Merge lp:~smb/ubuntu/oneiric/iscsitarget/proposed into lp:ubuntu/oneiric/iscsitarget

[Stefan Bader]

This brings the package to the version in Debian experimental which fixes compile failures with kernels after 2.6.38 for the produced dkms package. There is one additional change which modifies the dkms.conf file (missing upstream as well) to make it compile on kernels before 2.6.39.

[Colin Watson (cjwatson)]

review: Approve

=== added directory '.pc'
=== added file '.pc/.version'
--- .pc/.version	1970-01-01 00:00:00 +0000
+++ .pc/.version	2011-06-29 14:00:59 +0000
@@ -0,0 +1,1 @@
+2
 
=== added file '.pc/applied-patches'
--- .pc/applied-patches	1970-01-01 00:00:00 +0000
+++ .pc/applied-patches	2011-06-29 14:00:59 +0000
@@ -0,0 +1,2 @@
+debian-changes-1.4.20.2-5
+ubuntu-changes-1.4.20.2-5
 
=== added directory '.pc/debian-changes-1.4.20.2-5'
=== added file '.pc/debian-changes-1.4.20.2-5/ChangeLog'
--- .pc/debian-changes-1.4.20.2-5/ChangeLog	1970-01-01 00:00:00 +0000
+++ .pc/debian-changes-1.4.20.2-5/ChangeLog	2011-06-29 14:00:59 +0000
@@ -0,0 +1,652 @@
+Summary of changes from v1.4.20.1 to v1.4.20.2
+=================================
+Arne Redlich
+  o uninitialized data fixes pointed out by valgrind
+  o fix invalid read in iSNS code pointed out by valgrind
+
+Ross Walker
+  o check buffer boundaries in iSNS code (CVE-2010-2221)
+  o fix memleak in iSNS ACL code
+  o clean up iSCSI login response error handling
+  o rework ietd startup to avoid concurrent ietds
+  o fix soft lockup during UA allocation
+  o change DISTDIR to DESTDIR
+  o avoid deletion of targets with active sessions
+  o byte swap fix in MD5 / SHA1 code (thanks to Harshal Shete for testing on PPC)
+
+Summary of changes from v1.4.20 to v1.4.20.1
+=================================
+Arne Redlich
+  o adapt backward compat mechanism in Makefile to kernels >= 2.6.33
+
+Ross Walker
+  o restore SCSI SN behaviour of IET < 1.4.20 (VMWare compatibility)
+
+Summary of changes from v1.4.19 to v1.4.20
+=================================
+Minh Tran
+  o fix TTT for final PDU in ietd
+
+Qinghua(Kevin) Ye
+  o release reservations during session delete
+
+Ross Walker
+  o multiple updates for abort task compliance
+  o replace sync_page_range with filemap_write_and_wait_range in fileio
+  o replace do_sync_read/write with vfs_read/write in fileio
+  o multiple updates for target_del_all, atomic deletion of targets,
+    sessions and connections
+  o install moves existing kernel module, uninstall moves it back
+  o cleaned up RPM and DKMS package building
+  o cleaned up usr-kernel ioctl calls, added version checking, semaphore
+    to prevent ioctls during kernel cleanup (target_del_all)
+  o cleaned up session and connection reinstatement
+  o setup IO contexts in wthread, fixes CFQ performance problem
+  o enabled multiple connections per session
+  o moved common parameters from iotypes to volume
+  o UA handling on aborted/terminated connections
+  o removed OpenSSL dependency by borrowing kernel md5/sha1 code
+  o code cleanups
+
+Arne Redlich
+  o multiple kernel compatibility updates
+  o code cleanups
+
+Lars Ellenberg
+  o support for logical sector sizes
+  o case insensitive volume options
+  o code cleanups
+
+Luca Berra
+  o fix warnings when compiling via gcc strict checks
+
+Kun Huang
+  o temporary target redirections via ietadm for existing targets
+
+Jagadish Kumar
+  o multiple code cleanups around network and worker thread handling
+
+Summary of changes from v1.4.18 to v1.4.19
+=================================
+Andreas Florath
+  o re-enable MaxSessions parameter
+  o code cleanups
+
+Arne Redlich
+  o NOP-In support
+
+Harshal Shete
+  o add notes about cross compilation to README
+
+SZÉKELYI Szabolcs
+  o fix ietadm manpage
+
+Ross S. W. Walker
+  o fix initiators/targets.allow bugs
+  o fix compiler warnings
+  o fix return value of session_add()
+  o update RPM specfile improving RedHat and SuSE compatibility
+    (based on a patch from Matthew Wild)
+  o update RELEASE NOTES
+
+Kevin Ye
+  o fix session cleanup in case of -EPIPE during login
+
+
+Summary of changes from v0.4.17 to v1.4.18
+=================================
+Lars Ellenberg
+  o improve IETs procfs support to allow a larger number of targets
+  o compatibility fix for 2.6.28 (independently also provided by Francois
+    Micaux)
+  o ignore SIGPIPE in ietd
+  o allow cleanup of stale targets in the kernel module
+
+Andreas Florath
+  o support for a global worker threadpool instead of per target pools
+    (configurable via the worker_thread_pool_size module parameter)
+
+Arne Redlich
+  o compatibility fixes for 2.6.30, 2.6.29
+  o in cooperation with Xie Gang: Unit Attention support (thanks to Stefan
+    Rubner for spotting a bug)
+  o in cooperation with Shreyansh Jain: support for splitting text messages into
+    several PDUs, allowing more targets to be reported during discovery
+    (thanks to Cheng Rengquan for bugfixes)
+  o clean up all connections, sessions and targets in the kernel module if the
+    daemon is gone
+  o rework sense data handling, plugging memleaks
+  o fix valgrind warnings of uses of uninitialized variables
+  o support for Reject PDUs (thanks to FUJITA Tomonori for a bugfix)
+  o fix SERVICE ACTION IN: IET only supports READ CAPACITY 16, return sense data
+    for all others
+  o allow READ CAPACITY even if a LU is RESERVEd
+  o iSNS: report the actual target port instead of the default iSCSI port
+  o fix list corruption if target thread creation fails
+  o fix stopping of threads that have never been awakened before
+  o fix CHAP account handling bugs
+  o fix netlink socket cleanup
+  o code cleanups
+
+Cheng Renquan
+  o remove superfluous linefeed from debug messages
+  o enable runtime switching of debug levels for the kernel module and
+    provide description for the module's debug flags parameter
+
+sdrb
+  o return proper status class in case of login failures
+
+Ross S. W. Walker
+  o support more than one portal per target during discovery, and allow
+    configuration of the presented portals using targets.allow - previously only
+    the incoming interface was reported (thanks to David Shirley for
+    reporting a bug)
+  o deprecate initiators.allow
+  o regex based filtering in initiators.allow and targets.allow
+  o move config files to /etc/iet
+  o add version and author info and description to kernel module
+  o fix debug log levels
+  o fix ietd.conf permissions
+  o move kernel module to /lib/modules/<kver>/extra where it belongs
+  o improve init scripts, using common names and functions across distributions
+  o improve usage comments in ietd.conf
+  o specfile to build RPMs, including weak module support
+  o DKMS support
+
+Ming Zhang
+  o Use the LU's sector size in the format mode page instead of defaulting to 512
+
+Summary of changes from v0.4.16 to v0.4.17
+=================================
+
+Charley Cheng
+  o handle the \0-termination of strings properly when building iSNS TLVs
+
+Denis ChengRq
+  o fix stopping of wthreads that have never been awakened
+
+Shreyansh Jain
+  o fix debugging code that hexdumps PDU content to the syslog
+
+Jagadish Kumar
+  o avoid busy looping if a connection's socket is out of wmem
+
+Arne Redlich
+  o add support for "ietadm --op show --user"
+  o SCSI fixes: MODE SENSE handling, INQUIRY CmdDt, HiSup, WCE and RCD handling
+  o fix handling of NOP-Out w/ ping data
+  o fix compilation against glibc >= 2.8
+  o fix potential NULL-pointer derefences and resource leaks in the LUN param
+    parsing code
+  o documentation updates
+
+Ross S. W. Walker
+  o 2.6.26 and SLES compile fix
+  o SLES 10 SP2 compat patch
+
+Ming Zhang
+  o make needlessly global function static
+
+
+Summary of changes from v0.4.15 to v0.4.16
+=================================
+
+Arne Redlich
+  o fix overzealous assert() in digest_data()
+  o add checking on return value of ISCSI_PARAM_GET
+  o 2.6.22, 2.6.23 and 2.6.24 compile fixes
+  o add conn->rwsize check
+  o avoid potential NULL-ptr dereferences in rx and tx buffer
+  o fix the shell syntax in init scripts
+
+Dave Jiang
+  o fix digest endieness on LE archs
+
+FUJITA Tomonori
+  o fix SPARC alignement issues (based on a patch from [e-mail])
+
+Ross S. W. Walker
+  o fix DISTDIR in Makefile for /etc install
+  o add support to nullio for volumes > 2TB
+  o remove init.d memory size adjustment
+  o add error code reporting to blockio_open_path
+  o blockio gen_scsiid bug fix
+  o add verbosity to kernel output and task management
+
+
+Summary of changes from v0.4.14 to v0.4.15
+=================================
+
+Juhani Rautiainen
+ o Add RELEASE/RESERVE support
+
+Ross S. W. Walker
+ o Improve the build system to support several kernel versions
+ o Add block-io support
+
+
+Summary of changes from v0.4.13 to v0.4.14
+=================================
+
+Arne Redlich
+  o Kill unused "state" in struct iscsi_cmnd.
+  o Fixed fileio_sync() to propagate error to the caller (and initiator).
+  o Don't attempt to show target/session params if communication with ietd
+  fails.
+  o Fixes to ietadm parameters handling.
+
+FUJITA Tomonori
+  o rewritten iSNS code, many iSNS fixes.
+  o added iSNS SCN support.
+  o IPv6 fixes to userspace.
+
+Ming Zhang
+  o Fix the READ_* commands error handling bug.
+  o fix the mode sense response.
+  o wrong #endif sequence in misc.h
+
+Richard Bollinger
+  o add a patch to ietd.c that allows defunct sessions to go away.
+  o add write-back cache and read-only support.
+
+Frederic Temporelli
+  o Fix for the combination of 32-bit userland and 64-bit kernel on mips.
+
+Henry Liu
+  o corrected many task management functions, prevent crashing on
+  LUN RESET, TARGET WARM RESET.
+
+K Chapman
+  o Fixed a typo in check_segment_length().
+
+Emmanuel Florac
+  o Add ietadm manpage.
+
+
+Summary of changes from v0.4.12 to v0.4.13
+=================================
+Arne Redlich
+  o patch to avoid digest calculation for PDUs whose data has been skipped
+  already for various reasons.
+  o Correct a bug managing non-default MaxRxDSL.
+  o added to ietadm ability to show target parameters.
+  o add on the workaround to AIX initiator MaxCmdSN bug.
+
+FUJITA Tomonori
+  o added to ietadm ability to show the iSCSI parameters for an established
+  session.
+
+Ming Zhang
+  o Fixed this bug : ietd should manage the iscsi name in a case insensitive
+  way to conform to the RFC.
+  o workaround to AIX initiator MaxCmdSN bug.
+  o Fixed socket() return value judgment.
+
+Bastiaan Bakker
+  o add 'condrestart' command to the RedHat initscript.
+
+Robert Whitehead
+  o correct the bug that prevents iet to start if there isn't
+  an /etc/ietd.conf file.
+
+
+Summary of changes from v0.4.11 to v0.4.12
+=================================
+
+Arne Redlich
+  o Fix MaxRecvDataSegmentLength handling.
+  o Fix login parameter handling.
+  o Update man pages.
+
+Bastiaan Bakker
+  o Add features to specify the listen address and port.
+  o Fix setuid and setgid bugs in ietd daemon.
+
+FUJITA Tomonori
+  o Add IPv6 support.
+
+Junjiro Okajima
+  o Fix a bug about getting parameters from kernel space.
+
+Krzysztof Blaszkowski
+  o Fix for requests with unaligned to 4 length.
+
+
+Summary of changes from v0.4.10 to v0.4.11
+=================================
+
+FUJITA Tomonori
+  o Fix Task Management Function bugs.
+
+Ming Zhang
+  o Update man pages.
+
+
+Summary of changes from v0.4.9 to v0.4.10
+=================================
+
+Arne Redlich
+  o Fix 0x83 inquiry output.
+  o Fix iSCSI parameter handling bugs.
+
+FUJITA Tomonori
+  o Add the access control based on initiator address
+    and target name patterns.
+
+Junjiro Okajima
+  o Fix parameter checking bugs.
+
+Ming Zhang
+  o Add the nullio mode (only useful for performance evaluation).
+
+
+Summary of changes from v0.4.8 to v0.4.9
+=================================
+
+FUJITA Tomonori
+  o Fix parameter negotiation handling bugs.
+
+Wang Zhenyu
+  o Fix digest negotiation handling bugs.
+
+
+Summary of changes from v0.4.7 to v0.4.8
+=================================
+
+FUJITA Tomonori
+  o Fix unsolicited data handling bugs.
+  o Rewrite parameter handling code.
+  o Rewrite ietadm tool.
+  o Improve dynamic configuration support.
+  o Cleanups on the kernel-user interface.
+  o Improve wrong PDU handling.
+  o Implement a framework to handle multiple configuration methods.
+  o Implement basic access control support.
+
+
+Summary of changes from v0.4.6 to v0.4.7
+=================================
+
+Florian Zierer
+  o Add the startup script for Gentoo.
+
+FUJITA Tomonori
+  o Rewrite parameter handling code.
+  o Fix task management code bug.
+  o Fix 0x83 inquiry output (Thanks to Christophe Varoqui).
+
+Ming Zhang
+  o Acquire T10 ID.
+  o Fix parameter handling bugs.
+  o Some user-space cleanups.
+
+Philipp Hug
+  o Fix ietd.8 man typo.
+
+
+Summary of changes from v0.4.5 to v0.4.6
+=================================
+
+FUJITA Tomonori
+  o Replace the makeshift event notification code with netlink.
+  o Add task management code except for ACA and reassign stuff.
+  o Fix r2t lun bug (Thanks to Ming Zhang).
+
+
+Summary of changes from v0.4.4 to v0.4.5
+=================================
+
+FUJITA Tomonori
+  o Rewrite the iSCSI command handling code.
+  o Rewrite the I/O data handling code.
+  o Fix worker thread.
+  o Several cleanups.
+
+
+Summary of changes from v0.4.3 to v0.4.4
+=================================
+
+Krzysztof Blaszkowski
+  o Fix an out-of-memory bug.
+
+
+Summary of changes from v0.4.2 to v0.4.3
+=================================
+
+Arne Redlich
+  o Fix header digest bug.
+  o Fix unexpected closed connection bug.
+  o Fix iSCSI parameter bug.
+
+FUJITA Tomonori
+  o Fix network thread.
+
+
+Summary of changes from v0.4.1 to v0.4.2
+=================================
+
+FUJITA Tomonori
+  o Fix network thread.
+  o Fix MaxOutstandingR2T handling.
+
+Ming Zhang
+  o Add large volume support (over 2TB).
+
+
+Summary of changes from v0.4.0 to v0.4.1
+=================================
+
+Arne Redlich
+  o Add mutual CHAP support. Note that you need to replace "User"
+    with "IncomingUser" in ietd.conf.
+
+FUJITA Tomonori
+  o Fix InitialR2T=No support.
+  o Fix INQUIRY command handling.
+  o Fix network and worker thread.
+  o Start to split SCSI stuff.
+  o Rewrite the R2T handling code.
+  o Several cleanups.
+
+
+Summary of changes from v0.3.8 to v0.4.0
+=================================
+
+Arne Redlich
+  o iSNS bug fix.
+
+FUJITA Tomonori
+  o Move to 2.6 kernels.
+  o Rewrite the kernel thread performing network I/O.
+  o Add header and data digests (Thanks to Arne Redlich).
+
+Ming Zhang
+  o Add mode sense page 0x3 and 0x4 support (Thanks to K Chapman).
+  o iSNS bug fix.
+
+
+Summary of changes from v0.3.7 to v0.3.8
+=================================
+
+Arne Redlich
+  o Fix ietadm global option bug.
+
+FUJITA Tomonori
+  o Fix TCP option bugs (Thanks to Chuck Berg).
+  o Fix REPORT LUN (handling lots of LUs).
+
+
+Summary of changes from v0.3.6 to v0.3.7
+=================================
+
+Arne Redlich
+  o Fix target_alloc_pages().
+
+FUJITA Tomonori
+  o Fix REPORT LUN bug.
+
+
+Summary of changes from v0.3.5 to v0.3.6
+=================================
+
+Arne Redlich
+  o Fix bugs about rejecting PDUs.
+
+FUJITA Tomonori
+  o Cleanups on target_cmnd structure.
+  o Kill highmem stuff.
+  o Fix REPORT LUN (handling lots of LUs).
+
+
+Summary of changes from v0.3.4 to v0.3.5
+=================================
+
+Arne Redlich
+  o Fix ietd security hole.
+  o Fix REPORT LUN bug.
+  o FIX NOOP_OUT padding bug.
+
+FUJITA Tomonori
+  o Rewrite event notification code.
+
+Libor Vanek
+  o Add max_sessions option.
+  o Fix command parsing bug.
+
+Ming Zhang
+  o Cleanups for 64-bit architectures.
+
+
+Summary of changes from v0.3.3 to v0.3.4
+=================================
+
+FUJITA Tomonori
+  o Improve dynamic configuration support (adding targets and users).
+
+
+Summary of changes from v0.3.2 to v0.3.3
+=================================
+
+FUJITA Tomonori
+  o Fix Makefile for the startup script.
+
+
+Summary of changes from v0.3.1 to v0.3.2
+=================================
+
+Ali Lehmann
+  o Add a new startup script for Debian.
+
+FUJITA Tomonori
+  o Fix the istd's handling of connections in out-of-memory situations.
+  o Fix bugs in regular file support.
+  o Fix `ietadm --mode del all`.
+
+Libor Vanek
+  o Add a new startup script for RedHat.
+
+Ming Zhang
+  o Add uid/gid option to ietd daemon.
+  o Fix a access freed-memory bug in kernel/daemon.c.
+
+
+Summary of changes from v0.3.0 to v0.3.1
+=================================
+
+FUJITA Tomonori
+  o Fix memory leaks in ietd daemon (Thanks to Ming).
+  o Fix bugs about REPORT_LUNS commands (Thanks to Ming).
+  o Fix a bug about Target Task Tag.
+  o Add regular file support to fileio mode.
+
+
+Summary of changes from v0.2.6 to v0.3.0
+=================================
+
+Ali Lehmann
+  o Update ietd.8 man page.
+
+FUJITA Tomonori
+  o Fix shutdown code.
+  o Fix istd kernel thread bugs.
+  o Replace procfs interface with ioctl.
+  o Add dynamic configuration support.
+  o Update README and the boot script.
+
+Ming Zhang
+  o Add config option to ietd daemon.
+
+
+Summary of changes from v0.2.5 to v0.2.6
+=================================
+
+Ali Lehmann
+  o Add ietd.8 and ietd.conf.5 man pages.
+
+FUJITA Tomonori
+  o Update README, Makefile, and the boot script.
+
+
+Summary of changes from v0.2.4 to v0.2.5
+=================================
+
+FUJITA Tomonori
+  o Update README.
+
+
+Summary of changes from v0.2.3 to v0.2.4
+=================================
+
+Ming Zhang
+  o Add a preliminary iSNS support.
+  o Fix merge mistakes that I made at the previous release.
+
+
+Summary of changes from v0.2.2 to v0.2.3
+=================================
+
+Ming Zhang
+  o Improve INQUIRY, REQUEST_SENSE, and MODE_SENSE command supports
+  o Add fake RESERVE* and RELEASE* command supports
+
+
+Summary of changes from v0.2.1 to v0.2.2
+=================================
+
+FUJITA Tomonori
+  o Improve the write performance of the file IO mode
+
+Ming Zhang
+  o Fix unupdated pg_cnt when allocating a new tcmnd
+  o Several cleanups
+
+
+Summary of changes from v0.2.0 to v0.2.1
+=================================
+
+FUJITA Tomonori
+  o Fix a bug that makes the target use CPU unnecessarily
+  o Add a feature that enable you to pass options to an IO mode
+
+
+Summary of changes from v0.1.0 to v0.2.0
+=================================
+
+FUJITA Tomonori
+  o Rewrite read and write kernel threads which perform network IO
+  o Fix race issues in the proc interface
+  o Fix shutdown code
+
+Ming Zhang
+  o Fix memory leaks in file and block IO modes
+
+
+Summary of changes from the ardis v20040211 to v0.1.0
+=================================
+
+FUJITA Tomonori
+  o Remove a kernel patch. Multiple threads execute I/O operations
+  o Replace IO functions with the kernel starndard functions
+  o Add multiple IO modes feature
+  o Fix several race issues
+  o Fix several out-of-memory situation bugs
 
