NI */
+void osd_req_set_master_key(struct osd_request *, ...);/* NI */
+
+void osd_req_format(struct osd_request *, u64 tot_capacity);
+
+/* list all partitions
+ * @list header must be initialized to zero on first run.
+ *
+ * Call osd_is_obj_list_done() to find if we got the complete list.
+ */
+int osd_req_list_dev_partitions(struct osd_request *,
+	osd_id initial_id, struct osd_obj_id_list *list, unsigned nelem);
+
+void osd_req_flush_obsd(struct osd_request *,
+	enum osd_options_flush_scope_values);
+
+void osd_req_perform_scsi_command(struct osd_request *,
+	const u8 *cdb, ...);/* NI */
+void osd_req_task_management(struct osd_request *, ...);/* NI */
+
+/*
+ * Partition commands
+ */
+void osd_req_create_partition(struct osd_request *, osd_id partition);
+void osd_req_remove_partition(struct osd_request *, osd_id partition);
+
+void osd_req_set_partition_key(struct osd_request *,
+	osd_id partition, u8 new_key_id[OSD_CRYPTO_KEYID_SIZE],
+	u8 seed[OSD_CRYPTO_SEED_SIZE]);/* NI */
+
+/* list all collections in the partition
+ * @list header must be init to zero on first run.
+ *
+ * Call osd_is_obj_list_done() to find if we got the complete list.
+ */
+int osd_req_list_partition_collections(struct osd_request *,
+	osd_id partition, osd_id initial_id, struct osd_obj_id_list *list,
+	unsigned nelem);
+
+/* list all objects in the partition
+ * @list header must be init to zero on first run.
+ *
+ * Call osd_is_obj_list_done() to find if we got the complete list.
+ */
+int osd_req_list_partition_objects(struct osd_request *,
+	osd_id partition, osd_id initial_id, struct osd_obj_id_list *list,
+	unsigned nelem);
+
+void osd_req_flush_partition(struct osd_request *,
+	osd_id partition, enum osd_options_flush_scope_values);
+
+/*
+ * Collection commands
+ */
+void osd_req_create_collection(struct osd_request *,
+	const struct osd_obj_id *);/* NI */
+void osd_req_remove_collection(struct osd_request *,
+	const struct osd_obj_id *);/* NI */
+
+/* list all objects in the collection */
+int osd_req_list_collection_objects(struct osd_request *,
+	const struct osd_obj_id *, osd_id initial_id,
+	struct osd_obj_id_list *list, unsigned nelem);
+
+/* V2 only filtered list of objects in the collection */
+void osd_req_query(struct osd_request *, ...);/* NI */
+
+void osd_req_flush_collection(struct osd_request *,
+	const struct osd_obj_id *, enum osd_options_flush_scope_values);
+
+void osd_req_get_member_attrs(struct osd_request *, ...);/* V2-only NI */
+void osd_req_set_member_attrs(struct osd_request *, ...);/* V2-only NI */
+
+/*
+ * Object commands
+ */
+void osd_req_create_object(struct osd_request *, struct osd_obj_id *);
+void osd_req_remove_object(struct osd_request *, struct osd_obj_id *);
+
+void osd_req_write(struct osd_request *,
+	const struct osd_obj_id *, struct bio *data_out, u64 offset);
+void osd_req_append(struct osd_request *,
+	const struct osd_obj_id *, struct bio *data_out);/* NI */
+void osd_req_create_write(struct osd_request *,
+	const struct osd_obj_id *, struct bio *data_out, u64 offset);/* NI */
+void osd_req_clear(struct osd_request *,
+	const struct osd_obj_id *, u64 offset, u64 len);/* NI */
+void osd_req_punch(struct osd_request *,
+	const struct osd_obj_id *, u64 offset, u64 len);/* V2-only NI */
+
+void osd_req_flush_object(struct osd_request *,
+	const struct osd_obj_id *, enum osd_options_flush_scope_values,
+	/*V2*/ u64 offset, /*V2*/ u64 len);
+
+void osd_req_read(struct osd_request *,
+	const struct osd_obj_id *, struct bio *data_in, u64 offset);
+
+/*
+ * Root/Partition/Collection/Object Attributes commands
+ */
+
+/* get before set */
+void osd_req_get_attributes(struct osd_request *, const struct osd_obj_id *);
+
+/* set before get */
+void osd_req_set_attributes(struct osd_request *, const struct osd_obj_id *);
+
+/*
+ * Attributes appended to most commands
+ */
+
+/* Attributes List mode (or V2 CDB) */
+  /*
+   * TODO: In ver2 if at finalize time only one attr was set and no gets,
+   * then the Attributes CDB mode is used automatically to save IO.
+   */
+
+/* set a list of attributes. */
+int osd_req_add_set_attr_list(struct osd_request *,
+	const struct osd_attr *, unsigned nelem);
+
+/* get a list of attributes */
+int osd_req_add_get_attr_list(struct osd_request *,
+	const struct osd_attr *, unsigned nelem);
+
+/*
+ * Attributes list decoding
+ * Must be called after osd_request.request was executed
+ * It is called in a loop to decode the returned get_attr
+ * (see osd_add_get_attr)
+ */
+int osd_req_decode_get_attr_list(struct osd_request *,
+	struct osd_attr *, int *nelem, void **iterator);
+
+/* Attributes Page mode */
+
+/*
+ * Read an attribute page and optionally set one attribute
+ *
+ * Retrieves the attribute page directly to a user buffer.
+ * @attr_page_data shall stay valid until end of execution.
+ * See osd_attributes.h for common page structures
+ */
+int osd_req_add_get_attr_page(struct osd_request *,
+	u32 page_id, void *attr_page_data, unsigned max_page_len,
+	const struct osd_attr *set_one);
+
+#endif /* __OSD_LIB_H__ */
diff --git a/include/scsi/osd_protocol.h b/include/scsi/osd_protocol.h
new file mode 100644
index 000000000000..ef7505e4b9ff
--- /dev/null
+++ b/include/scsi/osd_protocol.h
@@ -0,0 +1,575 @@
+/*
+ * osd_protocol.h - OSD T10 standard C definitions.
+ *
+ * Copyright (C) 2008 Panasas Inc.  All rights reserved.
+ *
+ * Authors:
+ *   Boaz Harrosh <bharrosh@panasas.com>
+ *   Benny Halevy <bhalevy@panasas.com>
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License version 2
+ *
+ * This file contains types and constants that are defined by the protocol
+ * Note: All names and symbols are taken from the OSD standard's text.
+ */
+#ifndef __OSD_PROTOCOL_H__
+#define __OSD_PROTOCOL_H__
+
+#include <linux/types.h>
+#include <asm/unaligned.h>
+#include <scsi/scsi.h>
+
+enum {
+	OSDv1_ADDITIONAL_CDB_LENGTH = 192,
+	OSDv1_TOTAL_CDB_LEN = OSDv1_ADDITIONAL_CDB_LENGTH + 8,
+	OSDv1_CAP_LEN = 80,
+	/* Latest supported version */
+/* 	OSD_ADDITIONAL_CDB_LENGTH = 216,*/
+	OSD_ADDITIONAL_CDB_LENGTH =
+		OSDv1_ADDITIONAL_CDB_LENGTH, /* FIXME: Pete rev-001 sup */
+	OSD_TOTAL_CDB_LEN = OSD_ADDITIONAL_CDB_LENGTH + 8,
+/* 	OSD_CAP_LEN = 104,*/
+	OSD_CAP_LEN = OSDv1_CAP_LEN,/* FIXME: Pete rev-001 sup */
+
+	OSD_SYSTEMID_LEN = 20,
+	OSD_CRYPTO_KEYID_SIZE = 20,
+	OSD_CRYPTO_SEED_SIZE = 4,
+	OSD_CRYPTO_NONCE_SIZE = 12,
+	OSD_MAX_SENSE_LEN = 252, /* from SPC-3 */
+
+	OSD_PARTITION_FIRST_ID = 0x10000,
+	OSD_OBJECT_FIRST_ID = 0x10000,
+};
+
+/* (osd-r10 5.2.4)
+ * osd2r03: 5.2.3 Caching control bits
+ */
+enum osd_options_byte {
+	OSD_CDB_FUA = 0x08,	/* Force Unit Access */
+	OSD_CDB_DPO = 0x10,	/* Disable Page Out */
+};
+
+/*
+ * osd2r03: 5.2.5 Isolation.
+ * First 3 bits, V2-only.