=== added file '.pc/debian-changes-1.4.20.2-5/Makefile'
--- .pc/debian-changes-1.4.20.2-5/Makefile	1970-01-01 00:00:00 +0000
+++ .pc/debian-changes-1.4.20.2-5/Makefile	2011-06-29 14:00:59 +0000
@@ -0,0 +1,352 @@
+#
+# Makefile for the Linux kernel device drivers.
+#
+# Note! Dependencies are done automagically by 'make dep', which also
+# removes any old dependencies. DON'T put your own dependencies here
+# unless it's something special (not a .c file).
+#
+# Note 2! The CFLAGS definitions are now in the main makefile.
+
+#export KSRC := /usr/src/linux
+
+SUBDIRS := $(shell pwd)
+
+ifeq ($(KSRC),)
+	KSRC ?= /lib/modules/$(shell uname -r)/build
+endif
+
+
+ifneq ($(wildcard $(KSRC)/include/generated/utsrelease.h),)
+	VERSION_FILE := $(KSRC)/include/generated/utsrelease.h
+else
+  ifneq ($(wildcard $(KSRC)/include/linux/utsrelease.h),)
+	  VERSION_FILE := $(KSRC)/include/linux/utsrelease.h
+  else
+	  VERSION_FILE := $(KSRC)/include/linux/version.h
+  endif
+endif
+
+KVER := $(shell $(CC) $(CFLAGS) -E -dM $(VERSION_FILE) | \
+	grep UTS_RELEASE | awk '{ print $$3 }' | sed 's/\"//g')
+
+KMOD := /lib/modules/$(KVER)/extra
+
+KMAJ := $(shell echo $(KVER) | \
+	sed -e 's/^\([0-9][0-9]*\)\.[0-9][0-9]*\.[0-9][0-9]*.*/\1/')
+KMIN := $(shell echo $(KVER) | \
+	sed -e 's/^[0-9][0-9]*\.\([0-9][0-9]*\)\.[0-9][0-9]*.*/\1/')
+KREV := $(shell echo $(KVER) | \
+	sed -e 's/^[0-9][0-9]*\.[0-9][0-9]*\.\([0-9][0-9]*\).*/\1/')
+
+kver_eq = $(shell [ $(KMAJ)$(KMIN)$(KREV) -eq $(1)$(2)$(3) ] && \
+	echo 1 || echo 0)
+kver_lt = $(shell [ $(KMAJ)$(KMIN)$(KREV) -lt $(1)$(2)$(3) ] && \
+	echo 1 || echo 0)
+kver_le = $(shell [ $(KMAJ)$(KMIN)$(KREV) -le $(1)$(2)$(3) ] && \
+	echo 1 || echo 0)
+kver_gt = $(shell [ $(KMAJ)$(KMIN)$(KREV) -gt $(1)$(2)$(3) ] && \
+	echo 1 || echo 0)
+kver_ge = $(shell [ $(KMAJ)$(KMIN)$(KREV) -ge $(1)$(2)$(3) ] && \
+	echo 1 || echo 0)
+kver_lk = $(shell [ `echo $(KVER) | egrep $(1)` ] && echo 1 || echo 0)
+
+#
+# Please when adding patch sets start with the latest to the earliest
+# the idea behind this is that by properly patching the latest code
+# base first the earlier patch sets will not need to be modified.
+#
+
+# Compatibility patch for kernels <= 2.6.32
+ifeq ($(call kver_le,2,6,32),1)
+	PATCHES := $(PATCHES) compat-2.6.32.patch
+endif
+
+# Compatibility patch for kernels <= 2.6.31
+ifeq ($(call kver_le,2,6,31),1)
+	PATCHES := $(PATCHES) compat-2.6.31.patch
+endif
+
+# Compatibility patch for kernels <= 2.6.30
+ifeq ($(call kver_le,2,6,30),1)
+	PATCHES := $(PATCHES) compat-2.6.30.patch
+endif
+
+# Compatibility patch for kernels <= 2.6.29
+ifeq ($(call kver_le,2,6,29),1)
+	PATCHES := $(PATCHES) compat-2.6.29.patch
+endif
+
+# Compatibility patch for kernels <= 2.6.28
+ifeq ($(call kver_le,2,6,28),1)
+	PATCHES := $(PATCHES) compat-2.6.28.patch
+endif
+
+# Compatibility patch for kernels >= 2.6.25 and <= 2.6.27
+ifeq ($(call kver_le,2,6,27),1)
+	PATCHES := $(PATCHES) compat-2.6.25-2.6.27.patch
+endif
+
+# Compatibility patch for kernels <= 2.6.24
+ifeq ($(call kver_le,2,6,24),1)
+	PATCHES := $(PATCHES) compat-2.6.24.patch
+endif
+
+# Compatibility patch for kernels <= 2.6.23
+ifeq ($(call kver_le,2,6,23),1)
+	PATCHES := $(PATCHES) compat-2.6.23.patch
+endif
+
+# Compatibility patch for kernels <= 2.6.22
+ifeq ($(call kver_le,2,6,22),1)
+	PATCHES := $(PATCHES) compat-2.6.22.patch
+endif
+
+# Compatibility patch for kernels >= 2.6.19 and <= 2.6.21
+ifeq ($(call kver_le,2,6,21),1)
+	PATCHES := $(PATCHES) compat-2.6.19-2.6.21.patch
+endif
+
+# Compatibility patch for kernels >= 2.6.14 and <= 2.6.18
+ifeq ($(call kver_le,2,6,18),1)
+	PATCHES := $(PATCHES) compat-2.6.14-2.6.18.patch
+endif
+
+# We don't support kernels < 2.6.14 except for explicit distros
+ifeq ($(call kver_lt,2,6,14),1)
+	UNSUPPORTED := true
+endif
+
+# Compatibility patches for SuSE distros
+ifneq ($(wildcard /etc/SuSE-release),)
+	# Compatibility patch for SLES 10 SP2
+	ifeq ($(call kver_lk,"2\.6\.16\.60-.*"),1)
+		PATCHES += compat-sles10sp2.patch
+		UNSUPPORTED :=
+	endif
+endif
+
+# Compatibility patches for Redhat distros
+ifneq ($(wildcard /etc/redhat-release),)
+	# Compatibility patch for RHEL4/CentOS4
+	ifeq ($(call kver_lk,"2\.6\.9-.*\.(EL|plus\.c4)"),1)
+		PATCHES += compat-rhel4.patch
+		UNSUPPORTED :=
+	endif
+endif
+
+MANPAGES:= ietadm.8 ietd.8 ietd.conf.5
+
+ifeq ($(MANDIR),)
+	MANPATH := $(shell (manpath 2>/dev/null || \
+		echo $MANPATH) | sed 's/:/ /g')
+	ifneq ($(MANPATH),)
+		test_dir = $(findstring $(dir), $(MANPATH))
+	else
+		test_dir = $(shell [ -e $(dir) ] && echo $(dir))
+	endif
+	MANDIR := /usr/share/man /usr/man
+	MANDIR := $(foreach dir, $(MANDIR), $(test_dir))
+	MANDIR := $(firstword $(MANDIR))
+endif
+
+ifeq ($(MANDIR),)
+	MANDIR := /usr/share/man
+endif
+
+DOCS:= ChangeLog COPYING RELEASE_NOTES README README.vmware README.initiators
+
+ifeq ($(DOCDIR),)
+	DOCDIR := /usr/share/doc/iscsitarget
+endif
+
+all: usr kernel
+
+usr: patch
+	$(MAKE) -C usr
+
+kernel: patch
+	$(MAKE) -C $(KSRC) SUBDIRS=$(shell pwd)/kernel modules
+
+patch: $(UNSUPPORTED) integ_check $(PATCHES)
+
+$(UNSUPPORTED):
+	@echo "Sorry, your kernel version and/or distribution is currently"
+	@echo "not supported."
+	@echo ""
+	@echo "Please read the README file for information on how you can"
+	@echo "contribute compatibility/bug fixes to the IET project."
+	@exit 1
+
+integ_check:
+	@if [ -e .patched.* -a ! -e .patched.$(KVER) ]; then \
+		$(MAKE) unpatch; \
+	fi
+
+$(PATCHES): .patched.$(KVER)
+
+.patched.$(KVER):
+	@set -e; \
+	if [ ! -e .patched.* ]; then \
+		for p in $(PATCHES); do \
+			echo "Applying Patch $$p"; \
+			patch -p1 < patches/$$p; \
+			echo $$p >>.patched.$(KVER); \
+		done; \
+	fi
+
+unpatch:
+	@set -e; \
+	if [ -e .patched.* ]; then \
+		for p in `cat .patched.*`; do \
+			reverse="$$p $$reverse"; \
+		done; \
+		for r in $$reverse; do \
+			echo "Reversing patch $$r"; \
+			patch -p1 -R < patches/$$r; \
+		done; \
+		rm -f .patched.*; \
+	fi
+
+depmod:
+	@echo "Running depmod"
+	@if [ x$(DESTDIR) != x -o x$(INSTALL_MOD_PATH) != x ]; then \
+		depmod -aq -b $(DESTDIR)$(INSTALL_MOD_PATH) $(KVER); \
+	else \
+		depmod -aq $(KVER); \
+	fi
+
+install-files: install-usr install-etc install-doc install-kernel
+
+install: install-files depmod
+
+install-kernel: kernel/iscsi_trgt.ko
+	@if [ -d $(DESTDIR)$(INSTALL_MOD_PATH)/lib/modules/$(KVER) ]; then \
+		if [ -f /etc/debian_version ]; then \
+			find $(DESTDIR)$(INSTALL_MOD_PATH)/lib/modules/$(KVER) \
+				-name iscsi_trgt.ko -type f \
+				-exec /bin/sh -c "dpkg-divert --rename {}" \;; \
+		else \
+			find $(DESTDIR)$(INSTALL_MOD_PATH)/lib/modules/$(KVER) \
+				-name iscsi_trgt.ko -type f \
+				-execdir mv \{\} \{\}.orig \;; \
+		fi \
+	fi
+	@install -vD kernel/iscsi_trgt.ko \
+		$(DESTDIR)$(INSTALL_MOD_PATH)$(KMOD)/iscsi/iscsi_trgt.ko
+
+install-usr: usr/ietd usr/ietadm
+	@install -vD usr/ietd $(DESTDIR)/usr/sbin/ietd
+	@install -vD usr/ietadm $(DESTDIR)/usr/sbin/ietadm
+
+install-etc: install-initd
+	@if [ ! -e $(DESTDIR)/etc/ietd.conf ]; then \
+		if [ ! -e $(DESTDIR)/etc/iet/ietd.conf ]; then \
+			install -vD -m 640 etc/ietd.conf \
+				$(DESTDIR)/etc/iet/ietd.conf; \
+		fi \
+	fi
+	@if [ ! -e $(DESTDIR)/etc/initiators.allow ]; then \
+		if [ ! -e $(DESTDIR)/etc/iet/initiators.allow ]; then \
+			install -vD -m 644 etc/initiators.allow \
+				$(DESTDIR)/etc/iet/initiators.allow; \
+		fi \
+	fi
+	@if [ ! -e $(DESTDIR)/etc/targets.allow ]; then \
+		if [ ! -e $(DESTDIR)/etc/iet/targets.allow ]; then \
+			install -vD -m 644 etc/targets.allow \
+				$(DESTDIR)/etc/iet/targets.allow; \
+		fi \
+	fi
+
+install-initd:
+	@if [ -f /etc/debian_version ]; then \
+		install -vD -m 755 etc/initd/initd.debian \
+			$(DESTDIR)/etc/init.d/iscsi-target; \
+	elif [ -f /etc/redhat-release ]; then \
+		install -vD -m 755 etc/initd/initd.redhat \
+			$(DESTDIR)/etc/rc.d/init.d/iscsi-target; \
+	elif [ -f /etc/gentoo-release ]; then \
+		install -vD -m 755 etc/initd/initd.gentoo \
+			$(DESTDIR)/etc/init.d/iscsi-target; \
+	elif [ -f /etc/slackware-version ]; then \
+		install -vD -m 755 etc/initd/initd \
+			$(DESTDIR)/etc/rc.d/iscsi-target; \
+	else \
+		install -vD -m 755 etc/initd/initd \
+			$(DESTDIR)/etc/init.d/iscsi-target; \
+	fi
+
+install-doc: install-man
+	@ok=true; for f in $(DOCS) ; \
+		do [ -e $$f ] || \
+			{ echo $$f missing ; ok=false; } ; \
+	done ; $$ok
+	@set -e; for f in $(DOCS) ; do \
+		install -v -D -m 644 $$f \
+			$(DESTDIR)$(DOCDIR)/$$f ; \
+	done
+
+install-man:
+	@ok=true; for f in $(MANPAGES) ; \
+		do [ -e doc/manpages/$$f ] || \
+			{ echo doc/manpages/$$f missing ; ok=false; } ; \
+	done ; $$ok
+	@set -e; for f in $(MANPAGES) ; do \
+		s=$${f##*.}; \
+		install -v -D -m 644 doc/manpages/$$f \
+			$(DESTDIR)$(MANDIR)/man$$s/$$f ; \
+	done
+
+uninstall: uninstall-kernel depmod uninstall-usr uninstall-etc uninstall-doc
+
+uninstall-kernel:
+	rm -f $(DESTDIR)$(INSTALL_MOD_PATH)$(KMOD)/iscsi/iscsi_trgt.ko
+	@if [ -f /etc/debian_version ]; then \
+		find $(DESTDIR)$(INSTALL_MOD_PATH)/lib/modules/$(KVER) \
+			-name iscsi_trgt.ko.distrib -type f \
+			-exec /bin/sh -c "dpkg-divert --remove --rename \
+				\`dirname {}\`/iscsi_trgt.ko" \;; \
+	else \
+		find $(DESTDIR)$(INSTALL_MOD_PATH)/lib/modules/$(KVER) \
+			-name iscsi_trgt.ko.orig -type f \
+			-execdir mv \{\} iscsi_trgt.ko \;; \
+        fi
+
+uninstall-usr:
+	@rm -f $(DESTDIR)/usr/sbin/ietd
+	@rm -f $(DESTDIR)/usr/sbin/ietadm
+
+uninstall-etc: uninstall-initd
+
+uninstall-initd:
+	if [ -f /etc/debian_version ]; then \
+		rm -f $(DESTDIR)/etc/init.d/iscsi-target; \
+	elif [ -f /etc/redhat-release ]; then \
+		rm -f $(DESTDIR)/etc/rc.d/init.d/iscsi-target; \
+	elif [ -f /etc/gentoo-release ]; then \
+		rm -f $(DESTDIR)/etc/init.d/iscsi-target; \
+	elif [ -f /etc/slackware-version ]; then \
+		rm -f $(DESTDIR)/etc/rc.d/iscsi-target; \
+	else \
+		rm -f $(DESTDIR)/etc/init.d/iscsi-target; \
+	fi
+
+uninstall-doc: uninstall-man
+	rm -rf $(DESTDIR)$(DOCDIR)
+
+uninstall-man:
+	set -e; for f in $(MANPAGES) ; do \
+		s=$${f##*.}; \
+		rm -f $(DESTDIR)$(MANDIR)/man$$s/$$f ; \
+	done
+
+clean:
+	$(MAKE) -C usr clean
+	$(MAKE) -C $(KSRC) SUBDIRS=$(shell pwd)/kernel clean
+
+distclean: unpatch clean
+	find . -name \*.orig -exec rm -f \{\} \;
+	find . -name \*.rej -exec rm -f \{\} \;
+	find . -name \*~ -exec rm -f \{\} \;
+	find . -name Module.symvers -exec rm -f \{\} \;
+
 
=== added file '.pc/debian-changes-1.4.20.2-5/RELEASE_NOTES'
--- .pc/debian-changes-1.4.20.2-5/RELEASE_NOTES	1970-01-01 00:00:00 +0000
+++ .pc/debian-changes-1.4.20.2-5/RELEASE_NOTES	2011-06-29 14:00:59 +0000
@@ -0,0 +1,176 @@
+               =========================================
+               Release Notes for iSCSI Enterprise Target
+               =========================================
+
+                            July 14th, 2010
+                            Version 1.4.20.2
+                            ----------------
+
+The IET development team is pleased to announce the release of version
+1.4.20.2 of the iSCSI Enterprise Target.
+
+This release is a bugfix release. The following issues were fixed:
+- various issues in the iSNS code resulting from buffer boundaries not being
+  checked (CVE-2010-2221)
+- memory leak in the iSNS ACL code
+- the kernel module allocating UAs in atomic context could cause a soft lockup
+- handling iSCSI logins was faulty, causing issues with QLogic HBAs
+- ietd was not sufficiently protected against multiple instances
+- ietd exited when trying to remove targets with active sessions
+- MD5/SHA-1 used for CHAP authentication were faulty on big endian platforms
+
+               =========================================
+               Release Notes for iSCSI Enterprise Target
+               =========================================
+
+                            April 25th, 2010
+                            Version 1.4.20.1
+                            ----------------
+
+The IET development team is pleased to announce the release of version
+1.4.20.1 of the iSCSI Enterprise Target.
+
+This release is a bugfix release. The following issues were fixed:
+- VMWare ESX does not recognize existing IET LUs after upgrading from older IET
+  versions to 1.4.20 because of a change in the SCSI SN. The old behaviour from
+  IET versions < 1.4.20 is restored with this release. Thanks to Steffen Plotner
+  for pointing it out and helping in debugging / testing.
+- With kernel 2.6.33 the utsrelease.h used by IETs backward compatibility
+  infrastructure was moved to a new location. This lead to compilation issues
+  with kernels >= 2.6.33.
+
+
+                            April 13th, 2010
+                             Version 1.4.20
+                             --------------
+
+The IET development team is pleased to announce the release of version
+1.4.20 of the iSCSI Enterprise Target.
+
+This release includes numerous bug fixes and compatibility improvements
+and it is highly recommended that all production servers be on this
+version for compatibility and stability reasons.
+
+We would like to thank all those that contributed for this release. As
+we are a small project we rely heavily on user contributions and we
+welcome all who wish to participate in improving IET.
+
+New since 1.4.19:
+
+New Features:
+- Added ability to atomically delete all targets/sessions/connections
+- Added ability to move existing kernel module on install and move
+  back on uninstall per distribution requirements (dpkg-divert)
+- Added support for logical sector sizes (512, 1024, 2048, 4096)
+  and devices whose native sector size > 512
+- Added case insensitive volume parameter processing
+- Added ability to set temporary target redirections via ietadm
+  for existing targets
+- Added support for auto-generating MD5 hash for SCSI ID, if not
+  specified, and auto-setting SCSI SN to the hex of the SCSI ID,
+  if not specified.
+
+Experimental Features:
+- Added experimental support for multiple connections per session
+
+Updated Features:
+- Updated kernel compatibility to 2.6.33
+- Updated RPM and DKMS build/install/maintenance handling
+- Updated kernel and usr ioctl functions to perform version checking
+  between kernel module and ietd
+
+Fixes:
+- Fixed bug in ietd text final text response
+- Fixed bug with reservation release during session deletion
+- Fixed compliance issues around ABORT TASK/ABORT TASK SET
+- Fixed fileio so it should work with all file system types
+- Fixed session and connection reinstatement to work per RFC
+- Fixed performance issue with CFQ scheduler
+- Fixed corner case with fileio_sync missing last page in a sync
+- Fixed bug in ietd where an interface without an address would
+  cause it to go into an infinite loop.
+- Fixed race condition between issuing ioctls during kernel cleanup
+- Fixed race condition between ietd shutdown and initiators which
+  would cause initiators to fail to reconnect properly
+- Fixed a race condition with data-out handling with multiple
+  connections per session
+- Fixed possible wthread lost wakeup condition when the number of
+  wthreads is less then the number of CPUs/cores.
+
+
+                          November 15th, 2009
+                             Version 1.4.19
+                             --------------
+
+The IET development team is pleased to announce the release of version
+1.4.19 of the iSCSI Enterprise Target.
+
+This release includes numerous bug fixes and compatibility improvements
+and it is highly recommended that all production servers be on this
+version for compatibility and stability reasons.
+
+We would like to thank all those that contributed for this release. As
+we are a small project we rely heavily on user contributions and we
+welcome all who wish to participate in improving IET.
+
+New since 1.4.18:
+
+- Added ability to limit number of sessions per target (MaxSessions)
+- Added NOP-In heartbeat for connections so dropped or abandoned
+  connections and their sessions close faster
+- Fixed a serious bug in initiators.allow code
+- Updated RPM .spec file with added SuSE compatibility and better
+  overall cross-platform friendliness.
+
+
+                           October 5th, 2009
+                             Version 1.4.18
+                             --------------
+
+The version numbering of IET has changed in this release from the 0.X.X
+series to the 1.X.X series. We did this because we felt that, since IET
+has been stable and in production use for many years now, it deserved a
+version number that better reflected that stability and maturity. We
+are preserving the minor/maintenance numbers though in order to help
+maintain some consistency between prior and future releases.
+
+New since 0.4.17:
+
+- Added support for recent kernel versions up to 2.6.31
+- Added support for even more targets via:
+	* multi PDU support during discovery
+	* improved procfs support
+	* global thread pool (module parameter)
+- Added support to list all target addresses during discovery, with
+  the ability to filter those addresses via targets.allow
+- Replaced initiators.deny with an implicit deny on a failure to find
+  a match in initiators.allow while preserving backwards compatiblity
+  with existing installations (iff initiators.deny doesn't exist)
+- Added ability to filter initiators by their IQNs as well as their
+  IP addresses, using basic regex pattern to specify IQNs
+- Moved the config files to /etc/iet/ while preserving backwards
+  compatiblity with existing installations (iff a valid config file
+  exists in the new location will it be used)
+- Added support for Unit Attention Conditions
+	* SCSI RESERVE/RELEASE now issues a UAC on reservation loss
+- Added support for DKMS and building RPM packages right from the
+  source archive (from either the source tar ball or subversion)
+
+Plus many many bug and compatibility fixes (see ChangeLog for details)
+
+
+                         Special thanks go to:
+                         ---------------------
+
+Shreyansh Jain, Lars Ellenberg, Ming Zhang, Francois Micaux, Cheng Renquan,
+Andreas Florath, sdrb, Xie Gang, Stefan Rubner, FUJITA Tomonori, Oliver R.,
+Matthew Wild, Kevin Ye, Min Tran, Luca Berra, Kun Huang, VMware, spren,
+Jagadish Kumar, Steven Umbehocker, Emmanuel Florac, Harshal Shete,
+Tim Westervoorde, Andrei Tanas, Chris Siebenmann, Manoj Iyer, Rob Caldwell
+Pasi Karkkainen, Eugen Rieck and to all list members who submitted bug
+reports, suggestions and comments.
+
+Without whose contributions IET would not have been possible.
+
+Arne Redlich & Ross Walker
+
 
=== added file '.pc/debian-changes-1.4.20.2-5/dkms.conf'
--- .pc/debian-changes-1.4.20.2-5/dkms.conf	1970-01-01 00:00:00 +0000
+++ .pc/debian-changes-1.4.20.2-5/dkms.conf	2011-06-29 14:00:59 +0000
@@ -0,0 +1,61 @@
+#
+# Master copy of dkms.conf for iscsitarget
+#
+
+PACKAGE_NAME="iscsitarget"
+PACKAGE_VERSION="1.4.20.2"
+MOD_PATH=${dkms_tree}/${PACKAGE_NAME}/${PACKAGE_VERSION}
+
+BUILT_MODULE_NAME="iscsi_trgt"
+BUILT_MODULE_LOCATION="kernel"
+DEST_MODULE_LOCATION="/kernel/iscsi"
+
+MAKE="make -C ${kernel_source_dir} SUBDIRS=${dkms_tree}/${PACKAGE_NAME}/${PACKAGE_VERSION}/build/kernel modules"
+
+CLEAN="make -C ${kernel_source_dir} SUBDIRS=${dkms_tree}/${PACKAGE_NAME}/${PACKAGE_VERSION}/build/kernel clean"
+
+AUTOINSTALL="yes"
+
+#
+# Patches newest - oldest, distro spec at bottom
+#
+
+PATCH[0]="compat-2.6.32.patch"
+PATCH_MATCH[0]="2\.6\.(9|14|15|16|17|18|19|20|21|22|23|24|25|26|27|28|29|30|31|32)"
+
+PATCH[1]="compat-2.6.31.patch"
+PATCH_MATCH[1]="2\.6\.(9|14|15|16|17|18|19|20|21|22|23|24|25|26|27|28|29|30|31)"
+
+PATCH[2]="compat-2.6.30.patch"
+PATCH_MATCH[2]="2\.6\.(9|14|15|16|17|18|19|20|21|22|23|24|25|26|27|28|29|30)"
+
+PATCH[3]="compat-2.6.29.patch"
+PATCH_MATCH[3]="2\.6\.(9|14|15|16|17|18|19|20|21|22|23|24|25|26|27|28|29)"
+
+PATCH[4]="compat-2.6.28.patch"
+PATCH_MATCH[4]="2\.6\.(9|14|15|16|17|18|19|20|21|22|23|24|25|26|27|28)"
+
+PATCH[5]="compat-2.6.25-2.6.27.patch"
+PATCH_MATCH[5]="2\.6\.(9|14|15|16|17|18|19|20|21|22|23|24|25|26|27)"
+
+PATCH[6]="compat-2.6.24.patch"
+PATCH_MATCH[6]="2\.6\.(9|14|15|16|17|18|19|20|21|22|23|24)"
+
+PATCH[7]="compat-2.6.23.patch"
+PATCH_MATCH[7]="2\.6\.(9|14|15|16|17|18|19|20|21|22|23)"
+
+PATCH[8]="compat-2.6.22.patch"
+PATCH_MATCH[8]="2\.6\.(9|14|15|16|17|18|19|20|21|22)"
+
+PATCH[9]="compat-2.6.19-2.6.21.patch"
+PATCH_MATCH[9]="2\.6\.(9|14|15|16|17|18|19|20|21)"
+
+PATCH[10]="compat-2.6.14-2.6.18.patch"
+PATCH_MATCH[10]="2\.6\.(9|14|15|16|17|18)"
+
+PATCH[11]="compat-sles10sp2.patch"
+PATCH_MATCH[11]="2\.6\.16\.60-.*"
+
+PATCH[12]="compat-rhel4.patch"
+PATCH_MATCH[12]="2\.6\.9-.*\.(el|plus\.c4)"
+
 
=== added directory '.pc/debian-changes-1.4.20.2-5/doc'
=== added directory '.pc/debian-changes-1.4.20.2-5/doc/manpages'
=== added file '.pc/debian-changes-1.4.20.2-5/doc/manpages/ietd.conf.5'
--- .pc/debian-changes-1.4.20.2-5/doc/manpages/ietd.conf.5	1970-01-01 00:00:00 +0000
+++ .pc/debian-changes-1.4.20.2-5/doc/manpages/ietd.conf.5	2011-06-29 14:00:59 +0000
@@ -0,0 +1,336 @@
+.\" Process this file with
+.\" groff -man -Tascii ietd.conf.5
+.\"
+.TH "IETD.CONF" "5" "27 July 2005" "A. Lehmann, M. Zhang and A. Redlich" "File formats"
+.SH "NAME"
+/etc/ietd.conf \- configuration for iSCSI Enterprise Target Daemon
+.SH "SYNOPSIS"
+/etc/ietd.conf
+.SH "DESCRIPTION"
+/etc/ietd.conf contains configuration information for the
+.B ietd (8)
+command. This is the place, where you configure your iSCSI targets and daemon defaults.
+.P
+Only lines starting with `#' are ignored. Putting '#' in the middle of a line is disallowed. A line may be extended across multiple lines by making the last character a backslash.
+.P
+The "Yes" and "No" for parameter values are case sensitive. The parameter names are case insensitive.
+.P
+The file consists of a global part and zero or more "Target" stanzas. Everything until the first target definition belongs to the global configuration.
+
+Here is an example:
+
+IncomingUser joe secret
+.br
+OutgoingUser jack secret2
+
+Target iqn.2001\-04.com.example:storage.disk2.sys1.xyz
+    IncomingUser jim othersecret
+    OutgoingUser james yetanothersecret
+    Lun 0 Path=/dev/sdc,Type=fileio
+    Lun 1 Blocks=10000,BlockSize=4096,Type=nullio
+    Alias Test
+    HeaderDigest None
+    DataDigest None
+    MaxConnections 1
+    MaxSessions 0
+    InitialR2T Yes
+    ImmediateData No
+    MaxRecvDataSegmentLength 8192
+    MaxXmitDataSegmentLength 8192
+    MaxBurstLength 262144
+    FirstBurstLength 65536
+    DefaultTime2Wait 2
+    DefaultTime2Retain 20
+    MaxOutstandingR2T 8
+    NOPInterval 0
+    NOPTimeout 0
+    DataPDUInOrder Yes
+    DataSequenceInOrder Yes
+    ErrorRecoveryLevel 0
+.P
+Stanzas start with the word "Target" and the target name. This name must be a globally unique name, as defined by the iSCSI standard : the "iSCSI Qualified Name". The daemon brings the targets up in the order listed.
+.SH "GLOBAL OPTIONS"
+Global Options are case sensitive.
+.TP
+.B [IncomingUser <username> <password>]
+The
+.I <username>
+and
+.I <password>
+used during discovery sessions to authenticate iSCSI initiators. Several of those can be specified for discovery. If no
+.B IncomingUser
+is specified, any initiator is allowed to open a discovery session.
+.RS
+HINT: RFC 3720 requires
+.I <password>
+to be 12 characters long. This is enforced e.g. by MS Initiator.
+.RE
+.TP
+.B [OutgoingUser <username> <password>]
+The
+.I <username>
+and
+.I <password>
+used during discovery sessions to authenticate the target to initiators. Only one outgoing
+.I <username>/<password>
+combination may be specified.
+.RS
+HINT: RFC 3720 requires
+.I <password>
+to be 12 characters long. This is enforced e.g. by MS Initiator.
+.RE
+.TP
+.B Target iqn.<yyyy\-mm>.<tld.domain.some.host>[:<identifier>]
+A target definition and the target name. The targets name (the
+.B iSCSI Qualified Name
+) must be a globally unique name (as defined by the iSCSI standard) and has to start with
+.I iqn
+followed by a single dot. The EUI\-64 form is not supported.
+.I <yyyy\-mm>
+is the date (year and month) at which the domain is valid. This has to be followed by a single dot and the reversed domain name.
+The optional
+.I <identifier>
+\- which is freely selectable \- has to be separated by a single colon. For further details please check the iSCSI spec.
+
+Here is an example:
+
+Target iqn.2004\-07.com.example.host:storage.disk2.sys1.xyz
+.SH "TARGET OPTIONS"
+Target options are also case sensitive.
+.TP
+.B [IncomingUser <username> <password>]
+The
+.I <username>
+and
+.I <password>
+used to authenticate the iSCSI initiators to this target. It may be different from the username and password in section GLOBAL OPTIONS, which is used for discovery. If you omit the
+.B IncomingUser
+Option, connections are allowed without authentication. A
+.I <password>
+has to be provided, if there is a
+.I <username>
+given. Specifying several different
+.B IncomingUser
+accounts is supported.
+.TP
+.B [OutgoingUser <username> <password>]
+The
+.I <username>
+and
+.I <password>
+used to authenticate this iSCSI target to initiators. Only one
+.B
+OutgoingUser
+per target is supported. It may be different from the username and password in section GLOBAL OPTIONS, which is used for discovery. A
+.I <password>
+has to be provided, if there is a
+.I <username>
+given.
+.TP
+.nf
+.B Lun <lun> Type=(fileio|blockio),Path=<device>[,ScsiId=<scsi_id>][,ScsiSN=<scsi_sn>][,IOMode=(wb|ro|wt)][,BlockSize=<size>]
+.TP
+.B Lun <lun> Type=nullio[,Blocks=<count>][,BlockSize=<size>]
+.fi
+.RS
+Parameters after <lun> should not contain any blank space characters except the first blank space after <lun> is needed.
+.P
+The value of
+.I <lun>
+can be from 0 to 16384. The first
+.I <lun>
+defined MUST be 0.
+.P
+In
+.I fileio
+mode (default), it defines a mapping between a "Logical Unit Number"
+.I <lun>
+and a given device
+.I <device>
+, which can be any block device (including regular block devices like hdX and sdX and virtual block devices like LVM and Software RAID devices) or regular files.
+.P
+In
+.I blockio
+mode, it defines a mapping between a "Logical Unit Number"
+.I <lun>
+and a given block device
+.I <device>.
+This mode will perform direct block i/o with the device, bypassing page-cache for all operations. This allows for efficient handling of non-aligned sector transfers (virtualized environments) and large block transfers (media servers). This mode works ideally with high-end storage HBAs and for applications that either do not need caching between application and disk or need the large block throughput.
+.P
+A
+.I ScsiId
+can be specified to assign a unique SCSI ID (VPD 0x83) to an iSCSI volume. This is used by initiator hosts to uniquely identify a SCSI volume. This is necessary with multipath\-aware hosts accessing the same
+.I <device>
+through multiple iSCSI sessions. The
+.I <scsi_id>
+must not exceed 16 characters.
+.P
+Also a
+.I ScsiSN
+can be specified to assign a unique serial number (VPD 0x80) to an iSCSI volume. This is used by certain initiator hosts (e.g. VMware ESX) to uniquely identify a SCSI volume. This is essential when used in conjunction with VMware ESX hosts accessing the same
+.I <device>
+through multiple iSCSI sessions. The
+.I <scsi_sn>
+must not exceed 16 characters.
+.P
+By default LUNs are writable, employing write-through caching. By setting
+.I IOMode
+to "ro" a LUN can be set to read only mode. Setting
+.I IOMode
+to "wb" will enable write-back caching of a LUN. Setting
+.I IOMode
+to "wt" is only symbolic as that is the default behavior.
+.P
+.B NOTE:
+.I IOMode
+"wb" is ignored when employing blockio as it performs no caching.
+.P
+.B WARNING: IOMode=wb could lead to serious data loss from an unexpected system failure (power loss, system crash). Use at your own risk!
+.P
+You can specify a logical
+.I BlockSize
+for an iSCSI volume. This
+.I <size>
+must be one of (512, 1024, 2048, 4096). If
+.I BlockSize
+isn't specified the default is 512 bytes for fileio, and the logical block size of the lower level device for blockio (which typically is 512 bytes).
+.P
+.B NOTE:
+For blockio, you CANNOT specify a
+.I BlockSize
+less than the lower level device's logical block size which is the minimum size the device can handle. An error will be reported and the LUN will fail to attach if you try. You CAN use fileio though to export a LUN with a
+.I BlockSize
+less than the lower level device's logical block size.
+.P
+.B WARNING: Once your data is written at a given BlockSize you cannot change this BlockSize without risking corruption of your existing data.
+.P
+In
+.I nullio
+mode, it defines a mapping between a "Logical Unit Number"
+.I <lun>
+and an unnamed virtual device with a specified number of
+.I Blocks
+of
+.I BlockSize
+bytes. If
+.I Blocks
+is not specified then it will default to 64GB's worth given the curent
+.I BlockSize
+which defaults to 512 bytes. This is ONLY useful for performance measurement purposes. All writes to this virtual device will be discarded and all reads will return random data.
+.P
+.B WARNING: By sending random kernel memory over the IP network you can potentially expose sensitive information.
+.RE
+.TP
+.B [Alias <aliasname>]
+This assigns an optional
+.I <aliasname>
+to the target.
+.TP
+.B [HeaderDigest <CRC32C|None>]
+Optional. If set to "CRC32C" and the initiator is configured accordingly, the integrity of an iSCSI PDU's header segments will be protected by a CRC32C checksum. The default is "None". Note that header digests are not supported during discovery sessions.
+.TP
+.B [DataDigest <CRC32C|None>]
+Optional. If set to "CRC32C" and the initiator is configured accordingly, the integrity of an iSCSI PDU's data segment will be protected by a CRC32C checksum. The default is "None". Note that data digests are not supported during discovery sessions.
+.TP
+.B [MaxConnections <value>]
+Optional. The number of connections within a session. Has to be set to "1" (in words: one), which is also the default since MC/S is not supported.
+.TP
+.B [MaxSessions <value>]
+Optional. The maximum number of sessions for this target. If this is set to 0 (wich is the default) there is no explicit session limit.
+.TP
+.B [InitialR2T <Yes|No>]
+Optional. If set to "Yes" (default), the initiator has to wait for the target to solicit SCSI data before sending it. Setting it to "No" allows the initiator to send a burst of
+.B FirstBurstLength
+bytes unsolicited right after and/or (depending on the setting of
+.B ImmediateData
+) together with the command. Thus setting it to "No" may improve performance.
+.TP
+.B [ImmediateData <Yes|No>]
+Optional. This allows the initiator to append unsolicited data to a command. To achieve better performance, this should be set to "Yes". The default is "No".
+.TP
+.B [MaxRecvDataSegmentLength <value>]
+Optional. Sets the maximum data segment length that can be received. The
+.I <value>
+should be set to multiples of PAGE_SIZE. Currently the maximum supported value is 64 * PAGE_SIZE, e.g. 262144 if PAGE_SIZE is 4kB. Configuring too large values may lead to problems allocating sufficient memory, which in turn may lead to SCSI commands timing out at the initiator host. The default value is 8192.
+.TP
+.B [MaxXmitDataSegmentLength <value>]
+Optional. Sets the maximum data segment length that can be sent. The
+.I <value>
+actually used is the minimum of
+.B MaxXmitDataSegmentLength
+and the
+.B MaxRecvDataSegmentLength
+announced by the initiator. The
+.I <value>
+should be set to multiples of PAGE_SIZE. Currently the maximum supported value is 64 * PAGE_SIZE, e.g. 262144 if PAGE_SIZE is 4kB. Configuring too large values may lead to problems allocating sufficient memory, which in turn may lead to SCSI commands timing out at the initiator host. The default value is 8192.
+.TP
+.B [MaxBurstLength <value>]
+Optional. Sets the maximum amount of either unsolicited or solicited data the initiator may send in a single burst. Any amount of data exceeding this value must be explicitly solicited by the target. The
+.I <value>
+should be set to multiples of PAGE_SIZE. Configuring too large values may lead to problems allocating sufficient memory, which in turn may lead to SCSI commands timing out at the initiator host. The default value is 262144.
+.TP
+.B [FirstBurstLength <value>]
+Optional. Sets the amount of unsolicited data the initiator may transmit in the first burst of a transfer either with and/or right after the command, depending on the settings of
+.B InitialR2T
+and
+.B ImmediateData
+.
+.I <value>
+should be set to multiples of PAGE_SIZE. Configuring too large values may lead to problems allocating sufficient memory, which in turn may lead to SCSI commands timing out at the initiator host. The default value is 65536.
+.TP
+.B [DefaultTime2Wait <value>]
+Currently not supported.
+.TP
+.B [DefaultTime2Retain <value>]
+Currently not supported.
+.TP
+.B [MaxOutstandingR2T <value>]
+Optional. Controls the maximum number of data transfers the target may request at once, each of up to
+.B MaxBurstLength
+bytes. The default is 1.
+.TP
+.B [DataPDUInOrder <Yes|No>]
+Optional. Has to be set to "Yes" \- which is also the default.
+.TP
+.B [DataSequenceInOrder <Yes|No>]
+Optional. Has to be set to "Yes" \- which is also the default.
+.TP
+.B [ErrorRecoveryLevel <value>]
+Optional. Has to be set to "0" (in words: zero), which is also the default.
+.TP
+.B [NOPInterval <value>]
+Optional. If
+.I value
+is non-zero, the initiator will be "ping"ed during phases of inactivity (i.e. no data transfers) every
+.I value
+seconds to verify the connection is still alive. If the initiator fails to respond within
+.B NOPTimeout
+seconds, the connection will be closed.
+.TP
+.B [NOPTimeout <value>]
+Optional. If a non-zero
+.B NOPInterval
+is used to periodically "ping" the initiator during phases of inactivity (i.e. no data transfers), the initiator must respond within
+.I value
+seconds, otherwise the connection will be closed. If
+.I value
+is set to zero or if it exceeds
+.B NOPInterval
+, it will be set to
+.B NOPInterval.
+.TP
+.B [Wthreads <value>]
+Optional. The iSCSI target employs several threads to perform the actual block I/O to the device. Depending on your hardware and your (expected) workload, the number of these threads may be carefully adjusted. The default value of 8 should be sufficient for most purposes.
+.TP
+.B [QueuedCommands <value>]
+Optional. This parameter defines a window of commands an initiator may send and that will be buffered by the target. Depending on your hardware and your (expected) workload, the
+.I value
+may be carefully adjusted. The default value of 32 should be sufficient for most purposes.
+.SH "KNOWN BUGS/LIMITATIONS"
+Currently (as of 0.4.11) not all iSCSI target parameters are used. Header and data digests are not supported during discovery sessions.
+.SH "SEE ALSO"
+.B ietd (8)
+.TP
+You should have a look at
+.B RFC 3720
+for all the glory details.
 
=== added directory '.pc/debian-changes-1.4.20.2-5/etc'
=== added file '.pc/debian-changes-1.4.20.2-5/etc/ietd.conf'
--- .pc/debian-changes-1.4.20.2-5/etc/ietd.conf	1970-01-01 00:00:00 +0000
+++ .pc/debian-changes-1.4.20.2-5/etc/ietd.conf	2011-06-29 14:00:59 +0000
@@ -0,0 +1,110 @@
+# Example iscsi target configuration
+#
+# Everything until the first target definition belongs
+# to the global configuration.
+#
+# "iSNSServer" is the iSNS server you want your portal to register
+# with.
+#
+# "iSNSAccessControl" is for enabling initiator access control
+# through the iSNS server.
+#
+# "IncomingUser" specifies credentials the initiator has to provide -
+# several of these are supported. If mutual CHAP shall be employed,
+# "OutgoingUser" specifies the user/pass combination the target will
+# provide - only one is supported.
+#
+# Leave them alone (keep them commented out) if you don't want to use
+# authentication for discovery sessions.
+
+#iSNSServer 192.168.1.16
+#iSNSAccessControl No
+
+#IncomingUser joe secret
+#OutgoingUser jack 12charsecret
+
+# Targets definitions start with "Target" and the target name.
+# The target name must be a globally unique name, the iSCSI
+# standard defines the "iSCSI Qualified Name" as follows:
+#
+# iqn.yyyy-mm.<reversed domain name>[:identifier]
+#
+# "yyyy-mm" is the date at which the domain is valid and the identifier
+# is freely selectable. For further details please check the iSCSI spec.
+
+#Target iqn.2001-04.com.example:storage.disk2.sys1.xyz
+	# CHAP Users
+	#
+	# The same rules as for discovery users apply here.
+	#
+	# Don't set them if you don't want to use CHAP authentication.
+	#
+	#IncomingUser joe secret
+	#OutgoingUser jim 12charpasswd
+	#
+	# Logical Unit definition
+	#
+	# Block devices, regular files (fileio only), LVM, and RAID
+	# can be offered to the initiators as a block device.
+	#
+	# Lun numbers MUST start with zero (each target needs a Lun 0)
+	#
+	#Lun 0 Path=/dev/sdc,Type=fileio,ScsiId=xyz,ScsiSN=xyz
+	#
+	# Alias name for this target (Not Used)
+	#
+	#Alias Test
+	#
+	# Various iSCSI parameters
+	# (not all are used right now, see also iSCSI spec for details)
+	#
+	# Outgoing SCSI data (initiator to target user data or command
+	# parameters) is sent as either solicited data or unsolicited data.
+	# Solicited data is sent in response to R2T PDUs.  Unsolicited data
+	# can be sent as part of an iSCSI command PDU sequence
+	# ("Immediate Data") or as a separate iSCSI data PDU sequence.
+	#
+	#MaxConnections		1		# Number of connections/session
+						# We only support 1
+	#MaxSessions		0		# Number of sessions/target
+						# 0 = no explicit limit
+	#InitialR2T		Yes		# Wait first for R2T
+						# Yes = no unsolicited data
+	#ImmediateData		Yes		# Data can accompany command
+						# Yes = cmnd/data in same PDU
+	#MaxRecvDataSegmentLength 8192		# Max data per PDU to receive
+	#MaxXmitDataSegmentLength 8192		# Max data per PDU to transmit
+	#MaxBurstLength		262144		# Max data per sequence (R2T)
+	#FirstBurstLength	65536		# Max unsolicited data sequence
+	#DefaultTime2Wait	2		# Secs wait for ini to log out
+						# Not used
+	#DefaultTime2Retain	20		# Secs keep cmnds after log out
+						# Not used
+	#MaxOutstandingR2T	1		# Max outstanding R2Ts per cmnd
+	#DataPDUInOrder		Yes		# Data in PDUs is ordered
+						# We only support ordered
+	#DataSequenceInOrder	Yes		# PDUs in sequence are ordered
+						# We only support ordered
+	#ErrorRecoveryLevel	0		# We only support level 0
+	#HeaderDigest		None,CRC32C	# PDU header checksum algo list
+						# None or CRC32C
+						# If only one is set then the
+						# initiator must agree to it
+						# or the connection will fail
+	#DataDigest		None,CRC32C	# PDU data checksum algo list
+						# Same as above
+	#MaxSessions		0		# Maximum number of sessions to
+						# this target - 0 = unlimited
+	#NOPInterval		0		# Send a NOP-In ping each after
+						# that many seconds if the conn
+						# is otherwise idle - 0 = off
+	#NOPTimeout		0 		# Wait that many seconds for a
+						# response on a NOP-In ping
+						# If 0 or > NOPInterval, NOPInterval
+						# is used!
+	#
+	# Various target parameters
+	#
+	#Wthreads		8		# Number of IO threads
+	#QueuedCommands		32		# Number of queued commands
+
 
=== added directory '.pc/debian-changes-1.4.20.2-5/etc/initd'
=== added file '.pc/debian-changes-1.4.20.2-5/etc/initd/initd'
--- .pc/debian-changes-1.4.20.2-5/etc/initd/initd	1970-01-01 00:00:00 +0000
+++ .pc/debian-changes-1.4.20.2-5/etc/initd/initd	2011-06-29 14:00:59 +0000
@@ -0,0 +1,39 @@
+#!/bin/bash
+#
+# Start the iSCSI Enterprise Target.
+#
+
+PATH=/sbin:/bin:/usr/sbin:/usr/bin
+OPTIONS=""
+
+if [ -f /etc/sysconfig/iscsi-target ]; then
+	. /etc/sysconfig/iscsi-target
+fi
+
+start_server()
+{
+	modprobe -q crc32c
+	modprobe iscsi_trgt
+	/usr/sbin/ietd $OPTIONS
+}
+
+stop_server()
+{
+	ietadm --op delete
+	killall ietd
+}
+
+case "$1" in
+	start)
+		start_server
+		;;
+	stop)
+		stop_server
+		;;
+	*)
+		echo "Usage: {start|stop}" >&2
+		exit 1
+		;;
+esac
+
+exit 0
 