+ * Also for attr 110h "default isolation method" at Root Information page
+ */
+enum osd_options_byte_isolation {
+	OSD_ISOLATION_DEFAULT = 0,
+	OSD_ISOLATION_NONE = 1,
+	OSD_ISOLATION_STRICT = 2,
+	OSD_ISOLATION_RANGE = 4,
+	OSD_ISOLATION_FUNCTIONAL = 5,
+	OSD_ISOLATION_VENDOR = 7,
+};
+
+/* (osd-r10: 6.7)
+ * osd2r03: 6.8 FLUSH, FLUSH COLLECTION, FLUSH OSD, FLUSH PARTITION
+ */
+enum osd_options_flush_scope_values {
+	OSD_CDB_FLUSH_ALL = 0,
+	OSD_CDB_FLUSH_ATTR_ONLY = 1,
+
+	OSD_CDB_FLUSH_ALL_RECURSIVE = 2,
+	/* V2-only */
+	OSD_CDB_FLUSH_ALL_RANGE = 2,
+};
+
+/* osd2r03: 5.2.10 Timestamps control */
+enum {
+	OSD_CDB_NORMAL_TIMESTAMPS = 0,
+	OSD_CDB_BYPASS_TIMESTAMPS = 0x7f,
+};
+
+/* (osd-r10: 5.2.2.1)
+ * osd2r03: 5.2.4.1 Get and set attributes CDB format selection
+ *	2 bits at second nibble of command_specific_options byte
+ */
+enum osd_attributes_mode {
+	/* V2-only */
+	OSD_CDB_SET_ONE_ATTR = 0x10,
+
+	OSD_CDB_GET_ATTR_PAGE_SET_ONE = 0x20,
+	OSD_CDB_GET_SET_ATTR_LISTS = 0x30,
+
+	OSD_CDB_GET_SET_ATTR_MASK = 0x30,
+};
+
+/* (osd-r10: 4.12.5)
+ * osd2r03: 4.14.5 Data-In and Data-Out buffer offsets
+ *	byte offset = mantissa * (2^(exponent+8))
+ *	struct {
+ *		unsigned mantissa: 28;
+ *		int exponent: 04;
+ *	}
+ */
+typedef __be32 __bitwise osd_cdb_offset;
+
+enum {
+	OSD_OFFSET_UNUSED = 0xFFFFFFFF,
+	OSD_OFFSET_MAX_BITS = 28,
+
+	OSDv1_OFFSET_MIN_SHIFT = 8,
+	OSD_OFFSET_MIN_SHIFT = 3,
+	OSD_OFFSET_MAX_SHIFT = 16,
+};
+
+/* Return the smallest allowed encoded offset that contains @offset.
+ *
+ * The actual encoded offset returned is @offset + *padding.
+ * (up to max_shift, non-inclusive)
+ */
+osd_cdb_offset __osd_encode_offset(u64 offset, unsigned *padding,
+	int min_shift, int max_shift);
+
+/* Minimum alignment is 256 bytes
+ * Note: Seems from std v1 that exponent can be from 0+8 to 0xE+8 (inclusive)
+ * which is 8 to 23 but IBM code restricts it to 16, so be it.