=== added file '.pc/debian-changes-1.4.20.2-5/etc/initd/initd.debian'
--- .pc/debian-changes-1.4.20.2-5/etc/initd/initd.debian	1970-01-01 00:00:00 +0000
+++ .pc/debian-changes-1.4.20.2-5/etc/initd/initd.debian	2011-06-29 14:00:59 +0000
@@ -0,0 +1,75 @@
+#!/bin/sh
+#
+# chkconfig: - 39 35
+# description: Starts and stops the iSCSI target
+# debianized start-stop script
+
+PATH=/sbin:/bin:/usr/sbin:/usr/bin
+DAEMON=/usr/sbin/ietd
+OPTIONS=""
+PIDFILE=/var/run/iscsi_trgt.pid
+
+if [ -f /lib/init/vars.sh ]; then
+	. /lib/init/vars.sh
+fi
+
+if [ -f /lib/lsb/init-functions ]; then
+	. /lib/lsb/init-functions
+fi
+
+if [ -f /etc/sysconfig/iscsi-target ]; then
+	. /etc/sysconfig/iscsi-target
+fi
+
+RETVAL=0
+
+ietd_start()
+{
+	log_daemon_msg "Starting iSCSI Target" "ietd"
+	modprobe -q crc32c
+	modprobe iscsi_trgt
+	start-stop-daemon --start --exec $DAEMON --pidfile $PIDFILE -- $OPTIONS
+	log_end_msg $?
+}
+
+ietd_stop()
+{
+	log_daemon_msg "Stopping iSCSI Target" "ietd"
+	ietadm --op delete
+	start-stop-daemon --stop --exec $DAEMON --pidfile $PIDFILE
+	rm -f $PIDFILE
+	log_end_msg $?
+}
+
+ietd_status()
+{
+	PID=`pidof ietd`
+	if [ $PID ]; then
+		echo "iSCSI Target (pid $PID) is running..."
+	else
+		echo "iSCSI Target is stopped."
+		exit 1
+	fi
+}
+
+case "$1" in
+  start)
+        ietd_start
+        ;;
+  stop)
+        ietd_stop
+        ;;
+  restart)
+        ietd_stop
+	sleep 1
+	ietd_start
+        ;;
+  status)
+	ietd_status
+        ;;
+  *)
+        echo $"Usage: $0 {start|stop|restart|status}"
+        exit 1
+esac
+
+exit 0
 
=== added file '.pc/debian-changes-1.4.20.2-5/etc/initd/initd.gentoo'
--- .pc/debian-changes-1.4.20.2-5/etc/initd/initd.gentoo	1970-01-01 00:00:00 +0000
+++ .pc/debian-changes-1.4.20.2-5/etc/initd/initd.gentoo	2011-06-29 14:00:59 +0000
@@ -0,0 +1,39 @@
+#!/sbin/runscript
+#
+# Start the iSCSI Enterprise Target.
+#
+# To define start-up options, create a config file called
+# /etc/conf.d/iscsi-target, in it define OPTIONS="...".
+
+depend()
+{
+	use net
+	need localmount
+	after bootmisc
+}
+
+start()
+{
+	ebegin "Starting iSCSI Target"
+	modprobe -q crc32c
+	modprobe iscsi_trgt
+	start-stop-daemon --start --exec /usr/sbin/ietd --pidfile /var/run/iscsi_trgt.pid -- $OPTIONS
+	eend $?
+}
+
+stop()
+{
+	ebegin "Stopping iSCSI Target"
+	ietadm --op delete
+	start-stop-daemon --stop --exec /usr/sbin/ietd --pidfile /var/run/iscsi_trgt.pid
+	rm -f /var/run/iscsi_trgt.pid
+	eend $?
+}
+
+restart()
+{
+	stop
+	sleep 1
+	start
+}
+
 
=== added directory '.pc/debian-changes-1.4.20.2-5/include'
=== added file '.pc/debian-changes-1.4.20.2-5/include/iet_u.h'
--- .pc/debian-changes-1.4.20.2-5/include/iet_u.h	1970-01-01 00:00:00 +0000
+++ .pc/debian-changes-1.4.20.2-5/include/iet_u.h	2011-06-29 14:00:59 +0000
@@ -0,0 +1,151 @@
+#ifndef _IET_U_H
+#define _IET_U_H
+
+#define IET_VERSION_STRING	"1.4.20.2"
+
+/* The maximum length of 223 bytes in the RFC. */
+#define ISCSI_NAME_LEN	256
+#define ISCSI_ARGS_LEN	2048
+
+#define ISCSI_LISTEN_PORT	3260
+
+#define SCSI_ID_LEN	16
+#define SCSI_SN_LEN	(SCSI_ID_LEN * 2)
+
+#ifndef aligned_u64
+#define aligned_u64 unsigned long long __attribute__((aligned(8)))
+#endif
+
+struct module_info {
+	char version[128];
+};
+
+struct target_info {
+	u32 tid;
+	char name[ISCSI_NAME_LEN];
+};
+
+struct volume_info {
+	u32 tid;
+	u32 lun;
+	aligned_u64 args_ptr;
+	u32 args_len;
+};
+
+struct session_info {
+	u32 tid;
+
+	aligned_u64 sid;
+	char initiator_name[ISCSI_NAME_LEN];
+	u32 exp_cmd_sn;
+	u32 max_cmd_sn;
+};
+
+#define DIGEST_ALL	(DIGEST_NONE | DIGEST_CRC32C)
+#define DIGEST_NONE		(1 << 0)
+#define DIGEST_CRC32C           (1 << 1)
+
+struct conn_info {
+	u32 tid;
+	aligned_u64 sid;
+
+	u32 cid;
+	u32 stat_sn;
+	u32 exp_stat_sn;
+	int header_digest;
+	int data_digest;
+	int fd;
+};
+
+enum {
+	key_initial_r2t,
+	key_immediate_data,
+	key_max_connections,
+	key_max_recv_data_length,
+	key_max_xmit_data_length,
+	key_max_burst_length,
+	key_first_burst_length,
+	key_default_wait_time,
+	key_default_retain_time,
+	key_max_outstanding_r2t,
+	key_data_pdu_inorder,
+	key_data_sequence_inorder,
+	key_error_recovery_level,
+	key_header_digest,
+	key_data_digest,
+	key_ofmarker,
+	key_ifmarker,
+	key_ofmarkint,
+	key_ifmarkint,
+	session_key_last,
+};
+
+enum {
+	key_wthreads,
+	key_target_type,
+	key_queued_cmnds,
+	key_nop_interval,
+	key_nop_timeout,
+	target_key_last,
+};
+
+enum {
+	key_session,
+	key_target,
+};
+
+struct iscsi_param_info {
+	u32 tid;
+	aligned_u64 sid;
+
+	u32 param_type;
+	u32 partial;
+
+	u32 session_param[session_key_last];
+	u32 target_param[target_key_last];
+};
+
+enum iet_event_state {
+	E_CONN_CLOSE,
+};
+
+struct iet_event {
+	u32 tid;
+	aligned_u64 sid;
+	u32 cid;
+	u32 state;
+};
+
+#define	DEFAULT_NR_WTHREADS	8
+#define	MIN_NR_WTHREADS		1
+#define	MAX_NR_WTHREADS		128
+
+#define	DEFAULT_NR_QUEUED_CMNDS	32
+#define	MIN_NR_QUEUED_CMNDS	1
+#define	MAX_NR_QUEUED_CMNDS	256
+
+#define DEFAULT_NOP_INTERVAL 	0
+#define MIN_NOP_INTERVAL 	0
+#define MAX_NOP_INTERVAL 	90
+
+#define	DEFAULT_NOP_TIMEOUT 	0
+#define MIN_NOP_TIMEOUT 	0
+#define MAX_NOP_TIMEOUT		90
+
+#define NETLINK_IET	21
+
+#define GET_MODULE_INFO _IOW('i', 20, struct module_info)
+#define ADD_TARGET _IOWR('i', 21, struct target_info)
+#define DEL_TARGET _IOW('i', 22, struct target_info)
+#define ADD_VOLUME _IOW('i', 24, struct volume_info)
+#define DEL_VOLUME _IOW('i', 25, struct volume_info)
+#define ADD_SESSION _IOW('i', 26, struct session_info)
+#define DEL_SESSION _IOW('i', 27, struct session_info)
+#define GET_SESSION_INFO _IOWR('i', 28, struct session_info)
+#define ADD_CONN _IOW('i', 29, struct conn_info)
+#define DEL_CONN _IOW('i', 30, struct conn_info)
+#define GET_CONN_INFO _IOWR('i', 31, struct conn_info)
+#define ISCSI_PARAM_SET _IOW('i', 32, struct iscsi_param_info)
+#define ISCSI_PARAM_GET _IOWR('i', 33, struct iscsi_param_info)
+
+#endif
 
=== added file '.pc/debian-changes-1.4.20.2-5/iscsitarget.spec'
--- .pc/debian-changes-1.4.20.2-5/iscsitarget.spec	1970-01-01 00:00:00 +0000
+++ .pc/debian-changes-1.4.20.2-5/iscsitarget.spec	2011-06-29 14:00:59 +0000
@@ -0,0 +1,507 @@
+##
+## Global Package Definitions
+##
+
+## IET Release
+%define iet_version 1.4.20.2
+
+## Package Revision
+%define revision 1
+
+## Build Options (modify to tune your build)
+# Build DKMS kernel module
+#
+# Two passes through rpmbuild are required, once for the userland
+# binary package and another time with --target=noarch for the
+# kernel dkms package.
+%define dkms 0
+
+# Build weak modules (KABI tracking) if supported
+#
+# If it isn't supported it will figure that out below
+%define weak 1
+
+# Build from SVN repository
+%define svn 0
+%define svn_rev HEAD
+%define svn_repo https://iscsitarget.svn.sourceforge.net/svnroot/iscsitarget
+
+## Platform Definitions (you shouldn't need to modify these)
+# Determine distribution
+%define is_suse		!%(test -e /etc/SuSE-release; echo $?)
+%define is_fedora	!%(test -e /etc/fedora-release; echo $?)
+%define is_redhat	!%(test -e /etc/redhat-release; echo $?)
+%define is_mandrake	!%(test -e /etc/mandrake-release; echo $?)
+%define is_mandriva	!%(test -e /etc/mandriva-release; echo $?)
+
+# Define kernel version information
+%{!?kernel:	%define kernel %(uname -r)}
+
+%define kver	%(echo %{kernel} | sed -e 's/default//' -e 's/pae//i' -e 's/xen//' -e 's/smp//' -e 's/bigmem//' -e 's/hugemem//' -e 's/enterprise//' -e 's/-$//')
+%define krel	%(echo %{kver} | sed -e 's/-/_/g')
+%define ktype	%(echo kernel-%{kernel} | sed -e 's/%{kver}//' -e 's/--/-/' -e 's/-$//')
+
+%define moddir  /lib/modules/%{kernel}
+
+# Set location of tools
+%{!?__chkconfig:	%define __chkconfig	/sbin/chkconfig}
+%{!?__depmod:		%define __depmod	/sbin/depmod}
+%{!?__dkms:		%define __dkms		/usr/sbin/dkms}
+%{!?__find:		%define __find		/usr/bin/find}
+%{!?__modprobe:		%define __modprobe	/sbin/modprobe}
+%{!?__service:		%define __service	/sbin/service}
+%{!?__svn:		%define __svn		/usr/bin/svn}
+%if %is_suse
+%{!?__weak_modules:	%define __weak_modules	/usr/lib/module-init-tools/weak-modules}
+%else
+%{!?__weak_modules:	%define __weak_modules	/sbin/weak-modules}
+%endif
+
+# Subversion build information
+%if %svn
+%if !%(test "%{iet_version}" = "trunk"; echo $?)
+%define svn_url %{svn_repo}/trunk
+%define svn_ver %(%{__svn} info -r %{svn_rev} --non-interactive %{svn_url} | awk '{if ($1 == "Revision:") {print "r"$2}}')
+%else
+%define svn_url %{svn_repo}/tags/%{iet_version}
+%endif
+%endif
+
+# Build weak module (KABI Tracking) on platforms that support it
+%if %weak
+%define weak !%(test -x %{__weak_modules}; echo $?)
+%endif
+
+# Basic regex filters for unwanted dependencies (used for weak modules)
+%if %weak
+%define pro_filter ""
+%if %is_redhat
+%define req_filter "\\(fsync_bdev\\|sync_page_range\\|vfs_write\\|get_io_context\\|copy_io_context\\|put_io_context\\)"
+%else
+%define req_filter ""
+%endif
+%endif
+
+# Define build user
+%define user	%(whoami)
+
+
+##
+## Userland Package
+##
+
+## Information
+Summary: iSCSI Enterprise Target
+Name: iscsitarget
+Version: %{?svn_ver: %{iet_version}_%{svn_ver}}%{!?svn_ver: %{iet_version}}
+Release: %{?revision: %{revision}}%{!?revision: 1}
+License: GPL
+Group: System Environment/Daemons
+URL: http://sourceforge.net/projects/iscsitarget/
+Packager: IET development team <[e-mail]>
+
+## Source files
+%if !%svn
+Source0: %{name}-%{version}.tar.gz
+%endif
+
+## Patches
+#Patch0: %{name}-%{version}-example.p
+
+## Install Requirements
+Requires: %{name}-kmod = %{version}
+
+## Build Requirements
+BuildRequires: kernel >= 2.6
+BuildRequires: gcc, make, patch, binutils, /usr/bin/install, openssl-devel
+%if %is_suse
+BuildRequires: kernel-source = %{kver}
+%else
+BuildRequires: %{ktype}-devel = %{kver}
+%endif
+%if %svn
+BuildRequires: subversion
+%endif
+
+## Build Definitions
+BuildRoot: %{_tmppath}/%{name}-%{version}-%{release}-%{user}
+
+## Description
+%description
+iSCSI Enterprise Target
+
+
+##
+## Kernel Module Package
+##
+%if %dkms
+%ifarch noarch
+%package -n dkms-%{name}
+
+## Information
+Summary: iSCSI Enterprise Target kernel module
+Group: System Environment/Kernel
+Release: %{release}
+
+## Install Requirements
+Requires: dkms >= 2, gcc, make, patch, binutils
+%if %is_suse
+Requires: kernel-source
+%else
+Requires: %{ktype}-devel
+%endif
+
+## Install Provides
+Provides: kernel-modules
+Provides: %{name}-kmod = %{version}
+
+## Install Conflicts
+Conflicts: kmod-%{name}
+
+## Description
+%description -n dkms-%{name}
+iSCSI Enterprise Target kernel module
+%endif
+%else
+%package -n kmod-%{name}
+
+## Information
+Summary: iSCSI Enterprise Target kernel module
+Group: System Environment/Kernel
+Release: %{release}_%{krel}
+
+## Install Requirements
+%if %weak
+Requires: %{ktype}
+Requires(post): %{__weak_modules}
+Requires(postun): %{__weak_modules}
+%else
+Requires: %{ktype} = %{kver}
+%endif
+Requires(post): %{__depmod}
+Requires(postun): %{__depmod}
+
+## Install Provides
+%if !%weak
+Provides: kernel-modules = %{kver}
+%endif
+Provides: %{name}-kmod = %{version}
+
+## Install Conflicts
+Conflicts: dkms-%{name}
+
+## Description
+%description -n kmod-%{name}
+iSCSI Enterprise Target kernel module
+%endif
+
+
+##
+## Package Creation
+##
+
+## Preparation
+%prep
+
+
+## Setup
+%if %svn
+%setup -q -D -T -c -n %{name}-%{version}
+if [ ! -f include/iet_u.h ]; then
+%{__svn} export --force -r %{svn_rev} --non-interactive -q %{svn_url} .
+%{__sed} -i -e "s/\(#define IET_VERSION_STRING\).*/\1\t\"%{version}\"/" include/iet_u.h
+# Patches to apply to SVN
+#%patch0 -p0
+fi
+%else
+%setup -q -n %{name}-%{version}
+# Patches to apply to release
+#%patch0 -p0
+%endif
+
+
+## Build
+%build
+%{__make} distclean
+%ifnarch noarch
+%if %dkms
+%{__make} usr
+%else
+%{__make}
+%endif
+%endif
+
+
+## Installation
+%install
+%{__rm} -rf %{buildroot}
+
+%ifnarch noarch
+%if %dkms
+%{__make} install-usr install-doc install-etc DISTDIR=%{buildroot} DESTDIR=%{buildroot}
+%else
+%{__make} install-usr install-doc install-etc install-kernel DISTDIR=%{buildroot} DESTDIR=%{buildroot}
+%{__rm} -f %{buildroot}/lib/modules/%{kernel}/modules.*
+%if !%(test -d %{buildroot}%{moddir}/extra/iscsi; echo $?)
+%define module	%{moddir}/extra/iscsi/iscsi_trgt.ko
+%else
+%define module	%{moddir}/kernel/iscsi/iscsi_trgt.ko
+%endif
+%endif
+%if %is_redhat || %is_fedora
+if [ -d %{buildroot}/etc/init.d ]; then
+	%{__mkdir} %{buildroot}/etc/rc.d
+	%{__mv} %{buildroot}/etc/init.d %{buildroot}/etc/rc.d
+fi
+%endif
+%if %(test -d %{buildroot}/etc/iet; echo $?)
+%define ietd_conf %{_sysconfdir}/iet/ietd.conf
+%define initiators_allow %{_sysconfdir}/iet/initiators.allow
+%define targets_allow %{_sysconfdir}/iet/targets.allow
+%else
+%define ietd_conf %{_sysconfdir}/ietd.conf
+%define initiators_allow %{_sysconfdir}/initiators.allow
+%define targets_allow %{_sysconfdir}/initiators.deny
+%endif
+%elseif %dkms
+%{__mkdir} -p %{buildroot}/usr/src/%{name}-%{version}
+%{__cp} -r COPYING dkms.conf include kernel patches %{buildroot}/usr/src/%{name}-%{version}
+%{__sed} -i -e "s/\(PACKAGE_VERSION=\).*/\1\"%{version}\"/" %{buildroot}/usr/src/%{name}-%{version}/dkms.conf
+%endif
+
+# Ugly hack to filter out unwanted dependencies
+%if %weak
+%global _use_internal_dependency_generator 0
+%if !%(test -n "%{pro_filter}"; echo $?)
+%define iet_provides %{_tmppath}/iet_provides-%{user}
+%{__cat} << EOF > %{iet_provides}
+%{__find_provides} "\$@" | %{__grep} -v %{pro_filter}
+exit 0
+EOF
+%{__chmod} 755 %{iet_provides}
+%define __find_provides %{iet_provides}
+%endif
+%if !%(test -n "%{req_filter}"; echo $?)
+%define iet_requires %{_tmppath}/iet_requires-%{user}
+%{__cat} << EOF > %{iet_requires}
+%{__find_requires} "\$@" | %{__grep} -v %{req_filter}
+exit 0
+EOF
+%{__chmod} 755 %{iet_requires}
+%define __find_requires %{iet_requires}
+%endif
+%endif
+
+
+## Cleaning
+%clean
+%{__rm} -rf %{buildroot}
+%if %{?iet_provides:1}%{!?iet_provides:0}
+%{__rm} -f %{iet_provides}
+%endif
+%if %{?iet_requires:1}%{!?iet_requires:0}
+%{__rm} -f %{iet_requires}
+%endif
+
+
+## Post-Install Script
+%ifnarch noarch
+%post
+%if %is_suse
+%{__ln} -s %{_initrddir}/iscsi-target %{_sbindir}/rciscsi-target
+%endif
+%{__chkconfig} --add iscsi-target
+%endif
+
+
+## Pre-Uninstall Script
+%ifnarch noarch
+%preun
+if [ "$1" = 0 ]; then
+    %{__service} iscsi-target stop &>/dev/null
+    %{__chkconfig} --del iscsi-target &>/dev/null
+fi
+%if %is_suse
+%{__rm} -f %{_sbindir}/rciscsi-target
+%endif
+%endif
+
+
+## Post-Uninstall Script
+%ifnarch noarch
+%postun
+if [ "$1" != 0 ]; then
+    %{__service} iscsi-target condrestart &>/dev/null
+fi
+%endif
+
+
+## Post-Install Script (Kernel Module)
+%if %dkms
+%ifarch noarch
+%post -n dkms-%{name}
+%{__dkms} add -m %{name} -v %{version} --rpm_safe_upgrade
+%{__dkms} build -m %{name} -v %{version}
+%{__dkms} install -m %{name} -v %{version} --force
+%endif
+%else
+%post -n kmod-%{name}
+%{__depmod} %{kernel} -a
+%if %weak
+if [ -x %{__weak_modules} ]; then
+    echo %{module} | %{__weak_modules} --add-modules
+fi
+%endif
+%endif
+
+
+## Pre-Uninstall Script (Kernel Module)
+%if %dkms
+%ifarch noarch
+%preun -n dkms-%{name}
+%{__dkms} remove -m %{name} -v %{version} --all --rpm_safe_upgrade
+%endif
+%else
+%preun -n kmod-%{name}
+%{__modprobe} -r -q --set-version %{kernel} iscsi_trgt
+%if %weak
+if [ -x %{__weak_modules} ]; then
+    echo %{module} | %{__weak_modules} --remove-modules
+fi
+%endif
+%endif
+
+
+## Post-Uninstall Script (Kernel Module)
+%if !%dkms
+%postun -n kmod-%{name}
+%{__depmod} %{kernel} -a
+%endif
+
+
+## File Catalog
+%ifnarch noarch
+%files
+%defattr(-, root, root)
+%{_sbindir}
+%{_mandir}/man?
+%{_initrddir}
+%config(noreplace) %{ietd_conf}
+%config(noreplace) %{initiators_allow}
+%config(noreplace) %{targets_allow}
+%doc %{_docdir}
+%endif
+
+
+## File Catalog (Kernel Module)
+%if %dkms
+%ifarch noarch
+%files -n dkms-%{name}
+%defattr(-, root, root)
+/usr/src/%{name}-%{version}
+%endif
+%else
+%files -n kmod-%{name}
+%defattr(-, root, root)
+/lib/modules/%{kernel}
+%endif
+
+
+%changelog
+* Tue May 11 2010 Ross Walker <rswwalker at gmail dot com> - 1.4.20.2
+- fixed recursive KERNELSRC define problem on SLES
+- updated SVN building to handle tagged builds
+- updated iet_version for next release
+- fixed install and file catalogs to work with previous versions
+
+* Tue Mar 30 2010 Ross Walker <rswwalker at gmail dot com> - 1.4.20
+- Changed DKMS module name to dkms-iscsitarget from kmod-iscsitarget
+- Added Provides: kernel-modules to DKMS module so multiple versions can
+  be managed by DKMS at once.
+- Added conflicts to dkms and kmod for each other so only one type can
+  be installed at a time.
+- Added --rpm_safe_upgrade to DKMS add/remove for RPM-DKMS sync
+- Added --force to DKMS install in case version changed, but code didn't
+
+* Mon Mar 29 2010 Ross Walker <rswwalker at gmail dot com> - 1.4.20
+- Updated SVN build defines.
+- Added weak build define to manually choose KABI tracking or per-kernel
+- Fixed issue with upgrading KABI tracking module by removing
+  Provides: kernel-modules to weak builds, kept for per-kernel builds
+  if you install weak built module, all per-kernels are uninstalled,
+  you cannot install a per-kernel while a weak built is installed.
+
+* Mon Mar 08 2010 Ross Walker <rswwalker at gmail dot com> - 1.4.19
+- removed openssl-devel build requirement
+
+* Wed Oct 14 2009 Ross Walker <rswwalker at gmail dot com> - 1.4.18
+- Added more macros for better cross-distro support
+- Modified Matt's update for better Redhat-SuSE compatibility
+
+* Fri Oct 09 2009 Matthew Wild <matthew.wild at stfc dot ac dot uk> - 1.4.18
+- Added openSuSE specific configuration
+- Tidied up files section for init.d|rc.d
+- run depmod with -a rather than -A option
+
+* Wed Sep 25 2009 Ross Walker <rswwalker at gmail dot com> - 0.4.17-244
+- SuSE puts weak-modules under /usr/lib/module-init-tools
+- Kernel module now located in /lib/modules/<kernel>/extra/iscsi
+
+* Wed Sep 25 2009 Ross Walker <rswwalker at gmail dot com> - 0.4.17-242
+- Added ability to build weak modules for platforms that support them
+- Cleaned up logic a little
+- Made safe for multi-user builds
+- Redacted old spec file maintainer's email address
+
+* Wed Sep 22 2009 Ross Walker <rswwalker at gmail dot com> - 0.4.17-236
+- Updated file catalog for new config directory
+
+* Wed Sep 09 2009 Ross Walker <rswwalker at gmail dot com> - 0.4.17-226
+- Added ability to build directly from subversion repo
+- Added ability to build dkms kernel module
+
+* Mon Nov 10 2008 Ross Walker <rswwalker at gmail dot com> - 0.4.17-177
+- Changed kernel-module naming to kmod
+- Updated versioning
+
+* Fri Feb 16 2007 Ross Walker <rswwalker at gmail dot com> - 0.4.14-96
+- Reworked spec file for latest release
+- Commented and cleaned up sections
+- Added additional documents to %files
+
+* Mon Nov 21 2005 Bastiaan Bakker <redacted> - 0.4.13-0.1266.1
+- upstream snapshot 1266
+- added condrestart patch
+- stop and start service on update or removal
+
+* Sun Nov 13 2005 Bastiaan Bakker <redacted> - 0.4.13-0.1264.2
+- run %post and %preun for kernel package, not main package
+
+* Sun Nov 13 2005 Bastiaan Bakker <redacted> - 0.4.13-0.1264.1
+- updated to snapshot 1264
+
+* Thu Nov 03 2005 Bastiaan Bakker <redacted> - 0.4.12-6
+- added openssl-devel build requirement
+- removed '.ko' extension in modprobe command
+
+* Wed Nov 02 2005 Bastiaan Bakker <redacted> - 0.4.12-5
+- fixed kernel-devel BuildRequires
+
+* Fri Sep 23 2005 Bastiaan Bakker <redacted> - 0.4.12-4
+- fixed modprobe -r 'FATAL' message
+- run depmod with correct kernel version
+
+* Fri Sep 23 2005 Bastiaan Bakker <redacted> - 0.4.12-3
+- added config files
+- set kernel module file permissions to 744
+- fixed provides/requires of kernel module
+- removed BuildArch restriction
+
+* Thu Sep 22 2005 Bastiaan Bakker <redacted> - 0.4.12-2
+- create separate subpackage for kernel module
+- include man pages
+- added kernel compatibility patch for kernels < 2.6.11
+
+* Wed Aug 03 2005 Bastiaan Bakker <redacted>
+- First version.
+
 
=== added directory '.pc/debian-changes-1.4.20.2-5/kernel'
=== added file '.pc/debian-changes-1.4.20.2-5/kernel/block-io.c'
--- .pc/debian-changes-1.4.20.2-5/kernel/block-io.c	1970-01-01 00:00:00 +0000
+++ .pc/debian-changes-1.4.20.2-5/kernel/block-io.c	2011-06-29 14:00:59 +0000
@@ -0,0 +1,327 @@
+/*
+ * Target device block I/O.
+ *
+ * Based on file I/O driver from FUJITA Tomonori
+ * (C) 2004 - 2005 FUJITA Tomonori <[e-mail]>
+ * (C) 2006 Andre Brinkmann <brinkman at hni dot upb dot de>
+ * (C) 2007 Ross Walker <rswwalker at hotmail dot com>
+ * (C) 2007 Ming Zhang <blackmagic02881 at gmail dot com>
+ * This code is licenced under the GPL.
+ */
+
+#include <linux/types.h>
+#include <linux/blkdev.h>
+#include <linux/parser.h>
+#include <linux/buffer_head.h>
+
+#include "iscsi.h"
+#include "iscsi_dbg.h"
+#include "iotype.h"
+
+struct blockio_data {
+	char *path;
+	struct block_device *bdev;
+};
+
+struct tio_work {
+	atomic_t error;
+	atomic_t bios_remaining;
+	struct completion tio_complete;
+};
+
+static void blockio_bio_endio(struct bio *bio, int error)
+{
+	struct tio_work *tio_work = bio->bi_private;
+
+	error = test_bit(BIO_UPTODATE, &bio->bi_flags) ? error : -EIO;
+
+	if (error)
+		atomic_set(&tio_work->error, error);
+
+	/* If last bio signal completion */
+	if (atomic_dec_and_test(&tio_work->bios_remaining))
+		complete(&tio_work->tio_complete);
+
+	bio_put(bio);
+}
+
+/*
+ * Blockio_make_request(): The function translates an iscsi-request into
+ * a number of requests to the corresponding block device.
+ */
+static int
+blockio_make_request(struct iet_volume *volume, struct tio *tio, int rw)
+{
+	struct blockio_data *bio_data = volume->private;
+	struct request_queue *bdev_q = bdev_get_queue(bio_data->bdev);
+	struct tio_work *tio_work;
+	struct bio *tio_bio = NULL, *bio = NULL, *biotail = NULL;
+
+	u32 offset = tio->offset;
+	u32 size = tio->size;
+	u32 tio_index = 0;
+
+	int max_pages = 1;
+	int err = 0;
+
+	loff_t ppos = ((loff_t) tio->idx << PAGE_SHIFT) + offset;
+
+	/* Calculate max_pages for bio_alloc (memory saver) */
+	if (bdev_q)
+		max_pages = bio_get_nr_vecs(bio_data->bdev);
+
+	tio_work = kzalloc(sizeof (*tio_work), GFP_KERNEL);
+	if (!tio_work)
+		return -ENOMEM;
+
+	atomic_set(&tio_work->error, 0);
+	atomic_set(&tio_work->bios_remaining, 0);
+	init_completion(&tio_work->tio_complete);
+
+	/* Main processing loop, allocate and fill all bios */
+	while (tio_index < tio->pg_cnt) {
+		bio = bio_alloc(GFP_KERNEL, min(max_pages, BIO_MAX_PAGES));
+		if (!bio) {
+			err = -ENOMEM;
+			goto out;
+		}
+
+		/* bi_sector is ALWAYS in units of 512 bytes */
+		bio->bi_sector = ppos >> 9;
+		bio->bi_bdev = bio_data->bdev;
+		bio->bi_end_io = blockio_bio_endio;
+		bio->bi_private = tio_work;
+
+		if (tio_bio)
+			biotail = biotail->bi_next = bio;
+		else
+			tio_bio = biotail = bio;
+
+		atomic_inc(&tio_work->bios_remaining);
+
+		/* Loop for filling bio */
+		while (tio_index < tio->pg_cnt) {
+			unsigned int bytes = PAGE_SIZE - offset;
+
+			if (bytes > size)
+				bytes = size;
+
+			if (!bio_add_page(bio, tio->pvec[tio_index], bytes, offset))
+				break;
+
+			size -= bytes;
+			ppos += bytes;
+
+			offset = 0;
+
+			tio_index++;
+		}
+	}
+
+	/* Walk the list, submitting bios 1 by 1 */
+	while (tio_bio) {
+		bio = tio_bio;
+		tio_bio = tio_bio->bi_next;
+		bio->bi_next = NULL;
+
+		submit_bio(rw, bio);
+	}
+
+	if (bdev_q && bdev_q->unplug_fn)
+		bdev_q->unplug_fn(bdev_q);
+
+	wait_for_completion(&tio_work->tio_complete);
+
+	err = atomic_read(&tio_work->error);
+
+	kfree(tio_work);
+
+	return err;
+out:
+	while (tio_bio) {
+		bio = tio_bio;
+		tio_bio = tio_bio->bi_next;
+
+		bio_put(bio);
+	}
+
+	kfree(tio_work);
+
+	return err;
+}
+
+static int
+blockio_open_path(struct iet_volume *volume, const char *path)
+{
+	struct blockio_data *bio_data = volume->private;
+	struct block_device *bdev;
+	int flags = FMODE_READ | (LUReadonly(volume) ? 0 : FMODE_WRITE);
+	int err = 0;
+
+	bio_data->path = kstrdup(path, GFP_KERNEL);
+	if (!bio_data->path)
+		return -ENOMEM;
+
+	bdev = open_bdev_exclusive(path, flags, THIS_MODULE);
+	if (IS_ERR(bdev)) {
+		err = PTR_ERR(bdev);
+		eprintk("Can't open device %s, error %d\n", path, err);
+		bio_data->bdev = NULL;
+	} else {
+		bio_data->bdev = bdev;
+		fsync_bdev(bio_data->bdev);
+	}
+
+	return err;
+}
+
+/* Create an enumeration of our accepted actions */
+enum
+{
+	opt_path, opt_ignore, opt_err,
+};
+
+/* Create a match table using our action enums and their matching options */
+static match_table_t tokens = {
+	{opt_path, "path=%s"},
+	{opt_ignore, "scsiid=%s"},
+	{opt_ignore, "scsisn=%s"},
+	{opt_ignore, "type=%s"},
+	{opt_ignore, "iomode=%s"},
+	{opt_ignore, "blocksize=%s"},
+	{opt_err, NULL},
+};
+
+static int
+parse_blockio_params(struct iet_volume *volume, char *params)
+{
+	struct blockio_data *info = volume->private;
+	int err = 0;
+	char *p, *q;
+
+	/* Loop through parameters separated by commas, look up our
+	 * parameter in match table, return enumeration and arguments
+	 * select case based on the returned enum and run the action */
+	while ((p = strsep(&params, ",")) != NULL) {
+		substring_t args[MAX_OPT_ARGS];
+		int token;
+		if (!*p)
+			continue;
+		iet_strtolower(p);
+		token = match_token(p, tokens, args);
+		switch (token) {
+		case opt_path:
+			if (info->path) {
+				iprintk("Target %s, LUN %u: "
+					"duplicate \"Path\" param\n",
+					volume->target->name, volume->lun);
+				err = -EINVAL;
+				goto out;
+			}
+			if (!(q = match_strdup(&args[0]))) {
+				err = -ENOMEM;
+				goto out;
+			}
+			err = blockio_open_path(volume, q);
+			kfree(q);
+			if (err < 0)
+				goto out;
+			break;
+		case opt_ignore:
+			break;
+		default:
+			iprintk("Target %s, LUN %u: unknown param %s\n",
+				volume->target->name, volume->lun, p);
+			return -EINVAL;
+		}
+	}
+
+	if (!info->path) {
+		iprintk("Target %s, LUN %u: missing \"Path\" param\n",
+			volume->target->name, volume->lun);
+		err = -EINVAL;
+	}
+
+  out:
+	return err;
+}
+
+static void
+blockio_detach(struct iet_volume *volume)
+{
+	struct blockio_data *bio_data = volume->private;
+	int flags = FMODE_READ | (LUReadonly(volume) ? 0 : FMODE_WRITE);
+
+	if (bio_data->bdev)
+		close_bdev_exclusive(bio_data->bdev, flags);
+	kfree(bio_data->path);
+
+	kfree(volume->private);
+}
+
+static int
+blockio_attach(struct iet_volume *volume, char *args)
+{
+	struct blockio_data *bio_data;
+	int err = 0;
+
+	if (volume->private) {
+		eprintk("Lun %u already attached on Target %s \n",
+			volume->lun, volume->target->name);
+		return -EBUSY;
+	}
+
+	bio_data = kzalloc(sizeof (*bio_data), GFP_KERNEL);
+	if (!bio_data)
+		return -ENOMEM;
+
+	volume->private = bio_data;
+
+	err = parse_blockio_params(volume, args);
+	if (!err) {
+		/* see Documentation/ABI/testing/sysfs-block */
+		unsigned bsz = bdev_logical_block_size(bio_data->bdev);
+		if (!volume->blk_shift)
+			volume->blk_shift = blksize_bits(bsz);
+		else if (volume->blk_shift < blksize_bits(bsz)) {
+			eprintk("Specified block size (%u) smaller than "
+				"device %s logical block size (%u)\n",
+				(1 << volume->blk_shift), bio_data->path, bsz);
+			err = -EINVAL;
+		}
+	}
+	if (err < 0) {
+		eprintk("Error attaching Lun %u to Target %s \n",
+			volume->lun, volume->target->name);
+		goto out;
+	}
+
+	volume->blk_cnt = bio_data->bdev->bd_inode->i_size >> volume->blk_shift;
+
+	/* Offer neither write nor read caching */
+	ClearLURCache(volume);
+	ClearLUWCache(volume);
+
+  out:
+	if (err < 0)
+		blockio_detach(volume);
+
+	return err;
+}
+
+static void
+blockio_show(struct iet_volume *volume, struct seq_file *seq)
+{
+	struct blockio_data *bio_data = volume->private;
+
+	/* Used to display blockio volume info in /proc/net/iet/volumes */
+	seq_printf(seq, " path:%s\n", bio_data->path);
+}
+
+struct iotype blockio = {
+	.name = "blockio",
+	.attach = blockio_attach,
+	.make_request = blockio_make_request,
+	.detach = blockio_detach,
+	.show = blockio_show,
+};
 
=== added file '.pc/debian-changes-1.4.20.2-5/kernel/config.c'
--- .pc/debian-changes-1.4.20.2-5/kernel/config.c	1970-01-01 00:00:00 +0000
+++ .pc/debian-changes-1.4.20.2-5/kernel/config.c	2011-06-29 14:00:59 +0000
@@ -0,0 +1,362 @@
+/*
+ * (C) 2004 - 2005 FUJITA Tomonori <[e-mail]>
+ *
+ * This code is licenced under the GPL.
+ */
+
+#include <linux/proc_fs.h>
+
+#include "iscsi.h"
+#include "iscsi_dbg.h"
+
+static DECLARE_MUTEX(ioctl_sem);
+
+struct proc_entries {
+	const char *name;
+	struct file_operations *fops;
+};
+
+static struct proc_entries iet_proc_entries[] =
+{
+	{"volume", &volume_seq_fops},
+	{"session", &session_seq_fops},
+};
+
+static struct proc_dir_entry *proc_iet_dir;
+
+void iet_procfs_exit(void)
+{
+	int i;
+
+	if (!proc_iet_dir)
+		return;
+
+	for (i = 0; i < ARRAY_SIZE(iet_proc_entries); i++)
+		remove_proc_entry(iet_proc_entries[i].name, proc_iet_dir);
+
+	remove_proc_entry(proc_iet_dir->name, proc_iet_dir->parent);
+}
+
+int iet_procfs_init(void)
+{
+	int i;
+	struct proc_dir_entry *ent;
+
+	if (!(proc_iet_dir = proc_mkdir("iet", init_net.proc_net)))
+		goto err;
+
+	for (i = 0; i < ARRAY_SIZE(iet_proc_entries); i++) {
+		ent = create_proc_entry(iet_proc_entries[i].name, 0, proc_iet_dir);
+		if (ent)
+			ent->proc_fops = iet_proc_entries[i].fops;
+		else
+			goto err;
+	}
+
+	return 0;
+
+err:
+	if (proc_iet_dir)
+		iet_procfs_exit();
+
+	return -ENOMEM;
+}
+
+static int get_module_info(unsigned long ptr)
+{
+	struct module_info info;
+	int err;
+
+	snprintf(info.version, sizeof(info.version), "%s", IET_VERSION_STRING);
+
+	err = copy_to_user((void *) ptr, &info, sizeof(info));
+	if (err)
+		return -EFAULT;
+
+	return 0;
+}
+
+static int get_conn_info(struct iscsi_target *target, unsigned long ptr)
+{
+	struct iscsi_session *session;
+	struct iscsi_conn *conn;
+	struct conn_info info;
+	int err;
+
+	err = copy_from_user(&info, (void *) ptr, sizeof(info));
+	if (err)
+		return -EFAULT;
+
+	session = session_lookup(target, info.sid);
+	if (!session)
+		return -ENOENT;
+
+	conn = conn_lookup(session, info.cid);
+	if (!conn)
+		return -ENOENT;
+
+	info.cid = conn->cid;
+	info.stat_sn = conn->stat_sn;
+	info.exp_stat_sn = conn->exp_stat_sn;
+
+	err = copy_to_user((void *) ptr, &info, sizeof(info));
+	if (err)
+		return -EFAULT;
+
+	return 0;
+}
+
+static int add_conn(struct iscsi_target *target, unsigned long ptr)
+{
+	struct iscsi_session *session;
+	struct conn_info info;
+	int err;
+
+	err = copy_from_user(&info, (void *) ptr, sizeof(info));
+	if (err)
+		return -EFAULT;
+
+	session = session_lookup(target, info.sid);
+	if (!session)
+		return -ENOENT;
+
+	return conn_add(session, &info);
+}
+
+static int del_conn(struct iscsi_target *target, unsigned long ptr)
+{
+	struct iscsi_session *session;
+	struct conn_info info;
+	int err;
+
+	err = copy_from_user(&info, (void *) ptr, sizeof(info));
+	if (err)
+		return -EFAULT;
+
+	session = session_lookup(target, info.sid);
+	if (!session)
+		return -ENOENT;
+
+	return conn_del(session, &info);
+}
+
+static int get_session_info(struct iscsi_target *target, unsigned long ptr)
+{
+	struct iscsi_session *session;
+	struct session_info info;
+	int err;
+
+	err = copy_from_user(&info, (void *) ptr, sizeof(info));
+	if (err)
+		return -EFAULT;
+
+	session = session_lookup(target, info.sid);
+	if (!session)
+		return -ENOENT;
+
+	info.exp_cmd_sn = session->exp_cmd_sn;
+	info.max_cmd_sn = session->max_cmd_sn;
+
+	err = copy_to_user((void *) ptr, &info, sizeof(info));
+	if (err)
+		return -EFAULT;
+
+	return 0;
+}
+
+static int add_session(struct iscsi_target *target, unsigned long ptr)
+{
+	struct session_info info;
+	int err;
+
+	err = copy_from_user(&info, (void *) ptr, sizeof(info));
+	if (err)
+		return -EFAULT;
+
+	return session_add(target, &info);
+}
+
+static int del_session(struct iscsi_target *target, unsigned long ptr)
+{
+	struct session_info info;
+	int err;
+
+	err = copy_from_user(&info, (void *) ptr, sizeof(info));
+	if (err)
+		return -EFAULT;
+
+	return session_del(target, info.sid);
+}
+
+static int add_volume(struct iscsi_target *target, unsigned long ptr)
+{
+	struct volume_info info;
+	int err;
+
+	err = copy_from_user(&info, (void *) ptr, sizeof(info));
+	if (err)
+		return -EFAULT;
+
+	return volume_add(target, &info);
+}
+
+static int del_volume(struct iscsi_target *target, unsigned long ptr)
+{
+	struct volume_info info;
+	int err;
+
+	err = copy_from_user(&info, (void *) ptr, sizeof(info));
+	if (err)
+		return -EFAULT;
+
+	return iscsi_volume_del(target, &info);
+}
+
+static int iscsi_param_config(struct iscsi_target *target, unsigned long ptr, int set)
+{
+	struct iscsi_param_info info;
+	int err;
+
+	err = copy_from_user(&info, (void *) ptr, sizeof(info));
+	if (err)
+		return -EFAULT;
+
+	err = iscsi_param_set(target, &info, set);
+	if (err < 0 || set)
+		return err;
+
+	err = copy_to_user((void *) ptr, &info, sizeof(info));
+	if (err)
+		return -EFAULT;
+
+	return 0;
+}
+
+static int add_target(unsigned long ptr)
+{
+	struct target_info info;
+	int err;
+
+	err = copy_from_user(&info, (void *) ptr, sizeof(info));
+	if (err)
+		return -EFAULT;
+
+	err = target_add(&info);
+	if (err < 0)
+		return err;
+
+	err = copy_to_user((void *) ptr, &info, sizeof(info));
+	if (err)
+		return -EFAULT;
+
+	return 0;
+}
+
+static long ioctl(struct file *file, unsigned int cmd, unsigned long arg)
+{
+	struct iscsi_target *target = NULL;
+	long err;
+	u32 id;
+
+	err = down_interruptible(&ioctl_sem);
+	if (err < 0)
+		return err;
+
+	if (cmd == GET_MODULE_INFO) {
+		err = get_module_info(arg);
+		goto done;
+	}
+
+	if (cmd == ADD_TARGET) {
+		err = add_target(arg);
+		goto done;
+	}
+
+	err = get_user(id, (u32 *) arg);
+	if (err < 0)
+		goto done;
+
+	/* locking handled in target_del */
+	if (cmd == DEL_TARGET) {
+		err = target_del(id);
+		goto done;
+	}
+
+	target = target_lookup_by_id(id);
+	if (!target) {
+		err = -ENOENT;
+		goto done;
+	}
+
+	err = target_lock(target, 1);
+	if (err < 0)
+		goto done;
+
+	switch (cmd) {
+	case ADD_VOLUME:
+		err = add_volume(target, arg);
+		break;
+
+	case DEL_VOLUME:
+		err = del_volume(target, arg);
+		break;
+
+	case ADD_SESSION:
+		err = add_session(target, arg);
+		break;
+
+	case DEL_SESSION:
+		err = del_session(target, arg);
+		break;
+
+	case GET_SESSION_INFO:
+		err = get_session_info(target, arg);
+		break;
+
+	case ISCSI_PARAM_SET:
+		err = iscsi_param_config(target, arg, 1);
+		break;
+
+	case ISCSI_PARAM_GET:
+		err = iscsi_param_config(target, arg, 0);
+		break;
+
+	case ADD_CONN:
+		err = add_conn(target, arg);
+		break;
+
+	case DEL_CONN:
+		err = del_conn(target, arg);
+		break;
+
+	case GET_CONN_INFO:
+		err = get_conn_info(target, arg);
+		break;
+	default:
+		eprintk("invalid ioctl cmd %x\n", cmd);
+		err = -EINVAL;
+	}
+
+	target_unlock(target);
+done:
+	up(&ioctl_sem);
+
+	return err;
+}
+
+static int release(struct inode *i __attribute__((unused)),
+		   struct file *f __attribute__((unused)))
+{
+	down(&ioctl_sem);
+	target_del_all();
+	up(&ioctl_sem);
+
+	return 0;
+}
+
+struct file_operations ctr_fops = {
+	.owner		= THIS_MODULE,
+	.unlocked_ioctl	= ioctl,
+	.compat_ioctl	= ioctl,
+	.release	= release
+};
 