+ */
+static inline osd_cdb_offset osd_encode_offset_v1(u64 offset, unsigned *padding)
+{
+	return __osd_encode_offset(offset, padding,
+				OSDv1_OFFSET_MIN_SHIFT, OSD_OFFSET_MAX_SHIFT);
+}
+
+/* Minimum 8 bytes alignment
+ * Same as v1 but since exponent can be signed than a less than
+ * 256 alignment can be reached with small offsets (<2GB)
+ */
+static inline osd_cdb_offset osd_encode_offset_v2(u64 offset, unsigned *padding)
+{
+	return __osd_encode_offset(offset, padding,
+				   OSD_OFFSET_MIN_SHIFT, OSD_OFFSET_MAX_SHIFT);
+}
+
+/* osd2r03: 5.2.1 Overview */
+struct osd_cdb_head {
+	struct scsi_varlen_cdb_hdr varlen_cdb;
+/*10*/	u8		options;
+	u8		command_specific_options;
+	u8		timestamp_control;
+/*13*/	u8		reserved1[3];
+/*16*/	__be64		partition;
+/*24*/	__be64		object;
+/*32*/	union { /* V1 vs V2 alignment differences */
+		struct __osdv1_cdb_addr_len {
+/*32*/			__be32 		list_identifier;/* Rarely used */
+/*36*/			__be64		length;
+/*44*/			__be64		start_address;
+		} __packed v1;
+
+		struct __osdv2_cdb_addr_len {
+			/* called allocation_length in some commands */
+/*32*/			__be64	length;
+/*40*/			__be64	start_address;
+/*48*/			__be32 list_identifier;/* Rarely used */
+		} __packed v2;
+	};
+/*52*/	union { /* selected attributes mode Page/List/Single */
+		struct osd_attributes_page_mode {
+/*52*/			__be32		get_attr_page;
+/*56*/			__be32		get_attr_alloc_length;
+/*60*/			osd_cdb_offset	get_attr_offset;
+
+/*64*/			__be32		set_attr_page;
+/*68*/			__be32		set_attr_id;
+/*72*/			__be32		set_attr_length;
+/*76*/			osd_cdb_offset	set_attr_offset;
+		} __packed attrs_page;
+
+		struct osd_attributes_list_mode {
+/*52*/			__be32		get_attr_desc_bytes;
+/*56*/			osd_cdb_offset	get_attr_desc_offset;
+
+/*60*/			__be32		get_attr_alloc_length;
+/*64*/			osd_cdb_offset	get_attr_offset;
+
+/*68*/			__be32		set_attr_bytes;
+/*72*/			osd_cdb_offset	set_attr_offset;
+			__be32 not_used;
+		} __packed attrs_list;
+
+		/* osd2r03:5.2.4.2 Set one attribute value using CDB fields */
+		struct osd_attributes_cdb_mode {
+/*52*/			__be32		set_attr_page;
+/*56*/			__be32		set_attr_id;
+/*60*/			__be16		set_attr_len;
+/*62*/			u8		set_attr_val[80-62];
+		} __packed attrs_cdb;
+/*52*/		u8 get_set_attributes_parameters[80-52];
+	};
+} __packed;
+/*80*/
+
+/*160 v1*/
+/*184 v2*/
+struct osd_security_parameters {
+/*160*/u8	integrity_check_value[OSD_CRYPTO_KEYID_SIZE];
+/*180*/u8	request_nonce[OSD_CRYPTO_NONCE_SIZE];
+/*192*/osd_cdb_offset	data_in_integrity_check_offset;
+/*196*/osd_cdb_offset	data_out_integrity_check_offset;
+} __packed;
+/*200 v1*/
+/*224 v2*/
+
+struct osdv1_cdb {
+	struct osd_cdb_head h;
+	u8 caps[OSDv1_CAP_LEN];
+	struct osd_security_parameters sec_params;
+} __packed;
+
+struct osdv2_cdb {
+	struct osd_cdb_head h;
+	u8 caps[OSD_CAP_LEN];
+	struct osd_security_parameters sec_params;
+} __packed;
+
+struct osd_cdb {
+	union {
+		struct osdv1_cdb v1;
+		struct osdv2_cdb v2;
+		u8 buff[OSD_TOTAL_CDB_LEN];
+	};
+} __packed;
+
+static inline struct osd_cdb_head *osd_cdb_head(struct osd_cdb *ocdb)
+{