=== added file '.pc/debian-changes-1.4.20.2-5/kernel/conn.c'
--- .pc/debian-changes-1.4.20.2-5/kernel/conn.c	1970-01-01 00:00:00 +0000
+++ .pc/debian-changes-1.4.20.2-5/kernel/conn.c	2011-06-29 14:00:59 +0000
@@ -0,0 +1,262 @@
+/*
+ * Copyright (C) 2002-2003 Ardis Technolgies <[e-mail]>
+ *
+ * Released under the terms of the GNU GPL v2.0.
+ */
+
+#include <linux/file.h>
+#include <linux/ip.h>
+#include <net/tcp.h>
+#include <scsi/scsi.h>
+
+#include "iscsi.h"
+#include "iscsi_dbg.h"
+#include "digest.h"
+
+static void print_conn_state(char *p, size_t size, unsigned long state)
+{
+	if (test_bit(CONN_ACTIVE, &state))
+		snprintf(p, size, "%s", "active");
+	else if (test_bit(CONN_CLOSING, &state))
+		snprintf(p, size, "%s", "closing");
+	else
+		snprintf(p, size, "%s", "unknown");
+}
+
+static void print_digest_state(char *p, size_t size, unsigned long flags)
+{
+	if (DIGEST_NONE & flags)
+		snprintf(p, size, "%s", "none");
+	else if (DIGEST_CRC32C & flags)
+		snprintf(p, size, "%s", "crc32c");
+	else
+		snprintf(p, size, "%s", "unknown");
+}
+
+void conn_info_show(struct seq_file *seq, struct iscsi_session *session)
+{
+	struct iscsi_conn *conn;
+	struct sock *sk;
+	char buf[64];
+
+	list_for_each_entry(conn, &session->conn_list, list) {
+		sk = conn->sock->sk;
+		switch (sk->sk_family) {
+		case AF_INET:
+			snprintf(buf, sizeof(buf),
+				 "%u.%u.%u.%u", NIPQUAD(inet_sk(sk)->inet_daddr));
+			break;
+		case AF_INET6:
+			snprintf(buf, sizeof(buf), "[%pI6]",
+				 &inet6_sk(sk)->daddr);
+			break;
+		default:
+			break;
+		}
+		seq_printf(seq, "\t\tcid:%u ip:%s ", conn->cid, buf);
+		print_conn_state(buf, sizeof(buf), conn->state);
+		seq_printf(seq, "state:%s ", buf);
+		print_digest_state(buf, sizeof(buf), conn->hdigest_type);
+		seq_printf(seq, "hd:%s ", buf);
+		print_digest_state(buf, sizeof(buf), conn->ddigest_type);
+		seq_printf(seq, "dd:%s\n", buf);
+	}
+}
+
+struct iscsi_conn *conn_lookup(struct iscsi_session *session, u16 cid)
+{
+	struct iscsi_conn *conn;
+
+	list_for_each_entry(conn, &session->conn_list, list) {
+		if (conn->cid == cid)
+			return conn;
+	}
+	return NULL;
+}
+
+static void iet_state_change(struct sock *sk)
+{
+	struct iscsi_conn *conn = sk->sk_user_data;
+	struct iscsi_target *target = conn->session->target;
+
+	if (sk->sk_state != TCP_ESTABLISHED)
+		conn_close(conn);
+	else
+		nthread_wakeup(target);
+
+	target->nthread_info.old_state_change(sk);
+}
+
+static void iet_data_ready(struct sock *sk, int len)
+{
+	struct iscsi_conn *conn = sk->sk_user_data;
+	struct iscsi_target *target = conn->session->target;
+
+	nthread_wakeup(target);
+	target->nthread_info.old_data_ready(sk, len);
+}
+
+/*
+ * @locking: grabs the target's nthread_lock to protect it from races with
+ * set_conn_wspace_wait()
+ */
+static void iet_write_space(struct sock *sk)
+{
+	struct iscsi_conn *conn = sk->sk_user_data;
+	struct network_thread_info *info = &conn->session->target->nthread_info;
+
+	spin_lock_bh(&info->nthread_lock);
+
+	if (sk_stream_wspace(sk) >= sk_stream_min_wspace(sk) &&
+	    test_bit(CONN_WSPACE_WAIT, &conn->state)) {
+		clear_bit(CONN_WSPACE_WAIT, &conn->state);
+		__nthread_wakeup(info);
+	}
+
+	spin_unlock_bh(&info->nthread_lock);
+
+	info->old_write_space(sk);
+}
+
+static void iet_socket_bind(struct iscsi_conn *conn)
+{
+	int opt = 1;
+	mm_segment_t oldfs;
+	struct iscsi_session *session = conn->session;
+	struct iscsi_target *target = session->target;
+
+	dprintk(D_GENERIC, "%llu\n", (unsigned long long) session->sid);
+
+	conn->sock = SOCKET_I(conn->file->f_dentry->d_inode);
+	conn->sock->sk->sk_user_data = conn;
+
+	write_lock_bh(&conn->sock->sk->sk_callback_lock);
+	target->nthread_info.old_state_change = conn->sock->sk->sk_state_change;
+	conn->sock->sk->sk_state_change = iet_state_change;
+
+	target->nthread_info.old_data_ready = conn->sock->sk->sk_data_ready;
+	conn->sock->sk->sk_data_ready = iet_data_ready;
+
+	target->nthread_info.old_write_space = conn->sock->sk->sk_write_space;
+	conn->sock->sk->sk_write_space = iet_write_space;
+	write_unlock_bh(&conn->sock->sk->sk_callback_lock);
+
+	oldfs = get_fs();
+	set_fs(get_ds());
+	conn->sock->ops->setsockopt(conn->sock, SOL_TCP, TCP_NODELAY, (void *)&opt, sizeof(opt));
+	set_fs(oldfs);
+}
+
+int conn_free(struct iscsi_conn *conn)
+{
+	dprintk(D_GENERIC, "%p %#Lx %u\n", conn->session,
+		(unsigned long long) conn->session->sid, conn->cid);
+
+	assert(atomic_read(&conn->nr_cmnds) == 0);
+	assert(list_empty(&conn->pdu_list));
+	assert(list_empty(&conn->write_list));
+
+	list_del(&conn->list);
+	list_del(&conn->poll_list);
+
+	del_timer_sync(&conn->nop_timer);
+	digest_cleanup(conn);
+	kfree(conn);
+
+	return 0;
+}
+
+static int iet_conn_alloc(struct iscsi_session *session, struct conn_info *info)
+{
+	struct iscsi_conn *conn;
+
+	dprintk(D_SETUP, "%#Lx:%u\n", (unsigned long long) session->sid, info->cid);
+
+	conn = kzalloc(sizeof(*conn), GFP_KERNEL);
+	if (!conn)
+		return -ENOMEM;
+
+	conn->session = session;
+	conn->cid = info->cid;
+	conn->stat_sn = info->stat_sn;
+	conn->exp_stat_sn = info->exp_stat_sn;
+
+	conn->hdigest_type = info->header_digest;
+	conn->ddigest_type = info->data_digest;
+	if (digest_init(conn) < 0) {
+		kfree(conn);
+		return -ENOMEM;
+	}
+
+	spin_lock_init(&conn->list_lock);
+	atomic_set(&conn->nr_cmnds, 0);
+	atomic_set(&conn->nr_busy_cmnds, 0);
+	INIT_LIST_HEAD(&conn->pdu_list);
+	INIT_LIST_HEAD(&conn->write_list);
+	INIT_LIST_HEAD(&conn->poll_list);
+	init_timer(&conn->nop_timer);
+
+	list_add(&conn->list, &session->conn_list);
+
+	set_bit(CONN_ACTIVE, &conn->state);
+
+	conn->file = fget(info->fd);
+	iet_socket_bind(conn);
+
+	list_add(&conn->poll_list, &session->target->nthread_info.active_conns);
+
+	nthread_wakeup(conn->session->target);
+
+	return 0;
+}
+
+void conn_close(struct iscsi_conn *conn)
+{
+	struct iscsi_cmnd *cmnd;
+	struct iscsi_session *session = conn->session;
+
+	if (test_and_clear_bit(CONN_ACTIVE, &conn->state))
+		set_bit(CONN_CLOSING, &conn->state);
+
+	spin_lock(&conn->list_lock);
+	list_for_each_entry(cmnd, &conn->pdu_list, conn_list) {
+		set_cmnd_tmfabort(cmnd);
+		if (cmnd->lun) {
+			ua_establish_for_session(session, cmnd->lun->lun, 0x47, 0x7f);
+			iscsi_cmnd_set_sense(cmnd, UNIT_ATTENTION, 0x6e, 0x0);
+		}
+	}
+	spin_unlock(&conn->list_lock);
+
+	nthread_wakeup(conn->session->target);
+}
+
+int conn_add(struct iscsi_session *session, struct conn_info *info)
+{
+	struct iscsi_conn *conn;
+	int err;
+
+	conn = conn_lookup(session, info->cid);
+	if (conn)
+		conn_close(conn);
+
+	err = iet_conn_alloc(session, info);
+	if (!err && conn)
+		err = -EEXIST;
+
+	return err;
+}
+
+int conn_del(struct iscsi_session *session, struct conn_info *info)
+{
+	struct iscsi_conn *conn;
+	int err = -EEXIST;
+
+	conn = conn_lookup(session, info->cid);
+	if (!conn)
+		return err;
+
+	conn_close(conn);
+
+	return 0;
+}
 
=== added file '.pc/debian-changes-1.4.20.2-5/kernel/iotype.c'
--- .pc/debian-changes-1.4.20.2-5/kernel/iotype.c	1970-01-01 00:00:00 +0000
+++ .pc/debian-changes-1.4.20.2-5/kernel/iotype.c	2011-06-29 14:00:59 +0000
@@ -0,0 +1,110 @@
+/*
+ * Manager for various I/O types.
+ * (C) 2004 - 2005 FUJITA Tomonori <[e-mail]>
+ * This code is licenced under the GPL.
+ */
+
+#include "iscsi.h"
+#include "iotype.h"
+#include "iscsi_dbg.h"
+
+static LIST_HEAD(iotypes);
+static rwlock_t iotypes_lock = RW_LOCK_UNLOCKED;
+
+static struct iotype *find_iotype(const char *name)
+{
+	struct iotype *iot = NULL;
+
+	list_for_each_entry(iot, &iotypes, iot_list) {
+		if (strcmp(iot->name, name) == 0)
+			return iot;
+	}
+	return NULL;
+}
+
+struct iotype *get_iotype(const char *name)
+{
+	struct iotype *iot;
+
+	read_lock(&iotypes_lock);
+	iot = find_iotype(name);
+	read_unlock(&iotypes_lock);
+
+	return iot;
+}
+
+void put_iotype(struct iotype *iot)
+{
+	if (!iot)
+		return;
+	return;
+}
+
+static int register_iotype(struct iotype *iot)
+{
+	int err = 0;
+	struct iotype *p;
+
+	write_lock(&iotypes_lock);
+
+	p = find_iotype(iot->name);
+	if (p)
+		err = -EBUSY;
+	else
+		list_add_tail(&iot->iot_list, &iotypes);
+
+	write_unlock(&iotypes_lock);
+
+	return err;
+}
+
+static int unregister_iotype(struct iotype *iot)
+{
+	int err = 0;
+	struct iotype *p;
+
+	write_lock(&iotypes_lock);
+
+	p = find_iotype(iot->name);
+	if (p && p == iot)
+		list_del_init(&iot->iot_list);
+	else
+		err = -EINVAL;
+
+	write_unlock(&iotypes_lock);
+
+
+	return err;
+}
+
+struct iotype *iotype_array[] = {
+	&fileio,
+	&blockio,
+	&nullio,
+};
+
+int iotype_init(void)
+{
+	int i, err;
+
+	for (i = 0; i < ARRAY_SIZE(iotype_array); i++) {
+		if (!(err = register_iotype(iotype_array[i])))
+			iprintk("Registered io type %s\n",
+						iotype_array[i]->name);
+		else {
+			eprintk("Failed to register io type %s\n",
+						iotype_array[i]->name);
+			break;
+		}
+	}
+
+	return err;
+}
+
+void iotype_exit(void)
+{
+	int i;
+
+	for (i = 0; i < ARRAY_SIZE(iotype_array); i++)
+		unregister_iotype(iotype_array[i]);
+}
 
=== added file '.pc/debian-changes-1.4.20.2-5/kernel/iscsi.c'
--- .pc/debian-changes-1.4.20.2-5/kernel/iscsi.c	1970-01-01 00:00:00 +0000
+++ .pc/debian-changes-1.4.20.2-5/kernel/iscsi.c	2011-06-29 14:00:59 +0000
@@ -0,0 +1,1967 @@
+/*
+ * Copyright (C) 2002-2003 Ardis Technolgies <[e-mail]>
+ * Copyright (C) 2008 Arne Redlich <[e-mail]>
+ *
+ * Released under the terms of the GNU GPL v2.0.
+ */
+
+#include <linux/module.h>
+#include <linux/hash.h>
+#include <net/tcp.h>
+#include <scsi/scsi.h>
+
+#include "iscsi.h"
+#include "iscsi_dbg.h"
+#include "iotype.h"
+
+unsigned long debug_enable_flags;
+unsigned long worker_thread_pool_size;
+
+static struct kmem_cache *iscsi_cmnd_cache;
+static u8 dummy_data[PAGE_SIZE];
+
+static int ctr_major;
+static char ctr_name[] = "ietctl";
+extern struct file_operations ctr_fops;
+
+static u32 cmnd_write_size(struct iscsi_cmnd *cmnd)
+{
+	struct iscsi_scsi_cmd_hdr *hdr = cmnd_hdr(cmnd);
+
+	if (hdr->flags & ISCSI_CMD_WRITE)
+		return be32_to_cpu(hdr->data_length);
+	return 0;
+}
+
+static u32 cmnd_read_size(struct iscsi_cmnd *cmnd)
+{
+	struct iscsi_scsi_cmd_hdr *hdr = cmnd_hdr(cmnd);
+
+	if (hdr->flags & ISCSI_CMD_READ) {
+		struct iscsi_rlength_ahdr *ahdr =
+			(struct iscsi_rlength_ahdr *)cmnd->pdu.ahs;
+
+		if (!(hdr->flags & ISCSI_CMD_WRITE))
+			return be32_to_cpu(hdr->data_length);
+		if (ahdr && ahdr->ahstype == ISCSI_AHSTYPE_RLENGTH)
+			return be32_to_cpu(ahdr->read_length);
+	}
+	return 0;
+}
+
+static void iscsi_device_queue_cmnd(struct iscsi_cmnd *cmnd)
+{
+	set_cmnd_waitio(cmnd);
+	wthread_queue(cmnd);
+}
+
+static void iscsi_scsi_queuecmnd(struct iscsi_cmnd *cmnd)
+{
+	struct iscsi_queue *queue = &cmnd->lun->queue;
+
+	dprintk(D_GENERIC, "%p\n", cmnd);
+
+	if ((cmnd->pdu.bhs.flags & ISCSI_CMD_ATTR_MASK) != ISCSI_CMD_UNTAGGED &&
+	    (cmnd->pdu.bhs.flags & ISCSI_CMD_ATTR_MASK) != ISCSI_CMD_SIMPLE) {
+		cmnd->pdu.bhs.flags &= ~ISCSI_CMD_ATTR_MASK;
+		cmnd->pdu.bhs.flags |= ISCSI_CMD_UNTAGGED;
+	}
+
+	spin_lock(&queue->queue_lock);
+
+	set_cmnd_queued(cmnd);
+
+	switch (cmnd->pdu.bhs.flags & ISCSI_CMD_ATTR_MASK) {
+	case ISCSI_CMD_UNTAGGED:
+	case ISCSI_CMD_SIMPLE:
+		if (!list_empty(&queue->wait_list) || queue->ordered_cmnd)
+			goto pending;
+		queue->active_cnt++;
+		break;
+
+	default:
+		BUG();
+	}
+	spin_unlock(&queue->queue_lock);
+
+	iscsi_device_queue_cmnd(cmnd);
+	return;
+ pending:
+	assert(list_empty(&cmnd->list));
+
+	list_add_tail(&cmnd->list, &queue->wait_list);
+	spin_unlock(&queue->queue_lock);
+	return;
+}
+
+static void iscsi_scsi_dequeuecmnd(struct iscsi_cmnd *cmnd)
+{
+	struct iscsi_queue *queue;
+
+	if (!cmnd->lun)
+		return;
+	queue = &cmnd->lun->queue;
+	spin_lock(&queue->queue_lock);
+	switch (cmnd->pdu.bhs.flags & ISCSI_CMD_ATTR_MASK) {
+	case ISCSI_CMD_UNTAGGED:
+	case ISCSI_CMD_SIMPLE:
+		--queue->active_cnt;
+		break;
+	case ISCSI_CMD_ORDERED:
+	case ISCSI_CMD_HEAD_OF_QUEUE:
+	case ISCSI_CMD_ACA:
+		BUG();
+	default:
+		/* Should the iscsi_scsi_queuecmnd func reject this ? */
+		break;
+	}
+
+	while (!list_empty(&queue->wait_list)) {
+		cmnd = list_entry(queue->wait_list.next, struct iscsi_cmnd, list);
+		switch ((cmnd->pdu.bhs.flags & ISCSI_CMD_ATTR_MASK)) {
+		case ISCSI_CMD_UNTAGGED:
+		case ISCSI_CMD_SIMPLE:
+			list_del_init(&cmnd->list);
+			queue->active_cnt++;
+			iscsi_device_queue_cmnd(cmnd);
+			break;
+		case ISCSI_CMD_ORDERED:
+		case ISCSI_CMD_HEAD_OF_QUEUE:
+		case ISCSI_CMD_ACA:
+			BUG();
+		}
+	}
+
+	spin_unlock(&queue->queue_lock);
+
+	return;
+}
+
+/**
+ * create a new command.
+ *
+ * iscsi_cmnd_create -
+ * @conn: ptr to connection (for i/o)
+ *
+ * @return    ptr to command or NULL
+ */
+
+struct iscsi_cmnd *cmnd_alloc(struct iscsi_conn *conn, int req)
+{
+	struct iscsi_cmnd *cmnd;
+
+	/* TODO: async interface is necessary ? */
+	cmnd = kmem_cache_alloc(iscsi_cmnd_cache, GFP_KERNEL|__GFP_NOFAIL);
+
+	memset(cmnd, 0, sizeof(*cmnd));
+	INIT_LIST_HEAD(&cmnd->list);
+	INIT_LIST_HEAD(&cmnd->pdu_list);
+	INIT_LIST_HEAD(&cmnd->conn_list);
+	INIT_LIST_HEAD(&cmnd->hash_list);
+	cmnd->conn = conn;
+	spin_lock(&conn->list_lock);
+	atomic_inc(&conn->nr_cmnds);
+	if (req)
+		list_add_tail(&cmnd->conn_list, &conn->pdu_list);
+	spin_unlock(&conn->list_lock);
+	cmnd->tio = NULL;
+
+	dprintk(D_GENERIC, "%p:%p\n", conn, cmnd);
+
+	return cmnd;
+}
+
+/**
+ * create a new command used as response.
+ *
+ * iscsi_cmnd_create_rsp_cmnd -
+ * @cmnd: ptr to request command
+ *
+ * @return    ptr to response command or NULL
+ */
+
+static struct iscsi_cmnd *iscsi_cmnd_create_rsp_cmnd(struct iscsi_cmnd *cmnd, int final)
+{
+	struct iscsi_cmnd *rsp = cmnd_alloc(cmnd->conn, 0);
+
+	if (final)
+		set_cmnd_final(rsp);
+	list_add_tail(&rsp->pdu_list, &cmnd->pdu_list);
+	rsp->req = cmnd;
+	return rsp;
+}
+
+static struct iscsi_cmnd *get_rsp_cmnd(struct iscsi_cmnd *req)
+{
+	return list_entry(req->pdu_list.prev, struct iscsi_cmnd, pdu_list);
+}
+
+static void iscsi_cmnds_init_write(struct list_head *send)
+{
+	struct iscsi_cmnd *cmnd = list_entry(send->next, struct iscsi_cmnd, list);
+	struct iscsi_conn *conn = cmnd->conn;
+	struct list_head *pos, *next;
+
+	spin_lock(&conn->list_lock);
+
+	list_for_each_safe(pos, next, send) {
+		cmnd = list_entry(pos, struct iscsi_cmnd, list);
+
+		dprintk(D_GENERIC, "%p:%x\n", cmnd, cmnd_opcode(cmnd));
+
+		list_del_init(&cmnd->list);
+		assert(conn == cmnd->conn);
+		list_add_tail(&cmnd->list, &conn->write_list);
+	}
+
+	spin_unlock(&conn->list_lock);
+
+	nthread_wakeup(conn->session->target);
+}
+
+static void iscsi_cmnd_init_write(struct iscsi_cmnd *cmnd)
+{
+	LIST_HEAD(head);
+
+	if (!list_empty(&cmnd->list)) {
+		eprintk("%x %x %x %x %lx %u %u %u %u %u %u %u %d %d\n",
+			cmnd_itt(cmnd), cmnd_ttt(cmnd), cmnd_opcode(cmnd),
+			cmnd_scsicode(cmnd), cmnd->flags, cmnd->r2t_sn,
+			cmnd->r2t_length, cmnd->is_unsolicited_data,
+			cmnd->target_task_tag, cmnd->outstanding_r2t,
+			cmnd->hdigest, cmnd->ddigest,
+			list_empty(&cmnd->pdu_list), list_empty(&cmnd->hash_list));
+
+		assert(list_empty(&cmnd->list));
+	}
+	list_add(&cmnd->list, &head);
+	iscsi_cmnds_init_write(&head);
+}
+
+static void do_send_data_rsp(struct iscsi_cmnd *cmnd)
+{
+	struct iscsi_conn *conn = cmnd->conn;
+	struct iscsi_cmnd *data_cmnd;
+	struct tio *tio = cmnd->tio;
+	struct iscsi_scsi_cmd_hdr *req = cmnd_hdr(cmnd);
+	struct iscsi_data_in_hdr *rsp;
+	u32 pdusize, expsize, scsisize, size, offset, sn;
+	LIST_HEAD(send);
+
+	dprintk(D_GENERIC, "%p\n", cmnd);
+	pdusize = conn->session->param.max_xmit_data_length;
+	expsize = cmnd_read_size(cmnd);
+	size = min(expsize, tio->size);
+	offset = 0;
+	sn = 0;
+
+	while (1) {
+		data_cmnd = iscsi_cmnd_create_rsp_cmnd(cmnd, size <= pdusize);
+		tio_get(tio);
+		data_cmnd->tio = tio;
+		rsp = (struct iscsi_data_in_hdr *)&data_cmnd->pdu.bhs;
+
+		rsp->opcode = ISCSI_OP_SCSI_DATA_IN;
+		rsp->itt = req->itt;
+		rsp->ttt = cpu_to_be32(ISCSI_RESERVED_TAG);
+		rsp->buffer_offset = offset;
+		rsp->data_sn = cpu_to_be32(sn);
+
+		if (size <= pdusize) {
+			data_cmnd->pdu.datasize = size;
+			rsp->flags = ISCSI_FLG_FINAL | ISCSI_FLG_STATUS;
+
+			scsisize = tio->size;
+			if (scsisize < expsize) {
+				rsp->flags |= ISCSI_FLG_RESIDUAL_UNDERFLOW;
+				size = expsize - scsisize;
+			} else if (scsisize > expsize) {
+				rsp->flags |= ISCSI_FLG_RESIDUAL_OVERFLOW;
+				size = scsisize - expsize;
+			} else
+				size = 0;
+			rsp->residual_count = cpu_to_be32(size);
+			list_add_tail(&data_cmnd->list, &send);
+
+			break;
+		}
+
+		data_cmnd->pdu.datasize = pdusize;
+
+		size -= pdusize;
+		offset += pdusize;
+		sn++;
+
+		list_add_tail(&data_cmnd->list, &send);
+	}
+
+	iscsi_cmnds_init_write(&send);
+}
+
+static struct iscsi_cmnd *create_scsi_rsp(struct iscsi_cmnd *req)
+{
+	struct iscsi_cmnd *rsp;
+	struct iscsi_scsi_cmd_hdr *req_hdr = cmnd_hdr(req);
+	struct iscsi_scsi_rsp_hdr *rsp_hdr;
+	struct iscsi_sense_data *sense;
+
+	rsp = iscsi_cmnd_create_rsp_cmnd(req, 1);
+
+	rsp_hdr = (struct iscsi_scsi_rsp_hdr *)&rsp->pdu.bhs;
+	rsp_hdr->opcode = ISCSI_OP_SCSI_RSP;
+	rsp_hdr->flags = ISCSI_FLG_FINAL;
+	rsp_hdr->response = ISCSI_RESPONSE_COMMAND_COMPLETED;
+	rsp_hdr->cmd_status = req->status;
+	rsp_hdr->itt = req_hdr->itt;
+
+	if (req->status == SAM_STAT_CHECK_CONDITION) {
+		assert(!rsp->tio);
+		rsp->tio = tio_alloc(1);
+		sense = (struct iscsi_sense_data *)
+			page_address(rsp->tio->pvec[0]);
+
+		assert(sense);
+		clear_page(sense);
+		sense->length = cpu_to_be16(IET_SENSE_BUF_SIZE);
+
+		memcpy(sense->data, req->sense_buf, IET_SENSE_BUF_SIZE);
+		rsp->pdu.datasize = sizeof(struct iscsi_sense_data) +
+			IET_SENSE_BUF_SIZE;
+
+		rsp->tio->size = (rsp->pdu.datasize + 3) & -4;
+		rsp->tio->offset = 0;
+	}
+
+	return rsp;
+}
+
+void iscsi_cmnd_set_sense(struct iscsi_cmnd *cmnd, u8 sense_key, u8 asc,
+			  u8 ascq)
+{
+	cmnd->status = SAM_STAT_CHECK_CONDITION;
+
+	cmnd->sense_buf[0] = 0xf0;
+	cmnd->sense_buf[2] = sense_key;
+	cmnd->sense_buf[7] = 6;	// Additional sense length
+	cmnd->sense_buf[12] = asc;
+	cmnd->sense_buf[13] = ascq;
+
+	/* Call to ACA/UAI handler */
+}
+
+static struct iscsi_cmnd *create_sense_rsp(struct iscsi_cmnd *req,
+					   u8 sense_key, u8 asc, u8 ascq)
+{
+	iscsi_cmnd_set_sense(req, sense_key, asc, ascq);
+	return create_scsi_rsp(req);
+}
+
+void send_scsi_rsp(struct iscsi_cmnd *req, void (*func)(struct iscsi_cmnd *))
+{
+	struct iscsi_cmnd *rsp;
+	struct iscsi_scsi_rsp_hdr *rsp_hdr;
+	u32 size;
+
+	func(req);
+	rsp = create_scsi_rsp(req);
+
+	switch (req->status) {
+	case SAM_STAT_GOOD:
+	case SAM_STAT_RESERVATION_CONFLICT:
+		rsp_hdr = (struct iscsi_scsi_rsp_hdr *) &rsp->pdu.bhs;
+		if ((size = cmnd_read_size(req)) != 0) {
+			rsp_hdr->flags |= ISCSI_FLG_RESIDUAL_UNDERFLOW;
+			rsp_hdr->residual_count = cpu_to_be32(size);
+		}
+		break;
+	default:
+		break;
+	}
+
+	iscsi_cmnd_init_write(rsp);
+}
+
+void send_data_rsp(struct iscsi_cmnd *req, void (*func)(struct iscsi_cmnd *))
+{
+	struct iscsi_cmnd *rsp;
+
+	func(req);
+
+	if (req->status == SAM_STAT_GOOD)
+		do_send_data_rsp(req);
+	else {
+		rsp = create_scsi_rsp(req);
+		iscsi_cmnd_init_write(rsp);
+	}
+}
+
+/**
+ * Free a command.
+ * Also frees the additional header.
+ *
+ * iscsi_cmnd_remove -
+ * @cmnd: ptr to command
+ */
+
+static void iscsi_cmnd_remove(struct iscsi_cmnd *cmnd)
+{
+	struct iscsi_conn *conn;
+
+	if (!cmnd)
+		return;
+
+	if (cmnd_timer_active(cmnd)) {
+		clear_cmnd_timer_active(cmnd);
+		del_timer_sync(&cmnd->timer);
+	}
+
+	dprintk(D_GENERIC, "%p\n", cmnd);
+	conn = cmnd->conn;
+	kfree(cmnd->pdu.ahs);
+
+	if (!list_empty(&cmnd->list)) {
+		struct iscsi_scsi_cmd_hdr *req = cmnd_hdr(cmnd);
+
+		eprintk("cmnd %p still on some list?, %x, %x, %x, %x, %x, %x, %x %lx\n",
+			cmnd, req->opcode, req->scb[0], req->flags, req->itt,
+			be32_to_cpu(req->data_length),
+			req->cmd_sn, be32_to_cpu(cmnd->pdu.datasize),
+			conn->state);
+
+		if (cmnd->req) {
+			struct iscsi_scsi_cmd_hdr *req = cmnd_hdr(cmnd->req);
+			eprintk("%p %x %u\n", req, req->opcode, req->scb[0]);
+		}
+		dump_stack();
+		BUG();
+	}
+	list_del(&cmnd->list);
+	spin_lock(&conn->list_lock);
+	atomic_dec(&conn->nr_cmnds);
+	list_del(&cmnd->conn_list);
+	spin_unlock(&conn->list_lock);
+
+	if (cmnd->tio)
+		tio_put(cmnd->tio);
+
+	kmem_cache_free(iscsi_cmnd_cache, cmnd);
+}
+
+static void cmnd_skip_pdu(struct iscsi_cmnd *cmnd)
+{
+	struct iscsi_conn *conn = cmnd->conn;
+	struct tio *tio = cmnd->tio;
+	char *addr;
+	u32 size;
+	int i;
+
+	eprintk("%x %x %x %u\n", cmnd_itt(cmnd), cmnd_opcode(cmnd),
+		cmnd_hdr(cmnd)->scb[0], cmnd->pdu.datasize);
+
+	if (!(size = cmnd->pdu.datasize))
+		return;
+
+	if (tio)
+		assert(tio->pg_cnt > 0);
+	else
+		tio = cmnd->tio = tio_alloc(1);
+
+	addr = page_address(tio->pvec[0]);
+	assert(addr);
+	size = (size + 3) & -4;
+	conn->read_size = size;
+	for (i = 0; size > PAGE_CACHE_SIZE; i++, size -= PAGE_CACHE_SIZE) {
+		assert(i < ISCSI_CONN_IOV_MAX);
+		conn->read_iov[i].iov_base = addr;
+		conn->read_iov[i].iov_len = PAGE_CACHE_SIZE;
+	}
+	conn->read_iov[i].iov_base = addr;
+	conn->read_iov[i].iov_len = size;
+	conn->read_msg.msg_iov = conn->read_iov;
+	conn->read_msg.msg_iovlen = ++i;
+}
+
+static void iscsi_cmnd_reject(struct iscsi_cmnd *req, int reason)
+{
+	struct iscsi_cmnd *rsp;
+	struct iscsi_reject_hdr *rsp_hdr;
+	struct tio *tio;
+	char *addr;
+
+	rsp = iscsi_cmnd_create_rsp_cmnd(req, 1);
+	rsp_hdr = (struct iscsi_reject_hdr *)&rsp->pdu.bhs;
+
+	rsp_hdr->opcode = ISCSI_OP_REJECT;
+	rsp_hdr->ffffffff = ISCSI_RESERVED_TAG;
+	rsp_hdr->reason = reason;
+
+	rsp->tio = tio = tio_alloc(1);
+	addr = page_address(tio->pvec[0]);
+	clear_page(addr);
+	memcpy(addr, &req->pdu.bhs, sizeof(struct iscsi_hdr));
+	tio->size = rsp->pdu.datasize = sizeof(struct iscsi_hdr);
+	cmnd_skip_pdu(req);
+
+	req->pdu.bhs.opcode = ISCSI_OP_PDU_REJECT;
+}
+
+static void cmnd_set_sn(struct iscsi_cmnd *cmnd, int set_stat_sn)
+{
+	struct iscsi_conn *conn = cmnd->conn;
+	struct iscsi_session *sess = conn->session;
+
+	if (set_stat_sn)
+		cmnd->pdu.bhs.sn = cpu_to_be32(conn->stat_sn++);
+	cmnd->pdu.bhs.exp_sn = cpu_to_be32(sess->exp_cmd_sn);
+	cmnd->pdu.bhs.max_sn = cpu_to_be32(sess->exp_cmd_sn + sess->max_queued_cmnds);
+}
+
+static void update_stat_sn(struct iscsi_cmnd *cmnd)
+{
+	struct iscsi_conn *conn = cmnd->conn;
+	u32 exp_stat_sn;
+
+	cmnd->pdu.bhs.exp_sn = exp_stat_sn = be32_to_cpu(cmnd->pdu.bhs.exp_sn);
+	dprintk(D_GENERIC, "%x,%x\n", cmnd_opcode(cmnd), exp_stat_sn);
+	if ((int)(exp_stat_sn - conn->exp_stat_sn) > 0 &&
+	    (int)(exp_stat_sn - conn->stat_sn) <= 0) {
+		// free pdu resources
+		cmnd->conn->exp_stat_sn = exp_stat_sn;
+	}
+}
+
+static int check_cmd_sn(struct iscsi_cmnd *cmnd)
+{
+	struct iscsi_session *session = cmnd->conn->session;
+	u32 cmd_sn;
+
+	cmnd->pdu.bhs.sn = cmd_sn = be32_to_cpu(cmnd->pdu.bhs.sn);
+
+	dprintk(D_GENERIC, "cmd_sn(%u) exp_cmd_sn(%u) max_cmd_sn(%u)\n",
+		cmd_sn, session->exp_cmd_sn, session->max_cmd_sn);
+
+	if  (between(cmd_sn, session->exp_cmd_sn, session->max_cmd_sn))
+		return 0;
+	else if (cmnd_immediate(cmnd))
+		return 0;
+
+	eprintk("sequence error: cmd_sn(%u) exp_cmd_sn(%u) max_cmd_sn(%u)\n",
+		cmd_sn, session->exp_cmd_sn, session->max_cmd_sn);
+
+	set_cmnd_tmfabort(cmnd);
+
+	return -ISCSI_REASON_PROTOCOL_ERROR;
+}
+
+static struct iscsi_cmnd *__cmnd_find_hash(struct iscsi_session *session, u32 itt, u32 ttt)
+{
+	struct list_head *head;
+	struct iscsi_cmnd *cmnd;
+
+	head = &session->cmnd_hash[cmnd_hashfn(itt)];
+
+	list_for_each_entry(cmnd, head, hash_list) {
+		if (cmnd->pdu.bhs.itt == itt) {
+			if ((ttt != ISCSI_RESERVED_TAG) && (ttt != cmnd->target_task_tag))
+				continue;
+			return cmnd;
+		}
+	}
+
+	return NULL;
+}
+
+static struct iscsi_cmnd *cmnd_find_hash(struct iscsi_session *session, u32 itt, u32 ttt)
+{
+	struct iscsi_cmnd *cmnd;
+
+	spin_lock(&session->cmnd_hash_lock);
+
+	cmnd = __cmnd_find_hash(session, itt, ttt);
+
+	spin_unlock(&session->cmnd_hash_lock);
+
+	return cmnd;
+}
+
+static int cmnd_insert_hash_ttt(struct iscsi_cmnd *cmnd, u32 ttt)
+{
+	struct iscsi_session *session = cmnd->conn->session;
+	struct iscsi_cmnd *tmp;
+	struct list_head *head;
+	int err = 0;
+	u32 itt = cmnd->pdu.bhs.itt;
+
+	head = &session->cmnd_hash[cmnd_hashfn(itt)];
+
+	spin_lock(&session->cmnd_hash_lock);
+
+	tmp = __cmnd_find_hash(session, itt, ttt);
+	if (!tmp) {
+		list_add_tail(&cmnd->hash_list, head);
+		set_cmnd_hashed(cmnd);
+	} else
+		err = -ISCSI_REASON_TASK_IN_PROGRESS;
+
+	spin_unlock(&session->cmnd_hash_lock);
+
+	return err;
+}
+
+static int cmnd_insert_hash(struct iscsi_cmnd *cmnd)
+{
+	int err;
+
+	dprintk(D_GENERIC, "%p:%x\n", cmnd, cmnd->pdu.bhs.itt);
+
+	if (cmnd->pdu.bhs.itt == ISCSI_RESERVED_TAG)
+		return -ISCSI_REASON_PROTOCOL_ERROR;
+
+	err = cmnd_insert_hash_ttt(cmnd, ISCSI_RESERVED_TAG);
+	if (!err) {
+		update_stat_sn(cmnd);
+		err = check_cmd_sn(cmnd);
+	} else if (!cmnd_immediate(cmnd))
+		set_cmnd_tmfabort(cmnd);
+
+	return err;
+}
+
+static void __cmnd_remove_hash(struct iscsi_cmnd *cmnd)
+{
+	list_del(&cmnd->hash_list);
+}
+
+static void cmnd_remove_hash(struct iscsi_cmnd *cmnd)
+{
+	struct iscsi_session *session = cmnd->conn->session;
+	struct iscsi_cmnd *tmp;
+
+	spin_lock(&session->cmnd_hash_lock);
+
+	tmp = __cmnd_find_hash(session, cmnd->pdu.bhs.itt,
+			       cmnd->target_task_tag);
+	if (tmp && tmp == cmnd)
+		__cmnd_remove_hash(tmp);
+	else
+		eprintk("%p:%x not found\n", cmnd, cmnd_itt(cmnd));
+
+	spin_unlock(&session->cmnd_hash_lock);
+}
+
+static void cmnd_skip_data(struct iscsi_cmnd *req)
+{
+	struct iscsi_cmnd *rsp;
+	struct iscsi_scsi_rsp_hdr *rsp_hdr;
+	u32 size;
+
+	rsp = get_rsp_cmnd(req);
+	rsp_hdr = (struct iscsi_scsi_rsp_hdr *)&rsp->pdu.bhs;
+	if (cmnd_opcode(rsp) != ISCSI_OP_SCSI_RSP) {
+		eprintk("unexpected response command %u\n", cmnd_opcode(rsp));
+		return;
+	}
+
+	size = cmnd_write_size(req);
+	if (size) {
+		rsp_hdr->flags |= ISCSI_FLG_RESIDUAL_UNDERFLOW;
+		rsp_hdr->residual_count = cpu_to_be32(size);
+	}
+	size = cmnd_read_size(req);
+	if (size) {
+		if (cmnd_hdr(req)->flags & ISCSI_CMD_WRITE) {
+			rsp_hdr->flags |= ISCSI_FLG_BIRESIDUAL_UNDERFLOW;
+			rsp_hdr->bi_residual_count = cpu_to_be32(size);
+		} else {
+			rsp_hdr->flags |= ISCSI_FLG_RESIDUAL_UNDERFLOW;
+			rsp_hdr->residual_count = cpu_to_be32(size);
+		}
+	}
+	req->pdu.bhs.opcode =
+		(req->pdu.bhs.opcode & ~ISCSI_OPCODE_MASK) | ISCSI_OP_SCSI_REJECT;
+