+	return (struct osd_cdb_head *)ocdb->buff;
+}
+
+/* define both version actions
+ * Ex name = FORMAT_OSD we have OSD_ACT_FORMAT_OSD && OSDv1_ACT_FORMAT_OSD
+ */
+#define OSD_ACT___(Name, Num) \
+	OSD_ACT_##Name = __constant_cpu_to_be16(0x8880 + Num), \
+	OSDv1_ACT_##Name = __constant_cpu_to_be16(0x8800 + Num),
+
+/* V2 only actions */
+#define OSD_ACT_V2(Name, Num) \
+	OSD_ACT_##Name = __constant_cpu_to_be16(0x8880 + Num),
+
+#define OSD_ACT_V1_V2(Name, Num1, Num2) \
+	OSD_ACT_##Name = __constant_cpu_to_be16(Num2), \
+	OSDv1_ACT_##Name = __constant_cpu_to_be16(Num1),
+
+enum osd_service_actions {
+	OSD_ACT_V2(OBJECT_STRUCTURE_CHECK,	0x00)
+	OSD_ACT___(FORMAT_OSD,			0x01)
+	OSD_ACT___(CREATE,			0x02)
+	OSD_ACT___(LIST,			0x03)
+	OSD_ACT_V2(PUNCH,			0x04)
+	OSD_ACT___(READ,			0x05)
+	OSD_ACT___(WRITE,			0x06)
+	OSD_ACT___(APPEND,			0x07)
+	OSD_ACT___(FLUSH,			0x08)
+	OSD_ACT_V2(CLEAR,			0x09)
+	OSD_ACT___(REMOVE,			0x0A)
+	OSD_ACT___(CREATE_PARTITION,		0x0B)
+	OSD_ACT___(REMOVE_PARTITION,		0x0C)
+	OSD_ACT___(GET_ATTRIBUTES,		0x0E)
+	OSD_ACT___(SET_ATTRIBUTES,		0x0F)
+	OSD_ACT___(CREATE_AND_WRITE,		0x12)
+	OSD_ACT___(CREATE_COLLECTION,		0x15)
+	OSD_ACT___(REMOVE_COLLECTION,		0x16)
+	OSD_ACT___(LIST_COLLECTION,		0x17)
+	OSD_ACT___(SET_KEY,			0x18)
+	OSD_ACT___(SET_MASTER_KEY,		0x19)
+	OSD_ACT___(FLUSH_COLLECTION,		0x1A)
+	OSD_ACT___(FLUSH_PARTITION,		0x1B)
+	OSD_ACT___(FLUSH_OSD,			0x1C)
+
+	OSD_ACT_V2(QUERY,			0x20)
+	OSD_ACT_V2(REMOVE_MEMBER_OBJECTS,	0x21)
+	OSD_ACT_V2(GET_MEMBER_ATTRIBUTES,	0x22)
+	OSD_ACT_V2(SET_MEMBER_ATTRIBUTES,	0x23)
+	OSD_ACT_V2(READ_MAP,			0x31)
+
+	OSD_ACT_V1_V2(PERFORM_SCSI_COMMAND,	0x8F7E, 0x8F7C)
+	OSD_ACT_V1_V2(SCSI_TASK_MANAGEMENT,	0x8F7F, 0x8F7D)
+	/* 0x8F80 to 0x8FFF are Vendor specific */
+};
+
+/* osd2r03: 7.1.3.2 List entry format for retrieving attributes */
+struct osd_attributes_list_attrid {
+	__be32 page;
+	__be32 attr_id;
+} __packed;
+
+/*
+ * osd2r03: 7.1.3.3 List entry format for retrieved attributes and
+ *                  for setting attributes
+ * NOTE: v2 is 8-bytes aligned, v1 is not aligned.
+ */
+struct osd_attributes_list_element {
+	__be32 page;
+	__be32 attr_id;
+	__be16 attr_bytes;
+	u8 attr_val[0];
+} __packed;
+
+enum {
+	OSDv1_ATTRIBUTES_ELEM_ALIGN = 1,
+	OSD_ATTRIBUTES_ELEM_ALIGN = 8,
+};
+
+enum {
+	OSD_ATTR_LIST_ALL_PAGES = 0xFFFFFFFF,
+	OSD_ATTR_LIST_ALL_IN_PAGE = 0xFFFFFFFF,
+};
+
+static inline unsigned osdv1_attr_list_elem_size(unsigned len)
+{
+	return ALIGN(len + sizeof(struct osd_attributes_list_element),
+		     OSDv1_ATTRIBUTES_ELEM_ALIGN);
+}
+
+static inline unsigned osdv2_attr_list_elem_size(unsigned len)
+{
+	return ALIGN(len + sizeof(struct osd_attributes_list_element),
+		     OSD_ATTRIBUTES_ELEM_ALIGN);
+}
+
+/*
+ * osd2r03: 7.1.3 OSD attributes lists (Table 184) — List type values
+ */
+enum osd_attr_list_types {
+	OSD_ATTR_LIST_GET = 0x1, 	/* descriptors only */
+	OSD_ATTR_LIST_SET_RETRIEVE = 0x9, /*descriptors/values variable-length*/
+	OSD_V2_ATTR_LIST_MULTIPLE = 0xE,  /* ver2, Multiple Objects lists*/
+	OSD_V1_ATTR_LIST_CREATE_MULTIPLE = 0xF,/*ver1, used by create_multple*/
+};
+
+/* osd2r03: 7.1.3.4 Multi-object retrieved attributes format */
+struct osd_attributes_list_multi_header {
+	__be64 object_id;
+	u8 object_type; /* object_type enum below */
+	u8 reserved[5];
+	__be16 list_bytes;
+	/* followed by struct osd_attributes_list_element's */
+};
+
+struct osdv1_attributes_list_header {
+	u8 type;	/* low 4-bit only */
+	u8 pad;
+	__be16 list_bytes; /* Initiator shall set to Zero. Only set by target */
+	/*
+	 * type=9 followed by struct osd_attributes_list_element's
+	 * type=E followed by struct osd_attributes_list_multi_header's
+	 */
+} __packed;
+
+static inline unsigned osdv1_list_size(struct osdv1_attributes_list_header *h)
+{
+	return be16_to_cpu(h->list_bytes);
+}
+
+struct osdv2_attributes_list_header {
+	u8 type;	/* lower 4-bits only */
+	u8 pad[3];
+/*4*/	__be32 list_bytes; /* Initiator shall set to zero. Only set by target */
+	/*
+	 * type=9 followed by struct osd_attributes_list_element's
+	 * type=E followed by struct osd_attributes_list_multi_header's
+	 */
+} __packed;
+
+static inline unsigned osdv2_list_size(struct osdv2_attributes_list_header *h)
+{
+	return be32_to_cpu(h->list_bytes);
+}
+
+/* (osd-r10 6.13)
+ * osd2r03: 6.15 LIST (Table 79) LIST command parameter data.
+ *	for root_lstchg below
+ */
+enum {
+	OSD_OBJ_ID_LIST_PAR = 0x1, /* V1-only. Not used in V2 */
+	OSD_OBJ_ID_LIST_LSTCHG = 0x2,
+};
+
+/*
+ * osd2r03: 6.15.2 LIST command parameter data
+ * (Also for LIST COLLECTION)
+ */
+struct osd_obj_id_list {
+	__be64 list_bytes; /* bytes in list excluding list_bytes (-8) */
+	__be64 continuation_id;
+	__be32 list_identifier;
+	u8 pad[3];
+	u8 root_lstchg;
+	__be64 object_ids[0];
+} __packed;
+
+static inline bool osd_is_obj_list_done(struct osd_obj_id_list *list,
+	bool *is_changed)
+{
+	*is_changed = (0 != (list->root_lstchg & OSD_OBJ_ID_LIST_LSTCHG));
+	return 0 != list->continuation_id;
+}
+
+/*
+ * osd2r03: 4.12.4.5 The ALLDATA security method
+ */
+struct osd_data_out_integrity_info {
+	__be64 data_bytes;
+	__be64 set_attributes_bytes;
+	__be64 get_attributes_bytes;
+	__be64 integrity_check_value;
+} __packed;
+
+struct osd_data_in_integrity_info {
+	__be64 data_bytes;
+	__be64 retrieved_attributes_bytes;
+	__be64 integrity_check_value;
+} __packed;
+
+struct osd_timestamp {
+	u8 time[6]; /* number of milliseconds since 1/1/1970 UT (big endian) */
+} __packed;
+/* FIXME: define helper functions to convert to/from osd time format */
+
+/*
+ * Capability & Security definitions
+ * osd2r03: 4.11.2.2 Capability format
+ * osd2r03: 5.2.8 Security parameters
+ */
+
+struct osd_key_identifier {
+	u8 id[7]; /* if you know why 7 please email bharrosh@panasas.com */
+} __packed;
+
+/* for osd_capability.format */
+enum {
+	OSD_SEC_CAP_FORMAT_NO_CAPS = 0,
+	OSD_SEC_CAP_FORMAT_VER1 = 1,
+	OSD_SEC_CAP_FORMAT_VER2 = 2,
+};
+
+/* security_method */
+enum {
+	OSD_SEC_NOSEC = 0,
+	OSD_SEC_CAPKEY = 1,
+	OSD_SEC_CMDRSP = 2,
+	OSD_SEC_ALLDATA = 3,
+};
+
+enum object_type {
+	OSD_SEC_OBJ_ROOT = 0x1,
+	OSD_SEC_OBJ_PARTITION = 0x2,
+	OSD_SEC_OBJ_COLLECTION = 0x40,