+	cmnd_skip_pdu(req);
+}
+
+static int cmnd_recv_pdu(struct iscsi_conn *conn, struct tio *tio, u32 offset, u32 size)
+{
+	int idx, i;
+	char *addr;
+
+	dprintk(D_GENERIC, "%p %u,%u\n", tio, offset, size);
+	offset += tio->offset;
+
+	if (!(offset < tio->offset + tio->size) ||
+	    !(offset + size <= tio->offset + tio->size)) {
+		eprintk("%u %u %u %u", offset, size, tio->offset, tio->size);
+		return -EIO;
+	}
+	assert(offset < tio->offset + tio->size);
+	assert(offset + size <= tio->offset + tio->size);
+
+	idx = offset >> PAGE_CACHE_SHIFT;
+	offset &= ~PAGE_CACHE_MASK;
+
+	conn->read_msg.msg_iov = conn->read_iov;
+	conn->read_size = size = (size + 3) & -4;
+	conn->read_overflow = 0;
+
+	i = 0;
+	while (1) {
+		assert(tio->pvec[idx]);
+		addr = page_address(tio->pvec[idx]);
+		assert(addr);
+		conn->read_iov[i].iov_base =  addr + offset;
+		if (offset + size <= PAGE_CACHE_SIZE) {
+			conn->read_iov[i].iov_len = size;
+			conn->read_msg.msg_iovlen = ++i;
+			break;
+		}
+		conn->read_iov[i].iov_len = PAGE_CACHE_SIZE - offset;
+		size -= conn->read_iov[i].iov_len;
+		offset = 0;
+		if (++i >= ISCSI_CONN_IOV_MAX) {
+			conn->read_msg.msg_iovlen = i;
+			conn->read_overflow = size;
+			conn->read_size -= size;
+			break;
+		}
+
+		idx++;
+	}
+
+	return 0;
+}
+
+static void set_offset_and_length(struct iet_volume *lu, u8 *cmd, loff_t *off, u32 *len)
+{
+	assert(lu);
+
+	switch (cmd[0]) {
+	case READ_6:
+	case WRITE_6:
+		*off = ((cmd[1] & 0x1f) << 16) + (cmd[2] << 8) + cmd[3];
+		*len = cmd[4];
+		if (!*len)
+			*len = 256;
+		break;
+	case READ_10:
+	case WRITE_10:
+	case WRITE_VERIFY:
+		*off = (u32)cmd[2] << 24 | (u32)cmd[3] << 16 |
+			(u32)cmd[4] << 8 | (u32)cmd[5];
+		*len = (cmd[7] << 8) + cmd[8];
+		break;
+	case READ_16:
+	case WRITE_16:
+		*off = (u64)cmd[2] << 56 | (u64)cmd[3] << 48 |
+			(u64)cmd[4] << 40 | (u64)cmd[5] << 32 |
+			(u64)cmd[6] << 24 | (u64)cmd[7] << 16 |
+			(u64)cmd[8] << 8 | (u64)cmd[9];
+		*len = (u32)cmd[10] << 24 | (u32)cmd[11] << 16 |
+			(u32)cmd[12] << 8 | (u32)cmd[13];
+		break;
+	default:
+		BUG();
+	}
+
+	*off <<= lu->blk_shift;
+	*len <<= lu->blk_shift;
+}
+
+static u32 translate_lun(u16 * data)
+{
+	u8 *p = (u8 *) data;
+	u32 lun = ~0U;
+
+	switch (*p >> 6) {
+	case 0:
+		lun = p[1];
+		break;
+	case 1:
+		lun = (0x3f & p[0]) << 8 | p[1];
+		break;
+	case 2:
+	case 3:
+	default:
+		eprintk("%u %u %u %u\n", data[0], data[1], data[2], data[3]);
+		break;
+	}
+
+	return lun;
+}
+
+static void send_r2t(struct iscsi_cmnd *req)
+{
+	struct iscsi_cmnd *rsp;
+	struct iscsi_r2t_hdr *rsp_hdr;
+	u32 length, offset, burst;
+	LIST_HEAD(send);
+
+	length = req->r2t_length;
+	burst = req->conn->session->param.max_burst_length;
+	offset = be32_to_cpu(cmnd_hdr(req)->data_length) - length;
+
+	do {
+		rsp = iscsi_cmnd_create_rsp_cmnd(req, 0);
+		rsp->pdu.bhs.ttt = req->target_task_tag;
+
+		rsp_hdr = (struct iscsi_r2t_hdr *)&rsp->pdu.bhs;
+		rsp_hdr->opcode = ISCSI_OP_R2T;
+		rsp_hdr->flags = ISCSI_FLG_FINAL;
+		memcpy(rsp_hdr->lun, cmnd_hdr(req)->lun, 8);
+		rsp_hdr->itt = cmnd_hdr(req)->itt;
+		rsp_hdr->r2t_sn = cpu_to_be32(req->r2t_sn++);
+		rsp_hdr->buffer_offset = cpu_to_be32(offset);
+		if (length > burst) {
+			rsp_hdr->data_length = cpu_to_be32(burst);
+			length -= burst;
+			offset += burst;
+		} else {
+			rsp_hdr->data_length = cpu_to_be32(length);
+			length = 0;
+		}
+
+		dprintk(D_WRITE, "%x %u %u %u %u\n", cmnd_itt(req),
+			be32_to_cpu(rsp_hdr->data_length),
+			be32_to_cpu(rsp_hdr->buffer_offset),
+			be32_to_cpu(rsp_hdr->r2t_sn), req->outstanding_r2t);
+
+		list_add_tail(&rsp->list, &send);
+
+		if (++req->outstanding_r2t >= req->conn->session->param.max_outstanding_r2t)
+			break;
+
+	} while (length);
+
+	iscsi_cmnds_init_write(&send);
+}
+
+static void scsi_cmnd_exec(struct iscsi_cmnd *cmnd)
+{
+	assert(!cmnd->r2t_length);
+
+	if (cmnd->lun) {
+		iscsi_scsi_queuecmnd(cmnd);
+	} else {
+		iscsi_device_queue_cmnd(cmnd);
+	}
+}
+
+static int nop_out_start(struct iscsi_conn *conn, struct iscsi_cmnd *cmnd)
+{
+	u32 size, tmp;
+	int i, err = 0;
+
+	if (cmnd_ttt(cmnd) != cpu_to_be32(ISCSI_RESERVED_TAG)) {
+		cmnd->req = cmnd_find_hash(conn->session, cmnd->pdu.bhs.itt,
+					   cmnd->pdu.bhs.ttt);
+		if (!cmnd->req) {
+			/*
+			 * We didn't request this NOP-Out (by sending a
+			 * NOP-In, see 10.18.2 of the RFC) or our fake NOP-Out
+			 * timed out.
+			 */
+			eprintk("initiator bug %x\n", cmnd_itt(cmnd));
+			err = -ISCSI_REASON_PROTOCOL_ERROR;
+			goto out;
+		}
+
+		del_timer_sync(&cmnd->req->timer);
+		clear_cmnd_timer_active(cmnd->req);
+		dprintk(D_GENERIC, "NOP-Out: %p, ttt %x, timer %p\n",
+			cmnd->req, cmnd_ttt(cmnd->req), &cmnd->req->timer);
+	}
+
+	if (cmnd_itt(cmnd) == cpu_to_be32(ISCSI_RESERVED_TAG)) {
+		if (!cmnd_immediate(cmnd))
+			eprintk("%s\n", "initiator bug!");
+		update_stat_sn(cmnd);
+		err = check_cmd_sn(cmnd);
+		if (err)
+			goto out;
+	} else if ((err = cmnd_insert_hash(cmnd)) < 0) {
+		eprintk("ignore this request %x\n", cmnd_itt(cmnd));
+		goto out;
+	}
+
+	if ((size = cmnd->pdu.datasize)) {
+		size = (size + 3) & -4;
+		conn->read_msg.msg_iov = conn->read_iov;
+		if (cmnd->pdu.bhs.itt != cpu_to_be32(ISCSI_RESERVED_TAG)) {
+			struct tio *tio;
+			int pg_cnt = get_pgcnt(size, 0);
+
+			assert(pg_cnt < ISCSI_CONN_IOV_MAX);
+			cmnd->tio = tio = tio_alloc(pg_cnt);
+			tio_set(tio, size, 0);
+
+			for (i = 0; i < pg_cnt; i++) {
+				conn->read_iov[i].iov_base
+					= page_address(tio->pvec[i]);
+				tmp = min_t(u32, size, PAGE_CACHE_SIZE);
+				conn->read_iov[i].iov_len = tmp;
+				conn->read_size += tmp;
+				size -= tmp;
+			}
+		} else {
+			for (i = 0; i < ISCSI_CONN_IOV_MAX; i++) {
+				conn->read_iov[i].iov_base = dummy_data;
+				tmp = min_t(u32, size, sizeof(dummy_data));
+				conn->read_iov[i].iov_len = tmp;
+				conn->read_size += tmp;
+				size -= tmp;
+			}
+		}
+		assert(!size);
+		conn->read_overflow = size;
+		conn->read_msg.msg_iovlen = i;
+	}
+
+out:
+	return err;
+}
+
+static u32 get_next_ttt(struct iscsi_session *session)
+{
+	u32 ttt;
+
+	if (session->next_ttt == ISCSI_RESERVED_TAG)
+		session->next_ttt++;
+	ttt = session->next_ttt++;
+
+	return cpu_to_be32(ttt);
+}
+
+static void scsi_cmnd_start(struct iscsi_conn *conn, struct iscsi_cmnd *req)
+{
+	struct iscsi_scsi_cmd_hdr *req_hdr = cmnd_hdr(req);
+
+	dprintk(D_GENERIC, "scsi command: %02x\n", req_hdr->scb[0]);
+
+	req->lun = volume_get(conn->session->target, translate_lun(req_hdr->lun));
+	if (!req->lun) {
+		switch (req_hdr->scb[0]) {
+		case INQUIRY:
+		case REPORT_LUNS:
+			break;
+		default:
+			eprintk("%x %x\n", cmnd_itt(req), req_hdr->scb[0]);
+			create_sense_rsp(req, ILLEGAL_REQUEST, 0x25, 0x0);
+			cmnd_skip_data(req);
+			goto out;
+		}
+	} else
+		set_cmnd_lunit(req);
+
+	switch (req_hdr->scb[0]) {
+	case SERVICE_ACTION_IN:
+		if ((req_hdr->scb[1] & 0x1f) != 0x10)
+			goto error;
+	case INQUIRY:
+	case REPORT_LUNS:
+	case TEST_UNIT_READY:
+	case SYNCHRONIZE_CACHE:
+	case VERIFY:
+	case VERIFY_16:
+	case START_STOP:
+	case READ_CAPACITY:
+	case MODE_SENSE:
+	case REQUEST_SENSE:
+	case RESERVE:
+	case RELEASE:
+	{
+		if (!(req_hdr->flags & ISCSI_CMD_FINAL) || req->pdu.datasize) {
+			/* unexpected unsolicited data */
+			eprintk("%x %x\n", cmnd_itt(req), req_hdr->scb[0]);
+			create_sense_rsp(req, ABORTED_COMMAND, 0xc, 0xc);
+			cmnd_skip_data(req);
+		}
+		break;
+	}
+	case READ_6:
+	case READ_10:
+	case READ_16:
+	{
+		loff_t offset;
+		u32 length;
+
+		if (!(req_hdr->flags & ISCSI_CMD_FINAL) || req->pdu.datasize) {
+			/* unexpected unsolicited data */
+			eprintk("%x %x\n", cmnd_itt(req), req_hdr->scb[0]);
+			create_sense_rsp(req, ABORTED_COMMAND, 0xc, 0xc);
+			cmnd_skip_data(req);
+			break;
+		}
+
+		set_offset_and_length(req->lun, req_hdr->scb, &offset, &length);
+		req->tio = tio_alloc(get_pgcnt(length, offset));
+		tio_set(req->tio, length, offset);
+		break;
+	}
+	case WRITE_6:
+	case WRITE_10:
+	case WRITE_16:
+	case WRITE_VERIFY:
+	{
+		struct iscsi_sess_param *param = &conn->session->param;
+		loff_t offset;
+		u32 length;
+
+		req->r2t_length = be32_to_cpu(req_hdr->data_length) - req->pdu.datasize;
+		req->is_unsolicited_data = !(req_hdr->flags & ISCSI_CMD_FINAL);
+		req->target_task_tag = get_next_ttt(conn->session);
+
+		if (LUReadonly(req->lun)) {
+			create_sense_rsp(req, DATA_PROTECT, 0x27, 0x0);
+			cmnd_skip_data(req);
+			break;
+		}
+
+		if (!param->immediate_data && req->pdu.datasize)
+			eprintk("%x %x\n", cmnd_itt(req), req_hdr->scb[0]);
+
+		if (param->initial_r2t && !(req_hdr->flags & ISCSI_CMD_FINAL))
+			eprintk("%x %x\n", cmnd_itt(req), req_hdr->scb[0]);
+
+		if (req_hdr->scb[0] == WRITE_VERIFY && req_hdr->scb[1] & 0x02)
+			eprintk("Verification is ignored %x\n", cmnd_itt(req));
+
+		set_offset_and_length(req->lun, req_hdr->scb, &offset, &length);
+		if (cmnd_write_size(req) != length)
+			eprintk("%x %u %u\n", cmnd_itt(req), cmnd_write_size(req), length);
+
+		req->tio = tio_alloc(get_pgcnt(length, offset));
+		tio_set(req->tio, length, offset);
+
+		if (req->pdu.datasize) {
+			if (cmnd_recv_pdu(conn, req->tio, 0, req->pdu.datasize) < 0)
+				assert(0);
+		}
+		break;
+	}
+	error:
+	default:
+		eprintk("Unsupported %x\n", req_hdr->scb[0]);
+		create_sense_rsp(req, ILLEGAL_REQUEST, 0x20, 0x0);
+		cmnd_skip_data(req);
+		break;
+	}
+
+out:
+	return;
+}
+
+static void data_out_start(struct iscsi_conn *conn, struct iscsi_cmnd *cmnd)
+{
+	struct iscsi_data_out_hdr *req = (struct iscsi_data_out_hdr *)&cmnd->pdu.bhs;
+	struct iscsi_cmnd *scsi_cmnd = NULL;
+	u32 offset = be32_to_cpu(req->buffer_offset);
+
+	update_stat_sn(cmnd);
+
+	cmnd->req = scsi_cmnd = cmnd_find_hash(conn->session, req->itt, req->ttt);
+	if (!scsi_cmnd) {
+		eprintk("unable to find scsi task %x %x\n",
+			cmnd_itt(cmnd), cmnd_ttt(cmnd));
+		goto skip_data;
+	}
+
+	if (scsi_cmnd->r2t_length < cmnd->pdu.datasize) {
+		eprintk("invalid data len %x %u %u\n",
+			cmnd_itt(scsi_cmnd), cmnd->pdu.datasize, scsi_cmnd->r2t_length);
+		goto skip_data;
+	}
+
+	if (scsi_cmnd->r2t_length + offset != cmnd_write_size(scsi_cmnd)) {
+		eprintk("%x %u %u %u\n", cmnd_itt(scsi_cmnd), scsi_cmnd->r2t_length,
+			offset,	cmnd_write_size(scsi_cmnd));
+		goto skip_data;
+	}
+
+	scsi_cmnd->r2t_length -= cmnd->pdu.datasize;
+
+	if (req->ttt == cpu_to_be32(ISCSI_RESERVED_TAG)) {
+		/* unsolicited burst data */
+		if (scsi_cmnd->pdu.bhs.flags & ISCSI_FLG_FINAL) {
+			eprintk("unexpected data from %x %x\n",
+				cmnd_itt(cmnd), cmnd_ttt(cmnd));
+			goto skip_data;
+		}
+	}
+
+	dprintk(D_WRITE, "%u %p %p %p %u %u\n", req->ttt, cmnd, scsi_cmnd,
+		scsi_cmnd->tio, offset, cmnd->pdu.datasize);
+
+	if (cmnd_recv_pdu(conn, scsi_cmnd->tio, offset, cmnd->pdu.datasize) < 0)
+		goto skip_data;
+	return;
+
+skip_data:
+	cmnd->pdu.bhs.opcode = ISCSI_OP_DATA_REJECT;
+	cmnd_skip_pdu(cmnd);
+	return;
+}
+
+static void iscsi_session_push_cmnd(struct iscsi_cmnd *cmnd);
+
+static void data_out_end(struct iscsi_conn *conn, struct iscsi_cmnd *cmnd)
+{
+	struct iscsi_data_out_hdr *req = (struct iscsi_data_out_hdr *) &cmnd->pdu.bhs;
+	struct iscsi_cmnd *scsi_cmnd;
+	u32 offset;
+
+	assert(cmnd);
+	scsi_cmnd = cmnd->req;
+	assert(scsi_cmnd);
+
+	if (conn->read_overflow) {
+		eprintk("%x %u\n", cmnd_itt(cmnd), conn->read_overflow);
+		assert(scsi_cmnd->tio);
+		offset = be32_to_cpu(req->buffer_offset);
+		offset += cmnd->pdu.datasize - conn->read_overflow;
+		if (cmnd_recv_pdu(conn, scsi_cmnd->tio, offset, conn->read_overflow) < 0)
+			assert(0);
+		return;
+	}
+
+	if (req->ttt == cpu_to_be32(ISCSI_RESERVED_TAG)) {
+		if (req->flags & ISCSI_FLG_FINAL) {
+			scsi_cmnd->is_unsolicited_data = 0;
+			iscsi_session_push_cmnd(scsi_cmnd);
+		}
+	} else {
+		/* TODO : proper error handling */
+		if (!(req->flags & ISCSI_FLG_FINAL) && scsi_cmnd->r2t_length == 0)
+			eprintk("initiator error %x\n", cmnd_itt(scsi_cmnd));
+
+		if (!(req->flags & ISCSI_FLG_FINAL))
+			goto out;
+
+		scsi_cmnd->outstanding_r2t--;
+
+		if (scsi_cmnd->r2t_length == 0)
+			assert(list_empty(&scsi_cmnd->pdu_list));
+
+		iscsi_session_push_cmnd(scsi_cmnd);
+	}
+
+out:
+	iscsi_cmnd_remove(cmnd);
+	return;
+}
+
+static void __cmnd_abort(struct iscsi_cmnd *cmnd)
+{
+	if (cmnd_rxstart(cmnd))
+		set_cmnd_tmfabort(cmnd);
+
+	if (cmnd_waitio(cmnd))
+		return;
+
+	if (cmnd->conn->read_cmnd != cmnd)
+		cmnd_release(cmnd, 1);
+}
+
+static int cmnd_abort(struct iscsi_session *session, struct iscsi_cmnd *req)
+{
+	struct iscsi_task_mgt_hdr *req_hdr =
+				(struct iscsi_task_mgt_hdr *)&req->pdu.bhs;
+	struct iscsi_cmnd *cmnd;
+
+	u32 min_cmd_sn = req_hdr->cmd_sn - session->max_queued_cmnds;
+
+	req_hdr->ref_cmd_sn = be32_to_cpu(req_hdr->ref_cmd_sn);
+
+	dprintk(D_GENERIC, "cmd_sn(%u) ref_cmd_sn(%u) min_cmd_sn(%u) rtt(%x)"
+		" lun(%d) cid(%u)\n",
+		req_hdr->cmd_sn, req_hdr->ref_cmd_sn, min_cmd_sn, req_hdr->rtt,
+		translate_lun(req_hdr->lun), req->conn->cid);
+
+	if (after(req_hdr->ref_cmd_sn, req_hdr->cmd_sn))
+		return ISCSI_RESPONSE_FUNCTION_REJECTED;
+
+	if (!(cmnd = cmnd_find_hash(session, req_hdr->rtt, ISCSI_RESERVED_TAG))) {
+		if (between(req_hdr->ref_cmd_sn, min_cmd_sn, req_hdr->cmd_sn))
+			return ISCSI_RESPONSE_FUNCTION_COMPLETE;
+		else
+			return ISCSI_RESPONSE_UNKNOWN_TASK;
+	}
+
+	dprintk(D_GENERIC, "itt(%x) opcode(%x) scsicode(%x) lun(%d) cid(%u)\n",
+		cmnd_itt(cmnd), cmnd_opcode(cmnd), cmnd_scsicode(cmnd),
+		translate_lun(cmnd_hdr(cmnd)->lun), cmnd->conn->cid);
+
+	if (cmnd_opcode(cmnd) == ISCSI_OP_SCSI_TASK_MGT_MSG)
+		return ISCSI_RESPONSE_FUNCTION_REJECTED;
+
+	if (translate_lun(cmnd_hdr(cmnd)->lun) !=
+						translate_lun(req_hdr->lun))
+		return ISCSI_RESPONSE_FUNCTION_REJECTED;
+
+	if (cmnd->conn && test_bit(CONN_ACTIVE, &cmnd->conn->state)) {
+		if (cmnd->conn->cid != req->conn->cid)
+			return ISCSI_RESPONSE_FUNCTION_REJECTED;
+	} else {
+		/* Switch cmnd connection allegiance */
+	}
+
+	__cmnd_abort(cmnd);
+
+	return ISCSI_RESPONSE_FUNCTION_COMPLETE;
+}
+
+static int target_reset(struct iscsi_cmnd *req, u32 lun, int all)
+{
+	struct iscsi_target *target = req->conn->session->target;
+	struct iscsi_session *session;
+	struct iscsi_conn *conn;
+	struct iscsi_cmnd *cmnd, *tmp;
+	struct iet_volume *volume;
+
+	list_for_each_entry(session, &target->session_list, list) {
+		list_for_each_entry(conn, &session->conn_list, list) {
+			list_for_each_entry_safe(cmnd, tmp, &conn->pdu_list, conn_list) {
+				if (cmnd == req)
+					continue;
+
+				if (all)
+					__cmnd_abort(cmnd);
+				else if (translate_lun(cmnd_hdr(cmnd)->lun)
+									== lun)
+					__cmnd_abort(cmnd);
+			}
+		}
+	}
+
+	list_for_each_entry(volume, &target->volumes, list) {
+		if (all || volume->lun == lun) {
+			/* force release */
+			volume_release(volume, 0, 1);
+			/* power-on, reset, or bus device reset occurred */
+			ua_establish_for_all_sessions(target, volume->lun,
+						      0x29, 0x0);
+		}
+	}
+
+	return 0;
+}
+
+static void task_set_abort(struct iscsi_cmnd *req)
+{
+	struct iscsi_session *session = req->conn->session;
+	struct iscsi_conn *conn;
+	struct iscsi_cmnd *cmnd, *tmp;
+
+	list_for_each_entry(conn, &session->conn_list, list) {
+		list_for_each_entry_safe(cmnd, tmp, &conn->pdu_list, conn_list) {
+			if (translate_lun(cmnd_hdr(cmnd)->lun)
+					!= translate_lun(cmnd_hdr(req)->lun))
+				continue;
+
+			if (before(cmnd_hdr(cmnd)->cmd_sn,
+					cmnd_hdr(req)->cmd_sn))
+				__cmnd_abort(cmnd);
+		}
+	}
+}
+
+static inline char *tmf_desc(int fun)
+{
+	static char *tmf_desc[] = {
+		"Unknown Function",
+		"Abort Task",
+		"Abort Task Set",
+		"Clear ACA",
+		"Clear Task Set",
+		"Logical Unit Reset",
+		"Target Warm Reset",
+		"Target Cold Reset",
+		"Task Reassign",
+        };
+
+	if ((fun < ISCSI_FUNCTION_ABORT_TASK) ||
+				(fun > ISCSI_FUNCTION_TASK_REASSIGN))
+		fun = 0;
+
+	return tmf_desc[fun];
+}
+
+static inline char *rsp_desc(int rsp)
+{
+	static char *rsp_desc[] = {
+		"Function Complete",
+		"Unknown Task",
+		"Unknown LUN",
+		"Task Allegiant",
+		"Failover Unsupported",
+		"Function Unsupported",
+		"No Authorization",
+		"Function Rejected",
+		"Unknown Response",
+	};
+
+	if (((rsp < ISCSI_RESPONSE_FUNCTION_COMPLETE) ||
+			(rsp > ISCSI_RESPONSE_NO_AUTHORIZATION)) &&
+			(rsp != ISCSI_RESPONSE_FUNCTION_REJECTED))
+		rsp = 8;
+	else if (rsp == ISCSI_RESPONSE_FUNCTION_REJECTED)
+		rsp = 7;
+
+	return rsp_desc[rsp];
+}
+
+static void execute_task_management(struct iscsi_cmnd *req)
+{
+	struct iscsi_conn *conn = req->conn;
+	struct iscsi_session *session = conn->session;
+	struct iscsi_target *target = session->target;
+	struct iscsi_cmnd *rsp;
+	struct iscsi_task_mgt_hdr *req_hdr = (struct iscsi_task_mgt_hdr *)&req->pdu.bhs;
+	struct iscsi_task_rsp_hdr *rsp_hdr;
+	u32 lun;
+	int function = req_hdr->function & ISCSI_FUNCTION_MASK;
+
+	rsp = iscsi_cmnd_create_rsp_cmnd(req, 1);
+	rsp_hdr = (struct iscsi_task_rsp_hdr *)&rsp->pdu.bhs;
+
+	rsp_hdr->opcode = ISCSI_OP_SCSI_TASK_MGT_RSP;
+	rsp_hdr->flags = ISCSI_FLG_FINAL;
+	rsp_hdr->itt = req_hdr->itt;
+	rsp_hdr->response = ISCSI_RESPONSE_FUNCTION_COMPLETE;
+
+	switch (function) {
+	case ISCSI_FUNCTION_ABORT_TASK:
+	case ISCSI_FUNCTION_ABORT_TASK_SET:
+	case ISCSI_FUNCTION_CLEAR_ACA:
+	case ISCSI_FUNCTION_CLEAR_TASK_SET:
+	case ISCSI_FUNCTION_LOGICAL_UNIT_RESET:
+		lun = translate_lun(req_hdr->lun);
+		if (!volume_lookup(target, lun)) {
+			rsp_hdr->response = ISCSI_RESPONSE_UNKNOWN_LUN;