+	OSD_SEC_OBJ_USER = 0x80,
+};
+
+enum osd_capability_bit_masks {
+	OSD_SEC_CAP_APPEND	= (1 << 0),
+	OSD_SEC_CAP_OBJ_MGMT	= (1 << 1),
+	OSD_SEC_CAP_REMOVE	= (1 << 2),
+	OSD_SEC_CAP_CREATE	= (1 << 3),
+	OSD_SEC_CAP_SET_ATTR	= (1 << 4),
+	OSD_SEC_CAP_GET_ATTR	= (1 << 5),
+	OSD_SEC_CAP_WRITE	= (1 << 6),
+	OSD_SEC_CAP_READ	= (1 << 7),
+
+	OSD_SEC_CAP_NONE1	= (1 << 8),
+	OSD_SEC_CAP_NONE2	= (1 << 9),
+	OSD_SEC_CAP_NONE3	= (1 << 10),
+	OSD_SEC_CAP_QUERY	= (1 << 11), /*v2 only*/
+	OSD_SEC_CAP_M_OBJECT	= (1 << 12), /*v2 only*/
+	OSD_SEC_CAP_POL_SEC	= (1 << 13),
+	OSD_SEC_CAP_GLOBAL	= (1 << 14),
+	OSD_SEC_CAP_DEV_MGMT	= (1 << 15),
+};
+
+/* for object_descriptor_type (hi nibble used) */
+enum {
+	OSD_SEC_OBJ_DESC_NONE = 0,     /* Not allowed */
+	OSD_SEC_OBJ_DESC_OBJ = 1 << 4, /* v1: also collection */
+	OSD_SEC_OBJ_DESC_PAR = 2 << 4, /* also root */
+	OSD_SEC_OBJ_DESC_COL = 3 << 4, /* v2 only */
+};
+
+/* (osd-r10:4.9.2.2)
+ * osd2r03:4.11.2.2 Capability format
+ */
+struct osd_capability_head {
+	u8 format; /* low nibble */
+	u8 integrity_algorithm__key_version; /* MAKE_BYTE(integ_alg, key_ver) */
+	u8 security_method;
+	u8 reserved1;
+/*04*/	struct osd_timestamp expiration_time;
+/*10*/	u8 audit[30-10];
+/*30*/	u8 discriminator[42-30];
+/*42*/	struct osd_timestamp object_created_time;
+/*48*/	u8 object_type;
+	u8 permissions_bit_mask[54-49];
+/*54*/	u8 reserved2;
+/*55*/	u8 object_descriptor_type; /* high nibble */
+} __packed;
+
+/*56 v1*/
+struct osdv1_cap_object_descriptor {
+	union {
+		struct {
+/*56*/			__be32 policy_access_tag;
+/*60*/			__be64 allowed_partition_id;
+/*68*/			__be64 allowed_object_id;
+/*76*/			__be32 reserved;
+		} __packed obj_desc;
+
+		u8 object_descriptor[80-56];/*24*/
+	};
+} __packed;
+/*80 v1*/
+
+/*56 v2*/
+struct osd_cap_object_descriptor {
+	union {
+		struct {
+			__be32 allowed_attributes_access;
+/*60*/			__be32 policy_access_tag;
+/*64*/			__be16 boot_epoch;
+/*66*/			u8 reserved[72-66];
+/*72*/			__be64 allowed_partition_id;
+/*80*/			__be64 allowed_object_id;
+/*88*/			__be64 allowed_range_length;
+/*96*/			__be64 allowed_range_start;
+		} __packed obj_desc;
+
+		u8 object_descriptor[104-56]; /*48*/
+	};
+} __packed;
+/*104 v2*/
+
+struct osdv1_capability {
+	struct osd_capability_head h;
+	struct osdv1_cap_object_descriptor od;
+} __packed;
+
+struct osd_capability {
+	struct osd_capability_head h;
+/* 	struct osd_cap_object_descriptor od;*/
+	struct osdv1_cap_object_descriptor od; /* FIXME: Pete rev-001 sup */
+} __packed;
+
+/**
+ * osd_sec_set_caps - set cap-bits into the capabilities header
+ *
+ * @cap:	The osd_capability_head to set cap bits to.
+ * @bit_mask: 	Use an ORed list of enum osd_capability_bit_masks values
+ *
+ * permissions_bit_mask is unaligned use below to set into caps
+ * in a version independent way
+ */
+static inline void osd_sec_set_caps(struct osd_capability_head *cap,
+	u16 bit_mask)
+{
+	/*
+	 *Note: The bits above are defined LE order this is because this way
+	 *      they can grow in the future to more then 16, and still retain
+	 *      there constant values.
+	 */
+	put_unaligned_le16(bit_mask, &cap->permissions_bit_mask);
+}
+
+#endif /* ndef __OSD_PROTOCOL_H__ */
diff --git a/include/scsi/osd_sec.h b/include/scsi/osd_sec.h
new file mode 100644
index 000000000000..4c09fee8ae1e
--- /dev/null
+++ b/include/scsi/osd_sec.h
@@ -0,0 +1,45 @@
+/*
+ * osd_sec.h - OSD security manager API
+ *
+ * Copyright (C) 2008 Panasas Inc.  All rights reserved.
+ *
+ * Authors:
+ *   Boaz Harrosh <bharrosh@panasas.com>
+ *   Benny Halevy <bhalevy@panasas.com>
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License version 2
+ *
+ */
+#ifndef __OSD_SEC_H__
+#define __OSD_SEC_H__
+
+#include "osd_protocol.h"
+#include "osd_types.h"
+
+/*
+ * Contains types and constants of osd capabilities and security
+ * encoding/decoding.
+ * API is trying to keep security abstract so initiator of an object
+ * based pNFS client knows as little as possible about security and
+ * capabilities. It is the Server's osd-initiator place to know more.
+ * Also can be used by osd-target.
+ */
+void osd_sec_encode_caps(void *caps, ...);/* NI */
+void osd_sec_init_nosec_doall_caps(void *caps,
+	const struct osd_obj_id *obj, bool is_collection, const bool is_v1);
+
+bool osd_is_sec_alldata(struct osd_security_parameters *sec_params);
+
+/* Conditionally sign the CDB according to security setting in ocdb
+ * with cap_key */
+void osd_sec_sign_cdb(struct osd_cdb *ocdb, const u8 *cap_key);
+
+/* Unconditionally sign the BIO data with cap_key.
+ * Check for osd_is_sec_alldata() was done prior to calling this. */
+void osd_sec_sign_data(void *data_integ, struct bio *bio, const u8 *cap_key);
+
+/* Version independent copy of caps into the cdb */
+void osd_set_caps(struct osd_cdb *cdb, const void *caps);
+
+#endif /* ndef __OSD_SEC_H__ */
diff --git a/include/scsi/osd_types.h b/include/scsi/osd_types.h
new file mode 100644
index 000000000000..ea5372da0a32
--- /dev/null
+++ b/include/scsi/osd_types.h
@@ -0,0 +1,40 @@
+/*
+ * osd_types.h - Types and constants which are not part of the protocol.
+ *
+ * Copyright (C) 2008 Panasas Inc.  All rights reserved.
+ *
+ * Authors:
+ *   Boaz Harrosh <bharrosh@panasas.com>
+ *   Benny Halevy <bhalevy@panasas.com>
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License version 2
+ *
+ * Contains types and constants that are implementation specific and are
+ * used by more than one part of the osd library.
+ *     (Eg initiator/target/security_manager/...)
+ */
+#ifndef __OSD_TYPES_H__
+#define __OSD_TYPES_H__
+
+struct osd_systemid {
+	u8 data[OSD_SYSTEMID_LEN];
+};
+
+typedef u64 __bitwise osd_id;
+
+struct osd_obj_id {
+	osd_id partition;
+	osd_id id;
+};
+
+static const struct __weak osd_obj_id osd_root_object = {0, 0};
+
+struct osd_attr {
+	u32 page;
+	u32 attr_id;
+	u16 len;		/* byte count of operand */
+	void *val_ptr;		/* in network order */
+};
+
+#endif /* ndef __OSD_TYPES_H__ */
diff --git a/include/scsi/scsi.h b/include/scsi/scsi.h
index a109165714d6..de1cef275826 100644
--- a/include/scsi/scsi.h
+++ b/include/scsi/scsi.h
@@ -263,6 +263,7 @@ static inline int scsi_status_is_good(int status)
 #define TYPE_RAID           0x0c
 #define TYPE_ENCLOSURE      0x0d    /* Enclosure Services Device */
 #define TYPE_RBC	    0x0e
+#define TYPE_OSD            0x11
 #define TYPE_NO_LUN         0x7f
 
 /* SCSI protocols; these are taken from SPC-3 section 7.5 */